Ana Pueyo Velasco

Departamento de Ingeniería Química Industrial y Medio Ambiente
Escuela Técnica Superior de Ingenieros Industriales
Climate change technology transfer to developing

countries: evidence analysis and policy
recommendations
TESIS DOCTORAL
Ana Pueyo Velasco

Licenciada en Administración y Dirección de Empresas
MSc Analysis, Design and Management of Information Systems
Madrid 2012
Codirectores
Julio Lumbreras Martín Pedro Linares

Doctor Ingeniero Industrial Doctor Ingeniero Agrónomo
Ingeniero Industrial Ingeniero Agrónomo
i
Tribunal nombrado por el Magnifico y Excelentísimo Sr. Rector de la
Universidad Politécnica de Madrid, el día 31 de enero de 2012.
Presidente: Dra. Encarnación Rodríguez Hurtado
Secretario: Dra. Mª Jesús Sánchez Naranjo
Vocal: Dr. José Ignacio Pérez Arriaga
Vocal: Dr. Jim Watson
Vocal: Dr. Xabier Labandeira
Suplente: Dr. Walter Leal
Suplente: Dr. Pablo Martínez de Anguita
Realizado el acto de defensa y lectura de la tesis el día 5 de marzo de 2012

en la E.T.S. Ingenieros Industriales.
Calificación: __________________________________________
EL PRESIDENTE LOS VOCALES
EL SECRETARIO
ii
Agradecimientos
Desde los primeros pasos en la elaboración de mi tesis hasta la meta final han sido muchas las
personas que me han apoyado y desde muchos lugares, a las que recuerdo por orden
cronológico.
Mi interés por la transferencia de tecnología para la mitigación del cambio climático comenzó
durante mis años como consultora en Garrigues Medio Ambiente. Durante esos años
compaginé mi trabajo con los cursos de Doctorado y la elaboración de la DEA. Agradezco a mis
compañeros de Garrigues la primera inspiración para emprender mi investigación y la
flexibilidad para compaginar largas horas de trabajo con mi labor académica. Asimismo, mi
investigación no habría podido comenzar sin el apoyo de Pablo Martínez de Anguita, de la
Universidad Rey Juan Carlos, mi Director durante la fase de la DEA. Gracias Pablo por tu gran
generosidad durante los primeros años de Doctorado.
Tras varios años de trabajo en Londres, la segunda y más dura fase de elaboración de la tesis
doctoral no habría sido posible sin mis dos directores Julio Lumbreras y Pedro Linares. Gracias
a Julio por su apoyo durante el proceso. Dados los largos horarios del trabajo en consultoría y
lo absorbentes que pueden llegar a ser los proyectos, esta tesis no habría visto la luz si Julio no
me hubiera ofrecido la oportunidad de aparcar mi trabajo durante una temporada para
dedicarme casi por entero al Doctorado. Pedro me ha proporcionado la orientación que
necesitaba para enfocar mi investigación, mejorar la calidad de su metodología y abordar la
gran complejidad del estudio de la transferencia tecnológica. Con la lucidez de sus comentarios
y su incansable perfeccionismo Pedro me ha ayudado a ser una mejor investigadora.
Me gustaría también dar las gracias a Mª Jesús Sánchez Naranjo, por su dirección al comienzo
de la tesis, su gran ayuda para mejorar mis conocimientos de técnicas de análisis cuantitativo,
su entusiasmo y buenos consejos si alguna vez caía en el desánimo. Otra influencia crucial
durante mi tesis ha sido María Mendiluce. María ha revisado minuciosamente muchos de mis
textos y ha aportado a mi trabajo la conexión con la realidad empresarial. Durante los últimos
años de Doctorado hemos comenzado una fructífera colaboración académica, profesional y
personal que espero que continúe durante mucho tiempo.
Para el trabajo de campo en Chile ha sido inestimable la ayuda de Rodrigo García Palma, ex-
compañero de Ecofys y Gerente Técnico del CER en Chile. Rodrigo ha compartido conmigo su
gran sabiduría sobre los desafíos de la transferencia de tecnologías de energía renovable a
países en desarrollo y me ha abierto las puertas a las experiencias reales de transferencia que
constituyen los casos de estudio de mi tesis. Gracias Rodrigo por esas maravillosas semanas en
Chile y por los piscos compartidos para alegrar nuestras discusiones. Gracias también a Darío
Morales, de CORFO, por compartir sus amplios conocimientos, por nuestra colaboración y los
buenos momentos pasados en Chile. Agradezco también a todos los profesionales chilenos
que aceptaron participar en mi investigación, no nombrados aquí, pero sí como parte de la
tesis.
Quisiera también agradecer el apoyo de Walter Leal y Erik Haites. Walter Leal ha dado
visibilidad a mi investigación mediante su proyecto JELARE y me ha mostrado las enormes
posibilidades que existen para mejorar los niveles de transferencia tecnológica a
Latinoamérica. Erik Haites ha revisado mis artículos y contribuido a su mejora con su crítica
constructiva. Erik Haites y Steve Seres también han compartido con generosidad sus datos del
estudio de transferencia de tecnología en el CDM.
Gracias también a mi amiga de toda la vida y Doctora ya desde hace tiempo Beatriz Alonso, por
su gran interés en mi tesis y sus sabios consejos. En el camino recorrido he hecho nuevos
amigos, comenzando por Maruxa en la Universidad Rey Juan Carlos y Luz en la Politécnica. Una
mención especial a Amandine Ody por su ayuda con los aspectos econométricos y el software
iii
de análisis cuantitativo, así como por sus recomendaciones en la búsqueda de trabajo
académico. También gracias a Penélope Woods, por los agradables días de estudio y las
comidas en la Biblioteca Nacional de Londres.
Finalmente, sin duda quienes merecen mi mayor agradecimiento son mi familia. Gracias Chris
por aguantar estoicamente todos estos años de investigación desde que nos conocimos y te
dije que me quedaban “unos meses” para terminar la tesis. Gracias por tu paciencia, por
aceptar que muchos fines de semana no podíamos hacer planes porque los pasaría en la
biblioteca, por leer y corregir muchos capítulos y por los continuos ánimos. Gracias también a
mis padres por su apoyo e ilusión con mi proyecto y un especial agradecimiento a mi padre por
inculcarme su amor por el trabajo y el estudio.
iv
Abstract
Developing countries are experiencing unprecedented levels of economic growth. As a result,
they will be responsible for most of the future growth in energy demand and greenhouse gas
(GHG) emissions. Curbing GHG emissions in developing countries has become one of the
cornerstones of a future international agreement under the United Nations Framework
Convention for Climate Change (UNFCCC). However, setting caps for developing countries’
GHG emissions has encountered strong resistance in the current round of negotiations.
Continued economic growth that allows poverty eradication is still the main priority for most
developing countries, and caps are perceived as a constraint to future growth prospects. The
development, transfer and use of low-carbon technologies have more positive connotations,
and are seen as the potential path towards low-carbon development.
So far, the success of the UNFCCC process in improving the levels of technology transfer (TT) to
developing countries has been limited. This thesis analyses the causes for such limited success
and seeks to improve on the understanding about what constitutes TT in the field of climate
change, establish the factors that enable them in developing countries and determine which
policies could be implemented to reinforce these factors.
Despite the wide recognition of the importance of technology and knowledge transfer to
developing countries in the climate change mitigation policy agenda, this issue has not
received sufficient attention in academic research. Current definitions of climate change TT
barely take into account the perspective of actors involved in actual climate change TT
activities, while respective measurements do not bear in mind the diversity of channels
through which these happen and the outputs and effects that they convey. Furthermore, the
enabling factors for TT in non-BRIC (Brazil, Russia, India, China) developing countries have been
seldom investigated, and policy recommendations to improve the level and quality of TTs to
developing countries have not been adapted to the specific needs of highly heterogeneous
countries, commonly denominated as “developing countries”. This thesis contributes to
enriching the climate change TT debate from the perspective of a smaller emerging economy
(Chile) and by undertaking a quantitative analysis of enabling factors for TT in a large sample of
developing countries.
Two methodological approaches are used to study climate change TT: comparative case study
analysis and quantitative analysis. Comparative case studies analyse TT processes in ten cases
based in Chile, all of which share the same economic, technological and policy frameworks,
thus enabling us to draw conclusions on the enabling factors and obstacles operating in TT
processes. The quantitative analysis uses three methodologies – principal component analysis,
multiple regression analysis and cluster analysis – to assess the performance of developing
countries in a number of enabling factors and the relationship between these factors and
indicators of TT, as well as to create groups of developing countries with similar performances.
The findings of this thesis are structured to provide responses to four main research questions:
What constitutes technology transfer and how does it happen? Is it possible to measure
technology transfer, and what are the main challenges in doing so? Which factors enable
climate change technology transfer to developing countries? And how do different developing
countries perform in these enabling factors, and how can differentiated policy priorities be
defined accordingly?
v
Resumen
Los países en desarrollo están experimentando niveles de crecimiento económico sin
precedentes. Como consecuencia, se espera que sean responsables de la mayor parte del
futuro crecimiento global en demanda energética y emisiones de Gases de Efecto de
Invernadero (GEI). Reducir las emisiones de GEI en los países en desarrollo es por tanto uno de
los pilares de un futuro acuerdo internacional en el marco de la Convención Marco de las
Naciones Unidas para el Cambio Climático (UNFCCC). La posibilidad de compromisos
vinculantes de reducción de emisiones de GEI ha sido rechazada por los países en desarrollo,
que perciben estos límites como frenos a su desarrollo económico y a su prioridad principal de
erradicación de la pobreza. El desarrollo, transferencia y uso de tecnologías bajas en carbono
tiene connotaciones más positivas y se percibe como la vía hacia un crecimiento bajo en
carbono.
Hasta el momento, la UNFCCC ha tenido un éxito limitado en la promoción de transferencias
de tecnología (TT) a países en desarrollo. Esta tesis analiza las causas de este resultado y busca
mejorar la comprensión sobre qué constituye transferencia de tecnología en el área de cambio
climático, cuales son los factores que la facilitan en países en desarrollo y qué políticas podrían
implementarse para reforzar dichos factores.
A pesar del extendido reconocimiento sobre la importancia de la transferencia de tecnología a
países en desarrollo en la agenda política de cambio climático, esta cuestión no ha sido
suficientemente atendida por la investigación existente. Las definiciones actuales de
transferencia de tecnología relacionada con la mitigación del cambio climático no tienen en
cuenta la diversidad de canales por las que se manifiestan o los efectos que consiguen. Los
factores facilitadores de TT en países en desarrollo no BRIC (Brasil, Rusia, India y China) apenas
han sido investigados, y las recomendaciones políticas para aumentar el nivel y la calidad de la
TT no se han adaptado a las necesidades específicas de países muy heterogéneos aglutinados
bajo el denominado grupo de "países en desarrollo". Esta tesis contribuye a enriquecer el
debate sobre la TT de cambio climático con la perspectiva de una economía emergente de
pequeño tamaño (Chile) y el análisis cuantitativo de factores que facilitan la TT en una amplia
muestra de países en desarrollo.
Se utilizan dos metodologías para el estudio de la TT a países en desarrollo: análisis
comparativo de casos de estudio y análisis cuantitativo basado en métodos multivariantes. Los
casos de estudio analizan procesos de TT en diez casos basados en Chile, para derivar
conclusiones sobre los factores que facilitan u obstaculizan el proceso de transferencia. El
análisis cuantitativo multivariante utiliza tres metodologías: regresión múltiple, análisis de
componentes principales y análisis cluster. Con dichas metodologías se busca analizar el
posicionamiento de diversos países en cuanto a factores que facilitan la TT; las relaciones entre
dichos factores e indicadores de transferencia tecnológica; y crear grupos de países con
características similares que podrían beneficiarse de políticas similares para la promoción de la
transferencia de tecnología.
Los resultados de la tesis se estructuran en torno a cuatro preguntas de investigación: ¿Que es
la transferencia de tecnología y cómo ocurre?; ¿Es posible medir la transferencia de
tecnologías de bajo carbono?; ¿Qué factores facilitan la transferencia de tecnologías de bajo
carbono a países en desarrollo? y ¿Cómo se puede agrupar a los países en desarrollo en
función de sus necesidades políticas para la promoción de la transferencia de tecnologías de
bajo carbono?
vi
Contents
1 INTRODUCTION ............................................................................................................................ 1
1.1 INTRODUCTION .................................................................................................................................. 2
1.2 INTERNATIONAL POLICY CONTEXT: TECHNOLOGY TRANSFER IN THE UNFCCC .................................................. 3
1.3 OBJECTIVES AND RESEARCH QUESTIONS ................................................................................................ 11
1.4 STRUCTURE ..................................................................................................................................... 13
2 LITERATURE ON CLIMATE CHANGE TECHNOLOGY TRANSFER TO DEVELOPING COUNTRIES ....... 15
2.1 CONCEPTUALISING TECHNOLOGY TRANSFER ........................................................................................... 16
2.2 CHANNELS OF TECHNOLOGY TRANSFER.................................................................................................. 17
2.3 MEASURING TECHNOLOGY TRANSFER ................................................................................................... 18
2.4 ENABLING FRAMEWORKS FOR CLIMATE CHANGE TECHNOLOGY TRANSFER..................................................... 31
2.5 CONCLUSIONS.................................................................................................................................. 43
3 CHILE’S ENABLING ENVIRONMENT FOR CLIMATE CHANGE TECHNOLOGY TRANSFER................. 46
3.1 INTRODUCTION ................................................................................................................................ 47
3.2 OVERVIEW OF CHILEAN ENERGY SYSTEM AND GHG EMISSIONS.................................................................. 47
3.3 ECONOMIC AND INSTITUTIONAL FRAMEWORK ........................................................................................ 60
3.4 TECHNOLOGY DEMAND FACTORS AND POLICIES....................................................................................... 62
3.5 TECHNOLOGY SUPPLY FACTORS AND POLICIES ......................................................................................... 71
3.6 INDUSTRIAL DEVELOPMENT FACTORS .................................................................................................... 79
3.7 CONCLUSIONS.................................................................................................................................. 84
4 CASE STUDIES ON CLIMATE CHANGE TECHNOLOGY TRANSFER TO CHILE ................................... 87
4.1 INTRODUCTION ................................................................................................................................ 88
4.2 METHODOLOGY ............................................................................................................................... 89
4.3 CASE STUDIES .................................................................................................................................. 97
4.4 DISCUSSION................................................................................................................................... 143
4.5 CONCLUSIONS................................................................................................................................ 170
5 QUANTITATIVE ANALYSIS TO ASSESS DEVELOPING COUNTRIES’ POLICY NEEDS FOR CLEAN
ENERGY TECHNOLOGY TRANSFER ..................................................................................................... 171
5.1 INTRODUCTION .............................................................................................................................. 172
5.2 DATA AVAILABILITY ......................................................................................................................... 176
5.3 EXPLORING INDICATORS OF ENABLING FRAMEWORKS FOR TECHNOLOGY TRANSFER ...................................... 190
5.4 ANALYSING THE RELATIONSHIP BETWEEN ENABLING FRAMEWORKS AND INDICATORS OF CLEAN ENERGY
TECHNOLOGY TRANSFER ............................................................................................................................ 203
5.5 DEFINING GROUPS OF DEVELOPING COUNTRIES PER TECHNOLOGY TRANSFER POLICY PRIORITY ........................ 214
5.6 DISCUSSION AND CONCLUSIONS......................................................................................................... 242
6 SUMMARY AND CONCLUSIONS................................................................................................ 248
6.1 INTRODUCTION .............................................................................................................................. 249
6.2 MAIN FINDINGS ............................................................................................................................. 250
6.3 CONTRIBUTIONS AND IMPLICATIONS ................................................................................................... 264
6.4 LIMITATIONS AND FUTURE RESEARCH .................................................................................................. 267
7 REFERENCES ............................................................................................................................. 269
8 ANNEXES .................................................................................................................................. 282
8.1 CHAPTER 2: TABLES ........................................................................................................................ 283
8.1 CHAPTER 3: TABLES ........................................................................................................................ 289
8.2 CHAPTER 4. ................................................................................................................................... 290
8.3 CHAPTER 5 .................................................................................................................................... 314
vii
Tables
TABLE 1- EFFECTIVENESS OF UNFCCC MECHANISMS IN PROMOTING TECHNOLOGY TRANSFERS ............. 6
TABLE 2- SUMMARY OF REVIEWED APPROACHES FOR MEASURING TECHNOLOGY TRANSFER ............... 29
TABLE 3- OVERVIEW OF PRODUCTIVITY MEASURES ................................................................................. 38
TABLE 4- SUMMARY OF ENABLING FACTORS FOR CLIMATE CHANGE TECHNOLOGY TRANSFER.............. 41
TABLE 5- DETAILS OF INTERVIEWS WITH CHILEAN EXPERTS ..................................................................... 47
TABLE 6- TOTAL INSTALLED CAPACITY IN THE SIC, 2010 ........................................................................... 58
TABLE 7- TOTAL INSTALLED CAPACITY IN THE SING, 2010 ........................................................................ 58
TABLE 8- NCRE PROJECTS APPROVED IN THE SEIA AT JUNE 2010 ............................................................. 59
TABLE 9- NCRE POTENTIALS IN THE SIC IN 2025 ........................................................................................ 59
TABLE 10- AVERAGE ELECTRICITY CONSUMPTION AND TPEC GROWTH BETWEEN 1990-2008 AND 2000-
2008 IN A SAMPLE OF LATIN AMERICAN AND WORLD COUNTRIES ................................................ 64
TABLE 11- SELECTION OF CASE STUDIES .................................................................................................... 94
TABLE 12- INTERVIEW DETAILS .................................................................................................................. 96
TABLE 13- APPROACHES TO MEASURE CLIMATE CHANGE TECHNOLOGY TRANSFER BASED ON CASE
STUDY ANALYSIS ............................................................................................................................. 151
TABLE 14- INDUSTRIAL POLICY MEASURES INFLUENCING LINKAGES AND LOCAL INNOVATION ............ 163
TABLE 15- ENABLING FACTORS FOR CLIMATE CHANGE TECHNOLOGY TRANSFER ................................. 168
TABLE 16- MEASUREMENTS OF CLIMATE CHANGE TECHNOLOGY TRANSFER SUGGESTED BY THE
LITERATURE REVIEW AND CASE STUDY ANALYSIS ......................................................................... 173
TABLE 17- ENABLING FACTORS SUGGESTED BY THE LITERATURE REVIEW AND THE CASE STUDY ANALYSIS
........................................................................................................................................................ 174
TABLE 18-COMTRADE CLEAN ENERGY TECHNOLOGY IMPORT AND EXPORT DATA CODE DESCRIPTION 176
TABLE 19- DESCRITIVE STATISTICS FOR RENEWABLE ENERGY TECHNOLOGY IMPORT VARIABLES ......... 177
TABLE 20- DESCRIPTIVE STATISTICS FOR RENEWABLE ENERGY TECHNOLOGY EXPORTS ........................ 178
TABLE 21-CORRELATIONS BETWEEN RE CAPACITY, RE ELECTRICITY GENERATION, CDM RE PROJECTS AND
CDM RE PROJECTS CO2 EMISSION REDUCTIONS. .......................................................................... 179
TABLE 22- EXTREME VALUES OF CLAIMS OF TECHNOLOGY TRANSFER IN RENEWABLE ENERGY AND NON-
HYDRO RENEWABLE ENERGY CDM PROJECTS ............................................................................... 179
TABLE 23- DESCRIPTIVE STATISTICS OF RENEWABLE ENERGY CAPACITY INVOLVING TECHNOLOGY
TRANSFER ....................................................................................................................................... 180
TABLE 24- SELECTED TECHNOLOGY TRANSFER VARIABLES ..................................................................... 181
TABLE 25- TECHNOLOGY TRANSFER VARIABLES THAT COULD NOT BE INCLUDED ................................. 181
TABLE 26- CORRELATIONS OF CLEAN ENERGY TECHNOLOGY TRANSFER VARIABLES ............................. 183
TABLE 27-FINAL SET OF ENABLING FACTORS .......................................................................................... 187
TABLE 28-FINAL SET OF EXPLANATORY VARIABLES ................................................................................. 190
TABLE 29- SUITABILITY OF PCA THROUGH THE ANALYSIS OF SAMPLE SIZE AND CORRELATIONS AMONG
VARIABLES ...................................................................................................................................... 193
TABLE 30- CORRELATIONS OF VARIABLES EXPRESSED IN VALUES PER GDP ........................................... 200
TABLE 31- NUMBER OF VALID CASES FOR DEPENDENT VARIABLES ........................................................ 204
TABLE 32- SUMMARY OF REGRESSION ANALYSIS ................................................................................... 212
TABLE 33- AGGLOMERATION SCHEDULE IN WARD’S CLUSTERING METHOD ......................................... 216
TABLE 34- CLUSTER STRUCTURE’S WARDS METHOD WITH LOGS ........................................................... 219
TABLE 35- ANOVA .................................................................................................................................... 220
TABLE 36-CLUSTER CHARACTERISATION ................................................................................................. 221
TABLE 37- NUMBER OF CASES IN EACH CLUSTER .................................................................................... 223
TABLE 38- ANOVA .................................................................................................................................... 223
TABLE 39-CLUSTER FORMATION WITH K-MEANS.................................................................................... 224
TABLE 40-MEMBERS OF CLUSTER 3 IN THE FIRST ITERATIONS OF WARD’S AND K-MEANS CLUSTERING
METHODS ....................................................................................................................................... 228
TABLE 41- MEMBERS OF CLUSTER 2 IN THE FIRST ITERATIONS OF WARD’S AND K-MEANS CLUSTERING
METHODS ....................................................................................................................................... 228
METHODS ....................................................................................................................................... 229
METHODS ....................................................................................................................................... 230
viii
TABLE 44-VALUE OF TECHNOLOGY TRANSFER VARIABLES FOR MEMBERS OF WARD’S CLUSTERS WITH
LOGS ............................................................................................................................................... 235
TABLE 45-VALUE OF TT VARIABLES FOR MEMBERS OF K-MEANS CLUSTERS WITH LOGS ....................... 238
LITERATURE REVIEW AND CASE STUDY ANALYSIS ......................................................................... 253
TABLE 47- APPROACHES TO MEASURE PROJECT-LEVEL CLIMATE CHANGE TECHNOLOGY TRANSFER
BASED ON CASE STUDY ANALYSIS .................................................................................................. 254
........................................................................................................................................................ 256
TABLE 49- CHANNELS FOR PRIVATE SECTOR CLIMATE CHANGE TECHNOLOGY TRANSFER .................... 283
TABLE 50- KNOWLEDGE TRANSFER METRICS IN HOI ET AL. (2008)......................................................... 284
TABLE 51- KNOWLEDGE TRANSFER METRICS IN JENSEN ET AL (2009) .................................................... 285
TABLE 52-UNFCCC PERFORMANCE INDICATORS ON TECHNOLOGY TRANSFER ...................................... 286
TABLE 53- UNCTAD INNOVATION CAPABILITY INDEX .............................................................................. 289
TABLE 54- COST OF ELECTRICITY PRODUCTION WITH DIFFERENT TECHNOLOGIES ................................ 300
TABLE 55- LITHIUM-ION BATTERY COST BREAKDOWN ........................................................................... 301
FIGURE 72- WIND POWER CAPACITY, TOP 10 COUNTRIES ...................................................................... 302
FIGURE 73- SOLAR PV EXISTING CAPACITY, TOP SIX COUNTRIES, 2009 .................................................. 304
FIGURE 75- PV SUPPLY CHAIN.................................................................................................................. 307
FIGURE 77- BASIC STRUCTURE OF THE CSP CORE VALUE CHAIN............................................................. 308
FIGURE 79- GLOBAL LITHIUM-ION BATTERY MARKET SHARE, BY COUNTRY AND FIRM ......................... 310
FIGURE 80- PATENTING GROWTH RATES FOR SELECTED CLEAN ENERGY TECHNOLOGIES..................... 311
FIGURE 82- PATENTS AND RESEARCH PAPERS RELATED TO LITHIUM-ION BATTERIES 1998-2007, BY
COUNTRY ........................................................................................................................................ 312
TABLE 56- ESTIMATED GLOBAL GAPS IN PUBLIC RD&D SPENDING AND MAIN PRIORITIES FOR
TECHNOLOGIES COVERED IN CASE STUDIES .................................................................................. 313
TABLE 57- MODEL RESULTS: EXPORTS OF RENEWABLE ENERGY TECHNOLOGIES PER CAPITA ............... 335
TABLE 58- MODEL RESULTS: IMPORTS OF RENEWABLE ENERGY TECHNOLOGIES PER CAPITA .............. 336
TABLE 59- MODEL RESULTS: RENEWABLE GENERATION CAPACITY WITH TT PER CAPITA ..................... 337
TABLE 60- AGGLOMERATION SCHEDULE ................................................................................................. 338
TABLE 61- CLUSTER STRUCTURE SECOND ITERATION ............................................................................. 341
TABLE 62- ANOVA .................................................................................................................................... 342
FIGURE 86- MEANS PLOT ......................................................................................................................... 343
TABLE 63- CLUSTER CHARACTERISATION ................................................................................................ 343
TABLE 64-CLUSTERS FORMATION IN HIERARCHICAL CLUSTER ANALYSIS ............................................... 345
TABLE 65- NUMBER OF CASES IN EACH CLUSTER .................................................................................... 347
TABLE 66- ANOVA .................................................................................................................................... 347
TABLE 67-CLUSTER FORMATION WITH K-MEANS.................................................................................... 347
TABLE 68- CLUSTERS FORMATION IN NON-HIERARCHICAL, K-MEANS CLUSTER ANALYSIS .................... 350
FIGURE 88- NON-HIERARCHICAL (K-MEANS) CLUSTERING MEANS PLOT FOR SECOND ITERATION WITH
NO LOGS ......................................................................................................................................... 352
TABLE 69-MEMBERS OF CLUSTER 1 IN THE SECOND ITERATIONS OF WARDS AND K-MEANS CLUSTERING
METHODS ....................................................................................................................................... 353
METHODS ....................................................................................................................................... 353
METHODS ....................................................................................................................................... 354
METHODS ....................................................................................................................................... 354
TABLE 73- CLUSTER DESCRIPTIVES- HIERARCHICAL CLUSTER ANALYSIS FIRST ITERATION ..................... 355
TABLE 74- TUKEY POST-HOC TEST, HIERARCHICAL ANALYSIS, FIRST ITERATION..................................... 356
TABLE 75-CLUSTER DESCRIPTIVES WITH ANOVA, NON-HIERARCHICAL CLUSTER ANALYSIS, FIRST
ITERATION ...................................................................................................................................... 358
TABLE 76- POST-HOC TUKEY TEST, NON HIERARCHICAL CLUSTER, FIRST ITERATION ............................. 359
Figures
ix
FIGURE 1- CO2 EMISSIONS FROM ENERGY CONSUMPTION PER COUNTRY .............................................. 48
FIGURE 2 - EVOLUTION OF CHILEAN CARBON INTENSITY ......................................................................... 48
FIGURE 3 - CHILEAN CARBON INTENSITY COMPARED TO CHINA, INDIA, BRAZIL AND SOUTH AFRICA ..... 49
FIGURE 4- CHILE’S PRIMARY ENERGY CONSUMPTION (TERACAL) ............................................................ 49
FIGURE 5- EVOLUTION OF PRIMARY ENERGY CONSUMPTION IN CHILE 2000-2008 ................................. 50
FIGURE 6- SOURCES OF FINAL ENERGY CONSUMPTION IN CHILE 2008 .................................................... 51
FIGURE 7- ENERGY CONSUMPTION PER SECTOR IN CHILE 2008 ............................................................... 51
FIGURE 8- SOURCE OF ENERGY CONSUMPTION PER SECTOR IN CHILE 2008............................................ 52
FIGURE 9- ENERGY CONSUMPTION IN CHILEAN INDUSTRY AND MINING SECTORS 2008 ........................ 53
FIGURE 10- EVOLUTION OF ELECTRICITY GENERATION IN CHILE 1998-2010 ............................................ 53
FIGURE 11- FINAL CONSUMPTION OF ELECTRICITY IN CHILE, 2010 .......................................................... 54
FIGURE 12- FINAL CONSUMPTION OF ELECTRICITY IN CHILE BY MINING AND INDUSTRY SECTORS, 201054
FIGURE 13- ELECTRICITY GENERATION MARKET SHARES IN THE SIC, 2008 .............................................. 55
FIGURE 14- REMUNERATION IN THE CHILEAN ELECTRICITY MARKET ....................................................... 57
FIGURE 15- ECONOMY AND POPULATION SIZES OF A SAMPLE OF LATIN AMERICAN AND WORLD
COUNTRIES IN 2008.......................................................................................................................... 62
FIGURE 16- TPES AND ELECTRICITY CONSUMPTION FOR A SAMPLE OF WORLD AND LATIN AMERICAN
COUNTRIES IN 2008.......................................................................................................................... 63
FIGURE 17- 2008 TPES PER CAPITA AND PER GDP FOR A SAMPLE OF WORLD AND LATIN AMERICAN
COUNTRIES ....................................................................................................................................... 63
FIGURE 18- 2008 ELECTRICITY CONSUMPTION FOR A SAMPLE OF WORLD AND LATIN AMERICAN
COUNTRIES ....................................................................................................................................... 64
FIGURE 19- NODAL PRICE 1982-2010 IN CHILE .......................................................................................... 65
FIGURE 20- ELECTRICITY PRICES FOR INDUSTRY IN THE OECD AND A SAMPLE OF DEVELOPING
COUNTRIES, 2008 AND 2009 ............................................................................................................ 66
FIGURE 21- ELECTRICITY PRICES FOR HOUSEHOLDS IN THE OECD AND A SAMPLE OF DEVELOPING
COUNTRIES, 2008 AND 2009 ............................................................................................................ 66
FIGURE 22- DISTRIBUTION OF CHILEAN CDM PROJECTS AND GHG EMISSION REDUCTIONS PER
TECHNOLOGY ................................................................................................................................... 69
FIGURE 23- POPULATION AGED 25-34 AND 55-64 THAT HAS ATTAINED UPPER SECONDARY AND
TERTIARY EDUCATION, 2007 (PERCENTAGE) ................................................................................... 71
FIGURE 24 - PISA SCIENCE MEAN SCORE ................................................................................................... 71
FIGURE 25- GROSS EXPENDITURE ON R&D AS A PERCENTAGE OF GDP, 2008 .......................................... 72
FIGURE 26- RESEARCHERS PER THOUSAND EMPLOYED ............................................................................ 73
FIGURE 27- TRIADIC PATENT FAMILIES PER MILLION INHABITANTS IN EMERGING COUNTRIES 2007 ..... 73
FIGURE 28- SCIENCE AND INNOVATION PROFILE OF CHILE....................................................................... 74
FIGURE 29- SCIENCE AND INNOVATION PROFILES OF BRAZIL, CHINA, INDIA AND SOUTH AFRICA .......... 75
FIGURE 30- CHILE´S EXPORT COMPOSITION IN 2006 ................................................................................ 80
FIGURE 31- INDUSTRIAL DEVELOPMENT POSITIONING OF UPPER-MIDDLE INCOME ECONOMIES, 1993 82
FIGURE 34- STEPS IN CASE STUDY RESEARCH ............................................................................................ 90
FIGURE 35- CASE STUDY CHARACTERISTICS .............................................................................................. 95
FIGURE 36- FRAMEWORK OF ANALYSIS FOR CASE STUDIES OF TECHNOLOGY TRANSFER ....................... 97
FIGURE 37- FIBROVENT´S TECHNOLOGY TRANSFER PROCESS .................................................................. 99
FIGURE 38 – FIBROVENT´S WIND BLADE PRODUCTION PLANT PROJECT ................................................ 103
FIGURE 39 - MOLTEN SALTS PROCESSING SYSTEM ................................................................................. 105
FIGURE 40- SQM TECHNOLOGY TRANSFER PROCESS .............................................................................. 105
FIGURE 41- -THE LITHIUM INNOVATION CENTER TECHNOLOGY TRANSFER PROCESS............................ 110
FIGURE 42- MICRO-HYDRO PLUG & PLAY TECHNOLOGY TRANSFER PROCESS ........................................ 116
FIGURE 43- MICRO-HYDRO PLUG & PLAY TURBINE CONCEPT ................................................................ 118
FIGURE 44- KDM TECHNOLOGY TRANSFER PROCESS .............................................................................. 121
FIGURE 45- GENERATORS AT LOMA LOS COLORADOS I POWER PLANT ................................................. 123
FIGURE 46- ENERGÍA DEL SUR TECHNOLOGY TRANSFER PROCESS ......................................................... 125
FIGURE 47- KWB AND BINDER BIOMASS BOILERS DISTRIBUTED BY ENERGIAS DEL SUR ........................ 127
FIGURE 48- THE TECHNOLOGY TRANSFER PROCESS IN ELECTRICA NUEVA ENERGÍA ............................. 129
FIGURE 49- BIOMASA CHILE TECHNOLOGY TRANSFER PROCESS ............................................................ 131
x
FIGURE 50- BIOMASA CHILE´S WOOD CRUSHING EQUIPMENT .............................................................. 133
FIGURE 51- MNC-GEN TECHNOLOGY TRANSFER PROCESS...................................................................... 135
FIGURE 52- MAINSTREAM CHILE TECHNOLOGY TRANSFER PROCESS ..................................................... 140
FIGURE 53- NEW FINANCIAL SECTOR INVESTMENTS IN CLEAN ENERGY, 2009, US$ BILLION ................. 177
FIGURE 54- LOAD FACTORS OF PRINCIPAL COMPONENTS 1 AND 2 ........................................................ 197
FIGURE 55: LOAD FACTORS OF PRINCIPAL COMPONENTS 1 AND 3 ........................................................ 198
FIGURE 56 - COUNTRY PERFORMANCE PRINCIPAL COMPONENTS 1 AND 2 ........................................... 202
FIGURE 57 - COUNTRY PERFORMANCE PRINCIPAL COMPONENTS 1 AND 3 ........................................... 202
FIGURE 58- AGGLOMERATION COEFFICIENTS ......................................................................................... 217
FIGURE 59- HIERARCHICAL CLUSTERING MEANS PLOT ........................................................................... 221
FIGURE 60- NON-HIERARCHICAL CLUSTERING MEANS PLOT .................................................................. 224
FIGURE 61 HIERARCHICAL (WARD’S) CLUSTERING MEANS PLOT FOR FIRST ITERATION WITH LOGS ..... 227
FIGURE 62- NON-HIERARCHICAL (K-MEANS) CLUSTERING MEANS PLOT FOR FIRST ITERATION WITH LOGS
........................................................................................................................................................ 227
FIGURE 63- MEANS PLOT OF TT VARIABLES FOR WARD’S CLUSTERS WITH LOGS .................................. 233
FIGURE 64- MEDIANS PLOT OF TT VARIABLES FOR WARD’S CLUSTERS WITH LOGS ............................... 234
FIGURE 65- MEANS PLOT OF TT VARIABLES FOR K-MEANS CLUSTERS WITH LOGS ................................ 236
FIGURE 66- MEDIANS PLOT OF TT VARIABLES FOR K-MEANS CLUSTERS WITH LOGS ............................. 237
FIGURE 67- CLUSTER SELECTION FOR TECHNOLOGY TRANSFER RECIPIENT COUNTRIES ........................ 240
FIGURE 68- CLUSTER SELECTION FOR TECHNOLOGY TRANSFER RECIPIENT COUNTRIES ........................ 262
FIGURE 69- ELECTRICITY PRODUCTION COSTS IN 2007, 2020 AND 2030 ................................................ 299
FIGURE 70 - RENEWABLE ENERGY SHARE OF GLOBAL FINAL ENERGY CONSUMPTION, 2008 ................ 301
FIGURE 83- RESIDUALS VS PREDICTED VALUES PLOT OF THE SELECTED MODEL TO EXPLAIN REIMPPCLOG
........................................................................................................................................................ 336
........................................................................................................................................................ 337
FIGURE 85- RESIDUALS VS PREDICTED VALUES PLOT OF THE SELECTED MODEL TO EXPLAIN
RECAPTTPCLOG .............................................................................................................................. 338
xi
1 Introduction1
1.1. Introduction
1.2. International policy context
1.3. Objective and research
questions
1.4. Thesis structure
1
Part of this chapter is published in Climate Policy (in Press, available online: 29 Sep 2011). “How
to increase technology transfers to developing countries: a synthesis of the evidence”. Authors: Ana
Pueyo, Maria Mendiluce, Maria Jesus Sanchez Naranjo and Julio Lumbreras.
1
1.1 Introduction
Developing countries are experiencing unprecedented levels of economic growth. As a
result they will be responsible for most of the future growth in energy demand and
greenhouse gas (GHG) emissions (IEA, 2010). The largest fast-growing countries, such
as Brazil, China and India, will cover most of this growth.
Therefore, curbing GHG emissions in developing countries has become one of the
cornerstones of a future international climate change agreement under the United
Nations Framework Convention for Climate Change (UNFCCC). However, setting caps
for developing countries’ GHG emissions is facing strong resistance in the current
round of negotiations. Continued economic growth that allows poverty eradication is
still the main priority of most developing countries, and caps are perceived as a
constraint to future growth prospects. The development, transfer and use of low-
carbon technologies have more positive connotations. Technology could guide the
path towards achieving sustained growth without compromising the climate.
Since its inception, the UNFCCC has recognised the importance of technology transfers
(TTs) in achieving the stabilisation of global emissions. In the 13th Conference of the
Parties (COP-13), held in 2007 in Bali, technology became one of the four pillars of an
expected post-2012 climate change regime. More recently, at COP-16, held in
December 2010, the Cancun Agreements decided to establish a technology mechanism
(TM) that contains a technology executive committee (TEC) and a climate technology
centre and network (CTCN). The objective of the TM is to enhance clean technology
development and diffusion. The COP17, held in December 2011 in Durban achieved
more concrete outcomes regarding the type of body that the TEC will be, with the
adoption of its modalities and procedures. COP17 also adopted the terms of reference
(ToR) for the CTCN and requested the CTCN to elaborate its modalities and procedures
based in these ToR once it is operational. A decision by the COP on the modalities and
procedures of the CTCN is not expected until 2013 in COP19. Besides, the relationship
between the TEC and the CTCN and the essential linkage between the TM and the
financial mechanism still need to be further elaborated.
Other significant initiatives outside of the UNFCCC, such as the ongoing creation of
climate innovation centres in developing countries, promoted by the World Bank, also
call for research that informs the design of effective TT policies adapted to each
country´s circumstances.
Unfortunately, so far the success of the UNFCCC process in promoting technology
transfer has been limited because the Convention has been entangled in long
discussions that have delayed action, while the mechanisms it has created have either
failed to materialise in actual TT or have led to progress on a project-by-project basis
that has been unable to scale-up to the level required. Additionally, TTs are inherently
difficult to define and measure (IPCC, 2000), which makes it difficult to assess the
extent of the transfers and their effectiveness in achieving actual emissions reductions
and contributing to the technological development of recipient countries. As a result,
businesses and developing country policymakers often complain about the long
distance between the bureaucratic UNFCCC processes and their actual and urgent
needs.
2
This thesis is an effort to shed more light on the gaps of the current UNFCCC approach
to TT, and to improve understanding of the concept and the measurement of TT, as
well as the factors and policies that can enhance them. The thesis focuses on the
international transfer of mitigation technologies and particularly on clean energy
technologies. It avoids entering the debate about transfers from R&D to commercial
stages. Central to this research is the understanding that developing countries are
highly heterogeneous and international policies can only be effective if they target the
specific needs of each country.
1.2 International policy context: technology transfer in the UNFCCC
1.2.1 Status of technology transfer in international climate change negotiations

TT is an important element of the UNFCCC, is included in several articles of the
Convention text (mainly articles 4.1, 4.3, 4.5 and 4.7) and has been part of the agenda
at every Conference of the Parties (COP). Since its adoption in 1992, the UNFCCC
approach to TT has been based in two notions:
• Division of the world into developed and developing countries. The Convention
assumes that developed countries have the capacity to undertake research,
development, deployment and the transfer of technologies (Article 4.5), while
developing country parties are seen as recipients of technologies, know-how
and finance as a precondition for action (Article 4.7).
• Requirement of broad technological support from developed country parties to
developing country parties. This includes the transfer of equipment and know-
how, and support for the development and enhancement of endogenous
capacities and technologies in developing countries (article 4.5).
At COP 13, held in Bali, technology was identified as one of the four building blocks for
a future climate change regime as outlined in the Bali Action Plan. COP 15, held at
Copenhagen in 2009, failed to deliver the new binding and ambitious international
climate agreement that had been outlined in Bali. Instead, it contributed to fracture
static UNFCCC preconceptions about the developing world and to reveal new
structures of power (Grubb, 2010), with China particularly, but also India and Brazil,
playing a prominent role in drafting the non-binding Copenhagen Accord. COP 15 also
portrayed the limitations of multilateral treaties for securing consensus on mitigation
commitments, thus unleashing a renewed interest in bilateral agreements and small
group agreements including the most powerful countries2.
However, Copenhagen achieved some progress in the area of technology because the
Copenhagen Accord proposed a technology mechanism (TM) as the basis for
subsequent negotiations. Its aim was “to accelerate technology development and
transfer in support of action on adaptation and mitigation that will be guided by a
country-driven approach and be based on national circumstances and priorities”. This
new institution was mainly brought forward by developing countries, specifically the
G77 and China, while Annex I countries preferred using or modifying existing
institutions (Staley and Freeman, 2009). Financial support was also mentioned in the
2
As declared by Nitin Desai, from the Indian delegation in ‘When Two's Company’, Times of India, 4
January 2010.
3
Copenhagen Accord through $30 billion in fast-track funding from developed to
developing countries for the period 2010-2012 and the creation of a green fund,
mobilising $100 billion a year by 2020. However, it was not clear what part of these
funds would be available for TT.
The Cancun Agreements of COP 16 in December 2010 placed the Copenhagen Accord
under the auspices of the UN, with the approval of the 193 countries working under
the Convention. Much work still needs to be done to establish a comprehensive, long-
term framework for controlling GHG emissions, particularly on the definitions of
emission reduction targets and timetables. Nonetheless, significant outcomes were
achieved, including the decision to establish a TM, including a TEC and a CTCN that
should be fully operational by 2012. Provisions relating to the establishment of the TM
are contained in Section IV B of Decision 1/CP.16 of COP 16 of the “Cancun
Agreements” (UNFCCC, 2010a). Its objective is to “facilitate the implementation of
actions” and “to accelerate action consistent with international obligations” at
different stages of the technology cycle (UNFCCC, 2010a). Funding availability for the
TM, as well as the institutional arrangements for allocating these funds, are still under
discussion, and eligibility criteria for countries and technologies have not yet been
addressed. COP 17 held in Durban in November-December 2011 achieved some
progress by approving the modalities and procedures of the TEC and approving the
ToR of the CTCN, encouraging the latter to elaborate its modalities and procedures.
However, the issues of finance availability and relationships between different
institutions still need to be resolved.
After some divergences between developing and developed country views on the
impact of intellectual property rights (IPRs) in TT, there was no mention about IPRs in
the final text approved in Cancun, i.e. they were not central to any country’s proposals.
However, developing and developed countries have offered passionate views on this
matter in that developed countries consider IPRs as essential in promoting innovation,
while in contrast some developing countries proclaim that IPRs deter TT and have
proposed compulsory licensing of certain clean technologies (defended by China and
India). Still, some of the emergent economies have softened their stance towards IPRs,
as they have increased their ownership of climate change technologies (Staley and
Freeman, 2009)
Throughout the negotiations, the significant role of the private sector and the need of
policy reforms in developing countries to encourage private investment have been
stressed by developed countries. Developing country submissions to the UNFCCC
negotiations have instead focused on requesting developed countries to meet their
responsibility to provide finance and technology (WRI, 2009; Marcellino and
Gerstetter, 2010). Annex IV of the Cancun Agreements mentions that the Technology
Executive Committee should seek input from the private sector and civil society, which
has moved the need to include the private sector from the main text of the Bali Action
Plan to an Annex that suggests that it is not foreseen as a priority.
1.2.2 The Effectiveness of existing UNFCCC technology transfer catalysts

The UNFCCC has used three main types of catalysts of climate change TT to developing
countries:
• Institutions, namely the Expert Group of Technology Transfer (EGTT);
4
• The creation and diffusion of information through technology needs
assessments (TNAs) and the technology transfer information clearing house
TT:CLEAR; and
• Financial vehicles like the global environment facility (GEF) and the Clean
Development Mechanism (CDM).
The EGTT was created in 2001 to report to the Subsidiary Body for Technological and
Scientific Advice (SBSTA), with the objective of enhancing the implementation of
Article 4.5 of the Convention, including, inter alia, analysing and identifying ways to
facilitate and advance TT activities. In Bali, the EGTT was asked to also report to the
Subsidiary Body for Implementation (SBI) and was given the task of preparing reports
on performance indicators (FCCC/SB/2009/4), financial needs (FCCC/SB/2009/2) and
future strategy (FCCC/SB/2009/3). At the request of the G77 and China (IISD, 2006), a
new technology mechanism was considered in the Copenhagen Accord and decided in
the Cancun Agreements to succeed the EGTT.
TNA is a systematic approach used to “identify, evaluate, and prioritise technological
means for achieving sustainable development in developing countries, increasing
resilience to climate change, and avoiding dangerous anthropogenic climate change”
(UNDP, 2009). TNAs are country-driven activities, undertaken within the UNFCCC
framework and supported by the GEF, the United Nations Development Programme
(UNDP) and the United Nations Environment Programme (UNEP) as funders and
coordinators.
The CDM was created by the Kyoto Protocol (1997) to reduce the cost of compliance
for Annex I countries by taking advantage of cheaper emission reduction opportunities
in non-Annex I parties and to support sustainable development in these host countries.
The GEF is the financial arm of the UNFCCC, as well as other UN Conventions, and it
provides grants to developing countries and nations with economies in transition for
projects related to biodiversity, climate change, international waters, land
degradation, ozone layer depletion and persistent organic pollutants. Neither the CDM
nor the GEF were created with the aim of funding TT, but they have done so indirectly.
The success of UNFCCC efforts to promote TT has been limited. The EGTT has been
criticised for delaying hard but necessary decisions to enhance TT and for the lack of
expertise of its political representatives (SC & CIEL, 2008). TNAs have identified
potential projects but these have failed to materialise in implementation. The GEF has
only a limited budget, which has resulted in a lower scale than the market-based CDM.
A recent study shows that 40% of CDM projects accounting for 59% of estimated
emission reductions (roughly 335 MtCO2/ year) claim to involve TT3 (Seres et al., 2010).
The CDM has contributed to technology diffusion, reducing the payback period and
improving the internal rate of return (IRR) of clean technology projects (Hansen, 2008;
Ang, 2009). However, the CDM as a vehicle for technological change has been widely
3
Data in Seres et al. (2010) about TT in CDM projects rely on TT claims made by project participants in
the Project Design Documents (PDDs) of 4,984 projects that were in the CDM pipeline as of 30 June
2010. Only the projects for which the PDD explictly states whether or not TT is expected or not are
considered to calculate percentages. These claims are subject to uncertainty as they are not based on a
common definition of TT and they have not gone through verification unless they belong to the
additionality test. Also, they refer to projects often at the design stage, which may not be successfully
implemented and hence may not deliver emission reductions.
5
criticised in the literature. Firstly, it has not built to the scale required to meet the
stabilisation challenge (UNFCCC, 2008; McKinsey & Company, 2009). Secondly, its
project-based nature does not foster the large-scale deployment of mitigation
technologies or the promotion of innovation in host countries (Staley and Freeman,
2009). Thirdly, many of the emission reductions achieved by the CDM are not
“additional”, meaning that they would have happened anyway and should not be
financially supported (Wara, 2009). Several studies of TT in Chinese, Indian, Brazilian
and Malaysian CDM projects show that certified emissions reductions (CERs) income
was rarely the primary reason why the projects were developed, because of the
uncertainty of carbon income and long CDM registration time lags (Hansen, 2008;
Wang, 2010; Lewis, 2007; Hultman et al., 2010; He and Morse, 2010). Finally, CDM
projects have concentrated on the largest emerging economies, while African
countries and other least developed countries (LCDs) have been largely ignored (Unep
Risøe, 2010). Table 1 summarises the main achievements and pitfalls of these
mechanisms in the promotion of TT.
TABLE 1- EFFECTIVENESS OF UNFCCC MECHANISMS IN PROMOTING TECHNOLOGY TRANSFERS
Instrument Achievements Pitfalls

• Improvement in the • Members are political appointees instead of
EGTT
understanding of financial and experts in TT, IPR or clean technologies (South
capability gaps to enhance TT Centre and CIEL, 2008)
through a number of reports • Considered by the G77 plus China (IISD, 2006) as a
and workshops mere fact-finding mechanism that has delayed
• Enhanced dialogue with the hard but necessary decisions to enhance TT
private sector since 2009
• 68 TNAs prepared and more • No projects or programmes from the TNAs have
TNA
than 200 project proposals and yet been implemented
ideas • Lack of coordination between the TNA process
• Identification of in-country and national planning processes
capacity gaps in developing • Absence of a systematic approach for financing
countries
• Attempt to turn proposals into
specific projects through a
recent call for proposals for TT
pilot projects
• 40% of CDM projects (1,516 • Data on funding and emission reductions of CDM
CDM
projects), accounting for 59% of projects involving TT are not systematically
estimated emission reductions collected
4
(roughly 335 MtCO2/ year) • Quality of CDM TT is not systematically assessed
claim to involve TT (Seres et al., • Finance flows channelled by the CDM account for
2010). a small share of the estimated funding needed to
• The CDM has contributed to reduce emissions in developing countries
technology diffusion, reducing (UNFCCC 2010)
the payback period and • The reductions achieved by the CDM are not
improving the internal rate of always “additional”, meaning that they would
return (IRR) of clean technology have happened anyway and should not be
projects (Hansen, 2008; Ang, financially supported (Wara, 2009)
2009) • The CDM usually contributes to diffuse low-risk
technologies at the later stages of maturity
4
TT quality can be defined as the degree to which it raises the recipient´s technological know-how and
capacity to innovate (Popp, 2008; Schneider et al., 2008)
6
Instrument Achievements Pitfalls
• CDM financing has not filled the gaps left by the
private sector, concentrating on China, India and
Brazil and in low risk technologies (UNEP Risø,
2010)
• The project-based nature of the CDM does not
foster the large-scale deployment of mitigation
technologies or the promotion of innovation in
host countries (Staley and Freeman, 2009)
• A significant share of the • Small budget (190 MUSD in 2009) compared to
GEF
budget is allocated to projects CDM financing and to other funds
in least-developed countries • Most emission reductions achieved in big projects
and innovative technologies in China, Brazil, India and Russia, therefore not
(Peterson, 2008). filling the gaps left by the private sector
(Peterson, 2008).
1.2.3 The Technology Mechanism

The Technology Mechanism is ultimately placed “under the guidance of the COP”. In
general, the wording of the technology mechanism’s mandate appears rather intricate
and convoluted compared to previous formulations in the negotiations. In the
definition of the role of the TM, The COP has underlined “the importance of nationally
determined technology needs, based on national circumstances and priorities, the
setting of appropriate enabling environments to scale up the development and
transfer of technologies in developing countries and the need to accelerate action at
different stages of the technology cycle”5
The COP 16 decision lists a number of ‘priority areas’ to be considered under the
Convention:
• Development and enhancement of endogenous capacities and technologies of
developing country parties, including cooperative research, development and
demonstration programmes;
• Deployment and diffusion of environmentally sound technologies and know-
how in developing country parties;
• Increased public and private investment in technology development,
deployment, diffusion and transfer;
• Deployment of soft and hard technologies for the implementation of
adaptation and mitigation actions;
• Improved climate change observation systems and related information
management;
• Strengthening of national systems of innovation and technology innovation
centres; and
• Development and implementation of national technology plans for mitigation
and adaptation
5
Draft decision [-/CP.17] on the Outcome of the work of the Ad Hoc Working Group on Long-term
Cooperative Action under the Convention. Durban, 2011.
7
In order to operationalise the mechanism, the parties established the Technology
Executive Committee (TEC) to oversee it, to facilitate between governments and the
private sector and to recommend actions to address barriers to technology
development and transfer. Furthermore, the parties established the Climate
Technology Centre and Network (CTCN) to “facilitate a network of national, regional,
sectoral and international technology networks, organisations and initiatives”.
At COP 16 in Cancun, the parties decided that the TEC would comprise twenty experts
elected by the COP, serving in their personal capacity and nominated by the parties.
Nine of the twenty experts should come from Annex I countries, nine from non-Annex I
countries, three per region (Africa; Asia and the Pacific; and Latin America and the
Caribbean), one from a small island developing state and one from an LDC. The TEC
would make decisions drawing upon outside expertise, including that provided by the
CTCN and the private sector.
Ahead of COP17 in December 2011, The Technology Executive Committee had its first
meeting from 1st-3rd September 2011, in order to elect its members and elaborate on
its modalities and procedures, which would be considered at COP 17 in Durban. A
background paper on its modalities was published for this meeting (TEC, 2011).
The modalities and procedures of the TEC were adopted by COP17 in Durban6. They
establish the following six key elements of the modalities of the TEC:
a. Analysis and synthesis; including mainly the elaboration of technology outlooks,
technical papers on specific policies and technical issues and regular overview
of existing technology development and transfer initiatives, activities and
programmes.
b. Policy recommendations; mainly to the COP as regards actions, policies and
programmes to promote technology development and transfer.
c. Facilitation and catalysing; through the promotion of collaborations with
relevant organisations, organising forums and workshops to share knowledge.
d. Linkage with other institutional arrangements; inside and outside the
Convention.
e. Engagement of stakeholders; through issue-based engagement channelled
through work programmes. Stakeholders can be involved through their
participation in TEC meetings as observers or expert advisors or through other
models such as consultative groups or stakeholder forums.
f. Information ad knowledge sharing; including the upgrade of TT:CLEAR with an
expanded and more strategic focus.
As defined in the adopted modalities and procedures of the TEC, it will mainly have an
advisory role. Its functions are quite general and its mandate does not seem to go far
beyond what has already been done by the existing EGTT, which was considered by
developing countries as a “mere fact finding mechanism that has delayed hard but
necessary decisions to enhance TT” (IISD, 2006). Besides, the relationships between
the TEC and the CTCN, and the linkage between the technology mechanism and the
financial mechanism, are still under negotiation.
6
Draft decision -/CP.17- Technology Executive Committee – modalities and procedures. Advanced
unedited version, Durban, December 2011.
8
The CTCN has not yet developed its modalities and procedures. COP 17 agreed the ToR
of the CTCN and requested it to elaborate its modalities and procedures once it would
be fully operational, presumably by 2012. According to the agreed text in Durban, the
COP is not expected to make a decision on the modalities and procedures of the CTCN
until COP19 to be held by the end of 2013.
The main aim of the CTCN is to provide services to developing countries at their
request, consistently with their respective capabilities and national circumstances and
priorities. Some of the roles considered for the CTCN in the ToR approved in Durban
are:
a) Identifying currently available climate-friendly technologies for mitigation and
adaptation that meet their key low-carbon and climate-resilient development
needs;
b) Facilitating the preparation of project proposals for the deployment, utilization
and financing of existing technologies for mitigation and adaptation;
c) Facilitating adaptation and the deployment of currently available technologies
to meet local needs and circumstances;
d) Facilitating research, development and demonstration of new climate-friendly
technologies for mitigation and adaptation, which are required to meet the key
objectives of sustainable development;
e) Enhancing national and regional human and institutional capacity to manage
the technology cycle;
f) Helping to facilitate the financing of the previous activities through various
sources including the financial mechanism of the Convention, bilateral,
multilateral and private sector channels, philanthropic sources as well as
financial and in-kind contributions from the host of organization and
participants in the Network;
g) The most original and challenging function of the CTCN is the facilitation of a
network of national, regional, sectoral and international technology centres,
networks, organisations and initiatives. However, there is ambiguity in the term
“facilitate” in this and all the above functions, and it is not clear what the role
of the CTCN will be in achieving such a network. Moreover, there are some
overlaps between the activities of the CTCN and the TEC.
.
1.2.4 The Prospects of UNFCCC instruments in promoting climate change TT to developing

countries post-2012
Lack of clarity about the functions, institutional design and funding of the new TM
decided in the Cancun Agreements may render it futile in enhancing TT beyond what
the existing instruments of the UNFCCC have been able to achieve. The new TM could
potentially achieve higher rates of TT to developing countries if some gaps in the
current UNFCCC approach were addressed.
First, the UNFCCC deals with TT as a government-to-government process, urging
developed country parties to support technological development in developing
countries (as Article 4.5 of the Convention insists) and providing guidelines to
developing countries on how to prioritise technological means for achieving low-
carbon development paths (through the Technology Needs Assessment handbook).
9
Decisions in the EGTT were made by political appointees instead of experts with first-
hand experience in the development and transfer of climate change technologies, and
no significant efforts have been made to engage the private sector in the definition of
an effective technology policy (UNCTAD, 2010; WBCSD 2010). This approach neglects
the key role of the private sector as the owner of most of the climate change
mitigation technology and responsible for most international TT through trade and
foreign direct investment (FDI) (Stern, 2007; Brewer, 2009; WB, 2009). The new
technology mechanism will seek expertise from external parties, including all kinds of
observers, many of whom will be from the private sector. However, no specific
position is given to the private sector in the institutional arrangements of the TM,
which is therefore expected to provide ad hoc advice when required. The members of
the TM will still be political appointees, as in the preceding EGTT, and its functions are
vague and confusing.
The UNFCCC process appears disconnected from the enabling frameworks that
facilitate private investment, and until recently it has not attempted to understand
strategic business decisions regarding TT. On the one hand, existing instruments used
to promote TT (EGTT, TNA, CDM, GEF) are not integrated into national planning
processes. When they materialise on specific TT, it is done on a project basis that does
not permeate the rest of the economy. These mechanisms seem insufficient for
removing existing barriers, leverage a great amount of private investment and
promote endogenous technologies in developing countries. The updated TNA
handbook recognises the need to step away from the project-based approach and
considers TT in light of long-term visions (UNDP, 2009). In addition, the new TM does
not directly support the development and implementation of national policies for
technology development and adoption by developing countries themselves. Besides,
the absence of globally binding emission reduction targets after the COPs in
Copenhagen, Cancun and Durban, and the sluggishness and inefficiency of
international climate change negotiations, have not provided the strong signals
necessary to stimulate private sector investment in low-carbon technologies in
Second, the UNFCCC discourse is based on a north-south TT paradigm that divides the
world into two blocks: technologically active and passive (Brewer, 2008; Cannady,
2009). Observation of actual technology flows shows that this simplistic paradigm no
longer holds. In an increasingly interconnected world, a group of large and dynamic
developing countries are actively absorbing foreign technologies and developing and
transferring endogenous ones.
FDI data compiled by the World Investment Report 2010 (UNCTAD, 2010) shows that
over a quarter of greenfield investments in alternative/renewable power generation
were placed in developing economies. Nearly 10 per cent of renewable power
generation investment projects were made by trans-national companies (TNCs) from
developing countries, the bulk of which were south-south oriented. More illustrative is
the case of greenfield investments in the manufacturing of clean energy technology
products such as wind turbines, solar panels or biodiesel plants. Developing economies
attracted nearly half of the projects in this industry over 2003-2009 and have
surpassed developed countries in the last two years (UNCTAD, 2010). Still,
performance across developing countries is highly unequal, with most investment
10
concentrated in China, India and Brazil, three of the so-called BRIC economies7. The
first two countries host nearly 20 per cent of the world’s wind power generation
capacity (WB, 2009) and are world leaders in a variety of climate-friendly technologies.
The diverse performance of the CDM across developing countries reflects a similar
pattern to FDI. China has the largest amount of CDM projects, distantly followed by
India and Brazil. These three countries held 72% of the CDM projects and 77% of the
associated GHG emission reductions as of June 2010 (Unep Risøe, 2010), while African
countries and LDCs were largely ignored. As regards TT facilitated by the CDM, a recent
study shows that China, India and Brazil have a lower than average TT rate (Seres et al.,
2010), and while BRIC countries use mostly local knowledge or equipment, the rest of
the host countries are more dependent on foreign technology to implement CDM
projects. The same study shows that non-Annex I countries supply technology to 15%
of CDM projects involving TT. China is among the top five technology suppliers, ahead
of many developed economies, while India and Brazil also hold a significant share
(Seres et al., 2010). These figures show that the Convention mechanisms have not
addressed the specific need of different developing countries to attract foreign climate
change technologies; instead they have behaved similarly to international private
investment flows.
The decision creating the technology mechanism emphasises the need for country-
specific policies to promote technology transfer through multiple references to
national needs, circumstances and country-driven approaches, and by declaring that
“technology needs must be nationally determined, based on national circumstances
and priorities”(UNFCCC, 2010), although its functions do not directly support national
action.
Additionally, there are no systematic measurements for the magnitude and
effectiveness of climate change TT to developing countries. Clear measurements of
technology transfer are necessary to assess the effectiveness with which different
developing countries assimilate foreign technologies. They can also provide
policymakers with valuable information about the most appropriate channels for low-
carbon technology transfer. In the current stage of the international climate change
negotiations, where developed countries have committed funds to support low-carbon
development in developing countries, the measurement of technology transfer will
also be essential because it will enable the definition of eligibility criteria for
technology funds, the assessment of their effectiveness and support policymakers to
make informed decisions about where to allocate available finance.
1.3 Objectives and research questions

The broad objective of this doctoral thesis is to improve current understanding about
what constitutes TT in the field of climate change, what are the factors that enable TT
in developing countries and which policies could be implemented to reinforce these
factors. It also analyses different approaches to measuring TT and its enabling factors,
assesses existing statistical sources and shows their limitations.
This broad objective can be divided into a set of research questions:
7
The term “BRIC countries” was coined by Goldman Sachs to refer to Brazil, Russia, India and China as a
group of large and fast-growing economies.
11
1. What constitutes climate change technology transfer and how does it happen?
Technology transfer has been defined in many different ways and by a wide range
of disciplines. The concept of technology is elusive because it does not always have
a measurable presence and its transfer goes beyond market-like exchanges. A clear
definition of what constitutes technology transfer is required to avoid overlaps in
the “mitigation” and ”technology” negotiation tracks of the UNFCCC, and to be
able to design targeted policies.
Section 2 reviews a wide range of literature about the concept of “technology
transfer” and the channels through which it takes place. Section 4 addresses the
lack of engagement of the UNFCCC process with the private sector by providing
first-hand information about how TT happens in ten case studies located in Chile.
The case studies offer a definition about what constitutes TT for the private sector
and shows the diversity of channels used to incorporate foreign knowledge into
Chilean low-carbon activities. The case of Chile was considered relevant for our
analysis because it fills a gap in the literature of climate change technology
transfer, which has so far focused on large emerging countries such as China or
India. Chile is also interesting because it has a number of enabling frameworks to
attract climate change TT, but their penetration has so far been slow. Therefore, an
analysis of the factors that have prevented further TT could be useful in drawing
policy conclusions applicable to other small and medium-sized developing
countries.
2. Is it possible to measure technology transfer, and what are the main challenges
in doing so?
Measuring TT is important for assessing the performance of the very policies
designed to support it. However, it is not an easy task due to the elusiveness of the
concept of technology. Current measurements of technology transfer available in
the UNFCCC provide a very limited view of the process. The design of effective
policies to support technology transfer in developing countries will require that
clear measurements are defined to assess their effectiveness.
Section 2 includes a revision of the different approaches used by the literature to
measure technology transfer inputs, outputs and effects, showing their main
limitations. Section 3 carries out a diagnosis of climate change TT in Chile and
assesses the limitations of existing data. Section 4 assesses the possibility of
measuring TT inputs, outputs and effects in specific case studies.
3. Empirically, which factors enable climate change technology transfer to
developing countries?
One of the limitations of the existing UNFCCC framework in promoting TT is its
disconnection with the enabling frameworks that facilitate private investment in
climate change technologies. Section 2 reviews the factors that have contributed to
promoting technology transfer to developing countries, as identified in the existing
literature.
Economic models employed to analyse the impact of enabling frameworks on TT
often present a narrow perspective of technology and cannot account for specific
country characteristics and different types of policies. They are also limited by a
12
lack of data on developing countries. Case studies can show a richer perspective,
but they are mostly focused on the success stories of BRIC economies. However,
these countries do not represent the average developing economy. The sheer sizes
of their financial systems attract foreign investors, and their solid industrial base
and targeted industrial policies allow the emergence of local renewable energy
industries. Smaller developing countries would struggle to replicate BRIC countries’
success stories, so further case studies are necessary to identify the enabling
factors that operate in these minor nations. Section 3 analyses the case of Chile to
show the perspective of a smaller developing country. Section 4 analyses the
enabling frameworks for technology transfer in ten specific case studies from Chile.
The aim of the analysis is to assess the importance of the defined enabling
frameworks for the success or failure of TT in real experiences of low-carbon
technology adoption.
4. How do different developing countries perform in these enabling factors, and
which differentiated policy priorities can be defined accordingly?
The group of developing countries jointly considered as non-Annex I by the
UNFCCC is highly heterogeneous, as each country has different technological needs
depending on factors such as size, income level, growth rate, technological
capabilities or industrial development. Targeted policies to enhance climate change
TT would maximise their impact on GHG emissions reductions and their
contribution to increasing recipient countries capabilities. Instead, UNFCCC
mechanisms have applied an undifferentiated approach for all developing
countries.
Section 5 will analyse the current status of a group of developing countries as
regards their level of climate change TT and their performance in the predefined
set of enabling factors. Using multivariate analysis techniques, we will analyse the
strengths and weaknesses of these countries and create clusters according to their
policy priorities.
1.4 Structure
Section 2 reviews the existing literature about the concept, channels, measurement
and enabling frameworks for climate change technology transfer to developing
countries. It attempts to answer the research questions and highlight any gaps in the
literature which require further research.
Section 3 carries out a diagnosis of the specific case of Chile, as regards its current
adoption of low-carbon technologies and its performance in the enabling factors
compiled in the literature review. It highlights strengths and weakness of the Chilean
case and limitations regarding data availability.
Section 4 looks at ten case studies on low-carbon technology research, development,
deployment and commercialisation in Chile, looking at the role that international
climate change TT has played, its enabling frameworks and barriers. Policy
recommendations are made on the basis of the expert opinion of private sector actors
involved in the cases.
Section 5 maps the current status of a group of developing countries as regards their
current levels of foreign climate change technology transfer and their performance in a
13
predefined set of indicators about enabling frameworks. Using multivariate analysis
techniques, it creates clusters of countries based on their international policy priorities
in an attempt to increase the volume and effect of climate change TT.
Section 6 concludes and provides international policy recommendations for improving
the level and quality of TT to developing countries.
14
2 Literature on climate change technology
transfer to developing countries8
2.1. Conceptualising technology transfer

2.2. Channels of technology transfer
2.3. Measuring technology transfer
2.4. Enabling frameworks for climate
change technology transfer
2.5. Conclusions
8
Part of this chapter is published in Climate Policy (in press, available online: 29 Sep 2011). “How
to increase technology transfers to developing countries: a synthesis of the evidence”. Authors: Ana
Pueyo, Maria Mendiluce, Maria Jesus Sanchez Naranjo and Julio Lumbreras.
15
2.1 Conceptualising technology transfer
The term “technology transfer” has been defined and measured in many different
ways and by a wide range of disciplines. This doctoral thesis is concerned with the
“horizontal” or international perspective of technology transfer that enables
developing countries to acquire, adapt, diffuse and promote climate change mitigation
technologies from overseas. Technology transfer is also defined by other publications
from a vertical perspective as the transfer of knowledge or inventions from the R&D
stage through to commercialisation.
Early research provided a narrow definition of technology as scientific and engineering
knowledge and blueprints or their manifestation in artefacts. The transfer of this
codified knowledge or its manufactured materialisation then constituted technology
transfer. The concept has evolved and now technology is defined in broader terms as
encompassing the corporate capacity to operationalise and use this knowledge
effectively in production (Cantwell, 2009). Technology in this broader sense has two
components: the potentially public element of technology, encompassing codifiable
items as presented in scientific publications and engineering blueprints and designs,
and the tacit element of technology that refers to firm-specific competence in
production. Tacit corporate technological capabilities cannot be transferred through
market-like exchanges and must instead be internally learned, with or without external
assistance (Cantwell, 2009). There is therefore a clear difference between technology
trade and real technology transfer, as the former is merely the import of equipment or
the execution of projects on a turnkey basis, while the latter involves mastering the
imported know-how of core technologies and the development and generation of
technologies utilising scientific and technological capacities (Cohen, 2004).
The importance of knowledge-based capacities within developing countries has made
some authors feel uncomfortable with the term “technology transfer”, which they see
as encouraging a view of technology as an object and transfer as a one-off transaction
that maintains dependency on host country suppliers (IPCC, 2000). “Technology
diffusion” is one of the proposed alternative terms (Grubler and Nakicenovic, 1991;
WBCSD, 2010), as it is seen to imply a two-way relationship between technology
suppliers and importers (IPCC, 2000, Section 1.4). This thesis uses indistinctively the
terms “technology transfer” and “technology diffusion”.
Several authors have distinguished three different flows of transferred technological
content, from lower to higher impact on the technological capabilities of the recipient
(Bell, 1987; Wei, 1995; Ockwell et al., 2008). The first flow encompasses capital goods
and equipment; increases the production capacity of the recipient but on its own does
not enable the recipient to use the imported facilities efficiently or to generate
technological change. The second flow includes skills and know-how for operating and
maintaining equipment. It places the human resources of the importer at the
technological level required to operate the imported technology efficiently, but
without indigenous efforts beyond learning how to use the technology it would not
enable technological change. The third flow encompasses knowledge and expertise for
generating and managing technological change. It creates new technological capacity
through technology transfers and active independent learning, creation and innovation
of the recipient.
16
Organisational learning is therefore an essential factor in the success of technology
transfer. Learning processes are broader when foreign knowledge is linked to the
wider structure of the host country economy through backward (with suppliers) and
forward (with clients) inter-firm linkages (Wei, 1995), which in turn are strong when
the host country has a network of local suppliers capable of supporting the operation
and maintenance of this new foreign technology (Ivarsson and Alvstam, 2005). The
creation of backward linkages requires that local suppliers learn how to produce inputs
or services for the imported technology at the required specifications and standards.
The creation of forward linkages requires that local companies learn to use the output
of the imported technology in their own production processes. Without these linkages
the importing economy will not benefit from knowledge spillovers, and instead will
become dependent on foreign suppliers for imported technologies inputs, or on
foreign customers for the sale of technology outputs.
In the field of climate change, the Intergovernmental Panel on Climate Change (IPCC)
provides a frequently quoted definition of climate change TT as “a broad set of
processes covering the flows of know-how, experience and equipment for mitigating
and adapting to climate change amongst different stakeholders such as governments,
private sector entities, financial institutions, non-governmental organisations, and
research or education institutions” (IPCC, 2000). Under this broad definition, TT could
provide the recipient country with the capacity to install, operate, maintain and repair
imported technologies, produce lower cost versions of imported technologies, adapt
imported technologies to domestic markets and circumstances and develop new
technologies, whilst respecting relevant intellectual property rights (UNFCCC, 2009).
2.2 Channels of technology transfer

Technologies can be transferred through several mechanisms. There is a link made in
the literature between different transfer mechanisms and the outcome of the transfer
process in terms of the impact on recipient country capacities. Ivarsson and Alvstam
(2005) characterise TT mechanisms by their level of integration as internalised or
externalised transfers. Internalised transfers are those in which the technology
supplier transfers a full technological package as part of foreign direct investment in
the recipient country. These transfers happen mostly through FDI in wholly-owned
subsidiaries and majority joint ventures. Externalised TTs are those made to firms
outside direct ownership and/or the control of the technology provider, such as in
cases of minority joint ventures, franchising, sales of capital goods, licences,
subcontracting or original equipment manufacturing arrangements. Externalised TTs
are more likely to generate new technological capacity in recipient countries than
internalised transfers, if in the latter the main motivation of the foreign multinational
companies (MNCs) is to exploit static location advantages such as cheaper production
costs, rather than engaging in deeper learning processes. External TT can have a higher
potential to contribute to technology upgrading, since local firms have to make an
effort to select, absorb, adapt to and improve the acquired technology, given that they
have the capabilities to undertake efficient learning. Still, MNCs are a dominant source
of innovation, and their direct investment is an important mode of international TT
that contributes to local innovative activities in host countries (UNCTAD, 2003).
17
Lema and Lema (2011) distinguish between “conventional” and “unconventional”
technology transfer on the basis of the degree of interaction between suppliers and
importers and the degree of internal effort and investment of the recipient.
Conventional TT requires low interaction and low internal effort, as it involves
technological artefacts available “off the shelf” and at a continuously declining cost.
Some examples of conventional TT are imports of capital equipment, the licensing of
technology and foreign majority joint ventures. Unconventional TT is highly interactive
and requires substantial investments from the recipient. Some examples of
unconventional TT are the acquisition of foreign technology owners, the establishment
of R&D departments overseas, collaborative R&D projects or joint ventures in foreign
countries. Unconventional TT allows developing countries to access global pools of
knowledge for innovation.
Some climate change TT mechanisms commonly used in China and India have been
licensing, joint development and the acquisition of foreign companies. For example,
the leading Chinese wind turbine manufacturer Goldwind produced its first 660-750
kW turbines using a licence from a German company and its 1.5 MW turbines in joint-
development with another German company, which it subsequently acquired to
produce the most advanced 2.5-5MW models. Goldwind actively pursues patenting all
new products to ensure market power (Lewis, 2007; Wang, 2010; Zhang et al. 2009).
Suzlon, the leading Indian wind turbine manufacturer, wind farm project developer
and operator built its capacity upon licensing agreements complemented with internal
R&D. Suzlon located R&D centres at knowledge hubs in Germany and the Netherlands,
and its international headquarters in Denmark. The company also acquired foreign
manufacturers of generators, gearboxes and wind turbines (Lewis, 2007). The Chinese
solar PV industry has instead relied on purchases of manufacturing equipment,
particularly turnkey production lines, and the recruitment of skilled Chinese
entrepreneurs from the Chinese diaspora (de la Tour et al., 2010).
Table 49 (in the Annex of Chapter 2) presents the main channels for private sector TT
identified by the IPCC (2000), and also highlights key issues and factors affecting the
choice of mechanism, the degree of integration and characterisation as conventional
or unconventional sources of TT. The table shows that unconventional TT mechanisms
had not yet been considered significant by the IPCC in its ‘Special Report on
Technology Transfer’.
2.3 Measuring technology transfer

Given the tacit nature of technology in its broader sense, measuring technology
transfers is inherently difficult because technology has no measurable physical
presence or a well-defined price (IPCC, 2000). Rather, it is embodied in products,
intermediate inputs and processes (World Bank, 2008). Moreover, as described in the
previous section, TT can occur through a diversity of channels for which data are not
always available.
Across different disciplines, three types of indicators can be defined, namely input,
output and effect (Neuhoff et al., 2009). If applied to the measurement of technology
transfer, input indicators can provide information on resources spent on activities to
facilitate TT activities or on the channels that make foreign technological inputs
available, some examples of which are expenditures in collaborative R&D, imports of
18
equipment or the salaries of foreign staff. Output indicators can measure the results of
technological inputs, for example, the number of patents issued, installed capacity of
RE projects, production volume or value of low-carbon technologies. Finally, effect
indicators can quantify the achievement of the long-term goals of the technology
transfer, such as CO2 emission reductions, technology cost reductions or knowledge
spillovers through backward and forward linkages with local suppliers and clients. The
problem with output and effect indicators is how to attribute the part of them that is
enabled by foreign, rather than local, technologies.
2.3.1 Economics of technology diffusion

Economic studies about technology diffusion across countries have attempted to
model the knowledge spillovers of foreign activities as the impact of foreign
technology on domestic productivity growth. They have also analysed the factors that
make some countries more effective than others in adopting foreign technology. Keller
(2004) and Saggi (2004) provide comprehensive reviews of this literature.
These studies use input, output and effect indicators to measure technology transfer.
The first is linked to expenditures on R&D for the development of technology, or to
foreign technology inputs such as imported equipment. Output indicators mainly refer
to the commercial result of the use of foreign technologies, such as patents issued and
final technologies produced. Finally, effect indicators usually refer to productivity
growth as a result of foreign activity.
Some technology transfer measurements based in input indicators relate foreign R&D
to its channels of diffusion. TT can be proxied as the bilateral import-share weighted
R&D stocks of trade partners (Coe and Helpman, 1995; Xu and Wang, 1999; Coe et al.,
1997; Hoekman and Javorcik, 2006) or through regressions where productivity is
jointly explained by foreign R&D and exposure to trade (Bayoumi et al., 1999).
However, imports are a weak form of technology transfer because the technology as
such is not available domestically – only its manufactured outcome. Accordingly, other
studies consider that import shares are not necessary for measuring the extent of TT,
as transfers do not necessary happen through trade (Keller, 1998; Keller, 1999;
Cameron, 2005). TT could also occur through internal efforts enabled by ‘absorptive
capacity’ in the form of human capital and internal R&D investments (Cameron, 2005).
The use of R&D investments as a proxy for technology has some limitations in that
internationally comparable data on R&D expenditures is published by the OECD, but
does not include most developing countries. The definition of R&D expenditures
focuses on resources spent on innovation and not those spent on imitation and
technology adoption, which is very important in developing countries. An additional
drawback of this indicator is that it ignores the stochastic nature of the process of
innovation. The relationship between R&D expenditure and innovation is not
deterministic, so domestic R&D can be treated as an important explanatory variable of
technology adoption and innovation, and not as a proxy for them. The return to R&D
expenditures varies substantially across agents and over time, which limits
comparability. Also, the return to publicly-funded R&D is lower than the return to
privately-funded R&D, which explains why many studies focus on business R&D
spending more than government spending to reflect technology development (Keller,
2004).
19
Patent counts are frequently used as a proxy of technology output (Eaton and Kortum,
1996; Archibugi and Michie, 1997; Branstetter, 2001). The benefits of patent data
relative to R&D are that they have been collected for a longer time and include
developing countries. Some problems with this indicator, though, are that simple
patent counts do not measure technology output well because a small number of
patents account for most of the value of all patents, although citation-weighted patent
counts could solve this problem. Another problem is that a large set of innovations,
particularly in developing countries, are not patented. Also, if technology is in part
non-codifiable, patent statistics miss that part (Keller, 2004). The number or share of
patent applications registered in a country that come from abroad can be used as a
proxy of diffusion (Archibugi and Pianta, 1996; Archibugi and Michie, 1997). Patent
files by foreign actors are a measurement of the global exploitation of technology, as
firms undertake the cost and effort involved in extending a patent abroad if they
expect to be compensated either by trading the invention in a disembodied way or by
exporting products which embody it. Patent citations have also been used as a
measure of innovation diffusion (Jaffe et al., 1993; Jaffe and Tratjenberg, 2000).
Citations delimit the scope of the property right conveyed by the patent and identify
the previous knowledge that inspired the innovations included in that patent.
Productivity indicators are used by several studies as a proxy of the effect of
technology. Total factor productivity (TFP) is usually the selected productivity
indicator, and it reflects the gross output divided by combined inputs (labour, capital,
energy and services)9. Controlling for the effect of other variables, differences in TFP
between countries will be due to spillovers created by the factor technology (Saggi,
2004). To explain the effect of TT on productivity, economic studies use regressions
where TFP is explained by domestic and foreign technology, the latter of which is
proxied through FDI (Blalock, 2001; Borensztein et al., 1998; Xu, 2000) or imports
(Eaton and Kortum, 2001; Xu and Wang, 1999; Mazumdar, 2001; Lichtenberg and
Potterie, 2000). Controlling for the impact of other variables, productivity changes are
a proxy for the knowledge spillovers created by the absorbed technologies.
2.3.2 Vertical technology transfer: knowledge transfer metrics

Literature concerned with the vertical perspective of technology transfer has
developed knowledge transfer metrics to allow public research organisations and
countries to monitor and compare their achievements in transferring knowledge to the
commercial sphere (EC, 2009; Holi et al., 2008; Jensen et al., 2009). Given the
centrality of knowledge transfer in international technology transfer, some of these
metrics are also applicable to our horizontal TT perspective. According to this literature
there are three major forms in which knowledge can be carried: as codified knowledge
(mainly through publications or patents), as internalised by people or as embedded in
artefacts such as machinery, software, new materials or modified organisms (EC,
2009). Knowledge transfer can have economic impacts in terms of turnover and
profits. Many factors, apart from the original invention, have an influence in the final
9
TFP is defined by the OECD (2001) as gross output divided by combined inputs (labour, capital, energy
services). Labour productivity is defined as the total output divided by the total number of workers or
the number of hours worked. Economist, T. (2010). "The Economist economic dictionary".
20
economic outcome; therefore, it is difficult to calculate any one part of the economic
result due to the knowledge transferred.
Although measuring the amount of knowledge transferred from public research
organisations is virtually impossible, two proxies can solve this impasse. Firstly,
knowledge transfers can be approximated by their economic value or price. In some
cases, it is transferred free of charge, such as in scientific publications put into the
public domain. In such circumstances, the price is not related to the value of
knowledge. In some other cases, public research organisations sell licences to use the
knowledge, and the price of the licence reflects the future value expected by the
prospective customer. The licence could also be payable through for example equity
shares in spin-offs. A second option for measuring knowledge transfers is to count the
number of manifestations of knowledge transfer as activities in various transfer
channels, such as: the number of established spin-off firms derived from the invention,
the number of lectures given in network seminars, the number of patents applied for
or granted or the number of licences issued. Apart from the number of manifestations
their quality should also be assessed.
A European Commission report (EC, 2009) selects seven core knowledge transfer
indicators: the number of research agreements with private companies, the number of
invention disclosures evaluated, the number of priority patent applications, the
number of patents granted, the number of licences executed, licence income earned
and the number of spin-offs expressly established to develop or exploit intellectual
property (IP) or know-how created by the public research organisation. All the
indicators except licence income earned follow the approach of counting knowledge
transfer manifestations.
A report commissioned by UNICO in the UK (Hoi et al., 2008) establishes a larger
number of indicators of the quantity and quality of nine different “knowledge transfer
activities”, or channels, as shown in Table 50 (in the Annex of Chapter 2). Indicators
included can be value or manifestation-related. The report uses these metrics to
perform a benchmarking analysis of universities in the UK, US and Canada.
The report by the Intellectual Property Research Institute of Australia (Jensen et al.,
2009) uses the UNICO framework as a starting point. It proposes indicators classified
per the same types of knowledge transfer activities but is not only interested in the
knowledge transfer activities per se, but also considers the economic effects of the
commercialisation of this knowledge. Therefore, for each activity it includes some
indicators of their effect. The proposed indicators can be seen in Table 51.
2.3.3 Climate change technology transfer

Economic literature focusing specifically on the international transfer of low-carbon
technologies is still scarce in comparison to the general literature on technology
diffusion. Pizer and Popp (2008) identify two potential avenues in which foreign
knowledge can have an influence on the domestic economy. First, foreign inventions
could increase the productivity of domestic R&D by inspiring domestic inventors. In
this case, foreign technology could be represented by foreign patents and technology
transfers by the “knowledge spillover” effects of these patents in terms of R&D
productivity increases. A second avenue for foreign knowledge to flow through would
be the diffusion and adoption of the manufactured technologies or the codified
21
knowledge representing them. In this case, indicators of the commercial output of
technology, such as imports or licences, could proxy technology transfer.
The IPCC (2000) recommends the use, with caution, of several types of international
financial flows as indicators about the levels of international TT and how these levels
are changing over time. Among the main types of financial flows are: official
development assistance (ODA), loans at market rates, foreign direct investment (FDI),
commercial sales, foreign portfolio equity investment (FPEI) and venture capital.
However, these financial flows cannot be taken as an indication of the level of skills
and knowledge that have been transferred alongside equipment imports and capital
investments (Mallet et al., 2009).
Evidence-based studies of climate change technology transfer across countries have
used two main types of indicators to proxy technology transfers – output indicators
and effect indicators. Studies measuring the output of TT have used TT claims in CDM
projects that use foreign equipment or knowledge (Pueyo, 2007; Dechezleprêtre et al.,
2008; Doranova, 2009; Seres et al., 2009; Seres et al., 2010) or patenting data. TT can
be measured through patents data as the number of foreign patents filed in a country
and related to climate change technologies (Hascic and Johnstone, 2009;
Dechezleprêtre et al., 2010) or as the count of citations received by patents originating
in one country by patents originating in another country (Verdolini and Galeotti, 2011).
Foreign holders of patents gain the exclusive right to exploit commercially their
technology in the granting country. Hence, this measurement can be used as a proxy
for market-driven knowledge flows. Patent citations are instead used as proxies of
general knowledge diffusion at the international level.
The use of claims in CDM projects and of patent data as proxies for technology transfer
has some limitations because TT claims in CDM projects are subject to uncertainty, as
they are not based on a common definition of TT and have not gone through
verification unless they belong to the additionality test. Also, they refer to projects
often at the design stage, which may not be successfully implemented and hence may
not involve technology transfer. In addition, they present a narrow perspective of
technology as equipment or engineering knowledge and cannot be taken as an
indication of the levels of skills and knowledge that have been transferred alongside.
Some of the limitations of patent data, as highlighted by Dechezlepretre et al. (2010)
and Verdolini and Galeotti (2011), are that they do not reflect the whole extent of
innovation, as they are only one of several means of protection; the propensity to
patent differs highly between sectors and countries; certain forms of knowledge are
not patentable, such as learning by doing; the value of patented inventions is highly
heterogeneous; the amount of technology covered by a patent varies significantly
across countries and foreign companies could file patents abroad to block competitors
and to prevent other firms from invading their own markets, in which case they would
not reflect technology transfer, but the opposite. Some of these limitations can be
corrected in modelling exercises, for example by adjusting patent data per countries’
patent breadth or propensity to patent, but the main limitation of the use of patent
data and CDM claims as proxies of TT is that they do not reflect the transfer process
and its success in leading to knowledge spillovers.
Verdolini and Galeotti (2011) construct a model to quantify energy related knowledge
spillovers in which a country’s annual production of knowledge, measured as annual
22
patent counts, is explained by technology demand and supply determinants. Demand
determinants are local (electricity price, value-added and environmental policy of the
local economy) but supply determinants include the stocks of both internal and
external knowledge. External knowledge is proxied as the sum of the knowledge
produced abroad, measured as the stock of patents adjusted using the perpetual
inventory method, that has crossed the border of the recipient country i. This flow is
measured as the probability that a patent from country j is cited in recipient country i.
The resulting model indicates that knowledge spillovers are a powerful source of
inducement: a 10% increase from the average foreign knowledge stock value increases
by 9.6% the probability of domestic innovation. The impact of foreign knowledge
stocks in the local production of knowledge is much higher than the impact of demand
factors, and it is also higher than the own stock of knowledge for less innovative
countries and similar to the own stock for top innovating countries. This is a valuable
model showing the link between foreign knowledge flows and local knowledge
production; however, patents are only a narrow reflection of the wider TT process.
Besides, the data available only allows a sample of 17 developed countries, therefore
omitting the particularities of developing and emerging economies.
Models based on the effect of climate change TT measure the knowledge spillovers of
foreign activities as the resulting changes in relative10 CO2 emissions or energy
consumption (Blackman and Wu, 1999; Frankel and Rose, 2002; Mielnik and
Goldemberg, 2002; Cole and Elliott, 2003; Cole, 2006; Ang, 2009; Hubler and Keller,
2010), or cost reductions of low-carbon technologies through joint learning-by-doing11
(LBD) and learning-by-searching12 (LBS) (Qiu and Anadon, 2011). The main problem
with the first type of studies measuring the effects of TT on CO2 emissions is that it is
difficult to attribute a specific part of the emission reductions that is due to TT, as
there are many factors that have an impact on these emissions. In models based on
the effect of climate change TT, foreign flows of technology are proxied as
international economic flows, mostly trade or FDI or as “technology adoptions” (Qiu
and Anadon, 2011).
A technology adoption indicator was created by Qiu and Anadon (2011) to account for
the LBS effect. This indicator is defined as “the number of technologies being adopted
by Chinese turbine manufacturers plus the technologies transferred to foreign
subsidiaries in China from their parent companies each year”. Technologies are
considered as adopted from foreign suppliers if commercialised by licences, joint
design, the opening of foreign subsidiaries or the constitution of joint ventures. They
can also be adopted from local suppliers through licensing, transfer from domestic
research institutes or in-house R&D. In all, 67% of the 43 technology adoptions by wind
turbine manufacturers in China originated from foreign suppliers.
10
Per capita or per unit of output.
11
Learning by doing is the process by which the accumulation of experience in manufacturing and/or
project development leads to productivity increases and cost reductions. A learning by doing rate shows
the percentage of reduction in unit costs occurring with a doubling of cumulative experience, where the
latter is usually measured as installed capacity (Qiu and Anadon, 2011).
12
Learning by searching is a process that accounts for cost reductions stemming from improvements in
technology (Qiu and Anadon, 2011).
23
Qiu and Anadon’s (2011) model is particularly interesting because it devises, even if it
is not the aim of the paper, a potential new way of measuring knowledge spillovers
from foreign technology adoptions. The average normalised cost per unit of electricity
produced by Chinese wind projects is explained by the country’s joint learning rate,
which captures LBD and LBS; the scale of the wind parks, capturing the effect of
economies of scale; manufacturing localisation rates and two control variables: steel
price and wind resource quality. LBD and LBS are considered jointly due to their high
correlation coefficient of 0.9967, which causes collinearity in the regression.
Consequently, the model only includes the variable cumulative capacity measuring the
LBD effect to also represent the LBS effect. In any case, the paper cannot measure the
impact of foreign technologies because their technology adoption indicator considers
jointly foreign and local sources of technology. Some of the problems in using
technology adoption as an indicator are similar to those in patent data, as the count
nature of the technology adoption indicator cannot reflect the highly heterogeneous
value of the different “technology adoptions” and the differences in the amount of
technology covered by one “adoption”. However, the author argues that an interview
with an expert in the wind industry shows that it is difficult to weight technology
adoptions by their impact on the production process, as this influence very much
depends on the specific context of the projects13.
Studies using financial flows as proxies of foreign technology analyse the aggregate
effect of foreign activity instead of the specific effect of climate change-related
technologies, which is due mostly to the unavailability of historic data on climate
change-related financial flows for most countries. Aggregate FDI, trade or ODA data
are not useful per se for estimating the knowledge spillover effects of foreign
technologies because foreign economic flows actually encompass three different
effects on the carbon or energy intensity of the recipient country – scale, composition
and technique effects – where only the technique effect is related to TT14. As
described by Peterson (2008), the scale effect entails that more economic activity
resulting from foreign activity leads to higher energy use and GHG emissions, and it is
only relevant when carbon emissions or energy consumption are measured in
absolute terms. The composition effect implies that foreign inflows may change the
industrial structure of an economy and result in positive or negative consequences
depending on the country’s circumstances. For example, countries with lax
environmental regulation or high capital to labour ratios may attract pollution-
intensive activities, raising energy use and emissions. On the other hand, economic
development leads to structural shifts, first from agriculture to industry that leads to
higher emissions, and then from industrial to service activities, or from heavier to
lighter industry, which leads to emissions reductions. Therefore, the final sign of the
composition effect depends on specific country circumstances. The technique effect
conveys a direct and indirect effect on GHG emissions reductions. The direct effect
comes from increased openness to the diffusion of cleaner and efficient production
methods, which relates to TT. The income-induced indirect effect happens as
openness leads to increased wealth and a rising demand for stricter environmental
13 th
Writen communication with Laura Diaz Anadon, 10 March 2011.
14
A direct explanation of carbon or energy intensity by foreign flows would lead to spurious
correlations, as it happens in Mielnik and Goldemberg (2002).
24
regulations, which also encourages cleaner technologies, and corresponds with the
literature on the Kuznets environmental curve that predicts an inverted U relationship
between per capita income and pollution. Some authors believe that the translation of
wealth into higher environmental conscience and subsequent environmental
regulation happens more clearly in democratic societies (Frankel and Rose, 2002).
However, some studies contradict the Kuznets curve, showing that income per capita
does not have any significant impact on the reduction of CO2 emissions (Ang, 2009).
The few studies that have attempted the econometric study of knowledge spillovers in
energy consumption and carbon emissions through foreign financial flows have
therefore been concerned mostly with isolating the direct technique effect of foreign
activity from the indirect technique effect and the scale and composition effects. The
scale effect is usually neutralised by using variables relative to income levels or
population (i.e. carbon intensity or carbon emissions per capita) instead of gross
variables, whereas the composition effect is modelled through indicators of a
country’s economic structure, such as the capital/labour ratio or the share of industrial
value-added in GDP. Finally, the indirect technique effect can be modelled by
introducing income per capita variables.
Hubler and Keller (2010) built a model to explain changes through time in the energy
intensity of a set of countries by their foreign financial flows (FDI, imports and ODA),
income per capita, GDP share of industrial value-added and GDP share of total
investments. Income per capita would represent the income-induced technique effect
and GDP share of industrial value-added, the part of the composition effect related to
increased industrialisation. Foreign financial flows would account for the TT effect and
the part of the composition effect that relates to a shift in the production pattern
within the agricultural, industrial or service sectors – the authors’ caveat for their
inability to completely isolate TT in their empirical model. In fact, the model shows an
insignificant effect of FDI, imports, income per capita and GDP share of industrial
value-added. However, ODA flows are significant and negative, which would be in line
with the expectations that industrialised donor countries promote clean technologies
in developing countries. The investment share of GDP is significant and positive, which
indicates that rising investments lead to more energy-intensive processes. In any case,
the model does not find a robust energy-reducing effect of FDI and imports. The
authors interpret that the effects of FDI inflows on energy use depend on country-
specific characteristics and policies that cannot be captured at the macro level.
Besides, the use of aggregated FDI precludes one from identifying the particular effect
of low-carbon technology-related FDI and isolating it from the composition effect of
FDI related to energy-intensive production.
Cole and Elliott (2003) defined a model to explain the effect of trade on emissions per
capita and per unit of GDP of various pollutants, among which was CO2. Their model
captures scale, composition and technique effects through income, capital to labour
ratio and trade openness (the ratio of imports plus exports to GDP) variables. The
model with emissions per capita as the dependent variable confirms positive and
significant elasticity for scale and technique effects, the composition effect and a small,
weakly significant and positive direct trade effect. An analysis of the variability of
country-specific trade elasticities suggests that there is no obvious relationship
between the trade elasticity of emissions per capita and the level of income or capital-
25
labour ratio. Hence, the relationship between trade liberalisation and CO2 emissions
depends on country characteristics other than income and factor endowments. A
significant and negative coefficient of the time trend for emissions per capita indicates
that factors common to all countries which change over time, such as technological
advances, are reducing emissions.
Another attempt to model the effect of trade on CO2 emissions was provided by
Frankel and Rose (2002). Their model uses income per capita and its square, trade
openness, an index of democracy and population density, as explanatory variables of
per capita emissions for a number of pollutants including CO2. Their results show an
increase in CO2 emissions per capita for all income levels, contradicting the Kuznets’s
curve hypothesis, while trade openness does not show a clear significance.
In a more recent application to the case of China, income per capita does not have any
significant impact on CO2 emissions per capita but trade openness is found to have a
significant positive effect on emissions (Ang, 2009), showing the predominance of the
scale and composition effects of trade over the technique effect.
2.3.4 Technology transfer metrics in the framework of the UNFCCC

In 2009, the Expert Group on Technology Transfer (EGTT) presented a report to the
Subsidiary Body for Scientific and Technological Advice (SBSTA) and the Subsidiary
Body for Implementation (SBI) on “Performance indicators to monitor and evaluate the
effectiveness of the implementation of the technology transfer framework”(UNFCCC,
2009). The objective of the report was “to develop and test a balanced and robust set
of performance indicators that could be used [...] to monitor and evaluate the
effectiveness of the implementation of the framework for meaningful and effective
actions to enhance the implementation of Article 4, paragraph 5, of the Convention”.
Article 4, paragraph 5, of the Convention states that “Developed countries shall
support not only the transfer of low-carbon technologies, but also the development
and enhancement of endogenous capacities and technologies of developing country
parties”.
The report includes a total of forty indicators classified in six categories: technology
needs and needs assessments; technology information; enabling environments;
capacity building; mechanisms for technology transfer and finance. Table 52 in the
Annex of Chapter 2 shows the proposed indicators, as well as an assessment made by
the UNFCCC of their relevance and data availability and a classification made by the
author of the indicators per type as a technology input, output or effect15, as well as
per type of metric used, namely value or manifestation.
The analysis made in Table 52 shows that most indicators refer to manifestations of
UNFCCC input activities in the technology transfer process, with uncertain effects on
the actual transfer of technologies. No indicators about the effect of technology
transfer are included in the list. As analysed in the Introduction, UNFCCC activities have
in fact had a limited effect on the transfer of technologies to developing countries.
Therefore, input indicators are not very appropriate for assessing the contribution of
UNFCCC activities to generating knowledge spillovers in the recipient countries.
15
Those indicators that do not refer to technology itself but to some of its demand or supply
explanatory variables, such as “investment environment”, are classified as “explanatory”.
26
Nevertheless, some of the proposed indicators are still very valuable for future
economic analyses of technology transfer, mainly those included in the Finance
category about the value of global investment and financial flows in climate change
technologies. In fact, the lack of a specific indicator on the financial flows of climate
change-related technologies was one of the highlighted limitations of the econometric
studies of the effect of climate change technology transfer. However, in spite of the
favourable assessment of availability in the initial EGTT report, these data are not
readily available. A note by the SBSTA about the information required for using the
performance indicators recognises that data on global investment and financial flows
for climate change mitigation technologies are not systematically collected. Previous
UNFCCC reports had listed a number of estimates, based on disparate sources and
using different methods and assumptions, with a significant likelihood of double
counting between the different sources and at the same time a high probability that
some financial flows were unaccounted for. The biggest gap in the estimates is in
private financing for the deployment of technologies. These figures are possibly
underestimated because internal funding for energy efficiency by large firms is not
included in the estimates. The data for developing countries are also incomplete,
especially for deployment and diffusion, while agreed estimates of total annual
investment and financial flows in climate change technologies through the Kyoto
Protocol’s flexibility mechanisms are not available either. The OECD (Buchner et al.,
2011) and UNCTAD (2010) are currently working to improve data availability on
climate change-related FDI.
2.3.5 Summary and identification of gaps

The literature review establishes that there is not an overarching approach to
measuring TT but instead a wide range of frameworks and models. Table 2 summarises
these approaches and classifies the different indicators proposed by their type as
input, output or effect, and their metering approach as count or value. It also classifies
them per body of literature proposing them.
The review has also shown some gaps in the literature and several promising avenues
to contribute to climate change-related research on TT. Climate change literature lacks
solid studies about the output and the effect of TT, and those studies measuring the TT
effect as changes in relative CO2 emissions or energy consumption due to foreign
activities have faced the limitation of unavailability of financial information related to
low-carbon technologies. Their modelling results have been therefore “contaminated”
by the scale, composition and indirect technique effects of foreign activity, which have
precluded an assessment of the actual effect of technology transfer. The availability of
data about climate change-specific foreign financial flows could contribute highly to
the success of future modelling exercises. The UNFCCC proposes “global investment
and financial flows in climate change technologies” among its performance indicators,
which shows willingness to obtain the data. However, these data are still not
systematically collected and the UNFCCC has not provided any methodology to start
doing so, although UNCTAD and the OECD are currently working to improve data
availability on climate change-related FDI.
Even though it is not the aim of their paper, Qiu and Anadon (2011) devise an
interesting approach to measuring the effect of foreign climate change technology
27
adoptions as their impact on electricity cost reductions. The referred paper does not
perform this analysis, as it considers the joint impact of foreign and local technology
adoptions in the cost of electricity production. The analysis of this learning by
searching effect on a country-by-country and technology-by-technology basis could be
useful in analysing the factors that make some countries more successful in translating
access to foreign technology into cost reductions, controlling for other variables. Some
limitations of this analysis come, again, from data unavailability. Data on the
“technology adoption” indicator used by the authors was provided by the Wind Energy
Equipment sub-association of the Chinese Agricultural Machinery Industry Association
(CAMIA), but is most likely not available in most developing countries. Furthermore,
electricity generation costs required for their analysis were obtained from the Chinese
wind power government concession programme, through personal communications
with the chief wind energy engineer at China Huaneng Group and through a wind
power report by Li et al. (2008). Obtaining such information for other developing
countries could prove complicated.
Other effect indicators that could be worth pursuing relate to the linking effects of the
adopted technologies. Local learning processes are essential in TT activities, and they
are broader when foreign knowledge is linked into the wider structure of the host
country’s economy through backward and forward inter-firm linkages.
Effect measurements present a number of problems for informing policymaking, as it
is difficult to isolate the part of an effect like CO2 emission reductions or cost
reductions that are due to technology transfer. The effect is shared across a number of
factors, and econometric studies are often unable to attribute any one part that is
caused by the adoption of foreign technologies. An additional problem relates to the
importance of specific host country characteristics and policies used to translate
foreign technologies into actual emission reductions. Very often, such specificities
cannot be captured by macro-level models. Case study approaches therefore play an
important role in informing policymaking in the area of technology transfer.
In relation to the output indicators of climate change technology transfer, current
studies use climate change-related patent data, CDM TT claims and financial flows
related to climate change technologies, mainly imports and FDI. Section 2.2, which
discusses TT channels, shows that there are many other ways in which TT can happen,
which are not accounted for by these indicators. These ways include licences for
foreign technology, the acquisition of foreign companies, setting up R&D activities
overseas, collaborative R&D programmes, hiring foreign staff or training locals abroad.
Additional indicators that the field of climate change could borrow from the study of
vertical technology transfers relate to the number, value or revenues generated by
new companies or business divisions to exploit a foreign technology. Linking output to
input indicators can show the effectiveness of different countries in adopting foreign
technologies.
Input indicators are the main proposal of the UNFCCC. Their main advantages are their
ease of measurement and their availability against short timeframes. However, on
their own, input indicators cannot be used to assess the success or failure of
technology policies, hence the need to link them to output or effect indicators.
28
TABLE 2- SUMMARY OF REVIEWED APPROACHES FOR MEASURING TECHNOLOGY TRANSFER
Type of Measure- Body of literature

proxy ment
approach Economics of technology Vertical technology Climate change technology transfer UNFCCC technology performance
diffusion transfer indicators
Input Value Foreign R&D expenditure Value of collaborative International financial flows: FDI, imports, ODA Financial resources provided for
weighted by the bilateral research agreements technology needs assessments
import-share
Financial resources provided for
International financial flows: capacity building programmes
FDI, imports, ODA
Value of joint public-private climate
change R&D programmes
International financial flows (FDI,
imports, ODA...) related to climate
change technologies
Count Research agreements Number of foreign patents filed related to climate Several indicators of the number of
with private change technologies technology-related activities
companies performed by the UNFCCC, such as
Adoptions of foreign technologies
reports written or capacity building
programmes organised
Output Value Licence revenues

Profits from inventions
Count Patents filed by foreign Scientific publications Local patent production explained by the stock of
actors external knowledge.
Patents applied and/or
Foreign patent cites by local granted CDM projects that incorporate foreign
patents technologies or foreign knowledge
Licences issued
Citations received by patents originating in a
29
Type of Measure- Body of literature
proxy ment
approach Economics of technology Vertical technology Climate change technology transfer UNFCCC technology performance
diffusion transfer indicators
Spin-offs established country by patents originating in another country,
related to climate change technologies
Invention disclosures
evaluated
Effect Value Local TFP growth explained CO2 emissions or energy intensities explained per
by foreign R&D, openness to international financial flows (FDI, imports, ODA)
trade and control variables and control variables
Local TFP growth explained Technology cost reductions explained by local and
by FDI and import flows and foreign technology adoption and control variables
control variables
30
2.4 Enabling frameworks for climate change technology transfer
The reviewed literature and discussions between the public and private sectors
(WBCSD, 2010) often conclude that enabling frameworks are needed to accelerate
climate change technology diffusion to developing countries. Five types of factors have
been identified in the literature as creating these enabling frameworks for technology
transfer:
• Economic and institutional frameworks
• Technology demand
• Technology supply
• Industry development
• Bilateral factors promoting cooperation between technology supplier and
recipient
These enabling frameworks can operate at different levels of the technology transfer
process defined in the previous chapter (inputs, channels, outputs, effects). For
example, they can lower the transaction costs of TT channels such as imports, foreign
direct investment or hiring foreign staff, and can also enable foreign technologies to be
used efficiently and absorbed locally. Enabling factors are also required for
technological inputs to become commercial outputs and to facilitate knowledge
spillovers through backward and forward links with local companies.
A government can affect these factors through appropriate policies. For example,
technology demand and supply factors can be enhanced through the so-called
demand-pull or technology-push policies, both of which are crucial to understanding
the innovation and diffusion processes and the role that technology transfer plays
therein (Mowery and Rosenberg, 1979). Factors promoting investment can be
enhanced through sound macroeconomic policies and solid institutions, while factors
promoting cooperation between countries can be enhanced, for example, through
trade openness, regional integration or the government brokerage of TT processes.
Industry-specific factors preventing the emergence of local technologies have been
addressed through industrial policy. The following subsections provide a more detailed
account of these enabling frameworks.
2.4.1 Economic and institutional frameworks promoting private investment

The private sector owns most of the technologies required for a shift to low-carbon
paths and is responsible for most international TT through mostly trade and FDI (Stern,
2007; Brewer, 2009; WB, 2009). Ease of doing business indicators can show the
general state of business regulation and have been used in models to explain levels of
international cooperation in CDM projects (Dinar et al., 2008). Institutional and
economic frameworks that have been proved to enable private investment, leading to
transfers of climate change technologies, include:
• Stable policies, transparent investment regulation and conductive local
conditions (WBCSD, 2010). For example, uncoordinated energy policies, the
uncertainty of institutional frameworks, excessive bureaucracy or
inappropriate power purchase agreements (PPA) are behind some failures to
transfer renewable energy technologies to Ghana and South Africa (Gboney,
31
2009; Grant, 2009). The level of corruption also has a negative impact on
foreign technology suppliers.
• Protection of intellectual property rights (IPRs) (ACTI, 2009). Strict IPRs have a
positive influence on a climate change technology transfer model based on
patent data (Dechezleprêtre, Glachant et al., 2010), and they are not found to
impede the emergence of local renewable energy industries in developing
countries (Barton, 2007). For example, the Chinese solar PV and wind industries
have been able to thrive, even though most technology has been patented by
Western countries (Lewis, 2007; de la Tour et al., 2010; Lema and Lema, 2011).
In fact, Chinese measures to strengthen IPR laws and implementation have
been found to catalyse technology adoption through joint design with foreign
companies (Qiu and Anadon, 2010).
• Open trade policies that allow engaging in international economic activities
that can channel TT, mostly imports and FDI. For example, the success of Chile
in attracting CDM projects is attributed to its export-oriented, open and
liberalised economy, political stability and business-friendly politics (Rindefjäll
et al., 2010). Dechezleprêtre et al. (2010) find a significant negative impact of
high tariffs when modelling the determinants of climate change TT using
patent data.
2.4.2 Technology demand factors and policies

The transfer of foreign technologies is influenced heavily by the size and growth of
demand in the recipient country. The success of China, India and Brazil in attracting
and deploying foreign climate change technologies, and growing domestic industries,
seems to confirm this point. A large market allows technology businesses to build a
significant production scale and achieve lower production costs as a result of
economies of scale and technological learning curves (Wei, 1995; Stern, 2007). It also
provides scope to develop a wide portfolio of low-carbon technologies. Demand
factors explain incremental technological change better than discontinuous change
(Mowery and Rosenberg, 1979), so they can play an important role in steering
transfers of low-carbon technologies at the commercial stage.
Some indicators of technology demand that have shown high and statistically
significant coefficients in several econometric models of climate change technology
transfer are income per capita, population (Dechezleprêtre et al., 2010; Doranova,
2009), GDP growth (Dechezleprêtre et al., 2008), electricity consumption (Hascic and
Johnstone, 2009) and electricity prices, signalling the expected cost of fossil fuel-based
technologies (Verdolini and Galeotti, 2011). In a previous study, not directly concerned
with international technology transfer, Popp (2002) analysed the influence of both
energy prices and the stock of knowledge available to scientists on innovative activity
related to energy efficiency, by measuring innovative activity through U.S. patent data
from 1970 to 1994. His most important result was the strong, positive impact of
energy prices on new innovations, but the results also showed that prices on their own
were not enough to spur technological change. The supply of ideas also played an
important role in shaping the direction of innovation.
Experience of climate change technology adoption in China, India and Brazil has shown
that, in addition to their large market sizes, successful technology transfers have
32
required government intervention to steer the demand for renewable energy
technologies, which are often more expensive than conventional fossil fuel solutions.
So-called “demand-pull” policies affect the size of the market for a new technology by
raising the payoffs of innovation and deployment (Nemet, 2009). Some examples of
demand-pull policies in the field of climate change include carbon markets, tax credits
and rebates for consumers of new technologies, technology mandates, energy
efficiency standards, feed-in tariffs, renewable energy portfolios, taxes on competing
technologies and government procurement (Nemet, 2009; Coninck et al., 2008). The
rationale of government intervention is the expectation of cost reductions through a
variety of learning processes as the installed capacity increases (Grubb, 2004).
The effect of demand-pull policies has been assessed in some of the reviewed
econometric literature. Some examples of policy variables are dummies for the
existence of at least one policy targeting energy efficiency (Verdolini and Galeotti,
2011); for the ratification of the Kyoto Protocol (Dechezlepretre et al., 2011) or for the
adoption of the Environmental Protection Law in China (Ang, 2009); year dummies to
capture changing policy environments, bidding terms or any other exogenous technical
change in China (Qiu and Anadon, 2011) and a variable showing the degree of
involvement in CDM activities, measured as CERs obtained between pairs of
technology providers and recipients (Hascic and Johnstone, 2009). The variables are
positive and significant in some of the regressions, but they are unsophisticated and
cannot capture the diversity of policies existing in different countries and their impact
on technology demand.
In China, several authors agree that the radical development of climate change
technologies, with a particular emphasis on wind, has been driven by demand-pull
policies such as the Renewable Energy Law, approved in 2005; Renewable Energy
Pricing and Cost Sharing Management and Trial Methods, issued in 2006; the Mid-and
Long-term Development Plan for Renewable Energy, issued in 2007; wind energy feed-
in tariffs effective from August 2009, with varying tariff levels per region depending on
wind resource class; and wind concession programmes, where wind farm developers
compete in a bidding process through proposals to develop wind farms. The winners of
the bid are awarded PPAs at the fixed tariff they included in their bid, guaranteed
interconnection, financial support for grid extension and access roads and preferential
tax and loan conditions (Lewis, 2007; Wang, 2010; Zhang et al,. 2009; Qiu and Anadon,
2011; Lema and Lema, 2011; Lewis, 2011).
The development of a leading solar PV industry in China follows a different pattern, in
that it was not induced by national but international demand-pull policies. China is the
largest solar PV cell producer in the world but accounts for a very small share of global
PV demand (de la Tour et al., 2010). The growth of the Chinese PV industry has been
driven by incentive policies implemented in a limited number of developed countries,
mainly Germany, Japan, Spain and the US. Feed-in-tariffs have been the main support
mechanism for solar PV in these countries, and generous feed-in-tariffs, particularly in
Spain, have led to a market boom that spurred Chinese development in the most
energy- and labour-intensive segments of the industry, taking advantage of its low
labour and energy costs (de la Tour et al., 2010). Technology transfer was essential for
entry into the solar PV market but it was possible through the purchase of equipment
and hiring staff trained abroad.
33
The Indian Government has a stated target for renewable energy to contribute 10% of
total power generation capacity by 2012 (Lewis, 2011). Early support for wind
technologies included the National Guidelines for Clearance of Wind Power Projects, in
1995, which stipulated that all state electricity boards should take measures to ensure
grid compatibility. The Electricity Act of 2003 required all state-level energy regulatory
commissions to encourage renewable energy generation quotas among electricity
distributors. Subsequently, many states set up renewable energy targets and policy
support mechanisms, including feed-in tariffs, guaranteeing long-term contracts at
declining subsidised prices (Lewis, 2007; Lema and Lema, 2011).
South Korea, home to several wind turbine technology developers and one of the
world’s largest innovators in clean energy technologies, set a target of 5% renewable
energy share in primary consumption by 2011. South Korean incentive programmes to
achieve these targets include feed-in tariffs guaranteed per 15 years, tax incentives
and subsidies for the local wind market and demonstration projects (Lewis, 2011).
The development of a competitive ethanol industry in Brazil was spurred by the
Brazilian Alcohol Programme (PROACOOL), established in 1975 with the purpose of
reducing oil imports by producing ethanol from sugarcane. The government
guaranteed a market and subsidised the price of ethanol to make it competitive
against fossil fuels, which were also subsidised. Subsidies given to ethanol production
were eliminated in 1997, as its efficiency and cost competitiveness evolved and fuel
prices were liberalised (Goldemberg et al., 2004; Ribeiro y de Abreu, 2008).
Studies showing demand-pull policies in developing countries beyond China, India and
Brazil are very limited, but they include the case of South Africa (Grant, 2009), a highly
carbon-intensive country with low renewable energies penetration. In 2003, South
Africa released a ‘White Paper on Renewable Energy’, outlining the government’s
objective of achieving 10,000 GWh of renewable energy contribution to final energy
consumption by 2013, or approximately 3% of final electricity demand. The
government also released its official ‘Energy Efficiency Strategy’ in 2005, setting a
target of a 12% average reduction in final energy demand by 2015. The aspirational
character of these policies, and the low electricity tariffs, prevented the creation of a
significant market for renewable energies. New policy developments in South Africa
include new renewable energy feed-in-tariff guidelines. Some recommendations to
accelerate deployment are the implementation of feed-in-tariffs jointly with robust
and transparent tender processes, which would need to be managed by Eskom, the
sole buyer of electricity in the country (Grant, 2009).
A study about Ghana (Gboney, 2009) highlighted the key domestic policies and
institutional set up related to renewable energies. Ghana set up a national target of
10% RE in the country’s energy mix by 2020 and sought to enhance energy efficiency
through demand-side management. However, the country has been unsuccessful so
far in achieving the significant penetration of renewable energies. From a purely
demand-side perspective, some barriers come from the absence of transparent
network pricing networks, well-documented tariff principles or rural electrification
policies.
Several studies point to the limited technology demand-pull effect of international
carbon markets (through the CDM) for developing countries. Evidence has shown that
34
successful TT through the CDM has often been based on pre-existing enabling
frameworks in recipient countries. For example, Hansen (2008) observed that transfers
of biomass and biogas technologies to Malaysia through the CDM were done by
foreign companies that already had a strong presence in the Malaysian market. Several
studies of TT in Chinese CDM projects show that certified emissions reductions (CERs)
income was rarely the primary reason why the projects were developed, due to the
uncertainty of carbon income and long CDM registration time lags (Wang, 2010; Lewis
2010). A study based in India and Brazil demonstrated that managers considering
investments in the CDM tended to favour projects that were non-additional or only
marginally additional, due to the uncertain future revenue streams of the carbon
market and the certain immediate costs (Hultman et al., 2010).
2.4.3 Technology supply factors and policies

Technology supply factors refer to the available stock of knowledge in recipient
countries. They determine the absorptive capacity of recipient organisations to
efficiently operate the transferred technologies and to exploit opportunities emerging
from the state-of-the art elsewhere. General research on technological development
has shown that local industries with a strong technological tradition are able to catch
up with technologically superior, foreign-owned companies (Cantwell, 2009).
Modelling exercises have pointed at the importance of internal technological
capabilities for developing countries. Controlling for country size, the smaller, poorer
and further from the technological frontier (measured as distanced to the US TFP) a
country is, the more intensively it relies on domestic technology sources to increase
productivity (Keller, 2004) or to reduce CO2 emissions per capita (Ang, 2009). This is
because there is no automatic learning, and internal capabilities increase the ability to
assimilate foreign technologies. The impact of the assimilation of foreign technologies
in terms of CO2 emission reductions is higher the further away a country sits from the
technological frontier due to diminishing returns of foreign technology as countries get
closer to the technological leaders. TT models through the CDM which do not control
for country size show that smaller and less capable countries are more likely to use
foreign technologies in their projects (Doranova, 2009; Seres, 2010) due to the
difference in the size of their R&D investments and production capacities. The impact
of technological capabilities on the likelihood of TT in CDM projects also depends on
the type of technology used by the projects. In the energy and chemicals sectors, high
capabilities increase the likelihood of TT, while in more technologically mature sectors
like agriculture and waste management, high capabilities reduce the likelihood,
meaning that the technologies would already be available locally (Dechezlepretre et
al., 2008;)
In the reviewed econometric literature, technology supply factors have been
accounted for through indicators of the quantity and quality of human capital, local
R&D spending and the stock and quality of local patents. Although not included in the
reviewed modelling exercises, several technology indexes can provide more
comprehensive measurements of a country’s technological development by using a
variety of indicators. Each of these indexes emphasises some aspects of technological
development over other aspects. For example, the index of innovation capability,
published by the United Nations Conference on Trade and Development (UNCTAD
35
2005), emphasises education levels, numbers of scientists and engineers and
expenditures on research and development (R&D) or R&D personnel. The Technology
Achievement Index (TAI), published by the United Nations Development Programme
(UNDP, 2001), also incorporates information on the diffusion of technologies and on
indicators of innovation such as the number of patents granted. The ArCo indicator of
technological capabilities considers three main components: the creation of
technology, technological infrastructures and the development of human skills
(Archibugi and Coco, 2004), while the World Bank’s technology index considers three
dimensions of technological achievement: the extent of scientific innovation and
invention, the diffusion of older technologies and the diffusion of newer technologies
(WB, 2008).
Dechezlepretre et al. (2010) differentiated between general and specialised knowledge
stocks to explain their impact on climate change TT measured through patent data.
They found that the impact of the knowledge stock is significant and positive when
measured as specialised knowledge through the discounted stock of previously filed
patents by local inventors in the transferred technology. However, general knowledge,
measured as a tertiary gross enrolment ratio16, does not have a significant impact.
Models show that both technology demand and supply factors are necessary for
spurring innovation. Supply factors have been found to explain better discontinuous
technological change (Mowery and Rosenberg, 1979), while demand factors explain
better incremental change. A mix of both factors enables market opportunities to
connect with opportunities to exploit them locally, improving the chances of successful
technology transfer.
Governments can intervene to increase the level of local technology supply through
so-called “technology-push” policies, the most common examples of which are
government-sponsored R&D, tax credits for companies that invest in R&D, support for
education and training, infrastructure development and funding demonstration
projects (Nemet, 2009). Positive knowledge spillover externalities provide the rationale
for government intervention.
Technology-push policies have been essential for China, India, South Korea and Brazil
to create technological change through foreign climate change technologies. The
Chinese Ministry of Science and Technology (MOST) has subsidised renewable energy
R&D expenditures through the Renewable Energy Fund, created in 1996. The national
basic research programme (973 programme), the national high-tech R&D programme
(863 programme) and the national key technology R&D programme have set research
priorities and indicated allocations of funds in China. The wind industry particularly has
highly benefitted from these programmes. Large-sized wind turbines, variable speed
technologies, variable pitch distance, offshore turbines and wind farm design have
been the main priorities in wind technology research. Domestic research institutes
have received significant subsidies to undertake these developments, which has
enabled national champions like Goldwind and Sinovel to compete with international
providers of large-sized turbines (Lewis, 2007; Zhang et al., 2009; Qiu and Anadon,
2011). Indian support for the technological development of the wind industry started
in the mid-1980s with the establishment of demonstration projects in the states of
16
Number of students in tertiary education.
36
Gujarat and Maharashtra (Mallet et al., 2009). Since then, the government has
contributed to improve the quality of domestic turbine manufacture through a
national certification programme for wind turbines, based on international testing and
certification standards, administered by the Ministry of New and Renewable Energy
and with technological support from the Danish Risoe National Energy Laboratory
(Lewis, 2007). South Korea has provided large amounts of funding for renewable
energy development through its January 2009 ‘Green New Deal’ stimulus package and
has set a ‘Comprehensive R&D Plan on Green Technology’ to define priority areas by
2012 (Lewis, 2011).
Studies of countries that have not achieved a significant penetration of climate change
technologies point at weak domestic capacities as one of the most important barriers.
For example, in Ghana the lack of capacity in financial institutions with regard to
renewable energy and energy efficient technologies, infrastructure project finance, risk
assessment and power purchase agreements was behind the difficulties in accessing
credit for these projects. Another barrier in the same country was the weak technical
capacity of private entrepreneurs to install and maintain low-carbon technologies
(Gboney, 2009). Experiences in South Africa also show the need to build institutional
and technical capacities to develop a robust policy framework that supports the
transition to a low-carbon economy, and to absorb and adapt technologies to the local
context (Grant, 2009).
2.4.4 Industrial development factors

Development typically involves moving from low-productivity to high-productivity
activities. Industrialisation contributes to this upgrade by raising output per worker
through specialisation, economies of scale and the application of increasingly
sophisticated technologies (Akyüz, 2009; The Economist, 2011). Access to foreign
technology is a prerequisite for moving from one stage to the next, and knowledge
spillovers from these transfers are greater when more advanced industries create
backward linkages with a solid base of lower value-added suppliers (Di Maio, 2009).
Industrial development can be measured through several indicators. Productivity
indicators show the efficiency with which inputs of capital and/or labour are turned
into gross output or value-added (OECD, 2001), while value-added itself is usually
measured as the selling price of a product minus the cost of all the materials that went
into making it (The Economist, 2004). The best comparable figures of total factor
productivity are provided by the OECD (2001) and labour and capital productivity, as
well as multifactor productivity measures (MFP), either in the form of capital-labour
MFP, based on a value-added concept of output, or in the form of capital-labour-
energy-materials MFP (KLEMS), based on a concept of gross output. Among those
measures, value-added-based labour productivity is the single most frequently
computed productivity statistic, followed by capital-labour MFP and KLEMS MFP. Table
3 shows the main productivity measures as defined in the OECD Manual (2001)
37
TABLE 3- OVERVIEW OF PRODUCTIVITY
PRODUCTI MEASURES
Source: OECD, 2001

TFP increases when workers can use more efficient equipment for product
transformation, and these
hese increases in turn enable rises in real income and living
standards.
A more sophisticated indicator of industrial development is the Competitive Industrial
Performance Index (CIPI), published by the United Nations Industrial Development
Organisation (UNIDO),, which combines
ombines data about manufacturing value-added,
manufactured exports and the share of medium and high-tech high value
value-added and
exports to show the strength of different countries’ industrial sectors
sector (UNIDO, 2009).
Without a solid industrial base, many developing countries face high barriers of entry
to create their own indigenous low-carbon technologies or to develop backward and
forward linkages with foreign investors. Countries at the beginning of their
development processes predominantly buy from abroad, rather rather than produce locally
(Amsden, 2001). Barriers of entry are higher in industries characterised by the
existence of market power,
power identified by a high level of profit rates of incumbents and
high market concentration (Cabral, 2000). Companies with market power can protect
their position through highly integrated TT, where knowledge remains internally.
There are several successful cases of emerging countries creating their own renewable
industries and achieving
chieving international competitiveness, such as ethanol in Brazil, wind
and solar power in China, wind power in India or biomass plants in Malaysia. The
market power of incumbents has also had an impact on these developments. For
example, Chinese PV technology
technology producers “are mostly active in downstream
segments of the PV production chain – cell production and module assembling – where
barriers to entry are low, competition is tough and profit margins are thin”, while
Western companies lead the upstream, more more lucrative segments (de la Tour et al.,
al.
2010). Chinese companies,, in this case, have taken advantage of their low labour and
energy costs to penetrate this competitive market. In the wind industry, leading wind
turbine manufacturers have been understandably reluctant to licence proprietary
information to companies in developing countries that could subsequently become
competitors. Chinese and Indian turbine manufacturers have obtained
obtained technology
from small wind power companies with low market power and little to lose in terms of
international competition (Lewis, 2007). Country leaders Suzlon and Goldwind have
also purchased majority control of European wind turbine and components suppliers
38
to overcome barriers to accessing foreign technologies (Lewis, 2007; Lema and Lema,
2011).
Government strategies can shape the form in which technology transfer happens to
maximise possibilities of developing domestic technological capabilities. In their
catching-up processes, Chinese and Indian wind industries have required government
interventions in trade and industry to overcome barriers of entry (Lewis, 2007; Wang,
2010; Zhang et al., 2009). China supports local producers and promotes technology
transfer through differentiated fiscal and tariff regulations on foreign technology and
local content requirements. Under the differentiated tariff regulations, only advanced
equipment without domestic substitutes is eligible for exemption from import tariffs
and value-added tax returns, so importers of complete technology systems are subject
to tariff levels of around 10% (Wang, 2010; Zhang et al. 2009). Local content
requirements are used as part of the Chinese Wind Concession Programme to
strengthen the capacity of local wind turbine manufacturers and to encourage foreign
technology providers to localise production of their advanced technologies.
Requirements began mandating 50% local content in 2003, which was increased to
70% in 2004, with any percentage above the minimum mandatory increasing the score
of the bidder. The policy seems to have been very successful, and it has increased the
participation of domestic and joint venture manufacturers of turbine installations from
25% in 2004 to 76% in 2008 (Zhang et al., 2009). This has caused many foreign-owned
companies to shift their manufacturing to China (Lewis, 2007) and enabled Chinese-
owned companies to obtain IPRs through their cooperation with foreign producers
(Wang, 2010). The localisation policy has also reduced the costs of wind turbines in the
country (Zhang et al., 2009). The doubling of localisation rates is associated with
reductions in wind electricity costs ranging from 19.7% to 20.6% (Qiu and Anadon,
2011). Cost reductions are caused by lower labour, material and transportation costs
of Chinese components. In contrast, India has been more hands-off than China in
promoting the use of domestically manufactured wind turbines, but it has also set
customs and excise duties that favour importing wind turbine components over
complete machines (Lewis, 2007; Lewis, 2011).
After the failure of orthodox trade liberalisation policies dominant among international
policymakers in the last part of the twentieth century, there has been a renewed view
of industrial policies as “intrinsic fundamental ingredients of all development
processes” (Cimoli et al., 2009). Industrial policies are indeed present in every
experience of successful industrialisation, from developed countries like Germany and
the USA to “latecomers” like South Korea, Taiwan, Brazil, China and India. However,
not all countries can afford to implement the same industrial policies that have fuelled
successful renewable energy industries in China and India, as these two countries are
different to most other developing nations due to their high bargaining power in the
global trade sphere, their large markets that make domestic industries viable and can
induce foreign investors to accept local content policies and their significant industrial
base that enables backward linkages with MNC. Some empirical evidence about local
content and related policies in other technologies and countries suggests that they are
costly to the economy, that often they do not achieve the desired backward and
forward linkages, they encourage inefficient foreign entry and can create potential
problems for future liberalisation (Saggi, 2004). Besides, local content requirements
39
are forbidden at the multilateral level by the WTO Trade-Related Investment Measures
(TRIMs) Agreement, which considers them trade distorting (UNCTAD, 2001). Further
research about the kinds of policies that can trigger local low-carbon technology
development outside BRIC countries could be highly valuable for informing policy.
2.4.5 Bilateral country characteristics

Bilateral technology transfer models using climate change-related patent flows
between country dyads have found that the probability that one country learns an
idea originating from another country (country i citing a patent from country j), or that
one country decides to file a patent in another country, depends on a series of
bilateral characteristics (Dechezleprêtre et al., 2010; Verdolini and Galeotti, 2011).
These bilateral factors influence the cost of the TT transaction and the diffusion of
ideas between countries. Bilateral characteristics have already been used to explain
joint investment in CDM projects by pairs of countries (Dinar et al., 2008).
Annual bilateral trade between countries was found to be the most significant variable
in Dinar et al.’s (2008) model of cooperation in CDM projects between pairs of
countries. The model showed that a 1% change in the dyad’s trade importance,
expressing total trade between a country pair as a fraction of the sum of their GDPs,
increases the likelihood of CDM cooperation incidence by close to 5%. However, the
model does not explicitly refer to technology transfer, but to non-Annex I countries’
investment decisions in CDM projects, which may or may not lead to TT.
Studies using international patent flows use some common bilateral variables to
explain the incidence of TT, including the geographical distance between country pairs,
belonging to the same trade area and having the same official language
(Dechezleprêtre et al., 2010; Verdolini and Galeotti, 2011). The most recent study also
incorporates two bilateral variables showing the technological similarities between the
country pairs: a technological index about the characteristics of the patents portfolio,
showing patent shares by technological fields, and a measure of distance in the
technological development of the country pairs, using information on average forward
citations received by the dyad’s patents. The first study incorporates 76 developed and
developing countries in the analysis, while the second refers only to 17 developed
countries.
Results in Dechezlepretre et al. (2010) differ according to technology type. Being part
of the same trade block significantly increases patent flows in seven out of twelve
technology fields, while the geographic distance between countries significantly
decreases the flow of patents in the regressions for eight technology fields. A common
language has a significant positive impact in ten regressions.
Results in Verdolini and Galeotti (2011) indicate that technological distance is more
important than geographic distance in that the latter is not significant, but completely
different patenting patterns drop the probability of inter-country citations to 18.3%.
Citation between countries speaking different languages is 82% as likely as citations
between countries with the same language, while countries that belong to different
trade blocks are 71% as likely to cite patents as those that belong to different trade
blocks.
40
Government policies can contribute to bilateral cooperation for technology transfer.
For example, the Chinese government served as a TT broker and initiator in order to
promote the transfer of Japanese cement waste heat recovery systems, solving the
asymmetrical information disadvantages of the local companies (Wang, 2010).
Regional integration could also increase opportunities to interact with technology for
small countries. The WBCSD (2010) recommends that those with similar needs around
a specific technology type group together to present larger opportunities for
technology providers.
2.4.6 Summary and identification of gaps

The review of the literature on enabling frameworks shows that a wide range of
country and country-pair characteristics and policies have a role in the extent of TT and
its impact on achieving internal technological change.
Some of the identified factors have opposite effects depending on whether their final
aim is to increase the penetration of low-carbon technologies or to promote the
creation of a domestic industry. For example, trade openness is considered to have a
positive effect in TT, as it eases access to foreign technologies. However, differentiated
tariffs for targeted technologies have had a positive effect in the creation of Chinese
and Indian renewable energy industries. Furthermore, high energy prices create
demand for low-carbon technologies, but low energy costs have been behind the
success of local Chinese PV manufacturers in achieving international competitiveness.
Table 4 summarises the identified enabling factors, showing indicators for their
measurement and their expected effect in promoting technology transfer.
TABLE 4- SUMMARY OF ENABLING FACTORS FOR CLIMATE CHANGE TECHNOLOGY TRANSFER
Type of factor Enabling factor Indicator Expected

effect
Economic and Ease of doing Ease of doing business index Positive

institutional business
frameworks
Stable, transparent Qualitative Positive
and coordinated
policies
Protection of IPRs Index of IPRs strictness Positive
Open trade • Average tariffs Positive

• Trade openness indicators (X+M/GDP)
Low corruption Corruption index Negative
Technology Demand size • Population Positive

demand • GDP
• Energy consumption
Demand growth • Annual GDP growth Positive

• Annual energy consumption growth
Electricity prices Electricity prices Positive
41
effect
Demand pull policies • Qualitative or dummies for: Positive

• Feed-in-tariffs
• Renewable energy targets
• Carbon markets
• Tax credits and rebates
• Technology mandates
• Energy efficiency standard
• Taxes on competing technologies
• Government procurement
Technology Human capital • Enrolment in tertiary education Positive

supply • % of degrees in engineering and science
• Annual number of graduates in science and
engineering
• Literacy rates
R&D • General R&D spending Positive

• R&D spending in energy-related
technologies
Patent stock • Stock of patents of specific technology fields Positive

• Quality of the stock of patents, measured
through citation levels
Technology • UNCTAD index of innovation capability Positive

development indexes • UNDP technology achievement index
• ArCo indicator of technological capabilities
• World Bank technology index
Technology push • Government-sponsored R&D Positive

policies • Tax credits for R&D investments
• Education and training expenses
• Infrastructure development expenses
• Funding for demonstration projects
• National certification programmes
Industrial Industrial • Competitive industrial performance index Positive

development development (CIPI)
• Total factor productivity (TFP)
Market power in • Profit levels of incumbent companies Negative

related industries • Market concentration
Cost of factors • Labour cost Negative

• Energy cost
Industrial policy • Local content requirements Positive in

• Differentiated tariffs for targeted BRICs,
technologies Uncertain
• Differentiated fiscal treatment for targeted in smaller
technologies economies
42
effect
Bilateral Trade relationships • Trade importance (total dyad trade/sum of Positive

their GDPs)
• Trade dependence (total dyad trade/sum of
their trade with other parties)
• Dummy belonging to the same trade area
Geographical • Km between country pairs Negative

distance
Language • Dummy having the same language Positive
Technological • Index about the characteristics of the Negative

distance patents portfolio, showing patent shares per
technological fields
• Difference between average forward
citations received by the patents of the dyad
• Difference between the countries TFP
2.5 Conclusions
The literature review has provided a number of answers to the initial research
questions, but also has highlighted some remaining gaps. This section summarises both
answers and gaps.
2.5.1 What constitutes climate change technology transfer, and how does it happen?
The literature reviewed shows a diversity of perspectives used to define TT, but it
seems clear that it is understood as going beyond the mere transmission of foreign
equipment, as it also encompasses the required knowledge to operate equipment
efficiently, the capacity to absorb knowledge created abroad and to exploit it to bring
about technological change.
The reviewed literature does not however show the perspective of actors involved in
real climate change technology transfer activities. Section 4 of this thesis will present
an example of this viewpoint by analysing the role that TT has played in several low-
carbon projects undertaken in Chile. The aim of this chapter is to contrast the
theoretical framework with the actual TT process, the underlying motivations of
recipient and supplier companies and the channels through which knowledge flows
and the barriers it encounters.
2.5.2 Is it possible to measure technology transfer, and what are the main challenges in
doing so?
The literature has clarified three ways to measure technology transfer: through inputs
into its process, outputs and effects. It has also shown two quantification approaches
as the number of manifestations of TT activities or as their value.
The different measurements used by the literature do not show the diversity of
channels through which knowledge flows and must restrict themselves to narrow
perspectives of technology, mainly due to data unavailability. Data is even scarcer
43
when it relates to developing countries. Very few models measure climate change TT
to developing countries, and when they do so they use very narrow measurements,
mainly claims in CDM projects and international flows of patents. Chapter 5 will show
the difficulties in obtaining data for the compiled TT indicators in a wide sample of
developing countries. Chapter 4 will point at the problems of using quantitative
approaches to measure TT and instead use a qualitative study of specific cases to show
the actual process in which TT takes place and the difficulties in measuring its effects.
Our review has shown some limitations of input, output and effect measurements and
recommended their joint use to enable an assessment of policy effectiveness. The
indicators proposed by the UNFCCC are instead too concerned with input data, which
cannot show the actual effect of efforts to improve TT. We recommend an extension
of technology performance indicators in the UNFCCC framework. The review has also
shown the difficulty of using output and effect indicators of TT in which the part of the
output or effect of the TT that is attributable to foreign technologies cannot be
isolated from other influential factors. This is particularly the case when CO2 emission
reductions are used as an indicator of the effect of TT.
2.5.3 Empirically, which factors enable climate change technology transfer to developing
countries?
The literature review has delivered a useful framework of indicators to assess the
status of developing countries and to define policy priorities accordingly. These include
both qualitative and quantitative indicators in five areas: macroeconomic and
institutional frameworks; technology demand; technology supply; industrial
development and the bilateral characteristics of the supplier and recipient.
Existing literature has mostly analysed the effect of these enabling frameworks to
explain the success stories of China, India and Brazil’s renewable energy industries.
However, these countries are hardly “average” developing countries, as the sheer sizes
of their economies attract foreign investors looking for large profits, and their power in
the world trade scene enables them to apply industrial policies which probably would
not be possible elsewhere. This thesis will use the identified enabling factors to analyse
the different challenges faced by a smaller economy. Chile is considered a relevant
case because many of the enabling frameworks are in place, such as stable economic
and institutional frameworks, a growing economy, an open trade philosophy and
relatively high capabilities. However, the penetration of low-carbon technologies is still
low. Our diagnosis will be mainly qualitative, but will also quantify the enabling factors
identified in the literature review. The results show which factors prevent a higher
penetration of foreign low-carbon technologies and their adoption by local actors. The
analysis of ten specific case studies also provides the first-hand perspective of the
private sector about the barriers and enablers to achieving higher levels of technology
transfer.
2.5.4 How do different developing countries perform in these enabling factors, and which
differentiated policy priorities can be defined accordingly?
Existing literature does not quantitatively assess the different performances of
developing countries as regards their enabling frameworks for climate change
technology transfer; instead, the identification of enabling frameworks is based on
individual case studies (mostly in China, India and Brazil) and on econometric models.
44
As such, the literature cannot provide a basis upon which to define differentiated
policy priorities to improve technology transfers to developing countries according to
country-specific circumstances.
Chapter 5 of this thesis uses multivariate analysis to identify the main strengths and
weaknesses of developing countries. Clusters of countries with similar performances
will be created, which will allow for the definition of policy priorities for groups of
countries with similar needs.
45
3 Chile’s enabling environment for climate
change technology transfer17
3.1. Introduction
3.2. Chilean energy system and GHG emissions
3.3. Economic and institutional framework
3.4. Technology demand
3.5. Technology supply
3.6. Industrial development
3.7. Conclusions
17
Part of this chapter is published in Energy Policy, Vol. 39, issue 7, pages 4274-4283. “The role of
technology transfer for the development of a local wind component industry in Chile”. Authors: Ana
Pueyo, Rodrigo Garcia, Maria Mendiluce and Darío Morales.
46
3.1 Introduction
The review of the literature in Chapter 2 has identified a number of elements in a
recipient country that can enhance climate change technology transfer. Chapter 4
analyses a set of case studies about climate change TT processes in Chile in order to
provide answers to the research questions presented in the introduction to this thesis.
The purpose of this chapter about Chile’s enabling environments for climate change
technology transfer is to provide the background necessary to interpret the evidence
collected in the case studies. With this aim, the chapter starts by introducing the
Chilean energy system. It shows the carbon intensity of the country’s energy supply,
the functioning of its electricity system, the penetration and potential of renewable
energies and the level of implementation of the Clean Development Mechanism. The
chapter continues by providing data about the enabling factors for TT in Chile, which
based on the findings of the literature review, are classified in economic and
institutional; technology demand factors and policies; technology supply factors and
policies and industrial development factors.
Data have been mostly collected through the analysis of relevant documentation, such
as energy balances, reports from the OECD, the IEA, the CNE (National Electricity
Commission) and the UN. In addition, four experts in the Chilean renewable energy
market and its technology policy have been interviewed to increase familiarity with the
Chilean context. Table 5 shows details of these interviews.
TABLE 5- DETAILS OF INTERVIEWS WITH CHILEAN EXPERTS
Name of Date of
Position Organisation
interviewee interview
Fernando Molina Associate Environment Lawyer Barros y Errazuriz 22-Nov-10
Conrad von Igel Director of the Department of Innovation Ministry of Economy of Chile 30-Nov-10
Darío Morales Manager of InnovaChile Industrial Development Agency of 03-Dec-10
Chile- CORFO
Rodrigo García Technical Director Renewable Energies Center of 03-Dec-10
Chile- CER
Fernando Cubillos Independent consultant ex Morgan Stanley Latin America 25-Nov-10
3.2 Overview of Chilean energy system and GHG emissions
3.2.1 Energy consumption and GHG emissions in Chile

Chilean CO2 emissions from the consumption of energy reached 62.55 Mt in 2009,
which represented 0.2% of global emissions for this concept. As shown in Figure 1, the
country’s emissions are well behind the BRICS, which occupy the first (China), third
(India), fourth (Russia), eleventh (South Africa) and fourteenth (Brazil) positions in the
world. BRICS are jointly responsible for 39% of global CO2 emissions from energy
consumption. Chilean emissions are also well behind many other developing countries.
47
FIGURE 1- CO2 EMISSIONS FROM ENERGY CONSUMPTION PER COUNTRY
Total CO2 emissions from the consumption of energy (Mt )- 2009

9000
8000
7000
6000
5000
4000
3000
2000
1000
0
Philippines
Argentina
Turkey
Belgium
Israel
Russia
Algeria
Greece
South Africa
Saudi Arabia
Australia
France
Egypt
Mexico
Iran
Iraq
Czech Rep
Brazil
Spain
Poland
UK
China
Canada
Venezuela
Nigeria
Austria
Chile
Germany
UAE
Other
US
India
Indonesia
Netherlands
Hong Kong
Kuwait
Thailand
Vietnam
Singapore
Colombia
Japan
Taiwan
Kazakhstan
Ukraine
Malaysia
Romania
Italy
Korea, North
Korea, South
Pakistan
Uzbekistan
Qatar
Source: US Energy Information Administration (EIA), International Energy Statistics. http://tonto.eia.doe.gov
According to EIA data, Chilean emissions grew by 111% in the period 1989-2009. Most
of the growth took place in the 1990s, reaching 100%, but it slowed down
subsequently, with a growth rate of only 5% between 1999 and 2009 and a decrease in
emissions of 8% in the period 2005-2009.
Absolute CO2 emissions from energy consumption have grown at a lower rate than
GDP. As shown by Figure 2 in the 20 years between 1989 and 2009, Chile reduced its
carbon intensity18 by 20%. After some increase in carbon intensity during the ‘90s, the
figure was reduced by 27% between 1999 and 2009. Energy demand has closely
followed GDP growth.
FIGURE 2 - EVOLUTION OF CHILEAN CARBON INTENSITY
Chile's Carbon Intensity using PPP

(tCO2 per Thousand 2005 US$)
0.45
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
Figure 3 shows that the Chilean performance as regards the reduction of its carbon
intensity during 1989-2009 is better than the 15% reductions of Brazil and India, but
not as good as China, which achieved carbon intensity reductions of 51% in the same
period, although departing from a much more carbon-intensive economy. Chilean
18
Carbon intensities calculated as Metric Tons of Carbon Dioxide per Thousand, Year 2005, U.S. Dollars,
at power purchasing parities.
48
carbon intensity in 2009, at 0.28 tCO2 per thousand US$ at PPP, is below the world
average of 0.46 tCO2 per thousand US$, in line with the average found in Central and
South America, and significantly lower than that of China, India and South Africa. This
shows a different economic structure, where heavy industry does not have an
important role.
FIGURE 3 - CHILEAN CARBON INTENSITY COMPARED TO CHINA, INDIA, BRAZIL AND SOUTH AFRICA
Carbon Intensities of BRICs and Chile

2
1.8
tCO2 per thousand 2005 US$ ppp
1.6
1.4 Brazil
1.2 Chile
1 South Africa
0.8 China
0.6 India
0.4
0.2
0
1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009
Chilean energy consumption per capita is still well below that of more developed
regions, but total consumption is expected to increase as the country´s income levels
rise.
According to the latest Chilean energy balance, the main primary energy source in
2008 was oil, followed by hydroelectricity and wood. This is illustrated by Figure 4.
FIGURE 4- CHILE’S PRIMARY ENERGY CONSUMPTION (TERACAL)
Gross primary energy consumption 2008

(Teracal)
PETROLEO CRUDO
(a)
18% GAS NATURAL
38% (b,c,d,e)
CARBON
21%
HIDROELECTRICIDA
D
15% 8%
LEÑA Y OTROS
Source: National Energy Commission, Chilean Government. National Energy Balance 2008.
49
Note: the calorific power used for hydroelectricity is 2,504 Kcal/Kwh (thermal equivalence of the national
power generation matrix). The standard calorific power using the international methodology for energy
balances is 860 Kcal/Kwh. Using this second measurement, hydroelectricity would represent 8% of
primary consumption.
Most oil, gas and coal products consumed in Chile are imported, and the nation
depends on external sources of energy for 70% of its energy consumption, which
makes it very vulnerable to external shocks. Such was the case in April 2004, when
Argentina severely restricted exports of natural gas to Chile, thus causing the so-called
“gas crisis”. Since 2004, natural gas consumption has decreased steadily and has been
substituted mostly by more carbon-intensive fuels such as coal and diesel. To
encourage investments in new generation plants, the Chilean Government introduced
changes to electricity tariffs regulations, allowing generators to sign long-term
contracts with distributors at a fixed price to reduce the risk of tariff reductions. As a
result, the prices agreed with distributors increased from close to US$ 60/MWh to
more than US$ 300/MWh, which encouraged investments in around 6GW of new
capacity between 2007 and 2011. Most new investments consider coal and
hydroelectricity as energy sources.
FIGURE 5- EVOLUTION OF PRIMARY ENERGY CONSUMPTION IN CHILE 2000-2008
Primary energy consumption

140,000
120,000
OIL
100,000
NATURAL GAS
Teracalories
80,000 COAL
60,000 HYDROELECTRICITY
40,000 WOOD
BIOGAS
20,000
0
2000 2001 2002 2003 2004 2005 2006 2007 2008
Source: National Energy Commission, Chilean Government. National Energy Balances 2000 to 2008.
Note: the calorific power used for hydroelectricity is 860 Kcal/Kwh according to the international energy
balances methodology.
Chile is looking to diversify its energy matrix to avoid situations of vulnerability such as
the one resulting from Argentina’s natural gas cuts. Coal and hydroelectricity are the
sources most often considered, but they have faced public protests due to their
environmental impact. The contribution of renewable energies is still very small,
though, and is expected to reach just 10% of electricity generation by 2020, which
would not solve the country’s energy needs. Chile has also considered nuclear power
to diversify its energy matrix, signaled by the Chilean Government sponsoring a
number of studies about its feasibility. It has also signed nuclear cooperation
agreements with the United States and France; however, the recent events that
caused a nuclear crisis in Japan have exposed the vulnerability of earthquake-prone
Chile and delayed the government’s decisions about the development of nuclear
power in the country.
50
As regards secondary energy consumption, oil derivatives are the most important
energy source, followed by electricity and wood, with similar shares. Shares of the
different secondary energy sources are shown in Figure 6.
FIGURE 6- SOURCES OF FINAL ENERGY CONSUMPTION IN CHILE 2008
Final energy consumption 2008 (Teracal)
OIL DERIVATIVES
ELECTRICITY
18%
COAL (**)
2%
COKE
1%
1% TARS (***)
2%
GAS
57%
BLAST-FURNACE GAS
19%
NATURAL GAS (**)
METHANOL
WOOD
Notes:
The calorific power used for hydroelectricity is 860 Kcal/Kwh according to the international energy
balances methodology.
(**) Imports and exports are considered in primary energy balance.
(***) Energetic tars (as produced in the steel industry).
Energy consumption is quite evenly distributed among industry and mining,

transportation, the tertiary and residential sectors and energy transformation centres,
as shown in Figure 7.
FIGURE 7- ENERGY CONSUMPTION PER SECTOR IN CHILE 2008
Energy consumption per sector (Teracal)
23% Transport
28%
Industry and Mining
Commercial, Institutional,
20% Residential
Transformation Centers
29%
Transportation relies almost exclusively on imported oil, with road transport

responsible for most consumption. Energy intensity in transport activities declined by
51
10% from 1999 to 2003 and has remained stable since that year (APEC, 2009)19.
Industry and mining sectors use electricity and oil derivatives in similar shares, with
biomass holding also a significant proportion. Biomass is also a very important energy
source for the residential sector, while the tertiary sector depends mostly on
electricity. Energy transformation plants are heavy users of coal, with oil derivatives
and natural gas following suit.
FIGURE 8- SOURCE OF ENERGY CONSUMPTION PER SECTOR IN CHILE 2008
Energy consumption per sector and energy source

(teracal) 2008
120,000
BIOGAS
100,000 WOOD
80,000 METHANOL
NATURAL GAS (**)
60,000
STEEL PLANT GAS
40,000 GAS
20,000 TAR (***)
COKE
0
COAL (**)
ELECTRICITY
OIL DERIVATIVES
Source: Comision Nacional de Energia Chile (CNE), National Energy Balance 2008.
Notes:
Inside the industry and mining sectors, copper mining stands out as the highest
energy-consuming sector, as it is one of the pillars of the Chilean economy and needs
vast amounts of energy to operate. Its main energy source is electricity, followed by oil
derivatives. The pulp and paper sector is also a significant energy consumer, but it
relies mostly on indigenous biomass resources. Oil derivatives are a very significant
energy source in other industrial sectors.
19
52
FIGURE 9- ENERGY CONSUMPTION IN CHILEAN INDUSTRY AND MINING SECTORS 2008
Energy consumption industry and mining sectors (Teracal) 2008
30,000
WOOD
25,000 METHANOL
NATURAL GAS (**)
20,000
GAS STEEL PLANT
GAS
15,000
TAR (***)
10,000 COKE
COAL (**)
5,000 ELECTRICITY
OIL DERIVATIVES
0
SALTS
COPPER
PAPER
IRON
SUGAR
FISHING
CEMENT
STEEL
OTHER IND
OTHER MIN
PETROCHEMIC
ALS
Source: Comision Nacional de Energia Chile (CNE), National Energy Balance 2008.
Notes:
3.2.2 The Chilean electricity system

Electricity provides nearly 20% of Chilean final energy consumption, as measured by
heating power. In 2010, Chile had an installed electricity capacity of 16,415MW in the
national grid and produced 58,660 GWh, selling 55,210 GWh. Electricity generation
increased by 77% between 1998 and 2010, at an annual average of 5%. Increases in
capacity to meet this growing demand have been met mostly with traditional energy
sources, namely fossil fuels and hydroelectricity. Chile’s non-diversified electricity
generation matrix has made it vulnerable to restrictions in Argentina’s gas supply and
low rainfall. As a result, Chile has experienced serious periods of electricity shortages
during the first decade of this century, as well as a sustained increase in electricity
prices.
FIGURE 10- EVOLUTION OF ELECTRICITY GENERATION IN CHILE 1998-2010
Total electricity generation in Chile 1998-2010

70,000.0
60,000.0
50,000.0
40,000.0
Gwh
30,000.0
20,000.0
10,000.0
0.0
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Source: CNE (2011)
53
Industry and mining sectors are the main consumers of electricity in Chile, above the
tertiary and residential sectors.
FIGURE 11- FINAL CONSUMPTION OF ELECTRICITY IN CHILE, 2010
Final consumption of electricity in Chile 2010

4%
Energy sector: self supply

29%
Industry and mining
Transport
1%
66% Commercial, Institutional

and Residential
Source: CNE, 2010
Within mining and industry, copper mining stands out as the main consumer, as it is
the pillar of the Chilean economy.
FIGURE 12- FINAL CONSUMPTION OF ELECTRICITY IN CHILE BY MINING AND INDUSTRY SECTORS, 2010
Industry and mining electricity consumption 2010
Copper
4%
Salts
Iron
25% Pulp and paper
Iron and Steel
49%
Petrochemicals
Cement
2%
Sugar
1%
2% Fishing
15%
Other industry
Other mining
1% 1%
Source: CNE, 2010
The Chilean electricity system consists of four major electricity sub-systems supplying
four different regions (CNE, 2010):
• The Northern Interconnected System (Sistema Interconectado del Norte
Grande- SING) supplies the northern zone of Chile where the mining industry is
located. With 4,103 MW20 in December 2010 it represented 25% of Chile´s total
installed capacity. However, it only supplies electricity to 5.8% of the
population, as its main customer is the mining industry.
• The Central Interconnected System (Sistema Interconectado Central- SIC)
supplies the central zone of Chile, which concentrates 90% of the Chilean
20
Including power generation units undertaking grid connection tests in December 2010
54
population. In December 2010, it had 12,180 installed MW and represented
74% of the country’s total installed capacity.
• The Electricity System of Aysen caters for a small proportion of the Chilean
population and represents only 0.3% of Chilean installed capacity.
• The Electricity System of Magallanes supplies the most southerly part of the
country and represents 0.6% of Chilean installed capacity.
Fossil fuels represent 99.6% of installed capacity in the SING and around 50% in the
SIC, where hydroelectricity has a strong participation. The Magallanes Electricity
System uses only fossil fuels and the Aysen Electricity System is split as 57% fossil fuels,
39% hydro and 4% wind. In 2010, 42% of electricity generation in Chile came from
hydroelectricity plants, 56.4% from fossil fuels, mostly coal and diesel oil, and the
remaining 1.6% from biomass and other renewable energies.
The Chilean electricity sector has a high level of market concentration. In 2008, three
companies and their subsidiaries owned 83% of the installed public service power in
the SIC, with a single company, Endesa, supplying almost half of the electricity needs of
the country. Market shares are shown in Figure 13. Chile has the most concentrated
electricity market among a group of Latin American countries including Argentina,
Brazil, Colombia and Peru (Maldonado and Palma, 2004).
FIGURE 13- ELECTRICITY GENERATION MARKET SHARES IN THE SIC, 2008
Source: Pontificia Universidad Católica de Chile, Departamento de Ingeniería Eléctrica, IEE3272 Mercados Eléctricos
II-2009. http://web.ing.puc.cl/~power/alumno09/concentra/conc_chile/con_chile.htm
In 1982, Chile was a world pioneer in the liberalisation of its electricity market. The
government delegated to private sector actors the responsibility to supply electricity,
and only kept the responsibilities of regulating the prices of distribution and
transmission, taxation and indicative planning of investments in generation and
transmission. Since then, decisions to invest in new capacity have been based only on
the marginal cost of electricity production for the available technology portfolio. A
reduced short-term energy price is the main objective, above other long-term
considerations such as limiting local greenhouse gas emissions or enhancing local
technology development and creating high-skilled jobs. The Chilean electricity system
55
is therefore founded on the premise that the best ways to meet energy demand at
affordable prices are to rely on competition between private companies, regulate
natural monopolies (distribution and transmission) and limit the role of the state in
business decisions. Embedded in this approach is the assumption that competitive
markets will deliver security of supply. However, the “gas crisis” caused by gas supply
cuts from Argentina and low rainfall have contradicted this assumption and put the
Chilean system under stress.
In the Chilean liberalised electricity system, electricity can be valued following four
different approaches (Universidad de Chile and Universidad Técnica Federico Santa
María, 2008):
• The spot or marginal price is calculated hourly by the Economic Load Dispatch
Centre (CDEC), the private organisation in charge of coordinating the operation
of the electricity system. Spot or marginal prices are calculated following a
marginalist approach. Marginal cost pricing, also called “peak load pricing”
involves consumers paying the marginal cost of the last power plant to enter
the dispatch order every hour, which would be the least efficient plant.
Customers also pay a price for the capacity associated with peak demand
hours, which ensures that revenues from electricity sales and from the sale of
capacity cover the investment and operational costs of producers as a whole.
• Free prices are independently agreed between generators and free or non-
regulated customers.
• Nodal prices represent the price of energy and peak capacity for each of the
substations supplying electricity. They are defined by the National Electricity
Commission (CNE) according to an indicative investment plan, and they
represent the expected value of the marginal costs in a minimum horizon of 36
months. These reference prices are then used by distribution companies in
electricity supply tenders for regulated customers. The CNE is a government
organisation in charge of producing and coordinating the plans, policies and
standards for the correct operation and development of the energy sector.
• Distribution prices are set by distribution companies for the sale of energy and
capacity to regulated customers. These prices have two components: the nodal
price (tender price), reflecting the average price that distributors pay for energy
to generators, and the distribution value-added (VAD), reflecting the
distribution costs of an efficient company.
Electricity consumers in Chile are grouped into two main segments: regulated
consumers and non-regulated consumers, the former of which are those with a
connected capacity equal to or less than 2MW. They pay electricity tariffs at the
“distribution price”, calculated as the nodal price plus the distribution value-added.
The regulated segment represents approximately 66% of consumption in the SIC and
10% of consumption in the SING (CNE and GTZ, 2009). Free or non-regulated
customers are those with a connected capacity higher than 2MW. Customers with a
connected capacity higher than 0.5 MW also have the option of selecting this method,
and they represent approximately 90% of consumption in the SING and 34% in the SIC.
Free customers usually belong to the mining or industrial sectors and can freely
negotiate electricity supply prices and conditions with the generation and distribution
56
companies through bilateral financial contracts. Figure 14 shows schematically the
operation of the Chilean electricity market.
FIGURE 14- REMUNERATION IN THE CHILEAN ELECTRICITY MARKET
Source: CNE and GTZ (2009)
3.2.3 Penetration and the potential of renewable energies in Chile

Biomass and hydroelectricity are very important energy sources for the Chilean energy
system. Jointly, they provided 40% of primary energy consumption in 2008, when
considering a heating power for hydroelectricity equivalent to that of the Chilean
electricity generation matrix, or 28% using the lower international standard heating
power for hydro. In the electricity generation matrix, hydropower plants can provide
up to 50% of the country’s electricity needs in years with good rainfall (such as in 2006)
and had 35% of the installed capacity in 2008.
However, the penetration of so-called “non-conventional renewable energies” (NCRE)
in the Chilean electricity system is very low. NCRE is defined by the Chilean government
as electricity generated through non-conventional, renewable, primary energy sources,
that have a low environmental impact, including:
• Biomass, used as the primary energy source, obtained from organic
biodegradable matter
• Hydropower, with a maximum electric potential of less than 20MW
• Geothermal energy
• Solar energy
57
• Wind
• Marine power
• Other means of renewable generation determined by the CNE
In December 2010, NCRE represented 3.42% of the installed capacity in the SIC and
0.4% of installed capacity in the SING, as shown in Table 6 and Table 7. The small Aysen
system contributed with 21MW of NCRE to the Chilean system, while Magallanes was
100% fossil fuelled. The penetration of NCRE in Chile is therefore very small, with only
452 MW of installed capacity.
TABLE 6- TOTAL INSTALLED CAPACITY IN THE SIC, 2010
Net installed Net installed

Type of fuel
power (MW) power (%)
Hydro dam 3,768.1 30.93%

Hydro run of river >20 MW 1,762.8 14.47%
Coal 1,286.1 10.56%
Natural Gas 2,721.1 22.34%
Oil 2,226.5 18.28%
Total Conventional 11,656.4 96.58%
Hydro <20 MW 123.6 1.01%
Wind 160.5 1.97%
Biomass 240 1.08%
Total NCRE 416.1 3.42%
Total installed power 12,180.8 100.00%
Source: CNE, 2010

th
Note: Figures include plants performing grid connection tests on 30 December 2010.
TABLE 7- TOTAL INSTALLED CAPACITY IN THE SING, 2010
Net installed Net installed

Type of fuel
power (MW) power (%)
Coal 1,666.1 40.6%

Natural Gas 2,073.9 8.5%
Oil 348.2 50.5%
Total Conventional 3,560 99.6%
Hydro <20 MW 14.9 0.4%
Total NCRE 14.9 0.4%
Total installed power 3,575 100.00%
Source: CNE, 2010

th
Note: Figures include plants performing grid connection tests on 30 December 2010.
In spite of this small penetration of NCRE in Chile, the private sector has shown
considerable interest in undertaking such projects, which is confirmed by the number
of NCRE projects that have been presented for approval to the Environmental Impact
Assessment System (SEIA) of the Chilean Government. In June 2010, there were 62
approved projects representing 2,392 MW in the SEIA, mostly wind, and 16 projects
58
under review with 413 MW, but only 19% of the approved projects have actually been
implemented.
TABLE 8- NCRE PROJECTS APPROVED IN THE SEIA AT JUNE 2010
Type Approved Under review Total

Projects MW Projects MW Projects MW
Hydro <20 MW 34 276 10 135 44 411
Wind 27 1,827 4 265 31 2,092
Biomass 10 280 1 4 11 284
Solar 1 9 1 9 1 18
Total NCRE 62 2,392 16 413 88 2,805
Source: SEIA, 2010
Several studies present projections of the potential development of renewable

energies in Chile. A study by Universidad de Chile and Universidad Federico Santa
Maria about the NCRE potential in the SIC estimates a gross potential of 191,000 MW,
according to its physical availability in the territory of the SIC. The technically feasible
potential would be 10,803 MW. The economically feasible potential in 2025 would be
between 3,332 MW under a conservative energy prices scenario, and 5,753 MW under
a dynamic energy prices scenario. This economically feasible potential would generate
between 17,743 GWh and 29,652 GWh, representing between 16.8% and 28.1% of the
expected demand of the SIC in 2025, at 105,560 GWh. These estimations are based on
a set of assumptions including 10% discount rates; three electricity price scenarios:
conservative with a price of US$75/MWh, dynamic and dynamic-plus both at
$US102/MWh; a 1% annual increase in energy prices for conservative and dynamic
scenarios and a 3.5% annual increase for the dynamic-plus scenario; and a 5% annual
increase in energy demand. Table 9 summarises the results of this study.
TABLE 9- NCRE POTENTIALS IN THE SIC IN 2025
Economic potential (MW)

Gross
Technical Dynamic-
Type potential
potential (MW) Conservative Dynamic plus
(MW)
Hydro <20 MW 20,392 3,003 1,421 1,653 1,850
Wind 40,000 1,500 330 998 1,200
Biomass 13,675 3,249 461 501 903
40,000-
Solar thermal 1,051 210 210 250
100,000
Solar PV 1,000 500 100 100 150
Geothermal 16,000 1,500 810 940 1,400
Total NCRE 191,067 10,803 3,332 4,402 5,753
SIC (2025) MW 22,736
% SIC capacity 840% 47,5% 14,7% 19% 25,3%
SIC (2025) GWh 105,560
% SIC generation 565% 53% 17% 21% 28,1%
Source: Universidad de Chile and Universidad Federico Santa María (2010)
59
Some conclusions by the aforementioned study are that the economically feasible
potential for electricity generation with ERNC is much higher than the current targets
set by the government. As regards sources of energy, hydro projects have the highest
potential, due to their wide availability, large capacity and low cost. Geothermal
projects also have a strong potential but still need to reduce exploration risks and deal
with social acceptance when they are located close to indigenous land, natural
reserves or touristic areas.
The CNE (2010) report on nodal electricity prices for the period 2010-2020 shows that
102 MW of new NCRE plants are under construction in the SIC, expecting to enter the
system in 2010 and 2012. A total of 51 MW will come from small hydropower plants,
31MW from biomass and 20 MW from wind power plants. NCRE will represent 4% of
new capacity additions to the SIC in 2010 and 2012. In this period 47% of new capacity
in the SIC will be provided by coal plants and 41% by large hydro plants. In addition,
the CNE recommends 3,776 MW of additional capacity to meet electricity demand
between 2011 and 2020. NCRE represents 27% of the CNE’s recommended capacity,
with the strong participation of wind power, while the remaining 73% will be mostly
met by large hydroelectric projects.
The CNE (2010) nodal prices report for the SING shows 760 MW of new capacity under
construction, planning to start operations in 2010 or 2011. All this capacity is met by
coal plants. The CNE recommends some additional 1,670 MW to cover demand in the
period 2011-2019. NCRE would represent 16% of recommended capacity and consist
of wind (190 MW) and geothermal (80 MW) plants, while the remaining recommended
capacity would be met by coal plants.
3.3 Economic and institutional framework

Chile joined the selective group of OECD members in May 2010, being the first South
American member and the first country to join the organisation in ten years. However,
Chile still endures a significant income gap against other OECD countries, with a GDP
per capita around one-third of the level in advanced OECD countries.
Chile implemented a number of macroeconomic reforms from the mid-1980s onwards
to liberalise foreign trade, improve the functioning of labour, product and financial
markets and to restructure the pension system. Its macroeconomic policy framework
has achieved sound stability through a structural fiscal rule with an independent
inflation-targeting central bank and a flexible exchange rate regime. Chile has a
reputation for strong financial institutions and sound policy, and has the strongest
sovereign bond rating in South America (OECD, 2010).
These aspects have led to a favourable environment for private investment. The
Chilean investment framework is characterised by high economic growth, the sharp
drop of public debt, the stabilisation of external accounts and favourable conditions
for international trade, amongst others.
3.3.1 Ease of doing business

The World Bank indicator on “Ease of doing business” placed Chile in the 43rd position
amongst a total of 183 countries in 2011 (World Bank, 2011). Among Latin American
and Caribbean countries, Chile is the fourth best country in terms of ease of doing
60
business, after Mexico, Peru and Colombia. This indicator evaluates each country
according to nine topics: starting a business, dealing with construction permits,
registering property, getting credit, protecting investors, paying taxes, trading across
borders, enforcing contracts and closing a business. The ranking on each topic is the
simple average of the percentile rankings on its component indicators. Chile is
particularly well positioned in registering property, protecting investors, paying taxes
and enforcing contracts; however, it does not score well on starting a business, dealing
with construction permits, getting credit, closing a business and, surprisingly, given the
country’s free trade economy, trading across borders. The trading indicator is
constructed by compiling procedural requirements for exporting and importing a
standardised cargo of goods by ocean transport. Documents associated with every
official procedure are counted – from the contractual agreement between the two
parties to the delivery of goods – along with the time and cost necessary for
completion (World Bank, 2011).
3.3.2 Open trade

Chile has unilaterally liberalised trade, applying a single general import tariff of 6%,
and reduced import taxes. It has also signed numerous free trade agreements, reduced
non-tariff barriers and signed investment promotion and protection agreements with a
number of countries. Free trade agreements allow companies in Chile to access 86% of
the world’s GDP (Invest Chile, 2011), and the country aspires to form a common
market in South America with Colombia and Peru, joined by Mexico, all of which enjoy
a growing affinity as fast-growing, medium-sized countries with Pacific coastlines,
betting on market economies, foreign investment and trade with Asia to achieve
development (The Economist, 2011). All of these policies have made Chile one of the
most liberal economies in the world and encouraged a heavy dependence of the
economy on exports, which contributed 20% to GDP in 2009.
Chile was ranked 11th in the Heritage Foundation and The Wall Street Journal’s 2009
Index of Economic Freedom. This score placed the nation above nearly all of the
European economies and only one position behind the United Kingdom. In the Global
Trade Enabling Report, issued by the 2008 World Economic Forum, Chile was ranked as
one of the world’s most open economies in terms of international trade and
investment – 27th out of 118 countries (WEF, 2010). Meanwhile, in the 2010 World
Competitiveness Yearbook, published by the Institute for Management Development
(IMD) in May 2010, Chile took 28th place out of 58 economies. The 2008 World
Investment Report, published by the United Nations Conference on Trade and
Development (UNCTAD, 2008), placed Chile in third position among the ten most
successful economies for attracting foreign investment in Latin America and the
Caribbean.
Foreign investment has in fact become an essential part of the country´s national
development strategy. Over the past decade, FDI has represented an annual average
6.5% of GDP, rising to an average of 8% in the period 2007-2009 (UNCTAD, 2009). FDI
inflows actually quadrupled in the five years to 2008.
61
3.3.3 Protection of IPR
As a member of the World Trade Organisation (WTO), Chile must comply with the
obligations imposed by the Agreement of Trade-Related Aspects of Intellectual
Property Rights (TRIPs). Accordingly, the country passed a law in 2005 to reinforce
intellectual property rights according to TRIPS standards.
The 2010 IPRI (International Property Rights Index) ranking, which measures the
relative strength of a property rights regime in 129 economies, places Chile in 28th
position, ahead of every Latin American and Caribbean country (Property Rights
Alliance, 2010). However, the Government of the United States has raised a number of
concerns about Chile´s lack of commitment to prosecute those contravening IPR law,
mostly related to pharmaceutical products and clinical examination data (IISD, 2010).
3.3.4 Low corruption

The Corruption Perceptions Index, elaborated by Transparency International for 178
countries, places Chile in 21st position, ahead of every Latin American and Caribbean
country, many European countries and even the United States (Transparency
International, 2011)
3.4 Technology demand factors and policies
3.4.1 Demand size

Chile has a population of 16.7 million and a GDP of 104.6 billion 2000 USD or 198.1
billion 2000 USD at purchase power parity (PPP). As shown by Figure 15, its economy is
therefore small in comparison with the so-called BRICs and with many of its Latin
American neighbours. Brazil, Mexico, Argentina, Colombia and Venezuela have larger
economies and all of them, plus Peru, also have larger populations.
FIGURE 15- ECONOMY AND POPULATION SIZES OF A SAMPLE OF LATIN AMERICAN AND WORLD
COUNTRIES IN 2008
2606
3000
2500
2000
1326
1140
1500
854
826
769
1000
430
395
192
183
167
142
134
107
105
500
84
49
45
40
29
28
17
Population (Million) GDP (billion 2000 USD)
Source: IEA 2010, Key World Energy Statistics
Chilean total primary energy supply (TPES) and electricity consumption are accordingly
behind most of these larger countries, as shown in Figure 16.
62
FIGURE 16- TPES AND ELECTRICITY CONSUMPTION FOR A SAMPLE OF WORLD AND LATIN AMERICAN
COUNTRIES IN 2008
3252
3500
3000
2116
2500
2000
1500 914
687
645
1000 621
429
249
232
215
181
134
111
500
86
76
64
56
44
31
31
30
15
0
TPES (Mtoe) Elec cons (TWh)
In relative terms, Chile shows energy intensity similar to that of OECD countries, below
dynamic emerging economies, except India, but higher than all Latin American
countries except Venezuela. Energy consumption per capita is lower than that of OECD
countries, but higher than China and India and most Latin American countries.
FIGURE 17- 2008 TPES PER CAPITA AND PER GDP FOR A SAMPLE OF WORLD AND LATIN AMERICAN
COUNTRIES
6
5
4
3
2
1
0
TPES/pop TPES*10/GDP PPP
Electricity consumption per capita is higher than in any other Latin American country
selected in the sample, but still far from OECD levels.
63
FIGURE 18- 2008 ELECTRICITY CONSUMPTION FOR A SAMPLE OF WORLD AND LATIN AMERICAN
COUNTRIES
9000
8000
7000
6000
5000
4000
3000
2000
1000
0
Elec cons/pop
3.4.2 Demand growth

The Chilean economy has experienced a fast recovery from the global recession and
the damage caused by the earthquake in 2010. Growth rates of up to 5 percent were
expected for 2010 and up to 6 to 7 percent for 2011 (Bloomberg, 2010). Income per
capita grew on average by 4.3% per year between 1986 and 2007, increasing from 18%
relative to the United States in 1986 to 31% in 2007 (OECD, 2010). However, growth
has substantially slowed since the end of the 1990s, mainly as a result of slow
productivity growth (OECD, 2010).
Chile has shown very high electricity and total primary energy consumption (TPEC)
growth rates in the last two decades, as shown in Table 10. Growth rates are above
Central and South America averages and among the highest in the region; however,
they do not exceed the dynamic Chinese and Indian economies.
TABLE 10- AVERAGE ELECTRICITY CONSUMPTION AND TPEC GROWTH BETWEEN 1990-2008 AND 2000-
2008 IN A SAMPLE OF LATIN AMERICAN AND WORLD COUNTRIES
Electricity consumption TPEC

Country Av. 90-08 Av. 00-08 Av. 90-08 Av. 00-08
Argentina 6% 5% 3% 3%
Brazil 4% 3% 4% 3%
Chile 7% 6% 5% 3%
Colombia 5% 5% 3% 2%
Mexico 4% 3% 3% 2%
Peru 6% 7% 3% 3%
Uruguay 3% 3% 2% 1%
Venezuela 3% 4% 3% 2%
Central & South America 4% 4% 3% 3%
South Africa 2% 2% 3% 3%
China 10% 12% 7% 10%
India 7% 6% 5% 5%
Source: US Energy Information Administration, International Energy Statistics, 2011, www.eia.gov
64
3.4.3 Electricity prices
Nodal prices are the most commonly used as a reference for analysing the evolution of
electricity prices in Chile, as they reflect electricity supply costs associated with current
and planned capacity, with a horizon of around ten years. Nodal prices are calculated
using the current and expected marginal costs of electricity. Before restrictions to
imports of gas from Argentina, marginal costs were usually determined by natural gas-
fired power plants, which operated at the peak of demand. After the gas crisis, peak
load was usually supplied by diesel plants, which involved significant increases in
marginal costs, so generators transferred cost increases to non-regulated clients,
increasing their prices. Nodal prices, using marginal costs as a reference, have also
significantly increased. Figure 19 illustrates this trend, showing a steep price increase
since gas restrictions in 2004 in both the SIC and the SING, with peaks due, among
other reasons, to high oil prices. Electricity prices in Chile are significantly high, which
increases the cost of industrial production but sends the right signals to investors
willing to invest in additional capacity.
FIGURE 19- NODAL PRICE 1982-2010 IN CHILE
NODAL PRICE (US$)
150
130
110
[mills/kWh]
90
70
50
30
10
OCTUBRE ABRIL 1985 ABRIL 1988 ABRIL 1991 ABRIL 1994 ABRIL 1997 ABRIL 2000 INDEX. A OCTUBRE MOD JUNIO INDEX. A OCTUBRE INDEX A INDEX A INDEX A INDEX A
1982 DIC-01 2003 2005 SEPT-06 2007 OCT-2008 ABR-2009 OCT-2009 AGO-2010
SIC - Santiago SING - Crucero
Source: CNE, 2011
The latest available data in the IEA database on electricity prices for a selection of
developing countries show that Chile has comparatively high prices for industry and
households, which would attract foreign investors in renewable electricity generation
technologies. Additionally, electricity prices are significantly higher than the OECD
average for both households and industry. In 2009, household prices were higher than
any other country in the sample except Uruguay, and in the same year industry prices
were only higher in Turkey, Brazil and El Salvador.
65
FIGURE 20- ELECTRICITY PRICES FOR INDUSTRY IN THE OECD AND A SAMPLE OF DEVELOPING
COUNTRIES, 2008 AND 2009
Electricity prices industry (US$/KWh)

0.25
0.2
0.15
0.1
2008
0.05
2009
0
Source: IEA, 2011
FIGURE 21- ELECTRICITY PRICES FOR HOUSEHOLDS IN THE OECD AND A SAMPLE OF DEVELOPING
COUNTRIES, 2008 AND 2009
Electricity prices households (US$/KWh)

0.25
0.2
0.15
0.1 2008
2009
0.05
Source: IEA, 2011
3.4.4 Demand-pull policies

For the past 30 years, Chilean energy policy has been founded on the premise that the
best ways to meet energy demand at affordable prices are to rely on competition
between private companies, regulate natural monopolies and limit the role of the
state in business decisions. In 1982, Chile pioneered the privatisation of the electricity
market, since when average technology costs and technical reliability have been the
only relevant variables taken into consideration for decisions on capacity expansion,
ranking above other considerations such as diversifying the generation portfolio,
66
limiting greenhouse gas emissions, enhancing local technology development or
creating high-skilled jobs.
The assumption that competitive markets would deliver security of supply was
embedded in this private approach to the electricity system. However, heavy
dependence on imported fossil fuels created periods of electricity shortages during the
first decade of this century. Chile’s energy sector is characterised by limited indigenous
fossil energy resources, unlike many of its South American neighbours. On the other
hand, the country’s geography is gifted with a significant renewable energy potential,
which has not been fully harnessed to increase security of supply because long-term
planning on capacity expansion to incorporate renewable energies is not easily
compatible with the existing model, which is better suited for short-term decisions on
dispatch through merit order.
Over the past five years, conditions for the development of non-conventional
renewable energy (NCRE) in Chile have improved significantly. This has been achieved
through new laws, the creation of instruments for direct support to investments (loans
and grants for pre-investment studies), better information on renewable resources,
the implementation of investment projects and recognising the diversification of the
energy mix as one of the core objectives of the current energy policy (CNE and GTZ,
2009).
The 2004 Law 19.940 opened up the spot market to small-scale plants, guaranteeing
them the right to connect to distribution networks and exempting them from main
transmission tolls. The 2008 Law 20.257 defined the NCRE and sought to create
favourable conditions for investment projects in this kind of energy through a quota-
based system. From 2010, all electricity generators of the SIC and the SING that
operate over 200 MW installed capacity must accredit that 5 percent of the electricity
sold to the grid comes from NCRE. From 2014, this percentage will be gradually
increased by 0.5 annually to reach 10 percent in 2024. Any electricity company failing
to fulfill this obligation will have to pay a surcharge of 0.4 UTM21 for each MWh in
deficit. If a company fails to fulfill its obligation again in the three years after the first
deficit, it will have to pay a penalty of 0.6 UTM per each MWh in deficit. The penalty
value is approximately 30 USD/MWh. The quota system established by Law 20.257 is
not specifically a tradable green certificate (TGC) system, because no tradable
certificates are issued and there is not a parallel market for the NCRE generation
attribute. Electricity generators show compliance with their obligation through the
Injections and Purchases Statement, kept by the Economic Load Dispatch Centre
(CDEC), which is the organisation in charge of coordinating each interconnected
system. Generators must show every year that they have injected into the system the
required percentage of NCRE through their own generation plants or through other
generators, with which they can sign bilateral contracts to purchase their NCRE
attribute, currently valued at around 12USD/MWh22.
The 5% target set by Law 20.257 was easily met in 2010. However, the quota system
has not so far incentivated a significant growth of NCRE projects. The law applies only
21
A UTM is a “Monthly Tax Unit” defined every month by the Treasury Department. www.sii.cl.
22
Information provided by Rodrigo Garcia, Technical Director of the CER, as written communication on
st
21 June 2011.
67
to PPAs signed after August 2007, and most PPAs signed before that date were still
applicable in 2010. Only around 21% of electricity commercialised between January
and June 2010 was covered by the NCRE quota obligation (Galaz, 2010), while NCRE
allowed for accreditation represented around 8% of the electricity commercialised in
2010, which was covered by Law 20.257. However, NCRE allowed for accreditation
represented only 1% of the total electricity commercialised in that year – far behind
the 5% target.
The deadline for pre-August 2007 PPAs is not known, and therefore it is difficult to
estimate the future demand of new NCRE projects. Consequently, this situation
creates uncertainty regarding the real value of the renewable energy attribute and
makes it difficult to implement the legislation.
Some of the main critiques of the quota system established by Law 20.257 (ACERA,
2011; and Galaz, 2010) are:
• The target is not ambitious enough. The economically feasible NCRE potential
in Chile is estimated at between 16.8% and 28.1% in 2025. Quotas should be
reviewed periodically as technologies advance through the learning curve.
• The obligation falls on electricity generators instead of consumers, which
creates some difficulties for implementation. When consumers have the
obligation, a market can be created to provide the attribute. When generators
have to provide a share of NCRE, they are likely to implement their own
projects instead of paying competitors for them. When the NCRE attribute is
presented to other suppliers, it is done through bilateral transactions instead of
through a transparent market mechanism. As a result, the law gives too much
power to large generators over the development of the NCRE sector.
• There is not a liquid market for a homogeneous NCRE certificate. As a result,
there is not a common price for the NCRE attribute, which is defined on an ad
hoc basis in each bilateral contract.
• In terms of the asymetries of information, only a few large generators know the
deadlines for large PPAs signed before August 2007, and therefore they can
predict the future demand of NCRE certificates. Potential competitors in the
NCRE market do not count on this information to plan their projects or
estimate the value of the NCRE attribute.
• Technological neutrality, which benefits mature and low-cost technologies,
particularly in the mini-hydro field, does not pursue the objectives of energy
sources diversification, the compensation of social costs of conventional
sources, a reduction in energy dependence, electricity price stabilisation,
technology transfer increases or the creation of green jobs.
• The certification of renewable energy generation that can be used to comply
with the quotas is undertaken by the CDEC, which in Chile is controlled by large
conventional electricity generators.
In addition to Law 20.257, the current government aspires to incorporate at least 20%
of NCRE in total installed capacity by 2020. However, at the date of writing this thesis,
no roadmap or plan had been made public to illustrate how this aspiration would be
met. Furthermore, for local and international investors, an aspiration is not a strong
signal. In the competitive Chilean market it is still very hard for a new independent
68
producer to dispatch power to the grid, especially considering the high levels of
uncertainty behind the projected marginal costs of electricity. As a consequence, this
commercial risk can only be managed by traditional large power companies or global
investors with solid financial profiles, who distribute the risk across a wide portfolio
and therefore concentrate decisions on new generation capacity.
Other relevant regulation is the modification made to Decree 291 in 2009, requiring
that small-scale generators, with less than 300 MW installed capacity, be included in
the Directory of the Economic Load Dispatching Centre, which improves the
negotiating positions of small-scale generators. They now have direct access to
distributors, instead of depending on larger generators to negotiate dispatch
conditions.
The Clean Development Mechanism of the Kyoto Protocol is another demand-pull
policy applicable to Chilean renewable energy generators. In December 2010, the
country’s CDM portfolio consisted of 68 projects, 36 registered and 32 at validation,
with associated annual emissions reductions of 6 million tons of CO2e (Unep Risoe,
2010). Chile holds 8% of CDM projects and associated emission reductions in Latin
America, and is the third-placed country in the region with the highest participation in
the CDM, after Brazil, with 42% of projects, and Mexico, with 20%.
The Chilean CDM portfolio is dominated by hydroelectricity, landfill gas capture and
biomass energy projects, as shown in Figure 22. GHG emission reductions by CDM
projects are dominated by biogas from landfill gas or manure, hydroelectricity and
biomass, as shown in Figure 22.
FIGURE 22- DISTRIBUTION OF CHILEAN CDM PROJECTS AND GHG EMISSION REDUCTIONS PER
TECHNOLOGY
GHG emissions reductions by CDM projects, Dec 2010
2%
0%
0% 0%
0% Afforestation
11% 13%
Biomass energy
CO2 capture
EE supply side
16%
17% Fossil fuel switch
Hydro
Landfill gas
Methane avoidance
N2O
41% Wind
Source: UNEP Risoe CDM JI pipeline, December 2010

Chile does not produce its own electricity generation technologies, and therefore must
import all equipment into its renewable energy plants and other clean energy projects.
An analysis of TT claims in Chile’s CDM projects registered or under validation up to
June 2010 shows that 52% of the 69 projects analysed claim TT, representing 72% of
the GHG emission reductions by CDM projects. The analysis is based on a database
developed by Seres et al. (2010) at the request of the UNFCCC, which records evidence
of TT by analysing the project design documents of all CDM projects registered or
under validation as of June 2010. Even though all Chilean generation projects use
69
foreign equipment, agriculture, hydroelectricity and biomass-biogas CDM projects do
not tend to claim TT, because these technologies are already widespread throughout
the country. Projects using technologies new to the country, such as wind, landfill gas
capture or N2O destruction, acknowledge more frequently the need of TT. In any case,
TT claims are not based on a common definition of TT among CDM developers, so for
this reason these data are not totally reliable.
Experts interviewed, as well as the relevant literature, argue that the tangled
bureaucracy of the CDM system has meant that, except for landfill gas capture and
manure management projects, it has not acted as a financial incentive for companies
(Dufey, 2010). Landfill gas and biogas from manure capture projects are highly
profitable CDM projects due to the high global warming potential of methane, which
yields a high amount of carbon credits, while biomass and hydroelectricity projects,
the other dominant technology types, are traditional sources of energy that have not
required support in the past in order to be economically feasible.
Interviews with carbon finance expert Fernando Cubillos , with the Technical Director
of the CER Rodrigo García and with Darío Morales from CORFO confirm this
perspective. Fernando Cubillos explains that the carbon market provided a powerful
signal for the promotion of renewable energies in Chile, leading to a boom in
renewable energy projects presented for approval in the SEIA. However, in most of the
cases this signal did not materialise in the actual implementation of projects. The
carbon market did not deliver to expectations due to the highly bureaucratic
registration process and uncertainty about the future prices of carbon credits, given
the financial crisis and the uncertain outcomes of a post-2012 climate regime.
Fernando Cubillos believes that the role of the CDM was side-tracked towards strict
assurance of the additionality of emission reductions, instead of focusing on being a
channel of technology transfer and technological transformation for the recipient
countries. In his experience, all projects registered as CDM had already guaranteed
their financial feasibility and did not need carbon credits to be implemented.
Rodrigo García from the CER also believes that the CDM has had a very limited impact
on technology transfer, and he considers that there is a deficit of effective demand-
pull policies in Chile to encourage investment in renewable energy technologies. The
lack of clear support mechanisms means that technology-intensive foreign investors
are not attracted to the Chilean market. Besides, large companies control the
electricity market and have no interest in investing in risky new technologies when
their current business model is highly profitable. On the other hand, the government
still does not have a clear renewable energy policy that sets priorities or targets
specific technologies. The liberal Chilean Ministry of Economy is not interested in
“picking winners”, and therefore does not want to select the specific niches that
require support for technology transfer; instead, they are more interested in increasing
the total electricity generation capacity of the country at the lowest cost to fuel-
ambitious economic growth targets. A carbon market is seen as a potentially attractive
policy for reducing the carbon intensity of the electricity generation matrix.
70
3.5 Technology supply factors and policies
3.5.1 Human capital

Education statistics show that Chile is well positioned among emerging and developed
economies, as it has made great progress in terms of educational attainment, which is
the highest educational degree that its citizens have obtained. The coverage of primary
education is almost universal and secondary and tertiary attainment rates have
increased rapidly. Figure 23 shows that the educational attainment of Chileans aged
25-34 is now higher than that in emerging economies like Turkey, Mexico or Brazil, or
developed countries like Portugal, and is above the attainment of OECD populations
aged 55-64. However, this improvement comes against the backdrop of a relatively
low level of skills shared between the entire population. Scores of 15-year-olds in
science, reading and mathematics are still well below the OECD average, even
adjusting for the lower socio-economic background of Chilean students, according to
PISA results presented in Figure 24. However, they are ahead of other emerging
economies like Mexico, Turkey or Brazil (OECD, 2010).
FIGURE 23- POPULATION AGED 25-34 AND 55-64 FIGURE 24 - PISA SCIENCE MEAN SCORE
THAT HAS ATTAINED UPPER SECONDARY AND
TERTIARY EDUCATION, 2007 (PERCENTAGE)
Source: OECD, 2010

Education spending per pupil is low at the primary and secondary school levels, behind
all OECD countries except Mexico. However, the share of private spending on
education is the highest among all OECD countries at all levels of education, which
poses equity issues in a country with a highly unequal income distribution. In
particular, the gap in tertiary education between the highest and lowest income
groups is wide and has expanded in the last few years (OECD, 2010).
A relatively high 18% of all new degrees, close to the OECD average, were in science
and engineering disciplines in 2007, which shows the potential for technological
innovation. Some of the efforts made by the Chilean Government to increase the
employability of students are reflected by indicators of school internet access, which is
the highest in Latin America, investments in English language proficiency and digital
literacy (Invest Chile - CORFO).
71
3.5.2 R&D and innovation
According to the OECD, Chile had a global expenditure in research and development
(GERD) intensity of 0.7% in 2008, which is well below the OECD average and that of
emerging countries like China, India, Brazil, South Africa and Turkey. However, it
exceeds that of some OECD countries like Greece, Mexico or the Slovak Republic.
Business expenditure on R&D (BERD), at 0.3%, is also low because of the country’s
economic structure, in which the services sector makes up 64% of GDP, agriculture
15% and a low-technology manufacturing industry, including energy, comprises 23%.
Commodities account for almost three-quarters of total exports and high-tech
manufacturing industries’ participation is irrelevant.
The private sector financed 44% and executed 40% of R&D in 2008 (Ministerio de
Economia, 2010). Despite a relatively large stock of FDI, foreign affiliates of
multinational companies only account for a very small share of total business R&D,
recorded at 3.6% in 2002 compared to 47.9% in Brazil, 32.5% in Mexico and 23.2% in
Argentina (OECD, 2010). Per discipline, the private sector invests mostly in technology
and engineering R&D, while public research and education institutions invest most
resources in natural sciences.
FIGURE 25- GROSS EXPENDITURE ON R&D AS A PERCENTAGE OF GDP, 2008
2008 or latest available year 1999 or first available year

2.50
2.00
1.50
1.00
0.50
0.00
Source: OECD Factbook, 2010
Research capabilities in Chile are also low. According to OECD data, in 2008, Chile had
2.3 researchers per thousand employed, a figure similar to Brazil but far below the
OECD average of 7.6 employed. Chile produced only 185 scientific articles per million
people in 2008, although these had grown by a robust 10% per year since 1998. A
comparatively low 12% of firms introduced new to-market product innovations during
2004-06, and a below average 22% of firms engaged in non-technological innovation.
Chilean universities are active in research about renewable energy applications in the
country, with Universidad de Chile and Universidad de Concepción being mentioned by
the Technical Director of the CER as high performers in renewable energy R&D.
72
FIGURE 26- RESEARCHERS PER THOUSAND EMPLOYED
2008 or latest available year 1999 or first available year

9
Source: OECD factbook 2010

Chile produced 0.34 triadic patent23 families per million people in 2007. This figure is
behind emerging economies like South Africa, China, Brazil or Turkey and very far
behind the OECD average, which stands at 42.3 triadic patent families per million
inhabitants. Patent data show the wide innovation gap between developing and
developed countries and make evident the remaining distance between Chile and
other OECD countries. Both Darío Morales, from CORFO, and Conrad von Igel, from the
Ministry of Economy, indicate that awareness about intellectual property rights is very
low in Chile, particularly as entrepreneurs and scientific researchers do not understand
the benefits of patenting. Innova Chile is initiating a capacity-building programme to
promote the protection of knowledge generated in Chile.
FIGURE 27- TRIADIC PATENT FAMILIES PER MILLION INHABITANTS IN EMERGING COUNTRIES 2007
0.7
0.6
0.5
0.4
0.3
0.2
0.1
Source: OECD factbook 2010

Per industrial sector, mining is a significant source of Chilean technological capabilities,
as it represents 19% of the nation’s GDP and 60% of its exports on average over the
23
Patents that are registered simultaneously in the United States, the European Union and Japan.
73
last five years (SONAMI, 2011; Banco Central, 2010). The budget for the national
innovation system, mainly grants provided to innovative entrepreneurship and local
R&D, comes directly from royalties and production taxes from mining operations.
Furthermore, the mining industry is the main source of local expenditure on R&D, the
fourth highest sector in relation to innovation practices and the main patent owner in
the Chilean economy (21% of mining companies hold patents). The mining sector also
has a strong foreign presence, but significant technological capabilities have developed
among local suppliers of goods and services, who need to operate in a global market
and have incentives to develop world-class technology to increase their
competitiveness.
Some positive indicators about innovation show that during 1990-2005, the number of
patents per million of inhabitants grew from 10 to 40 and was expected to achieve
more than 100 in 2010. In the same period, the number of scientific publications grew
from 1,200 to 3,500. However, recent recommendations by the OECD suggest a need
for stronger links between scientific development and industry needs, as well as a
diversification of the economy and a transition towards higher value-added activities
(OECD and IDB, 2010).
3.5.3 Technology development indexes and scorecards

The OECD science and innovation scorecard for Chile shows its dependence on foreign
funding, with a relatively high percentage of GERD being funded from abroad and a
very high percentage of patents filed with foreign co-inventors. Chile also shows an
above average percentage of firms collaborating on innovation activities. All the other
indicators are below OECD average, although educational attainment and the
percentage of science and engineering degrees are close to OECD levels.
FIGURE 28- SCIENCE AND INNOVATION PROFILE OF CHILE
Source: OECD, 2010, Science, Technology and Industry Outlook 2010. Country notes.
In comparison with other emerging economies, Chile shows better science and
innovation performance than Brazil in all indicators except GERD and BERD intensities.
74
China performs impressively in the percentage of science and engineering degrees, the
percentage of firms with new to-market innovations and high GERD and BERD
intensities in comparison with other emerging economies. India performs
exceptionally well in terms of average annual growth rate (AAGR) of patenting
activities and South Africa presents a good innovation profile for its private businesses.
FIGURE 29- SCIENCE AND INNOVATION PROFILES OF BRAZIL, CHINA, INDIA AND SOUTH AFRICA
Source: OECD, 2010, Science, Technology and Industry Outlook 2010. Country notes.
The 2001 UNCTAD innovation capability index places Chile among the set of countries
with “medium innovation capabilities”, after Argentina but ahead of many emerging
economies. Data from the index is shown in Table 53, in the Annex of Chapter 3;
however, the index is now a bit outdated, as dramatic changes have taken place in
innovation capabilities during the last ten years. The UNDP technology achievement
index placed Chile in 37th position among 74 countries, while it was positioned 41st
among 162 countries in the Arco indicator of technological capabilities. However, both
indexes are also constructed with data more than a decade old.
3.5.4 Technology push policies

Chile has significantly increased the financial resources it uses to support innovation,
with a 74% increase in the innovation budget from 2005 to 2008. However, the budget
for investment in innovation is still far from OECD standards (OECD and IDB, 2010).
A decrease in economic growth in the last decade, mainly as a result of a slowdown in
total factor productivity (TFP), led to a rising consensus about the role of innovation as
a driver for Chilean development. The liberal approach of the Chilean economy has
precluded public intervention to support specific industries, but in the last five years
there has been a transition and innovation has become a legitimised space for public
intervention.
The National Council for Innovation and Competitiveness (CNIC) was created in 2005,
and it contributed to starting up a public policy reflection on innovation with a medium
75
and long-term view, and also elaborated the national strategy for innovation 2006-
2020. The strategy took a broad approach to innovation by identifying three cross-
sectional areas for action: human capital development, mission-oriented science and
business innovation. It also went beyond the traditional horizontal approach with the
selection of sectors with high potential impacts and spillovers in the economy, defined
as “clusters”. The underlying rationale for selecting sectors was that Chile is a small
economy, with limited resources, and therefore could not successfully compete
internationally in all economic sectors. Clusters were selected on the basis of their
potential to be important players in the global market. Eight priority sectors were
selected: fish farming, tourism, copper mining, offshoring, processed food, fruit,
poultry and pig farming. The clusters policy included the following core elements: the
elaboration and implementation of roadmaps to increase the competitiveness of the
sectors; favouring clusters in the public support instruments to business innovation;
supporting research related to the selected industries and attracting foreign capital to
complement national knowledge in the priority sectors. Five cross-sectional platforms
were also identified: human capital; R&D; infrastructure and natural resources;
regulatory, legal and political frameworks and financial services. Based on this strategy,
the Ministerial Committee for Innovation expounded a biannual action plan that
reduced priority clusters to five: the food industry, fish farming, mining, special
interest tourism and global services. It also redefined priority transversal areas as the
environment and water resources, biotechnology, renewable energies and ICT.
Conrad von Igel, in charge of the Innovation Division of the Ministry of Economy, as
part of the new and conservative government, still shows reluctance to “pick winners”,
indicating that the state is not qualified to point at sectors better placed to compete
internationally. He considers that the private sector should be in charge of innovation
and decide by itself in which technologies it wants to place its investments. The
policies that the new government is planning to utilise in order to promote innovation
again follow a horizontal approach and consist mainly of the following:
1. Attracting innovation talent from abroad. They have a target of attracting 1,000
entrepreneurs to the country by 2014, by offering them seed capital, easing the
administrative process of creating a business and granting work visas. A pilot
programme has already attracted 25 entrepreneurs, mostly from the United
States, but including also Chinese and German applicants. Most of the
proposed projects have been related to information technologies, as they
deliver results in the short-term and require low seed capital.
2. Global research networking. Chile is a small economy with limited resources. To
be able to compete globally it either needs to focus on a reduced number of
clusters, as defined in the national innovation strategy, or to spread wider by
creating alliances with the leading global institutions in each subject. The
government believes that it makes more sense to create partnerships between
Chilean research centres and leading foreign research centres.
3. Improve literacy about regarding intellectual property rights. To do so they are
financing Chilean researchers to be trained in United States technology transfer
offices, so that they can bring specific proposals to Chilean universities and
thereby increase transfers of technology from the university to the private
sector.
76
The implementing institutions of Chile’s innovation policy are CONICYT (Science and
Technology Council) and CORFO (Chilean Economic Development Agency). In addition,
the Fund of Innovation and Competitiveness (FIC) was created in 2006 to channel
public income from mining royalties towards financing the innovation policy. In general
terms, two-thirds of the funds used to support and encourage innovation come from
the public sector, while one-third comes from the private sector. CONICYT covers
individual basic research and has more than twice the budget allocated to CORFO for
industrial innovation and applied R&D.
InnovaChile is the implementing agency for CORFO’s innovation policies and supports
Chilean firms to improve their competitiveness in national and international markets
by promoting the development of innovative processes. Energy-related small and
medium size companies, in cooperation with university research centres, are some of
the main beneficiaries. In the 2005-2010 period, InnovaChile supported more than 120
innovation projects on renewable energy, allocating more than U$40 million to
transfer, improve or develop technology. The agency’s support is channeled through
four main programmes addressed at helping four different types of innovating actors:
• The pre-competitive innovation programme supports investments in
technologies that have not yet reached commercial feasibility in Chile. This
programme has financed an inventory of renewable energy resources for
marine and solar technologies, research on the production of biofuels with
microalgae and forestry waste. InnovaChile usually collaborates with
universities to provide technical support.
• The prototype development programme provides subsidies to private business
to develop prototypes.
• The entrepreneurs programme provides seed capital to individual
entrepreneurs.
• The technology transfer programme uses several instruments, including
technology internships, where CORFO partly finances companies to send their
employees abroad to learn about a new technology, specialised consulting by
foreign experts, where companies obtain subsidies to bring foreign experts to
their companies over a specific period of time, technological missions, where a
group of companies visit fairs and companies abroad to learn about the latest
technological developments in a specific industry, and technology diffusion,
where companies in the same industry undertake capacity-building activities
with the support of foreign experts.
Darío Morales, manager of InnovaChile, indicates that they do not assess the
performance of the funds in achieving their final aim, but they are currently designing
a new system of indicators to solve this issue.
CORFO’s high technology programme, also known as INVEST Chile, is one of the
instruments used for attracting FDI that can strengthen the growth capacities of the
Chilean economy and diversify both the production and export base. INVEST Chile
supports foreign companies or foreign majority joint ventures carrying out research
activities or technological development projects that invest in national value to the
tune of over US$500,000. In addition, in order to support to organisations during the
77
administrative process involved in starting up a business in Chile, the programme
provides substantial financial incentives, including (Invest Chile, 2011)24:
• Up to 60% of the pre-investment study cost, to a maximum of US$30,000
• Up to US$ 30,000 for start-up activities
• Up to 50% of annual salaries during the investment phase, with a maximum of
US$25,000 per person
• Up to 40% of the total investment in fixed assets, with a maximum of
US$2million
• Up to 40% of total property lease amount during the first five years, with a
maximum of US$500,000
• Up to 50% of specialised training or recruitment, with a maximum of
US$100,000
Mr von Igel indicates that low private R&D is partly due to previous counterproductive
regulations that penalised R&D expenses by eliminating them from the profit and loss
statement for the calculation of income taxes. This was changed by Law 20.241,
entering into force in 2008, which aimed at improving incentives for innovation in the
private sector by introducing a tax credit for 35% of R&D expenses in projects carried
out by or in collaboration with research centres certified and included in a registry by
CORFO. The government is currently launching a new R&D law in Congress to further
incentivise investment. As regards foreign R&D investment, Conrad von Igel indicates
that it is still quite low and focused mostly in the agro industry, biotechnology and
mining.
Chile lacks a clear RD&D (Research, Development and Deployment) strategy at the
Ministry of Energy (IEA, 2009), which is a significant weakness for local innovation in
energy technologies. The merit of individual projects is the main basis for the approval
of basic research funding. As a result, activities are dispersed, collaboration between
institutions is lacking and research is project-driven rather than linked to the country’s
needs. Renewable energy is not considered a priority sector under the Chilean
innovation strategy; rather, it is considered a transversal technology platform relevant
to all niches of the economy. Using competitiveness as the only measure might be a
narrow goal for future energy R&D priorities, because other factors such as
environmental sustainability and energy security should also be taken into account
(IEA, 2009).
Some institutional developments favouring a renewable energy technology push are
the creation of the Ministry of Energy and the Centre of Renewable Energies (CER) in
2009. The purpose of the CER is to work alongside the CNE and CORFO to accelerate
investment in NCRE and to become a knowledge and technology transfer hub. At its
inception, the CER was expected to have a lead role in promoting renewable energy
technologies in the market and to serve as a clearing house, connecting research
entities and private companies to the international network of renewable energy
technologies. One of the key activities that the CER is planning to implement is a series
of NCRE pilot projects that demonstrate the technical feasibility of specific
technologies in Chile and create precedents that invite large-scale deployment.
According to Rodrigo García, from the CER, to be eligible for funding, candidate pilot
24
http://www.investchile.com/investchile_services/incentives/fixed_assets
78
projects should be commercially replicable, and therefore have a low technical risk.
However, they should also build some gaps and go beyond the business-as-usual
situation in Chile, for example by opening transmission lines for other projects,
showing the feasibility of using a renewable energy resource or proposing new
business models. The pilot projects should also have the potential to create spillovers
in the economy, by encouraging the creation of new businesses that provide required
services or components. The pilot project programme is technology neutral, and no
specific amount is allocated to specific technology types. The CER estimates that each
pilot project will receive up to US$ 2.5 million.
Renewable energy projects can benefit from the general support instruments of
CORFO described in the previous section. CORFO has also set up diverse instruments to
support specific investments in the development of clean energy technologies. The
most important of these are:
• Technical and financial support for pre-feasibility studies for NCRE projects.
Over US$ 400,000 have been use to subsidise 50% of the cost of the study.
• A programme of pre-investment for advanced studies in NCRE co-finances part
of the basic and detailed engineering costs, electricity connection studies and
assessments, and the environmental impact assessment of projects that have
passed a pre-feasibility study. It covers up to 50% of the related costs.
• Low-interest loans for renewable energy investments are co-financed by the
German bank KfW. Loans are designed for refinancing long-term credit and
leasing operations for investments contributing to environmental
improvements, with a maximum loan value of US$ 10,000 per project.
• Capital guarantee and risk capital funds for clean energy and energy efficiency
are provided up to a total of US$7.5 million.
3.6 Industrial development factors

Since the mid-1980s, Chile’s economy has been fuelled by natural resources-based
export markets, with significant concentration on production in a few sectors: mining,
pulp and paper and fruits. High commodity prices allowed a 7.1% average growth rate
in the ‘84-‘98 period; however, since 1998, economic growth rates have decreased
significantly. The contribution of total factor productivity growth (TFP) to GDP growth
accounted for a substantial part of GDP growth from the late 1980s to the mid-1990s
but has been close to zero since 1998. Capital stock growth represents the major
growth component since 1998 (OECD, 2010).
Chile’s income per capita gap with other OECD countries is mostly attributable to lower
labour productivity. Transitory gains in labour productivity from the movement from
agriculture or personal services sectors to higher productivity sectors like
manufacturing or financial services have been nearly exhausted. Future productivity
gains may be enhanced by the reallocation of resources from low productivity firms to
higher productivity industrial firms. Entrepreneurship and innovation can facilitate the
reallocation of factors and the development of new products, the improvement of
production processes and the adoption of new marketing or organisational techniques
(OECD, 2010).
79
The country’s exports are concentrated in primary products (mainly copper) and
resource-based manufacturing. This pattern reflects an approach to industrial policy
that favours static comparative advantages and export-led growth instead of selective
industrial policies for the development of industries with higher value-added
attributes. Being a natural resources abundant country, Chile’s policy has led to a
specialisation in resource-intensive activities in the primary sector and the resource
processing manufacturing subsectors. The sophistication of Chile’s export basket is
lower than would be expected when considering its income per capita, and it also
lower than other resource-rich countries in the OECD like Canada, Australia and New
Zealand (OECD, 2010). Chile’s export vector is also characterised by a low income
elasticity of demand and a production structure suffering from a strong de-linking25
process (Cimoli and Di Maio, 2004).
FIGURE 30- CHILE´S EXPORT COMPOSITION IN 2006
Source: OECD, 2010
An analysis of input-output tables by Cimoli and Di Maio (2004) shows that the
multiplier effect on the domestic production structure, associated with internal
demand growth, decreased between 1986 and 1998. Trade liberalisation policies have
had a substitution effect of domestic producers by foreign ones. The most affected
sectors have been the manufacturing industries, as the requirement for imports per
unit of export has considerably increased in the last decades. Exporting sectors have
gone through a process of modernisation by importing new capital goods. Their
growth has not increased labour income but has been instead used mainly to pay rents
for the use of natural resources. The transformation pattern of Chile’s domestic
industrial structure is therefore characterised by an impoverishment of domestic
25
Cimoli and Di Maio (2004) refer with this term to “the process of weakening of the linkages among the
domestic sectors measured as the decrease of the domestic leontevian multiplier associated with a
unity increase of the domestic demand for a given sector between two points in time”.
80
linkages, which is caused by the replacement of domestic production capacity by
foreign producers (Cimoli and Di Maio, 2004).
In such a dis-linked production structure, technology transfers have a low potential to
generate significant knowledge spillovers, which has implications for low-carbon TT
because higher demand for low-carbon energy will increase imports of foreign
equipment, without creating significant backward and forward links required for the
national economy to absorb the technology. In order to reduce the technological gap
with developed countries, Chile will need to increase the sectoral industrial
interdependence and the technological content of the produced goods (Di Maio,
2004).
An analysis by UNIDO (2009) maps countries of different income levels in terms of
their transactional capacity (TAC) and their transformational capacity (TFC). The TAC is
measured through an index of manufactured exports per capita and manufacturing
export quality, where quality is represented by the share of manufactured exports and
the share of medium- and high-tech exports among total exports. The TFC is measured
through the manufactured value-added per capita and industrialisation intensity, and
it represents the capability of a country to innovate, i.e. to transform (in complex
terms), raw materials, other commodities and intermediate goods into medium- and
high-tech products. Values of the TAC and the TFC indexes are between 0 and 1. The
two dimensions – TAC on the x-axis and TFC on the y axis – are represented visually in
Figures 31 to 33 for the years 1993, 1998 and 2003. The median (0.5) for all countries
of all levels of income for both TAC and TFC is plotted. The arithmetic mean of both
indices respectively for each income group is also plotted. This illustrates the position
of the country with respect to the median of all countries as well as to the group
average – whether it is faring much better than the group average or lagging far
behind – and represents a significant point of reference for policymakers.
The median (0.5) for both TAC and TFC respectively divides the graphs into four
quadrants:
• Quadrant I – ‘Laggards’ – countries in this quadrant have low levels of both TAC
and TFC.
• Quadrant II – ‘Intermediators’ – countries in this quadrant possess high
transactional, intermediation or exporting capacity and low transformational or
value-adding capability.
• Quadrant III – ‘Innovators’ – countries have high TFC, but low TAC.
• Quadrant IV – ‘Innovating Intermediators’ – whose countries possess high
innovativeness which is manifested as transformational value-adding capability
as well as high intermediating capacity.
Chile is positioned in the ‘Laggards’ quadrant in all three years. It is also a laggard as
compared with countries of the same income level, except in 2003, when its TFC
increases above the arithmetic mean of other upper-middle income economies. Chile
is therefore not on a path towards industrial development as shown by other emerging
economies, particularly Mexico and Malaysia. As a result, low-carbon technology
transfers may not permeate the industrial fabric of the country and instead make Chile
dependent on the leading developers of these technologies, either from developed or
81
FIGURE 31- INDUSTRIAL DEVELOPMENT POSITIONING OF UPPER-MIDDLE INCOME ECONOMIES, 1993
Source: UNIDO, 2009
Source: UNIDO, 2009
82
Source: UNIDO, 2009
3.6.1 Industrial policy

Significant empirical evidence supports the view that the existence of a large and
articulated industrial sector seems to be a necessary condition of a steady growth
path. In addition, sectoral domestic linkages developed by the exporting sectors
determine the growth possibilities of a country characterised by an export-led growth
model (Cimoli and Di Maio, 2004). Chile could therefore benefit from the increased
diversification and sophistication of its export basket, leading to higher productivity
goods. However, there are risks associated with actively targeting specific sectors
through public policy, as public resources may be wasted on sectors that turn out to be
unviable or on sectors that would have developed even in the absence of public
support (OECD, 2010).
Chile has been particularly reluctant to target specific industrial sectors through
industrial policy, although in some cases, like the salmon industry, it has done so with a
considerable level of success. In the last few years, the authorities have tried to move
away from this purely horizontal approach to innovation policy and towards a more
vertical approach, under which some sectors are singled out for priority support. With
this aim, the National Innovation Council strategy defines some clusters for priority
support to diversify into more knowledge-based activities, as described in Section 3.6.5
about technology-push policies. Even so, the cluster strategy explicitly rejects the view
that resource-abundant countries should move away from resource-intensive
production irrespective of their comparative advantage. However, the government has
defined priority knowledge-intensive transversal areas, which shows a will to support
knowledge-based industries, notwithstanding Chile’s comparative advantage. Other
examples of targeted support are InvestChile’s grants for high-technology sectors and
lower tax rates for repatriated earnings from knowledge-intensive sectors.
In any case, Mr von Igel insists on the incapacity of the government to “pick-winners”
and the reliance on the private sector to decide which sectors have the highest market
potential. The approach of the new government is therefore to wait for the private
sector to take risks and then provide them with some complementary support. This
83
willingness for the private sector to take risks would indicate that they see a clear
market potential in a specific niche. Mr Garcia, from the CER, has a different
perspective and indicates that a hands-off approach cannot lead to long-term private
investment decisions. Industry support decisions are made ad hoc in the short term,
such as the renewable energy pilot projects or support for microalgae research. The
perspective of Mr Morales, from CORFO, is that the state should only allocate funds to
technologies where there are still not a large number of global competitors. In those
technologies that are still at the earlier stage of development, Chilean companies
would still have an opportunity to create their own niche in the global market. Among
these emergent technologies, Chile should support those where there is already a
significant knowledge base in the country. Such would be the case of biofuels, as Chile
has strong capabilities in the biosciences which have so far been used in the fish
farming, agriculture, forestry and viticulture sectors. An additional criterion to access
public support would be the potential impact of the technology on developing local
capacity and generating new high-skilled jobs.
3.7 Conclusions
Chile’s absolute GHG emissions from energy consumption are not significant in
comparison with those of large emerging economies and many other medium-sized
developing countries. However, Chile´s emissions have considerably increased in the
last few decades and are expected to continue to do so as the country gets closer to
the welfare levels of developed countries. Carbon intensity is also low for developed
countries’ standards and in comparison with large emerging economies like China,
India and South Africa. For these reasons, Chile is not a key actor for the global
reduction of GHG emissions.
The country depends heavily on imported fossil fuels, which constitute 70% of its
energy consumption and makes it vulnerable to external shocks. Chile´s main interest
in renewable sources of energy stems derives from its concern about security of
supply. Abundant local renewable energy sources like hydro and biomass are already
being extensively used and make up 30% of primary energy consumption, or up to 40%
of primary consumption when applying calorific power for a hydroelectricity
equivalent to the national power generation matrix.
Oil derivatives are the main source of final energy consumption, with 57% mostly for
the transport sector, while electricity provides 19% of final energy consumption and
wood provides 18%. Industry and mining consume 29% of the final energy and 66% of
electricity, with the copper mining sector standing out as the highest energy consumer.
The copper industry’s main energy sources are electricity from fossil-fuelled plants
from the SING and the direct combustion of oil derivatives.
The two main electricity systems are the SING, with 4,103 MW, which supplies the
northern zone of the country where the mining industry is located, and the SIC with
12,180 MW, supplying the central region where most of the population is located.
Fossil fuels represent 99.6% of the installed capacity in the SING and 50% in the SIC,
where hydroelectricity has strong participation. So-called Non-Conventional
Renewable Energies (NCRE), which are all types of renewable sources excluding large
hydro-power plants, have a very small participation in the system, representing 3.42%
of the installed capacity in the SIC and 0.4% in the SING in December 2010.
84
The penetration of NCRE technologies has been very slow in Chile, despite a large
renewable energy potential and the interest of the private sector, which has submitted
2,805MW in total of renewable energy projects for their approval by the
Environmental Impact Assessment System, and has presented 77 projects for their
registration as CDMs. The nation has accumulated considerable experience in biomass
and hydroelectricity technologies but still needs to transfer all the equipment from
abroad, as it does not count with a domestic equipment industry. In other technology
types, Chile needs not only foreign equipment but also the knowledge to operate it
and use it to drive organisational change. Technology transfer processes have so far
consisted mostly of imports of equipment for hydro and biomass projects.
Consequently, more profound TT processes are required to drive Chile towards a low-
carbon development path.
An analysis of the enabling environments for TT in Chile provides some explanations
for the low penetration of NCRE technologies and the lack of domestic capabilities to
develop their own low-carbon technologies.
Among the positive aspects of Chile’s enabling environment for TT is its favourable
conditions for foreign investment, as shown by the high share of FDI in national GDP
and the increasing amount of FDI inflows in the latest years. A stable macroeconomic
framework, strong institutions, a liberal economy, open trade, low corruption, the
strength of the property rights regime and ease of doing business attract foreign
investors.
Chilean total energy demand is small in comparison with other Latin American
countries like Brazil, Mexico, Argentina and Venezuela, but it has experienced strong
growth rates. Electricity consumption per capita is higher than any other Latin
American country and has also experienced strong growth. Also, electricity prices are
high for Latin American standards, which should attract foreign investment in
alternative energy technologies.
Regarding technology supply factors, Chile has almost universal primary education and
is increasing secondary and tertiary attainment rates. It also shows a high percentage
of science and engineering degrees, demonstrating the potential for technological
innovation. However, the quality of education is still low in comparison with developed
countries and shows a wide gap in tertiary education between the highest and lowest
income groups. The mining sector is a significant source of technological capabilities
and has created strong links with local suppliers, while the country’s science and
innovation sectors depend on foreign funding and a high percentage of patents are
filed with foreign co-inventors.
These positive aspects have not led to high levels of low-carbon energy technology
transfer, if we look at the low implementation of NCRE projects and the lack of local
renewable energy technologies. FDI has flown into the country, though, but not as
technology-related investment.
Some explanations for the technology demand side come from the liberal character of
the electricity system, which benefits the lowest cost technologies. On the technology
supply side, R&D intensity and business expenditure on R&D are very low. Despite
their high participation in the national GDP, foreign subsidiaries of multinational
companies account for a very small share of R&D. Research capabilities are low, with
85
innovation levels, measured through patents per capita, far behind emerging
economies from BRIC countries. On the other side, Chile lacks a strong industrial sector
that could manufacture its own renewable energy technologies or provide inputs into
foreign suppliers because its economy is based in natural resources-based export
markets with a significant concentration in mining, pulp and paper, fruit and fishing.
Further, static comparative advantages provided by the abundance of resources over
specialisation in higher value-added activities in the industrial sector are favoured.
Domestic manufacturers have been substituted by foreign ones, so that the industrial
fabric is weak and delinked. As a result, foreign technology transfers have a low
potential to generate knowledge spillovers through backward and forward links with
domestic suppliers.
Chile has designed and implemented several policies to improve its enabling
environment for TT. On the demand side, Law 20.257 established a renewable quota
system to achieve a 5% participation of NCRE in electricity sales from 2010, which
would reach 10% by 2024. However, this law has not so far incentivised a significant
growth in NCRE projects due to some flaws in its design and implementation, mainly
the lack of a liquid market for green certificates and the law’s applicability to only one
share of the total electricity generation. Additionally, the high volatility of electricity
prices increases the risks of investing in electricity generation projects and makes it
difficult to obtain credit. The Clean Development Mechanism of the Kyoto Protocol has
supported several low-carbon projects, making Chile the third highest Latin American
country in relation to the number of CDM projects in its portfolio. However, experts
argue that the CDM has not supported “additional” carbon emission reductions,
meaning that the supported projects would have taken place anyway. Interviewed
experts consider that there is a deficit of effective demand-pull policies in existence.
On the technology supply side, Chile is making an effort to improve its productivity,
with a national strategy for innovation and some specific support lines for innovation
in renewable energies. However, it lacks a clear RD&D strategy for the energy sector,
and its economic liberalism makes it reluctant to support specific sectors and
technologies, or to “pick winners”. The general approach of the government is to
concentrate in what they do better, buy from abroad at a competitive price what they
cannot produce and provide some support to the local private sector, but only once
they have taken their own risks. Support for endogenous innovation is therefore made
following an ad hoc approach, which does not contribute to creating hubs of
excellence in specific low-carbon technologies.
86
4 Case studies on climate change technology
transfer to Chile26
4.1. Introduction
4.2. Methodology
4.3. Case studies
4.3.1. Case 1: Fibrovent
4.3.2. Case 2: SQM
4.3.3. Case 3: Lithium Innovation Centre
4.3.4. Case 4: UChile micro-hydro
4.3.5. Case 5: KDM
4.3.6. Case 6: Energia del Sur
4.3.7. Case 7: Electrica Nueva Energia
4.3.8. Case 8: Biomasa Chile
4.3.9. Case 9: MNC-Gen
4.3.10. Case 10: Mainstream
4.4. Discussion
4.5. Conclusion
26
One of the case studies in this chapter is published in Energy Policy, Vol. 39, issue 7, pages 4274-4283.
“The role of technology transfer for the development of a local wind component industry in Chile”.
Authors: Ana Pueyo, Rodrigo Garcia, Maria Mendiluce and Darío Morales.
87
4.1 Introduction
Measuring TT is inherently difficult, as highlighted in Section 2 of this thesis. Climate
change TT data do not exist as such and only some imperfect proxies can be used to
provide an idea of the extent of these transfers. Quantitative studies that define and
measure TT and explain the factors that drive and enable them are therefore limited
by lack of data. Nevertheless, case studies can complement quantitative studies by
showing the actual TT process in its context and by providing a focused, localised,
empirical and qualitative approach to understanding the variables that govern
successful TT.
Case study literature on TT is scarce and has mostly focused on the success stories of
BRIC economies, such as the emergence of leading wind turbine manufacturers in India
and China (Lewis, 2007; Wang, 2010; Zhang et al., 2009; Lewis, 2011), world-leading
Chinese photovoltaic technology (de la Tour et al., 2010) or Brazilian biofuels
production (Hira and de Oliveira, 2009). However, these countries do not represent the
average developing economy, as they offer potential investors the prospect of large
profits and have an advanced domestic technological base that allows for the rapid
adoption of foreign technologies. In addition, their large size allows them to specialise
in a wide range of technologies, so smaller developing countries would struggle to
replicate BRIC countries’ success stories given their smaller bargaining power towards
foreign technology providers and smaller resources, which require selecting niches for
specialisation. This calls for further case study research to identify national policies and
business strategies that have worked or failed in non-BRIC economies.
This thesis contributes to the TT debate by presenting ten new case studies of Chilean
organisations directly involved in TT processes to start clean energy technology
ventures. Selecting Chile is relevant in that it shows the challenges faced by actors in
non-BRIC countries in attracting and successfully absorbing foreign technologies. The
case studies can provide qualitative information about what constitutes successful TT
and what are the key factors that enable them. They highlight the particularities of
low-carbon TT in the context of a relatively small developing economy with limited
resources.
The final aim of the case study analysis is to provide answers to the research questions
posed in Section 1 of this thesis, namely:
1. What constitutes climate change technology transfer, and how does it happen?
The case studies describe the technology transfer processes taking place in Chilean
organisations, as well as the stages and elements of the process and the different
channels through which they can take place.
2. Is it possible to measure technology transfer, and what are the main challenges in
doing so?
The case studies reveal potential methods for measuring TT inputs, channels,
outputs, effects and enabling factors in specific situations.
3. Empirically, what factors enable climate change technology transfer to developing
countries?
88
The case studies show the factors that enabled low-carbon technology transfer to
Chile and their importance for the success or failure of the process. The case
studies also show the barriers that prevented or delayed TT processes. Policy
recommendations are derived to improve the enabling environment for climate
change technology transfer in Chile.
4. How do different developing countries perform in these enabling factors, and which
This chapter focuses on a single country, Chile, which is special due to its sustained
growth, stable macroeconomic conditions, liberal economy and OECD
membership. However, some of the policy recommendations derived from the
analysis of Chile’s enabling factors and barriers to TT can be applied to other
developing countries. These findings complement the results of the literature
review in Chapter 2 of this thesis.
4.2 Methodology
Technology transfer processes are complex, so their understanding requires a holistic
approach; qualitative research methods can provide such a comprehensive view. The
case method is said to be appropriate for “describing, explaining, predicting or
controlling processes associated with a variety of phenomena at the individual, group
and organisational levels” (Woodside and Wilson, 2003). Describing means answering
the questions who, what, when and how (Eisenhardt, 1989; Kidder, 1982), while
explaining involves answering the question why and predicting means anticipating the
future in the light of historic events. Controlling means trying to influence processes
once the factors that have an impact on them have been identified (Woodside and
Wilson, 2003).
The main advantage of case research is that it provides an in-depth analysis of
phenomena in context. One of its main weaknesses is that it is not always possible to
infer theories from case studies that are applicable to an entire population, as
evidence is obtained under very specific circumstances. However, they can help to
refine a theory by adding details to it or by testing its validity in a specific case. Another
difficulty lies in the external validity of the results, as it may not be possible for another
researcher to reproduce a case study (Gagnon, 2010).
Case studies must be founded on systematic procedures that allow for demonstrating
the validity and reliability of the evidence collected. The following steps were followed
to ensure this objective, as suggested by Gagnon (2010):
89
FIGURE 34- STEPS IN CASE STUDY RESEARCH
1.
Assess Objective: Determine whether the case method is
appropriateness relevant and appropriate
and uselfulness
2. Objective: Show that the results are rigorous and

Ensure accuracy of representative and that they correspond to reality
results
Objective: Have a sufficiently elaborate and

3. precise research framework to ensure rigourous
Preparation data collection
Objective: Find enough cases that meet the

4. criteria and make sure the study can be
Case selection conducted within budget and on schedule
Objective: Gather rich, credible raw data, while

5.
observing rules of ethics
Data collection
Objective: Perform a systematic and fruitful

6. analysis of the collected data
Data analysis
7. Objective: Produce tested, plausible theoretical

Data explanations of the phenomenon of interest
interpretation
Objective: Make a contribution to the body of

8.
knowledge on the phenomenon and share it with
Report results
the scientific and professional communities
Source: Adapted from Gagnon (2010)
The approach followed to comply with each of these steps is described in the following
subsections.
4.2.1 Assess appropriateness and usefulness of case study research

According to Gagnon (2010), researchers should be able to respond to the following
questions before applying case study research:
• Can the phenomenon of interest be studied outside its natural setting?
The difficulties in describing and measuring TT, as highlighted in Chapter 2 in the
literature review, show that TT cannot be properly understood without taking into
account the interaction between the technology and the environment in which it is
being introduced.
• Must the study focus on contemporary events?
Most of our case studies focus on contemporary TT processes, which enabled us to
identify the main barriers and drivers of the processes as they were taking place.
However, some of them show TT processes that have already taken place.
• Is the control or manipulation of subjects or events unnecessary?
Control and manipulation of the subjects and events was not necessary or possible.
• Does the phenomenon of interest enjoy an established theoretical base?
90
There is a significant theoretical base for technology transfer, as reviewed in Chapter 2,
inasmuch that it provided the structure for the case study analysis, which tackles the
main themes of climate change technology transfer from the points of view of its
conceptualisation, channels, measurement and enabling factors.
4.2.2 Ensure accuracy of results

The demonstration of accuracy of the results includes two components: reliability and
validity. Reliability relates to the consistency of the observations, meaning the
replicability of the results, whereas validity relates to the connection between the
results and reality.
The following actions were taken to ensure the reliability of data:
• Use of questionnaires with concrete and precise questions, common to all
interviewees. These questionnaires are included in Annex 1 of Chapter 4.
• Safeguard the raw data by recording and transcribing all the interviews. This
ensures that other researchers can check and confirm the accuracy of the
interpretations.
• Confirmation and review of the collected data. The text of the case studies was
sent to the interviewees for their review and confirmation of accuracy.
Interviewees were also contacted after the interview when there were pending
questions that had not been solved in the interview.
• Clarification of the researcher´s position. The researcher’s position is
independent from the phenomenon being studied and it should not influence
the results of the data collection. In some of the case studies, the interviewees
were introduced to the researcher by the Technical Director of Chile’s Centro
de Energias Renovables (CER), which may have caused bias in the responses of
the interviewees if they wanted to influence the behaviour of the policymaker.
To avoid this, the academic character of the research was highlighted when
contacting the research subjects and the researcher became the only point of
contact after the introduction was made.
The following actions were taken to ensure the validity of data:
• Select cases relevant to the research study and representative of the
population. The approach to select the relevant cases is presented in Step 4
about case selection.
• Study several cases. Ten cases were analysed, which enables drawing general
conclusions beyond the specificities of a single case.
4.2.3 Preparation
Preparation for data collection involved several activities:
1. Framing the research question. The main objective of this case study research,
as stated in the introduction to this thesis, is to improve the understanding
about what constitutes TT in the field of climate change, what are the factors
that enable them in developing countries and which policies could be
implemented to reinforce these factors. This broad objective is divided into
four research questions, as stated in the introduction to this chapter. The
questionnaire used to collect data reflects these predefined research questions.
91
2. Choose between a single or multiple-case study. We opted for a multiple-case
study, which would make it possible to draw conclusions from a set of
experiences and reduce the problem of generalisation that affects the case
method as a whole. Theory on case-study research recommends that four to
ten cases are studied (Eisenhardt, 1989). This thesis analyses ten case studies.
3. Determine the main data collection technique and potential data sources. Data
collection involved a combination of documentary analysis and interviews with
the actors to obtain their interpretations of the facts and events. Among the
key actors, we targeted the managers of projects involving TT, as they were
involved directly in the technology adoption processes and negotiation with
foreign providers. Documentation was mostly obtained from public sources,
such as companies’ annual reports and corporate websites, public
presentations, government documents and written press.
4. Identify the target population and establish the case selection criteria. The
target population for our study was made up of foreign or local organisations,
based in Chile and involved in processes of low-carbon technology transfer
from foreign providers. The selected sample shows the diversity of low-carbon
technologies, TT channels and organisations involved in TT processes, and it
also includes technologies at different stages of development, from R&D to
commercialisation, as well as those targeting different renewable energy
resources. The most common types of renewable energy in Chile, namely
biomass and hydro, were included, as well as less mature technology types like
wind and solar. Case studies are also diverse in terms of the size and ownership
of the organisations analysed. Different TT channels are presented, such as
foreign direct investment (FDI), imports, hiring foreign staff or collaboration
with foreign institutions. All case studies were focused on Chile to ensure that
they shared the same set of national enabling factors for TT.
5. Become familiar with the phenomenon of interest. To become familiar with the
phenomenon of interest, the researcher met with experts in the Chilean
renewable energy and technology framework, including Fernando Molina,
Environmental Lawyer at the Chilean law firm Barros y Errazuriz, Fernando
Cubillos, ex-manager of Morgan-Stanley Latin America, Rodrigo García,
Technical Director of the CER, Conrad von Igel, Director of the Department of
Innovation in Chile´s Ministry of Economy, and Darío Morales, manager of the
programme Innova-Chile of CORFO. In addition, a thorough study of the
enabling frameworks for low-carbon TT to Chile (Chapter 3 of this thesis), a
review of the literature on climate change TT (Chapter 2) and an analysis of the
technologies covered in the case studies (In Annex 2 of Chapter 4) were
undertaken, to provide the necessary background for the analysis.
4.2.4 Selecting cases

Taking into consideration the criteria above, a total of 16 potential case studies were
identified. We contacted more companies than required in the expectation that some
of them would not be available to or be interested in participating. Six of the contacts
were made with the intermediation of Rodrigo Garcia, and then two with the
intermediation of Fernando Molina. In one case, a personal contact introduced us to a
potential interviewee, and in another one of the interviewees provided us with the
92
details of another company involved in TT processes. Finally, in five cases we obtained
the company’s details from information included in the project design document (PDD)
of projects that had gone through the CDM registration process. The highest response
rate was achieved when contacts were made through the intermediation of Rodrigo
Garcia, and the lowest when companies were directly contacted without the
intermediation of an actor in the Chilean renewable energies market. Table 11 shows
the details of the companies contacted, including the name of the organisation, its
size, type of ownership (foreign or local), type and stage of maturity of the technology,
main channel of TT, approach used to contact the company and positive or negative
response.
93
TABLE 11- SELECTION OF CASE STUDIES
Stage of maturity Facilitator of first Agree to

Organisation Size Ownership Technology type Main TT channel
of technology contact collaborate
Licenses, foreign staff,
Fibrovent Medium Local, private Wind components Diffusion Rodrigo Garcia, CER yes
imports
Imports, collaboration with
SQM Large Local, private CSP raw materials Deployment Rodrigo Garcia, CER yes
foreign organisations
Universidad de Chile Medium Local, public Micro-Hydro systems R&D Import Rodrigo Garcia, CER yes
Lithium-ion batteries Imports, collaboration with

Universidad de Chile Medium Local, public R&D Rodrigo Garcia, CER yes
for EV foreign organisations
KDM Large Foreign, private Landfill gas Deployment Import Fernando Molina, B&E yes
Wind Diffusion
Solar CSP Deployment
MNC-Gen Large Foreign, private FDI, imports, foreign staff Rodrigo Garcia, CER yes
Biodiesel from algae R&D
Biomass co-firing Commercial
Biomasa Chile Medium Local, private Forestry biomass Commercial Import Rodrigo Garcia, CER yes
Energías del Sur Small Foreign, private Biomass Commercial Imports, foreign staff Personal contact yes
Forestry and
Direct approach by
Eléctrica Nueva Energía Medium Local, private industrial waste Commercial Import yes
email, details in PDD
biomass
Alejandro Keller, from
Mainstream Chile Large Foreign, private Wind Diffusion FDI, foreign staff yes
KDM
Direct approach by
Cristalerias Toro Medium Local, private Wind Commercial N.A. no
Direct approach by
MASISA Medium Local, private Forestry biomass Commercial N.A. no
Celulosa Arauco y Direct approach by
Medium Local, private Forestry biomass Commercial N.A. no
Constitucion email, details in PDD
Direct approach by
Solarpack Medium Foreign, private Solar PV Deployment FDI, imports, foreign staff no
email, details in EIA
Yuki Kashiyama-
Small Foreign, private Biomass micro-algae R&D FDI, imports, foreign staff Fernando Molina, B&E no
microalgas
Note: MNC-GEN is a fictitious name for the company, as the interviewed company representative requested that their real name remains confidential.
94
The resulting sample of ten companies is quite diverse in terms of the type and stage
of maturity of the technologies involved in comparison with the organisation’s size and
ownership. These characteristics are shown in Figure 35. Some companies are involved
in more than one technology and therefore the sum in the figures can be higher than
ten.
FIGURE 35- CASE STUDY CHARACTERISTICS
Technology type Technology stage of maturity
Wind, 3 Commercial Diffusion, 3

Biomass, 3 ,3
Biogas, 1
Solar CSP,
2 Deploymen R&D, 2
Electric t, 3
cars, 1
Hydro, 1
Ownership of the organisation
Foreign, Local,
private, 4 private, 4
Local,
public, 2
4.2.5 Data collection

Data were collected through semi-structured interviews and complemented with
organisational documentation. Details of the interviews are included in Table 12 and
the specific questionnaires are included in Annex 1 of Chapter 4. The interviews
included precise questions but allowed the respondents considerable freedom. The
questionnaire initially included structured questions for which the respondent was
requested to evaluate quantitatively different aspects of the TT process, but these
types of questions were found not to provide reliable information. More interesting
information could be gathered when the respondents were given flexibility in their
responses. Three of the interviews, with representatives of Mainstream, Energía del
Sur and Eléctrica Nueva Energía, were conducted by telephone, while the remaining
seven were face-to-face interviews in the premises of the relevant organisations. All
interviews were recorded and each lasted for approximately one hour.
Reviewed documentation included press releases and project design documents for
CDM projects, environmental impact assessments (EIAs) for renewable energy projects
95
developed by the organisation, as well as annual reports, company webpages and
company presentations.
TABLE 12- INTERVIEW DETAILS
Date of
Interviewee name Position Organisation
interview
Victor Poblete Fibrovent's Director Fibrovent 20-Aug-10
Jose Francisco Technical Director Fibrovent Wind 26-Nov-10
Antunes
Pablo Altimiras Manager SQM 26-Nov-10
Guillermo Jimenez Researcher, professor Universidad de Chile 02-Dec-10
Jaime Alée Manager Lithium Innovation Center Universidad de Chile 02-Dec-10
Alejandro Keller Manager KDM 03-Dec-10
Antoine Ferrant Manager MNC-GEN 30-Nov-10
Alonso Gomez Manager Biomasa Chile 25-Nov-10
Bernhard Partner and Director Energías del Sur 29-Nov-10
José Bertrán Partner and Director Eléctrica Nueva Energía 29-Nov-10
Jose Ignacio Partner and Director Mainstream Chile 10-Dec-10
Escobar
Note: MNC-GEN is a fictitious name for the company, as the interviewed company representative
requested that their real name remains confidential.
4.2.6 Data analysis

The first step for data analysis was the transcription of all recorded interviews.
Evidence collected through the interviews showed a common pattern for the
homogenous description of each case study, which were therefore structured around
the following elements of the technology transfer process:
• Introduction to the case study
• Local inputs into the TT process
• International inputs into the TT process
• Channels for the transfer of foreign inputs
• Output of the TT
• Effects of the TT process
Enabling factors operated to facilitate each of these elements of the TT transfer and
included the economic and institutional frameworks of Chile, technology demand
factors, technology supply factors, industrial development factors and supportive
public policies.
The case study descriptions were submitted to the informants, who were then asked
to comment on whether the document accurately reflected reality. They were
corrected according to their comments.
4.2.7 Data interpretation

The data interpretation stage involved compiling the evidence from the 10 case studies
to provide answers to our research questions and eventually apply a theoretical
approach to interpret evidence. In order to define theories from the evidence
collected, the particularities of each case study needed to be considered at a higher
level of abstraction.
96
4.3 Case studies
In every process of technology transfer analysed through the selected
selected case studies we
find a common pattern with the following elements:
• A final, direct objective of the technology transfer, defined as the “output” of
the TT process.
• Technological “inputs” to achieve the desired output. They refer to the inputs
for the specific
pecific process of technology transfer and are a combination of local
and foreign inputs.
• Technology transfer “channels”,
“channels” set up to make foreign inputs available locally.
• Technology effects, beyond the production of an output.
output
• Enabling factors and policies
policie operate at all the levels of the technology transfer
process through enabling the supply of local and foreign technology inputs, the
transfer of foreign inputs through TT channels, the materialisation of local and
foreign inputs into to technological outputss and the achievement of further
knowledge spillovers benefitting the recipient economy.
Figure 36 illustrates the framework of analysis that will be used to describe each of the
case studies. When some of the elements or enabling frameworks of the TT process
are not present or have not yet been achieved,
achieved they are represented in a text box
framed with a dotted line.
FIGURE 36- FRAMEWORK OF ANALYSIS
ANAL FOR CASE STUDIES OF TECHNOLOGY TRANSFER
SFER
97
4.3.1 CASE 1- Fibrovent: developing an indigenous wind component industry in Chile
Introduction to the case study
Fibrovent Wind27 (FW) is a Chilean start-up company that produces wind blades. FW
was created as a partnership between the Spanish wind turbine provider Eozen and
the Chilean company Fibrovent. The partnership had been initially conceived as a joint
venture but has subsequently developed into a purely commercial relationship where
Eozen sell Fibrovent blade moulds and blade design licences.
Fibrovent is part of the SAME Group, a Chilean holding covering a wide range of
services for the mining industry. The SAME group was founded in 1982 to provide
industrial ventilation systems in Chile, and in 1992 it developed its first project for the
mining industry. Fibrovent was created in 2002 to provide the mining industry with
composite materials technologies for ventilation systems, and it is now the leading
Chilean supplier of these technologies and has gained international reputation after
winning projects in South Africa and Laos. In 2009, Fibrovent adopted the strategic
goal of becoming the first Chilean wind blade manufacturer.
Eozen is a Spanish wind turbine manufacturer founded in 1999. It is a licensee of
Vensys wind turbine technology, which uses direct drive permanent magnet
synchronous turbines with no gear box, converting 100% of their machine power.
Eozen has developed and certified its own blade technology in a specially designed
way to optimise the output of Vensys wind turbines. In 2007, Eozen finalised the
construction of its production plant in Spain but was soon hit by the contraction of the
Spanish wind market, after which it looked for opportunities abroad to sell its wind
turbines. In April 2010, Eozen announced a redundancy plan for its 90 employees and
in January 2011 new shareholders entered the company, providing 31.33% of equity to
avoid the collapse of the company (Alimarket, 2011)28.
Figure 37 shows schematically Fibrovent’s TT process. The information included in the
figure was retrieved through interviews with Fibrovent representatives. The following
subsections explain in more detail each element of the TT process as follows: inputs,
channels, outputs, effects and enablers.
27
Fibrovent Wind is not the real name of the new company. A hypothetical name has been created by
the researchers because the final name of the wind blades manufacturer had not yet been decided at
the date of writing this thesis.
28
“Los directivos de la eólica en Andalucía al rescate de Eozen”, ALIMARKET Fabricación de Equipos 14
de Enero de 2011.
98
FIGURE 37- FIBROVENT´S TECHNOLOGY TRANSFER PROCESS
Local technology inputs and their enablers

Fibrovent is an experienced industrial manufacturer of ventilation equipment for the
mining sector. Having gained significant manufacturing experience with composites,
the company searched for business growth opportunities through new applications of
this material. After attending an international congress on renewable energies, the
company’s CEO was convinced that business opportunities in the wind market were
plentiful and that Fibrovent had the capability to start this new venture. The main
technological inputs that Fibrovent could provide were:
• Know-how on industrial production with composites. Fibrovent is a pioneering
company in the application of state-of-the-art injection technologies for the
composite materials industry in Chile. It has successfully delivered products to
the national mining industry since 2002 and has been exporting its products for
five years, which has contributed to building its international reputation.
Previous experience in the mining industry has provided Fibrovent with a well-
functioning, trusted, flexible and low-cost supply chain. The company regularly
undertakes training sessions for local suppliers to ensure the quality of
procured goods. As a distributor of several foreign suppliers, it can ensure low-
cost materials. The steep learning curve that Fibrovent has gone through to
achieve excellence in the mining industry can be now capitalised into the new
wind blades business, which requires a strong level of precision and expertise.
• Highly skilled professionals, with the number of engineers growing from one to
ten since the company was born.
• Profound internal learning processes to penetrate a new market, including
expanding their suppliers’ networks, getting to know and influence the Chilean
99
renewable energy market, learning about wind generation technologies and
networking with foreign providers of wind blades technology.
• Specific wind technology R&D in its own laboratories as well as in partnerships
with universities. The aims of these partnerships are to develop patents jointly,
to create spin-offs as a result of successful research projects and to provide
research and testing services to other institutions. One of the ongoing
partnerships with the Chilean Universidad de La Frontera is currently working
on the design of a wind turbine for low wind speed in isolated regions.
The enabling factors that have made these inputs available are, mainly:
• A solid mining sector. Fibrovent was born as a supplier of ancillary products and
services to Chile´s mining sector and could as such become an expert in
composite materials. A mature industry of suppliers for the mining sector has
enabled cross-sectoral technology transfer for the development of wind blade
manufacturing.
• Availability and quality of Chilean engineers. The good education levels of
Chilean engineers were mentioned as an important enabler. However, specific
capabilities on wind technology are not available in Chile, which was
considered a barrier.
• A national innovation system including qualified universities to undertake joint
R&D projects
Foreign technology inputs and enablers
In spite of its good acumen in industrial manufacture with composites, Fibrovent
lacked specific technical and market experience in the wind industry. Knowledge of
wind blade production was not available in Chile, and therefore foreign technological
support was required. The following foreign technological inputs were particularly
needed for the success of the new business:
• Know-how for the design and construction of a wind blades production plant.
• Equipment for the production of wind blades and know-how to select the right
providers.
• Know-how for the efficient operation of the plant and the equipment.
• Know-how about the wind energy industry.
• Certified blades design, which is a precondition for commercial relationships
with potential customers.
Channels of foreign technology transfer
The following channels were used to structure the required foreign technology
transfers, with the first two in the list considered the most significant:
• Incorporation of a Brazilian expert with more than 20 years’ experience in the
manufacture of wind blades and generators, gained in the Brazilian affiliate of
the German company ENERCON. The Brazilian expert joined Fibrovent in May
2009, initially through a specialised consultancy assignment financed by the
public programme “InnovaChile” and subsequently on a permanent basis as a
partner of FW. The role of this expert has been key for the technical and
organisational design of the new company, the selection of technology
100
providers, the definition of a business plan and capacity building among local
staff. The permanent contact of the expert with Fibrovent’s staff has enabled
profound TT involving technological capabilities acquired through hands-on
experience in the Brazilian subsidiary of a German wind turbine manufacturer.
Fibrovent has gone through a process of internal learning assisted by this
foreign expert. The fact that this constituted a south-south relationship seems
to have eased the transfer process, with recipient and provider acknowledging
a high degree of affinity.
• Licences for wind blade designs and blade moulds bought to the Spanish
company Eozen. Although Fibrovent and Eozen had initially agreed to
constitute a joint venture for the production of wind blades, the technology
transfer was subsequently conceived as a client-supplier relationship.
• Imports of foreign capital goods. Equipment for wind blade manufacture was
sourced from a German company. The selection was made on the basis of
previous experience from the Brazilian expert and the Spanish company Eozen.
Raw materials for composites are also imported from Germany, China and the
United States.
• Participation in several international congresses. Fibrovent´s staff are
continuously involved in international events to keep up to date with
technological and market developments in the wind industry.
Although it did not ultimately take place, the channel initially conceived for the TT was
a joint venture between Fibrovent and Eozen that would give the Spanish company
50% of FW’s shares and 45% to Fibrovent, with the remaining 5% allocated to the
Brazilian partner. A pre-agreement was made in November 2009 defining that FW
would commercialise Eozen generators and use Eozen’s licensed wind blade design at
no cost. The initial target market would be Chile, but FW would expect to resume
exports to the rest of Latin America and beyond after successful deployment in Chile.
The new company would be partly financed by public subsidies and partly by a private
loan endorsed by the SAME group. Negotiations to conclude the joint venture stalled
due to the financial weakness of Eozen and to the high uncertainty regarding Chile’s
demand for wind blades. The TT was subsequently conceived as an arm’s length
supplier-client relationship. Such an arrangement considerably reduces the risk for
Eozen as well as the potential gains, and it provides Fibrovent with more freedom to
decide the structure of the new company and its potential market. Geographical
restrictions for the exports of wind blades were among the critical issues under
discussion in the joint venture agreement, but have been eliminated from the current
negotiation.
Eozen was one of the very few foreign technology providers that showed some
interest in starting a wind blade production venture in Chile. All other potential
partners were interested in penetrating larger markets such as the US, China, India or
Brazil. In other cases, wind turbine manufacturers were simply not interested in
sharing their knowledge. The choice of Eozen as technology partner was motivated by
its openness to collaboration and by technological affinity with Fibrovent’s in-house
expertise. After the joint venture negotiations were interrupted and replaced by a
commercial relationship, more foreign parties became interested. Fibrovent reports
101
that now it does not fully depend on Eozen, as German and Dutch companies would
also be willing to sell them wind blade moulds and licences to produce in Chile. This
increased Fibrovent’s bargaining power in the TT negotiation.
The following factors enabled the technology transfer through the mentioned
channels:
• Public funds for the consultancy appointment of the Brazilian expert, through
the instrument InnovaChile, managed by CORFO. This external consultant later
became a key member of the new business activity.
• Public funds for a large share of the initial equipment and human capital costs,
through the public instrument InvestChile, also managed by CORFO. InvestChile
is considered the best available support instrument in Chile for starting
technological businesses requiring TT. It aims to attract highly specialised
foreign investors by providing the most generous subsidies, faster and more
flexibly than for instruments aimed at local innovation. One of the
requirements for accessing these funds is the majority participation of a foreign
company. For this reason, Fibrovent initially decided to structure the TT as a
joint venture with a foreign majority.
• Low country risk profile of Chile, as a result of its stable macroeconomic
conditions, minimal corruption and a liberal economy that welcomes foreign
investors. The decline of Spanish demand forced Eozen to look for fast growth
markets abroad, and Latin America is a natural expansion region for Spanish
companies.
• Low income tax enabled competitive remuneration for the Brazilian expert.
• Free trade agreements lower the costs of wind blades exports and equipment
imports.
The technology-push policies implemented by CORFO played a relevant role in
ensuring access to foreign access. However, the requirement of foreign majority
participation to access InvestChile’s funds precluded Eozen and Fibrovent from
choosing the most appropriate TT channel. Given the high capabilities of Fibrovent, the
weak financial situation of Eozen, the lack of any other foreign companies interested in
constituting a joint venture and the uncertain demand for wind technology in Chile, an
arm’s length relationship presented a number of advantages over a joint venture, as it
increased the bargaining power of Fibrovent and reduced the risk for Eozen. Joint
venture negotiations were eventually stopped and replaced by a purely commercial
relationship but this will not have severe implications for FW’s access to InvestChile’s
funds, because the instrument is currently being modified to allow the participation of
national companies independently from foreign investors. A second significant barrier
to the project has been the unstable financial position of Eozen, the Spanish partner.
Outputs and enablers
FW plans to replicate European processes for the manufacture of blades for 1.5MW
and 2.5MW wind turbines in Chile. The company expected to start operations in 2011,
producing 150 blades per year to supply a wind capacity of 75-100 MW. A second
production plant would start operating in 2012, producing 300 blades per year for a
wind capacity of 150-200 MW, and would be expanded in 2013 to produce 600 blades
per year for a 300-400 MW wind capacity. At the date of the latest contact with the
102
company representatives (April 2011), the production plant had still not started
operations, the main reason for which was the lack of a significant demand for wind
energy in Chile. Uncertain local demand also created difficulties in the relationship
with the foreign partner Eozen, who became more protective of its technology.
Chile currently has 172 MW of wind capacity and only needs to reach 202 MW by 2020
to meet the renewable energy targets set by Law 20.257. A significant increase in
these targets, or an improvement in the implementation of the legislation, would be
required to achieve commercial feasibility. Although FW´s business plan relies on
exporting a large share of its production, Chilean demand is key for the company to be
able to grow and compete internationally, as access to international markets depends
on credibility gained through experience; consequently, FW blades need to be
operating in a number of wind power projects before they can bid for international
projects. Demonstration projects will be much easier to get in the Chilean market.
Besides, Fibrovent needed foreign partners that could provide the technological
knowledge base. While potential foreign partners were not interested in creating an
additional global competitor, they were attracted by the possibility of penetrating a
new market. The existence of a minimum local market size, as well as the possibility of
entering the Latin American market, was a prerequisite for Eozen to enter into
negotiations for joint blade production. International partners expect that the new
products will cover the Chilean demand for blades, before other international suppliers
can access this market. Therefore, although a small potential demand, Chile was
considered an interesting and reliable market to start accessing Latin America.
FIGURE 38 – FIBROVENT´S WIND BLADE PRODUCTION PLANT PROJECT
Source: Fibrovent, 2010
The interviewees believe that a more effective implementation of the renewable

energy quota, as well as policies providing price stability, will boost wind energy
demand. Price stabilisation rather than subsidies through feed-in tariffs is more likely
to work in Chile, given the high price of Chilean electricity and its free market
economy. Fibrovent is actively involved in ACERA, a lobby requesting regulatory
amendments that extend the demand of renewable energy in Chile. The CDM is not
considered relevant for providing incentives to boost local demand. Fibrovent did not
factor in their business plan the developments of the CDM in the international climate
change negotiations, as they did not consider this instrument to have a significant
impact on their activities.
103
Effects and enablers
The project has so far enabled knowledge spillovers beyond the company involved, to
local universities with whom Fibrovent is undertaking collaborative R&D on wind
technology. Once the plant starts operations and the first Chilean blades can be
installed, further technology effects are expected, including:
• Technology cost reductions. Company representatives estimate that the wind
blades produced would be around 40% cheaper than European blades and 20%
cheaper than Brazilian blades. This will be enabled by the low cost of labour,
low taxes and low tariffs in comparison to Europe and Brazil. In addition,
production for local use would avoid the high transport costs of imported
blades. As the company accumulates production experience, costs would be
further reduced by learning effects. The lower costs of the technology would
enable its wider diffusion.
• Emission reductions. Once wind blades are part of wind turbines implemented
in Chilean wind parks, they will enable emission reductions by replacing
electricity produced from fossil fuel-based plants.
• National productivity gains. This new industrial activity can contribute to Chile’s
upgrade from low-productivity, natural resources-based industries, to higher-
productivity manufacturing industries.
4.3.2 CASE 2- SQM: the challenge of adding value to a natural resources-based industry
Sociedad Quimica y Minera de Chile S.A. (SQM) is a Chilean company operating in the
non-metal mining and chemistry sectors. Its five main business lines are specialty plant
nutrition, iodine and derivatives, lithium and derivatives, potassium chloride and
industrial chemicals. Presently, SQM is the world’s largest producer in the niche
markets of specialty plant nutrition, iodine and lithium, and it is directly involved in the
low-carbon energy business as a key supplier of raw materials such as sodium and
potassium nitrates for the thermal storage systems of CSP plants and lithium for
electric car batteries. This case study focuses on salts for heat storage in CSP plants.
CSP plants with storage systems need sodium nitrate and potassium nitrate to absorb
heat that can then be released to generate electricity when the sun is not shining.
Before they can be fed to CSP storage systems, sodium and potassium nitrates are
mixed in a specific proportion and melted in high-temperature melting furnaces.
Special pumps are used to feed molten salts into the storage tanks. After the system
has started operation, the salt is heated up during the day – in addition to the normal
generation of electricity – to around 400°C. Heat is then taken from the storage tanks
during the hours without sunlight and fed into a steam generator. The salt melt which
as a result is cooled to around 300°C is heated again to 400°C during hours with
sunlight (Durferrit, 2010). The heat storage in the salt bath enables the 24-hour supply
of electricity, therefore reducing the inherent variability of solar energy. Molten salts
are also being used as heat transfer fluid instead of synthetic oil, which increases the
efficiency of CSP plants.
104
FIGURE 39 - MOLTEN SALTS PROCESSING SYSTEM
Source: Durferrit, 2010
The results of the interview are summarised in Figure 40. They will be further analysed
in the following subsections.
FIGURE 40- SQM TECHNOLOGY TRANSFER PROCESS
Local technology inputs and enablers

SQM became involved in the CSP industry in 1995, when it was contacted by the US
Department of Energy (DoE) and the US-based Sandia National Laboratories to provide
salts for their first CSP plant prototype developed in the so-called “Solar Two” project.
It has since then participated in a number of major CSP projects in Spain and the USA.
The main technological inputs that SQM provides for CSP plants with heat storage are:
105
• Access to the unique natural resources of “caliche” ores and salt flat brines in
the Atacama Desert. These resources are unique in the world due to their ease
of extraction and chemical composition29. They constitute SQM´s main source
of competitive advantage.
• Know-how for the mining and production of sodium and potassium nitrates.
Previous to its use for heat storage, SQM had long been producing these salts
for the fertiliser industry, so participation in the solar industry did not require
structural changes to its production plants, even though some new processes
were added and others were modified to increase the purity of the final
product, as required by the solar industry.
• Long experience in the mining sector. SQM is one of the most emblematic non-
metal mining companies in Chile. It was founded in 1968 to organise the
Chilean nitrates industry and has since then become a global leader in several
business areas. In addition, the company has gained international experience
through exports and participation in projects abroad, which has exposed it to
new uses of materials already produced for other industries, as happened with
the salts used for CSP thermal storage.
• Highly qualified employees, including many chemists and doctors in chemistry,
as well as engineers.
• Internal R&D activities and several patented chemical treatment processes.
SQM undertakes specific R&D on the properties of sodium nitrate and
potassium nitrate at high temperatures, relevant to their use as molten salts for
storage in CSP plants. It is also conducting measurements on solar radiation in
the north of Chile, in collaboration with DICTUC, a spin-off of the Catholic
University of Chile. DICTUC will design and install the measurement equipment
and train SQM staff to operate it, and will interpret and audit the obtained data
on a monthly basis. This will help SQM decide about the feasibility and the
technical characteristics of a future CSP plant.
The main enabling factors for SQM´s technology inputs are Chile´s rich mining
resources, a well-established mining sector and the quality and availability of chemists
and engineers.
International technology inputs
SQM’s current solar business has not required significant foreign technology inputs.
The products supplied for the renewable energy market are similar to those already
supplied for agriculture and other industries, with the addition of some processes to
increase the purity of the materials and to deliver them under the required
specifications. Most of the effort exerted to fit SQM´s processes to the new
specifications has been made in-house, on the basis of its long experience in mining
and processing sodium and potassium nitrates, as well as dealing with international
customers.
In any case, technology transfer has been required to start this new product line and to
devise new ways to add value to their natural resources provision business. To be able
to provide salts for heat storage in CSP plants, SQM required the following foreign
technological inputs not available in Chile:
29
SQM Annual Report, 2009.
106
• Information and know-how about the characteristics of heat storage tanks and
the specifications that the salts should meet to perform at high temperatures.
Several tests needed to be performed in collaboration with foreign researchers
to meet product specifications.
• Information and know-how about the operation of CSP plants to design the
most efficient supply system.
• Equipment. Chile does not produce the necessary equipment for salt
processing, which therefore needs to be imported. The equipment required for
the production of salts for CSP plants was already within the company, as it is
similar to that used for salts sold to the fertilising industry.
SQM is considering undertaking higher value-added activities, one step ahead of the
supply of raw materials in the CSP technology value chain. For this, it will require the
following international technology input:
• Know-how about salts melting and feeding services for CSP plants.
• Capital goods and equipment for mixing, melting and feeding salts into storage
tanks.
Additionally, SQM is considering developing a CSP plant on its own land, once technical
and financial barriers are overcome. For this project, the company will need foreign
equipment and know-how about the development, operation and maintenance of CSP
plants.
Channels of international technology inputs
Foreign technological inputs have been mainly channelled through:
• Imports of foreign equipment. Most of the equipment is already available as
part of their fertiliser business line.
• Collaborative R&D with foreign institutions. SQM started its involvement in CSP
technology in collaboration with the US Sandia National Laboratory and the
DoE. It still collaborates with the US DoE in an R&D project, “Advanced Heat
Transfer Fluid Development”, where SQM performs physical property and
corrosion studies on molten salt heat transfer fluids (HTFs). SQM has also
sought the advice of the NREL before installing radiation measuring stations on
its land, with the prospect of developing a CSP plant in the future, if a number
of enabling conditions are met. Additionally, the organisation participates in a
European Union research initiative on the physical properties of molten salt
HTFs.
• Learning by exporting. SQM has gone through an intense learning process to fit
its products to international clients’ requirements. Learning has also taken
place through contact with other supplying companies. For example, SQM has
gone through a process of joint learning with providers of storage tanks to
ensure that metals in the tank will resist the chemical processes resulting from
salt evaporation.
Through its experience in CSP projects in Spain, SQM has learnt about the possibility of
adding value to its activities by providing salt mixing, melting and feeding. These
services would require collaboration with foreign suppliers, as SQM currently lacks the
know-how and equipment.
107
Technology outputs and enablers
Currently SQM only participates in the lower level of the CSP value chain as a provider
of raw materials, namely sodium and potassium nitrates. SQM produces around 70%
of the world’s sodium nitrate and potassium nitrate required for heat storage tanks in
CSP plants. It has participated in many successful projects in Spain (Andasol I y II,
Manchasol, La Dehesa, Valle, etc.), at around 50 MW, and in the US, where it will soon
provide salts for a very large CSP plant.
SQM delivers sodium and potassium nitrates as separate products in a compact, solid
form and packaged to avoid contamination during transport. It has also created
intelligent systems to feed CSP plants and has advised clients about the salt melting
processes.
SQM is looking for new ways to increase the value-added aspects of its renewable
energy business, but in an incremental way and without losing focus on its core mining
and chemistry businesses. The next activities in the CSP value chain would be blending
sodium and potassium nitrates to provide a single mix, melting the salts and then
feeding them to the storage tanks. These activities are usually undertaken by other
suppliers with whom SQM is considering collaborating in order to provide the
complete package of molten salts provision to CSP plant developers.
Venturing into higher value-added activities, such as the production of components or
the development and operation of CSP plants, could yield larger profits, but SQM does
not want to leapfrog to activities that are not related to its core business. The energy
industry is seen as complex and with high barriers of entry, while high uncertainty
about the technical and financial feasibility of CSP technologies in Chile also prevents
them from considering a higher value-added involvement in the renewables industry.
However, they do not discard the future development of a CSP plant to supply
electricity to its facilities, or a partnership with a plant developer and operator to
install a plant on their land in the north of Chile. Before attempting this step, they
expect cost reductions as a result of mass deployment, as well as clear signals from the
Chilean government favouring this technology and lower uncertainty as regards the
techno-economic feasibility of CSP power plants in the country.
SQM´s participation in the renewable energies business is driven purely by demand
pull policies in Spain and the US. Chile has a huge potential for CSP technology,
particularly in its northern region, where SQM carries out its activities, because the
area has one of the highest solar radiation levels in the world and a large energy
demand from mining companies. However, the lack of public policies to support these
technologies prevents their development until significant cost reductions have been
achieved elsewhere in the world.
Technology effects and enablers
SQM’s supply of salts for CSP plants does not have significant technological effects for
Chile, as emission reductions are achieved abroad and Chile does not directly benefit
from the knowledge acquired through participation in these international projects.
Knowledge spillovers are difficult to achieve because Chile lacks a manufacturing
sector capable of undertaking the next activities in the CSP technology value chain.
SQM has needed to collaborate with foreign companies for most of the technology-
108
intensive stages involved in supplying molten salts, and it has also imported all
equipment for its activity.
Participation in the next stages of the CSP value chain could bring further technology
effects, as it would allow SQM to upgrade to higher value-added activity, with
subsequent productivity gains, and then experience in these activities would enable
cost reductions through learning effects. Besides, if CSP projects where implemented
in Chile, SQM´s participation could also lead to emission reductions for the country.
The main enabling factors required to bring about these further technological effects
are:
• Industrial policies that signal renewable energy technology development
priorities.
• Demand-pull policies that allow CSPs to compete with fossil fuel electricity
generation. The government has made some weak attempts to support this
technology, such as the announcement in early 2010 of their intention to open
a bid for a pilot CSP plant. The bidding process has been delayed, which has
increased uncertainty in the market. SQM considers that a functioning CSP
plant in Chile would provide the right signals to investors and finance providers.
4.3.3 CASE 3- The Lithium Innovation Centre: Increasing Chile’s share of a growing clean
industry
The Lithium Innovation Centre (LIC) was created in October 2010 to undertake applied
research on the use of lithium for electricity storage. It is a public-private initiative led
by the Energy Centre at the public University of Chile (UChile) and co-founded with
three private companies: Chemetall (Germany), SQM (Chile) and Marubeni (Japan).
SQM and Chemetall are world leaders in lithium production. We have just discussed
SQM, but Chemetall is a Germany-headquartered company owned by the US-based
Rockwood Holdings, a global specialty chemicals and advanced materials company.
Chemetall’s two major businesses are surface treatment and fine chemicals, including
lithium, special metals and metal sulphides, and it has several lithium production
facilities in the US, Taiwan and Chile (Chemetall, 2009). Sociedad Chilena del Litio (SCL)
is a subsidiary of Chemetall in charge of operations in Chile. Most of Chemetall’s
lithium is produced in Chile, where SCL’s activities include extraction from the Atacama
Salt Flat, purification and crystallisation. Chemetall operates through a decentralised
organisation that gives each local company the freedom to respond quickly and flexibly
to local conditions.
Marubeni is a general trading Japanese company involved in the handling of products
and the provision of services in a broad range of sectors. It has approximately 30,000
employees and has been selected as a global Fortune 500 company. It conducts
business throughout the world together with many partners and customers, making
use of a network of 118 overseas branches and offices in 71 countries30. Marubeni´s
management style is bottom-up, with local subsidiaries having a high degree of
freedom to find business opportunities and undertake R&D activities in their territory.
30
www.marubeni.com
109
Marubeni’s operations in Chile focus on three main business areas: copper mining,
water management, and the official distribution of Nissan automobiles (Marubeni,
2010). The latter activity links Marubeni to the lithium business, as Nissan is actively
involved in developing and commercialising electric vehicles (EV) with lithium-ion
batteries. Additionally, Marubeni has invested in the EV business by participating in a
venture initiated by Japanese Keio University jointly with four other companies to
provide the SIM-Drive technology for EV and cooperate with a wide range of
companies that manufacture EV and electric vehicle parts (Bloomberg, 2009).
Marubeni has also entered a partnership with Spain's Endesa to produce rapid
chargers for EV. The two companies will jointly develop systems that employ chargers
manufactured by France's SGTE Power, with whom Marubeni has a sales agency
agreement31.
The LIC is not a closed organisation. Co-founders have provided the seed capital but
the aim of the Centre is to incorporate more private partners and researchers. With
this aim, the first months of operation of the Centre will be spent visiting a number of
world corporations and research Centres, mainly Japanese such as Nissan, NEC, NTT
and Keio University. Members of the LIC are also building contacts with other
departments of UChile and with Universidad de Antofagasta, located in the North of
the country, close to the Atacama salt flats.
The results of the interview to the Director of the LIC are summarised in Figure 41 and
further explained in the following subsections.
FIGURE 41- -THE LITHIUM INNOVATION CENTER TECHNOLOGY TRANSFER PROCESS
Local inputs to the technology transfer process
31
Automated Trader, Tuesday, 6 July 2010
110
The Lithium Innovation Centre was launched by professors at the Energy Centre of
UChile, who observed the strong growth of lithium demand and the limited
participation of Chile in the lithium ion (Li-ion) battery business for EV. The main
technological inputs provided locally are:
• Access to lithium resources. SQM and Chemetall concentrate 60% of the
world’s lithium production and 80% of the lithium used in batteries. Most of
their lithium production comes from the Atacama region’s salt flats.
• Skilled human resources. The LIC has been mostly driven by enthusiastic and
highly skilled researchers in UChile, bringing in a significant network of Chilean
lithium companies and foreign institutions active in EV R&D. The Director has
extensive experience as a manager in the Japanese multinational NEC. This
experience has enabled networking with Japanese technology providers, and
provided skills to negotiate TT deals with this country. Beyond the Energy
Centre of UChile, the LIC also counts on other departments with significant
expertise in mining, chemical engineering and electrical engineering.
• R&D infrastructure and capabilities. UChile can provide a scientific global
network, labs and research groups to the centre (Universidad de Chile, 2010).
• R&D about clean energies in UChile has so far focused mainly on mature
technologies such as hydro or biomass. The centre has undertaken some
specific research about EV in collaboration with Marubeni, but not at a very
high profile level. The first EV R&D project consisted of a techno-economic
feasibility study of massive EV penetration in Chile, while the second project
involved technological research on multi-standard chargers for EV adapted to
the conditions of Santiago de Chile. The Japanese company Marubeni brought a
Japanese charger that would be modified by UChile to enable charging under
several standards. This project is frozen at the moment due to the unavailability
of funding.
Among the main enabling factors for the provision of these inputs are:
• Rich Chilean mining resources. The LIC is a resource-pushed activity, driven by
Chile´s access to 23% of the world’s lithium reserves (COCHILCO, 2009). Lithium
can be obtained either through salt flat brines or through mineral processing.
Chile is rich in brines lithium, which is easier and cheaper to extract.
• Availability and quality of engineers.
• Experienced mining industry. Chile is a world leader in lithium production, with
44% of the world’s production, followed by Australia with 25%, China, 13% and
Argentina 12% (COCHILCO, 2009).
The LIC´s local technology inputs face considerable shortfalls. Firstly, Chile does not
have previous R&D experience in lithium applications for EV batteries. R&D in the
global EV industry is advancing quickly and it may be too late to catch up with
technological leaders. Several private companies and research institutions are
investing vast amounts of resources in R&D to be ready for massive deployment of EV.
The leading countries, mainly the United States (MIT, 2009), Germany (Green Car
Congress, 2010), Japan (Reuters, 2007), South Korea (Green Car Congress, 2010) and
China (China Automotive Review, 2009), have implemented industrial policies to
111
promote the production of lithium-ion batteries. Even Bolivia has unveiled a strategic
plan to develop a national lithium industry (La Razon, 2010).
Chile does not have an industrial policy aimed at extracting higher value from its
lithium resources, instead focusing on its applications for nuclear energy, which qualify
it as “strategic mineral resource”, in the royalties generated by extraction companies
and in improving lithium extraction capabilities. In 1979, Law 2886 decreed that the
state would control all lithium except for deposits discovered and owned before the
law entered into force. In 1983, Law 18.097 established that lithium deposits could not
be given in concession to mining companies, except for those deposits already
conceded before the law was issued. Currently, all lithium deposits that can be used
for production in the north of Chile belong to CORFO, which leases them to SCL and
SQM. SCL and SQM´s contracts will expire in 2014 and 2016, respectively, but are
expected to be renewed. The Chilean government receives royalties from SQM and
SCL at 6.8% of the value of lithium exports, which reached 263 MUS$ in 2008 and
delivered around 16.5 MUS$ royalties to the state from lithium production (COCHILCO,
2009).
A 2009 report by the Chilean Copper Commission (COCHILCO) recommended active
public policies to increase the government’s role in the lithium market beyond the
mere reception of royalties. COCHILCO is a technical organisation depending of the
Chilean Mining Ministry, and it was created in 1976 to advise the government on
issues related to copper and all other metal and non-metal minerals, except coal and
hydrocarbons. Most of COCHILCO’s recommendations focused on improving
knowledge about the size and ownership of existing deposits and facilitating
technology transfers to small and medium mining companies that allow them to
extract the mineral. Such technology transfers would have a low technology
component, as Chilean lithium can be extracted at a low cost with basic technologies.
COCHILCO’s report also referred to the possibility of creating a “Lithium Institute” that
would be financed by the leasing fees and taxes paid by SQM. Its aim would be to
centralise lithium-related information, specifically about its new applications and
technological developments. The “Lithium Institute” would not have any role in the
promotion of high value-added innovations in the later stages of the lithium batteries
value chain. COCHILCO´s report also recommended creating a “lithium regional
cluster” that would include Chile as well as its lithium-rich neighbours Bolivia and
Argentina. The cluster would explore the development of high value-added industries
in the region, such as the production of metallic lithium or even batteries for EV. Such
activities could find a potential market in EV assembling units announced by ZAP in
Uruguay (ZAP, 2008), while regional integration would allow for exploiting synergies in
lithium production, financial resources for entrepreneurs and market size.
The creation of this “lithium regional cluster”, or the “Lithium Institute”, was not
accomplished. Representatives of the Ministry of Mining, the Chilean Nuclear Energy
Commission, the National Service of Geology and Mining, the Public Companies
Organisation and COCHILCO met in October 2009 to discuss the recommendations of
COCHILCO’s lithium report (COCHILCO, 2009). The meeting confirmed Chilean interest
in the extraction and initial processing of raw materials and a lack of interest in
undertaking higher value-added activities. Attendants at the meeting concluded that
the three main priorities of the Chilean Government as regards lithium exploitation
112
were updating information about lithium deposits, improving available information
about expected growth in the lithium market and improving understanding of the
current extraction methods employed by SQM and Chemetall, to ensure that efficient
methods are used.
Foreign inputs into the technology transfer process
Chilean distance to the technological frontier in the Li-ion batteries field indicates that
foreign technology transfers are essential in achieving results that can compete with
the fast technological pace of leading research institutes and private companies. The
LIC needs particularly:
• Know-how about lithium application to batteries for EV. Research centres
around the world are focusing on the four main priorities of Li-ion batteries to
enable the large deployment of EV: reducing battery cost, increasing their
storage capacity, reducing the charging time and increasing their life. The LIC is
currently trying to find its niche in these main issues of global research, on the
basis of Chilean resources and capabilities.
• Know-how about EV technology. Battery cell and pack companies and
automotive firms often enter in partnerships or joint ventures to develop
battery technology for vehicles. This allows joint R&D and production facilities
for vehicle and batteries as well as agile and reliable suppliers nearby for
automakers. These alliances are particularly strong in Japan, the leading
innovator in EV and lithium-ion batteries. Marubeni, one of the members of the
LIC, can provide first-hand experience about EV technology, as it participates in
a number of initiatives for its development in Japan and commercialises Nissan
EV. However, Marubeni does not have R&D capabilities in Japan, where its
automobile business is limited to the distribution of Nissan automobiles. The
centre is contacting other leading Japanese organisations that can provide the
required know-how in EV technology.
Channels of technology transfer
The main channels of TT in the LIC are:
• Exposure to foreign R&D. Members of the LIC are continuously updated with
research developments in foreign laboratories, mainly through publications or
direct contact with foreign researchers. This allows them to build from previous
R&D investments, instead of starting their activities from scratch. LIC has
particularly good links with Japanese research institutions due to the
networking skills of its Director.
• Staff experience in foreign multinationals. Previously to being the Director of
the LIC, Jaime Alée, from UChile, worked for 10 years as the General Manager
of the Japanese multinational NEC, which is currently one of the leading
producers of batteries for EV. Japan is a leader in research on EV and lithium-
ion batteries, and the experience of LIC´s Director has allowed the centre to
build an extensive network with Japanese institutions involved in this
technology.
• Foreign direct investment. The Japanese multinational Marubeni, one of LIC´s
co-founders, joined this initiative through its Chilean subsidiary. Marubeni is
particularly interested in the EV business because it is the official distributor of
113
Nissan automobiles in Chile. In turn, Nissan is strongly involved in the
development of EV through partnerships with Renault and NEC and expects to
start mass production of these vehicles in the short term. Nissan´s EV will only
be sold in Chile if the necessary charging infrastructure is in place. According to
our interviewee, Marubeni has plans to sell a Nissan EV in Chile by 2012, so it is
in Marubeni´s best interest to develop enabling conditions for the adoption of
this technology in the country.
• Partnerships with foreign research institutions and private companies. The
centre has still not reached this stage, but it plans to attract foreign research
groups with extensive experience in lithium-ion batteries.
Trade openness in the Chilean economy is seen as an enabling factor, as it has allowed
exposure to foreign multinationals in the EV business and access to imported
technology.
Some of the barriers to the TT process are high entry barriers in the automobile
industry and the non-technological profile of FDI in Chile in the automobile sector,
which is characterised by strong competition in order to respond to changing
consumer demands for low-cost products. Several partnerships between battery and
automobile manufacturers are already operating in the joint development of EV, and
these partnerships need to protect their knowledge assets and operate with high
levels of secrecy. FDI in the industry, as shown by the case of Marubeni, is mainly
concerned with selling, not developing, vehicles, which prevents cross-sectoral
technology transfers for battery manufacturing.
Outputs of the technology transfer process
The LIC is still in its early stages and has not delivered any technology output. Its
founders expect that the joint use of local and foreign technological inputs will deliver
new business activities, scientific publications, patents and technological
developments related to the value chain of lithium-ion batteries. Their expected
activities will not only include basic scientific research, but also new business solutions
such as battery recycling and reutilisation for electricity storage in, for example, solar
and wind power plants. The aim of these final products is to create value beyond the
basic activity of extracting and processing a raw material.
To progress towards the delivery of technological outputs the LIC first needs to decide
the research niches on which it wants to focus. LIC’s preliminary analysis has defined
five potential research layers. The lowest value-added, but easiest to attain, would
consist of research to produce lithium compounds in Chile for their use in batteries.
This step would allow the lithium production companies Chemetall and SQM to deliver
higher value compounds by combining several minerals instead of the basic
combinations currently produced. Knowledge necessary for this step would be easy to
obtain, considering the wide experience of lithium producers. However, value-added
benefits would be limited. A higher value-added activity would consist of the
production of cells and batteries. The LIC is preparing a techno-economic feasibility
study to help decision-making in this research area. After the first six to nine months of
operation, and once the research niches are decided, LIC will seek finance to start its
first research projects. Expected sources of financing are additional members, public
subsidies or further contributions from the co-founders.
114
The lack of demand in Chile is seen as a barrier for the development of a local industry
concentrating li-ion battery production. The country does not count automobile
manufacturing facilities in its land which could demand the product and create
synergies with batteries manufacturers. Besides, Chilean authorities have not shown a
clear stand on EV technology and have instead manifested an interest in biofuels as a
clean alternative for transport, which could reduce the involvement of Marubeni in the
initiative. Marubeni is mainly interested in expanding the commercialisation of
Nissan’s electric cars in Chile, so a lack of government will to facilitate the introduction
of the necessary infrastructure for the massive deployment of electric cars would
potentially make it lose interest in the initiative. The centre’s Director anticipates that
if the LIC leads to a spin-off for the production of lithium-ion batteries, all of them will
be exported. Therefore, the activities of the centre would be driven by international
demand-pull policies.
Effects of technology transfer
No technological effects have been achieved so far by the LIC, as it has just started its
operations. Nonetheless, some of the expected effects if the LIC succeeded in
delivering technological outputs would be:
• Increased productivity in the country. The activities of the centre could lead to
the production of higher value-added products by existing mining companies,
such as lithium compounds, at the lower end of the Li-ion battery supply chain.
They could also promote spin-offs for Li-ion battery production at the higher
end of the value chain.
• Knowledge-spillovers. The centre could improve the technological capabilities
of the country through forward linkages with local manufacturing companies,
including mining companies like SQM and Chemetall, and potentially industrial
manufacturing companies.
The lack of a significant industrial fabric that allows backward and forward linkages
between the mining, industrial and automobile distribution sectors is perceived as one
of the main barriers to the success of TT in this project. SQM and Chemetall are
interested in expanding their lithium extraction businesses and do not seem ready to
undertake higher value-added activities such as the manufacture of batteries, which
would pertain to the industrial component manufacturing businesses. The lack of this
intermediate industrial layer precludes knowledge spillovers between the mining and
industrial sectors. On the other hand, automobile distributor Marubeni is interested in
the penetration of EV in Chile and has relevant knowledge in its Japanese
headquarters. However, the absence of an infrastructure for EV would preclude
knowledge transfer to Chilean industries.
4.3.4 CASE 4- Universidad de Chile’s micro-hydro plug & play: new uses and functionalities
for a mature technology
The Energy Centre of Universidad de Chile is developing a micro-hydroelectric turbine
for both on- and off-grid operation that is easy to install and operate. The turbine is
named the “micro-hydro ‘plug & play’ turbine”, to reflect its ease of installation and
use. The objective of this new development is to exploit abundant hydro energy
115
resources in small streams that are currently underutilised in Chile. The main project
participants are:
1. Energy Centre (EC) of UChile, which is a technology developer. The EC was
created in June 2009 with the aim of developing and introducing high quality
technological solutions relevant for national development and being
internationally competitive.
2. Apacomint is a private company created by a professor at UChile. The
company´s main activities are importing and distributing a number of products,
mainly from China. In this project, it was in charge of finding and financing a
suitable low-cost turbine in China, meeting the requirements set by engineers
at UChile.
3. Chinese technology providers have provided a low-cost turbine, which will be
adapted to Chilean conditions and upgraded with new functionalities by the
researchers at UChile.
4. The Joint European Latin American Universities Renewable Energy project
(JELARE) is a co-operation scheme involving universities from Germany, Latvia,
Bolivia, Brazil, Chile and Guatemala with the aim of fostering innovative labour-
and market-oriented educational and research approaches in the field of
renewable energies at Latin American and European institutes of higher
education. It has provided finance to UChile’s project through ALFA III, an EU
programme for co-operation between the European Union and Latin America,
in the framework of higher education and training.
Figure 42 identifies the main elements of the TT process in this case study, as
established in the interview with UChile’s representatives. Each of the elements is
described in more detail in the following subsections.
FIGURE 42- MICRO-HYDRO PLUG & PLAY TECHNOLOGY TRANSFER PROCESS
116
Local technology input
The project was the idea of the Director of the Energy Centre (EC) of UChile. It is in line
with the main objective of the EC, which is contributing to energy technology
development in Chile with innovations that are competitive internationally. The
existing technological capabilities of the EC were key to the initiation and success of
the project. The university has considerable research acumen in hydro technology,
which has been driven by large hydro energy resources in Chile. During the last eight
years, several graduate and under-graduate dissertations have focused on the diverse
characteristics of micro-hydro technology, such as monitoring systems and grid
synchronisation. There is also a specialised laboratory focused solely on hydro
technology research.
The EC team at UChile developed a micro-hydro plug & play turbine concept, which
required a basic turbine to which a number of functionalities would be added. EC
researchers prepared specifications for the basic turbine including weight, power,
water level, turbine type and cost requirements. Apacomint used its network of
equipment suppliers, mostly from China, to find the lowest cost alternative and yet
meet the specifications provided. Apacomint visited several Chinese turbine
manufacturers and selected the most appropriate equipment. This is the point at
which the project stood at the moment of the interview. The next step after receipt of
the imported turbine would be its adaptation to Chilean conditions and the addition of
new functionalities, with the first prototype expected to take one year to develop.
Engineers at UChile will also undertake market research to explore the demand
prospects and pricing options for their new technological development.
The project could be implemented with the financial support of private Chilean
company Apacomint and from European Union funds through the JELARE network.
UChile expects to access government funds for the prototype development stage,
which will be cost-intensive. With this aim they have just prepared a proposal to obtain
R&D funds from CONICYT and the Ministry of Energy, through a programme aimed at
small-scale renewable energy projects in rural areas.
Foreign technology inputs
The interviewed researcher considered that there has not been a relevant TT in this
project, as all the technological development to add new functionalities to a basic
hydro turbine has been done in-house by UChile engineers. However, the new
technology would have not been possible without an imported Chinese turbine, as
Chile does not have a local industry producing this basic equipment. Chinese products
were selected because UChile´s partner Apacomint had a previous trade relationship
with Chinese suppliers, which enabled them to negotiate favourable conditions. The
low cost of these turbines was another reason for their choice. There has not been
technology transfer from China beyond hardware and written instructions for its
operation.
In addition to hardware, knowledge transfers may have occurred through the training
of EC members in foreign universities. In particular, the EC Director, who is the main
architect of the innovation, gained his doctorate at Dortmund University, Germany. In
any case, it is difficult to assess whether and how his experience in Germany
influenced the innovation.
117
Channels of foreign technology
The main channel for foreign technological inputs has been the importing of Chinese
equipment. Communication problems with the Chinese supplier were mentioned as
one of the barriers to the TT process because all of the instructions were written in
Chinese, which required intermediaries for their translation and for additional
questions to the providers. Communication with Brazilian providers, who are also very
competitive in hydro technology, is considered easier than with Chinese providers
because there are very few language barriers with Brazil and their geographic
proximity enables more direct contact.
Technology outputs
The project had not yet achieved the intended technological output at the date of the
interview, as it was still in its early stages of development. The shorter term expected
outputs are a patent and a prototype of the plug & play micro-hydro turbine. The new
technology will be adapted to the conditions of rural Chile, solving two main
drawbacks of existing off-grid hydro technologies: costly and complex installation and
potential phase-out of the technology if the rural area is connected to the grid. The
modular structure of the turbine allows easy assembling and minimises engineering
work. The automatic control of energy supply and demand allows operation on-grid,
when there is a surplus of power generated, and off-grid, when the power is required
on site. The turbine is user-friendly and allows its owners to monitor turbine variables
easily, even if they are not engineers or technology experts.
After the successful development of a prototype, researchers will present the new
technology to several companies to attract interest for local production. The
subsequent technology output will be the commercial production of the turbine, in
partnership with a private company. Several studies about the mini hydro energy
potential in Chile indicate that installed capacity could reach 20,000 MW. The micro-
hydro plug & play initiative expects to reach several MW of installed capacity, targeting
individuals with access to hydro resources, mainly in the central and southern regions
of the country. The estimated payback for the turbine would be around three to four
years.
FIGURE 43- MICRO-HYDRO PLUG & PLAY TURBINE CONCEPT
Source: Centro de Energía, Universidad de Chile
118
The significant size and low cost of hydroelectricity resources in Chile, high electricity
prices and the possibility to expand into other South American countries with
abundant hydro resources are considered the main enablers of the project.
Some identified barriers to turn technological inputs into outputs are:
• Lack of a competitive industrial equipment sector that could take up
commercial production of the innovation. There are also some gaps in the
whole value chain of the new turbine, including maintenance, servicing and
spare parts. Public policy is necessary to promote industrial development in
Chile; otherwise, the country will continue to rely on foreign equipment, and
technology transfers will not help to create an industrial sector.
• Unavailability of finance for commercial production, once a successful
prototype has been tested. This is expected to be the strongest barrier to
massive deployment of the micro-hydro plug & play technology, which will not
be competitive at the start, although costs will decrease as manufacturers
advance through the learning curve. They anticipate that public support will be
required for local producers that initiate commercial production of the
technology. Such has been the approach in China or Brazil. However, Guillermo
Jimenez doubts that the liberal Chilean economy would be willing to support
this incipient industry in the future.
• Policy barriers related to the lack of clear regulations about grid connection for
micro-hydro generators. There is uncertainty about the possibility of selling
electricity to the grid, so the government is currently working to solve this legal
gap.
The CDM has not played any role in promoting their innovation, because transaction
costs for registration are too high for small-scale technologies like this.
Technology effects
The expected effects of the technology transfer process are:
• Increased productivity. The production of micro-hydro plug & play turbines
would introduce Chile into a high value-added industrial activity beyond its
currently natural resources-based industry. Chile would not be able to provide
basic hydro turbines able to compete with the low-cost Chinese alternatives,
and therefore should incorporate some differentiating characteristics that
allow for charging higher prices. This is the intention of the turbine under
development.
• Knowledge spillovers through forward links. If the project reaches the stage of
commercial production, the knowledge acquired by EC researchers will be
transferred to local manufacturing companies, increasing the technological
capabilities of the country.
• Technology cost reductions. The use of Chinese low-cost turbines and the mass
production and deployment of the improved technology would enable cost
reductions.
• Emissions reductions in Chile. Mass deployment of the technology would
provide higher access to clean electricity for rural areas, which would then
replace electricity generation from diesel generators.
119
As in the technology outputs, the missing enablers to achieve these effects are a local
industrial sector and finance for the commercial production of the innovation.
4.3.5 CASE 5: KDM: pioneering power generation with biogas from urban waste
KDM, S.A. is a company specialising in the handling, treatment and final disposition of
solid waste in Chile. KDM is fully owned by foreign investors Urbaser, from Spain, and
The Danner Company, from the United States. KDM started operations in 1995 when a
10-90 joint venture between a Chilean and a United States partner won a bid to
manage waste produced by the northern region of Santiago. As part of their project
they developed the Quilicura Waste Transfer Station and the “Loma Los Colorados”
landfill site, 63 km north of Santiago, which is the largest in the country. It has been
operating since 1996 and has so far received more than 16 million metric tons of
residential solid waste from the Greater Santiago metro area. Its total capacity is 100
million tons and it is expected to operate until 2045, when it will reach full capacity.
Urbaser entered the company in 1999 with a 50% participation share. Therefore, KDM
is now jointly owned on a 50-50 basis by the Spanish and American partners32.
KDM now manages four residential solid waste landfills in several regions of the
country. It has also recently opened a new division, KDM Energia, to develop
renewable energy generation projects. Its first project, and the first biogas generation
project in Chile, is a power generation plant in Loma Los Colorados, which was
conceived as a CDM project for landfill gas collection, flaring and electricity generation.
The project achieved emission reductions through methane destruction by flaring and
combustion and through the substitution of fossil fuel-based electricity from the grid
with renewable power.
KDM´s TT process involving implementing and operating a new low-carbon technology
is illustrated in Figure 44 further described in the subsequent subsections.
32
Declaracion de impacto ambiental ampliación del sistema de abatimiento de biogás; sistema de
captación, termodegradacion y utilización energética, en el marco del mecanismo para un desarrollo
limipio, en el relleno sanitario Loma Los Colorados, KDM, S.A. January 2006.
120
FIGURE 44- KDM TECHNOLOGY TRANSFER PROCESS
4.3.6 Local inputs to the TT process
KDM did not receive technical support from Urbaser or The Danner Company for the
implementation of the project because it was initiated and managed by the Chilean
subsidiary on its own. The technological capabilities of the local staff were essential for
the success of the TT process, and they were very involved in the power generation
project design and engineering phases, as the equipment suppliers did not have
enough experience to solve several technical conundrums faced during
implementation. Chilean engineers were also in charge of the installation of the
electricity network, LFG capture and flaring system integration and the related
construction works.
KDM lacked experience in biogas power generation when they implemented the
project and went through a profound learning process. Distributed power generation
was a technological challenge too, as the national grid was not ready for them to be
both providers and consumers of electricity. Moreover, gaining the necessary licences
for grid connection was a long and complex process.
The quality of Chilean engineering education is considered key for the successful
adoption of foreign equipment and the independent solution of technical challenges
faced through the project. The steep learning curve of the first project enables them to
undertake new renewable energy projects.
International inputs into the TT process
In order to implement its LFG capture, flaring and electricity generation project, KDM
required foreign equipment and know-how. It was the first company to implement
such a project in Chile, so the equipment and knowledge were not available in the
country and needed to be imported. Before undertaking the LFG project, foreign
121
knowledge had already been necessary to design and manage the landfill, thus
meeting developed country environmental standards.
Foreign investment provided KDM a solid financial base on which to start the project.
One of the main hurdles in Chile to starting a renewable energy business is obtaining
finance from commercial banks, who still consider these activities too uncertain. Banks
do not have technical capabilities to analyse the risk of renewable energy projects and
consider most of them as very risky. Additionally, their financial risk analysis is based
on future price projections that expect a steep decrease in electricity prices as a result
of the installation of large hydroelectricity projects in the south of Chile. Under a
scenario of low electricity prices, most renewable energy projects are not profitable,
but KDM could count on solid internal financial capabilities which could back the risk of
the operation and provide assurance for the bank to finance one share of the
investment costs.
Channels of international technology transfer
All the equipment necessary for the landfill gas capture, flaring and power generation
system has been imported:
• Loma Los Colorados landfill was designed by a United States engineering
company, meeting US environmental and safety standards. In 1995, when the
landfill was developed, Chile did not have the necessary local knowledge to
design landfills with these characteristics.
• The initial capture and flare system used before the project expansion was
imported from a Dutch company.
• The new biogas capture and flaring system uses flares, compressors and
capture pipelines from the United States. System integration was made by
Chilean engineers according to the instructions provided by the equipment
providers, while US technicians supervised the start-up of the system.
• The initial 2MW power generation system uses two Waukesha generators from
the United States. The choice of technology provider took into consideration
that the equipment was provided by a local distributor with relevant
experience in the area and the capacity to offer technical support locally.
Continuous support would reduce the risk of interruptions, which was their
main priority in this first stage of the project.
• The additional power generation system of up to 28MW installed capacity will
use GE Jenbacher generators produced in Austria. The choice of technology
provider was mainly driven by the high quality of the equipment. Additionally,
Jenbacher could offer technical support from a service provider in Argentina
with long experience in the installation of this equipment.
URBASER has similar biogas generation projects in Spanish landfills but their
contribution was not required beyond the approval of the project budget, as Chilean
staff could manage the project on their own, even though this required an intensive
learning process. The most relevant TT channels for KDM’s renewable energy activities
were importing equipment and codified instructions for their operation, as well as the
technical support for installation and maintenance. Free trade agreements were an
important enabler for the imports of equipment at low cost.
Technology outputs
122
The following outputs have been achieved through the TT process:
• Landfill gas (LFG) collection and flaring. The system started operations and was
registered as a CDM project in March 2007. The landfill had a small LFG
collection and flaring system with a capacity of 1,000 m3/h before the new
system was implemented. The expanded and improved system has a capacity
of 10,000 m3/h. The LFG captured increases as the landfill reaches its full
operating capacity. In 2011, 15 years after the start of the landfill operations,
the expected LFG capture is 9,000m3/h.
• Electricity generation plant “Loma Los Colorados I”, with 2MW installed
capacity, started operations in November 2009 and has also been included as
part of the registered CDM activity. The current biogas generation plant only
uses 1,000 of the 9,000m3/h of biogas generated. The remaining biogas is
flared.
• Electricity generation plant “Loma Los Colorados II” started with 10MW
installed capacity in July 2011 and will increase capacity every year as biogas
generation grows to reach 28MW by 2024. The aim of this project is to use as
much biogas as possible for electricity generation instead of flaring.
According to information published in the UNEP-Risoe CDM pipeline33, estimated
average annual emission reductions for the project are 582ktCO2e/y as a result of
methane destruction and the substitution of fossil fuel-based power generation.
FIGURE 45- GENERATORS AT LOMA LOS COLORADOS I POWER PLANT
Source: KDM, 2010

The main driver of the landfill gas capture and flaring project was the possibility of
obtaining carbon credits through the CDM. With estimated annual emission reductions
of 582 kCO2e, the project obtains a substantial income through the sale of carbon
credits. This project would have not been profitable without CDM income. The power
generation project was also part of the initial CDM project, but was not driven by
carbon credits income. In fact, the interviewee referred to other companies’ biogas
generation projects that have not been registered as CDM due to the high transaction
costs and delays in the process.
The power generation project was motivated instead by high electricity prices. It had
been under consideration by the company since 1996, when the landfill started
33
www.cdmpipeline.org
123
operations, as electricity could provide an additional income for the landfill operation.
The decision to invest in a power generation system came only in 2006, after the gas
distribution company Metrogas made KDM an offer to buy their landfill gas to
generate electricity. When KDM analysed the offer, they realised that they could start
a much more profitable business by generating electricity themselves. KDM Energia
operates in the liberalised electricity market without fixed contracts, selling its
electricity at the system’s marginal cost. The high electricity prices enable them to
make a significant profit, and they have also motivated further KDM’s investments in
biogas and hydro generation.
KDM is facing some legal problems as a result of its generation of biogas, as the power
plant exceeds the NOx emission limits of a new air quality regulation for the
metropolitan area of Santiago. To comply with the new legislation they have installed
some SCR (selective catalytic reduction) filters but they don’t have enough knowledge
about the performance of this technology and fear that their plant may be closed if the
filters don’t perform as expected. KDM believes that the new legislation severely limits
the potential for biogas electricity generation in Chile, as landfill gas is typically located
next to metropolitan areas.
Technology effects
The main effects of the TT process are:
• Emission reductions through the capture and combustion of landfill gas and the
replacement of fossil fuel-based electricity generation.
• Horizontal spillovers, as KDM Energia plans to leverage knowledge acquired
through the landfill gas generation project to expand its renewable energy
business. The company will mainly develop biogas projects. It is planning to
develop two landfill gas capture and generation projects in the south of Chile,
as well as biodigestors in biomass incinerators. They are also considering mini-
hydro projects, due to their low technological risk and abundant hydro
resources. Wind projects are still considered quite risky, particularly because
the largest wind project implemented in Chile, Canela I, has performed below
expectations.
• Technology cost reductions, as a result of learning effects.
• Increased technological capabilities of the recipient company, as a result of
internal learning processes.
High electricity prices and potential carbon income are enablers for KDM´s
development of additional renewable energy projects. However, there are not
sufficient policies in place to support renewable energy technologies that cannot
compete with cheaper coal, fuel-oil, gas or large hydropower plants. The current
legislation, based on a quota scheme, has failed to promote the renewable energies
industry, and it has not provided any certainty about the price for renewable electricity
nor created a liquid market for NCRE certificates. Additionally, renewable energy
projects would very much benefit from mechanisms that ensure a minimum price level
and therefore reduce their risk profile for commercial banks.
124
4.3.7 CASE 6: Energía del Sur: meeting European air quality standards in Chile
Energía del Sur (ES) is a Chilean company distributing biomass thermal generation
technologies for the residential, commercial and industrial sectors. ES installs the
biomass generation equipment and provides maintenance and calibration services.
The company also provides thermal biomass energy generated in its own boilers to
companies in the industrial and services sectors.
The company was created in 2006 by two Austrian partners who had been long
established in Chile. The idea to start up the business came from one of the partners
working in the design and construction of high-end houses in the south of Chile. He
observed that the residential sector did not have many options for the provision of
low-cost, safe, efficient and clean thermal energy, in spite of the widespread
availability of biomass resources in the southern region. The only residential heating
alternatives were gas, oil and wood. Wood was not very safe, while oil and gas were
costly, polluting and used very obsolete technologies. After analysing several
alternatives, both partners concluded that biomass thermal generation with pellets
and chips was the most appropriate alternative for the region. The partners´
knowledge of the Austrian experience gave them confidence about the technical and
economic feasibility of these technologies.
Figure 46 illustrates the TT process in Energía del Sur based on the findings of the
interview. Each of the elements depicted is described in the following subsections.
FIGURE 46- ENERGÍA DEL SUR TECHNOLOGY TRANSFER PROCESS
Local inputs into the technology transfer process

The wide availability of biomass resources in Chile has enabled the accumulation of
experience in biomass energy technologies. The general education level of the Chilean
125
workforce is also considered an enabling factor for technology transfer. The company
has 18 employees, all of them Chilean except for the two Austrian founders.
Lack of awareness and technological knowledge among local stakeholders, including
construction companies as clients, installation companies and pellet providers as
suppliers, and banks were mentioned as some of the main barriers to the TT process.
Construction companies usually prefer well-known, incumbent fossil-fuel technologies
with a lower initial cost. Pellet providers were unable to provide products to meet the
boiler´s specifications. Besides, installers and service providers were unable to operate
the technology. To solve this lack of awareness among their commercial partners,
Energia del Sur regularly attends construction fairs to increase awareness among
construction companies and is involved in training activities for their pellet suppliers
and installers. Local banks are reluctant to provide finance for a technology they do
not understand; consequently, the co-founders needed to provide themselves with
most of the financing to start their business.
International inputs into the technology transfer process
The Austrian background of the company founders exposed them to successful
experiences implementing biomass boilers for thermal energy generation in the
residential and industrial sectors. Based on the similarities of Austrian and southern
Chilean biomass resources, the company founders concluded that it would be
technically and economically feasible to implement similar biomass energy generation
technologies in Chile.
Chile does not have a boiler manufacturing industry, and therefore all equipment
needed to be imported. The know-how to install, operate and maintain the equipment
was also required, but the Austrian origin of the company´s founders made it easy to
access and negotiate favourable conditions with Austrian technology suppliers.
TT in ES has happened through the following channels:
• Imports of low emission equipment from Austria. Energía del Sur is the
exclusive distributor of the brands Binder (industrial biomass boilers) and KWB
(residential biomass boilers) in Chile.
• Training by Austrian technology providers. Austrian technicians regularly visit
Energía del Sur to train their staff. The Chilean staff are also sent to Austria
regularly to update their technical training.
• Foreign staff. The two Austrian co-founders, trained as engineers in their home
country, have contributed with their European training and environmental
awareness to the successful introduction of equipment with a higher
environmental performance than the Chilean baseline.
The favourable Chilean legal and macroeconomic environment, with free trade
agreements that enable low-cost imports and appropriate transport infrastructures,
was mentioned as one of the key enablers for TT in this project.
Technology output
ES sells, installs, maintains and calibrates Binder and KWB biomass boilers, but it also
sells biomass thermal energy generated in its own boilers to industrial and commercial
126
clients. The distributed boilers have the best air quality performance in the Chilean
market, the highest efficiency and the highest level of automation. At the date of the
interview, ES had sold around 450 boilers between 8kW and 2MW, 85% of which were
sold to the residential sector and the remaining 15% to the industrial sector, but in
terms of installed capacity the industrial sector surpasses the residential. ES does not
have any significant competitors in their niche, mainly because the boilers are
expensive and require technical knowledge for their installation.
FIGURE 47- KWB AND BINDER BIOMASS BOILERS DISTRIBUTED BY ENERGIAS DEL SUR
Source: www.energiadelsur.com
The high prices of fossil fuels and the new air quality legislation were the main drivers
of the increased demand for their biomass boilers. The tightened air quality standards
in the city of Santiago de Chile have significantly increased demand. ES is actively
lobbying for tighter air quality standards in other metropolitan regions, too.
Environmental performance was indeed one of the key features for the choice of
biomass technology providers. The selected technology met strict European emissions
standards at a time when Chilean regulation was lax. Foreign air quality legislation is
also a driver for their business, as it is seen as the benchmark that Chile should achieve
if it wants to become a developed country.
High fossil fuel prices enabled a relatively short payback period of one to two years in
the industrial sector and three to five in the residential sector, but even with this short
pay-back period, many construction companies were still reluctant to invest in biomass
boilers with higher initial costs than the incumbent fossil fuel-based boilers.
Construction companies usually look at reducing their short-term costs instead of the
lifecycle costs of the building, so their market is mostly limited to high value family
houses where the property owner is the person making investment decisions on the
heating equipment.
The approval of energy efficiency standards for buildings or for boilers themselves
could significantly increase the penetration of clean and efficient biomass boilers.
Furthermore, subsidies to residential consumers could be a good incentive but are not
considered a feasible option in Chile, given its liberal economic philosophy.
As regards international climate policies, at the moment they are relevant for their
industrial clients. International voluntary carbon footprinting initiatives could have a
significant impact for the agro-industry, as some European clients request carbon
footprint data from their suppliers.
127
Technology effects
The main technology effects resulting from the transfer of Austrian biomass energy
generation technologies have been:
• Emissions reductions in the industrial and residential sectors.
• Knowledge spillovers to local suppliers of pellets and chips, installers and
construction companies. Energías del Sur has trained local pellet providers to
improve their logistics systems and the quality of their products. The pellets
market was not very developed due to a lack of demand. They have also
trained installers and clients to operate the technology and repair the
equipment.
The effects of the technology could be enhanced through increased awareness of the
final consumers, i.e. construction companies or owners of buildings. ES regularly
attends construction fairs to increase awareness among construction companies
4.3.8 CASE 7: Eléctrica Nueva Energía: Brazilian technology for the widespread use of
biomass resources
Electrica Nueva Energia (ENE) operates a 15MW biomass cogeneration plant in the Bio
Bio region in the south of Chile. It uses biomass residues from the forestry company
CMPC Maderas, paper mill Forestal y Papelera Concepcion (FPC) and some additional
small forestry companies. Heat is sold to CMPC and FPC and electricity is sold to the
national grid. ENE was acquired in April 2010 by the private equity fund Americas
Energy Fund, managed by SCL EnergiaActiva. The private fund is interested not only in
renewable energies, but it also owns a coal mining company, a natural gas combined
cycle power plant and a gas thermal power plant in Colombia and is planning to buy
other gas and coal thermal plants in Peru, Guatemala and Chile.
The biomass cogeneration project was initially developed by the paper mill FPC to gain
a stable energy price and reduce the risk of its newsprint paper production activity.
The cogeneration plant was initially designed to cover half of the energy needs of FPC,
with a 12 MWe installed capacity, which was later increased to 15MWe. In 2007, due
to high electricity prices, it was more profitable for the cogeneration plant to sell
electricity to the grid than to use it to supply FPC’s energy-intensive newsprint
production. High electricity prices meant that their core business was no longer
profitable, so FPC temporarily closed its operations and sold the cogeneration plant to
ENE, which continued heat and power generation for sale to local forestry industries
and the national grid. FPC reopened in 2009 as a producer of recycled fibre-based
packaging paper, which was much less energy-intensive than newsprint paper.
Figure 48 shows the elements of ENE´s TT process, while the remainder of this section
will describe them more thoroughly.
128
FIGURE 48- THE TECHNOLOGY TRANSFER PROCESS IN ELECTRICA NUEVA ENERGÍA

Chile has abundant forestry biomass resources and has accumulated significant
experience in its energy use. Biomass cogeneration is quite common in pulp and paper
plants. Previous experience enables the fast adoption of imported technology and the
negotiation of favourable conditions with importers, while installation and servicing
are performed by Chilean companies that have a thorough understanding of the
technology.
Biomass cogeneration in Chile is a resource-pushed activity, mainly driven by the
widespread availability of forestry biomass. Forestry resources have also driven the
development of a solid pulp and paper industry, which is the main industrial client of
biomass energy.
Chile does not produce competitive cogeneration equipment, which therefore must be
imported from abroad. ENE´s cogeneration plant uses the widespread steam-Rankine
cycle technology that consists of the direct combustion of biomass in a boiler to
generate steam, which is then expanded through a turbine. The plant manager
considers that there has not been technology transfer in this project because the
imported technology is very mature – steam-Rankine cycle technology for biomass
cogeneration was introduced into the Chilean market about 100 years ago. ENE’s
interviewed plant manager considers that there are few opportunities for technology
transfer in this sector, as the last significant technology transfer happened when
European fluidised bed boilers were introduced in the 1990s. The interviewee believes
that biomass technology transfer opportunities reside in other sectors, such as
agriculture and urban waste.
129
Blueprints from the provider were required for installation, use and servicing, but all
these activities were carried out by Chilean engineers. Brazilian engineers are only
required to supervise the turbines’ overhaul approximately every five years.
The generation equipment was imported from Brazil because European technology
was too expensive for such a small-scale project. The most competitive options were
Brazilian and Chinese, but contact with the Brazilian technology providers was easier
due to geographic proximity and language similarities.
Technology outputs
The cogeneration plant produces heat for forestry, pulp and paper companies and
electricity to be sold to the grid. High and unstable electricity prices were the main
driver for FPC to develop this cogeneration project. In addition, several policies
contributed to its feasibility. Firstly, the biomass cogeneration plant benefits from Law
20.257, which sets renewable generation quotas. ENE has signed a 15-year bilateral
contract with the mining company CODELCO to sell them their NCRE attribute. Law
20.257 forces electricity generators, not consumers, to meet renewable generation
quotas, but generators have transferred that duty to large consumers, mainly mining
companies. Even though ENE has benefited from Law 20.257, the plant manager
considers that the law has had a small impact in promoting renewable energy
generation in the country. Its implementation is flawed, though, as it excludes a large
share of electricity contracts from compliance. Additionally, the high concentration of
the electricity generation industry means that the companies forced into having a
quota of renewable energy generation usually block the entry of new competitors, by
implementing their own low-cost renewable energy projects or by paying the penalties
instead. The sale of renewable electricity is not guaranteed and small renewable
energy companies face high risks due to price volatility. A mechanism designed to
stabilise electricity prices could contribute to boosting the sector, but it would
introduce rigidities in the market that would translate into higher electricity prices for
the consumer. In a developing country like Chile, this is considered an obstacle for
sustained economic growth.
Secondly, the project was registered as a CDM in July 2009, well after the start of
operations and after an administrative process that lasted more than two years. The
project has not yet requested the issue of carbon credits, and therefore has not yet
made any revenue from its CDM registration. Due to the small size of the project, the
transaction costs of carbon credits issuance are considered too high, so project owners
prefer to accumulate a significant amount of emission reductions before requesting
their first CER issuance. The project is currently profitable without the need for carbon
credits; however, the manager of ENE claims that the possibility of being able to obtain
carbon credits was key in the fund‘s decision to invest in the cogeneration plant.
Technology effects
The implementation of the project has not had any significant knowledge spillover
effects for Chile, as the technology used was already mature in the country. It has
achieved, nonetheless, emission reductions of around 34ktCO2e/year, by replacing
fossil fuel-based grid electricity with biomass CHP generation and by avoiding methane
130
emissions due to natural decay or the uncontrolled burning of biomass that would not
be used in the absence of the project activity.
4.3.9 CASE 8: Biomasa Chile: new energy uses for Chile’s vast forestry resources
Indef is a Chilean forestry engineering company that has been providing forest
management services since 1992. In 2002, it defined three business lines related to
renewable energy: forestry waste recovery, biomass recovery from the sustainable use
of native forests and dendroenergetic forest plantations for biomass production.
Biomasa Chile (BC) was created in 2005 to develop these business lines and is currently
exploring additional opportunities, such as the provision of biomass for co-firing in coal
power plants.
At the date of the interview, BC´s most developed business line was the recovery of
forestry waste. This activity consists of three main steps: biomass recovery from forest
plantations, biomass crushing and transportation to consumption centres. BC has a
wood crushing installed capacity of 180,000 m3/month, produces 1.2 million m3 of
biomass and expects to grow to 1.5 million m3 in 2011 and 2 million m3 by 2012.
Figure 49 illustrates the technology transfer process that BC has gone through. Each of
its elements is further described in the following subsections.
FIGURE 49- BIOMASA CHILE TECHNOLOGY TRANSFER PROCESS
Local technology inputs

BC was initiated on the solid basis created by Indef after around 15 years of experience
in the Chilean forestry sector. BC already had a client base which would provide access
to forestry biomass resources, all the necessary equipment except wood crushers, the
131
know-how to operate conventional forestry equipment, skilled local staff and an
extensive knowledge of the Chilean forestry sector.
BC had the capacity to undertake of its own accord two of the three steps of the new
biomass supply activity: biomass recovery and transportation to consumption centres.
The first step required imported excavators and skidders already available in the
company. For the transportation step, BC developed an efficient logistics system that
minimised loading and unloading times in the production and consumption sites. BC
required foreign support for the wood crushing step, but was involved in internal
learning processes for the operation of the imported machinery and its adaptation to
Chilean conditions. As a result, BC would often advise the technology provider on how
to improve the performance of the equipment, based on their hands-on experience.
BC considers that they have succeeded in becoming technologically independent, as
they don´t need the foreign supplier to operate the technology. However, they still
depend on the provider to procure spare parts, which are not available in Chile.
BC has gone through a steep learning curve to improve the quality of the biomass
produced and to adapt it to the requirements of its clients’ boilers. For example, at the
start of operations around 20% of their biomass load was sand, but this has been now
reduced to 5% of the load. Humidity content has also been significantly improved with
experience, as BM can now provide an optimal humidity content of around 40%, even
in the wet southern region, thanks to improvements in the logistics chain.
BC has an active approach to innovation and is currently involved in new R&D activities
to use native forest biomass as an energy source for co-firing in coal power plants. The
company has received some public funds for R&D on biomass for co-firing and has
approached several power generation companies with coal thermal plants. However,
the future of this new activity depends highly on the acceptance of co-firing as an
eligible NCRE by Law 20.257. If co-firing is not accepted in the renewable energy
quotas, power companies will not be interested in buying BC´s biomass due to high
shipping costs. At the date of the interview, the Ministry of Energy seemed unlikely to
accept co-firing to meet renewable power quotas.
The enabling conditions for BC´s provision of inputs for the TT process were the
widespread availability of forestry biomass resources in Chile, which has enabled
successful forestry, pulp and paper industries and the quality of Chilean forestry
engineers.
Foreign knowledge was essential, firstly for the business idea to be born, and secondly
for it to materialise in a business activity. The founders of BM came up with their
business idea after observing successful biomass production and transformation
companies in the United States and northern Europe. They attended several
international congresses and visited European and North American companies to learn
more about the possibility of producing biomass energy from forestry waste in Chile.
Once the investment decision was made, foreign equipment was required, as wood
crushing machines were not available in Chile. This equipment was imported from the
United States in 2005, and BC was the first company in South America to use wood-
132
crushing equipment considered the state-of-the-art in the industry. In addition to the
equipment, BC required the know-how to operate it efficiently.
Excavators and skidders to recover forestry biomass were also imported, but are
considered usual tools of business in the Chilean forestry industry.
FIGURE 50- BIOMASA CHILE´S WOOD CRUSHING EQUIPMENT
Source: Biomasa Chile (2010)
Channels of international technology

The main channels of foreign technology transfer for BC have been:
• Participation in specialised international industry events. These events have
exposed BC staff to the technological developments of forestry biomass-rich
countries and given birth to the idea of a new biomass recovery business line.
• Imports of wood crushing equipment from the United States. The market for
wood crushing equipment is quite specialised and there are very few providers,
who are mainly located in the United States. BC was the first company in Latin
America to import this equipment.
• Training for the efficient operation of the wood crushing equipment. The
United States technology providers trained Chilean staff to ensure that they
would be independently able to operate the machinery.
Technology outputs
BC produces forestry biomass feedstock that is supplied to large Chilean forestry
companies for energy use. The final output of BC´s activity is therefore a raw material.
High electricity prices have been the main driver of BC´s activity. Biomass cogeneration
could offer the forestry industry cheaper electricity and heat than the options provided
by fossil fuel alternatives; however, the high volatility of electricity prices created high
uncertainty at the start of BC´s operations, when its client base consisted of only one
customer. Under a scenario of low electricity prices, their clients would prefer to buy
electricity from the grid instead of using biomass resources to produce and sell their
own electricity, so sustained high electricity prices increased their client base. The risk
of changes in demand caused by electricity price volatility was reduced through long-
term biomass supply contracts with their clients, which were enabled by a good
previous relationship dating back to the start of INDEF activities in the forestry sector.
This security of demand enabled BC to invest in expensive new machinery using its
133
own resources. The company still has a large potential to expand further into the
Chilean market, where biomass from forestry waste and native forests is underutilised.
BC has not received public support for its new biomass production activity. Neither the
CDM nor Law 20.257 is considered a significant enabling factor for the TT process. In
fact, Law 20.257 is considered ineffective because the main power companies prefer
to pay a fine rather than allowing the entrance of new competitors in the market. A
mechanism to introduce stability in the electricity market could promote new
investments in renewable energy generation.
Technology effects
The main technology effects of BC TT process are:
• Emission reductions by consumers of forestry biomass feedstock.
• Cost reduction of biomass feedstock through learning-by-doing effects.
• Increased productivity for the Chilean economy, as a resource previously
considered “waste” is now used as an input for heat and electricity generation.
4.3.10 CASE 9: MNC-Gen: missing links for in-depth technology transfer34

MNC-Gen operates in Chile mainly through its control of SING-GEN, the leading power
company in the SING (Sistema Interconectado del Norte Grande). SING-GEN holds 49%
of the market share in the SING and is the fourth-largest electricity generator in the
country. MNC-Gen is a major industrial player in the global energy sector, with a total
world capacity of 72.7 GW. In Latin America it has an installed capacity of 10.7 GW and
5GW under construction. Its world installed capacity is mainly based on fossil fuels,
with 54% natural gas and 10% coal. It also has some renewable energy assets, mostly
hydro but also wind (3% capacity) and biomass (1% capacity).
SING-GEN supplies power to northern Chile´s vital mining region with an installed
capacity of 1,795 MW. Coal-fired plants make up 44% of its installed capacity in Chile
and around 52% of power generation, while gas-fired plants follow suit with 38% of
installed capacity and around 32% power generation. Nearly 15% of power generation
comes from diesel and fuel-oil and the remaining generation is from hydropower.
SING-GEN is currently building two new coal-fired power plants with a combined
capacity of 330 MW, which will bring its total installed capacity in Chile to 2,125 MW in
2011. Future expansion plans include 750MW of coal plants and 280 MW of non-
conventional renewable energies35.
SING-GEN controls the risk of its activity by signing long-term power purchase
agreements (PPAs) with final consumers, mainly mining companies. From 2012, 83% of
its sales will be through PPAs within industry and the remaining 17% will be sold on the
regulated market for residential customers. The PPAs are indexed to the price of a mix
of coal, LNG, fuel-oil and diesel prices, as well as the SING marginal energy cost and the
consumer price index (CPI). Tariffs also consider the effects of maintenance and
34
MNC-GEN and SING-GEN are fictitious names for the companies, as the interviewed company
representative requested that their real name remains confidential.
35
SING-GEN. “Company presentation” September 2011.
134
regulatory costs. SING-GEN follows a conservative commercial policy, with new
capacity investments always supported by contracts.
MNC-Gen activities in the Chilean renewable energies sector are just starting. The
company is well positioned in a diverse portfolio of technologies, but in most cases its
projects are at the feasibility analysis and development stages, with only one wind
farm in operation in the SIC. MNC-Gen’s portfolio of renewable energy technologies
consists of:
• Wind farms, including the operating 38 MW wind farm of Monte Redondo, in
the SIC are currently being expanded to a capacity of 48MW36, and an
additional planned capacity of 100 MW in the SING.
• Solar-thermal, including a planned 5 MW thermal concentrated solar power
(CSP) plant, which will supply superheated steam to SING-GEN´s Mejillones
coal-fired plant (150 MW) in the north of Chile.
• Biodiesel from algae, including an R&D project at an initial stage, to produce
second generation biofuels from microalgae.
• Biomass co-firing in coal-fired thermal plants. SING-GEN is planning to use
biomass jointly with coal in its two new thermal plants of 150 MW each.
Figure 51 summarises the elements of the TT process in the case of MNC-Gen activities
in Chile. These elements are described in more detail in the following subsections.
FIGURE 51- MNC-GEN TECHNOLOGY TRANSFER PROCESS
36
The expansion was planned to finalise by the first quarter of 2011.
135
Local technology inputs
MNC-Gen relies mostly on foreign equipment and human resources for the
implementation of renewable energy projects. The company considers that Chilean
engineering firms do not have a sufficient knowledge base in most renewable energy
technologies due to their lack of experience, and therefore cannot provide a high
quality service. Still, the company relies on local partners for its bio algae and biomass
co-firing projects, which are still in the early stages of development, as Chile has
significant knowledge in biomass energy.
The Chilean workforce is considered either over- or under-qualified for GDF´s needs.
Although Chilean universities provide highly trained superior engineers, GDF misses an
intermediate layer of technicians that can undertake product transformation activities.
Such intermediate profiles are key to developing a national industry but are not
provided by the Chilean education system, so public policies should be implemented to
promote intermediate technical education.
Linked to the lack of intermediate technicians is the lack of a solid industrial sector
related to energy generation. All equipment is imported and there are no capabilities
to develop, operate, repair and maintain endogenous technologies. GDF complains
that they need to source all their equipment from abroad and cannot count on local
technology providers and servicing companies, and they contrast this situation to the
case of Brazil, with a very strong local industry for basic equipment and skilled
technical staff.
The bio algae R&D project, done in unison with Copec, Pontificia Universidad Católica
de Chile, Rentapack and Bioscan, has received public financial support through
CORFO´s InnovaChile programme, which will provide US$3,245 million of the total
investment of US$6,836 million. However, Mr Ferrant considers that this support does
not respond to a clear country strategy, and it could make more sense for Chile to
support incremental innovation in components or support industries rather than
financing new and risky technologies.
MNC-Gen relies almost fully on foreign technology for its renewable energy projects,
as it considers local technology insufficient. The required foreign inputs include
equipment and knowledge for the design, installation, operation, repair and
maintenance of the renewable energy generation plants. MNC-Gen also requires
intermediately qualified technicians, which are not provided by the Chilean education
system.
The main TT channels for MNC-Gen in Chile are:
• Foreign direct investment. MNC-Gen transfers technology to Chile on a
continuous basis through its direct control of electricity generation companies
in the country, while its corporate structure is centralised in its European
headquarters. The technological knowledge acquired through projects
developed worldwide is centralised by a group of technology experts in the
headquarters, who can provide their expertise for projects developed in Chile.
136
• Imports of foreign equipment. All the equipment used or planned for MNC-
Gen’s renewable energy projects in Chile is imported. CSP technology is
imported from German company Solar Power Company, and the wind turbines
in Monte Redondo are imported from Danish company Vestas.
• Foreign services. MNC-Gen has hired foreign engineering firms to support its
own engineers in the design and implementation of the different renewable
energy projects in its portfolio.
• Collaboration with foreign companies. Their planned solar thermal project has
been entirely designed in Europe and all the equipment will be produced and
assembled in Europe by their German partner Solar Power Group.
• Training of local staff. Although still not implemented for renewable energy
generation, GDF runs its own education programme to train technical staff for
its coal plants while they are still at school. A similar scheme would be
potentially extended to their renewable power plants when the projects take
off. They could afford this investment because they have a solid business with
long-term PPA at high prices, but this would not be feasible for other
companies without public support.
The main enabling factors for GDF´s channels of TT to Chile have been its favourable
macroeconomic conditions, low political risk and openness to foreign trade and
investment.
Technology outputs
The end-product of MNC-Gen’s low-carbon TT activities is the generation of electricity
from renewable energy sources. So far, this output has only been achieved in wind
generation, where the 38MW Monte Redondo wind farm has been generating
electricity since December 2009, which will be extended to 48 MW by 2011. Before
starting operations, the company ensured the sale of electricity for 14 years, through a
PPA signed in February 2009 with distribution company CGE. The development of
some additional 100 MW in the SING is expected in the medium term. Wind is the
preferred technology for GDF´s expansion into the Chilean renewable energy market,
mainly because of its low technical risk and accumulated in-house experience.
The 5MW CSP project is still at the design stage and has not yet gone through the EIA’s
administrative process. MNC-Gen and the Belgian company Solar Power Group have
agreed to develop the project jointly, which will supply superheated steam to SING-
GEN´s Mejillones coal-fired plant (150 MW) in the north of Chile. It is expected that a
pilot plant will be in operation in 2012. The project consists of a solar boiler
constituted of modules developed by Solar Power Group and using Fresnel technology.
The boiler is connected to the coal-fired power plant which can reduce its coal
consumption, therefore reducing CO2 emissions and increasing fuel efficiency. MNC-
Gen regards the project as a R&D activity, and its main aim is to learn about the
technology for its future implementation on a larger scale.
The bio algae R&D project is currently at an initial stage, consisting of the selection of
algae cultures. Later on selected cultures will be produced next to SING-GEN’s coal
plant in Mejillones, where CO2 from the coal plant chimney will feed cultivation and
speed photosynthesis. Bio algae will be harvested and go through several chemical
processes to obtain biomass that can be processed for biofuels production.
137
The project for biomass co-firing in coal-fired thermal plants is only at the design stage.
Biomass will be used in two new coal thermal plants of 150 MW each, and it will make
up to 10% of the installed capacity, or 30 MW per plant. The feasibility of the project
depends on the inclusion of this technology with the NCRE sources allowed by Law
20.257, as co-firing is not currently permitted to meet renewable energy quota
obligations. The company considers that the costs of biomass transport from the south
to the north of Chile are too high without the incentive that Law 20.257 would create.
To avoid high transport costs they are considering the option of biomass provision
through dendroenergetic plantations in the north of Chile, in collaboration with the
local company Biomasa Chile.
The main enabling factors for MNC-Gen investments in renewable energy generation
in Chile are its fast economic growth and high electricity prices. The Chilean market is
very different to other Latin American markets where MNC-Gen operates, such as Peru
and Brazil, as the operations in each country are shaped by local regulations, local
incentives for renewable energies and local renewable energy resources. Chile’s liberal
economy calls for the implementation of the lower cost technologies, as no industrial
policy favours some technologies over others. Chile’s high electricity prices allow the
feasibility of lower cost renewable energy projects without public support.
Law 20.257 has had some impact on MNC-Gen’s renewable energy investments.
Renewable energy certificates from the Monte Redondo wind farm are used to meet
the 5% renewable energy quota requirement set by the law. They are also considering
further investments in wind and solar energy, but Law 20.257 falls short of
expectations and it is too weak to promote a real development or renewable energy
technologies. The interviewee believes that a national carbon market would have been
more effective in achieving large CO2 emission reductions.
At the international policy level, the interviewee considers that the impact of the CDM
on promoting renewable energies in Chile has been very limited. Carbon income is
currently considered a supplement to electricity sales but is not essential for
investment decisions. Carbon prices are very volatile and the process of CDM
registration is too long and costly to be taken for granted. They reduce the risk of
carbon price volatility by selling carbon credits to their European power generation
plants, which need them to comply with EU ETS. The Monte Redondo wind park
started the process of CDM registration in March 2009, when the PPA had already
been ensured.
Technology effects
The main spillover effects of GDF low-carbon activities in Chile will be technology cost
reductions, through learning-by-doing, and demonstration effects, by reducing the
technical uncertainty of technologies previously unknown in the country, and hence
reducing the cost of credit. The opportunities for knowledge spillovers to local
companies are limited, as there is not a local industrial sector that can provide the
high-tech and low-tech equipment needed for the new renewable energy initiatives.
As such, project developers need to rely on foreign providers, and therefore they do
not contribute to the development of a local industry.
Industrial policy is considered the most important policy gap in Chile. The Chilean
Government does not seem to have a clear view of the renewable energy areas where
138
it has a competitive advantage and where a local industry could be developed.
Government initiatives to support the development of renewable energy generation
will only benefit foreign companies if they do not come together with industrial policy
to support local innovation. Law 20.257 does not discriminate between technologies
and therefore incentivises the use of low-cost and mature foreign technologies,
instead of the development of a local industry. The interviewee believes low-cost
foreign technology developers from, for example, China or Brazil will benefit most
from the current Chilean policies.
Chile itself could benefit highly from a long-term industrial strategy for the renewable
energy sector. It should analyse opportunities to start-up transformation activities and
to support different technologies. The country could depart from something simple,
such as copper wiring, and build up from that activity so that it can reduce its foreign
industrial dependence and take advantage of opportunities to add value into the new
renewable energy industry.
4.3.11 CASE 10: Mainstream: raising awareness to unleash wind generation’s potential
Mainstream Chile was created in March 2009 as a joint venture between Ireland-based
Mainstream Renewable Power (MRP) and the Chilean company Andes Energy. The two
partners met at a renewables finance forum held in Brazil in 2008.
Andes Energy (AE) was founded in March 2007 by a group of Chilean partners. The
company required a strategic investment partner to develop a wind project portfolio
of around 300MW in the short term, and then to expand to a total of 800 MW in the
longer term. AE expected that their project portfolio would generate over US$ 106
million in cash flow, with an NPV of over US$ 30 million and an IRR of 15%.
Mainstream Renewable Power (MRP) was founded in Ireland in 2008 to develop, build
and operate wind energy, solar thermal and tidal plants. Once developed, plants are
usually sold to utilities, energy-intensive corporations or pension funds. MRP
sometimes continues to operate the plants after transferring the assets to their
owners. MRP partners with governments, utility companies, developers and investors
across Europe, North America, South America and Australia. Its partnership with AE
was the first of several other partnerships with local companies in Canada, South
Africa, the United States and Europe. MRP invested $2million to become the majority
shareholder of the joint venture, with 90% of the shares.
139
FIGURE 52- MAINSTREAM CHILE TECHNOLOGY TRANSFER PROCESS

Partnerships with local companies are crucial in Mainstream’s global investment
strategy if it is to gain access to local knowledge and expertise on the ground37.
Through their joint venture with AE, Mainstream could access the relevant know-how
about the Chilean electricity market and the regulatory framework for the
development of renewable energies. It could also access a wind project portfolio of
300MW, with the potential to increase to more than 800 MW. The Chilean team of
nine people is highly qualified, and the Chilean Director is heavily involved in lobbying
activities to improve the renewable energies regulatory framework through the
sectoral association ACERA.
The interviewee considers that the main enabling frameworks for local technological
inputs for wind power plant development are a good level of technical education and
previous experience in more complex energy generation and mining technologies.
AE did not have any hands-on experience in the implementation of wind power plants,
and therefore required not only a partner´s funding but also its knowledge in project
financing, project engineering, the procurement of equipment, construction, operation
& maintenance38. Being new to the renewable energies business, the Chilean partners
needed to work with an experienced partner to gain credibility among financial
37
“Andes Mainstream testing the winds of Valparaiso”, Green Momentum, December 16, 2008,
http://www.greenmomentum.com/wb3/wb/gm/gm_content?id_content=817
38 th
Andes Energy, S.A. Corporate presentation at EuromoneyEnergy Events, 7 April 2008.
140
institutions and potential clients. Mr Escobar considers that the wind market is difficult
to penetrate without the relevant experience.
Additionally, all the equipment they will need when they start construction of their
wind power plants will come from abroad, as Chile does not produce wind turbines.
Mainstream does not have agreements with any specific provider and selects the best
technology for local conditions at lower prices. Currently, their priority in Chile is to
create a critical mass of wind power plants that have a good technical performance, so
that the risk profile of these projects is reduced, thus making it easier to access private
bank finance. For this reason, they are interested in importing well-proved, mature 1-
3MW wind turbines rather than state-of-the-art 6MW turbines.
The main channels of TT in Mainstream Chile’s activities are:
• Foreign direct investment. TT from Mainstream headquarters to their Chilean
subsidiary happens continuously through their control of the company.
• Foreign staff. One Irish expert has been integrated in the Chilean team of nine
people.
• Equipment imports. Mainstream Chile has imported wind monitoring
equipment, as recommended by the Irish experts, in order to understand wind
conditions in the north of Chile. Additionally, they will import all wind turbines
required for the construction of their projects.
MRP´s investment in Chile was driven by an interest in the growth potential of
emerging markets. Latin America was one of its target markets and Chile was seen as a
low-risk country in which to start operations on the South American continent due to
its small size, good macroeconomic conditions, high growth, solid financial institutions,
low corruption, open economy and high electricity prices. The free market philosophy
of the Chilean economy is also positive for project developers in that it encourages
strong competition from foreign suppliers. European, Chinese and Indian wind turbine
providers have already opened commercial offices in Chile, offering very competitive
prices. As a result, high quality technologies are available at low prices, improving the
feasibility of renewable energy projects.
Technology output
The main objective of Mainstream Chile is to develop 400 MW of wind energy in the
country before 201539. The total investment commitment to develop this capacity was
$1billion. At the moment, Mainstream Chile has a wind projects portfolio of 860 MW,
with all the projects still under development. The four projects at the most advanced
stages are Alcones (35MW), Negrete (60 MW), Laguna Verde (27 MW) and Chiloe (100
MW), but none of them had started construction at the time of the interview. Laguna
Verde has had the environmental permit that authorises construction since 2007,
while the other projects have gone through feasibility studies but have not yet
received the permits required for construction.
39
“Andes Mainstream testing the winds of Valparaiso”, Green Momentum, December 16, 2008,
http://www.greenmomentum.com/wb3/wb/gm/gm_content?id_content=817
141
Mainstream Chile’s main barriers to achieving the intended output are the lack of
finance from local banks to develop the projects and the difficulty in signing PPAs with
final consumers or distributors. Enabling factors are required to build a substantial
demand for wind technology in Chile that allows TT processes to take place, but the
existing demand-pull policies are considered insufficient to promote a national wind-
power industry. Mr Escobar believes that the under-development of the wind industry
in Chile is due to the lack of local demand, not to the lack of local capabilities, whereby
high local demand would automatically create an industry of auxiliary services with
high value-added benefits.
Law 20.257 created an initial level of hype and attracted the attention of foreign
investors in renewable energies. However, it has failed to create a substantial market.
At the date of the interview, the only four wind power plants in operation belonged to
three large foreign companies: Endesa, SN Power and GDF-Suez. The oligopolistic
structure of the electricity generation industry is considered one of the main reasons
for the failure of Law 20.257,as it does not promote competition among renewable
energy project developers and instead requires that the large electricity companies
buy their 5% renewable energy quota through bilateral contracts instead of in a liquid
market. So, instead of paying their competitors for the renewable energy attribute,
oligopolistic companies usually prefer to pay a fine when they do not develop the
renewable energy project themselves. In addition, the law has set very low renewable
energy generation targets, especially when taking into account that only a share of the
total power supply contracts is required for the 5% of NCRE generation. The high
volatility of electricity prices in Chile also introduces a high level of uncertainty for
small generators, which also benefits large electricity generation companies.
Mr Escobar denounces the large electricity generation companies’ campaign to
discredit renewable energies by spreading misleading information about their
incompatibility with Chilean economic growth targets. Mainstream actively
participates in ACERA’s lobbying efforts to increase awareness and provide accurate
information about the benefits of renewable energies to the government and the
general public. He believes that these activities are working, as the general public
increasingly demands clean energy instead of coal- and gas-fired thermal plants.
Mr Escobar has a negative perspective about the CDM, which he believes has not
contributed to eliminate any of the barriers to wind development in Chile. CDM
registration is a long, costly and cumbersome process, and projects cannot count on
carbon income as part of their revenues. As an example, SN Power´s project, which
started operations in late 2009, was not registered as CDM until December 2010.
Credits will only be issued after this date, which could involve around two years of
waiting for a carbon income after the start of operations. Wind power companies can
therefore not count on carbon credits to top up electricity tariffs – only large
companies with solid financial back-up can afford the time and resources required to
register projects such as CDM, and these are the companies that need the additional
income the least. National policy is much more important than international
agreements for project developers, so Mr Escobar believes that international
negotiations will only have a significant impact in developing countries if they can
influence national policies.
142
Technology effects
If Mainstream projects get to be implemented, they could lead to:
• Technology cost reductions, through learning effects.
• Credit cost reduction, as financial institutions reduce the risk profile of wind
projects in Chile when they deliver a good technical performance.
• Emission reductions, as they displace electricity generated by fossil fuels.
Mr Escobar believes that an increasing demand for wind technology could create a
national industry of components and ancillary services. Chile should focus on those
products where it has a competitive advantage, such as the production of large
metallic components for wind towers, with high transportation costs, and the
production of copper rotors for gearless turbines. Both activities are not complex
technically but have a higher value-added than the natural resources-based industries
in which Chile currently specialises. Copper availability and the mining sector´s
experience in the metal-mechanic industry would provide a comparative advantage in
starting these activities. There are also opportunities for solar energy, as Chile has high
solar radiation and enough capabilities to produce solar PV modules. These activities
could take off if there was a significant local and regional market.
4.4 Discussion
4.4.1 What constitutes climate change technology transfer, and how does it happen?
The process of climate change technology transfer revealed by our ten case studies
invariably starts with the identification of an opportunity to generate new revenue
through technological activities that ultimately lead to the reduction of GHG emissions.
The mere identification of that opportunity requires in most cases direct exposure to
international technological developments. This happens through, for example,
participation in international fairs, reading international publications, visits to foreign
institutions, contacts with clients and competitors through exporting activities, or
contact with the head office in the case of foreign subsidiaries. Electrica Nueva
Energia’s biomass cogeneration and Universidad de Chile’s micro-hydro plug & play
projects were exceptions in which Chile’s accumulated know-how in biomass and
hydroelectricity technologies seemed sufficient enough to instigate new ideas for
technology development or implementation.
With the identification of new revenue-generating opportunities came the definition
of the intended outputs to be achieved. In our case studies, these included:
• The production of wind blades.
• The supply of sodium and potassium nitrates for storage tanks in CSP plants,
and the supply of more sophisticated products and services for CSP storage,
such as molten salts mixing, melting and feeding systems.
• Scientific publications and patents about the use of lithium for electric vehicles’
batteries, leading to commercial products such as Li-ion compounds for
batteries and ultimately Li-ion cells and batteries.
143
• Patents and prototypes for a new micro-hydro generation concept, leading to
the commercial production of enhanced micro-hydro turbines.
• Landfill gas capture, flare and electricity generation.
• The sale and installation of biomass boilers for the industrial and residential
sectors, as well as the sale of thermal energy generated.
• The generation of heat and electricity with forestry biomass for sale to
industrial clients and the national grid.
• The supply of biomass feed-stocks from forestry waste for their use by the
forestry, pulp and paper sectors.
• The generation of electricity with renewable energy sources, including solar
thermal, wind and biomass.
• The development of wind generation capacity.
To deliver the intended outputs, each organisation needed to take stock of the
technological inputs available. These included both local and foreign inputs. Foreign
inputs are required when local inputs are not available, when they cannot compete
with international alternatives or when foreign multinationals prefer to use their own
resources or trusted foreign product and service providers instead of unknown local
alternatives.
A solid base of available local technological inputs was crucial for local organisations
embarking in low-carbon technological activities, alone or in joint-ventures, with
foreign companies. Local technology supply elements included:
• Skilled human resources, including engineers and scientists, often with
qualifications obtained in foreign universities or through work experience
abroad. Chilean universities could also provide personnel with specific R&D
capabilities in renewable energy and R&D infrastructure for carrying out
performance tests and developing prototypes.
• Know-how in some of the aspects of the new technology. Know-how was
typically acquired through experience in well-established sectors of the Chilean
economy, mainly mining, forestry and pulp and paper. Activities in the biomass
and hydroelectricity sectors could benefit from extensive experience in Chile,
driven by the widespread availability of these resources. The pre-existing
knowledge base meant that the new activity could start from the solid base of a
reliable supply chain, a network of clients and employees familiar with similar
technologies. Additionally, the proven success in more mature industrial
sectors of Chile reduced the risk profile of local companies, giving them more
favourable access to credit.
• Access to natural resources. Being a natural resources-based economy, Chile’s
comparative advantage often originates from the availability of unique or low-
cost natural resources. Such was the main input for the technology transfer
processes related to molten salts for CSP plants and lithium uses for electric
vehicle batteries. Widespread access to hydro and biomass resources also
drove the initiatives related to biomass cogeneration, the supply of biomass
feed-stocks, the sale and installation of foreign biomass boilers and the
development of a micro-hydro turbine.
• In-depth internal learning processes, with or without external assistance. Local
companies expanding their activities to the renewable energies or low-carbon
144
sector typically went through intense learning processes about the renewable
energy market, the efficient operation and maintenance of foreign equipment,
the technological development of imported technology to adapt it to Chilean
conditions, product improvement and logistic systems to meet clients’
specifications or connection to the grid. Even though they come from very
different industrial sectors, many of the interviewed local companies belong to
the Chilean association ACERA, which gathers together the renewable energy
industry to influence policymaking and in turn to support renewable
generation.
The use of internal inputs seemed less important when the technology transfer
process was initiated by foreign MNCs operating in Chile. One MNC heavily relied on its
foreign workforce, expertise and suppliers to implement renewable energy projects,
and in two other cases MNCs showed a more de-centralised structure, which was very
reliant on Chilean staff’s knowledge about particular conditions of the Chilean market.
Internal inputs into the technology transfer process were very important in most of the
cases, but never sufficient to achieve the final intended output and additional
technology effects. Foreign technological inputs were always required, including:
• Production equipment. All the projects required foreign equipment, as Chile is
not a producer of heavy or light machinery. These include German equipment
for the wind blades manufacture, heavy machinery for the extraction and
refining of salts for CSP storage, Chinese hydro turbines, United States, Dutch
and Austrian equipment for landfill gas capture, flaring and electricity
generation, Austrian biomass boilers, a Brazilian biomass cogeneration system,
United States wood-crushing machines for biomass transformation, German
CSP technology and Danish wind turbines. Equipment came together with
engineering blueprints for their installation, operation and maintenance. In
some cases foreign engineers were required for the installation of the
equipment, such as in MNC-Gen’s wind farms or KDM’s landfill gas capture,
flare and electricity generation system. In cases such as Electrica Nueva
Energia’s cogeneration plant or Energia del Sur’s biomass boilers, installation
was carried out by Chilean companies.
• Skills and know-how for operation and maintenance. The most mature
technologies, such as the biomass cogeneration plant, the hydro-turbine or the
extraction and refining equipment for salts did not require significant foreign
support for their efficient operation, as the relevant knowledge was already
available in Chile. The less established technologies in the country required
foreign support to place local recipients at the technological level required to
operate the technology efficiently. Foreign support was facilitated through
several means such as consultancy assignments to foreign experts, partnerships
with foreign companies, training programmes by technology providers, or
through the continuous support of the head company in joint ventures or fully
owned foreign subsidiaries. In some cases the relevant knowledge stayed
abroad, when maintenance contracts were signed with foreign suppliers, or
when the installation, operation and servicing of the technology stayed within a
foreign MNC and did not permeate to local companies or staff.
145
• Product designs and specifications. The production of wind blades or molten
salts required an in-depth knowledge of clients’ specifications. A certified blade
design is a prerequisite to win contracts with wind turbine manufacturers.
Fibrovent needed a licence from the Spanish company Eozen to access a design
it could use for production, while molten salts specifications were developed by
SQM in collaboration with a United States foreign research laboratory or with
client CSP project developers.
• Knowledge, expertise and experience for generating and managing technical
change. This type of foreign input enables local companies to develop their
own technologies by using scientific and technological capabilities. Only a few
of the analysed case studies aimed at reaching the stage of developing their
own technologies, namely Fibrovent, Universidad de Chile’s hydro project and
the Lithium Innovation Centre. The remaining case studies either provided
processed raw materials (biomass feed-stocks or sodium and potassium
nitrates) or electricity using existing renewable energy generation technologies
(landfill gas, biomass cogeneration, biomass thermal energy generation, wind,
solar). Capabilities for generating and managing technology cannot be
transferred through market-like exchanges and must be internally learned. In
some of the case studies they are fully acquired through internal efforts, such
as in the micro-hydro plug & play project. In other cases external support is
needed through, for example, hiring foreign staff – as in the case of Fibrovent –
or collaborating with foreign institutions at the forefront of EV research, as
intended by the LIC.
The main channels for the transfer of foreign technological inputs have already been
mentioned above. Highly integrated channels include foreign direct investment with
centralised headquarters or turnkey contracts for equipment installation. In these
cases, the knowledge tends to stay with the technology provider, even though the
equipment is operated locally. Less integrated channels are the imports of equipment,
licensing, collaborative R&D with foreign institutions or companies and ad hoc
technical support from foreign experts. Disintegrated channels require the more active
participation of the recipient and typically lead to an improvement in its technological
capabilities.
Along with their intended output, technology transfer processes can deliver other
beneficial effects. The analysed case studies reveal the following:
• Emission reductions in the recipient’s economy, through the efficient use of the
transferred technology. In the cases where the intended output is the
generation of renewable energy or the capture of GHG gases, emissions
reductions are a direct effect of the process. In some other cases, they will only
happen when the final product is implemented in clean energy activities.
• Vertical knowledge spillovers through backward and forward linkages with local
suppliers and clients. These are non-pecuniary externalities that arise when
productive knowledge permeates into the local economy through, for example,
technological assistance provided by foreign clients to improve the quality of
the product provided by local suppliers. This was the case with SQM, which
received product specifications and regular feedback on product performance
from the US Sandia laboratory and the Spanish- and US-based CSP project
146
developers. Fibrovent can also expect such type of spillovers once it starts
producing wind blades and selling them to foreign turbine manufacturers.
Energia del Sur provided significant technological assistance to pellet providers
and installers to improve the quality of their products and services. Knowledge
spillovers can also lead to improved processes, as would be the case if SQM
undertakes the higher value-added activities mixing, melting and feeding salts
for development. In any case, backward linkages were rare in the analysed case
studies, as equipment and components were usually provided by foreign
companies. The de-linking of the Chilean economy is considered one of the
main barriers to TT.
• Horizontal knowledge spillovers, as the acquired capabilities are used to
implement other clean energy projects. These have taken place in the case of
KDM, which plans to capitalise learning in landfill gas generation with new
projects in other landfills as well as renewable energy generation with other
technologies, mainly hydro. Biomasa Chile has also devised an initiative for new
biomass uses in co-firing, based on its experience in forestry waste recovery
and transformation.
• Productivity increases in the recipient country. When the technology transfer
processes lead to an increase in the value-added by recipient technologies, or a
reduction of the production costs, the recipient country can benefit from
productivity increases which lead to increases in the real income. The use of
more efficient foreign technologies typically facilitates this process. Some of
the analysed case studies involve upgrades towards high productivity activities,
mainly those involving the manufacture of equipment or components. The
productive use of forestry waste by Biomasa Chile also involves an increase in
productivity, as well as SQM’s potential provision of services for CSP plants
beyond the mere supply of salts for heat storage.
• Technology cost reductions, through economies of scale, learning and
demonstration effects. Projects that introduce new technologies to Chile
contribute to lower production costs as a result of technological learning
curves. The good technical performance of pioneering projects also reduces the
risk profile of technologies unknown to local banks and reduces the cost of
getting credit. Technology cost reductions through learning effects were not
assessed in the case study analyses because none of the cases has achieved a
significant capacity allowing for these effects. However, learning effects are
validated by several publications (Qiu and Anadon, 2011; Grubler et al., 1999;
Goldemberg et al., 2004; Junginger et al., 2005). Technology cost reductions
can also be achieved through economies of scale when the implementing
company has a large production capacity. Furthermore, companies backed by
their financial strength or market power, such as MNC-Gen or KDM, are more
likely to implement large-scale renewable energy projects and hence achieve
technology cost reductions.
• An increase in local technological capabilities. The case studies in which local
actors are involved in intense learning projects deliver increased technological
capabilities that create the necessary knowledge base for transfers of
increasingly sophisticated technologies and the generation of endogenous
technologies. Highly integrated technology transfer processes do not achieve
147
this effect and can lead to the technological dependence on more developed
countries.
The technology transfer processes revealed by the case study analysis do not allow for
a straightforward conceptualisation. As defined in the literature, foreign technology
transfers can either increase the production capacity of the recipient with new
equipment and knowledge to operate it, or move a step ahead to enable technological
change (Bell, 1987; Wei, 1995; Ockwell et al., 2008). Both types of transfer, however,
are necessary to advance towards a low-carbon economy. The second type has a
higher potential to contribute to the sustained development of the recipient economy,
as it gets involved in higher value-added activities, leading to higher productivity and
living standards. It requires a solid technological base and intense learning processes in
the recipient. The first type of technology transfer, when foreign equipment and
operational know-how are transferred, is required to achieve a significant scale of low-
carbon technologies that can deliver emission reductions and cost reductions through
demonstration, economies of scale and learning effects. Foreign MNCs usually have
stronger financial and technological capabilities to deliver scale, particularly in the
most innovative technologies, even if their integrated technology transfers often do
not create knowledge spillovers in the recipient economy. The most mature
technologies are usually transferred through disintegrated channels, mainly imports,
and installed, operated and maintained locally. In our case studies, opportunities for
local technological change arise in the improvement of mature technologies or the
production of the less sophisticated components of new technologies.
Successful technology transfers require that foreign technology inputs, together with
local inputs if relevant, deliver their intended output and, even better, convey further
impacts to the recipient economy. If these outputs and effects are not realised, flows
of foreign technologies cannot be considered as “transferred”. Six of our ten case
studies managed to deliver their intended output. None of these cases involved the
creation of new technologies, but rather their implementation or the supply of raw
materials. Moreover, in most cases, a process of adaptation of the technology to
Chile’s conditions was required. The implementing companies were important players
in the Chilean energy, mining, forestry or waste management sectors (MNC-Gen, SQM,
Indef and KDM), or smaller companies selling or using biomass generation equipment
(Electrica Nueva Energia, Energias del Sur).
On the other hand, foreign technologies were only responsible for one part of those
outputs and effects achieved, with the remaining attributable to local technology
inputs, hence the difficulty of measuring technology transfers through inputs, outputs
or effects.
doing so?
As highlighted by the literature reviewed in Chapter 2, measuring technology transfer
is inherently difficult and only imperfect indicators can be used as proxies of its tacit
nature. Following the approach proposed in this chapter for the case study analysis,
we can measure three different aspects of the technology transfer process: the foreign
technology flows constituting inputs into the TT process, the technology outputs
148
resulting from these inputs and the technology effects conveyed beyond the intended
output.
Some potential indicators of the channels through which foreign climate change
technologies have entered Chile, as seen in our case studies, which can be used as
proxies of the foreign input element of TT, are:
• The value of imports of equipment or services for the generation of heat and
electricity from renewable sources, the reduction of energy consumption or the
capture and destruction of GHG emissions.
• The value of FDI specifically related to the activities above.
• Fees paid to foreign employees in industries related to the activities above.
• Fees paid to or the number of licences bought from foreign companies related
to the activities above.
• Travel expenses related to technological missions abroad.
• R&D expenses incurred by foreign companies or by local companies in
collaborative projects with foreign companies.
• Expenses in training programmes by foreign consultants.
• The number of joint ventures between local and foreign companies to develop
or implement low-carbon technologies.
Most of these indicators do not exist at the national level, especially for developing
countries, but could potentially be obtained at the project level. Besides, some other
channels of foreign technology input are difficult to value, such as access to foreign
technological information publicly available in scientific publications or the internet, or
the value of the education received abroad by some of the local staff. An additional
problem is that these inputs do not always deliver their intended outputs, yet when
they do so it is difficult to evaluate how important their contribution has been
compared to local inputs.
The output element of the TT process can be measured as the economic or unit value
of the products of services that have required foreign technological inputs. Some
suggested measurements would be:
• The MW of renewable electricity generation that have required foreign
technology input for their installation, operation and maintenance.
• The GWh or TJ generated from renewable energy sources that have required
foreign technology inputs for their installation, operation and maintenance.
• The GWh or TJ saved through the implementation of energy efficiency
technologies that required foreign input for their installation, operation and
maintenance.
• The tCO2e captured or destroyed with foreign equipment and know-how.
• The units or value of products and services that will be used for low-carbon
energy generation, energy efficiency or the capture and destruction of GHGs
that have required foreign technology inputs.
• The number or value of licence revenue from patents that required foreign
technology inputs into the R&D process.
The obvious problem with this approach to measuring TT is that it is not possible to
define the part of the output that is attributable to foreign technology transfer. In the
149
case of Chile, where virtually all equipment is imported from abroad, every output
would have had to be considered as a result of TT processes. However, some products
could have only required imports of machinery, while others would have required the
more intensive involvement of foreign companies in product design and the
installation, operation and maintenance of the production equipment. Both products
would therefore not incorporate the same amount of technology transfer. A potential
alternative is to measure TT in terms of value added to the cost of the original
equipment.
Technology transfers do not only convey the production of specific products or
services, as they have additional impacts such as enabling emission reductions,
increasing the productivity or the recipient as well as for other local companies
through knowledge spillovers, resulting in technology cost reductions, inspiring the
birth of local companies to deliver the same products or services. As with the output of
the TT process, attributing effects that are a result of TT is complicated because not
only foreign inputs, but also a number of local factors may have contributed to
bringing them about. Isolating the “foreign” part of the effect is difficult. In any case,
some of the proposed measurement approaches are:
• Emissions reductions. Some projects achieve emission reductions the moment
they start operating, either directly such as in the capture of landfill gas, or
indirectly by replacing electricity generated from fossil fuels with renewable
electricity. Their impact on emissions reductions can be measured as tCO2
reduced. In other cases, the projects enable emission reductions later on as
part of larger projects. This would be the case for the production of molten
salts for heat storage in CSP plants, the production of Li-ion batteries for EV,
the production of biomass feed-stocks, the sale of biomass boilers or the
production of wind blades. The amount of emission reductions they would
enable could be used as an indication of the effect of the technology transfer.
At the national level, previous models analysing the impact of foreign
technologies on carbon emissions reductions have not succeeded due to a lack
of specific data on foreign low-carbon technology inputs. If those data were
made available, they could provide valuable information about the effect of
foreign input of low-carbon technologies in countries with different conditions.
• Energy intensity reduction. As with emissions reductions, projects incorporating
foreign technologies can lead to improvements in the energy intensity of the
recipient country. The problem with its measurement is similar to the one
above.
• Productivity gains for the recipients. The implementation of foreign
technologies can improve the productivity of the recipient country, by helping
it to deliver more value but with less input of capital and labour. The impact of
foreign technology on productivity can be measured at the company level by
comparing the output generated before and after the implementation of
foreign technologies, keeping capital and labour constant. At the national level,
several publications were reviewed in Chapter 2 in order to explain national TFP
by domestic and foreign technology, where the latter was proxied through FDI
or imports, controlling for the impact of other variables. If climate change-
specific data on FDI and imports were made available, similar models could be
150
built to analyse the specific impact of foreign climate change technologies on
productivity.
• Knowledge spillovers through backward and forward linkages with domestic
actors. Local content is the most commonly used proxy for backward linkages
(UNCTAD, 2001), and it can be measured as the share of total components,
intermediate products and ancillary products and services used as inputs by the
company involved in TT processes that are produced locally.
• Horizontal spillovers. The impact of the TT process on encouraging new projects
or companies can be assessed ex-post as the number of new companies or new
installed capacity that have benefited from the previous experience.
• Technology cost reductions. The cost of the final product delivered40 when TT is
required could be compared to local alternatives. Several factors have an
impact on the cost of the final product. To understand the impact of foreign
activity in the cost of the product a model could be constructed, as devised in
Qiu and Anadon (2011), that would explain each provider’s cost per unit of
product though several variables, among which would be the adoption of
foreign technologies.
Evidence shows that measuring TT through only its input, output or effect elements
results in an incomplete picture of the process. None of these elements is able on its
own to represent what the TT process conveys, so a comprehensive approach is
recommended to measure and evaluate the TT component of low-carbon projects.
Such an approach would require creating a composite indicator that reflects the input,
output and effect elements of TT.
Table 13 compiles the suggested indicators necessary to construct the composite

measurement of TT at the project and national levels.
TABLE 13- APPROACHES TO MEASURE CLIMATE CHANGE TECHNOLOGY TRANSFER BASED ON CASE
STUDY ANALYSIS
INPUT OUTPUT EFFECT

• Value of imports • Installed capacity • tCO2e reduced
Project • Value of foreign investment • kWh of kJ generated or saved • Technology cost-
level • Fees paid to foreign service • tCO2e captured or destroyed efficiency differential
providers or employees • Units or value of with widely available
• Number or value of licences technological products and technologies locally
brought to foreign services • Productivity gains
companies • Number or licence revenue • Local content
• Travel expenses for from patents • Number of projects or
international technological • Number of scientific installed capacity
missions publications resulting from
• R&D expenses in learning effects of the
collaborative projects with project
foreign companies
• Value of CC-related imports • MW installed by foreign • tCO2e reductions
Country • Value of CC-related FDI owners or with foreign explained by foreign
level • Number of foreign workers technologies CC-related inputs and
40
Products could include installed capacity in kW, electricity generation in kWh, thermal generation in
kJ, energy savings in kJ or kWh, emissions reductions in tCO2e.
151
INPUT OUTPUT EFFECT
in the CC technology sector • kWh of kJ generated or saved control variables
• CC-related licence by foreign-owned projects or • Cost of clean energy
payments to foreign with foreign technologies explained by foreign
companies • tCO2e captured or destroyed CC-related inputs and
• Foreign companies R&D by foreign-owned projects or control variables
expenses in CC-related with foreign technologies • TFP of CC-related
sectors • Production of CC-related industrial sectors
technologies or products by explained by foreign
foreign owners or with CC-related inputs and
foreign technologies control variables
• Revenues from local patents
developed with foreign
support
• Number of local scientific
publications in collaboration
with foreign institutions
Under the proposed comprehensive approach, different ratios of outputs and effects
regarding foreign technology inputs would indicate the varying importance of local
technology inputs, as well as other control variables. Countries and companies with a
solid local technological basis would be more efficient in turning foreign technology
flows into final outputs and effects. To allow for comparability between different
projects, this measurement approach should be used separately for each technology
type and then use indicators with comparable units in each of the TT elements.
Data required for the national level analysis, on specific climate change technology
inputs and on the use of foreign technologies to deliver outputs, are often not
available, so it is not possible for most countries to undertake measurements of
climate change-related TT activities at the national level.
4.4.3 Empirically, which factors have enabled climate change technology transfer to Chile?
A set of factors have influenced each of the TT elements in our case studies, enabling
the availability of local technology inputs that support the efficient adoption of foreign
technologies, the host country’s access to foreign technology inputs through channels
of TT, the transformation of technology inputs into outputs and the achievement of
further technology spillovers beyond the intended outputs.
The main enabling factors of local technology inputs into the ten case studies have
been:
• Competitive industrial sectors enabling cross-sectoral knowledge spillovers.
Expertise gained in the mining, forestry, pulp and paper sectors was crucial for
the TT processes initiated by Fibrovent, SQM, the Lithium Innovation Centre,
Electrica Nueva Energia and Biomasa Chile. The mining sector is particularly
significant as a source of technological capabilities in the country. Competitive
local suppliers to the mining sector have the potential to start medium-tech
renewable industries, such as Fibrovent’s wind blades or the recently
announced wind turbine towers by the metal transformation company
Metalurgica Arrigoni (Renewable Energy Magazine, 2011). Both blades and
towers have high transportation costs and Chile has acquired a knowledge base
152
to produce them in-house through the supply of composites and metal
structures to the mining sector. For these cross-sectoral knowledge spillovers
to happen, it is necessary that strong industrial sectors in the economy develop
backward linkages with local suppliers.
• Access to natural resources. Many of the TT processes analysed were
“resource-pushed”. Access to a wide range of mineral reserves drove the
activities of the Lithium Innovation Centre and SQM’s production of salts for
the solar energy industry. The availability of vast, unused hydro resources
drove UChile’s micro-hydro plug & play project. Large biomass resources drove
the TT processes of Biomasa Chile, Electrica Nueva Energia and Energia del Sur.
In addition, high levels of solar radiation in the north of Chile have raised MNC-
Gen’s interest in solar thermal energy in the country. The Spanish company
Solarpack, which could not be interviewed for our analysis, has also been
attracted by Chile’s solar resources. It has developed a 1 MW solar PV plant and
is constructing two additional 9MW plants in the Atacama Desert41.
• Human capital. The availability and quality of locally trained engineers and
scientists was mentioned by all but one of the interviewees as one of the key
enablers for their TT activities. The interviewee that disagreed on this point
argued that, while Chile has very well trained engineers, it lacks intermediate
qualifications. Experienced local engineering firms supported the
implementation of KDM’s electricity generation with landfill gas and Electrica
Nueva Energia’s biomass cogeneration. Local firms also supported Universidad
de Chile in finding an appropriate supplier of a micro-hydro turbine. However,
MNC-Gen found that Chile could only deliver an either over- or under-qualified
workforce, missing the intermediate layer of technicians that is essential for
transformation activities.
• National innovation system. Several companies involved in TT processes
collaborate with local universities through joint R&D in the relevant
technologies. Fibrovent has signed a collaborative R&D agreement with a local
university to adapt its imported blade design to the conditions of the Chilean
wind. SQM collaborates with a university spin-off for the measurement of solar
radiation in the north of the country. MNC-Gen belongs to a consortium
including local universities for bioalgae R&D. The Lithium Innovation Centre is
established as a public-private partnership between Universidad de Chile and
several private companies and is looking to incorporate additional universities
into its membership. Biomasa Chile also collaborates with local universities for
R&D in the forestry sector, although it has not done so for the specific case
included in this thesis.
• Public funds to support R&D. Only the bio algae R&D project, with MNC-Gen
among the different participants, has so far received public funds to support
technological capabilities. Universidad de Chile expects to receive funds from
CONICYT to develop its micro-hydro plug & play project. In all the other cases,
R&D, when required, is privately funded.
41
“Waking up to the sun´s rich potential” 26 November 2010, Chile Special Report in Recharge News,
www.rechargenews.com
153
Several factors operated to facilitate the flow of foreign technology inputs through
different channels, mainly FDI, imports, licences, foreign workers or local workers
trained in foreign institutions. The main enablers at this level were:
• Chile’s openness to trade and foreign investment. Low tariffs and free-trade
agreements reduced the cost of importing equipment and services. Favourable
conditions for foreign investment attracted FDI, which sometimes led to TT.
• Low income tax. This allowed for competitive remuneration packages for
foreign workers, who can transfer their knowledge acquired in more
technology developed countries.
• Access to foreign education. Many of the local actors involved in TT had studied
abroad and could bring foreign expertise to their Chilean innovation activities.
In other cases, some Chilean workers had gained experience abroad by working
for foreign multinationals.
• Strong institutions and low corruption. Several foreign organisations investing
in Chile mentioned a favourable country risk profile as one of the main
attraction factors.
• Good transport infrastructure. It was mentioned by Energia del Sur as enabling
the imports of foreign equipment.
• Access to the Latin American market. The possibility of gaining access to the
wider Latin American market attracted foreign companies Mainstream and
Eozen to Chile. Free-trade agreements with most of their neighbours, and the
possibility of a future common market with Colombia, Peru and Mexico (The
Economist, 2011), is an incentive for investors looking for large markets from
the safe base of a low-risk country.
• Public support policies. Fibrovent counted on public support, channelled by the
programmes InnovaChile and InvestChile, to finance the consultancy
assignment of a foreign expert and initial capital investment. Funds for capital
investments were provided by InvestChile on the condition that the new
venture would have a majority foreign participation. For this reason, the TT
process was initially structured as a joint venture between Fibrovent and the
Spanish company Eozen, although this arrangement proved to be inappropriate
for their needs and was subsequently changed to an arm’s length commercial
relationship in which Eozen would sell Fibrovent a licence to use its blade
design. InvestChile was subsequently modified to allow support for local
innovators, without requiring a majority foreign participation. InnovaChile has
also supported the transfer of solar thermal energy technologies for buildings
financing a capacity building programme in which Spanish experts trained
Chilean companies to install and service the equipment. This programme
enabled a group of 50 to 60 companies to provide services beyond the import
of the equipment.
Foreign technology inputs could turn into outputs when the necessary enabling factors
were in place to create a significant demand. The main enablers in the six case studies
that reached the stage of delivering their intended output were:
• Strong foreign demand. This was the case for the molten salts produced by
SQM for storage in CSP plants built in the United States and Spain. Demand in
these countries was driven by demand-pull policies that subsidise electricity
154
generation with CSP technologies. Chile was selected as a supplier of these salts
for its unique access to mineral resources in the Atacama Desert. A large global
demand for wind technology was also considered by Fibrovent as one of its
main drivers for starting wind blade production in Chile.
• High electricity and fossil fuel prices. These elements drove all investments in
biomass and biogas heat and electricity generation. They were also mentioned
by MNC-Gen, Mainstream and Universidad de Chile as one of their main drivers
for investments in wind, solar and hydro technologies. As shown in Annex 2 of
Chapter 4, onshore wind generation, landfill biogas and large biomass
combustion in Europe are already competitive against oil-based electricity
generation, which makes up more than 50% of the generation capacity in the
SING and close to 20% in the SIC. Onshore wind generation is expected to be
competitive with oil, natural gas and coal by 2030 in Europe. High electricity
prices, together with high solar radiation in the Atacama Desert area, have
enabled the grid parity of Solarpack’s 1MW Solar PV plant. Solarpack has signed
a long-term PPA with the mining company CODELCO to sell them the electricity
generated at a stable price42.
• Fast economic growth. High economic growth rates of Chile in the last decade,
as well as ambitious economic growth targets for the future, require significant
increases in national electricity generation capacity. Energy sources should be
diversified to ensure the security of supply and to avoid shocks like the one
suffered when Argentina blocked gas imports. Sustained growth prospects
were one of the main reasons for MNC-Gen and Mainstream investing in
renewable energy generation capacity in Chile.
• International demand-pull policies. The CDM was essential to ensure
profitability in only one of the ten case studies, on the capture and destruction
of landfill gas by company KDM. Landfill gas destruction only has a value when
there is a price for the methane not released into the atmosphere. Carbon
credits from projects registered as CDM provided this price. KDM also
registered as CDM the activity of electricity generation with the captured
landfill gas. In this case, although carbon credits contributed to increase the
cash flow of the project, they were not essential to ensuring its feasibility, as
high electricity prices already made the project profitable. The biomass
cogeneration project owned by Electrica Nueva Energia has also been
registered as CDM. Although it has been able to receive carbon credits since
January 2009 it still has not requested their issuance, due to the high
transaction costs of the process. Heat and power generation are profitable
without the need of carbon credits. MNC-Gen has also requested CDM
registration of its Monte Redondo wind farm, but admits that carbon credits
are not a relevant factor when analysing the project feasibility.
• National demand-pull policies. Stricter air quality legislation in urban areas
boosted sales of Austrian biomass boilers imported by Energia del Sur. Law
20.257, establishing a green certificates system to promote renewable
electricity generation, was not considered a crucial factor for promoting
renewable electricity generation. Even MNC-Gen and ENE, which could ensure
42
www.solarpack.es
155
the sale of their green certificates, admit that the law has a very small impact
on promoting renewable energy generation in the country.
• Market power. Almost 2,000 MW of wind capacity submitted by several
companies has passed the SEIA process. However, the only functioning large-
scale wind farms in Chile are operated by two powerful incumbents in the
national electricity market – Endesa ECO and MNC-Gen – as well as SN Power, a
large Norwegian electricity generation company with global operations. MNC-
Gen’s powerful position in the market has facilitated a favourable long-term
PPA with a distribution company as well as access to credit to develop the
project. Global market leaders can also benefit from a higher bargaining power
with suppliers of equipment, regarding cost and waiting time for the supply of
equipment.
An enabling environment was also required to achieve technology effects beyond the
intended outputs. Technology spillovers were rare in the analysed case studies
because many projects had still not been able to deliver a technological output. In
other cases, the companies involved in TT processes had sourced equipment and
servicing from foreign companies, preventing knowledge spillovers to local suppliers.
The few projects that reached this stage mentioned as main enabling factors:
• A functioning national innovation system enabling links to local universities.
Fibrovent is transferring part of its acquired knowledge on wind blade design to
a local university, with which it collaborates on R&D. Industry-university
collaboration also has the potential to deliver knowledge spillovers in the cases
of the Lithium Innovation Centre and in the bio algae R&D consortium, once
both initiatives start delivering results.
• Absorptive capacity of the recipient. The previously described factors enabling
the supply of local technological inputs also make it possible for recipients to
absorb the transferred technologies. The absorptive capacity of the recipient
enhances productivity gains as a result of technology adoption and horizontal
knowledge spillovers. Such was the case for KDM, which plans to capitalise on
the steep learning curve it went through to generate electricity with landfill gas
through the development of new renewable energy projects.
• Qualified local suppliers. Energias del Sur could transfer knowledge about the
production of pellets for their use in biomass boilers to local suppliers of
pellets. Although their initial qualification was low and they could not produce
pellets at the required specification, high transport costs meant it was not
feasible to import pellets from abroad. It was therefore more cost-effective to
train local suppliers to meet their specifications, and they are now able to
produce high quality products.
• Local demand. Emissions reductions are a direct effect of the implementation
of low-carbon technologies. Technologies are implemented when there is a
significant demand, and there is demand for renewable electricity-enabled
emissions reductions in the projects implemented by MNC-Gen, Electrica
Nueva Energia, KDM, Energias del Sur and, indirectly, Biomasa Chile.
156
4.4.4 Empirically, which factors have hindered climate change technology transfer to
Chile?
Barriers, or missing enablers, have an impact on every stage of the technology transfer
process. They can deter the supply of local technology inputs that enable the adoption
of foreign technologies, block the flow of foreign technologies entering a country,
impede the achievement of technological outputs or limit the spillovers resulting from
the technology transfer process.
Our case studies showed that the following enablers were missing for the provision of
local technology inputs. These are complemented by barriers revealed in a
stakeholders’ meeting between the CER and the wind industry43:
• Industrial policy that supports the development of renewable energy
technologies. Chile lacks a strong industrial sector that could manufacture its
own renewable energy technologies or provide inputs to foreign suppliers.
Instead, the country favours the static comparative advantages provided by the
abundance of mining, fishing and forestry resources over specialisation in
higher value-added activities. There is a preference for buying foreign
equipment at a competitive price over supporting the higher prices of local one
until it reaches international competitiveness. Chile´s neoliberal economic
model makes it reluctant to support specific sectors and technologies, or to
“pick winners”. SQM, the Lithium Innovation Centre and MNC-Gen noted the
lack of a national strategy to promote renewable energy technologies. Rodrigo
Garcia, from the CER, also referred to the “ad hoc” approach of the Chilean
Government to promoting local renewable energy technologies. This precludes
long-term investment decisions by private companies and the creation of
knowledge clusters in specific technologies. SQM indicated that it would only
get involved in initiatives to implement CSP in Chile once the government
showed clear signals favouring this technology and lowered its techno-
economic risk. The vast solar radiation resources of the north of Chile, located
next to the energy-intensive mining industry and the reserves of molten salts
for heat storage, therefore stay untapped. The Lithium Innovation Centre has
emerged as an initiative of the Universidad de Chile and three private
companies, but it does not count on or receive public support. It must
therefore compete with heavily subsidised international institutions
researching Li-ion batteries for EV. The Chilean Government seems mainly
interested in the extraction and initial processing of minerals instead of related
higher value-added activities.
• A workforce with intermediate technical skills. MNC-Gen highlighted that,
whereas the level of engineering and science education is high, Chile lacks an
intermediate layer of technicians able to undertake the transformation
activities required for industrialisation. At the moment, Chile imports most of
its equipment and is involved in transformation in natural resources-based
industries. The Chilean Government should support education for a medium-
skilled layer of technicians to upgrade to higher value-added and technology
content activities.
43
Centro de Energias Renovables “Desarrollo internacional de la energía eólica y estado de la industria
en Chile”. Results of a stakeholdersmeetingheldonDecember 2010.
157
• A specific knowledge base in renewable energies and in the grid integration of
electricity generated from renewable sources. SQM, the Lithium Innovation
Centre, MNC-Gen and KDM mentioned this as a significant barrier. It was also
highlighted during the CER’s stakeholders’ meeting with the wind industry,
which was held in December 2010. MNC-Gen reacts to this lack of skills by
sourcing most of the design, implementation and servicing activities from
foreign staff or companies, although this precludes any knowledge spillovers to
the Chilean economy. The LIC notes that Chile’s technological distance from the
frontrunners in Li-ion battery research is one of the main barriers to the success
of the centre. SQM is investing to improve its knowledge of solar technologies
by measuring solar radiation levels in the north of Chile and collaborating in
solar energy storage research with US laboratories.
No significant barriers were found regarding the flow of foreign technologies into the
country. Chile’s favourable environment for foreign investment and its free trade
regime enabled exposure to foreign technologies. However, a favourable environment
for foreign investors does not always attract technology-related investments –
exemplified by the fact that the foreign companies analysed in our case studies
implement or commercialise, rather than produce, technologies in Chile. Chapter 3,
which discusses Chile’s enabling environment for technology transfer, shows that
despite a large stock of FDI, foreign affiliates of multinational companies in Chile only
account for a very small share of R&D, at 3.6%, compared to 48% in Brazil, 32% in
Mexico and 23% in Argentina44 (OECD, 2010). Low levels of R&D were partially caused
by flawed and old policies that penalised R&D expenditure in the country45. These
policies have been superseded by a new law to improve incentives for private sector
innovation. Chile has also introduced subsidies to support high technology FDI, but
these subsidies can crowd-out local innovators or increase their dependence on goods
and services from abroad, as shown in the case of Fibrovent, which initially needed to
give up control of its wind blade production venture to gain access to InvestChile
funds, which required a majority foreign participation. Intellectual property rights were
not considered by any of the interviewees as a barrier to the flow of foreign
technologies, partly because of a lack of awareness about IPRs and also because many
of the case studies did not involve the creation of new technologies but their use. IPRs
are relevant in the context of the passionate debate on IPRs confronting developed
and developing countries in the UNFCCC negotiations (as detailed Chapter 1’s
introduction to the international policy context).
Most of the interviewees considered barriers to deliver technology outputs as more
significant than barriers to local technology inputs. These barriers included:
• The small size of the Chilean market. Chilean energy demand is small in
comparison with other Latin American countries like Brazil, Mexico and
Argentina. Foreign investors in the production capacity of renewable energy
technologies are more attracted to large countries like Brazil, where there is a
large demand for their production. Wind turbine manufacturers like Enercon
44
2002 figures.
45
Interview with Mr Conrad von Igel, Manager of Innovation in Chile’s Ministry of Economy, December
2011.
158
have installed production plants in Brazil, from where they supply products to
the Latin American market. Labour mobility in Enercon Brazil has allowed
Fibrovent to gain first-hand knowledge about wind blade production. Other
multinationals like GE Jenbacher have their Latin American base in Argentina,
from where they offer technical support to neighbouring countries. KDM
selected GE Jenbacher generators for their quality and the availability of service
providers in neighbouring Argentina.
• Ineffective national demand-pull policies. Most interviewees considered that
Law 20.257 had failed to promote renewable energy technologies beyond the
cheapest and most mature hydropower. Some of the flaws mentioned by the
interviewees refer to the low targets set, considering that they only affect
electricity supply contracts signed after the 31st August 2007. Some
interviewees indicated further that many of the larger electricity supply
contracts were signed before this date. In addition, targets do not discriminate
per technology type and Law 20.257 does not create a liquid trading system for
renewable electricity generation certificates. Quotas are traded through
bilateral contracts between NCRE generators and either large electricity
generation companies or large consumers in the mining industry. According to
the interviewees, large generation companies obliged to buy NCRE generation
certificates often block the entry of new competitors by implementing their
own low-cost projects or by paying the penalties instead. The system of
bilateral contracts does not provide transparency about the price of the NCRE
attribute. As a result, the NCRE premium cannot be easily introduced in the
cash-flow projections of new projects, or evaluated by banks offering credit.
• The lack of an industrial fabric that can mass-produce local technologies. The
Lithium Innovation Centre and the micro-hydro plug & play projects of
Universidad de Chile are developing industrial concepts that will require
production capacity to be mass-produced and commercialised. The lack of a
solid industrial base in Chile is considered one of the main barriers preventing
these projects from leading to tangible technology outputs.
• A highly concentrated, oligopolistic electricity market. New entrants face
significant barriers to negotiating favourable PPAs with distribution companies,
as well as to obtaining favourable conditions for the supply of equipment.
• Unstable electricity prices. Many interviewees (Mainstream, Biomasa Chile,
Electrica Nueva Energia and Fibrovent) highlighted them as one of the main
barriers to investment in NCRE. Electricity prices are very high in Chile, which
should benefit expensive NCRE technologies. However, volatile prices and
projections of lower future prices create high uncertainty in the market and
block access to finance and long-term contracts for small generators. Large
electricity generation companies do not have incentives to provide stable
prices, which damages small competitors. Future electricity price projections
take into account the implementation of large hydroelectric projects in the
south of Chile, but they have not materialised for more than ten years. Recent
court decisions46 indicate that the construction of the large Hidroaysen
46
El Pais, 20/06/2011 “Un tribunal chileno suspende termporalmente la construccion de cinco presas en
la Patagonia”.
159
hydroelectricity project may be delayed even further due to environmental
impact concerns. Interviewees suggested the creation of a price stabilisation
mechanism to promote NCRE projects, while the renewable energy lobby group
ACERA has appointed an external consultant to study the feasibility and cost of
such a mechanism.
• Ineffective international demand-pull policies. Three of the ten analysed case
studies have registered or requested the registration of their projects as CDM.
Only in one case, landfill gas capture and destruction, were carbon credits
considered essential for the feasibility of the projects. In electricity generation
projects, carbon credits supplemented expected revenues but were not
necessary for their economic feasibility. The CDM did not seem to have an
impact on any of the identified barriers to the development of NCRE, and it
only provided additional revenues to companies that were already financially
solid and capable of gaining access to credit and negotiating a favourable PPA.
Companies found that the CDM registration process is too long, costly and
uncertain, and therefore revenues from carbon credits cannot be taken for
granted when assessing the economic feasibility of new projects. Our case
studies showed that effective national demand-pull policies have a greater
potential to promote NCRE than international carbon markets.
Technology spillovers or effects beyond the emission reductions resulting from project
implementation were rare in the case studies analysed. Foreign technology transfers
not always generated spillovers leading to the improvement of local technological
capabilities and productivity or the reduction of technology costs. The main barriers
preventing these spillovers were:
• Lack of qualified suppliers and clients benefiting from backward and forward
linkages with companies involved in TT processes. Fibrovent and SQM could not
transfer their knowledge of the wind and CSP industries to local manufacturing
companies, because no local companies undertake activities at the later stages
of the wind and solar value chains. The results of UChile’s research on a micro-
hydro plug & play concept will not be transferred to local companies if the
capacity to mass-produce the equipment cannot be provided locally. Foreign
companies MNC-Gen and KDM have outsourced most equipment and servicing
to suppliers abroad, due to their unavailability at the national level. As
described in Chapter 3, domestic manufacturers have been substituted by
foreign counterparts, so that Chile´s industrial fabric is weak and delinked. As a
result, foreign technology transfers have a low potential to generate knowledge
spillovers through backward and forward linkages with domestic suppliers.
• Industrial policy. Linked to the lack of a local industrial fabric is Chile’s hands-off
approach to industrial development, already described when analysing the
barriers to the provision of local technology inputs. MNC-Gen considers
industrial policy as the most important policy gap in Chile and anticipates that
further improvements to Law 20.257, promoting renewable energy generation,
will only benefit foreign companies.
• Small local demand. It prevents economies of scale and learning-by-doing
effects in the less mature technologies in Chile. Fibrovent and Mainstream have
160
not been able to exploit learning and economies of scale effects, as their
activities are interrupted until the market shows a significant scale.
4.4.5 Policy recommendations for Chile

Policy recommendations for the successful transfer of clean energy technologies to
Chile are based on the barriers previously highlighted and focus on two main areas:
nurturing a local industry made up of renewable energy technology suppliers, and
implementing more effective demand-pull policies.
Nurture a local industry made up of renewable energy technology suppliers
The pros and cons of developing industrial policy are complex issues and go beyond
the scope of this thesis. However, as the lack of an industrial fabric has been
mentioned as one of the missing enablers in most case studies, it is relevant to provide
a brief analysis of the issue.
Extensive empirical evidence on industrial development shows that the process
requires relatively high degrees of intervention to support the accumulation of
technological and organisational capabilities (Cimoli et al., 2009). The Chinese, Indian,
Brazilian and South Korean development of industrial leaders in renewable energy
technologies has required significant government intervention through local content
requirements (in China), differentiated fiscal and tariff regulations and subsidies. The
large economies of these countries have allowed them to become successful in many
different high-tech technologies whereas small economies like Chile, with lower
technological capabilities, may require instead to specialise in niche industries with a
medium technology content.
Chile has so far followed an export-led development model based on the static
comparative advantage provided by the availability of mining, forestry, agriculture and
fishing resources. This model provided impressive export and GDP growth rates in the
1980s and ‘90s; however, the growth rates seem to have been exhausted in the last
decade, mainly as a consequence of decreased productivity levels. Productivity gains
resulting from the move from agriculture to services seem to have been completed
and the capital-intensive natural resources-based industries do not allow for significant
productivity gains. Chile has not gone through a process of diversification to
commodities with higher technological content and income elasticity. As a result, the
sophistication of its industrial sector is very low, even in comparison with other natural
resources-based economies such as Australia, New Zealand or Canada. The recent
National Strategy for Innovation explicitly rejects the view that resource-abundant
countries should move away from resource-intensive production, and it identifies eight
clusters linked to fishing, mining, food and services as their priority areas for
development. However, the strategy also shows a will to support knowledge-based
industries, irrespective of Chile’s comparative advantages, by defining priority
transversal areas that include renewable energies and ICT.
Some authors warn of the perils of relying on natural resources availability to achieve
development (Cimoli et al., 2009). They posit that an economy based on natural
resources exports can suffer the so-called “Dutch-disease”, meaning competition
losses of domestic industry due to exchange rate appreciations. These economies are
typically capital-intensive with a reduced demand for skilled labour, which is not
161
necessarily the Chilean case. In addition, they favour polarisation in income
distribution and reduce learning potential. In order to avoid the resource curse “rents
have to be purposefully distributed against comparative advantages”, supporting the
diversification of production in knowledge-intensive activities (Cimoli et al., 2009)
Nevertheless, the case studies show that the Chilean model of development, focused
on openness to foreign trade and investment and the exploitation of natural resources
still offers some opportunities for the development of a domestic clean energy
industry that can contribute to achieve emissions reductions in the country.
Incremental innovation opportunities can come from the expertise gained in
manufacturing industries that extract or use natural resources in the mining and
forestry sectors. Suppliers to the large companies operating in the mining sector can
deliver medium-tech content products such as wind blades and wind towers and
services like drilling and the assessment of geothermal reservoirs, as well as civil and
electrical engineering. Suppliers to mining and forestry industries can also provide raw
materials like molten salts, for solar heat storage in CSP plants, or biomass feed-stocks.
Producers of raw materials have also identified ways to generate additional value-
added benefits to their basic products, such as lithium compounds, Li-ion batteries for
electric vehicles, or a full-service of mixing, melting and feeding molten salts to storage
tanks. A strategy identifying such opportunities for cross-sectoral technology transfer,
as well as incentives to support them, is necessary to build a local clean energies
industry; otherwise, Chile may be left as a simple user and not an active player in this
growing market.
A local industry of components and services for the clean energy sector could benefit
from knowledge spillovers through backward linkages with foreign investors. Chile has
an open door policy for foreign investment, which facilitates exposure to foreign
technologies. However, impoverished linkages in the domestic economy mean that
many MNCs source their inputs from abroad, preventing potential knowledge
spillovers to local suppliers. MNCs may be reluctant to invest in building local
capacities because the benefits leak to other buyers or because the starting point of
local capabilities is too low to make capacitation investments worthwhile. In other
cases, domestic suppliers do not have access to technology or the necessary finance to
deliver products or services that meet MNC specifications (UNCTAD, 2001). Public
policies can encourage the creation and deepening of backward linkages by lowering
the costs and raising the rewards of link formation for both MNCs and local companies.
Chile’s freedom to define industrial policies that nurture infant industries, and
strengthen backward linkages in the renewable energies sector, is limited by WTO
membership and a number of multilateral and bilateral free trade agreements. These
preclude the use of quotas, tariffs or other forms of non-tariff barriers such as local
content requirements, and they also place constraints in terms of what are admissible
subsidies or other forms of support on local industries. Table 14 compiles UNCTAD
(2001) recommendations, previous evidence of industrial development and successful
policies implemented in other developing countries, as reviewed in Chapter 2. Trade
and investment measures should go through a legal feasibility assessment that goes
beyond the aim of this thesis before they can be deemed appropriate for Chile.
162
TABLE 14- INDUSTRIAL POLICY MEASURES INFLUENCING LINKAGES AND LOCAL INNOVATION
Type of Measures
measures
Trade and • Targeted customs and excise duties

investment • Rules of origin, based on the domestic value-added or local content. However,
where local content capacity is weak, foreign affiliates will fill the gap, not
Note: Only
creating domestic linkages
applicable when they
don’t contravene • Joint venture requirements to foreign companies
free-trade • Export performance requirements. They do not directly create linkages, but if
agreements or other they do, they can have a higher quality because export markets are more
multilateral demanding
initiatives like TRIMS • Incentives to foreign affiliates to encourage the creation of linkages, such as
and the Agreement tax exemptions
on Subsidies and • Contractual arrangements with foreign investors including local linkages in the
Countervailing award procedures
Measures
Information and • Provision of information. Governments can act as facilitators by gathering and
matchmaking disseminated information on linkage opportunities
• Matchmaking. Involves a more active government role that would act as a
Note: Only make
broker to reach supply arrangements. Some related activities are facilitating
sense when there
one-to-one meetings, celebrating multilateral events to bring suppliers and
are viable suppliers
consumers together, carrying supportive activities like supplier duty diligence
or administrative contractual arrangements
Technology • Incentives to foreign companies for collaborative R&D with local companies or
upgrading research institutes
• Government partnerships with foreign affiliates to upgrade supplier
capabilities
Training • Promote supplier associations. These associations can help build training
linkages
• Support private sector training programmes, such as MNC-Gen’s programme
for the development of an intermediate layer of technicians for
transformation activities, currently done without support
Finance • Promote renewable energy-friendly financial institutions to facilitate credit to

technologies considered as too risky by conventional private banks
National linkage Programmes focusing on a limited number of industries and firms and using
promotion several of the previous specific measures. Most of these programmes are in
programmes countries with a high FDI presence and a strong local supplier base (Ireland,
Czech Republic, Singapore, Malaysia, Hungary)
The objectives of these programmes are: increase domestic production and
employment, improve the current account, make MNC more rooted in the local
economy, and upgrade the capabilities of domestic enterprises
Sources: UNCTAD (2001) and the author’s elaboration based on case studies
Chile does not apply trade and investment measures as they contradict its free market
economy. However, there are several initiatives in place that can facilitate information
and matchmaking, technology upgrading, training and financing.
163
Chile participates in the Subcontracting and Partnership Exchange (SPX) programme of
UNIDO, which is hosted by the “Corporacion Nacional de Ejecutivos de Abastecimiento
y Contratos de Chile-(CORPAC)”. It has close to 2,400 affiliated companies operating
mainly in the metal-mechanic, industrial services, textiles, plastics, electrical,
electronics, wood and glass sectors. CORPAC has not yet done any matchmaking in the
renewable energy sector in Chile47 but many of its members in the metal-mechanical
and industrial services sectors could be potential suppliers for the sector.
CORFO’s InnovaChile facilitates training and technological upgrading of local suppliers
through several programmes. Its human capital programme is specifically addressed to
professionals in the priority sectors identified by the National Strategy for Innovation:
fishing, food industry, mining, global services, tourism, forestry, metal-mechanic,
energy and transport. CORFO’s high technology investment programme, addressed to
foreign investment, also facilitates backward linkages with local suppliers, by arranging
meetings with equipment suppliers and professional service providers
The CER continuously organises workshops and events that facilitate networking
among Chilean and foreign companies in specific renewable energy areas. The Chilean
Association of Renewable Energies (ACERA) includes among its members large
international electricity generation companies (e.g. Endesa, SN Power, Enel),
international renewable energy technology developers (e.g. Gamesa, Vestas, Acciona,
Siemens), foreign renewable energy project developers (Acciona, Solarpack, Pacific
hydro, Mainstream, Eolic partners, Enhol) and some local suppliers of components and
services for the renewable energy sector, including Fibrovent for wind blades, several
engineering and consulting firms, construction firms and transport companies.
Chile does not have a national linkage promotion programme, but could take
advantage of its existing infrastructure and the lessons learnt in other countries to
implement a successful one. Strong political commitment will be required, as well as
effective public-private partnerships. ACERA and the CORPA can provide the private
sector side of public-private partnerships to improve the technological capabilities of
Chilean suppliers in the renewable energy sector.
Implement effective national renewable energy demand-pull policies
The lack of appropriate demand-pull policies was consistently mentioned as a barrier
to TT in the analysed case studies. Even if Chile is a small economy, focused on export-
led growth, a significant internal demand for renewable energy technologies is
required by domestic innovators to gain credibility in global markets with high barriers
to new entrants.
Chapter 3 described how Law 20.257 had failed to promote a certain, transparent and
competitive market for renewable energy certificates capable of encouraging
investment in renewable energy technologies. The experiences of other countries that
have implemented quota-based systems can provide valuable insights into improving
Chile’s scheme.
In Europe, tradable green certificates based on quota obligations for renewable
energies are now in place in Italy, the UK, Sweden, Belgium, Romania and Poland, and
in different variations (Haas et al., 2011). Only in Italy does the quota have to be
47 nd
Written communication with Amador Auad Herezi, representative of CORPAC-Chile, 22 June 2011.
164
fulfilled by the generators, whereas in all the other countries compliance has to be
demonstrated by electricity suppliers. In Sweden, up to 2006, compliance was required
from end users, but has since then changed to suppliers. Details about all EU schemes
can be found in Haas et al. (2011), the UK scheme is specifically described in Butler and
Neuhoff (2008) and Kelly (2008) and the Australian scheme is described in Kelly (2008).
Evidence in the EU has shown that TGC schemes have been less effective than feed-in
tariffs (FITs) schemes in achieving a significant increase in the deployment of capacity
from renewable energies. TGC had also less economic efficiency than FITs, with TGC
prices much higher than premiums received by the investors in guaranteed FITs
schemes, with the exception of Sweden. Belgium and Italy, both with TGC schemes,
were the EU countries with the highest support levels but also among those with the
lowest specific deployment. In Australia, targets were set as a fixed generation figure
by 2010 instead of a percentage, which as a result of high growth in electricity
generation led to a target of only 0.2% generation from renewables. Some interesting
lessons learnt by the EU (Haas et al., 2011; Butler and Neuhoff, 2008; Kelly, 2008) and
Australia (Kelly, 2008) about the implementations of a quota system are applicable to
Chile:
• Uncertainty about the price of renewable energy certificates in all schemes and the
amount of buy-out receipts specifically in the UK have deterred private investment
due to a high revenue risk that prevented access to credit and made it difficult to
sign long-term PPAs. Obtaining financial backing was easier and cheaper under
feed-in tariffs or public tendering schemes that could offer stable, long-term prices
for the energy generated. Uncertainty about the price of the renewable energy
attribute is even higher in Chile, where there is not a liquid market for tradable
certificates and where new NCRE capacity needed to meet the target is unknown
for all but the largest electricity generation companies, due to the unknown
deadlines of PPAs, as described in Chapter 3 of this thesis.
• Policy reviews and amendments cause disquiet among renewable energy investors,
introducing further uncertainty to that caused by unknown prices. Uncertainty
about policy continuity also deters the development of a domestic industry.
• Feed-in tariffs have facilitated the development of a renewable energy equipment
industry in Germany, Denmark and Spain, by encouraging market participants to
develop a long-term perspective. Quota schemes in the UK and Australia have
instead put the emphasis on reducing the price paid for renewable energy, creating
a volatile demand and hampering the growth of domestic equipment producers.
Uncertainty about the price of the NCRE attribute and the amount of NCRE
capacity required to meet targets is also likely to deter long-term investments in
Chile.
• Undifferentiated quota systems discourage the development of less mature
technologies. Having excluded large existing hydro projects from its scheme, the
UK concentrated most of its deployment on onshore wind, landfill gas and biomass
co-firing. Australia filled more than 50% of its renewable energy quota with large
existing hydro generators. Chile is expected to follow a similar path, with small
hydro projects likely to cope with most or all of the NCRE’s required quota.
Countries with feed-in tariffs have instead been able to move along the learning
curve of other technologies. Among countries with TGC schemes in the EU there is
165
currently a trend towards technology specification by technology weighting factors
for tradable green certificates, according to which more expensive technologies are
eligible for more than one green certificate.
• Quota systems’ revenue risks and the emphasis on achieving price reductions
preclude entry by small producers. Participation tends to be dominated by larger
electricity firms able to accept and internalise greater financial risks. This is even
more so in Chile, where the compliance obligation is placed on large electricity
generators and where asymmetries of information mean that only these large
generators know how much NCRE capacity is required to meet the targets. Smaller
renewable energy firms could support a wider market development, as they are
not subject to conflicts of interest with conventional fossil-fuelled generation and
they can address small-scale initiatives in more diverse geographical locations.
• Systems that do not differentiate by the quality of the resource in different
locations make sites with higher wind speed gain higher returns than necessary to
cover investment costs. The good resource scarcity rent would be captured by land
owners under strong competition or by the project developers with low
competition when, for example, there is no financing available due to high risk. In
Chile, where the renewable energy potential is still largely underutilised, this is not
expected to be a major problem, but it is an issue to consider as competition for
high quality sites increases.
Chile needs to amend its NCRE quota scheme to improve its effectiveness. It should
introduce further liquidity and transparency and facilitate entry for new producers. A
tradable green certificate (TGC) scheme with a specific market for the certificates
could introduce the transparency and liquidity required. Mechanisms for price
stabilisation were also suggested by many of the experts interviewed as a way to
reduce the current uncertainty of NCRE prices and promote long-term investments.
Price stabilisation rather than subsidies through feed-in tariffs are more acceptable in
Chile, given the high price of Chilean electricity and its free market economy. The CER
and ACERA are currently studying the feasibility of different potential mechanisms for
price stabilisation, mainly an electricity price insurance and an NCRE prices stabilisation
fund.
4.4.6 General policy recommendations for developing countries looking to increase

climate change technology transfer
The analysis of technology transfer case studies in Chile has helped in devising a
methodology to conceptualise and measure TT transfer processes, taking into
consideration technology inputs, the channels though which foreign technologies flow
into the country, technology outputs and effects. Shortfalls in any of these elements
lead to underperformance in TT processes. Several enabling factors operate to ensure
that these elements are in place so that successful TT can be achieved. The literature
review in Chapter 2 identified four types48 of enabling factors for climate change TT to
developing countries: economic and institutional, technology demand, technology
supply and industrial development. Findings about the strengths and weaknesses of
Chile’s enabling factors can complement those highlighted by the previous literature
48
The fifth type, “bilateral factors”, is not considered, as it applies to pairs of technology suppliers and
recipients, not to individual recipient countries.
166
and contribute to drawing general policy recommendations for other developing
countries. In addition, through the approach devised in this chapter, enabling factors
can be classified according to the element of the TT process where they have the
highest impact, which would help developing countries to focus their policy efforts
towards their main weaknesses. Table 5 presents a compilation of enabling factors for
climate change technology transfer, distributed per type of enabling factor and the
element of the TT process that they influence. It is worth noting that some enabling
factors can have contradictory impacts in the different elements of TT. For example, an
industrial policy used to support a local clean energy industry though targeted trade
and investment measures can benefit the development of local technology inputs and
the creation of backward linkages with local suppliers. However, it could also restrict
the flow of foreign technologies, therefore damaging the TT process. A fine balance
should be found through policies that attract foreign technologies but allow local
suppliers to catch up. The cases of India and China are clear examples of this balance,
where foreign suppliers were attracted, mainly due to the large size of the market, but
policies were in place to maximise technology spillovers of their activities. Chinese and
Indian market sizes were large enough to ensure substantial profits for foreign
suppliers, even in the case of competitiveness losses that may result from knowledge
leaks. This would not be the case in smaller economies like Chile, which could
therefore not apply the same types of policies.
Quantitative analysis mapping the performance of developing countries in these
enabling frameworks could be useful for setting policy priorities. However, it is
advisable to analyse qualitatively the situation of each recipient developing country to
identify the pitfalls of the technology transfer processes taking place. Enabling
frameworks, such as renewable energy demand-pull policies in Chile, may be in place
but ineffective due to flaws in their design or implementation.
167
TABLE 15- ENABLING FACTORS FOR CLIMATE CHANGE TECHNOLOGY TRANSFER
Enabling
factors Economic and institutional Technology demand Technology supply Industrial development
TT elements
Local Competitive labour costs Human capital Competitive industrial sectors
enabling cross-sectoral TT
technology General and clean energy technology
specific R&D capabilities Workforce with intermediate
inputs technical qualifications for
Access to natural resources transformative activities
General and clean technology specific Industrial policy specific for
knowledge stock (patents) clean energy technologies
Functioning NIS, including competent
universities and research centres
Technology-push policies: Tax credits
and subsidies to R&D, subsidies for
capacity building, demonstration
projects, national certification
programme
Channels of Openness to trade and FDI Good transport infrastructure

foreign Stable macroeconomic framework Public investment in infrastructure
technologies development
Low income tax
Strong governance
Low corruption
Ease of doing business
Regional economic integration
Specific High-technology FDI
attraction policies
168
Enabling
factors Economic and institutional Technology demand Technology supply Industrial development
TT elements
FDI attraction policies that do not
crowd out local innovators
Policies to support foreign
exposure at the corporate and
academic level
IPR protection
Technology Access to credit for renewable Large population Production capacity to bring
energy projects technology prototypes to market
outputs Large GDP pc
Low concentration in the
Large energy consumption pc
electricity generation system
High GDP growth
High energy consumption growth
High electricity and fossil fuel prices
Stable electricity prices
Demand-pull policies: FIT, TGC,
Concession programmes, tax credits
and rebates, energy efficiency
standards, air quality standards, ETS,
public procurement, taxes on fossil
fuels
Technology Significant local demand Functioning NIS enabling industry- Quality industrial local suppliers
effects university links
Linked domestic industrial sector
Industrial policy promoting
backward and forward linkages
169
4.5 Conclusions
This chapter has followed a qualitative approach to provide answers to the research
questions set at the introduction of this thesis, regarding the concept, measurement and
enabling factors of clean energy technology transfers to developing countries. The approach
has consisted of the analysis of ten case studies based in Chile. The case studies presented
processes of technology transfer for different types of technologies through different TT
channels and where different types of organisations were involved. Evidence was gathered
through interviews with representatives of the organisations involved and through the
review of company documentation.
The case study analysis has revealed a common pattern in which several elements interact
to achieve successful TT: local technology inputs, foreign technology inputs, channels
through which foreign inputs can flow into the local economy, technology outputs and
technology effects resulting from the transfer. All of these elements need to be in place in
successful TT processes. Therefore, an appropriate measurement of TT should take into
account all the elements involved.
Enabling factors operate at every element of the TT process. Each developing country
should assess its main weaknesses in the TT process and focus on the types of enabling
factors that can improve the performance of its weaker elements. The case studies analysed
highlighted the main weaknesses of Chile in achieving successful TT, namely the lack of a
solid industrial structure and ineffective demand-pull policies. Policy recommendations
were derived to improve Chile’s performance. A compilation of enabling factors for each of
the TT elements was presented to support the prioritisation of policies by developing
countries willing to attract and successfully absorb foreign technologies.
The proposed measurements of TT and its enabling factors set the foundation for the
quantitative approach of Chapter 5 to define policy priorities for a sample of developing
countries with a view to improving their performance in the adoption of foreign climate
change technologies.
170
5 Quantitative analysis to assess
developing countries’ policy needs for
clean energy technology transfer
5.1. Introduction
5.2. Data
5.3. Exploring indicators of enabling
frameworks for TT
5.4. Analysing the relationship between
enabling frameworks and TT
5.5. Defining groups of developing countries
5.6. Discussion and conclusions
171
5.1 Introduction
The UNFCCC has so far achieved limited success in the promotion of clean energy
technology transfer to developing countries, as explained in Chapter 1 of this thesis. The
new technology mechanism decided in the Cancun Agreements (UNFCCC, 2010) needs to
avoid previous flaws to contribute to higher levels of technology transfer. Firstly, it needs to
act on the enabling factors that facilitate technology transfer at the national level. Secondly,
it must adapt its support to the different circumstances of the recipient developing
countries. Finally, to allow for the implementation and evaluation of support instruments,
the flows of foreign technologies, their outputs and effects need to be monitored.
Monitoring technology transfer is a difficult task because it has no measurable physical
presence or a well-defined price (IPCC, 2000) and its transfer goes beyond the mere flow of
foreign technology inputs into a recipient country. Former approaches to measuring TT are
reviewed in Chapter 2, while Chapter 4 of this doctoral thesis complements the existing
literature with evidence provided through the analysis of ten case studies of climate change
TT in Chile. The case study analysis proposes in some cases measurements of climate change
technology transfer similar to those identified in the literature. It also provides some
additional measurements so far not analysed by the literature (these are shaded in green in
Table 16) and does not consider as relevant some of the previously suggested
measurements. Table 16 summarises the findings of Chapters 2 and 4, highlighting in green
possible measurements of technology transfer found through the case study analysis that
has not been previously used in the literature. The different measurements of technology
transfer are classified as inputs, if they refer to channels through which foreign technologies
flow into a country, outputs, when they refer to the products or projects directly achieved
through foreign technologies, or effects, when they refer to further spillovers from the
receipt of foreign technologies. The case study analysis suggests that technology transfer
should be measured by taking into account these three elements, but measuring TT through
only input, output or effect indicators results in an incomplete picture of the process.
172
TABLE 16- MEASUREMENTS OF CLIMATE CHANGE TECHNOLOGY TRANSFER SUGGESTED BY THE LITERATURE
REVIEW AND CASE STUDY ANALYSIS
Type of TT Literature review Case studies

indicator
Input Foreign R&D expenditure weighted by the R&D expenses of multinationals in CC-
bilateral import-share related sectors
Value of climate change-related FDI
Value of climate change-related imports
Value of climate change-related ODA
Value of collaborative research agreements
Financial resources provided by the UNFCCC for
TNA and capacity building
Value of joint public-private climate change
R&D programmes
Research agreements with private companies
Foreign patents filed related to climate change
technologies
Number of foreign workers in the CC
technology sector
CC-related licence payments to foreign
companies
Output Number and licence revenues of local patents produced with foreign support
Scientific publications
Number of spin-offs established
Invention disclosures evaluated
Citations received by patents originating in
a country by patents originating in another
country, related to climate change
technologies
Number of CDM projects that incorporate
foreign technologies or foreign knowledge
MW installed by foreign owners or with foreign
technologies
kWh of kJ generated or saved by foreign-
owned projects or with foreign technologies
tCO2e captured or destroyed by foreign-owned
projects or with foreign technologies
Production of CC-related technologies or
products by foreign owners or with foreign
technologies
Effect TFP of CC-related industrial sectors explained by foreign CC-related inputs
tCO2e reductions in the recipient economy explained by foreign CC-related inputs
Cost of clean energy explained by foreign CC-related inputs
Chapters 2 and 4 also indicate that successful technology transfer requires a set of enabling
factors to facilitate the flow of foreign technologies into the country, the provision of local
technological inputs so that the recipient country can efficiently absorb foreign technologies
and use them to generate technical change, the implementation of foreign technologies to
deliver their intended output, and channels for the achievement of technology spillovers in
the recipient economy. Technology transfer processes can be stalled or underperform when
some of these enabling factors are missing. Chapter 2 reviews the different enabling factors
identified in the literature and classifies them as four types: macroeconomic and
173
institutional frameworks, technology demand, technology supply and industrial
development. Chapter 4 analyses the importance of these factors and finds some additional
ones (highlighted in green in Table 17) through the study of ten cases of clean energy
technology transfer to Chile. It finds that each type of enabling factor has a predominant
impact on one of the elements of the TT process, and it then identifies which factors have
enabled TT processes in Chile and which would be required to achieve higher levels of TT.
Table 17 summarises the findings of both the literature review and the case study analysis.
Type of factor Literature review Case studies Expected TT process

and TT process effect element
Economic and Ease of doing business Positive Channels
institutional Corruption Negative of foreign
frameworks IPR protection Positive technology
Open trade Openness to trade and FDI Positive Channels
Stable macroeconomic Positive of foreign
framework technology
Regional economic integration Positive
Flows of high technology FDI Positive
Access to credit for private Positive Output
developers of clean energy
projects
Technology Demand size: Positive Output
demand Population
GDP absolute and per capita
Energy consumption absolute and per capita
Demand growth Positive
Annual GDP growth
Energy consumption growth
Electricity prices High electricity and fossil fuel Positive
prices
Demand-pull policies Depends on
policy type
and
implementat
ion
Stability of electricity prices Positive
Technology Human capital Positive Input
supply Enrolment in tertiary education
% of degrees in engineering and science
Annual number of graduates in science and engineering
Literacy rates
Quality of education
Specific capabilities in clean energy technology
Workforce with intermediate Positive Input
technical qualifications for
transformation activities
Policies to facilitate education Positive Channels
abroad of foreign
technology
R&D Positive Input
General R&D spending
R&D spending in energy-
related technologies
Patent stock: General and clean technology Positive Input
174
Stock of patents of Specific knowledge stock
specific technology fields
Quality of the stock of
patents, measured as
citation levels
Technology development Positive Input
indexes
Technology push policies: Positive Input
Government sponsored R&D
Tax credits for R&D
Subsidies to education
Infrastructure development
Funding for demonstration projects
National certification programmes
Functioning National Innovation Positive Input,
System effects
Good transport infrastructure Positive Input
Access to sources of renewable Positive Input
energy
Industrial Competitive industrial sector enabling cross-sectoral TT Positive Input,
development effect
Market power in related industries (electricity sector): Negative Output
Profit levels of incumbent companies
Market concentration
Cost of factors: labour and energy Negative Input
Industrial policy: Industrial policy : Positive in Input,
Local content Specific promotion of clean BRICs, effect
requirements energy industries Uncertain in
Differentiated tariffs for Attraction of high-technology FDI smaller
targeted technologies Specific policies for local economies
Differentiated fiscal innovators
treatment for targeted Promotion of backward and
technologies forward linkages
Large production capacity to Positive Output,
bring technology prototypes to effect
market
High quality of local suppliers Positive Effect
Linked domestic industrial sector Positive Effect
The aim of this chapter is to build upon the findings of the previous chapters to assess the
performance of a sample of developing countries regarding the levels of clean energy
technology transfer achieved, the enabling factors that make them possible and barriers or
missing enabling factors that could be preventing further TT. Firstly, the availability of data
to measure climate change TT and enabling factors found relevant after the case study
analysis will be assessed for a significant sample of developing countries. Secondly, the
principal components analysis technique will be used to explore the interrelationships
between the different indicators of enabling factors for TT and the performance of
developing countries on these aspects. Thirdly, the impact of enabling factors on actual
measurements of CC TT will be assessed through regression analysis. Finally, the cluster
analysis method will be used to group developing countries according to their policy
priorities in order to improve the levels of clean energy technology transfer.
175
5.2 Data availability
5.2.1 Dependent variables
5.2.1.1 Foreign flows of clean energy technology
Data on the imports of some clean energy-related technologies can be extracted through
the Commodity Trade Statistics Database of the United Nations (COMTRADE49). COMTRADE
contains data on the annual import and export values of different types of commodities for
139 countries, categorised according to the Commodity Description and Coding System
(HS1996). In order to find data on imports of clean energy technologies we selected a total
of seven commodities, as presented in Table 18. They do not represent the whole range of
clean energy technologies, because it is difficult to separate general energy from renewable
energy-related technologies from other COMTRADE commodity codes. Only 79 countries
present available data for these categories of commodities.
TABLE 18-COMTRADE CLEAN ENERGY TECHNOLOGY IMPORT AND EXPORT DATA CODE DESCRIPTION
Code Name and description of commodity

841011 Name: Hydraulic turbines, water wheels, power < 1000 kW
Description: Hydraulic turbines and water wheels :-- Of a power not exceeding 1,000 kW
841012 Name: Hydraulic turbines, water wheels, power 1000-10000 kW
Description: Hydraulic turbines and water wheels :-- Of a power exceeding 1,000 kW but not
exceeding 10,000 kW
841013 Name: Hydraulic turbines, water wheels, power > 10000 kW
Description: Hydraulic turbines and water wheels :-- Of a power exceeding 10,000 kW
841090 Name: Parts of hydraulic turbines and water wheels
Description: Parts, including regulators
841919 Name: Instantaneous/storage water heaters, not electric other
Description: Instantaneous or storage water heaters, non-electric – other [solar water heaters]
850231 Name: Wind-powered generating
Description: Other generating sets :-- Wind-powered
854140 Name: Photosensitive/photovoltaic/LED semiconductor devices
Description: Photosensitive semiconductor devices, including photovoltaic cells whether or not
assembled in modules or made up into panels, light emitting diodes
The distinction was made between imports of all the commodities above and imports
excluding hydro systems, but both variables were highly correlated. Therefore, it was
decided to keep only the variable REIMP defined as imports of clean energy technologies
and measured in US$ at purchasing prices for all the above categories. REIMP is expressed in
logarithms and then defined as REIMPlog to get a close to normal distribution N(15.55,
2.568)
We will also consider relative renewable energy import values per capita values (per
thousand population) and per GDP (billion US$) to avoid the scale effect. These variables are
expressed in logs to approximate a normal distribution and denominated REIMPPClog
(imports of renewables per capita in logs) and REIMPGDPlog (imports of renewable per GDP
in logs). Histograms and box plots of import variables in relative values are included in the
Annexes to Chapter 5, while Table 19 shows descriptive statistics for variables REIMP,
REIMPlog, REIMPPC, REIMPPClog, REIMPGDP and REIMPGDPlog.
49
http://comtrade.un.org/db/
176
TABLE 19- DESCRITIVE STATISTICS FOR RENEWABLE ENERGY TECHNOLOGY IMPORT VARIABLES
N Range Minimum Maximum Mean Std. Deviation Variance

REIMP 79 4.40E9 1559 4.40E9 1.0873E8 5.08307E8 2.584E17
REIMPlog 79 14.85 7.35 22.2 15.55 2.57 6.6
REIMPPC 79 19008.61 17.31 19025.92 1566.12 3053.96 9326696.51
REIMPPClog 79 6.95 2.91 9.85 6.27 1.49621 2.239
REIMPGDP 79 2751.93 .00 2751.93 270.05 458.37 210102.08
REIMPGDPlog 79 6.78 8.05 14.83 12.2064 1.19934 1.44
Valid N (list- 79
wise)
Data on foreign direct investment (FDI) related to renewable energy technologies are not
available and are difficult to obtain, as the process would involve scrutinising every
individual FDI case to separate out those which are definitively low-carbon from a total
number of investments of around 22,000 in 2009 alone (UNCTAD, 2010). The OECD
(Buchner et al., 2011) and UNCTAD (2010) are currently working to improve data availability
on climate change-related FDI. Data on the value of clean energy-related financial
investments, not distinguishing between local and foreign investment, are provided by
UNEP and Bloomberg New Energy Finance (2010) for 20 countries with more than 0.1US$
billion of investment, as shown in Figure 53. Financial investments are clearly dominated by
China, with the last six countries in the graph showing investment of around 0.1 US$ billion.
The sample of countries is very small and the figures do not differentiate between local and
foreign investment, which does not make them suitable for our analysis.
FIGURE 53- NEW FINANCIAL SECTOR INVESTMENTS IN CLEAN ENERGY, 2009, US$ BILLION
40
35
30
25
20
15
10
0
Peru
South Africa
Mexico
Philipines
China
India
Indonesia
Vietnam
Costa Rica
Uganda
Panama
Egypt
Sri Lanka
Brazil
Pakistan
Chile
Nicaragua
Ethiopia
Thailand
Data on climate change-related R&D expenses incurred by foreign multinationals in

developing countries, on the number and remuneration of foreign workers in clean energy-
related sectors and on payments for foreign licences are not systematically collected in
developing countries and cannot be taken into account for our analysis.
177
5.2.1.2 Outputs of foreign clean energy technology transfer
Production of clean energy technologies
Data on the production of clean energy technologies are not available for a significant
sample of developing countries. COMTRADE’s export data for a number of clean energy-
related commodities, as shown in Table 18 can be taken as a proxy for the production of
internationally competitive clean energy technologies employed by developing countries. It
is not possible to determine which share of these exports has relied on foreign technology
transfer for their production, but the high correlation between the imports and exports of
renewable energy technologies indicates that exporters are highly exposed to foreign
technologies. As with imports, the distinction between exports of all the commodities and
exports of only non-hydro technologies was not found to be relevant. Therefore, we only
take the variables REEXP (total exports of the equipment in Table 1), REEXPPC (total exports
per capita) and REEXPGDP (exports per billion dollars of GDP). All variables are expressed in
logs (REEXPlog, REEXPPClog, REEXPGDPlog) to approximate normal distributions.
Table 20 shows descriptive statistics for variables REEXP, REEXPlog, REEXPPC, REEXPPClog,
REEXPGDP and REEXPPGDPlog.
TABLE 20- DESCRIPTIVE STATISTICS FOR RENEWABLE ENERGY TECHNOLOGY EXPORTS

REEXP 67 1.13E10 1.00 1.13E10 2.1935E8 1.38926E9 1.930E18
REEXPlog 67 22.46 .69 23.15 12.6534 4.30001 18.490
REEXPPC 67 49171.8303 .0000 49171.8303 1713.956418 7087.1908687 50228274.410
REEXPPClog 66 10.80 .00 10.80 3.5476 2.79361 7.804
REEXPGDP 67 4358.6270 .0000 4358.6270 188.775853 626.9034094 393007.885
REEXPGDPlog 67 15.2695 .0181 15.2877 8.824871 3.0569672 9.345
Valid N (list- 66
wise)
Data on the production of biofuels can also be taken as an indicator of the ability to produce
clean energy technologies, but again there is a high number of zero values, with only 28
observations with values above zero; therefore, it will not be appropriate to perform
multivariate analysis.
Installed capacity and electricity generation with renewable energies
Data on installed renewable generation capacity are available in the US Energy Information
Database for a significant number of developing countries. However, it is not possible to
discern which share of renewable energy generation capacity and production has required
foreign technology transfer. Information of technology transfer for renewable electricity
generation projects is available for CDM projects because as part of a UNFCCC-sponsored
study on technology transfer in the CDM (Seres et al., 2010), a database was created
indicating for every CDM project whether it claims foreign technology transfer or not. The
database includes a total of 4,984 emission reduction projects, among which are 3,141
renewable energy-related projects, starting the CDM registration process between
December 2003 and June 2010. Some caveats regarding the use of these data are
highlighted in Chapter 2 about the literature review. There is a high correlation between the
renewable energy generation capacity of a country (RECAP), the number of renewable
energy CDM projects it hosts (CDMRE) and the estimated emissions reductions of these
CDM projects (CDMRECO2). Table 21 shows correlations between these variables. Emissions
178
data on CDM projects take into account the size of the projects, as emission reductions
depend on the amount of electricity produced.
TABLE 21-CORRELATIONS BETWEEN RE CAPACITY, RE ELECTRICITY GENERATION, CDM RE PROJECTS AND CDM
RE PROJECTS CO2 EMISSION REDUCTIONS.
RECAP CDMRE CDMRECO2

** **
Pearson 1 .922 .942
RECAP
N 130 73 73
** **
Pearson .922 1 .969
CDMRE
N 73 73 73
** **
CDMRECO Pearson .942 .969 1
2 N 73 73 73
The high correlation between RE capacity and emissions reductions produced by renewable
energy CDM projects allows us to use the technology transfer information on the CDM to
account for technology transfer in total installed renewable energy capacity. Data on
technology transfer in RE CDM projects are available for 54 developing countries.
Most countries claim foreign technology transfer in all of their RE CDM projects. Therefore,
the median of the share of emissions reductions by RE and non-hydro RE CDM projects that
claim TT (CDMRETTCO2p) is 100%. The mean for RE projects is 91.7% and for non-hydro RE
projects it is 93.3%.
Outliers claiming technology transfer well below 100% of the CDM projects are India, China,
Colombia, Brazil and Armenia, which indicates the higher capability of BRICs to develop their
own technologies. On the other hand, most renewable energy CDM projects in Colombia
and Armenia are hydro in nature, where the technology is considered normal in the country
and therefore technology transfer is not claimed.
TABLE 22- EXTREME VALUES OF CLAIMS OF TECHNOLOGY TRANSFER IN RENEWABLE ENERGY AND NON-
HYDRO RENEWABLE ENERGY CDM PROJECTS
The percentages of TT claims obtained from CDM projects will be used to adjust the
renewable energy capacity (RECAP) figures to reflect only the capacity estimated to have
required technology transfer. For all the countries for which information is not available, the
mean of TT claims in RE CDM projects, at 92%, will be used to adjust the figures for installed
capacity. Countries for which no data are available for CDM TT claims include one BRIC
country, Russia, and other emerging economies like Turkey, as well as the least developed
countries like Congo or Uganda. For this reason, the mean of 92% is considered more
179
appropriate than the median, 100%, for estimating the expected needs of TT. The values are
transformed with logarithms, and the new variable of adjusted renewable energy capacity
per estimated levels of technology transfer is called RECAPTTlog, when transformed with
logs.
TABLE 23- DESCRIPTIVE STATISTICS OF RENEWABLE ENERGY CAPACITY INVOLVING TECHNOLOGY TRANSFER

RECAPTT 105 51722.12 .92 51723.04 2789.1157 7403.93070 54818189.776
RECAPTTlog 105 10.20 .65 10.85 5.8804 2.46632 6.083
RECAPTTPC 105 2630.38 .07 2630.45 143.2973 356.36041 126992.744
RECAPTTPClog 105 7.80 .07 7.88 3.5705 1.74492 3.045
RECAPTTGDP 103 1165.21 .05 1165.26 58.5844 155.12317 24063.197
RECAPTTGDPlog 103 7.02 .05 7.06 2.9504 1.42052 2.018
Valid N (listwise) 103
5.2.1.3 Effects of clean energy technology transfers
The existing literature has analysed the knowledge spillovers of clean energy-related foreign
activities as the resulting changes in CO2 emissions or energy consumption per capita or per
unit of GDP. Existing studies analyse the impact of total flows of FDI or imports, instead of
the specific impact of climate change-related flows. As a result, existing models are
contaminated by the “scale, composition and technique effects” of foreign activities, as
detailed in the literature review chapter, and they do not reflect the specific impact of
climate change-related technologies in the economy. Furthermore, there are no studies
focusing on the impact of foreign climate change-related technologies on the TFP of the
recipient country. The required data required to analyse the effect of foreign technology
flows on CO2 emissions and productivity levels of the recipient countries are not available
for a significant sample of developing countries, and such a study cannot be undertaken by
this thesis with publicly available data.
Spillovers have also been measured as cost reductions of low-carbon technologies as a
result of foreign and local technology adoption and learning-by-doing effects. The only
known studies focusing on a developing countries focus in China´s wind technology (Qiu and
Anadon, 2011) and Brazil´s biofuels (Goldemberg et al, 2004). However, data on technology
costs and foreign technology adoption are not available for a significant number of
developing countries, and their creation and collection goes beyond the scope of this thesis.
5.2.1.4 Selection of indicators of clean energy technology transfer
The initial long list of indicators of clean energy technology transfer is presented in Annex 1
of Chapter 5. From this long list, three variables can be used to reflect the extent of clean
energy technology transfer in developing countries: imports and exports of renewable
energy technologies and renewable energy generation capacity adjusted per estimated
levels of technology transfer. The three variables are expressed in absolute, per capita and
per GDP values and are all transformed with logarithms, as they showed non-normal
distributions with long right-hand tails.
180
TABLE 24- SELECTED TECHNOLOGY TRANSFER VARIABLES
TT Variables Description Source Data sample

aspect (developing
countries)
Flows REIMPlog Ln of the value in US$ of imports UN Trade Statistics 79
of of a selected sample of RE database
foreign technologies in 2009 (COMTRADE)
inputs REIMPPClog Ln of the value of imports per UN Trade Statistics 79
thousand population of a database
selected sample of non-hydro (COMTRADE) and UN
RE technologies in 2009 population data
REIMPGDPlog Ln of the value in US$ of imports UN Trade Statistics 79
of renewable energy database
technologies per Million US$ of (COMTRADE) and UN
GDP in 2009 population data
Output REEXPlog Ln of the value in US$ of exports UN Trade Statistics 67
of a selected sample of RE database
technologies in 2009 (COMTRADE)
REEXPPClog Ln of the value in US$ of exports UN Trade Statistics 66
per thousand population of a database
selected sample of non-hydro (COMTRADE) and UN
RE technologies in 2009 population data
REEXPGDPlog Ln of the value in US$ of exports UN Trade Statistics 67
of renewable energy database
technologies per Million US$ of (COMTRADE) and
GDP in 2009 World Bank indicators
RECAPTTlog Ln of the installed MW of US Energy 105
renewable electricity in 2008, Information Agency
adjusted per the percentage of
TT claims in RE CDM projects
RECAPTTPClog Ln of installed MW of renewable US Energy 105
electricity generation adjusted Information Agency
per TT claims, per thousand And UN population
population data
RECAPTTGDPlog Ln of renewable electricity US Energy 103
generation MW adjusted per TT Information Agency
claims, per Billion US$ of GDP And World Bank
indicators data
Note: Due to the large number of zero values we added 1 to the variable before taking logs.
Indicators that could not be included due to the lack of available data are presented in the
table below.
TABLE 25- TECHNOLOGY TRANSFER VARIABLES THAT COULD NOT BE INCLUDED
Indicator
FLOWS OF Clean energy-related investments
FOREIGN INPUTS Climate change-related FDI.
Foreign workers in CC technology sectors
CC-related licence payments to foreign companies
CC-related patents filed by foreign actors
Foreign companies with R&D expenses in CC-related sectors
OUTPUT kWh of kJ generated or saved by foreign-owned projects or with foreign technologies
tCO2e captured or destroyed by foreign-owned projects or with foreign technologies
Revenues from local patents developed with foreign support
Number of local scientific publications in collaboration with foreign institutions
181
Indicator
EFFECT tCO2e reductions explained by foreign CC-related inputs
TFP of CC-related industrial sectors explained by foreign CC-related inputs
Correlations between the selected variables, including total generation capacity not
adjusted by TT claims, are flagged in Table 26, with significant correlations among variables
of the same type flagged in pale red and those of different variables shaded in green.
Renewable generation capacity in absolute value, adjusted or not for TT claims, has a
significant positive correlation with the absolute levels of imports and exports of related
technologies. This may be due to a scale effect, meaning that large countries are expected
to have large figures for capacity and foreign financial flows indicators, even if these are not
related to each other. In fact, the correlation does not exist between renewable energy
capacity and import and export variables when these are expressed in per capita values or
per GDP values. Exports per capita and per GDP are also correlated with absolute renewable
energy capacity, which may indicate that countries that have reached a significant scale of
installed capacity have the ability to produce more of their own technologies than countries
with smaller capacities.
Total renewable generation capacities per capita and GDP have a strong correlation with
total renewable generation, which indicates that the largest countries tend to achieve more
MW of clean energy per capita than smaller countries, probably facilitated by economies of
scale and learning effects.
Imports and exports of renewable energy technologies have a significant correlation, which
is weaker in per capita values and does not exist in values per GDP. This indicates that the
relationship is influenced by a scale effect, meaning that large countries would be expected
to have large foreign financial flows of all types, even if these are not correlated. In relative
values per GDP, importing renewable energy technologies is not related to a country’s
ability to produce and export its own technologies.
182
TABLE 26- CORRELATIONS OF CLEAN ENERGY TECHNOLOGY TRANSFER VARIABLES
Correlations
RECAPPCl RECAPGDPl RECAPTTl RECAPTT RECAPTT REIMPPCl REIMPGD REEXPlo REEXPPCl REEXPGDPl
RECAPlog REIMPlog
og og og TClog GDPlog og Plog g og og
** ** ** ** ** ** * ** ** **
RECAPlog P corr 1 .681 .646 .994 .630 .580 .673 .258 .106 .610 .506 .401
Sig .000 .000 .000 .000 .000 .000 .033 .385 .000 .000 .001
** ** ** ** ** * * **
RECAPPClog P corr .681 1 .807 .691 .988 .785 .026 .232 -.264 .327 .348 .206
Sig .000 .000 .000 .000 .000 .832 .055 .028 .011 .007 .114
** ** ** ** **
RECAPGDPlog P corr .646 .807 1 .656 .796 .983 -.110 -.216 -.122 .101 .037 .102
Sig .000 .000 .000 .000 .000 .368 .075 .316 .442 .781 .439
** ** ** ** ** ** * ** ** **
RECAPTTlog P corr .994 .691 .656 1 .655 .608 .654 .258 .101 .585 .493 .384
Sig .000 .000 .000 .000 .000 .000 .032 .409 .000 .000 .002
** ** ** ** ** * *
RECAPTTTClog P corr .630 .988 .796 .655 1 .802 -.048 .215 -.281 .240 .286 .146
Sig .000 .000 .000 .000 .000 .695 .076 .019 .065 .028 .265
** ** ** ** ** *
RECAPTTGDPlo P corr .580 .785 .983 .608 .802 1 -.199 -.238 -.140 -.010 -.043 .026
g Sig .000 .000 .000 .000 .000 .101 .049 .251 .942 .748 .843
** ** ** ** ** ** **
REIMPlog P corr .673 .026 -.110 .654 -.048 -.199 1 .573 .546 .728 .591 .539
Sig .000 .832 .368 .000 .695 .101 .000 .000 .000 .000 .000
* * * ** ** ** ** **
REIMPPClog P corr .258 .232 -.216 .258 .215 -.238 .573 1 .462 .437 .582 .341
Sig .033 .055 .075 .032 .076 .049 .000 .000 .000 .000 .005
* * ** ** *
REIMPGDPlog P corr .106 -.264 -.122 .101 -.281 -.140 .546 .462 1 .211 .100 .289
Sig .385 .028 .316 .409 .019 .251 .000 .000 .092 .432 .020
** * ** ** ** ** **
REEXPlog P corr .610 .327 .101 .585 .240 -.010 .728 .437 .211 1 .884 .937
Sig .000 .011 .442 .000 .065 .942 .000 .000 .092 .000 .000
** ** ** * ** ** ** **
REEXPPClog P corr .506 .348 .037 .493 .286 -.043 .591 .582 .100 .884 1 .860
Sig .000 .007 .781 .000 .028 .748 .000 .000 .432 .000 .000
** ** ** ** * ** **
REEXPGDPlog P corr .401 .206 .102 .384 .146 .026 .539 .341 .289 .937 .860 1
Sig .001 .114 .439 .002 .265 .843 .000 .005 .020 .000 .000
**. Correlation is significant at the 0.01 level (2-tailed).
183
Correlations
RECAPPCl RECAPGDPl RECAPTTl RECAPTT RECAPTT REIMPPCl REIMPGD REEXPlo REEXPPCl REEXPGDPl
RECAPlog REIMPlog
og og og TClog GDPlog og Plog g og og
** ** ** ** ** ** * ** ** **
RECAPlog P corr 1 .681 .646 .994 .630 .580 .673 .258 .106 .610 .506 .401
Sig .000 .000 .000 .000 .000 .000 .033 .385 .000 .000 .001
** ** ** ** ** * * **
RECAPPClog P corr .681 1 .807 .691 .988 .785 .026 .232 -.264 .327 .348 .206
Sig .000 .000 .000 .000 .000 .832 .055 .028 .011 .007 .114
** ** ** ** **
RECAPGDPlog P corr .646 .807 1 .656 .796 .983 -.110 -.216 -.122 .101 .037 .102
Sig .000 .000 .000 .000 .000 .368 .075 .316 .442 .781 .439
** ** ** ** ** ** * ** ** **
RECAPTTlog P corr .994 .691 .656 1 .655 .608 .654 .258 .101 .585 .493 .384
Sig .000 .000 .000 .000 .000 .000 .032 .409 .000 .000 .002
** ** ** ** ** * *
RECAPTTTClog P corr .630 .988 .796 .655 1 .802 -.048 .215 -.281 .240 .286 .146
Sig .000 .000 .000 .000 .000 .695 .076 .019 .065 .028 .265
** ** ** ** ** *
RECAPTTGDPlo P corr .580 .785 .983 .608 .802 1 -.199 -.238 -.140 -.010 -.043 .026
g Sig .000 .000 .000 .000 .000 .101 .049 .251 .942 .748 .843
** ** ** ** ** ** **
REIMPlog P corr .673 .026 -.110 .654 -.048 -.199 1 .573 .546 .728 .591 .539
Sig .000 .832 .368 .000 .695 .101 .000 .000 .000 .000 .000
* * * ** ** ** ** **
REIMPPClog P corr .258 .232 -.216 .258 .215 -.238 .573 1 .462 .437 .582 .341
Sig .033 .055 .075 .032 .076 .049 .000 .000 .000 .000 .005
* * ** ** *
REIMPGDPlog P corr .106 -.264 -.122 .101 -.281 -.140 .546 .462 1 .211 .100 .289
Sig .385 .028 .316 .409 .019 .251 .000 .000 .092 .432 .020
** * ** ** ** ** **
REEXPlog P corr .610 .327 .101 .585 .240 -.010 .728 .437 .211 1 .884 .937
Sig .000 .011 .442 .000 .065 .942 .000 .000 .092 .000 .000
** ** ** * ** ** ** **
REEXPPClog P corr .506 .348 .037 .493 .286 -.043 .591 .582 .100 .884 1 .860
Sig .000 .007 .781 .000 .028 .748 .000 .000 .432 .000 .000
** ** ** ** * ** **
REEXPGDPlog P corr .401 .206 .102 .384 .146 .026 .539 .341 .289 .937 .860 1
Sig .001 .114 .439 .002 .265 .843 .000 .005 .020 .000 .000
**. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).
184
The selected indicators cannot be taken as representing clean energy TT per se, as they only
provide an incomplete picture of the TT process. Only imports of some renewable energy
technologies could be included as inputs into the process, while as shown by this thesis
there are many other channels through which foreign technologies can flow. The literature
review showed that other studies had used flows of foreign patents filed in local patent
offices as indicators of foreign technology inputs (Hascic and Johnstone, 2009;
Dechezlepretre et al., 2010, Verdolini and Galeotti, 2011). As regards outputs, exports and
installed capacity are imperfect indicators because they are only proxies of the production
of internationally competitive clean energy technologies and their use for electricity
generation. As such, they cannot reflect many other outputs of the TT process. Also, it is
difficult to estimate the importance of technology transfer in the final outputs measured as
exports or renewable energy capacity. The literature review showed that other authors had
used claims of TT in CDM projects as a variable to measure TT outputs (Dechezlepretre et
al., 2008; Doranova, 2009; Seres et al., 2009, 2010). However, these studies represented TT
only by indicating whether or not they claimed TT, without giving an indication of the scale
of the transfer. Besides, there are not widely available data on the effects of the TT such as
technology cost reductions, improvements in productivity or emissions reductions in
developing countries. A complete measurement of the TT process is therefore not possible
due to lack of data. In any case, the existing indicators can be used to map the performance
of different developing countries in three particular measurements of technology transfer
and to analyse the impacts of different enabling factors for TT in the performance of
5.2.2 Independent variables

The initial long list of indicators of enabling factors for clean energy technology transfer was
presented in Table 17 in the introduction to this chapter, according to the findings of the
literature review and the case study analysis in previous sections of this thesis. This list was
then reduced to a more manageable short list, taking the following criteria into
consideration: data availability, number of valid cases and correlations among variables of
the same type. When variables were highly correlated (Pearson correlation r of 0.9 and
above) only one of the variables would be selected. Annex 2, among Chapter 5’s annexes,
presents the initial long list and details the reasons why some variables could not be
included in the analysis. Some of the variables suggested in Table 17 were not possible to
retrieve, which was particularly the case for variables related to policies (industrial,
technology-push and demand pull policies) that are not easily quantifiable and for which
there is very scarce information from developing countries. Only one policy variable, related
to the existence or not (dummy variable) of feed-in-tariffs or guaranteed subsidised prices
for renewable generation, could be included, as collected from the IEA Policies and
Measures database. Other important variables for which no data was publicly available are:
• Specific R&D expenditure on clean energy technologies. These data are collected by
the IEA for OECD countries, but they are not available for most developing countries.
• Electricity prices. The IEA database can provide electricity prices for only a small
sample of developing countries, which is not sufficient for modelling purposes.
• Concentration of the electricity generation industry.
185
Workforce characteristics per education levels. Data were found about tertiary
enrolment ratios and the percentage of degrees in engineering and science. The
former are included, but the latter had a large number of missing values.
Table 27 shows the final set of variables that could be collected. When transformation by
taking logarithms was required for the normality of the distributions, it is indicated, as well
as the performed calculation. The final variable used is indicated in pale red.
186
TABLE 27-FINAL SET OF ENABLING FACTORS
TT aspect Type of Variables Transformed Calculation Title and description Source
variable
Flows of Economic EDB - Title: Ease of Doing Business rank, 2011 The World Bank
foreign and Description: EDB measures a combination of nine aspects: Starting a Business, Dealing
inputs institutional with Construction Permits, Registering Property, Getting Credit, Protecting Investors,
framework Paying Taxes, Trading Across Borders, Enforcing Contracts, Closing a Business. Countries
with the lowest rank are the best performers.
CPI CPIlog Ln(CPI*100 Title: Corruption Perception Index score, 2010 Transparency
) Description: The CPI ranks countries according to perception of corruption in the public International
sector. The most corrupt countries have the lowest scores.
IPR - Title: Intellectual Property Rights index score, 2010 Property Rights
Description: The three components of the IPR index are: Protection of Intellectual Alliance
Property Rights, Patent Protection and Copyright Piracy. This index also feeds the more
general index of Property Rights. The higher the score of the IPR index, the higher the
protection.
INCOME - Title: Average income tax rate, 2011 Heritage
TAX Description: Rate of income tax used for the calculation of the Index of Economic Foundation and
Freedom. Wall Street Jour
CRED CREDlog Ln(CRED) Title: Domestic credit to private sector as a percentage of GDP, 2009 data The World Bank
Description: Domestic credit to private sector refers to financial resources provided to indicators
the private sector, such as through loans, purchases of non-equity securities and trade
credits and other accounts receivable, that establish a claim for repayment. For some
countries these claims include credit to public enterprises.
Foreign TARIFF - Title: Most Favoured Nation average applied tariff rates applied for non agricultural World Trade
trade and goods, 2009. Statistics
investment Description: A high rate represents high protection.
TRADEO TRADEOPlog Ln(TRADE Title: Trade openness, 2009 World Trade
P OP) Description: Own calculation as imports plus exports divided by GDP. Statistics
FDIOP FDIOPlog Ln(FDIOP* Title: Foreign Direct Investment openness, 2009 World Bank
100) Description: Own calculation as FDI net inflows divided by GDP. indicators
INVESTF - Title: Index of investment freedom, 2011 Heritage
REE Description: The Index evaluates a variety of restrictions typically imposed on Foundation and
investment. Points are deducted from the ideal score of 100 for each of the restrictions the Wall Street
found in a country’s investment regime. High scores mean high levels of freedom. Journal
PATLOCp PATLOCpclo Ln(PATLOC Title: Stock of patents filed by local inventors during the period 1883-2009. WIPO statistics
187
variable
c g PC+1) Description: The stock is calculated following the perpetual inventory method with a 10% database and
discount rate. Values expressed per capita as number of patents per million inhabitants. own calculation
LOG - Title: Logistics performance index: Overall (1=low to 5=high), 2009 World Bank
Description: The score reflects perceptions of a country's logistics based on efficiency of
customs clearance process, quality of trade- and transport-related infrastructure, ease of
arranging competitively priced shipments, quality of logistics services, ability to track and
trace consignments, and frequency with which shipments reach the consignee within the
scheduled time.
Output Demand size GDP GDPlog Ln(GDP) Title: Gross Domestic Product in current Million US$ at purchasers' prices, 2009
GDPg - Title: Average GDP Growth between 2005 and 2009 World Bank
GDPpc GDPPClog Ln(GDPpc) Title: GDP per capita in current US$ World Bank
CO2pc CO2pclog Ln(CO2pc+ Title: CO2 emissions per capita in metric tons, 2007 World Bank
1) Description: Carbon dioxide emissions are those stemming from the burning of fossil
fuels and the manufacture of cement. They include carbon dioxide produced during
consumption of solid, liquid, and gas fuels and gas flaring.
Energy PDIES - Title: Pump price for diesel fuel (US$ per litre) 2010 World Bank
prices Description: Fuel prices refer to the pump prices of the most widely sold grade of diesel Indicators
fuel. Prices have been converted from the local currency to U.S. dollars.
Availability FOSSILpc FOSSILpclog Ln(FOSSILP Title: Production of fossil fuels, expressed as tons per million people, 2009 US Energy
of fossil C+1) Description: Overall production of primary coal, dry natural gas and oil, converted to heat Information
fuels values by the author using the gross heat content values of every fuel per country Agency
Demand-pull FIT Dummy Title: Countries that have implemented Feed-in tariffs or that provide guaranteed , 2011 IEA Policies and
policies premiums to renewable electricity generation. 1=yes, and 0=no feed-in tariffs Measures
Description: Values taken from IEA policies and measures database, as countries that database,2011
have implemented these policies in or before 2011.
Local Industrial HTEXPC HTEXPClog Ln(HTEXPC Title: High-technology exports as a percentage of manufactured exports World Bank
inputs developmen *100) Description: High-technology exports are products with high R&D intensity, such as in indicators
Technol t aerospace, computers, pharmaceuticals, scientific instruments, and electrical machinery.
ogy ISO9pc ISO9pclog Ln(ISO9pc+ Title: Number of companies with ISO 9001 certification, 2008. Values expressed as The ISO 9001
effect 1) companies per Million people Survey
TFP TFPlog Ln(TFP*10 Title: Projections of Total Factor Productivity levels relative to the US for 2005. UNIDO World
0) Description: The most recent data available is from 2000, 2005 values were projections. Productivity
Correlation between real 2000 data and projections for 2005 is very high (Pearson database
correlation of 0.961 significant at the 0.01 level), therefore we take the latest data.
188
variable
UNIDO´s calculation based on the default capital stock (K06), based on the perpetual
inventory method (PIM) with an annual depreciation rate of 6% and an initial capital
stock including ten years of investment.
CIP - Title: Competitive Industrial Performance score, 2009. UNIDO World
Description: The CIP index combines four main dimensions of industrial competitiveness: Industrial
industrial capacity, manufactured export capacity, industrialization intensity and export Development
quality. A high value indicates good performance Report 2009
Technologic PATFORp PATFORPClo Ln(PATFOR Title: Total stock of patents filed by foreign inventors between 1883 and2009. WIPO statistics
al c g PC+1) Description: The stock is calculated using the perpetual inventory method with a 10% and own
developmen discount rate. Values are expressed per capita as number of patents per million calculation
t inhabitants.
Enrol3 ENROL3log Ln(ENROL3 Title: Tertiary education school enrolment ratio, as a percentage of population, 2008 World Bank
) Description: Gross enrolment ratio is the ratio of total enrolment, regardless of age, to
the population of the age group that officially corresponds to the level of education
shown. Tertiary education, whether or not to an advanced research qualification,
normally requires, as a minimum condition of admission, the successful completion of
education at the secondary level.
Access to REACPC REACPClog Ln(REACCE Title: Estimated annual renewable energy resources for solar, hydro, wind, different kinds Buys et al, 2007
renewables SS+1) of biomass and geothermal energy. Values expressed in toe per thousand people.
Description: For solar, wind and geothermal, low, intermediate and high scenarios are
available. Only intermediate scenarios are taken to estimate the potential availability.
WSHACC WSHACCPCl Ln(WSHAC Title: Estimated renewable energy potential for wind, solar and hydro sources Values Buys et al, 2007
PC og CPC+1) expressed in toe per thousand people
Description: The differentiation is made because we only count on imports data for wind,
hydro and solar technologies and it may be useful to count on the specific potential of
these sources to find relationships between the variables.
189
Descriptive statistics for this list of variables are presented in Table 28. Correlations are
presented as part of the chapter on principal components analysis.
TABLE 28-FINAL SET OF EXPLANATORY VARIABLES
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation

EDB 120 11 183 112.57 46.866
CPIlog 122 .34 2.04 1.0728 .35233
IPR 79 2.30 7.30 4.5443 .99779
CREDlog 107 1.57 4.99 3.3353 .76305
GDPlog 122 4.85 15.42 9.7930 2.06761
GDPpclog 122 5.07 11.15 7.6085 1.20167
CO2PClog 123 .02 4.03 .9652 .83004
PDIES 110 .01 2.03 .9022 .36764
HTEXPClog 87 .01 4.20 1.5296 .96382
ISOPClog 124 .00 8.59 2.1882 1.91104
TFPlog 81 .01 .73 .2444 .12939
CIP 74 .04 .47 .1974 .08600
ENROL3log 76 .15 4.80 2.6390 1.13522
PATLOCPClog 98 .00 8.06 2.2224 2.16733
PATFORPClog 98 .00 9.72 2.6480 2.46631
TRADEOPlog 115 3.12 6.36 4.3651 .52970
TARIFF 119 .00 25.59 10.1704 4.95389
FDIOPlog 122 -.01 .28 .0407 .04594
LOG 105 1.70 3.63 2.6242 .35949
INCOMETAX 120 .00 60.00 28.1125 11.37459
WSHACCPClog 124 .00 6.29 .9088 .89814
REACPClog 124 .00 14.24 7.6351 2.28096
FOSSILPClog 122 .00 15.01 3.9260 3.62730
GDPg 124 -4.93 21.21 5.2405 3.35853
INVESTFREE 120 .00 90.00 42.1250 21.13047
Valid N (listwise) 21
5.3 Exploring indicators of enabling frameworks for technology transfer

This section explores the correlations between the selected indicators of enabling
frameworks for technology transfer, and then summarises the information conveyed by
these indicators through principal component analysis. The aim of the section is to map the
performance of developing countries in terms of these principal components.
5.3.1 Principal Components Analysis

The aim of principal components analysis (PCA) is to describe the variation in a set of
correlated variables (in our case, the indicators of enabling factors for TT) in terms of a new
set of uncorrelated variables, which are called ‘principal components’ (PCs). Each principal
component is a linear combination of the original variables. The amount of information
conveyed by each principal component can be measured by its variance, which represents
the variation of the original variables that is captured by the new PC, and PCs are measured
in order of decreasing variance. The most informative PC is the first. A second PC can be
190
chosen to account for as much as possible of the remaining variation. Usually, two PCs are
enough to represent most of the variation in the original variables.
PCA is useful for reducing the dimensionality of our current representation of the enabling
factors technology transfer, but without losing too much of the information. It is an
exploratory technique that allows a better understanding of the interrelationships among
the variables. Three main steps are followed in performing PCA: assessing the suitability of
the data for PCA, the extraction of principal components and the interpretation of principal
components. The analysis will be useful to map how different developing countries perform
in the wide selection of indicators about enabling frameworks for low-carbon technology
transfer. The large numbers of indicators of enabling frameworks prevent a straightforward
interpretation of developing country performances, which is solved by summarising the
information conveyed by more than 20 indicators into just two or three.
Two main issues should be considered when determining if a particular dataset is suitable
for PCA analysis: the sample size and the strength of the relationship between the variables
(or items). Some authors suggest that the ratio of cases (number of countries in our study)
to variables should be ten to one (Nunnally, 1978), while others suggest that five cases for
each item is adequate in most cases (Tabachnick and Fidell, 2001). An inspection of the
correlation matrix is required for evidence of coefficients greater than 0.3, and if few
correlations above this level are found, then PCA may not be appropriate.
The extraction of principal components involves determining the smallest number of
principal components that can explain the variation of the original variables. All variables
will be standardised, since they were collected from different sources and endowed with
varying scales, units and ranges. Standardised data, with 0 mean and 1 standard deviation,
will be calculated according to the following formula:
The Kaiser Meyer Olkin measure of sampling adequacy (KMO) and the Barlett’s Test of
Sphericity value will be used to test the appropriateness of PCA for our sample50. For the
selection of components we will use the Kaiser’s criterion or the eigenvalue rule. The
eigenvalue of a factor represents the amount of the total variance explained by that factor.
Using this rule, only factors with an eigenvalue of 1.0 or more are retained for further
investigation. The choice of principal components can be validated by the Scree plot of the
eigenvalues. Catell (1966) recommended retaining all factors above the elbow, or a break in
the plot, as these factors contribute the most to explaining any variance in the dataset.
The Component Matrix shows the “load factors” or correlations of each of the initial
variables on the three components, while the component score coefficient matrix shows the
coefficients of the lineal combination of variables to calculate each principal component.
5.3.2 Suitability of the data for PCA

Table 29 shows the number of valid cases and the correlation with other variables for each
indicator. Correlations not significant at the 0.001 level are shaded in grey, and variables
with a low correlation with the rest of variables are shaded in grey in the headlines. In the
50
A KMO value above 0.6 and a p-value lower than 0.05 for the Barlett´s test indicate appropriateness.
191
meantime, variables with a large number of missing values are shaded in pale red and those
with both a high number of missing values and low correlation with the rest of variables are
shaded in bright red.
The variables related to access to renewable energy sources per capita do not have a highly
significant correlation with any of the economic, institutional, industrial and technology
variables, whereas those related to the trade and foreign investment openness only show
significant correlations between each other and another variable (CO2 emissions per capita
is correlated to trade openness and GDP is correlated to foreign investment openness). It is
interesting to note the positive correlation between trade openness and CO2 emissions per
capita. GDP growth, the level of IPR protection and the investment freedom performance of
the countries do not show many significant correlations either. The variables high
technology exports per capita, total factor productivity, competitive industrial performance
and enrolment in tertiary education have a high number of missing values.
All the shaded variables were removed for the PCA, which allowed for the number of
countries that would be included in the PCA to increase from 21 with 25 variables, to 61
with 14 variables highly correlated between each other and with a low number of missing
cases.
192
TABLE 29- SUITABILITY OF PCA THROUGH THE ANALYSIS OF SAMPLE SIZE AND CORRELATIONS AMONG VARIABLES
PATL PATF TRAD INVES INCO WSHA FOSSI
CPIlo CRED GDPl GDPp CO2P HTEX ISOP TFPlo ENRO OCPC ORPC EOPlo TARIF TFRE FDIO META CCPC REAC LPClo
EDB g IPR log og GDPg clog Clog PClog Clog g CIP L3log log log g F E Plog LOG X log PClog g PDIES
EDB P 1 -.560 ** -.272* -.637** -.297** -.085 -.564 ** -.521 ** .043 -.616** -.593** -.349 ** -.547 ** -.339 ** -.229* -.221 * .438** -.407 ** .005 -.439 ** .372 ** -.065 .099 -.218 * .159
Sig. (2t) .000 .016 .000 .001 .357 .000 .000 .694 .000 .000 .002 .000 .001 .025 .018 .000 .000 .961 .000 .000 .481 .280 .018 .103
CPIlog P -.560** 1 .632 ** .553** .073 -.070 .475 ** .367 ** -.073 .505** .505** .254 * .247 * .038 .082 .177 -.236 * .512 ** .056 .410 ** -.150 -.012 -.159 -.026 .133
Sig. (2t) .000 .000 .000 .425 .442 .000 .000 .510 .000 .000 .030 .032 .713 .425 .061 .011 .000 .543 .000 .104 .899 .080 .774 .166
IPR P -.272* .632 ** 1 .390** .256 * -.012 .333 ** .340 ** -.028 .340** .335** .231 .038 .030 .233* -.058 -.008 .254 * -.165 .473 ** -.035 .016 -.073 .193 .044
Sig. (2t) .016 .000 .001 .023 .916 .003 .002 .818 .002 .008 .067 .787 .801 .049 .618 .945 .024 .147 .000 .756 .892 .523 .093 .701
CREDlo P -.637** .553 ** .390 ** 1 .301** .006 .485 ** .484 ** .147 .659** .509** .548 ** .615 ** .380 ** .402 ** .244 * -.319** .173 -.059 .597 ** -.305 ** -.088 -.235 * .121 -.168
g Sig. (2t) .000 .000 .001 .002 .948 .000 .000 .199 .000 .000 .000 .000 .000 .000 .014 .001 .079 .549 .000 .002 .370 .015 .220 .105
GDPlog P -.297** .073 .256* .301** 1 .246 ** .458 ** .492 ** .342 ** .636** .366** .517 ** .583 ** .535 ** .801 ** -.253** -.126 .031 -.299 ** .615 ** -.099 .026 .245 ** .646 ** -.370 **
Sig. (2t) .001 .425 .023 .002 .006 .000 .000 .001 .000 .001 .000 .000 .000 .000 .007 .175 .739 .001 .000 .285 .773 .007 .000 .000
GDPg P -.085 -.070 -.012 .006 .246** 1 .139 .164 -.082 .020 -.124 -.013 .044 -.023 -.103 -.090 -.105 .004 .053 .020 -.191* -.036 .059 .245 ** -.248 **
Sig. (2t) .357 .442 .916 .948 .006 .126 .070 .451 .825 .269 .910 .704 .824 .312 .338 .256 .962 .561 .840 .037 .693 .517 .007 .009
GDPpcl P -.564** .475 ** .333 ** .485** .458** .139 1 .858 ** -.011 .732** .825** .377 ** .725 ** .381 ** .322 ** .235 * -.343** .187 * .023 .530 ** -.381 ** .166 -.036 .548 ** -.441 **
og Sig. (2t) .000 .000 .003 .000 .000 .126 .000 .923 .000 .000 .001 .000 .000 .001 .012 .000 .042 .803 .000 .000 .068 .692 .000 .000
CO2PCl P -.521** .367 ** .340 ** .484** .492** .164 .858 ** 1 .015 .707** .760** .384 ** .678 ** .498 ** .333 ** .313** -.327** -.028 .018 .539 ** -.522 ** .157 .001 .663 ** -.543 **
og Sig. (2t) .000 .000 .002 .000 .000 .070 .000 .891 .000 .000 .001 .000 .000 .001 .001 .000 .765 .846 .000 .000 .083 .989 .000 .000
HTEXPC P .043 -.073 -.028 .147 .342** -.082 -.011 .015 1 .179 .089 .582 ** .159 .252* .432 ** -.106 .038 -.091 -.078 .377 ** .216* -.190 -.094 .106 .136
log Sig. (2t) .694 .510 .818 .199 .001 .451 .923 .891 .098 .497 .000 .230 .027 .000 .339 .728 .403 .474 .001 .046 .079 .386 .336 .222
ISOPClo P -.616** .505 ** .340 ** .659** .636** .020 .732 ** .707 ** .179 1 .746** .534 ** .722 ** .624 ** .596 ** .098 -.313** .204 * -.129 .679 ** -.318 ** .146 .087 .499 ** -.270 **
g Sig. (2t) .000 .000 .002 .000 .000 .825 .000 .000 .098 .000 .000 .000 .000 .000 .295 .001 .025 .156 .000 .000 .105 .335 .000 .004
TFPlog P -.593** .505 ** .335 ** .509** .366** -.124 .825 ** .760 ** .089 .746** 1 .362 ** .702 ** .546 ** .381 ** .301** -.256 * .230 * .086 .425 ** -.232* .060 -.201 .374 ** -.367 **
Sig. (2t) .000 .000 .008 .000 .001 .269 .000 .000 .497 .000 .005 .000 .000 .002 .007 .021 .039 .446 .000 .037 .593 .072 .001 .002
CIP P -.349** .254 * .231 .548** .517** -.013 .377 ** .384 ** .582 ** .534** .362** 1 .341 * .435 ** .607 ** .166 -.288 * .096 -.128 .652 ** .086 -.179 -.282 * .256 * -.041
Sig. (2t) .002 .030 .067 .000 .000 .910 .001 .001 .000 .000 .005 .018 .000 .000 .161 .013 .418 .276 .000 .470 .127 .015 .030 .735
193
PATL PATF TRAD INVES INCO WSHA FOSSI
CPIlo CRED GDPl GDPp CO2P HTEX ISOP TFPlo ENRO OCPC ORPC EOPlo TARIF TFRE FDIO META CCPC REAC LPClo
EDB g IPR log og GDPg clog Clog PClog Clog g CIP L3log log log g F E Plog LOG X log PClog g PDIES
ENROL P -.547** .247 * .038 .615** .583** .044 .725 ** .678 ** .159 .722** .702** .341 * 1 .711 ** .529 ** .320** -.457** .060 -.014 .438 ** -.370 ** .036 .130 .453 ** -.384 **
3log Sig. (2t) .000 .032 .787 .000 .000 .704 .000 .000 .230 .000 .000 .018 .000 .000 .006 .000 .610 .906 .000 .001 .758 .264 .000 .001
PATLO P -.339** .038 .030 .380** .535** -.023 .381 ** .498 ** .252* .624** .546** .435 ** .711 ** 1 .713 ** -.019 -.180 -.076 -.119 .419 ** -.183 .237 * .327 ** .428 ** -.077
CPClog Sig. (2t) .001 .713 .801 .000 .000 .824 .000 .000 .027 .000 .000 .000 .000 .000 .854 .081 .464 .242 .000 .074 .019 .001 .000 .472
PATFO P -.229* .082 .233* .402** .801** -.103 .322 ** .333 ** .432 ** .596** .381** .607 ** .529 ** .713 ** 1 -.238 * .031 -.048 -.243 * .606 ** .138 .011 .125 .474 ** -.051
RPClog Sig. (2t) .025 .425 .049 .000 .000 .312 .001 .001 .000 .000 .002 .000 .000 .000 .021 .769 .639 .016 .000 .179 .913 .220 .000 .636
TRADE P -.221* .177 -.058 .244 * -.253** -.090 .235 * .313 ** -.106 .098 .301** .166 .320 ** -.019 -.238* 1 -.176 -.036 .508 ** .067 -.193* .036 -.118 -.007 -.148
OPlog Sig. (2t) .018 .061 .618 .014 .007 .338 .012 .001 .339 .295 .007 .161 .006 .854 .021 .064 .700 .000 .514 .040 .700 .209 .945 .138
TARIFF P .438** -.236 * -.008 -.319** -.126 -.105 -.343 ** -.327 ** .038 -.313** -.256 * -.288 * -.457 ** -.180 .031 -.176 1 -.334 ** -.053 -.228* .447 ** -.056 .050 -.031 .100
Sig. (2t) .000 .011 .945 .001 .175 .256 .000 .000 .728 .001 .021 .013 .000 .081 .769 .064 .000 .570 .021 .000 .545 .591 .741 .307
INVEST P -.407** .512 ** .254* .173 .031 .004 .187 * -.028 -.091 .204 * .230 * .096 .060 -.076 -.048 -.036 -.334** 1 .053 .257 ** -.130 -.099 -.116 -.249 ** .292 **
FREE Sig. (2t) .000 .000 .024 .079 .739 .962 .042 .765 .403 .025 .039 .418 .610 .464 .639 .700 .000 .569 .009 .156 .282 .207 .006 .002
FDIOPlo P .005 .056 -.165 -.059 -.299** .053 .023 .018 -.078 -.129 .086 -.128 -.014 -.119 -.243* .508** -.053 .053 1 -.098 .008 .053 -.086 -.140 .061
g Sig. (2t) .961 .543 .147 .549 .001 .561 .803 .846 .474 .156 .446 .276 .906 .242 .016 .000 .570 .569 .320 .930 .562 .348 .128 .530
LOG P -.439** .410 ** .473 ** .597** .615** .020 .530 ** .539 ** .377 ** .679** .425** .652 ** .438 ** .419 ** .606 ** .067 -.228 * .257 ** -.098 1 -.172 -.109 -.238 * .364 ** -.097
Sig. (2t) .000 .000 .000 .000 .000 .840 .000 .000 .001 .000 .000 .000 .000 .000 .000 .514 .021 .009 .320 .086 .266 .015 .000 .347
INCOME P .372** -.150 -.035 -.305** -.099 -.191 * -.381 ** -.522 ** .216* -.318** -.232 * .086 -.370 ** -.183 .138 -.193 * .447** -.130 .008 -.172 1 -.116 .031 -.251 ** .470 **
TAX Sig. (2t) .000 .104 .756 .002 .285 .037 .000 .000 .046 .000 .037 .470 .001 .074 .179 .040 .000 .156 .930 .086 .208 .737 .006 .000
WSHAC P -.065 -.012 .016 -.088 .026 -.036 .166 .157 -.190 .146 .060 -.179 .036 .237* .011 .036 -.056 -.099 .053 -.109 -.116 1 .725 ** .340 ** -.143
CPClog Sig. (2t) .481 .899 .892 .370 .773 .693 .068 .083 .079 .105 .593 .127 .758 .019 .913 .700 .545 .282 .562 .266 .208 .000 .000 .137
REACP P .099 -.159 -.073 -.235* .245** .059 -.036 .001 -.094 .087 -.201 -.282 * .130 .327 ** .125 -.118 .050 -.116 -.086 -.238* .031 .725** 1 .309 ** .046
Clog Sig. (2t) .280 .080 .523 .015 .007 .517 .692 .989 .386 .335 .072 .015 .264 .001 .220 .209 .591 .207 .348 .015 .737 .000 .001 .637
FOSSIL P -.218* -.026 .193 .121 .646** .245 ** .548 ** .663 ** .106 .499** .374** .256 * .453 ** .428 ** .474 ** -.007 -.031 -.249 ** -.140 .364 ** -.251 ** .340** .309 ** 1 -.610 **
PClog
Sig. (2t) .018 .774 .093 .220 .000 .007 .000 .000 .336 .000 .001 .030 .000 .000 .000 .945 .741 .006 .128 .000 .006 .000 .001 .000
PDIES P .159 .133 .044 -.168 -.370** -.248 ** -.441 ** -.543 ** .136 -.270** -.367** -.041 -.384 ** -.077 -.051 -.148 .100 .292 ** .061 -.097 .470 ** -.143 .046 -.610 ** 1
Sig. (2t) .103 .166 .701 .105 .000 .009 .000 .000 .222 .004 .002 .735 .001 .472 .636 .138 .307 .002 .530 .347 .000 .137 .637 .000
Valid cases 120 122 79 107 122 124 122 123 87 124 81 74 76 98 98 115 119 120 122 105 120 124 124 122 110
194
5.3.3 First iteration with per capita values
Our first iteration of the principal component analysis includes 14 highly correlated
variables with a low number of missing values and some variables expressed in per
capita values. The KMO and Barlett’s tests show that our sample is appropriate for
performing principal component analysis.
Three principal components can explain 72% of the total variance of the 14 variables,
as shown below.
The Scree plot confirms the selection of three principal components for further
analysis, as after the third component the curve of eigenvalues tends to flatten.
195
The matrix of communalities shows the part of each variable´s variance that can be
jointly explained by the three selected principal components.
The Component Matrix shows the “load factors” or correlations of each of the initial
variables with the three components. Only correlations above 0.3 are considered
significant. Those below are shaded in grey in the component matrix. Most of the
items load quite strongly (above 0.4) on the first component, and the number of
variables with strong loading is reduced to three in the last component.
196
Component
1 2 3
Zscore(ISOPClog) .906 -.007 .163
Zscore(CO2PClog) .859 -.077 -.308
Zscore(GDPpclog) .828 -.159 -.173
Zscore(LOG) .751 .074 .329
Zscore(GDPlog) .721 .525 .037
Zscore(CREDlog) .698 -.322 .314
Zscore(EDB) -.688 .438 -.148
Zscore(PATLOCPClog) .650 .354 .183
Zscore(FOSSILPClog) .644 .454 -.434
Zscore(PATFORPClog) .628 .607 .399
Zscore(CPIlog) .465 -.565 .407
Zscore(TARIFF) -.398 .507 .096
Zscore(PDIES) -.436 -.113 .767
Zscore(INCOMETAX) -.448 .429 .537
Extraction Method: Principal Component Analysis.
a. 3 components extracted.
The load factors of the different variables are presented graphically in Figure 54 and
Figure 55. Most items load quite strongly on the first component, which therefore
represents a combination of all the identified variables that can have an impact on
clean energy technology transfer. The second and third components capture the
variation of specific variables, as detailed afterwards.
FIGURE 54- LOAD FACTORS OF PRINCIPAL COMPONENTS 1 AND 2
197
FIGURE 55: LOAD FACTORS OF PRINCIPAL COMPONENTS 1 AND 3
The load factors of the different variables provide a straightforward interpretation of

the principal components obtained:
• The first principal component is higher, per order of importance, for countries
with a large number of high quality private businesses, high levels of CO2
emissions per capita, high levels of income per capita, a good logistics system, a
large economy, credit availability for the private sector, ease of doing business,
a large stock of patents filed by local inventors per capita, large fossil fuel
resources per capita, a large stock of foreign patents, low corruption, low
income taxes, low diesel prices and low tariffs. Countries with a high value for
the first PC are therefore expected to be particularly well suited to receive large
amounts of technology transfer in general.
• The second principal component is high, per order of importance, for countries
with a large stock of foreign patents per capita, high levels of corruption, large
economies, high tariffs, large fossil fuel resources per capita, where it is difficult
to do business, income taxes are high, there is a large stock of local patents and
it is difficult to get credit for the private sector. Countries with a high value for
the second PC would be expected to face some barriers to achieving large
levels of TT and would require reforms to reduce their level of corruption and
improve their environment for foreign investment. Besides, the large fossil fuel
resources per capita may render incompetitive alternative energies in the
absence of supportive demand-pull policies.
• The third principal component is higher, per order of importance, for countries
with high fossil fuel prices, high taxes on workers’ income, low fossil fuel
resources per capita, low corruption, a large stock of foreign patents per capita,
a functioning logistics system, access to credit and low CO2 emissions per
198
capita. These countries are expected to have a significant demand for clean
energy technologies, given the high prices of fossil fuels, as well as a favourable
business environment. They are countries that show favourable conditions to
receive significant levels of TT.
PC 1 PC 2 PC 3
Top 10 performers Saudi Arabia Iran Turkey

Qatar Russian Federation China
Malaysia China Chile
Bahrain Algeria South Africa
China Brazil Thailand
South Africa India Uruguay
Kazakhstan Congo, Rep. Brazil
Chile Argentina India
Russian Federation Mexico Mauritius
Thailand Indonesia Vietnam
Bottom 10 Zambia Lebanon Azerbaijan
performers Nepal Georgia Syria
Uganda El Salvador Bolivia
Madagascar Costa Rica Algeria
Tanzania Namibia Saudi Arabia
Cambodia Botswana Bahrain
Burkina Faso Oman Yemen
Congo, Rep. Bahrain Oman
Mali Qatar Qatar
Sierra Leone Mauritius Angola
5.3.4 Second iteration with per GDP values

The second iteration involves replacing variables expressed in per capita values per
variables expressed in values per unit of GDP. The correlations of these new variables,
presented in Table 30, show some differences with values per capita. In particular,
access to renewable energy resources per unit of GDP now has a significant negative
correlation with many variables, as they use the GDP value as a denominator, which is
positively correlated to many variables. This shows that correlations are only a result
of the GDP, and therefore they should not be included in the analysis. Correlations of
the rest of the variables are not very different to those in per capita values. Therefore,
it is not considered necessary to perform a second iteration using values relative to
GDP instead of per capita values.
199
TABLE 30- CORRELATIONS OF VARIABLES EXPRESSED IN VALUES PER GDP
5.3.5 Summary of results

The principal component analysis using per capita values has shown that the
performance of developing countries in 14 indicators of enabling frameworks for TT
200
can be summarised by three variables, called “principal components”. Figure 56 and
Figure 57 classify the sample of countries according to their rating in the first, second
and third principal components. Country codes used to interpret the data in the figures
can be found in Annex 3 of Chapter 5.
Figure 56 positions developing countries in our sample according to their performance
in the two first principal components. The graph is split into four quadrants to facilitate
the interpretation of country performance in the two principal components. Countries
positioned to the right of the first and second quadrants have a significant size and
good enabling factors for the transfer of technologies. Their only potential problem in
attracting foreign clean energy technology transfer could be the lack of demand, as
these countries have large production levels of fossil fuels, particularly those at the top
of quadrant 1. Additionally, countries in quadrant 2, or those in quadrant 1, that are
closer to the division line would be expected to have a more favourable environment
for private sector activities, due to their lower levels of corruption, tariffs, taxes and
ease of doing business.
Figure 57 positions the developing countries in the sample according to their
performance in principal components 1 and 3. High scores in PC1 show good enabling
conditions for TT in general, whereas high scores in PC3 show good enabling conditions
for low-carbon TT in particular, as it rates higher for countries with low production of
fossil fuels and high fossil fuel prices. Therefore, countries in quadrant 1 should be
considered the most likely to benefit from foreign clean energy technology transfer, in
per capita values. These include most clearly Malaysia, China, South Africa, Chile,
Thailand, Mexico, Brazil and Turkey. Uruguay, India, Argentina, Mauritius, Tunisia,
Colombia, Peru and Panama are also part of this quadrant, although they show a
worse performance in most enabling factors.
201
FIGURE 56 - COUNTRY PERFORMANCE PRINCIPAL COMPONENTS 1 AND 2
4 1
3 2
FIGURE 57 - COUNTRY PERFORMANCE PRINCIPAL COMPONENTS 1 AND 3
4 1
3 2
202
5.4 Analysing the relationship between enabling frameworks and indicators of
clean energy technology transfer
The previous analysis was useful for understanding the variables that could have an
impact on clean energy technology transfer and identifying the most attractive
countries according to their performance in these variables. The previous steps
allowed us to identify highly correlated independent variables with high missing
values, as well as groups of independent variables in relation to how they jointly affect
developing countries. However, PCA did not provide any information about their actual
effect on technology transfers.
We will carry out an OLS regression analysis to show the relationships between these
explanatory variables and some indicators of technology transfer: imports and exports
of a sample of renewable energy technologies and renewable generation installed
capacity. The dependent variables are expressed in per capita values to avoid the scale
effect.
A cross-sectional dataset is used for the analysis, with observations taken at a specific
moment in time (2009 for dependent variables REIMPPClog and REEXPPClog, and 2008
for dependent variable RECAPPClog) related to developing countries.
Several publications provide a good background for the proposed OLS method, such as
Dougherty (2011), Wooldridge (2006) or Peña (1989). This section provides a brief
introduction to the selected statistical approach, but further information can be
obtained from the referred literature.
Regression models are used to study the relationship between one dependent or
explained variable and one or several independent or explanatory variables. The
general multiple linear regression model relates to explained and explanatory variables
through the linear function
Y= β0 + β1x1+ β2 x2+...+ βkxk+u
where Y is the explained variable, β0 is the intercept, β1 is the parameter associated
with explanatory variable x1 and k is the number of independent variables in the
model. Variable u is the error term or disturbance and it collectively contains the
factors other than x1, x2, ...xk that affect y. The key assumptions for a regression model
using cross-sectional data with non-stochastic regressors (their values do not have
random components) are:
1. The model is linear in parameters and correctly specified.
2. There is some variation in the regressor in the sample.
3. The disturbance term has zero expectation- E (u) =0.
4. The disturbance term is homoskedastic. Var (ui) = σ2 = constant. This means
that the variance of the disturbance term is constant and does not depend on
the value of the explanatory variables.
203
5. The values of the disturbance term have independent distributions. E (ui, uj).
This means that the observations are independent and one observation does
not offer any information about the next observation.
6. The disturbance term has a normal distribution. Ui ~N (O, σ2).
7. An exact linear relationship between the regressors does not exist.
To estimate parameters β0 ... βk the ordinary least squares (OLS) method is used. This
method minimises the sum of squared vertical distances between the observed
responses in the dataset and the responses predicted by the linear approximation (the
residual sum of squares or RSS). The main properties of the OLS estimators are
unbiasedness, efficiency and precision. Unbiasedness of OLS estimators indicates that
the procedure by which OLS estimates are obtained is unbiased, when this procedure
is applied across all possible random samples. In our case, we expect to have a sample
of developing countries that gives us an estimate close to the population value,
although this cannot be assured. OLS estimators are precise because among linear and
unbiased estimators they have the smallest variance. The efficiency of OLS estimators
is demonstrated by the Gauss-Markov Theorem, which justifies the use of the OLS
method rather than using a variety of competing estimators, defining them as the best
linearly unbiased estimators (BLUE) of β0, β1, ...,βk. when the standard set of
assumptions previously defined holds (Wooldridge, 2006).
For social science research, about 15 subjects per explanatory variable are needed for
a reliable equation (Stevens, 2002). Table 31 shows the maximum advised number of
predictors for each dependent variable considering the number of valid cases.
TABLE 31- NUMBER OF VALID CASES FOR DEPENDENT VARIABLES
Maximum number
N of predictors
RECAPPClog 115 8
REIMPPClog 79 5
REEXPPClog 66 4
Valid N (list-wise) 63
For the selection of explanatory variables in our regression analysis, we will apply a
stepwise approach with multiple iterations until the best fitted model is defined.
5.4.1 Explaining exports of renewable energy technologies per capita

The best fitted model for the variable exports of hydro, wind and solar renewable
energy technologies per capita (REEXPPClog) is defined by the following equation:
REEXPPClog = -7.227 + 0.889 REACPClog + 0.589 IPR + 17.283 CIP + 0.396 GDPlog
This model was the best fit after carrying out several iterations that are detailed in
Annex 5 of Chapter 5.
The model explains 69% of the variation of REEXPPClog through the availability of
hydro, wind and solar resources per capita of the exporter, the size of its economy, the
level of intellectual property rights protection and the competitive industrial
performance.
204
The following tables show the characteristics of the model.
Model Summary
Adjusted R Std. Error of the

Model R R Square Square Estimate Durbin-Watson
a
1 .831 .691 .667 1.61318 1.783
a. Predictors: (Constant), GDPlog, WSHACCPClog, IPR, CIP
b. Dependent Variable: REEXPPClog
ANOVA
Model Sum of Squares df Mean Square F Sig.
a
1 Regression 291.313 4 72.828 27.986 .000
Residual 130.117 50 2.602
Total 421.430 54
a. Predictors: (Constant), GDPlog, WSHACCPClog, IPR, CIP
b. Dependent Variable: REEXPPClog
Coefficients
Unstandardised Standardised
Coefficients Coefficients Collinearity Statistics
Model B Std. Error Beta t Sig. Tolerance VIF
1 (Constant) -7.227 1.322 -5.466 .000
WSHACCPClog .889 .251 .286 3.541 .001 .947 1.056
IPR .589 .229 .210 2.569 .013 .920 1.087
CIP 17.283 3.085 .532 5.602 .000 .685 1.461
GDPlog .396 .127 .293 3.120 .003 .700 1.429
a. Dependent Variable: REEXPPClog
All the coefficients are significant at the 0.05 level, and all except IPR at the 0.01 level.
The coefficients indicate that:
• An increase of 1% in the endowment of wind, solar and hydro resources per
capita results in an increase of 0.9% in the exports of wind, solar and hydro
technologies per capita, all other factors being equal.
• An increase of 1 in the score of the IPR index leads to a 59% increase in the
exports of wind, solar and hydro technologies per capita, when all other
factors remain equal. The IPR index has a value between 8.5 for the best
performers in the world (Finland and Sweden), or 7.3 for the best performer in
our sample (South Africa), and 2.3 for the worst performers in our sample
(Georgia and Moldova).
• An increase of 0.1 in the competitive industrial performance score increases by
173% the exports of hydro, wind and solar technologies per capita, all other
factors remaining equal. The CIP index is a value between 0 and 1. The
minimum value in our sample is 0.04 for Ethiopia and the maximum 0.47 for
205
Malaysia. Increases of 0.1 in the CIP value would be equivalent to the
improvement from the situation of Ethiopia (0.04) to that of Kenya (0.14), from
Kenya to Morocco (0.242), from Morocco to Poland (0.33) or Mexico (0.379),
and from these to Malaysia.
• A 1% increase in the size of the exporting economy increases its exports of
clean energy technologies per capita by 0.4%, all other factors remaining
equal.
The plot of residuals vs predicted values, to test the hypothesis of linearity,
homoscedasticity and normality, is included as part of the annexes.
The ‘collinearity diagnostics’ on explanatory variables presented in the coefficients
table can pick up on problems with multicollinearity that may not be evident in the
correlation matrix. Two values are given, namely ‘Tolerance’ and ‘VIF’. Tolerance is an
indicator of how much of the variability of the specified independent is not explained
by the other independent variables in the model, and it is calculated using the formula
1–R2 for each variable. If this value is very small (less than 0.10), it indicates that the
multiple correlation with other variables is high, suggesting the possibility of
multicollinearity. The other value given is the VIF (variance inflation factor), which is
just the inverse of the Tolerance value (1 divided by Tolerance). VIF values above 10
would be a concern here, indicating multicollinearity. Our model is therefore
acceptable for multicollinearity. Also, the residuals present a normal distribution,
meeting the hypothesis of linear regression.
5.4.2 Explaining imports of renewable energy technologies per capita

The best fit model for the variable on imports of renewable energy technologies per
capita (REIMPPClog) is defined by the following equation:
REIMPPClog = -1.108 + 0.668GDPplog+ 0.688CREDlog
The model explains 59% of the variation of imports of renewable energy technologies
per capita through the income per capita of the importing economy and its access to
credit for the private sector. Variables such as access to renewable energy sources per
capita, the price and access to fossil fuels and the existence of feed-in-tariffs or other
policies guaranteeing a subsidised price for renewable energy generation are not
significant for explaining imports in developing countries, as shown in Annex 5 of
Chapter 5, with several iterations performed until arriving at the final model.
Model Summary
Model R R Square Square Estimate Durbin-Watson
a
1 .770 .593 .582 .96785 2.142
a. Predictors: (Constant), CREDlog, GDPpclog
b. Dependent Variable: REIMPPClog
b
ANOVA
a
1 Regression 94.309 2 47.154 50.340 .000
Residual 64.634 69 .937
206
Total 158.943 71
a. Predictors: (Constant), CREDlog, GDPpclog
b. Dependent Variable: REIMPPClog
a
Coefficients
Unstandardised Standardised
Coefficients Coefficients Collinearity Statistics
Model B Std. Error Beta t Sig. Tolerance VIF
1 (Constant) -1.108 .756 -1.466 .147
GDPpclog .668 .109 .536 6.112 .000 .765 1.307
CREDlog .688 .172 .351 3.998 .000 .765 1.307
a. Dependent Variable: REIMPPClog
All the coefficients are significant at the 0.001 level, and indicate that:
• An increase of 1% in the income per capita of the recipient economy increases
by 0.67% its imports of wind, solar and hydro technologies per capita, when all
other factors are kept equal. This may indicate a higher demand for
environmental technologies as income level increases.
• An increase of 1% in access to credit by the private sector in the recipient
economy increases by 0.69% the imports of wind, solar and hydro technologies
per capita, when all other factors are kept equal.
5.4.3 Explaining renewable electricity generation capacity with technology transfer per
capita
The best fit model for the variable renewable generation capacity with technology
transfer per capita (RECAPTTPC) is defined by the following equation:
RECAPTTPClog = -5.336 + 0.814GDPpcL+0.284 REACPClog – 0.679 IPR + 1.581 LOG -
0.132 FOSSILpclog
The model explains 40% of the variation in renewable electricity generation per capita
in developing countries, with four variables: GDP per capita, access to renewable
energy resources per capita, IPR protection, logistic performance and fossil fuels
production per capita.
b
Model Summary
Model R R Square Square Estimate
a
1 .637 .406 .360 1.39605
a. Predictors: (Constant), FOSSILPClog, IPR, REACPClog, GDPpclog,
LOG
b. Dependent Variable: RECAPTTPClog
b
ANOVA
207
a
1 Regression 86.449 5 17.290 8.871 .000
Residual 126.683 65 1.949
Total 213.132 70
a. Predictors: (Constant), FOSSILPClog, IPR, REACPClog, GDPpclog, LOG
b. Dependent Variable: RECAPTTPClog
a
Coefficients
Standardised
Unstandardised Coefficients Coefficients
Model B Std. Error Beta t Sig.
1 (Constant) -5.336 1.866 -2.859 .006
GDPpclog .814 .186 .561 4.385 .000

REACPClog .284 .084 .371 3.362 .001
IPR -.679 .191 -.388 -3.546 .001
LOG 1.581 .626 .326 2.523 .014
FOSSILPClog -.132 .062 -.275 -2.150 .035
a. Dependent Variable: RECAPTTPClog
a
Residuals Statistics
Minimum Maximum Mean Std. Deviation N

Predicted Value .8670 6.4504 3.9849 1.12349 71
Std. Predicted Value -2.433 2.591 .373 1.011 71
Standard Error of Pred. value .220 .744 .389 .113 71
Adjusted Predicted Value .7652 7.1555 3.9585 1.16013 63
Residual -3.98877 2.27546 -.19415 1.27778 63
Std. Residual -2.857 1.630 -.139 .915 63
Stud. Residual -3.003 1.682 -.151 .958 63
Deleted Residual -4.40729 2.42450 -.22822 1.40173 63
Stud. Deleted Residual -3.211 1.707 -.160 .983 63
Mahal. Distance .752 18.907 4.893 3.623 71
Cook's Distance .000 .177 .016 .034 63
Centred Leverage Value .011 .270 .070 .052 71
a. Dependent Variable: RECAPTTPClog
All the coefficients are significant at the 0.05 level and all except LOG and FOSSILPC are
significant at the 0.01 level. The values of the coefficients indicate that:
• A 1% increase in the GDP per capita increases renewable electricity capacity
with TT per capita by 0.81%, keeping all other factors constant.
• A 1% increase in access to renewable energy resources per capita increases
RECAPTTpc by 0.28%, keeping all other factors constant.
• An increase by one point in the intellectual property rights protection index
decreases by 68% RECAPTTpc, keeping all other factors constant.
• An increase by 1 point in the logistic performance index increases RECAPTTpc
by 159%, keeping all other factors constant. The logistics performance index
can take values from 1 to 5. An increase by 1 point would be equivalent to
moving from the situation of low performer Sierra Leone (1.97) to that of
Vietnam (2.96), or from Vietnam to Japan (3.96).
208
• An increase by 1% of the production of fossil fuels per capita decreases the
RECAPTTpc by 0.13%.
The same model is also useful for explaining the renewable electricity capacity per
capita, without adjustments per technology transfer. Some differences between the
coefficients in both models would be:
• GDPpc has a slightly higher impact on explaining installed renewable energy
capacity involving TT.
• IPR has a slightly higher impact on explaining installed renewable energy
capacity involving TT.
• LOG has a slightly higher impact on explaining installed renewable energy
capacity, without adjustments per technology transfer
• The fossil fuel production per capita variable would lose significance on
explaining installed renewable capacity, without adjustments per technology
transfer.
b
Model Summary
a
1 .630 .396 .350 1.42463
a. Predictors: (Constant), FOSSILPClog, IPR, REACPClog, GDPpclog,
LOG
b. Dependent Variable: RECAPPClog
a
Coefficients
Standardised
1 (Constant) -5.639 1.904 -2.961 .004
GDPpclog .767 .190 .521 4.045 .000

REACPClog .279 .086 .360 3.236 .002
IPR -.649 .195 -.366 -3.323 .001
LOG 1.821 .639 .371 2.849 .006
FOSSILPClog -.121 .063 -.249 -1.929 .058
a. Dependent Variable: RECAPPClog
5.4.4 Explaining exports of renewable energy technologies relative to GDP

Only the competitive industrial performance variable is significant for explaining
exports of renewable energy technologies relative to GDP, among all those that were
significant in the per capita model. A simple regression model with CIP as an
independent variable can explain 35% of the variation of exports of renewable energy
technologies per GDP. The coefficient of CIP is significant at the 0.01 level and
indicates that an increase of 0.1 in the CIP index leads to an increase of 2.1% in exports
of renewable energy technologies relative to the GDP.
209
b
Model Summary
Std. Error of
Mode R Adjusted R the
l R Square Square Estimate
a
1 .593 .352 .340 2.4833459
a. Predictors: (Constant), CIP
b. Dependent Variable: REEXPGDPlog
a
Coefficients
Standardise
Unstandardised d
Coefficients Coefficients
1 (Constan 4.661 .837 5.567 .000
t)
CIP 21.091 3.894 .593 5.417 .000
a. Dependent Variable: REEXPGDPlog
Several iterations of the regression analysis could not find any other significant
variables to explain the value of exports of renewable energy technologies relative to
GDP. As with imports, this could be due to the high correlation between some enabling
factors for TT and the size of the economy, which leads to opposite effects – on one
hand the variables increase exports of renewable energy technologies, but on the
other hand they also increase the value of GDP, therefore decreasing the value of
exports relative to GDP and leading to an indefinite result.
5.4.5 Explaining imports of renewable energy technologies relative to GDP

Explaining imports of renewable energy technologies relative to GDP instead of per
capita delivers a less powerful model and counterintuitive results.
b
Model Summary
a
1 .382 .146 .108 1.13261
a. Predictors: (Constant), GDPpclog, TRADEOPlog, CREDlog
b. Dependent Variable: REIMPGDPlog
The variable income per capita has a negative value, as higher absolute levels of GDP
increase the levels of income per capita (keeping the population constant) but
decrease the relative value of imports of renewables to GDP. Access to credit for the
private sector is, as in the per capita value, a positive and significant explanatory
variable. Several iterations with the available explanatory variables fail to find any
other significant relationships, which could be due to the high positive correlation
between the GDP value and many explanatory variables and leads to two contradictory
effects: the increase of exports of renewable energy technologies on the one hand,
and the increase of the value of GDP, leading to indefinite results in the value of
exports relative to GDP, on the other.
210
a
Coefficients
Standardised
Unstandardised Coefficients
Model Coefficients t Sig.
B Std. Error Beta
1 (Constant) 12.691 .881 14.402 .000
CREDlog .619 .201 .394 3.084 .003

GDPpclog -.335 .127 -.336 -2.629 .011
a. Dependent Variable: REIMPGDPlog
5.4.6 Explaining renewable generation capacity with TT relative to GDP

The best fit model explains 21% of the variation in renewable electricity capacity
involving TT, relative to GDP, with three variables: logistic performance, intellectual
property rights protection and access to renewable energy sources relative to GDP.
The signs and values of the coefficients of these variables are consistent with those
obtained in the per capita model. The variables fossil fuel production and GDP per
capita, which proved significant for the per capita model, are not significant in
explaining exports relative to GDP because of the high positive correlation between
fossil fuel production and GDP per capita with the denominator of the ratio of exports
to GDP.
b
Model Summary
a
1 .461 .212 .177 1.28885
a. Predictors: (Constant), LOG, IPR, REACGDPlog
b. Dependent Variable: RECAPTTGDPlog
b
ANOVA
a
1 Regression 29.956 3 9.985 6.011 .001
Residual 111.296 67 1.661
Total 141.251 70
b. Dependent Variable: RECAPTTGDPlog
a
Coefficients
Standardised
1 (Constant) 1.285 1.412 .910 .366
REACGDPlog .483 .169 .355 2.852 .006
IPR -.572 .175 -.402 -3.261 .002
LOG 1.415 .547 .358 2.585 .012
a. Dependent Variable: RECAPTTGDPlog
The same variables can explain the total levels of renewable energy electricity capacity
per GDP, without adjustments per estimated amount of technology transfer. However,
211
the model has a lower explanatory value, even though the value of the three
coefficients increases.
b
Model Summary
a
1 .449 .202 .166 1.30978
b. Dependent Variable: RECAPGDPlog
a
Coefficients
Standardised
1 (Constant) .733 1.435 .511 .611
REACGDPlog .487 .172 .355 2.833 .006

IPR -.546 .178 -.380 -3.064 .003
LOG 1.618 .556 .405 2.909 .005
a. Dependent Variable: RECAPGDPlog
5.4.7 Summary of results

Table 32 summarises the factors that have been found to have an impact on imports
and exports of renewable energy technologies and the renewable electricity
generation capacity that claims to have used foreign technology. The table also
indicates the variables that have a high correlation with those deemed to have a
significant impact on the level of the selected proxies of clean energy technology
transfer.
TABLE 32- SUMMARY OF REGRESSION ANALYSIS
Dependent Independent Sign Correlated variables

Imports of Income per capita EDB (-), CPIlog, CREDlog, GDPlog, CO2PClog,
renewable energy (GDPpclog) ISO9PClog, TFPlog, CIPlog, ENROL3log, PATLOC,
+
technologies per PATFOR, TARIFFS (-), LOG, INCOMETAX (-), FOSSILPC,
PDIES (-)
capita
Access to credit for the EDB (-), CPIlog, IPR, CREDlog, CO2PClog, ISO9PClog,
private sector + TFPlog, CIPlog, ENROL3log, PATLOC, PATFOR,
(CREDlog) TARIFFS (-), LOG, FOSSILPC, PDIES (-)
Exports of Wind, Solar and Hydro FOSSILPClog
renewable energy resources per capita +
technologies per (WSHACCPClog)
capita Intellectual Property CPIlog, CREDlog, INVESTFREE, LOG
+
Rights protection (IPR)
Competitive Industrial CREDlog, GDPlog, GDPPClog, CO2PClog, HTEXPClog,
+ ISOPClog, PATLOClog, PATFORlog, LOG
Performance (CIP)
Size of the economy EDB (-), GDPPClog, CO2PClog, HTEXPPClog,
(GDPlog) + ISO9PClog, TFPlog, CIP, ENROL3log, PATLOClog,
PATFORlog, FDIOPlog (-), LOG, FOSSILPClog, PDIES (-)
Renewable energy Income per capita (See above)
+
generation (GDPpclog)
capacity requiring Renewable energy FOSSILPClog
foreign technology potential per capita +
transfer per capita (REACPClog)
212
Dependent Independent Sign Correlated variables
Production of fossil GDPlog, GDPPClog, CO2PClog, ISOPClog, TFPlog,
fuels (FOSSILPClog) - ENROL3log, PATLOCPClog, PATFORPClog, LOG,
REACPClog, PDIES (-)
Intellectual Property (See above)
-
Rights protection (IPR)
Logistics performance EDB (-), CPIlog, IPR, CREDlog, GDPlog, GDPPClog,
(LOG) + CO2PClog, ISO9PClog, HTEEXPPClog, TFPlog, CIPlog,
ENROL3log, PATLOC, PATFOR, FOSSILPC
Imports of Access to credit for the (See above)
renewable energy private sector
+
technologies per (CREDlog)
GDP
Exports of Competitive Industrial (See above)
renewable energy Performance (CIP)
+
technologies per
GDP
Renewable energy Renewable energy EDB, CPIlog (-), CREDlog (-), GDPlog (-), GDPPClog (-),
generation potential per GDP + CO2PClog (-), ISO9PClog (-), TFPlog (-), CIPlog (-),
capacity claiming (REACPCGDP) ENROL3log (-), PATFOR (-), LOG (-), PDIES (-)
foreign technology Intellectual Property (See above)
-
transfer per GDP Rights protection (IPR)
Logistic performance (See above)
+
(LOG)
Note: The sign of correlations is positive unless otherwise indicated
The results show that different independent variables influence different aspects of
low-carbon technology transfer. Those related to foreign input-related aspects are
influenced by income levels per capita, which show higher demand for foreign low-
carbon technologies as welfare levels increase, and credit availability for the private
sector, which indicates the need for available finance to allow private sector
investment. Aspects related to outputs of the transfer process to meet local demand
(RECAPTTpc) indicate that income per capita is also important, as well as the
availability of renewable energy resources, good infrastructure and a lack of
indigenous fossil fuels. Outputs of the transfer process related to the ability to produce
and export low-carbon technologies are explained also by the availability of renewable
energy resources, but in this case competitive industrial performance and the size of
the producing economy are key to ensuring the competitiveness of the goods
produced.
An interesting point about the results for output-related variables RECAPTTpc and
REEXPpc is the different sign of the IPR variable, depending on whether it is used to
explain exports of renewable energy technologies or installed renewable electricity
capacity. IPR protection is positive for countries that have reached the stage of
producing their own renewable energy technologies and sell them internationally.
However, it has a negative effect on those countries that need to gain access to foreign
renewable energy technologies to increase their electricity generation capacity.
It is also worth noting that no policy-related variables could be introduced into the
models, mainly because data are not available for most developing countries about the
levels of public R&D spending in clean energy technologies or other policies in place to
promote them. The only available variable, namely the existence of feed-in-tariffs or
guaranteed price schemes, did not prove to be significant in explaining exports,
213
imports or the capacity of clean energy generation technologies, because the dummy
variable does not reflect the variety of schemes implemented in different developing
countries. A certain part of the variation of the dependent variables that could not be
explained by our explanatory variables may be explained by specific policies available
in developing countries. A qualitative country-by-country analysis would therefore be
required to understand fully the differences in performance between the countries in
the sample.
5.5 Defining groups of developing countries per technology transfer policy

priority
5.5.1 Cluster analysis

Section 5.3 explored the performance of developing countries across a set of 14 highly
correlated variables. In all, 72% of the information provided by these variables could
be summarised by three new variables, the principal components. By mapping the
countries in the sample according to their performance in the three principal
components, it was possible to differentiate several groups and infer how their
performance might affect their ability to attract and absorb foreign technologies. The
problem with this first grouping exercise was that separate groups were not clearly
defined and the definition of clear policy priorities was difficult, due to the combined
impact of all correlated variables in each of the principal components. Besides, it still
was not clear what the impact of these variables was on different proxies for
technology transfer. In this section, we will define clusters according to the similarity
of developing country performances in the eight variables that have been found to
have a significant and important impact on three proxies of technology transfer
through the regression analysis in Section 5.4. Subsection 5.5.4.2 will also compare the
results obtained through PCA to those obtained through cluster analysis.
Cluster analysis is a technique used for grouping individuals into unknown groups. The
number and characteristics of the groups are derived from the data and are not usually
known prior to the analysis. It is a useful technique for classifying developing countries
according to their similar performance in the variables that create or deter enabling
frameworks for clean energy technology transfer. This classification, in combination
with the results obtained through PCA, can be the basis for defining differentiated
policy priorities, with a view to increasing clean energy technology transfer in groups
of developing countries.
A cluster is a collection of data points that are similar to one another within the same
cluster and dissimilar to objects in other clusters. Similarity among observations is
assessed through distance measures, the most commonly used of which is the
Euclidian distance, calculated as the square root of the square of the difference
between the two observations. The Euclidian distance can only be used when all the
variables are continuous, so other distance measures can be used for interval, counts
and binary variables when using a hierarchical cluster analysis procedure.
Several types of clustering methods can be classified as hierarchical and non-
hierarchical (Afifi, 2004; Han and Kamber, 2006; Stevens, 2009; Tabachnick and Fidell,
2001). Hierarchical methods can be agglomerative or divisive. Agglomerative methods
begin with N clusters (each observation constitutes its own cluster) and combine, step
214
by step, the closest clusters ending in a single cluster containing all the observations.
Divisive methods perform in the opposite way, starting with a single cluster and
splitting off the cases most dissimilar to the rest in successive steps, ending up with N
clusters. Agglomerative methods are more widely used than divisive methods. As long
as all the variables are of the same type, the hierarchical cluster analysis procedure can
analyse interval (continuous), count or binary variables. The results of hierarchical
methods are sometimes difficult to interpret in large samples like the one we are
considering, as they involve a large number of divisions and do not provide a
procedure to select the number of clusters that could be significant. Ward’s is an
agglomerative hierarchical method in which incremental clustering is defined to
minimise the variance within the clusters. Cluster membership is assessed by
calculating the total sum of squared deviations from the mean of a cluster. The
criterion for fusion is that it should produce the smallest possible increase in the error
sum of squares. Ward’s method has been previously used in the climate change field
to classify developing countries according to their attractiveness for CDM projects
(Jung, 2006).
The k-means method is a popular non-hierarchical clustering technique, belonging to
the partitioning methods that sort the cases into a series of iterations until they
converge into a stable partition of k clusters. This method requires previous knowledge
about the number of clusters we wish to obtain. Once the desired number of clusters
is determined, the k-means method will produce the exact number of clusters
demanded, and with the greatest possible distinction. The k-means method has been
previously used in the climate change field to analyse the similarities among a group of
global climate change policy proposals (Gainza et al., 2010) and also to classify
developing countries according to their attractiveness for the CDM (Jung, 2006).
Using different methods leads to different clustering results. This thesis will try both
Ward’s and k-means on the assessed data to test the robustness of the results. The
best cluster structure will be assessed according to previous results in PCA and
regression analysis, and to the actual performance of developing countries in the three
existing proxies for clean energy technology transfer.
As the data used in the analysis are measured on different scales, they are
standardised using z-scores. Clustering techniques are particularly sensitive to outliers,
so in order to identify them we run the k-means method on all the cases and classify as
outliers those observations that are placed in a single-case cluster. When we use the
clustering variables without logarithmic transformation, four countries are placed in
single-case clusters, and are therefore considered outliers. These are Qatar, Saudi
Arabia, United Arab Emirates and China. Outliers disappear when we use logarithms
for the variables GDP, GDPpc, REACpc and FOSSILpc.
5.5.2 Hierarchical cluster analysis: Ward’s method

We will use the Ward’s method with the squared Euclidean distance as the distance or
dissimilarity measure. Two iterations were carried out. The first iteration used all the
countries in the sample and some clustering variables transformed with logarithms.
The second iteration included the variables without logarithmic transformations. The
aim of this second iteration was to test the sensitivity of the cluster formation to the
size of the countries, whose effect was previously softened by taking logarithms in
215
some variables. Three outliers were identified – Qatar, Saudi Arabia and China –
because they were placed in a one-case cluster.
The second iteration without logarithmic transformation delivered a more unbalanced
cluster formation, with a large number of countries placed in a single cluster and a
smaller number of countries spread across clusters with a very small number of
countries. The cluster structure with variables without logs therefore was not able to
differentiate between the large number of countries that were placed in the same
cluster. For this reason, it was decided to select the cluster structures created with the
logarithmic transformation of variables. The clustering results with untransformed
variables are presented in Annex 6 of Chapter 5.
Cluster analysis is undertaken using the hierarchical Ward’s method for a sample of 51
countries and by taking as differentiating variables the eight variables found significant
as part of the regression analysis in Section 5.4.
The “agglomeration schedule” in Table 33 helps to decide the optimum number of
clusters with which we should work, and it provides a solution for every possible
number of clusters from 1 to 51 (the number of valid cases). Each stage shows a
further joining of clusters, with stage 0 having one cluster per valid case and Stage 1
joining the two cases with the lowest Euclidean distance, which in this case are
Ecuador (case 35) and Syria (case 106), followed by Brazil (case 15) and Turkey (case
113), then by Benin (case 11) and Uganda (case 115) and so on. The “coefficients”
column shows the Euclidean distance between the clusters or cases joined. The
distance is higher as we progress through stages. The latest stage joining together the
two final clusters shows an Euclidean distance of 302.27 whereby the previous stage,
with two clusters, has a distance of 207.122, the one before, with three clusters has a
distance of 170.832 and the next, with four clusters, has a distance of 148.953. As
shown in Figure 58 after four clusters there are less pronounced leaps in the distances
between clusters, which indicates that they are not so clearly differentiated. We will
take four clusters as the optimal number, taking this into account as well as the need
to keep the number of clusters within a manageable level.
TABLE 33- AGGLOMERATION SCHEDULE IN WARD’S CLUSTERING METHOD
Cluster Combined Stage Cluster First Appears

Stage Cluster 1 Cluster 2 Coefficients Cluster 1 Cluster 2 Next Stage
1 35 106 .202 0 0 5
2 15 113 .616 0 0 30
3 11 115 1.071 0 0 9
4 25 36 1.577 0 0 22
5 35 89 2.149 1 0 32
6 30 46 2.724 0 0 17
7 57 112 3.321 0 0 25
8 83 85 3.934 0 0 17
9 11 108 4.624 3 0 21
10 19 123 5.398 0 0 37
11 29 37 6.188 0 0 35
12 44 74 6.992 0 0 41
13 59 88 7.802 0 0 14
14 51 59 8.629 0 13 18
15 56 118 9.594 0 0 26
16 54 92 10.747 0 0 22
17 30 83 12.009 6 8 36
18 51 97 13.275 14 0 39
19 4 73 14.693 0 0 34
216
20 24 109 16.127 0 0 27
21 11 67 17.586 9 0 37
22 25 54 19.052 4 16 28
23 77 80 20.587 0 0 39
24 2 13 22.212 0 0 32
25 23 57 23.910 0 7 38
26 14 56 25.634 0 15 44
27 24 69 27.373 20 0 42
28 25 53 29.124 22 0 40
29 8 121 30.935 0 0 36
30 15 52 32.918 2 0 34
31 91 96 34.911 0 0 33
32 2 35 37.208 24 5 40
33 84 91 40.182 0 31 48
34 4 15 43.168 19 30 46
35 29 63 46.872 11 0 43
36 8 30 50.749 29 17 45
37 11 19 54.786 21 10 47
38 23 86 59.038 25 0 43
39 51 77 63.510 18 23 41
40 2 25 69.649 32 28 49
41 44 51 75.862 12 39 45
42 24 101 82.323 27 0 46
43 23 29 88.884 38 35 44
44 14 23 96.316 26 43 49
45 8 44 104.392 36 41 47
46 4 24 117.160 34 42 48
47 8 11 131.441 45 37 50
48 4 84 148.953 46 33 51
49 2 14 170.832 40 44 50
50 2 8 207.122 49 47 51
51 2 4 302.270 50 48 0
FIGURE 58- AGGLOMERATION COEFFICIENTS
Agglomeration coefficients
350
300
250
200
150
100
50
0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51
The clustering process is presented in the dendrogram below, which shows how the
closest cases are joined step by step until a single cluster is reached in the final stage.
The dendrogram shows two large differentiated groups and then subsequent
subdivisions.
217
218
Cluster members are presented in Table 34.
TABLE 34- CLUSTER STRUCTURE’S WARDS METHOD WITH LOGS
Cluster
Members Subdivisions
number
1 2:DZA 53:IDN The dendrogram shows the following subdivisions in
the cluster:
13:BOL 54:IRN
• Ecuador, Syria, Peru, Algeria and Bolivia
25:COL 89:PER
• Egypt, Indonesia, Iran, Russia and Colombia
35:ECU 92:RUS
36:EGY 106:SYR
2 4:ARG 84:OMN The dendrogram shows the following subdivisions in

the cluster:
15:BRA 91:QAT
• China, Thailand, Malaysia and South Africa
24:CHN 96:SAU
• Argentina, Mexico, Brazil, Turkey and India
52:IND 101:ZAF • Oman, Qatar and Saudi Arabia
69:MYS 109:THA
73:MEX 113:TUR
3 8:BGD 77:MOZ The following subdivisions can be noticed in the

dendrogram:
11:BEN 80:NPL
• Cameroon, Zambia, Benin, Uganda, Tanzania,
19:CMR 83:NGA
Madagascar
30:CIV 85:PAK • Vietnam, Pakistan, Bangladesh, Guatemala,
Côte dÍvoire and Nigeria
44:GEO 88:PRY
• Georgia and Moldova
46:GTM 97:SEN • Kenya, Paraguay, Honduras and Senegal
51:HND 108:TZA • Mozambique and Nepal
59:KEN 115:UGA
67:MDG 121:VNM
74:MDA 123:ZMB
4 14:BWA 57:JOR The dendrogram shows the following subdivisions in

the cluster:
23:CHL 63:LVA
• Jamaica, Uruguay and Botswana
29:CRI 86:PAN
• Costa Rica, El Salvador and Libya
37:SLV 112:TUN • Chile, Jordan and Lebanon
56:JAM 118:URY • Panama stays on its own
A one-way analysis of variance (ANOVA) can help to understand the variables that are
significant to differentiate between the groups. Table 35 explores the differences
between the means of the different clusters, offering F values and significance levels to
show whether any of these mean differences are significant. The between groups
means are all significant except in the case of the access to renewable energies
REACCPClog, which indicates access to renewable energies cannot reliably distinguish
between the four clusters, but each of the remaining seven variables can clearly
differentiate the groups.
219
TABLE 35- ANOVA
Sum of Mean
Squares df Square F Sig.
Zscore(GDPlog) Between Groups 19.027 3 6.342 23.175 .000
Within Groups 13.136 48 .274
Total 32.163 51
Zscore(GDPpclog) Between Groups 29.026 3 9.675 32.809 .000
Total 43.182 51
Zscore(CREDlog) Between Groups 10.141 3 3.380 5.865 .002
Total 37.807 51
Zscore(IPR) Between Groups 22.412 3 7.471 14.325 .000
Total 47.444 51
Zscore(CIP) Between Groups 22.545 3 7.515 11.343 .000
Total 54.348 51
Zscore(REACPClo Between Groups .745 3 .248 .500 .684
g) Within Groups 23.830 48 .496
Total 24.575 51
Zscore(FOSSILPCl Between Groups 28.451 3 9.484 21.937 .000
og) Within Groups 20.751 48 .432
Total 49.201 51
Zscore(LOG) Between Groups 21.714 3 7.238 21.170 .000

Total 38.125 51
The Tukey post-hoc test detailed in Table 74 of Annex 7 of Chapter 5 shows the
following similarities and dissimilarities between scores of the differentiating variables
across clusters, according to which:
• Clusters 1 and 2 are not dissimilar in terms of total GDP.
• Clusters 1 and 3 are not dissimilar in terms of access to credit for the private
sector, IPR protection, competitive industrial performance, fossil fuel
production per capita and logistical performance.
• Clusters 1 and 4 are not dissimilar in terms of GDP per capita, access to credit
for the private sector, IPR protection, competitive industrial performance, fossil
fuel production per capita and logistical performance.
• Clusters 2 and 4 are not dissimilar in terms of GDP per capita, access to credit
for the private sector and IPR protection.
• Clusters 3 and 4 are not dissimilar in terms of total GDP, competitive industrial
performance, fossil fuels production per capita and logistical performance.
220
• None of the clusters is dissimilar in terms of access to renewable energies per
capita.
Table 73 in the Annex shows the cluster descriptives, indicating the characteristics of
each cluster according to their differentiating variables. The means plots allow one to
characterise each of the clusters in terms of their performance for each of the
variables. The variable REACCpclog is not shown, as it is not considered significant for
differentiating between clusters.
FIGURE 59- HIERARCHICAL CLUSTERING MEANS PLOT
Taking into consideration the previous analysis, Table 36 shows the main
characteristics of each of the clusters.
TABLE 36-CLUSTER CHARACTERISATION
Clus
Members* Characteristics Outliers**
t
• Algeria does not reach the cluster
1 DZA IDN • Second highest GDP, not dissimilar to
mean’s lower bound for CRED, IPR,
BOL IRN Cluster 2
CIP and LOG. It exceeds the cluster’s
• Third highest GDP per capita, not
COL PER mean upper bound for FOSSILpc
dissimilar to Cluster 4
ECU RUS • Bolivia does not reach the cluster’s
• Second lowest CRED, not dissimilar to
mean lower bound for GDP, GDPpc,
EGY SYR Cluster 3
IPR, CIP and LOG
• Second lowest IPR, not dissimilar to
• Russia exceeds the cluster’s mean
cluster 3
upper bound in GDP, GDPpc, CRED,
• Lowest CIP not dissimilar to Cluster 3
IPR and FOSSILpc
• Second highest FOSSILpc, not dissimilar
to Cluster 2
• Second lowest LOG, not dissimilar to
Cluster 3
221
Clus
Members* Characteristics Outliers**
t
• Argentina does not reach the cluster
2 ARG OMN • Highest GDP not dissimilar to Cluster 1
mean´s lower bound for CRED and
BRA QAT • Highest GDPpc, not dissimilar to Cluster
IPR
4
CHN SAU • China exceeds the cluster mean´s
• Highest CRED not dissimilar to Cluster 4
IND ZAF upper bound for GDP, CRED, CIP and
• Highest IPR not dissimilar to Cluster 4
LOG
MYS THA • Highest CIP
• Malaysia exceeds the cluster mean´s
MEX TUR • Highest FOSSILpc not dissimilar to
upper bound for CRED, CIP and LOG
Cluster 1
• South Africa exceeds the cluster
• Highest LOG
mean´s upper bound for CREDlog, IPR
and LOG
• Pakistan exceeds the upper bound of

3 BGD MOZ • Lowest GDP not dissimilar to Cluster 4
the cluster´s mean for GDPlog, CIP
BEN NPL • Lowest GDPpc
and FOSSILpclog
• Lowest CRED, not dissimilar to Cluster 1
CMR NGA • Vietnam exceeds the upper bound of
• Lowest IPR, not dissimilar to Cluster 1
CIV PAK the cluster´s mean for GDPlog,
• Lowest CIP, not dissimilar to Clusters 1
CREDlog, CIP, FOSSILpclog and LOG
GEO PRY and 4
• Nigeria exceeds the upper bound of
GTM SEN • Second lowest FOSSILpc, not dissimilar
the cluster´s mean for GDPlog,
to Cluster 4
HND TZA CREDlog and FOSSILpclog
• Lowest LOG, not dissimilar to clusters 1
• Guatemala exceeds the upper bound
KEN UGA and 4
of the cluster´s mean for GDPpclog
MDG VNM and CIP
MDA ZMB • Honduras exceeds the upper bound
of the cluster´s mean for GDPpc,
CREDlog and LOG
• Chile’s performance exceeds the

4 BWA JOR • Second lowest GDP, not dissimilar to
cluster’s mean upper bound for all
CHL LVA Cluster 3
variables except CIP, however it
• Second highest GDPpc, not dissimilar to
CRI PAN could not be part of Cluster 2
clusters 1 and 2
SLV TUN because size related variables (GDP,
• Second highest CRED, not dissimilar to
Fossil pc) as well as CIP are much
JAM URY Cluster 2
lower
• Second highest IPR, not dissimilar to
• El Salvador does not reach the
Cluster 2
cluster’s mean lower bound for
• Second highest CIP, but not dissimilar to
GDPpc, IPR and FOSSILpc, but
Clusters 1 and 3
exceeds the upper bound for CIP
• Lowest FOSSILpc, not dissimilar to
• Botswana does not reach the means’
Cluster 3
lower bound for GDP, CRED and LOG,
• Second highest LOG, but not dissimilar
but exceeds the upper bound for
to clusters 1 and 3
FOSSILpc
• Jamaica does not reach the mean’s
lower bound for GDP, CRED,
FOSSILpc and LOG, but exceeds the
means upper bound for IPR
222
*Note: Outliers are considered as those countries showing a consistent under- or over-performance as
compared to their peers.
We tested whether or not eliminating REACCpc from the clustering variables would
create a different clustering structure, as REACCpc was found to be not significant in
differentiating between clusters. The resulting cluster structure was the same when
this variable was not included.
5.5.3 Non-hierarchical cluster analysis: k-means

As with the hierarchical clustering method, two iterations were carried out, one with
differentiated variables transformed with logarithms and another with untransformed
variables. The first iteration delivered a better clustering structure and was the one
chosen. The results of the second iteration are included as part of Annex 6 of Chapter
5.
The k-means clustering method requires the previous selection of the number of
clusters. We select four clusters, as with the hierarchical method, and exclude cases
list-wise, in order to allow for comparison with the results in the hierarchical cluster
analysis. In this case, the four clusters delivered have 4, 10, 18 and 20 members.
TABLE 37- NUMBER OF CASES IN EACH CLUSTER
Cluster 1 4.000
2 10.000
3 18.000
4 20.000
Valid 52.000
Missing 125.000
The ANOVA table shows that all variables are significant in differentiating between
clusters, including REACCpc, which was deemed as not significant in the hierarchical
analysis.
TABLE 38- ANOVA
Cluster Error
Mean Square df Mean Square df F Sig.
Zscore(GDPlog) 6.100 3 .289 48 21.117 .000

Zscore(GDPpclog) 8.837 3 .347 48 25.446 .000
Zscore(CREDlog) 4.766 3 .490 48 9.730 .000
Zscore(IPR) 7.113 3 .544 48 13.080 .000
Zscore(CIP) 9.144 3 .561 48 16.308 .000
Zscore(REACPClog) 3.130 3 .316 48 9.893 .000
Zscore(FOSSILPClog) 6.475 3 .620 48 10.437 .000
Zscore(LOG) 7.583 3 .320 48 23.674 .000
The cluster descriptives that allow the characterisation of each cluster according to its
performance in each variable are presented in Table 75 of the Annex. The means plot
shows graphically the differentiating characteristics of each cluster.
223
FIGURE 60- NON-HIERARCHICAL CLUSTERING MEANS PLOT
Table 39 shows the cluster formation, indicating the characteristics of each cluster
according to the means comparison with the ANOVA.
TABLE 39-CLUSTER FORMATION WITH K-MEANS
Clus
Members Characteristics Outliers
ter
1 ARG RUS • Second highest GDP, not
OMN SAU
• Highest GDPpc, not dissimilar to
Cluster 2
• Medium CRED, not dissimilar to
any of the other clusters
• High IPR, not dissimilar to Cluster
2
• Low CIP, not dissimilar to
Clusters 3 and 4
• The highest REACCpc
• The highest FOSSILpc, but not
• Second highest LOG, not
dissimilar to Clusters 2 and 4
• Chile is a significantly lower performer
2 BRA MEX • The highest GDP, but not
in GDPlog, CIP, FOSSILpclog and LOG. It
CHL QAT outperforms the group in REACCpclog
• The second highest GDPpc, but
• Brazil outperforms the cluster in
CHN ZAF not dissimilar to Cluster 1
GDPlog and REACCpclog
IND THA • The highest CRED, but not
• China outperforms the group in
MYS TUR GDPlog, CREDlog, CIP and LOG
• The highest IPR, but not
224
Clus
Members Characteristics Outliers
ter
dissimilar to Cluster 1 • Malaysia outperforms the group in
• The highest CIP CREDlog, CIP and LOG
• Low REACCpc, not dissimilar to • Mexico is a lower performer in
Clusters 3 and 4 CREDlog, IPR and LOG
• Second highest FOSSILpc, not • South Africa outperforms the group in
dissimilar to Cluster 1 CREDlog, IPR and LOG
• Highest LOG, not dissimilar to • Turkey is a lower performer in
Cluster 1 CREDlog and FOSSILpclog
• Madagascar is an underperformer in
3 DZA MDA • The lowest GDP
GDPlog, GDPpclog, CREDlog, and
• The lowest GDPpc
BGD MOZ FOSSILpclog.
• The lowest CRED, not dissimilar
• The rest of the countries do not show
BEN NPL to Cluster 1
a clear over or under performing
BOL PAK • The lowest IPR
pattern as compared to their peers
• The lowest CIP, not dissimilar to
CMR PRY Clusters 1 and 4
CIV SEN • Low REACCpc, not dissimilar to
Clusters 2 and 4
GEO TZA • The lowest FOSSILpc, not
KEN UGA dissimilar to Cluster 4
• The lowest LOG
MDG ZMB
• Colombia is an over performer in
4 BWA JAM • Second lowest GDP
GDPlog, GDPpclog, IPR and
• Second lowest GDPpc, but not
COL JOR FOSSILpclog
• Costa Rica is an over performer in
CRI LVA • Medium CRED, not dissimilar to
GDPpclog and CIP, but
ECU NGA Cluster 1
underperforms in FOSSILpc
• Second lowest IPR, not dissimilar
EGY PAN • Jordan outperforms the cluster in
to cluster 1
CREDlog, IPR and CIP, but
SLV PER • Low CIP, not dissimilar to
underperforms in GDPlog
Clusters 1 and 3
GTM SYR • Tunisia outperforms the group in IPR
• The lowest REACCpc, not
and CIP
HND TUN dissimilar to Clusters 2 and 3
• El Salvador is an underperformer in
• Second lowest FOSSILpc, not
IDN URY GDPlog, CREDlog, REACCpclog and
dissimilar to Clusters 2 and 3
FOSSILpclog, but outperforms in CIP
IRN VNM • Second lowest LOG, not
• Honduras underperforms in GDPlog,
GDPpclog, CIP and FOSSILpclog, but
over performs in CREDlog
• Jamaica underperforms in GDPlog,
LOG, REACCpclog and FOSSILpclog
but over performs in IPR
• Syria underperforms in GDPpc,
CREDlog and CIP but outperforms in
FOSSILpclog
The post-hoc Tukey analysis, the results of which are presented in Table 76, is useful
for analysing similarities and dissimilarities between clusters for each variable. From
the post-hoc Tukey test we can conclude the following:
• In terms of GDPlog:
o Clusters 1 and 2 are not dissimilar
225
• In terms of GDPpc:
• In terms of CREDlog
o Cluster 1 is not dissimilar to clusters 2, 3 and 4
• For IPR
o Cluster 1 is not dissimilar to clusters 2 and 4
• For CIP
o Cluster 1 is not dissimilar to clusters 3 and 4
• For REACpclog
o Clusters 2, 3 and 4 are not dissimilar
• For FOSSILpclog
• For LOG
5.5.4 Selecting the most appropriate cluster formation
5.5.4.1 Stability of cluster formation
The cluster structures obtained with hierarchical and non-hierarchical methods show
that some countries tend to remain together in the same clusters, while others show a
higher mobility between clusters.
Both methods show a differentiated group of worse performers (Cluster 3 in both
cases), with 20 members in hierarchical (Ward’s) clustering and 18 members in non-
hierarchical (k-means) clustering. In the Ward’s hierarchical method, Cluster 2, with 12
members, represents the best performers in all variables. In the k-means method,
there are two groups of best performers with different characteristics. K-means Cluster
2, with 10 members, shows a better performance in GDPlog, CREDlog, IPR, CIP and
LOG. K-means Cluster 1, with only 4 members, performs better in GDPpclog,
FOSSILpclog and REACCpclog. Ward’s Cluster 1, with 10 members, includes relatively
large economies with high production of fossil fuels per capita, but bad performance in
all the other enabling factors. Finally, Ward’s Cluster 4, with 10 members, includes
relatively small economies with good performance (but lower than Cluster 2) in all
enabling factors for technology transfer and low production of fossil fuels per capita. K-
means Cluster 4, with 20 members, includes relatively small economies, with low
access to renewable energy and fossil fuel sources per capita, but better performance
than Cluster 3 in all enabling factors and better than Cluster 1 in CREDlog and CIP.
Figure 61 and Figure 4 show graphically the characteristics of Ward’s and k-means
cluster formations.
226
FIGURE 61 HIERARCHICAL (WARD’S) CLUSTERING MEANS PLOT FOR FIRST ITERATION WITH LOGS
FIGURE 62- NON-HIERARCHICAL (K-MEANS) CLUSTERING MEANS PLOT FOR FIRST ITERATION WITH
LOGS
Table 40 shows the members of Cluster 3 of worst performers according to both

clustering methods. Sixteen members are stable in their cluster positioning –
Guatemala, Honduras, Nigeria and Vietnam are positioned in Cluster 3 by the Ward’s
method, but belong to Cluster 4 in the k-means method. All these countries had been
identified as overachievers compared to their peers in Cluster 3 of the Ward’s’ method
for a number of variables. Therefore, it makes sense that they are placed in Cluster 4
by the k-means method, as it contains a group of countries that show better
performance than Cluster 3 in most variables but are still the second lowest
performers. Algeria and Bolivia are placed in Cluster 3 as worst performers by the k-
227
means method but belong to Cluster 1 in the Ward’s method, where Cluster 1 includes
relatively large economies with large fossil fuel production per capita but bad
performance in the remaining enabling factors for technology transfer.
TABLE 40-MEMBERS OF CLUSTER 3 IN THE FIRST ITERATIONS OF WARD’S AND K-MEANS CLUSTERING
METHODS
Members of Cluster 3 in both Members of Cluster 3 in Ward’s Members of Cluster 3 in k-
Ward’s and k-means methods method and Cluster 4 in k- means method and Cluster 1 in
means Ward’s
Bangladesh Guatemala Algeria
Benin Honduras Bolivia
Cameroon Nigeria
Cote d'Ivoire Vietnam
Georgia
Kenya
Madagascar
Moldova
Mozambique
Nepal
Pakistan
Paraguay
Senegal
Tanzania
Uganda
Zambia
Table 41 shows the members of Cluster 2, which represents the best performers,
according to both clustering methods. Nine countries remain stable in this position in
both clustering methods. Argentina, Oman and Saudi Arabia are placed in Cluster 2
according to the hierarchical Ward’s method, but belong to Cluster 1 according to the
k-means method. K-means cluster 1 includes relatively large and wealthy economies
with high production of fossil fuels per capita, but bad performance in some key
enabling factors. This is the case for Argentina, Oman and Saudi Arabia, which show
low levels of CIP as compared to other Cluster 2 members, and in the case of Argentina
also low access to credit for the private sector and low IPR protection. Chile is placed
among the group of best performers by the k-means method but belongs to Cluster 4
in the Ward’s method. Chile is a limit case and is considered an underperformer
compared to its peers in the k-means Cluster 2, with its smaller economy, low CIP, low
FOSSILpc and LOG. However, it is an overachiever among Ward’s Cluster 4, with
performance above the upper bound of the cluster´s mean for all variables except CIP.
METHODS
Members of Cluster 2 in both Members of Cluster 2 in Ward’s Members of Cluster 2 in k-
Ward’s and k-means methods method and Cluster 1 in k- means method and Cluster 4 in
means Ward’s
Brazil Argentina Chile
India Oman
Mexico Saudi Arabia
Turkey
228
Malaysia
South Africa
Thailand
China
Qatar
Cluster 1 shows high instability, as shown in Table 42. This cluster represents relatively
large countries with high production of fossil fuels per capita but low performance in
many enabling factors. Differences in how Cluster 1 is defined by Ward’s and k-means
methods contribute to the instability of the cluster formation. K-means Cluster 1 has
higher GDP per capita and IPR values compared to Ward’s Cluster 1, which may
contribute to differences in cluster formation by both methods, with k-means tending
to attract better performers for this group. Only Russia remains in Cluster 1 in both
hierarchical and non-hierarchical clustering methods, but it is an overachiever among
its peers in Ward’s Cluster 1. Algeria and Bolivia belong to Ward’s Cluster 1 but are
placed in k-means Cluster 3 among the lowest performers. Ecuador, Egypt, Iran, Peru,
Syria, Colombia and Indonesia belong to Ward’s Cluster 1 but are placed in k-means
Cluster 4 among second lowest performers. Finally, Argentina, Oman and Saudi Arabia,
which belong to Ward’s Cluster 2 of the best performers, are placed in k-means Cluster
1.
METHODS
Members of Cluster 1 Members of Cluster 1 Members of Cluster 1 Members of Cluster 1 in
in both Ward’s and k- in Ward’s method and in Ward’s method and k-means method and
means methods Cluster 3 in k-means Cluster 4 in k-means Cluster 2 in Ward’s
method
Russia Algeria Ecuador Argentina
Bolivia Egypt Oman
Iran Saudi Arabia
Peru
Syria
Colombia
Indonesia
Finally, Cluster 4 also shows instability, due to the different definition of the cluster by
both methods. Ward’s Cluster 4 includes higher achievers, with a considerably higher
average for GDPpclog, CREDlog, IPR, CIP and LOG. Ward’s Cluster 4 includes Chile,
which is placed among the highest performers in k-means Cluster 2. K-means Cluster 4
includes lower performing countries from Ward’s Cluster 3 such as Guatemala,
Honduras, Nigeria and Vietnam, but as previously highlighted these countries
outperform their Ward’s Cluster 3 peers in many variables. Finally, k-means Cluster 4
also includes countries from Ward’s Cluster 1 – Ecuador, Egypt, Iran, Peru, Syria,
Colombia and Indonesia – therefore raising the cluster´s average GDPlog and
FOSSILpclog compared to Ward’s Cluster 4.
229
METHODS
Members of Cluster 4 Members of Cluster 4 Members of Cluster 4 Members of Cluster 1 in
in both Ward’s and k- in Ward’s method and in k-means method Ward’s method and
means methods Cluster 2 in k-means and Cluster 3 in Cluster 4 in k-means
Ward’s
Botswana Chile Guatemala Ecuador
El Salvador Honduras Egypt
Jamaica Nigeria Iran
Uruguay Vietnam Peru
Costa Rica Syria
Jordan Colombia
Lebanon Indonesia
Panama
Tunisia
5.5.4.2 Comparison with results of principal components analysis
The principal components analysis in Section 3 indicated country performance in a set

of some highly correlated enabling factors for clean energy technology transfer. We
will compare the results achieved by PCA with the different cluster structures
obtained, in order to inform our decision about the best cluster structure. We will
analyse in particular how cluster structures relate to country performance in the first
and third principal components, as countries with high performance in both were
deemed the most likely to attract and absorb foreign technologies.
Firstly, the clustering results obtained with the k-means method, by taking logs for
some variables, show a good match with the results of PCA. Countries in Cluster 2 of
best performers are placed in the first quadrant (top-right), with the exception of
India, in the fourth quadrant (top left), but close to the divisive line and Qatar, in the
second quadrant (bottom-right). Countries in the first cluster are mostly placed in the
second quadrant, with the exception of Argentina in the fourth, which indicates that
factors such as low fossil fuel prices, high fossil fuels production or high corruption may
be preventing a higher uptake of clean energy technology transfer. Clusters 3 and 4 are
placed in the third and fourth quadrants, with countries in Cluster 4 performing better
in the first principal component, indicating better conditions overall for receiving and
absorbing clean energy technology transfer.
230
Clustering results using the Ward’s method show Cluster 2 members (best performers)
distributed across quadrants 1, 2 and 3, with most of them in the first quadrant. Most
members of the first cluster are placed in the third quadrant, except Colombia in the
fourth and Russia in the second. As shown in the cluster analysis, Russia outperformed
its peers in many variables, which is shown by its highest rating in the first principal
component. Cluster 3 members have the lowest ratings in the first principal
component and are therefore placed to the left of quadrants 3 and 4. Finally, Cluster 4
members are mostly placed to the right of quadrant 4, with the high performer Chile in
the first quadrant due to its higher rating in both principal components 1 and 3.
231
5.5.4.3 Clusters formation relevance in view of technology transfer data
In this subsection we analyse the relevance of the cluster structures when taking into
account the actual performance of their members in the proxy variables for clean
energy technology transfer, expressed in per capita values.
Firstly, a means plot of the technology transfer variables for Ward’s clusters shows the
following:
• Cluster 1, made up of large economies with poor performance in some enabling
factors, shows low levels of imports and exports of clean energy technologies,
which are not dissimilar to Cluster 3 of worst performers. It shows average
levels of renewable energy installed capacity, not dissimilar from any of the
other clusters. Russia outperforms other cluster members in terms of imports
and exports of clean technologies per capita, but with levels still far from those
of Cluster 2.
• Cluster 2, made up of high performers, shows the highest levels of imports and
exports of clean energy technologies per capita, and average levels of installed
renewable energy capacity requiring technology transfer, which are not
dissimilar to the other clusters. In Cluster 2, Mexico, China and particularly
Malaysia show much higher than average exports of clean energy technologies
per capita. Mexico, Turkey and Malaysia show considerably higher levels of
imports. Argentina and Brazil show higher than average levels of installed
renewable energy capacity per capita, both adjusted or not for technology
transfer. Conversely, Turkey and Oman are significant low performers in terms
of exports of technologies per capita. Argentina and India underperform in
232
imports of clean technologies per capita, and India, Thailand, South Africa and
China in installed renewable energy capacity per capita estimated to involve
technology transfer.
• Cluster 3, made up of low performers, shows the lowest levels for all variables,
but with levels of installed renewable capacity not dissimilar to the other
clusters and exports of renewable energy technologies not dissimilar to
Clusters 1 and 4. Vietnam clearly outperforms other cluster members in terms
of imports and exports per capita of clean technologies, while Paraguay and
Georgia are overachievers in their installed renewable energy capacity
estimated to involve technology transfer.
• Cluster 4, consisting of relatively small economies with good performance in
some enabling factors, shows low export levels, not dissimilar to Clusters 1 and
3, but a higher level of imports of clean energy technologies. Botswana, Chile
and Panama show significantly higher levels of imports per capita of clean
technologies than their peers, while Uruguay has a significantly higher
performance in installed renewable capacity per capita.
FIGURE 63- MEANS PLOT OF TT VARIABLES FOR WARD’S CLUSTERS WITH LOGS
6000
5000
4000
Cluster 1
3000 Cluster 2
Cluster 3
2000 Cluster 4
1000
0
RECAPpc RECAPTTpc REIMPpc REEXPpc
A plot of the median for each TT variable is also included, as given the high
heterogeneity among cluster members it can provide less skewed information. The
medians plot shows a similar pattern to the means plot, but with less clear differences
among Clusters 1 and 4, and with lower values than the means values.
233
FIGURE 64- MEDIANS PLOT OF TT VARIABLES FOR WARD’S CLUSTERS WITH LOGS
3000.0
2500.0
2000.0
Cluster 1
1500.0 Cluster 2
Cluster 3
1000.0 Cluster 4
500.0
0.0
234
TABLE 44-VALUE OF TECHNOLOGY TRANSFER VARIABLES FOR MEMBERS OF WARD’S CLUSTERS WITH
LOGS
Country Code RECAPpc RECAPTTpcREIMPpc REEXPpc WARDS 1
Algeria DZA 8.0 7.4 1334.3 1
Bolivia BOL 45.0 45.0 408.6 1
Ecuador ECU 142.8 140.5 601.6 2.1 1
Egypt, Arab Rep.
EGY 39.5 39.5 718.7 27.0 1
Iran, Islamic Rep.
IRN 106.0 106.0 1
Peru PER 112.7 74.5 574.9 10.7 1
Syrian Arab Republic
SYR 75.0 69.0 1
Colombia COL 195.9 71.5 815.1 8.2 1
Indonesia IDN 24.4 19.3 517.8 122.6 1
Russian Federation
RUS 329.2 302.9 744.3 295.1 1
Average Cluster 1 107.9 87.6 714.4 77.6
Argentina ARG 225.8 224.4 213.8 728.3 2
Brazil BRA 443.4 267.7 1831.8 596.9 2
India IND 42.5 5.7 364.7 668.3 2
Mexico MEX 115.6 115.6 6864.5 7559.0 2
Turkey TUR 199.8 183.8 9690.5 347.8 2
Oman OMN 1367.1 36.3 2
Malaysia MYS 75.9 75.5 13664.9 30112.3 2
Thailand THA 50.8 47.0 2657.7 2825.8 2
South Africa ZAF 13.6 12.5 2496.6 2781.6 2
Saudi Arabia SAU 2
China CHN 139.9 22.7 3295.7 8489.5 2
Qatar QAT 6049.3 2

Average Cluster 2 145.2 106.1 4244.7 5472.3
Bangladesh BGD 1.6 1.6 3
Côte d'Ivoire CIV 31.3 31.3 260.6 0.8 3
Pakistan PAK 37.9 34.4 65.3 1.0 3
Benin BEN 0.1 0.1 3
Cameroon CMR 42.0 38.6 3.2 3
Georgia GEO 651.6 651.6 3
Kenya KEN 21.8 21.8 308.0 48.4 3
Moldova MDA 17.8 16.3 156.9 10.8 3
Madagascar MDG 6.2 6.2 208.7 2.0 3
Mozambique MOZ 95.3 87.7 66.3 23.2 3
Nepal NPL 22.4 20.6 333.2 13.4 3
Paraguay PRY 1282.0 1179.4 90.9 250.7 3
Senegal SEN 0.1 0.1 196.4 1.9 3
Tanzania TZA 13.3 13.3 221.2 6.5 3
Uganda UGA 9.7 9.0 1363.8 2.7 3
Zambia ZMB 131.4 120.9 127.3 3.1 3
Guatemala GTM 58.7 55.4 588.9 90.6 3
Nigeria NGA 12.5 12.5 132.1 3.9 3
Honduras HND 70.1 70.1 183.9 0.0 3
Vietnam VNM 63.3 62.3 3077.4 1376.8 3
Average Cluster 3 128.5 121.7 461.3 108.2
Botsw ana BWA 4495.6 13.7 4
Uruguay URY 461.2 461.2 412.3 0.1 4
Jamaica JAM 15.7 15.7 942.8 110.7 4
El Salvador SLV 126.6 106.6 163.9 5.8 4
Chile CHL 292.7 285.4 9740.0 17.9 4
Costa Rica CRI 384.4 374.2 1141.1 38.6 4
Jordan JOR 2.2 2.2 823.2 58.5 4
Lebanon LVA 66.7 66.7 1039.2 113.7 4
Panama PAN 251.0 251.0 4616.2 50.0 4
Tunisia TUN 7.9 7.9 810.8 29.1 4
Average Cluster 4 178.7 174.6 2418.5 43.8
235
A similar analysis for the clustering structure obtained with the k-means method
highlights the following:
• Cluster 1 (not considering outlier Saudi Arabia) shows low levels of imports and
exports of clean technologies per capita, not dissimilar to Clusters 3 and 4, as
well as low levels of REIMPpc and RECAPTTpc. Oman outperforms other
members in terms of REIMPpc, but is far behind in REEXPpc.
• Cluster 2 shows the highest levels of imports and exports of clean energy
technologies per capita, but levels of renewable energy capacity not dissimilar
to the other clusters. Among them, Malaysia shows much higher levels than
average of REEXPpc, while Chile is far behind other cluster members for the
same concept. Malaysia also outperforms in terms of REIMPpc, with India
showing the lowest performance for this concept. Chile shows the highest level
of RECAPTTpc, followed by Brazil, with India and South Africa showing the
lowest values for this concept.
• Cluster 3 has the lowest average levels for all variables, with RECAPTTPClog and
REEXPPClog not dissimilar to those of Clusters 1 and 4. As in the Ward’s
method, Georgia and Paraguay over-perform against other cluster members in
RECAPTTpc, while Algeria and Uganda over-perform in REIMPpc.
• Cluster 4 has average values for RECAPTTPClog that are not dissimilar to the
other clusters. REIMPPClog is higher than Clusters 1 and 3, but REEXPpc is
lower than Clusters 1 and 2. Among the members of Cluster 4, Vietnam is a
high performer in REIMPpc and REEXPpc. Botswana and Panama also show
significantly higher than average levels of REIMPpc, while Uruguay shows high
levels of RECAPTTpc.
FIGURE 65- MEANS PLOT OF TT VARIABLES FOR K-MEANS CLUSTERS WITH LOGS
7000
6000
5000
Cluster 1
4000
Cluster 2
3000 Cluster 3
2000 Cluster 4
1000
0
The medians plot for TT variables is also presented, as it can provide more informative
data when the means are skewed by outliers. The pattern is quite similar to that of the
236
means plot, but in this case Clusters 4 and 1 perform similarly in terms of imports of
renewable energy technologies per capita (REIMPpc).
FIGURE 66- MEDIANS PLOT OF TT VARIABLES FOR K-MEANS CLUSTERS WITH LOGS
3500.0
3000.0
2500.0
Cluster 1
2000.0
Cluster 2
1500.0 Cluster 3
1000.0 Cluster 4
500.0
0.0
237
TABLE 45-VALUE OF TT VARIABLES FOR MEMBERS OF K-MEANS CLUSTERS WITH LOGS
Country Code RECAPpc RECAPTTpcREIMPpc REEXPpc KMEANS 1
Russian Federation
RUS 329.2 302.9 744.3 295.1 1
Argentina ARG 225.8 224.4 213.8 728.3 1
Oman OMN 1367.1 36.3 1
Saudi Arabia SAU 1
Average Average 277.5 263.6 775.1 353.2
Brazil BRA 443.4 267.7 1831.8 596.9 2
India IND 42.5 5.7 364.7 668.3 2
Mexico MEX 115.6 115.6 6864.5 7559.0 2
Turkey TUR 199.8 183.8 9690.5 347.8 2
Malaysia MYS 75.9 75.5 13664.9 30112.3 2
Thailand THA 50.8 47.0 2657.7 2825.8 2
South Africa ZAF 13.6 12.5 2496.6 2781.6 2
China CHN 139.9 22.7 3295.7 8489.5 2
Qatar QAT 6049.3 2
Chile CHL 292.7 285.4 9740.0 17.9 2
Average 152.7 112.9 5622.9 5944.8
Algeria DZA 8.0 7.4 1334.3 0.0 3
Bolivia BOL 45.0 45.0 408.6 0.0 3
Bangladesh BGD 1.6 1.6 3
Côte d'Ivoire CIV 31.3 31.3 260.6 0.8 3
Pakistan PAK 37.9 34.4 65.3 1.0 3
Benin BEN 0.1 0.1 3
Cameroon CMR 42.0 38.6 3.2 3
Georgia GEO 651.6 651.6 3
Kenya KEN 21.8 21.8 308.0 48.4 3
Moldova MDA 17.8 16.3 156.9 10.8 3
Madagascar MDG 6.2 6.2 208.7 2.0 3
Mozambique MOZ 95.3 87.7 66.3 23.2 3
Nepal NPL 22.4 20.6 333.2 13.4 3
Paraguay PRY 1282.0 1179.4 90.9 250.7 3
Senegal SEN 0.1 0.1 196.4 1.9 3
Tanzania TZA 13.3 13.3 221.2 6.5 3
Uganda UGA 9.7 9.0 1363.8 2.7 3
Zambia ZMB 131.4 120.9 127.3 3.1 3
Average Average 134.3 127.0 367.2 24.5
Ecuador ECU 142.8 140.5 601.6 2.1 4
Egypt, Arab Rep.
EGY 39.5 39.5 718.7 27.0 4
Iran, Islamic Rep.
IRN 106.0 106.0 4
Peru PER 112.7 74.5 574.9 10.7 4
Syrian Arab Republic
SYR 75.0 69.0 4
Colombia COL 195.9 71.5 815.1 8.2 4
Indonesia IDN 24.4 19.3 517.8 122.6 4
Guatemala GTM 58.7 55.4 588.9 90.6 4
Nigeria NGA 12.5 12.5 132.1 3.9 4
Honduras HND 70.1 70.1 183.9 0.0 4
Vietnam VNM 63.3 62.3 3077.4 1376.8 4
Botsw ana BWA 4495.6 13.7 4
Uruguay URY 461.2 461.2 412.3 0.1 4
Jamaica JAM 15.7 15.7 942.8 110.7 4
El Salvador SLV 126.6 106.6 163.9 5.8 4
Costa Rica CRI 384.4 374.2 1141.1 38.6 4
Jordan JOR 2.2 2.2 823.2 58.5 4
Lebanon LVA 66.7 66.7 1039.2 113.7 4
Panama PAN 251.0 251.0 4616.2 50.0 4
Tunisia TUN 7.9 7.9 810.8 29.1 4
Average Average 116.7 105.6 1203.1 114.6
238
5.5.4.4 Cluster selection
Cluster structures, with some variables transformed with logs, are preferred to those
with untransformed differentiating variables because they deliver more balanced
structures in terms of the number of countries per cluster. Cluster structures with
untransformed variables tend to concentrate a small number of overachievers in a
cluster and a large number of lower performers in another, with blurred clusters in the
middle.
Among cluster structures with differentiating variables transformed with logs, both
provide valid classifications of countries in four clusters, with two well-defined groups
of high and low performers. However, the two methods derive different compositions
for the two clusters placed between the high and low performers, which lend
themselves to different policy recommendations.
Resulting from the cluster and principal components analysis, as well as the actual data
for the technology transfer proxies, four main groups of host countries for clean
energy technology transfer can be identified: clean technology developers are
countries which are capable of succeeding in the three elements of technology
transfer, clean technology implementers are good performers in most enabling factors
but lack a large internal demand and industrial competitiveness, aid recipients are
countries needing foreign aid to create the building blocks for successful TT and
countries requiring structural changes to improve the business environment and
create clear demand signals favouring clean energy technologies over widely locally
available fossil fuels.
The cluster formation was derived through an exploratory analysis, combining the
results of Ward’s and k-means clustering methods as well as PCA. Although it includes
a certain level of judgement by the researcher, the quantitative basis for the cluster
selection is nevertheless provided by the results of the different multivariate analysis
techniques used in this section, as well as the observation of actual technology transfer
data.
These four differentiated groups are presented in Figure 68 which also shows those
countries that are not clearly attributable to one cluster, but are placed between two
of them.
239
FIGURE 67- CLUSTER SELECTION FOR TECHNOLOGY TRANSFER RECIPIENT COUNTRIES
TECHNOLOGY TECHNOLOGY STRUCTURAL AID RECIPIENTS

DEVELOPERS IMPLEMENTERS CHANGES
Bangladesh (L)
Bolivia (LM)
Brazil (UM) Botswana (UM) Algeria (UM) Benin (L)
China (LM) El Salvador (LM) Russia (UM) Cameroon (LM)
India (LM) Jamaica (UM) Oman (U) Côte dÍvoire (LM)
Mexico (UM) Uruguay (UM) Qatar (U) Georgia (LM)
Turkey (UM) Costa Rica (UM) Saudi Arabia (U) Guatemala (LM)
Jordan (LM) Honduras (LM)
Malaysia (UM) Ecuador (LM)
Lebanon (UM) Kenya (L)
South Africa (UM) Egypt (LM) Madagascar (L)
Thailand (L) Panama (UM) Iran (LM) Moldova (LM)
Tunisia (LM) Syria (LM) Mozambique (L)
Chile (UM)
Colombia (UM) Indonesia (LM) Nepal (L)
Argentina (UM) Nigeria (LM)
Vietnam (L) Argentina (UM) Pakistan (LM)
Paraguay (LM)
Chile (UM) Colombia (UM)
Senegal (L)
Tanzania (L)
Peru (UM) Peru (UM)
Uganda (L)
Zambia (L)
Vietnam (L)
Note: UM: Upper-middle income, LM: Lower-middle income, L: Low income
In the following, I will analyse and interpret the characteristics of each cluster
separately. Policy recommendations for each cluster will be provided in the discussion
section.
• Technology developers. These include those countries which are characterised
by large economies, with relatively high income per capita, the high availability
of credit for the private sector, high protection of IPR, good industrial
competitiveness, a well-functioning logistical infrastructure as well as a high
production of fossil fuels per capita. Among members of this cluster, Argentina
clearly underperforms in terms of credit for the private sector, and IPR
protection and is placed in between the group of technology developers and
that of countries in need of structural changes. China is a high achiever in terms
of total income, credit availability, industrial competitiveness and the quality of
its logistical infrastructure. Malaysia also outperforms most peers in credit
availability for the private sector, industrial competitiveness and logistical
performance as well as South Africa in the same categories, except industrial
competitiveness. All of these countries, except India and Argentina, showed
very high scores in the first principal component, indicating very good
performance in most enabling factors for technology transfer, including
variables such as the quality of their private businesses, ease of doing business
and the stock of local and foreign patents. Most countries also showed a good
score in principal component 3, showing that both supply and demand factors
are in place to enable clean energy technology transfer. These countries are
240
expected to be the main recipients of foreign technologies and to be capable of
using them to implement clean energy projects that achieve emission
reductions. These countries are also expected to absorb foreign technologies to
develop their own endogenous capabilities and create local clean energy
technologies, which seems to have been particularly the case in Malaysia, China
and Mexico, countries that export large volumes of clean energy technologies
per capita higher than their imports. However, the relatively high availability of
fossil fuels may place barriers to the competitiveness of clean energy
technologies.
• Technology implementers. These consist of countries with a relatively small
economy and low industrial competitiveness but high levels of income per
capita, good levels of credit availability for the private sector and IPR protection
and good logistical infrastructure. These countries also show very low levels of
fossil fuel production per capita, which would reflect in high fossil fuel prices
that would create the necessary demand-pull signals for investments in clean
energy technologies. Chile´s performance exceeds other cluster members for
all differentiating variables except industrial competitiveness, which is why it is
placed in between the group of technology developers and technology
implementers. This group of countries as a whole is expected to attract
significant levels of foreign clean energy technology transfer per capita, as a
result of its relative wealth and its need of energy security. This is already
shown in their current levels of imports of clean energy technologies. However,
the lack of a large internal demand and a competitive industry would prevent
them from becoming clean energy technology leaders.
• Countries in need of structural changes. These include relatively large
economies with high levels of fossil fuels production per capita, good levels of
income per capita but low industrial competitiveness, credit availability for the
private sector and in most cases low protection of IPR and low logistical
performance. Although many of these countries have economies large enough
and good levels of income per capita to attract foreign investment, they would
not be expected to attract large amounts of foreign clean energy technologies
per capita, due to the lack of clear demand signs. As large fossil fuel providers,
fossil fuel prices are low in comparison to other countries and they do not have
the incentive to promote clean energy for energy security and geopolitical
reasons, as many other developing countries have. Additionally, their
economies do not provide a good environment for private investment. Most
countries show low scores for principal component 3, indicating a bad business
environment with high corruption, as well as a large stock of foreign patents, a
variable highly correlated to the production of fossil fuels per capita. Russia
outperforms the rest of the cluster members in terms of total and per capita
income, credit availability and IPR protection. As a result, Russia also
outperforms other cluster members in terms of imports and exports of clean
technologies per capita.
• Aid recipients. These are characterised by their poor performance in all the
enabling factors considered for the cluster analysis. These countries lack a
sufficient demand and the economic and institutional frameworks that attract
241
private investment to clean energy technologies, as well as the technological
capabilities to implement foreign technologies and the industrial fabric to
develop their own technologies. They are countries placed as part of Cluster 3
in Ward’s and k-means methods and showing low scores for the principal
component 1, indicating bad performance in most of the highly correlated
enabling factors for technology transfer. Vietnam is an outlier in this cluster,
and could also belong to the “technology implementers” cluster, as it shows a
good performance in terms of economy size, credit availability for the private
sector, industrial competitiveness and logistic infrastructure. Accordingly,
Vietnam also outperforms its peers in terms of imports and exports per capita
of clean technologies. This cluster is expected to require higher levels of foreign
aid to build the necessary economic, institutional, technological and logistic
infrastructure required as a precondition for the attraction and implementation
of foreign clean energy technologies.
5.6 Discussion and conclusions

This chapter has followed a quantitative approach to complement answers to the
research questions previously provided by the literature review and the case study
analysis. Three multivariate analysis techniques have been used – multiple regression,
principal component analysis and cluster analysis – to analyse relationships between
enabling factors, unravel their relationship with technology transfer variables and
classify developing countries according to their performance in enabling factors for
technology transfer and their policy priorities. Publicly available data were used to
construct the models used in this section.
The results of the analysis could be used to define national policy priorities eligible for
support by the technology mechanism of the UNFCCC, in line with its manifested
“country-driven approach” and its seven priority areas, as described in the
introduction to this thesis. The analysis has several limitations, mainly due to the lack
of data, which prevent modelling the complexity of the technology transfer process
including all the flows through which technology can enter a country, their outcomes
and further spillovers. A multi-country quantitative analysis also misses many
specificities of the countries analysed, which should be assessed further through
detailed country-specific studies.
Answers to the research questions of this doctoral thesis are provided in the
remainder of this section.
5.6.1 Is it possible to measure climate change technology transfer to developing countries,

and what are the main challenges in doing so?
This chapter has dealt with the quantitative analysis of international clean energy
technology transfers. It has therefore involved measuring clean energy technology
transfer at the developing country level as well as the factors that drive, enable or
prevent it from occurring.
Previous chapters on the literature review and case study analysis had suggested a
number of measurements of clean energy technology transfer. Given the complexity of
the TT process, these suggested measurements included:
242
• Several measurements of the channels through which TT can enter a recipient
country, including imports, FDI, foreign R&D expenses and research
agreements, ODA, foreign patents and licences.
• Several measurements of the outputs that these foreign flows of technologies
can deliver, such as installed capacity of renewable energy generation using
foreign technologies, clean electricity generated using foreign technologies,
the production of clean energy technologies in the recipient country using
foreign knowledge or equipment, spin-offs established, scientific publications
and local patents generated with foreign support.
• Several measurements of the further spillovers foreign technologies can
achieve in the recipient economy, such as increases in TFP, cost reductions of
clean energy technologies and CO2 emission reductions beyond the specific
project implemented with foreign technologies.
Such a rich quantitative representation of the TT process in developing countries was
not possible due to the unavailability of data for most of the previous indicators. Only
three particular indicators were available for a significant sample of developing
countries: renewable energy generation capacity adjusted for the estimated
requirements of technology transfer, imports and exports of a selection of clean
energy technologies. Imports of clean energy technologies reflect the openness of the
recipient country to foreign flows of technology and its attractiveness as a location for
investments in the implementation of these technologies. The renewable generation
capacity using foreign knowledge and/or equipment shows the ability of the recipient
country to turn imported equipment and knowledge into functioning equipment.
Generation capacity and electricity generation are strongly correlated, so we can
assume that in principle installed capacity is effectively operated. Finally, exports of
clean energy technologies show the ability of developing countries to produce their
own technologies and achieve international competitiveness. Import and export data
values are for 2009, while the latest available renewable energy capacity data are for
2008.
An analysis of the correlations between the three proxies of clean energy technology
transfer in absolute and relative values (per capita and per GDP) shows a high
correlation between exports per capita and GDP and the absolute renewable energy
capacity, which may indicate that countries that have reached a significant scale of
installed capacity have the ability to produce more of their own technologies than
countries with smaller capacity. Correlations also show that countries with the largest
absolute renewable generation capacities also have large capacities per capita and per
income, indicating that economies of scale and learning effects may have facilitated a
larger deployment.
This thesis recognises that the three proxies of technology transfer used for the
analysis only provide an incomplete picture of the process, although they do offer a
good starting point for the quantitative study of international climate change
technology transfers beyond the CDM. The analysis should be further improved as new
data becomes available.
243
5.6.2 Empirically, which factors enable clean energy technology transfer to developing
countries?
As with measurements of climate change technology transfer, the previous sections
regarding the literature review and case study analysis delivered a long list of enabling
factors, classified under two criteria: the aspect of technology transfer they influence
(input, output or effect) and the type of factor (related to economic and institutional
frameworks, technology demand, technology supply or industrial development). The
initial long list of indicators was reduced to a more manageable short list according to
two criteria: data availability and correlations between variables. Variables for which
no data were available for a significant sample of developing countries, or were highly
correlated to other variables, were removed from the analysis. Consequently, a total of
26 enabling factors were used for the analysis. These are described in Table 27. Many
of the 26 variables are strongly correlated, as shown by the principal component
analysis that summarised the information provided by 14 highly correlated indicators
with three principal components, calculated as the linear combination of the original
variables and explaining 72% of their variance.
A multiple regression analysis was performed to explore the relationship between the
identified enabling factors and the actual values of technology transfer proxies. Several
iterations were performed with the 26 enabling factors to find the best fitted model,
following which three equations were derived, one per dependent variable: imports of
clean technologies, exports of clean technologies and installed renewable capacity
using TT, all of them expressed in per capita values to avoid the scale effect.
Four variables can jointly explain 69% of the variation of the exports of clean energy
technologies per capita, namely the exporter´s endowment of renewable energy
resources, the level of protection of IPR, the competitive industrial performance and
the size of its economy. The results indicate that countries with a favourable
renewable energy endowment may have developed a competitive advantage in the
production of technologies to exploit that potential at a low cost. The possibility of
demonstrating local technologies in their own territory at a low cost improves the
possibilities of learning by doing and scale effects benefitting further cost reductions.
The high protection of IPR provides the right signals to local and foreign innovators to
invest in new technologies that may not deliver profits at the earlier stages of
development. High industry competitiveness indicates that the exporter has in place
an industrial infrastructure capable of producing new equipment and creating
synergies across several industrial sectors. Finally, the size of the economy indicates a
local demand allowing for local demonstration and mass production to achieve cost
reductions through learning-by-doing and scale effects.
Imports of clean energy technologies per capita can be explained by two single
variables, which capture 59% of the variation of the explanatory variable. Firstly, a high
income per capita leads to higher imports of clean energy technologies, which
corresponds to the Kuznets hypothesis of higher demand for environmental quality as
income per capita increases. Secondly, the availability of credit for the private sector is
also essential in achieving high levels of clean energy imports, which points to the
importance of the private sector as a provider and consumer of clean energy
technology.
244
Five variables can explain 40% of the variation in renewable electricity generation
capacity involving TT per capita in developing countries, adjusted per the percentage
of capacity expected to have required foreign technology transfer. Firstly, income per
capita, which, as in the case of imports, may indicate a higher demand for
environmental quality in higher income countries. The renewable energy endowment
of the host country is also significant, indicating that countries that have access to
these resources at a lower cost are more likely to take advantage of their potential. IPR
protection has a negative coefficient, which shows that high protection may deter the
uptake of foreign clean renewable energy technologies. A good logistical infrastructure
also explains higher levels of renewable energy capacity, given the need to transport
often large and heavy equipment to remote locations where the renewable energy
resources reside. Finally, the production of fossil fuels per capita in the host countries
has a negative coefficient, indicating that the availability of cheap fossil fuels can leave
out of the market the often more expensive alternative renewable energies.
Identified explanatory variables with positive coefficients have a high positive
correlation with other enabling factors, such as the ease of doing business, the quality
of local businesses, the total factor productivity of the economy, the stock of local and
foreign patents, the levels of enrolment in tertiary education, the corruption
perception index, the freedom for foreign investment, the volume of high technology
exports per capita and CO2 per capita. Fossil fuel production per capita, explaining
installed renewable generation capacity with a negative coefficient, has a high
negative correlation to the price of fuels, but it is positively correlated to GDP and GDP
per capita, local and foreign patent stock, the logistical infrastructure of the country,
its TFP, the quality of its private businesses and levels of enrolment in tertiary
education.
An interesting outcome of the regression analysis was the different signs for IPR
protection, positive for exports of local clean energy technologies, but negative for the
local deployment of foreign clean energy technologies. The results indicate that IPR
protection can be positive for those countries that have reached a level of internal
capability that allows them to develop their own technologies. However, it could hold
back the adoption of foreign technologies for those countries at the stage of
implementation.
It is expected that an important part of the variation of the dependent variables that
could not be explained with the available explanatory variables may be explained
through the policy portfolio of host countries. A good mix of technology-push, market-
pull and industrial policy may be responsible for the better performance of some
countries. However, the effect of policies could not be quantified, as their diversity
could not be captured through dummy variables that showed significance. The effects
of policies would be better analysed through localised case studies in relevant
countries.
5.6.3 How do different developing countries perform in these enabling factors, and which
Two multivariate analysis techniques have been used to analyse the developing
country performance in technology transfer enabling factors: principal components
analysis and cluster analysis. The aim of the analysis was to classify developing
245
countries according to policy priorities in order to improve their levels of technology
transfer. This approach differs from the current neutral approach of the UNFCCC,
which has tended to consider the developing world as a single block. Four groups of
developing countries have been defined, as described in the previous section about
cluster selection:
1. Technology developers. These are those countries with the potential to excel in
all three aspects of clean energy technology transfer: the attraction of foreign
flows of technologies, the efficient operation and maintenance of foreign
equipment and the generation and management of technological change
through indigenous efforts to absorb foreign technologies. Countries in this
group are large economies, with relatively high incomes per capita, the high
availability of credit for the private sector, high IPR protection, good industrial
competitiveness and a well-functioning logistical infrastructure. The cluster
includes mostly upper-middle income countries but also lower-middle income
China and India and low-income Thailand, which shows an excellent industrial
competitiveness and credit availability for the private sector.
Policies in these countries depend on the specific stage of clean energy
technological development of each country, but in general terms they should
start with effective demand-pull policies that attract investments in clean
technology, complemented by technology-push policies that increase the local
capacity to use and maintain the technologies and transcend foreign
knowledge to create their own endogenous technologies. Some level of
industrial policy could be required to support local infant industries, as has
been shown by the experiences of China and India. The large demand size and
growth of these countries offer high potential gain for foreign technology
providers, which could be willing to accept restrictive industrial policies such as
those adopted in China and, to a lesser extent, in India. These policies should
only be temporary or otherwise risk creating uncompetitive industries. Other
countries in this cluster should learn from the success stories of exemplary
technology leaders of emerging economies. Given their favourable conditions,
countries in this cluster are probably in less need of international support than
the members of the remaining clusters.
2. Technology implementers. These are countries with relatively small economies
but high levels of income per capita, good levels of credit availability for the
private sector and IPR protection, even though their industrial competitiveness
and logistical infrastructure are still far from the levels of the group of
technology developers. These countries usually show very low levels of fossil
fuel production per capita, which would reflect in high fossil fuel prices,
creating the necessary demand-pull signals for investments in clean energy
technologies. The group includes mostly upper-middle income countries in
Africa, Latin America and the Middle East, but also lower-middle income Latin
American and Middle Eastern countries. It may also include Vietnam, a low
income country but with better performance in several enabling factors and
technology transfer indicators than low income countries in the group requiring
foreign aid. This cluster is not expected to develop high-tech clean energy
technologies, as it lacks a sufficient internal demand allowing scale effects and
learning-by-doing, or a competitive industrial sector. However, it is expected to
246
attract significant levels of foreign clean energy technology transfer per capita,
as a result of its relative wealth, its need of energy security and in many
instances, good renewable energy endowments. This is already shown in their
current levels of imports of clean energy technologies.
Their policy priorities should focus on demand-pull measures to increase
investment in clean technologies and improve internal capabilities through
learning-by-doing and the development of local support industries providing
services or low-tech components to the clean energy industry. These measures
could include clauses requiring from foreign investors the transfer of
technologies to local suppliers. Policies could also focus on supporting
demonstration projects to improve local capabilities.
3. Countries in need of structural changes. These are relatively large countries,
with high levels of fossil fuels production per capita and mostly high income or
upper-middle income, although they also include low income countries such as
Ecuador, Egypt, Iran and Indonesia. They show low industrial competitiveness
and, with the exception of rich Middle Eastern oil producers, low availability of
credit for the private sector, low IPR protection and a low logistical
performance. They also show significant levels of corruption and low ease of
doing business. This cluster´s main characteristic is the low fossil fuel prices
that render alternative renewable energies uncompetitive.
The policy priorities for these countries should therefore be to improve the
economic and institutional conditions favourable to private investment and to
create the appropriate demand signals for clean energy technologies, by
potentially eliminating subsidies for fossil fuels and using fossil fuel rents to
support investments in clean energy technologies.
4. Foreign aid recipients. This group is formed by mostly low income and lower-
middle income countries from Africa, as well as low and lower-middle income
countries from Asia and Latin America. Their poor performance in most of the
enabling factors for clean energy technology transfer indicates very low
attractiveness for foreign technology suppliers.
Policy priorities should focus on creating the appropriate capabilities,
institutions and infrastructure, before these technologies can be successfully
implemented. Given their low attractiveness for private investment, foreign aid
may be required to create these enabling conditions, and it could focus on
providing technological capabilities to local private companies and institutions,
supporting the reform of local institutions. Once appropriate institutions
capabilities and infrastructures are in place, an initial demand for clean energy
technologies could be created through, for example, demonstration projects.
The composition of the different groups is subject to changes as their performance
evolves and they implement policies to improve their main pitfalls. The analysis has
also shown that while some countries are very stable in their allocation to a group,
many others are very sensitive to the use of different clustering techniques or
explanatory variables. Therefore, the proposed cluster structure should be tested
through more in-depth country studies that back the defined policy priorities.
247
6 Summary and conclusions
6.1. Introduction
6.2. Major findings
6.3. Contributions and implications
6.4. Limitations and further
research
248
6.1 Introduction
This final chapter provides an overview of the most important conclusions of this
thesis. It seeks to answer the initial research questions and to reflect on the findings of
the analysis. Furthermore, it elaborates on the theoretical, methodological, empirical
and policy implications of the study, its contributions and limitations, and suggests
areas for future research.
The main purpose of this thesis is to gain a better insight into international TT to
developing countries in the field of climate change, and hence contribute to the design
of more effective policies. Chapter 1 provides an introduction to the international
policy background and highlights the main gaps of the current UNFCCC approach to TT.
The following gaps have been identified. Firstly, the UNFCCC´s disconnection from the
enabling factors that facilitate private investment leading to TT, secondly its uniform
approach across developing countries, which ignores the great diversity of the
developing world, and finally the lack of clear measurements of TT to assess the
magnitude of technology flows and their effectiveness in achieving knowledge
spillovers in recipient countries. Responding to these identified gaps, this thesis
analyses what constitutes TT, how it can be measured, what are its enabling factors in
developing countries and which context-specific policies can be implemented to
reinforce these factors. Central to this research is the understanding that developing
countries are highly heterogeneous and international policies can only be effective if
they target the specific needs of each country.
The outcomes of this thesis are particularly important for the current international
climate policy agenda. A technology mechanism (TM), consisting of a Technology
Executive Committee (TEC) and a Climate Technology Centre and Network (CTCN) are
being designed by the UNFCCC, after being agreed upon at COP16, which took place in
2010 in Cancun. The UNFCCC has stressed the importance of country-driven action to
promote the development and transfer of technologies in developing countries
through the TM, but it is not yet clear how the TM will contribute to setting
appropriate enabling environments in developing countries to achieve this aim. COP17
held in December 2011 in Durban has achieved some concrete outcomes regarding the
type of body that the TEC will be, with the adoption of its modalities and procedures.
The TEC has mainly an advisory role and some uncertainties remain as regards its
relationship with the CTCN and the essential linkage between the TM and the financial
mechanism. The CTCN is not yet fully operational and has not developed its modalities
and procedures. Other significant initiatives extra-UNFCCC, such as the ongoing
creation of climate innovation Centres in developing countries, promoted by the World
Bank, also call for research that informs the design of effective TT policies adapted to
each country´s circumstances.
Despite wide recognition of the importance of technology and knowledge transfer in
the climate change mitigation policy agenda, this issue has not received sufficient
attention in academic research. Chapter 2 of this thesis reviews the existing literature
and highlights existing gaps. Current definitions of climate change TT hardly take into
account the perspective of actors involved in actual climate change TT activities, while
existing measurements of climate change often TT do not consider the diversity of
channels through which these happen, nor take into account the outputs and effects
that these transfers convey. The enabling factors for TT in non-BRIC developing
249
countries have been seldom investigated, and policy recommendations to improve the
level and quality of TTs to developing countries have not been adapted to the specific
needs of highly heterogeneous countries commonly denominated as “developing”.
This thesis contributes to enriching the climate change TT debate from the perspective
of a smaller emerging economy (Chile) and the quantitative analysis of enabling factors
for TT in a large sample of developing countries.
This thesis uses two approaches to study climate change TT: comparative case study
analysis and quantitative analysis. Comparative case studies analysed TT processes in
ten cases based in Chile, which share the same economic, technological and policy
frameworks, thus enabling us to draw conclusions on the enabling factors and
obstacles operating in the TT processes. The quantitative analysis uses three
methodologies – principal component analysis, multiple regression analysis and cluster
analysis – to assess the performance of developing countries in a number of enabling
factors, the relationship between these factors and indicators of TT and to create
groups of developing countries with similar performances.
The following section summarises the main findings by responding to the four research
questions presented in the introduction to this study.
6.2 Main findings
6.2.1 What constitutes technology transfer, and how does it happen?

From a theoretical perspective, the concept of TT goes beyond the mere transmission
of equipment or codifiable knowledge, as it also involves the transfer of tacit
technological capabilities that cannot be subject to market-like exchanges. There is
therefore a clear differentiation between technology trade and the broader TT.
Organisational learning is a key element for the success of TT, and it is greater when
the recipient country has a strong network of local suppliers and research institutions
that can benefit from the transfer process through linkages with the recipient
organisation.
From an empirical perspective, a common pattern was identified in the ten case
studies of climate change TT analysed by this thesis. The process of climate change TT
consisted of the following elements:
• Inputs into the technology transfer process
These are a combination of local and foreign inputs. Local inputs were essential for the
success of every TT process. When the TT process involved a recipient local
organisation, the internal capabilities of the organisation enabled it to identify
opportunities to generate new revenue through technological activities, ultimately
leading to GHG emissions reduction. Internal capabilities subsequently enabled the
recipient to operate foreign equipment efficiently or to develop technologies based on
foreign designs. Foreign technologies could also spur local innovation when internal
learning process took place. When the TT process occurs through foreign direct
investment, local technology inputs allow for knowledge spillovers from the foreign
organisation to local agents, and it can prevent situations of technology dependence.
Some examples of local technology inputs identified through our case studies are
skilled local staff, in-house R&D, know-how about some of the aspects of the new
250
technology, local knowledge of the energy market or local technology needs, a well-
established network of suppliers or functioning production plants.
By definition, in a process of international TT, local inputs are not enough to deliver an
intended technology output efficiently. Foreign TT is hence required when local
capabilities and equipment are not sufficient, when they cannot compete with
international alternatives or when foreign multinationals prefer to use their own
resources or trusted foreign product and service providers instead of unknown local
alternatives. Some examples of required international technology inputs are
manufacturing equipment, skills and know-how for its operation and maintenance,
equipment or product designs and specifications and expertise to activate innovation
processes.
• Channels of foreign technology
International TT occurs when the required foreign technologies can flow into the local
economy. The mere identification of opportunities for TT by local organisations
requires direct exposure to international technological developments, which can
happen through, for example, participation in international fairs, access to
international publications, visits to foreign institutions, contacts with clients and
competitors through exporting activities, or contact with a head office in the case of
foreign subsidiaries. Furthermore, TT channels are required to make foreign input
available locally. Some examples of channels are imports of equipment, foreign
employees or consultants, foreign subsidiaries, international publications and
congresses, foreign patents, research partnerships with foreign organisations,
partnerships with foreign companies, training by foreign experts, or foreign
professional or educational experience of local employees.
Open economies have better chances of receiving foreign technology flows. The choice
of channel can also have an effect on the impact of the foreign technologies in the
local technological capabilities. In highly integrated channels, such as foreign direct
investment (FDI) with centralised headquarters, or turnkey contracts for equipment
installation, the knowledge tends to stay within the technology provider. Less
integrated channels such as imports of equipment, licensing, collaborative R&D with
foreign institutions or companies and ad hoc technical support from foreign experts
require the more active participation of the recipient and typically lead to an
improvement of the recipient´s technological capabilities. However, integrated
channels such as FDI also carry significant advantages, as foreign MNCs usually have
stronger financial and technological capabilities to deliver scale, particularly in the
most innovative technologies, and have the potential to deliver large emission and
cost reductions through demonstration, economies of scale and learning effects.
Accordingly, the most mature technologies are usually transferred through
disintegrated channels, mainly imports, and installed, operated and maintained locally,
whereas the most innovative and sophisticated technologies will be transferred
through integrated channels, mainly FDI.
• Output of the technology transfer process
The final, direct objective of the TT process is referred to as its output. Some examples
of outputs for a TT process are the local production of low-carbon technologies, the
251
implementation of renewable energy, energy efficiency or GHG reduction projects, the
supply of services, components or raw materials to low-carbon energy projects, the
publication of scientific research about low-carbon technologies, the generation of
electricity through low-carbon energy sources or local patents filed for low-carbon
technologies. These technology outputs must involve the transfer of foreign
technologies to be considered elements of a technology transfer process.
• Effects or spillovers of the technology transfer process
The effects or spillovers of TT processes go beyond the achievement of their intended
output. For example, they might involve knowledge spillovers to local organisations
through forward and backward linkages, which are non-pecuniary externalities that
arise when productive knowledge permeates into the local economy through, for
example, technological assistance given by foreign clients in order to improve the
quality of the product provided by local suppliers. Backward and forward linkages
happen in highly linked economies where local suppliers have the capacity to provide
equipment and components. Horizontal knowledge spillovers can also happen when
the acquired capabilities are used to implement other low-carbon energy projects.
Technology cost reductions through learning-by-doing are another example of the
effects of TT processes, which can also lead to improvements in national productivity
because local companies get involved in higher value-added activities or reduce their
production costs. The use of more efficient foreign technologies typically facilitates this
process.
According to this pattern identified through case studies, successful TTs require that
foreign technology inputs, together with local inputs, deliver their intended output
and, even better, convey further impacts to the recipient economy. If these outputs
and/or effects are not realised, flows of foreign technologies may not be considered
“transferred”.
doing so?
The measurement of TT processes should take into account all the elements identified
above, i.e. the channels through which foreign technology inputs flow, the technology
outputs resulting from these inputs and the technology effects conveyed beyond the
intended output. The different measurements used by the existing literature have
tended to focus only on channels of foreign technologies, which provide a narrow
perspective of the TT process.
Table 46 compiles the suggested indicators for the measurement of TT at the national
level, as provided by existing literature and complemented by our case studies. Our
suggested indicators provide a more holistic approach than the currently proposed at
the UNFCCC, which focuses primarily on input indicators.
252
LITERATURE REVIEW AND CASE STUDY ANALYSIS
Type of TT Literature review Case studies

indicator
Input- Foreign R&D expenditure weighted by the R&D expenses of multinationals in CC-
channel bilateral import-share related sectors
Value of climate change-related FDI
Value of climate change-related imports
Value of climate change-related ODA
Value of collaborative research agreements
Financial resources provided by the UNFCCC
for TNA and capacity building
Value of joint public-private climate change
R&D programmes
Research agreements with private
companies
Foreign patents filed related to climate
change technologies
Number of foreign workers in the CC
technology sector
CC-related licence payments to foreign
companies
Output Number and licence revenues of CC-related local patents produced with foreign support
CC-related scientific publications
Number of spin-offs established
Invention disclosures evaluated
Citations received by patents originating in a
country by patents originating in another
country, related to climate change
technologies
Number of CDM projects that incorporate
foreign technologies or foreign knowledge
MW installed by foreign owners or with
foreign technologies
kWh of kJ generated or saved by foreign-
owned projects or with foreign
technologies
tCO2e captured or destroyed by foreign-
owned projects or with foreign
technologies
Production of CC-related technologies or
products by foreign owners or with
Effect TFP of CC-related industrial sectors explained by foreign CC-related inputs
tCO2e reductions in the recipient economy explained by foreign CC-related inputs
The case study analysis also suggested the assessment of project-level TT through
indicators of technology inputs, outputs and effects, as presented in Table 47. Some of
these indicators are similar to those considered at the national level, whereas others,
such as fees paid to foreign service providers, travel expenses for technological
missions or technology cost-efficiency differential with local technologies, are more
appropriate for measurements at the project level.
253
TABLE 47- APPROACHES TO MEASURE PROJECT-LEVEL CLIMATE CHANGE TECHNOLOGY TRANSFER
BASED ON CASE STUDY ANALYSIS
Input-channel Output Effect

• Value of CC-related imports • Installed capacity • tCO2e reduced
• Value of CC-related FDI • kWh of kJ generated or saved • Technology cost-
• Fees paid to foreign service providers • tCO2e captured or destroyed efficiency differential
or employees in CC industries • Units or value of CC-related with widely available
• Number or value of CC-related technological products and technologies locally
licences bought to foreign companies services • Productivity gains
• Travel expenses for CC-related • Number or licence revenue • Local content
international technological missions from CC-related patents • Number of projects or
• R&D expenses in CC-related • Number of CC-related installed capacity
collaborative projects with foreign scientific publications resulting from learning
companies effects of the project
Under the proposed comprehensive approach, different ratios of outputs and effects
to foreign technology inputs would indicate the varying importance of local technology
inputs, as well as other control variables. Countries and companies with a solid local
technological basis would be more efficient in turning foreign technology flows into
final outputs and effects. To allow for comparability between different projects, this
measurement approach should be used separately for each technology type, along
with indicators with comparable units, in each of the TT elements.
Unfortunately, most data required for such a comprehensive approach to the
measurement of climate change TT at the national level are not available for a
significant sample of developing countries. Data availability was reviewed as part of
Chapter 5 of this thesis, reaching the following conclusions for each type of indicator.
• Indicators of channels for the flow of foreign technology inputs
Data on climate change-related FDI are not yet available for most developing
countries. Fortunately, significant progress is under way, mostly encouraged and
actioned by UNCTAD, the OECD and think tanks such as the Climate Policy Initiative
(Buchner et al., 2011; UNCTAD, 2010; Chaum et al., 2011). Progress is driven by the
need to monitor developed countries’ fulfilment of their climate finance provision
pledges as part of the Copenhagen Accord. Data on foreign workers in climate change-
related sectors, climate change-related licence payments and foreign R&D expenses in
developing countries are not available on international databases for a significant
sample of developing countries. Nevertheless, data on climate change-related patents
filed by foreign actors are available for an increasing number of countries, thanks to
efforts by the European Patent Office (EPO) and the OECD´s Directorate for Science,
Technology and Industry to develop a worldwide patent database (PATSTAT) which
indexes climate change mitigation-related patents. Finally, data on climate change-
related inputs are available for a selection of low-carbon technologies through the UN
Trade Statistics Database (COMTRADE).
• Indicators of technology transfer process outputs
Data on renewable energy capacity installed by foreign owners or with foreign
technologies do not exist as such. However, renewable capacity data in developing
countries are available, and they can be adjusted to account for the use of foreign
254
technology using claims of foreign TT in the CDM. Total renewable energy capacity in
developing countries and renewable capacity of CDM projects are strongly correlated.
Therefore, we can assume that the claims of TT in CDM projects can be extrapolated to
the total renewable capacity of a country.
Data on the production of climate change-related technologies or products in
developing countries by foreign owners or with foreign technologies is not available.
However, export values for renewable and energy efficient technologies by developing
countries can be taken as a proxy for the production of internationally competitive
low-carbon technologies.
Data on revenues from local patents developed with foreign support, and local
scientific publications in collaboration with foreign institutions, are not available for
most developing countries.
• Indicators of the effect of climate change-related technology transfer
None of the identified indicators of the effect/spillovers of climate change TT was
available for most developing countries. Measurements of tCO2e reductions, or TFP
increases explained by foreign CC-related inputs, are limited by the lack of input data.
Measurements of learning effects, or the reduction of technology costs as the
production levels or the installed capacity of low-carbon technology increase, are not
available for most developing countries and technologies either, due to the low
availability of data on accumulated production or capacity in developing countries. For
example, only two studies of learning effects for wind technology in China (Qiu and
Anadon, 2011) and for biofuels in Brazil (Goldemberg et al., 2004) are known. It is
expected that more of these studies will be feasible as low-carbon generation capacity
and the production of low-carbon technologies increase in developing countries.
The quantitative analysis in this thesis used three indicators of TT on the basis of their
availability for a significant sample of developing countries and their relevance after
the insights gained from the case study analysis. These indicators are renewable
energy generation capacity adjusted for the estimated requirements of TT and imports
and exports of a selection of clean energy technologies.
6.2.3 Which factors enable climate change technology transfer to developing countries?
A number of enabling factors were identified in the literature review of this thesis, and
then classified in five types: macroeconomic and institutional frameworks, technology
demand, technology supply, industrial development and bilateral characteristics of the
supplier and recipient.
The case study analysis has shown that enabling factors operate at all levels of the TT
process defined by our conceptual framework. They enable the supply of local and
foreign technology inputs, the transfer of foreign inputs through TT channels, the
materialisation of local and foreign inputs into technological outputs and the
achievement of further knowledge spillovers benefiting the recipient economy. The
case study analysis in Chile has also contributed to extend the list of enabling
frameworks identified by the existing literature.
255
A compilation of enabling factors gathered through the literature review and the case
study analysis is presented in Table 48, which also indicates the type of factor, their
expected effect on TT and the TT element they enable. Chapters 2 and 4 provide a
wider definition of these factors.
TABLE 48- ENABLING FACTORS SUGGESTED BY THE LITERATURE REVIEW AND THE CASE STUDY
ANALYSIS

Economic and Ease of doing business Positive Channels
institutional Corruption Negative of foreign
frameworks IPR protection Positive technology
Open trade Openness to trade and FDI Positive Channels
Stable macroeconomic framework Positive of foreign
Regional economic integration Positive technology
Flows of high technology FDI Positive
Access to credit for private Positive Output
developers of clean energy projects
Technology Demand size: Positive Output
demand • Population
• GDP absolute and per capita
• Energy consumption absolute and
per capita
Demand growth Positive
• Annual GDP growth
• Energy consumption growth
Electricity prices High electricity and fossil fuel Positive
prices
Demand-pull policies Depends
on policy
and
implement
ation
Stability of electricity prices Positive
Technology Human capital Positive Input
supply • Enrolment in tertiary education
• % of degrees in engineering and science
• Annual number of graduates in science
and engineering
• Literacy rates
• Quality of education
• Specific capabilities in clean energy
technology
• Workforce with intermediate Positive Input
technical qualifications for
transformation activities
• Policies to facilitate education Positive Channels
abroad of foreign
technology
R&D Positive Input
• General R&D
spending
• R&D spending in
energy-related
technologies
256
Patent stock: • General and clean technology Positive Input
• Stock of patents specific knowledge stock
of specific
technology fields
• Quality of the
stock of patents,
measured as
citation levels
Technology Positive Input
development
indexes
Technology push policies: Positive Input
• Government sponsored R&D
• Tax credits for R&D
• Subsidies to education
• Infrastructure development
• Funding for demonstration projects
• National certification programmes
• Functioning National Innovation Positive Input,
System effects
• Good transport infrastructure Positive Input
• Access to sources of renewable Positive Input
energy
Industrial Competitive industrial sector enabling cross- Positive Input,
development sectoral TT effect
Market power in related industries Negative Output
(electricity sector):
• Profit levels of incumbent companies
• Market concentration
Cost of factors: labour and energy Negative Input
Industrial policy: Industrial policy : Positive in Input,
• Local content • Specific promotion of clean BRICs, effect
requirements energy industries Uncertain
• Differentiated • Attraction of high-technology FDI in smaller
tariffs for • Specific policies for local economies
targeted innovators
technologies • Promotion of backward and
• Differentiated forward linkages
fiscal treatment
for targeted
technologies
• Large production capacity to Positive Output,
bring technology prototypes to effect
market
• High quality of local suppliers Positive Effect
• Linked domestic industrial sector Positive Effect
Chapters 3 and 4 analyse the specific enabling and deterring factors in operation in
Chile, which is a relatively small economy, with a very low share of global GHG
emissions, a high dependence on foreign fossil fuels and the high availability of
renewable energy sources. In spite of this large renewable energy endowment, and a
good environment for private investment, the penetration of non-conventional
renewable energy technologies (defined as all renewables except large hydro) has
257
been slow. This thesis has identified the main enabling factors of climate change TT to
Chile that are present or missing, for each of the TT elements in our conceptual
framework:
• Enablers of local technology inputs. The main enablers of local technology
inputs are the technology knowledge acquired by industries related to
extractive activities, mainly mining and forestry, the quality of the human
capital (even though Chile’s educational system has some problems such as
the wide gap between the highest and lowest income groups and the lack of
intermediary technical qualifications), the technological knowledge acquired in
biomass and hydro renewable energy technologies, the R&D capabilities of
some local universities and some national policies supporting R&D. Also, many
of the analysed TT processes were “resources-pushed”, driven mainly by the
availability of renewable energy sources or raw materials required for
technology development.
• Enablers of the flow of foreign technologies. Chile has excellent conditions for
foreign investment, with a stable macroeconomic framework, strong
institutions, a liberal economy and open trade agreements with many
countries, low corruption, low income tax that attracts foreign employees, a
good transport infrastructure, strong intellectual property rights and ease of
doing business. This enables the flow of foreign technologies through trade
and FDI, which represents a high share of national GDP. However, it is mostly
related to extractive activities (mining) and services, with a very low incidence
of productive and technological FDI. To solve this problem, the Chilean
Government has implemented policies to attract high-technology FDI, but their
results still have to be assessed.
• Enablers of TT outputs. Chilean demand is small in comparison to other Latin
American countries, but it has experienced strong growth rates. For instance,
electricity consumption per capita is higher than in any other Latin American
country and has experienced strong growth. Electricity prices are also high for
Latin American standards. All of this should attract investment in low-carbon
technologies. Chile has implemented several policies to increase demand for
renewable energy technologies. The most important is Law 20.257 on the
promotion of non-conventional renewable energy sources, as it establishes a
green tradable certificates system to achieve a 5% participation of NCRE in
electricity sales from 2010, which will reach 10% by 2024. However, so far this
law has not incentivised significant growth in NCRE projects due to a number
of flaws in its design and implementation, mainly the lack of a liquid market for
green certificates and the law’s applicability to only a share of the total
electricity generation. The high volatility of electricity prices increases the risks
of investing in electricity generation projects and makes it difficult to obtain
credit for renewable energy project developers. Besides, a highly concentrated
electricity supply market poses high barriers of entry for new clean energy
ventures. The Clean Development Mechanism of the Kyoto Protocol has
supported several low-carbon projects, though, making Chile the third Latin
American country with the highest number of CDM projects in its portfolio.
However, experts argue that the CDM has not supported “additional” carbon
emission reductions, meaning that the supported projects would have taken
258
place anyway. Interviewed experts consider that there is a deficit of effective
demand-pull policies in the country.
• Enablers of technology spillovers and effects. Chile lacks a strong industrial
sector that could manufacture its own renewable energy technologies or
provide inputs to foreign suppliers. Its economy is based on natural resources-
based export markets, with a significant concentration in mining, pulp and
paper, fruit and fishing. Chile favours the static comparative advantages
provided by the abundance of its resources over specialisation in higher value-
added activities in the industrial sector. Domestic manufacturers have been
substituted for their foreign counterparts, so the country’s industrial fabric has
become weak and delinked. As a result, foreign TTs have a low potential to
generate knowledge spillovers through backward and forward links with
domestic suppliers. The general approach of the government is to concentrate
on what they do better (natural resources-based industries), buy from abroad
at a competitive price what they cannot produce and provide support to the
local private sector, but only once they have taken their own risks.
All the results obtained through the analysis of Chile cannot be extrapolated directly to
all developing countries, as they apply only to the particular circumstances of this
country. However, some of the identified enabling factors would apply to most non-
BRIC developing countries, such as the importance of having some competitive
industries in place to enable cross-sectoral knowledge transfer, the importance of
favourable conditions for private investment and the limitations that small local
demand poses for the development of national renewable energy industries.
To provide a wider perspective of enabling frameworks for CC TT, beyond a single
country study, Chapter 5 undertakes quantitative analysis for a sample of up to 71
developing countries. The relationships between the identified enabling factors and
the available measurements of TT are analysed by multiple regressions. To do so, the
long list of potential enabling factors was reduced to a more manageable shortlist of
26 explanatory variables, taking into account data availability, the number of valid
cases and correlations among variables of the same type. When variables were highly
correlated, only one of them would be selected. The 26 enabling factors are described
in Table 27 in Chapter 5.
The analysis showed that the most important enabling factors for the import of low-
carbon technologies (in per capita values) were income per capita and the availability
of credit for the private sector. Both variables capture 59% of the variation of the
explanatory variable, which indicates that income levels and favourable conditions for
private investment are behind the flows of foreign technologies through a
disintegrated channel such as imports.
The most important enabling factors for the export of clean energy technologies by
developing countries (in per capita values) are the exporter´s endowment of
renewable energy resources per capita, the level of IPR protection, competitive
industrial performance and total GDP. These four variables jointly explain 69% of the
variation of clean energy technologies exports per capita. These results indicate the
importance of static comparative advantages related to the availability of renewable
energy sources employed to develop competitive renewable energy industries in
developing countries. The importance of economies of scale for the development of
259
local industries is highlighted by the variable size of the economy (GDP). A highly
competitive industrial performance indicates that the exporter has in place an
industrial infrastructure, capable of producing new equipment and creating synergies
across several industrial sectors. Finally, high IPR protection provides the right signal to
local and foreign innovators to invest in new technologies that may not deliver profits
during the earlier stages of development.
Finally, the most important enabling factors for renewable electricity generation
capacity with TT (in per capita values) are income per capita, renewable energy
resources endowment per capita, IPR protection (with a negative coefficient), logistical
infrastructure and the production of fossil fuels per capita with a negative coefficient.
These five variables can explain 40% of the variation in renewable electricity
generation capacity per capita in developing countries. The results are intuitively
correct, as countries with easier access to renewable energy resources are more likely
to take advantage of this potential. A good logistical infrastructure enables the
transportation of often large and heavy equipment to remote locations where the
renewable energy resources can be found. The availability of local fossil fuels would
lower the price of electricity generation reducing the incentives for investment in often
more expensive renewable energies. Finally, IPR protection has a negative impact on
the implementation of renewable energy projects, opposite to its impact on the
development of local renewable energy technology industries. This indicates that,
while IPR protection can be positive for countries with high internal capabilities to
develop endogenous technologies, it could hold back the adoption of foreign
technologies for those at the implementation stage.
The effect of policy variables could not be introduced in the models, due to the high
diversity of policy types and their different degrees of implementation, which
precluded their treatment as dummy variables. An important part of the variation of
the dependent variables that could not be explained by the models could potentially
be explained through these omitted policy variables.
It is worth noting that the significant explanatory variables in the model are strongly
correlated to other enabling factors, which would also have an impact on the levels of
TT. However, these variables could not be included in order to avoid collinearity.
An analysis of the correlation between the three indicators of clean energy TT in
absolute and relative values (per capita and per GDP) showed that the size of a country
is important in creating local productive capacity. The analysis showed a high
correlation between exports per capita and GDP and the absolute renewable energy
capacity, which may indicate that countries that have reached a significant scale of
installed capacity have the ability to produce more of their own technologies than
countries with smaller capacity. Correlations also show that countries with the largest
absolute renewable generation capacities also have large capacities per capita and per
income, indicating that economies of scale and learning effects may have facilitated a
larger deployment.
6.2.4 How do different developing countries perform in these enabling factors, and how
can differentiated policy priorities be defined accordingly?
The study of Chile provided some insights into policy priorities for a relatively small
economy with a very good environment for private investment, an open economy,
260
high energy prices, widely available renewable energy resources and good human
capital, but a weak industrial infrastructure, ineffective demand-pull policies and a
non-interventionist government favouring a neoliberal economic model.
The main policy priorities for Chile are, firstly, to boost demand, by solving the pitfalls
of its recently approved green tradable certificates law and promoting regional
integration with other fast-growing Latin American countries. Secondly, the lack of
some sort of industrial policy to support renewable energy technology initiatives was
consistently highlighted as an obstacle to technology transfer throughout the different
case studies. Even though it would be challenging for Chile to create global technology
leaders in sophisticated renewable energy technologies, such as India, China or
Malaysia have done, there are realistic opportunities for auxiliary or smaller scale
industries, such as wind blades, wind towers, CSP heat storage systems or mini-hydro
turbines, perforation equipment, pipes and valves for geothermal wells and
biodigesters, purifiers or collectors for biomass. There are also wide opportunities for
servicing companies to evaluate resource availability, carry out topographic and
sysmical terrain studies or operate generation plants. Industrial development requires
some level of intervention to support the accumulation of technological and
organisational capabilities, so the Chilean Government may therefore need to send the
right signals to the local private sector to show their support for these technologies;
otherwise, foreign investors may keep on sourcing their components and services from
foreign suppliers, preventing the achievement of the type of backward spillovers that
multiply the effect of TT. It is unlikely that Chile would apply trade and investment
measures to support local renewable energy industries, as they would contradict its
free market economy. However, there are other policies that can facilitate information
and matchmaking, the promotion of national linkages, technology upgrading, training
and the financing of initial capital investment or demonstration projects.
Policy recommendations for Chile show that its challenges are very different to those
of BRIC countries. However, they cannot be directly extrapolated to all non-BRIC
developing countries because developing countries per se are highly heterogeneous
and need technology policies tailored to their specific circumstances. The quantitative
analysis in Chapter 5 aimed at identifying the factors that determine the different
policy priorities of developing countries and at grouping them into clusters sharing
similar policy priorities. Principal components and cluster analysis techniques were
applied for this purpose. Four groups could be differentiated, as presented in Figure
68, which also shows those countries that are not clearly attributable to one cluster
but are instead placed between two of them.
261
FIGURE 68- CLUSTER SELECTION FOR TECHNOLOGY TRANSFER RECIPIENT COUNTRIES
TECHNOLOGY TECHNOLOGY STRUCTURAL AID RECIPIENTS

DEVELOPERS IMPLEMENTERS CHANGES
Bangladesh (L)
Bolivia (LM)
Brazil (UM) Botswana (UM) Algeria (UM) Benin (L)
China (LM) El Salvador (LM) Russia (UM) Cameroon (LM)
India (LM) Jamaica (UM) Oman (U) Côte dÍvoire (LM)
Mexico (UM) Uruguay (UM) Qatar (U) Georgia (LM)
Turkey (UM) Costa Rica (UM) Saudi Arabia (U) Guatemala (LM)
Jordan (LM) Honduras (LM)
Malaysia (UM) Ecuador (LM)
Lebanon (UM) Kenya (L)
South Africa (UM) Egypt (LM) Madagascar (L)
Thailand (L) Panama (UM) Iran (LM) Moldova (LM)
Tunisia (LM) Syria (LM) Mozambique (L)
Chile (UM)
Colombia (UM) Indonesia (LM) Nepal (L)
Argentina (UM) Nigeria (LM)
Vietnam (L) Argentina (UM) Pakistan (LM)
Paraguay (LM)
Chile (UM) Colombia (UM)
Senegal (L)
Tanzania (L)
Peru (UM) Peru (UM)
Uganda (L)
Zambia (L)
Vietnam (L)
Note: UM: Upper-middle income, LM: Lower-middle income, L: Low income
Each cluster has the following specific characteristics and policy priorities:
• Technology developers. These are those countries with the potential to excel
in all three aspects of clean energy TT, i.e. attracting foreign flows of
technologies, the efficient operation and maintenance of foreign equipment
and the generation and management of technological change through
indigenous efforts to absorb foreign technologies. Countries in this group are
characterised by large economies, with relatively high income per capita, high
availability of credit for the private sector, high IPR protection, good industrial
competitiveness and a well-functioning logistical infrastructure. The cluster
includes mostly upper-middle income countries but also lower-middle income
China and India and low-income Thailand, which shows an excellent industrial
competitiveness and credit availability for the private sector.
Policies in these countries depend on the specific stage of clean energy
technological development of each country, but in general terms they should
start with effective demand-pull policies that attract investments in clean
technology complemented by technology-push policies that increase the local
capacity to use and maintain the technologies and transcend foreign
knowledge to create their own endogenous technologies. These countries can
learn from the success stories of India, with leading wind turbine
manufacturers, China, with leading wind turbine and solar PV technologies
manufacturers, and Malaysia, leaders in biomass energy technologies. Thus,
some level of industrial policy could be required to support local infant
industries, as has been shown by the experiences of China and India. The large
262
demand size and growth of these countries offer high potential gain for foreign
technology providers, which could be willing to accept restrictive industrial
policies such as those adopted in China and, to a lesser extent, in India. These
policies should only be temporary or otherwise risk creating uncompetitive
industries. Among members of this cluster, Argentina clearly under-performs in
terms of credit for the private sector and IPR protection, and it is placed in
between the group of technology developers and that of countries in need of
structural changes. Chile under-performs in the size of its economy and its
industrial competitiveness, and it is placed between this group and the
“technology implementers” group.
• Technology implementers. These are countries with relatively small economies
but high levels of income per capita, good levels of credit availability for the
private sector, IPR protection and a good environment for private investment,
even though their industrial competitiveness and logistical infrastructure are
still far from the levels of the group of technology developers. These countries
usually show very low levels of fossil fuel production per capita, which would
reflect in high fossil fuel prices, creating the necessary demand-pull signals for
investments in clean energy technologies. The group includes mostly upper-
middle income countries in Africa, Latin America and the Middle East, but also
lower-middle income Latin American and Middle Eastern countries. It may also
include Vietnam, a low income country but with better performance in several
enabling factors and TT indicators than low income countries in the group
requiring foreign aid. Chile´s performance exceeds other cluster members for
all differentiating variables except industrial competitiveness, which is why it is
placed in between the group of technology developers and technology
implementers. This cluster is not expected to develop high-tech clean energy
technologies, as it lacks sufficient internal demand that allows scale effects and
learning-by-doing and lacks a competitive industrial sector. However, it is
expected to attract significant levels of foreign clean energy TT per capita, as a
result of its relative wealth, its need for energy security and, in many instances,
good renewable energy endowments. This is already shown in their current
levels of clean energy technologies imports.
Their policy priorities should focus on demand-pull measures to increase
investment in clean technologies and improve internal capabilities through
learning-by-doing, the financing of demonstration projects and the
development of local support industries providing services or low-tech
components to the clean energy industry.
• Countries in need of structural changes. This group includes relatively large
economies with high levels of fossil fuels production per capita, good levels of
income per capita, but low industrial competitiveness, credit availability for the
private sector and, in most cases, low IPR protection, low logistical
performance and an unfavourable environment for private investment. They
are mostly high income or upper-middle income countries, although they also
include lower-middle income countries such as Ecuador, Egypt, Iran and
Indonesia. Although many of these countries have economies large enough and
good levels of income per capita to attract foreign investment, they are not be
expected to attract large amounts of foreign clean energy technologies per
263
capita, due to the lack of clear demand signs. As large fossil fuel providers,
fossil fuel prices are low in comparison to other countries, which renders
renewable energies uncompetitive. Besides, there is not an incentive to
promote clean energy for energy security and geopolitical reasons.
Additionally, their economies do not provide a good environment for private
investment, showing high levels of corruption and low ease of doing business.
Russia outperforms the rest of the cluster members in terms of total and per
capita income, credit availability and IPR protection. As a result, it also
outperforms other cluster members in terms of imports and exports of clean
technologies per capita.
The policy priorities for these countries should therefore be to improve the
economic and institutional conditions favourable to private investment and to
create the appropriate demand signals for clean energy technologies, by
potentially eliminating subsidies for fossil fuels and using fossil fuel rents to
diversify their electricity generation portfolio.
• Aid recipients. This group is formed from mostly low income and lower-middle
income countries from Africa, as well as low and lower-middle income
countries from Asia and Latin America. Their poor performance in most of the
enabling factors for clean energy TT indicates very low attractiveness for
foreign technology suppliers. These countries lack a sufficient demand and the
economic and institutional frameworks that attract private investment to clean
energy technologies, as well as the technological capabilities to implement
foreign technologies and the industrial fabric to develop their own
technologies.
The appropriate capabilities and infrastructure need to be created before
climate change technologies can be successfully implemented. Given their low
attractiveness for private investment, foreign aid may be required to create
these enabling conditions, and it could focus on providing technological
capabilities to local private companies and institutions, supporting the reform
of local institutions and creating an initial demand for clean energy
technologies through, for example, demonstration projects. TT policy in most of
these countries should focus on creating these building blocks for the future
success of technological activities. Vietnam is an outlier in this cluster, and
could also belong to the “technology implementers” cluster. Vietnam shows
good performance in terms of economy size, credit availability for the private
sector, industrial competitiveness and logistic infrastructure, and it also
outperforms its peers in terms of imports and exports per capita of clean
technologies.
The composition of the different groups is subject to changes as their performance
evolves and they implement policies to improve their main pitfalls.
6.3 Contributions and implications

The contributions and implications of the research have been presented in each
chapter. Here, we provide a more general overview of theoretical, empirical,
methodological and policy implications.
264
6.3.1 Theoretical and empirical
As discussed in the literature review of Chapter 2, there is not an overarching approach
for the definition of technology transfer, its measurements and enabling frameworks
in the context of climate change. Existing literature on climate change technology
transfer to developing countries offers a narrow perspective of TT, a lack of case
studies beyond BRIC countries, very few quantitative studies and the lack of the
perspective of actors involved in real climate change technology transfer activities.
This research has contributed to filling some of these gaps through several means.
Firstly, it has reviewed a wide body of literature covering several disciplines such as
Management of Technology, International Business Management, Knowledge Transfer,
Economics of Technology Diffusion, International and Industrial and Development
Economics. This literature provides a solid theoretical background for the
conceptualisation of TT, its measurement and the identification of its enabling factors.
Building from the reviewed literature and from the evidence gathered through a
comparative case study analysis, this thesis proposes a conceptual framework for the
study of climate change TT. The TT process is dissected into several elements, all of
them necessary to achieve successful transfers: technology inputs, channels for the
transfer of foreign inputs, technology outputs and effects. Several enabling or support
factors are required by each of these elements. This research is also unique in
proposing a broad perspective for measuring climate change TT based on the
conceptual framework previously defined.
Empirically, this thesis has contributed to existing case study literature on climate
change TT to developing countries, which is scarce and has mostly focused on the
success stories of BRIC economies, such as the emergence of leading wind turbine
manufacturers in India and China, world-leading Chinese photovoltaic technology or
Brazilian biofuels production. However, these countries do not represent the average
developing economy because China, India and Brazil offer potential investors the
prospect of large profits and have an advanced domestic technological base that
allows for the rapid adoption of foreign technologies. Smaller developing countries
would struggle to replicate BRIC countries’ success stories, given their smaller
bargaining power towards foreign technology providers and their smaller internal
demand, which slows down economies of scale and learning effects. This paper
contributes to the TT debate with the country analysis of Chile and ten new case
studies of Chilean or foreign companies requiring TT to implement low-carbon energy
initiatives in the country. Through the case studies, the perspectives of those directly
involved in TT processes have been taken into account to assess the importance of
several enabling factors and to complement those identified by the literature.
Selecting Chile is relevant for showing the challenges faced by companies in non-BRIC
countries in attracting and successfully absorbing foreign technologies. Chile is,
nevertheless, a particular country due to its sustained growth, stable macroeconomic
conditions, liberal economy and OECD membership, which may limit the extrapolation
of our results to other developing countries.
Another significant empirical contribution of this study is the analysis of the
performance of a wide sample of developing countries as regards a set of enabling
factors for climate change TT. To our knowledge, no previous literature maps a large
265
sample of developing countries according to their performance in such a large number
of enabling factors for climate change TT. The main focus of previous quantitative
analysis was on the measurement of TT, instead of factors that can explain why some
countries are more successful than others in attracting, implementing and achieving
knowledge spillovers from foreign climate change technologies. This thesis can
contribute to international climate change technology policymaking, by suggesting that
efforts should focus on the specific priorities of relatively homogeneous groups of
countries.
6.3.2 Methodological approach

Methodologically, this thesis suggests a new approach to measuring TT and to
structuring case study analysis, based on the previously defined conceptual
framework. A methodology is also proposed to classify enabling factors according to
the TT element they support (inputs, channels, outputs, spillovers) and to their
typology (institutional and macroeconomic, technology demand, technology supply,
industrial).
The case study analysis methodology combines a high level analysis of the target
country with fieldwork to incorporate the perspective of actors directly involved in TT
processes. This approach proved particularly effective in identifying new enabling
factors not previously highlighted by the literature and to ponder their relevance in
specific TT processes.
One of the main methodological contributions of this thesis is the application of
principal components analysis (PCA) and cluster analysis (CA) to the study of climate
change TT. To our knowledge, this is the first study to use these techniques for the
specific field of climate change technology transfer. Measuring climate change TT is
challenging, because technology has no measurable physical presence or a well-
defined price, and its transfer goes beyond the mere flow of foreign technology inputs
into a recipient country. The broad approach to measuring climate change TT
proposed by this thesis could solve these problems, but it is limited by data
unavailability for most indicators of climate change TT. For this reason, quantitative
analysis techniques that focus on the analysis of explanatory variables, instead of the
relationship between dependent and independent variables, are particularly useful.
PCA and CA have been able to provide answers about the technology policy needs of
developing countries that previous econometric literature had left unresolved. This
thesis also uses multiple regression analysis to understand the relationship between
independent variables and three dependent variables representing different aspects of
climate change TT.
The proposed quantitative methodology combining three multivariate techniques –
PCA, CA and multiple regression analysis – has proved that the combined use of these
three methods can increase the robustness of results and frame the TT policy needs of
developing countries in a more comprehensive way. This methodology could
potentially be further used to study international policy priorities to promote TT to
allocate funds through the technology and/or financial mechanisms of the UNFCCC.
The methodology can be further improved as new indicators of climate change TT are
developed to inform international climate finance decisions.
266
6.3.3 Policy
The findings of the study have a number of policy implications for the promotion of
climate change TT to developing countries. Some of those derived by this thesis are
specifically related to Chile, for which an in-depth country analysis has been
performed. However, in these conclusions we refer to a wider range of policy
implications at the international level.
Firstly, international instruments used to promote climate change TT to developing
countries should be more clearly aligned to the enabling frameworks that facilitate
private sector investment. This thesis has identified a large set of national-level
enabling factors for climate change TT which the recently agreed technology
mechanism should aim to support. In doing so, the technology mechanism should
move from the existing project-by-project approach of the UNFCCC to a more
ambitious tactic that supports policymaking at the national level.
Secondly, the diversity of the developing world should be recognised. International TT
policies should be adapted to each country’s circumstances. The UNFCCC TT discourse
has been based on a north-south TT paradigm that divides the world into two blocks –
technologically active and passive – but observation of actual technology flows and the
performance of developing countries in a set of enabling factors shows that this
simplistic paradigm no longer holds. This thesis has classified the developing world into
four large groups according to their performance in TT metrics and their enabling
frameworks. Each of these groups has very different policy priorities and success
stories in one of the groups would not be easily replicated in another. The proposed
classification, which is expected to change as countries implement new policies, could
be the basis for setting technology-funding priorities by the technology mechanism of
the UNFCCC or by other international initiatives, such as the climate innovation centres
promoted by the World Bank.
Finally, this thesis has stressed the importance of appropriate measurements of
climate change TT to assess the performance of related policies and to decide the
allocation of technology funds. The existing metrics of TT at the UNFCCC level are
exclusively focused on UNFCCC input activities into the TT process, with uncertain
effects on the actual transfer of technologies. No indicators about the effect of TT are
included in the list of indicators used by the UNFCCC to monitor TT. UNFCCC activities
have in fact had a limited effect on the actual transfer of technologies to developing
countries. Input indicators are therefore considered insufficient for monitoring the
success of policies aiming at increasing the levels of TT. A broader approach for the
measurement of TT, such as the one proposed by this thesis, could contribute to the
effectiveness of future international efforts to promote climate change TT.
6.4 Limitations and future research

Along with some contributions and interesting findings, this doctoral thesis has a
number of limitations. Here, I would like to briefly summarise the most substantial
ones and suggest future areas of research to address them.
The qualitative analysis is limited by the study of a single country, Chile. While this
provided good insights into the different challenges faced by relatively small
developing or emerging economies, it still provides a partial view of the developing
267
world, given Chile´s particular characteristics, mainly its good economic performance
and strong institutions that have granted it a place among the restrictive club of
wealthy countries in the OECD. More case studies are therefore required that reflect
the diversity of challenges faced by different developing countries. As a starting point,
case studies should be selected to represent each of the groups defined: technology
developers, technology implementers, countries requiring structural changes and aid
recipients.
The quantitative analysis in this thesis is limited by the unavailability of data on climate
change TT. Only three variables were used to represent the different aspects of climate
change TT (imports and exports of renewable energy technologies and renewable
generation capacity involving TT) and to identify the most significant enabling
frameworks. These three variables only provide an incomplete picture of the process.
The quantitative analysis has the potential to be significantly improved as more data
become available on climate finance flows and on learning effects through the
implementation of low-carbon technologies. The regression analysis could be also
improved by identifying causalities and crossed effects among correlated variables.
Additionally, further research would be required about potential methods for
measuring the effectiveness of climate change technology transfer.
Finally, future research will be necessary to propose the most cost-effective way to
implement concrete policies that will improve TT in developing countries, particularly
within the framework of the UNFCCC mechanisms.
268
7 References
269
Afifi, A., 2004. Computer-Aided Multivariate Analysis. 4th Edition. Chapman & Hall CRC
Akyüz, Y., 2009, Industrial Tariffs, International Trade and Development. Chapter 6 of
“Industrial Policy and Development” Cimoli, M., Dosi, G. and Stiglitz, J.E.
ACTI, 2009. Policies for clean technology innovation and dissemination. Alliance for
Clean Technology Innovation.
Amsden, 2001. The Rise of the Rest: Challenges to the West from Late-Industrializing
Economies. Oxford: Oxford University Press
Ang, J. B., 2009. CO₂ emissions, research and technology transfer in China. Ecological
Economics 68 (10): 2658-2665.
APEC, 2009. Energy Efficiency in Chile- Background document for the APEC Peer
Review on Energy Efficiency (PREE), February 2009.
Archibugi, D. and Coco, A., 2004. "A New Indicator of Technological Capabilities for
Developed and Developing Countries (ArCo)." World Development 32(4): 629-654.
Archibugi, D. and Michie, J., 1997. Technological globalisation and national systems of
innovation: an introduction. In Technology, globalisation and economic performance.
D. Archibugi and J. Michie. Cambridge, Cambridge University Press.
Archibugi, D. and Pianta, M. 1996. "Measuring technological change through patents
and innovation surveys." Technovation 16(9): 451-468.
ACERA, 2011. Análisis de la ley ERNC (20.257). International Institute for Sustainable
Development, Santiago de Chile.
Banco Central de Chile, 2010. Cuentas Nacionales de Chile Año 2009.
www.bcentral.cl/prensa/notas-prensa/pdf/18032010ccnn.pdfS. Santiago de Chile.
Barton, J. H., 2007. Intellectual Property and Access to Clean Energy Technologies in
Developing Countries - An Analysis of Solar Photovoltaic, Biofuel and Wind
Technologies. ICTSD Programme on Trade and Environment. International Centre for
Trade and Sustainable Development, Geneva.
Bayoumi, T., Coe, D. T., Helpman, E. 1999. "R & D spillovers and global growth." Journal
of International Economics 47(2): 399-428.Bell, M. 1987, The Acquisition of Imported
Technology for Industrial Development: Problems of Strategies and Management in
Arab Region (Baghdad: ESCWA, UN, 1987), pp. 13-29.
Blackman, A. and Wu, X., 1999. Foreign direct investment in China's power sector:
trends, benefits and barriers. Energy Policy 27(12): 695-711.
Blalock, G. and Gertler, P. J., 2009. How firm capabilities affect who benefits from
foreign technology. Journal of Development Economics 90(2): 192-199.
Bloomberg, 2009.“Marubeni invests in new electric car-related joint venture” Monday,
24 August 2009.
Bloomberg, 2010. “Former Central Banker Corbo Sees Chile’s Rate Rising to 5.75% in
18 Months”. Bloomberg news at www.bloomberg.com/news2010-07-22/former-
central-banker-corbo-sees-chile-s-rate-rising-to-5-75-in-18-months.htm
270
Borensztein, E., De Gregorio, J. and Lee, J.W., 1998. How does foreign direct
investment affect economic growth?. Journal of International Economics 45(1): 115-
135
Branstetter, L. G., 2001. Are knowledge spillovers international or intranational in
scope? Microeconometric evidence from the US and Japan. Journal of International
Economics 53(1): 53-79.
Brewer, T. L., 2009. Technology Transfers and Climate Change: International Flows,
Barriers and Frameworks. In: Brainard, L. and Sorkin, I. (eds) Climate Change, Trade,
and Competitiveness: Is a Collision Inevitable?. Washington DC, The Brookings
Institution.
Buchner, B., Brown, J., Corfee-Morlot, J., 2011. Monitoring and tracking long-term
finance to support climate action. OECD/IEA, Paris.
Burns, R. and Burns, R., 2008. Business Research Methods and Statistics Using SPSS.
Sage Publications Ltd.
Butler, L. and Neuhoff, K., 2008. Comparison of feed-in tariff, quota and auction
mechanisms to support wind power development. Renewable Energy 33 1854–1867
Cabral, L. 2000. Introduction to Industrial Organisation. Cambridge, Massachusetts,
The MIT Press.
Cameron, 2005. Technological convergence, R&D, trade and productivity growth.
European Economic Review 49 (2005) 775 – 807
Cannady, C., 2009. Access to climate change technology by developing countries: A
practical strategy. Report for ICTSD.
Cantwell, J., 2009. Innovation and information technology in the MNE. In: A. M.
Rugman (Ed.) The Oxford handbook of international business. Oxford University Press.
Cattell, R.B., 1966. The scree test for the number of factors. Multivariate Behavioral
Research, 1, 245-76.
Chaum, M., Faris, C., Wagner, G., Buchner, B., Falconer, A., Trabacchi, C., Brown, J.,
Sierra, K., 2011. Improving the effectiveness of climate finance: key lessons. Climate
Policy Initiative, 23rd November 2011.
Chemetall, 2009. Chemetall Statement on Lithium Applications and availability, July
2009
China Automotive Review, 2009. “China to support lithium-ion batteries for EVs” Jul 8,
2009.
Cimoli, M. and Di Maio, M., 2004. Has the Chilean Neo-Liberal Experiment Run Out of
Fuel? A View on Specialisation, Technological Gaps and Catching-Up. Quaderni,
Università degli Studi di Siena, Dipartimento di Economia Politica. N. 426.
Cimoli, M., Dosi, G. and Stiglitz, J.E., 2009. Industrial Policy and Development. Oxford
University Press
Coe, D. T. and Helpman, E., 1995. International R&D spillovers. European Economic
Review 39(5): 859-887.
271
Coe, D.T., Helpman, E. and Hoffmaister, A.W., 1997. North-South R&D Spillovers. The
Economic Journal 107: 13-149.
Cohen, G., 2004. Technology transfer: strategic management in developing countries,
Sage publications.
Cole, M. A. and Elliott, R. J. R., 2003. Determining the trade-environment composition
effect: the role of capital, labour and environmental regulations. Journal of
Environmental Economics and Management 46(3): 363-383
Cole, M., 2006. Does trade liberalisation increase national energy use?. Econ Lett 92:
108-112.
CNE and GTZ, 2009. Non-conventional renewable energy in the Chilean electricity
market. Proyecto Energías renovables no convencionales. Comisión Nacional de la
Energía, Santiago de Chile.
CNE, 2010. Fijación de Precios de Nudo SIC and SING, April 2010. Comisión Nacional de
la Energía, Goverment of Chile, Santiago de Chile.
CNE, 2011. Capacidad Instalada por Sistema Eléctrico Nacional 2010. Comisión
Nacional de la Energía, Goverment of Chile, Santiago de Chile.
CNE, 2011. National Energy Balances 2000 to 2008. Comisión Nacional de la Energía,
Goverment of Chile, Santiago de Chile.
CNE, 2011. Precios de Nudo de la Energía. Comisión Nacional de la Energía, Goverment
of Chile, Santiago de Chile.
COCHILCO, 2009. “Minuta: Reunión de discusión estudio antecedentes para una
política pública en minerales estratégicos: el litio”, 1st October 2009. Comisisión
Chilena del Cobre, Santiago de Chile.
COCHILCO, 2009. Antecedentes para una Política Pública en Minerales Estratégicos:
Litio. Comisisión Chilena del Cobre, Santiago de Chile.
Corporación Nuclear Eléctrica Chile, S.A.,. 2009. Análisis relativo de impactos y riesgos
de la generación núcleo-eléctrica. June, 2009. Santiago de Chile.
de Coninck, H., Fischer, C. Newell, R. G. and Ueno, 2008. International technology-
oriented agreements to address climate change. Energy Policy 36(1): 335-356.
de la Tour, A., M. Glachant and Ménière, Y., 2010. Innovation and international
technology transfer: the case of the Chinese photovoltaic industry. CERNA working
paper series. Paris, Cerna, Centre d´économie industrielle, MINES Paris Tech.
Dechezleprêtre, A., Glachant, M. and Ménière,Y., 2008. The Clean Development
Mechanism and the international diffusion of technologies: An empirical study. Energy
Policy 36(4): 1273-1283
Dechezleprêtre, A., Glachant, M. and Ménière,Y., 2010. What drives the international
transfer of climate change mitigation technologies? Empirical evidence from patent
data, Nota di Laboro 12.2010, Fondazione Eni Enrico Mattei.
Deutsche Bank, 2009. Autos & auto parts Electric Cars: Plugged in 2. from
http://www.db.com
272
Di Maio, M., 2004. Industrial Policies in Developing Countries: History and
Perspectives. Quaderno di Dipartimento n.148. Università degli Studi di Macerata-
Dipartimento di Istituzioni Economiche e Finanziarie.
Dinar, A., Rahman, S.M., Larson, D. and Ambrosi, P., 2008. Factors Affecting Levels of
International Cooperation in Carbon Abatement Projects. Policy Research Working
Paper. T. W. Bank, Development Research Group- Sustainable Rural and Urban
Development Team: 38.
Doranova, A. ,2009. Technology transfer and learning under the Kyoto regime:
exploring the technological impact of CDM projects in developing countries, PhD
thesis, Maastricht University.
Dougherty, C., 2010. Introduction to Econometrics. 4th Edition. Oxford University Press
Dufey, A., 2010. Oportunidades y barreras domésticas a las energías limpias en Chile.
International Institute for Sustainable Development.
Durferrit, 2010. “A tank full of sun”, Durferrit promotional material, 2010
Eaton, J. and Kortum, S., 1999. International patenting and technology diffusion:
theory and measurement. International Economic Review 40: 537-70.
Eaton, J. and Kortum, S., 2002. Technology, geography, and trade. Econometrica 70(5):
1741-1779.
Eisenhardt, K.M., 1989. Building theories from case study research. Academy of
Management Review 14 (4): 532-550
Ernst&Young and Fraunhoffer Institute, 2011. “Middle East and North Africa Region
Assessment of the Local Manufacturing Potential for Concentrated Solar Power (CSP)
Projects”. Report for the World Bank .
European Commission, 2009. Metrics for Knowledge Transfer from Public Research
Organisations in Europe, Report from the European Commission’s Expert Group on
Knowledge Transfer Metrics
Frankel, J. A. and Rose, A. K., 2002. Is trade good or bad for the environment? Sorting
out the causality. NBER Working Paper no 9201. National Bureau of Economic
Research, Cambridge.
Gagnon, Y.C., 2010. The Case Study as Research Method- A Practical Handbook.
Presses de l’Université de Québec
Gainza-Carmenates, R., Altamirano-Cabrera, J.C., Thalmann, P. and Drouet, L., 2011.
Trade-offs and performances of a range of alternative global climate architectures for
post-2012, Environmental Science and Policy, Volume 13, Issue 1, Pages 63-71.
Galaz, R., 2010. Análisis crítico de la implementación de la Ley de Fomento de ERNC en
Chile y propuesta de mejoras. Report prepared for the CER.
Gboney, W., 2009. Policy and regulatory framework for renewable energy and energy
efficiency development in Ghana. Climate Policy 9(5): 508-516.
Goldemberg, J., Coelho,S.T., Nastari, P.M., Lucon,O., 2004. Ethanol learning curve- the
Brazilian experience. Biomass and Bioenergy, 26, 301-304.
273
Grant, K., 2009. Concentrated solar power in South Africa. Climate Policy 9(5): 544-
552.
Green Car Congress, 2009. “German Government to Award up to €9M to Support
Development of Battery-Electric Mercedes-Benz Vans”, 17 November 2009,
Green Car Congress, 2010. “South Korea to invest some $12.4B over 10 years to
become dominant in rechargeable li-ion batteries”. 13th July 2010
Grubb, M., 2004. "Technology, innovation and climate change policy: an overview of
issues and options." Keio Economic Studies 41(2): 103-132.
Grubb, M., 2010. Copenhagen: back to the future?. Climate Policy 10: 127-130.
Grubler, A. and Nakicenovic, N., 1991. Long waves, technology diffusion and
substitution. IIASA Research Report.
Grübler,A., Nakicenovic,N., Victor,D.G.,1999. Dynamics of energy technologies and
global change. Energy Policy 27,247-280.
Haas, R., Panzera, C.,Rescha, G.,Ragwitz, M., Reece, G., Held, A., 2011. A historical
review of promotion strategies for electricity from renewable energy sources in EU
countries, Renewable and Sustainable Energy Reviews 15 1003–1034
Han, J. and Kamber, M., 2006. Data mining: concepts and techniques. 2nd Edition.
Morgan Kauffman Publishers.
Hansen, U. E., 2008. Technology and knowledge transfer from Annex I countries to
non-Annex I countries under the Kyoto Protocol´s Clean Development Mechanism
(CDM) - An empirical case study of CDM projects implemented in Malaysia, Master´s
thesis, Department of Geography and Geology. Copenhagen, University of
Copenhagen.
Hascic, I. and Johnstone, N., 2009. The Kyoto Protocol and international technology
transfer: an empirical analysis using patent data. 17th Annual Conference of the
European Association of Environmental and Resource Economists. EAERE 2009
Amsterdam.
He, G. and Morse, R.K., 2010. Making carbon offsets work in the developing world:
lessons from the Chinese wind controversy. Working paper 90. Freeman Spogli
Institute for International Studies.
Hira, A. and de Oliveira, L. G., 2009. No substitute for oil? How Brazil developed its
ethanol industry. Energy Policy 37(6): 2450-2456.
Hoekman, B. and Javorcik, B. S., 2006. Global integration and technology transfer,
Palgrave Macmillan.
Holi MT, Wickramasinghe R and van Leeuwen M, 2008. Metrics for the evaluation of
knowledge transfer activities at universities. Cambridge: Library House, 2008.
Hubler, M. and Keller, A., 2010. Energy savings via FDI? Empirical evidence from
developing countries. Environment and Development Economics 15: 59-80.
Hultman, N. E., Pulver, S., Guimaraes, L., Deshmukh, R. and Kane, J., 2010. Carbon
market risks and rewards: Firm perceptions of CDM investment decisions in Brazil and
India. Energy Policy.
274
IMD, 2010. World Competitiveness Yearbook 2010, May 2010. Institute for
Management Development, Lausanne.
IPCC, 2000. Methodological and technological issues in technology transfer. Special
Report of Working Group III of the Intergovernmental Panel on Climate Change,
Geneva.
ICTSD, 2011, The Climate technology mechanism: Issues and Challenges. Information
Note no. 18. March 2011. International Centre for Trade and Sustainable
Development, Geneva.
IEA, 2009. Chile Energy Policy Review. International Energy Agency, Paris.
IEA, 2010, Global gaps in clean energy RD&D. International Energy Agency, Paris.
IEA, 2010. Energy Technology Perspectives. Paris. International Energy Agency, Paris.
IEA, 2011. Clean Energy Progress Report. International Energy Agency, Paris.
IEA, 2010. Key World Energy Statistics. International Energy Agency, Paris.
IEA. 2011. Electricity Prices Statistics. International Energy Agency, Paris.
IISD, 2006. Summary of the Twelfth Conference of the parties to the UNFCCC and
Second meeting of the parties to the Kyoto Protocol. Earth Negotiations Bulletin 12.
International Institute for Sustainable Development, Geneva.
InvestChile, 2011. Economic Indicators. CORFO. Santiago de Chile.
Ivarsson, I. and C. G. Alvstam, 2005. Technology transfer from TNCs to local suppliers in
developing countries: A study of AB Volvo's truck and bus plants in Brazil, China, India,
and Mexico. World Development 33(8): 1325-1344.
Jaffe, A. and Trajtenberg, M. 1996. Flows of knowledge from universities and federal
laboratories: modeling the flow of patent citations over time and across institutional
and geographic boundaries, Proceedings of the National Academy of Sciences of the
United States of America 93 (1996) 12671–12677.
Jaffe, A., Trajtenberg, M. and Henderson, R., 1993. Geographic localization of
knowledge: spillovers as evidenced by patent citations, Quarterly Journal of Economics
108 (1993) 577–598.
Jensen, PH, Palangkaraya, A, and Webster, E, 2009. A Guide to Metrics on Knowledge
Transfer from Universities to Business in Australia. Intelllectual Property Research
Institute of Australia. Occasional Paper No. 03/09.
Jung, M., 2006. Host country attractiveness for CDM non-sink projects, Energy Policy,
Volume 34, Issue 15, Pages 2173-2184
Junginger,M., Faaij,A., Turkenburg, W.C. ,2005. Global experience curves for wind
farms. Energy Policy 33,133–150.
Keller, W. , 2004. International technology diffusion. Journal of Economic Literature
42(3): 752-782.
Keller, W. 1998. Are International R&D Spillovers and Trade Related? Analyzing
Spillovers among Randomly Matched Trade Partners," European Economic Review 42,
pp. 1469-81.
275
Keller, W. 1999. How Trade Patterns and Technology Flows Affect Productivity Growth.
NBER Working Papers 6990, National Bureau of Economic Research.
Kelly, G., 2007. Renewable energy strategies in England, Australia and New
Zealand.Geoforum 38 326–338
Kidder, T., 1982. Soul of a new machine, New York. Avon.
La Razon, 2010. “Bolivia inicia sola el proceso para industrializar el litio”, 22nd October
2010,
Lema, R. and Lema, A. 2010. Whither technology transfer? The rise of China and India
in Green Technology Sectors. Whither technology transfer? the rise of China and India
in green technology sectors. Paper prepared for the 8th GLOBELICS International
Conference “Making Innovation Work for Society: Linking, Leveraging and Learning”,
Kuala Lumpur, 1-3 November
Lewis, J. I., 2010. The evolving role of carbon finance in promoting renewable energy
development in China. Energy Policy 38(6): 2875-2886.
Lewis, J., 2007. Technology Acquisition and Innovation in the Developing World: Wind
Turbine Development in China and India. Studies in Comparative International
Development (SCID) 42(3): 208-232
Lewis, J.I., 2011. Building a national wind turbine industry: experiences from China,
India and South Korea, Int. J. Technology and Globalisation, Vol. 5, Nos. 3/4, pp.281–
305
Li, J.F., Gao, H., Ma, L.J., Wang, Z.Y., Dong, L.Y., 2008. China Wind Power Report 2008.
China Environmental Science Press, Beijing
Lichtenberg, F. and Potterie, B., 2001. Does Foreign Direct investment Transfer
Technology Across Borders?. Review of Economics and Statistics 66: 304-7
Lowe, M., Tokuoka, S., Trig. g, T. and Gereffi, G., 2010. Lithium-ion Batteries for Electric
Vehicles: The U.S. Value Chain. Research on behalf of the Environmental Defence Fund,
published by the Centre on Globalization, Governance & Competitiveness of Duke
University.
Maldonado, P. and Palma, G., 2004. “Seguridad y calidad de abastecimiento eléctrico a
más de 10 años de la reforma de la industria eléctrica en países de América del Sur”,
Serie 72, CEPAL, julio de 2004.
Mallett, A., D. G. Ockwell, et al. 2009. UK-India collaborative study on the transfer of
low-carbon technology: Phase II Final report. University of Sussex Institute of
Development Studies TERI.
Marcellino, D. and Gerstetter, C., 2010. Technology transfer in the international
climate negotiations, Ecologic Institute, Washington DC.
Marubeni, 2010. Marubeni corporation Annual Report 2010.
Mazumdar, J., 2001. Imported machinery and growth in LDCs. Journal of Development
Economics 65(1): 209-224.
McKinsey&Company, 2009. Pathways to a low-carbon economy- Version 2 of the
global greenhouse gas abatement cost curve.
276
Mielnik, O. and Goldemberg, J., 2002. Foreign direct investment and decoupling
between energy and gross domestic product in developing countries. Energy Policy
30(2): 87-89.
Ministerio de Economía – Gobierno de Chile, 2010. Resultados de las Encuestas de
Innovación e I+D 2007-2008. División de Innovación. Santiago de Chile.
Massachussets Institute of Technology, 2009. “Stimulus big winner: battery
manufacturing” 17 February 2009, Technology Review, published by MIT.
www.technologyreview.com/energy/22188/
Mowery, D. and N. Rosenberg, 1979. "Influence of market demand upn innovation-
Critical review of some recent empirical studies" Research Policy 8(2): 102-153
Nemet, G. F., 2009. Demand-pull, technology-push, and government-led incentives for
non-incremental technical change. Research Policy 38: 700–709
Neuhoff, K., Cooper, S., Laing, T., Lester, S., and Rysanek, A., 2009. Indicator Choices
and Tradeoffs: Facilitating the Success of International Climate Policies and Projects.
University of Cambridge, EPRG Working Paper 09/32
Nunnally, J. C., 1978. Psychometric Theory (2nd ed.). New York: McGraw Hill
Ockwell, D. G., Watson, J., MacKerron, G., Pal, P. and Yamin, F., 2008. Key policy
considerations for facilitating low-carbon technology transfer to developing countries.
Energy Policy 36(11): 4104-4115
OECD, 2001. Measuring productivity- Measurement of aggregate and industry-level
productivity growth Statistics. Organisation of Economic Co-operation and
Development, Paris
OECD and IDB, 2010. Strengthening institutional capacities for innovation policy design
and implementation in Chile. Assessment and Proposals. Organisation of Economic Co-
operation and Development, Paris.
OECD, 2010. Science, Technology and Industry Outlook 2010. Organisation of
Economic Co-operation and Development, Paris.
OECD, 2010. Economic Surveys- Chile. Organisation of Economic Co-operation and
Development, Paris.
OECD, 2010. Factbook 2010.Organisation of Economic Co-operation and Development,
Paris.
Peña, D. 2002. Regresion y diseño de experimentos. Alianza Editorial, Madrid.
Peterson, S., 2008. Greenhouse gas mitigation in developing countries through
technology transfer?: a survey of empirical evidence. Mitigation and Adaptation
Strategies for Global Change 13(3): 283-305.
Pizer, W.A. and Popp, D., 2008. Endogenizing technological change: Matching empirical
evidence to modeling needs. Energy Economics, Volume 30, Issue 6: 2754-2770
Popp, D., 2008. International Technology Transfer for Climate Policy, Policy Brief
39/2009, Centre for Policy Research, Syracuse University- Maxwell School of
Citizenship and Public Affairs, Syracuse, New York.
277
Property Rights Alliance, 2011, International Property Rights Index 2011.
PRTM, 2010. Paving the Way for Electric Vehicles.www.prtm.com/uploadedFiles/
Thought_Leadership/Perspectives/PRTM_Paving_the_Way_for_Electric_Vehicles.pdf
Pueyo Velasco, A., 2007. Variables Underpinning Technology Transfers Through the
CDM. Joint Implementation Quarterly, 13(3):5-6.
Qiu, Y. and Anadon, L.D., 2011. The price of wind power in China during its expansion:
Technology adoption, learning-by-doing, economies of scale, and manufacturing
localization. Accepted manuscript. Available online, July 2011
Renewable Energy Magazine, 2011. “Santiago-based company to be first to
manufacture turbine towers in Latin America”
http://www.renewableenergymagazine.com
Reuters, 2007. “Japan looks to batteries to clean up cars” 3rd October 2007.
Ribeiro, S. K. and De Abreu, A. A., 2008, Brazilian transport initiatives with GHG
reductions co-benefit. Climate Policy 8(2): 220-240.
Rindefjäll, T., Lund, E. and Stripple, J., 2010. Wine, fruit and emission reductions:CDM
as development strategy in Chile. The Governance of Clean Development, Working
Paper Series, University of East Anglia.
Saggi, K., 2004. International technology transfer to developing countries.
Commonwealth Secretariat, London.
Schneider, M., Holzer, A. and Hoffman, V.H., 2008. Understanding the CDM's
contribution to technology transfer. Energy Policy 36(8): 2930-2938
Seres, S., 2008. Analysis of Technology Transfer in CDM Projects.Document for the
United Nations Framework Convention for Climate Change.
Seres, S., Haites, E. and Murphy, K., 2009, Analysis of technology transfer in CDM
projects: an update. Energy Policy 37(11): 4919-4926
Seres, S., Haites, E. and Murphy, K., 2010. The contribution of the Clean Development
Mechanism under the Kyoto Protocol to Technology Transfer. Document for the
United Nations Framework Convention for Climate Change.
Servicio de Evaluación Ambiental (SEIA), Gobierno de Chile, 2010. www.sea.gob.cl
SONAMI, 2011. Estadísticas. www.sonami.cl. Sociedad Nacional de Mineria, Santiago
de Chile.
South Centre (SC) and Centre for International Environmental Law (CIEL), 2008. The
technology transfer debate in the UNFCCC: politics, patents and confusion Intellectual
Property Quarterly.
Staley, B. C. and Freeman, C., 2009. Tick Tech Tick Tech: Coming to Agreement on
Technology in the Countdown to Copenhagen, World Resources Institute, Washington
DC.
Stern, N. H., 2007. The economics of climate change: the Stern review. Cambridge
University Press, Cambridge.
278
Stevens, J., 2009. Applied multivariate statistics for the social sciences. 5th ed.
Routledge Academics, London.
Tabachnick, B. G., & Fidell, L. S., 2001. Using Multivariate Statistics (4th ed.). Harper
Collins, New York.
TEC, 2011, Background paper on the modalities of the Technology Executive
Committee. First meeting 26th August 2011, TEC/2011/1/3. Technology Executive
Committee. United Nations Framework Convention for Climate Change, Bonn.
The Economist, 2004, “A productivity primer- The most common way to measure
economic efficiency is not the best”, print edition, Nov 4th 2004.
The Economist, 2010, "The Economist economic dictionary".
The Economist, 2011, “Can poor countries leapfrog manufacturing and grow rich on
services?”, print edition, May 19th 2011.
The Economist, 2011.Regional Integration in Latin America—The Pacific Players go to
Market April 9–15th 2011.
Transparency International, 2011, Corruption Perception Index 2011.
UNEP Risø Centre, 2010, The CDM pipeline. www.cdmpipeline.org.
UNEP, Bloomberg New Energy Finance and SEPI, 2010.: Global Trends in Sustainable
Energy Investment 2010.
Unep-Risoe, 2010, CDM pipeline. www.cdmpipeline.org
UNCTAD, 2001, World Investment Report 2001- Promoting Linkages. United Nations
Conference on Trade and Development, Geneva.
UNCTAD, 2003, World Investment Report 2003- FDI Policies for Development: National
and International Perspectives. United Nations Conference on Trade and
UNCTAD, 2005, World Investment Report 2005. Transnational Corporations and the
Internationalisation of R&D. United Nations Conference on Trade and Development,
Geneva.
UNCTAD, 2008. World Investment Report, 2008. United Nations Conference on Trade
and Development, Geneva.
UNCTAD, 2009 World Investment Report. United Nations Conference on Trade and
UNCTAD, 2010. World Investment Report 2010: Investing in a Low-carbon Economy,
United Nations. United Nations Conference on Trade and Development, Geneva.
UNDP,2009. Handbook for Conducting Technology Needs Assessment for Climate
Change. United Nations Development Programme, Nairobi.
UNFCCC, 2008. Investment and financial flows to address climate change- an update,
FCCC/TP/2008/7. United Nations Framework Convention for Climate Change, Bonn.
UNFCCC, 2009. Recommendations on future financing options for enhancing the
development, deployment, diffusion and transfer of technologies under the
279
Convention. FCCC/SB/2009/2. Subsidiary Body for Sciensce and Technological Advice,
United Nations Framework Convention for Climate Change, Bonn.
UNFCCC, 2009. Performance indicators to monitor and evaluate the effectiveness of
the implementation of the technology transfer framework. FCCC/SB/2009/1. United
Nations Framework Convention for Climate Change, Bonn.
UNFCCC, 2010. Report on the review and assessment of the effectiveness of the
implementation of Article 4, paragraphs 1(c) and 5, of the Convention- Note by the
secretariat. FCCC/SBI/2010/INF.4.. Subsidiary Body for Implementation, United Nations
Framework Convention for Climate Change, Bonn.
UNFCCC, 2010. Report on information required for using the performance indicators to
support the review of the implementation of Article 4, paragraphs 1(c) and 5, of the
Convention. FCCC/SBSTA/2010/INF.3, 25 May 2010. Report by the Subsidiary Body on
Scientific and Technological Advice, United Nations Framework Convention for Climate
Change, Bonn.
UNFCCC, 2010. Decisions adopted by the Conference of the parties on the Outcome of
the work of the Ad Hoc Working Group on long-term Cooperative Action.
FCCC/CP/2010/7/Add.1. United Nations Framework Convention on Climate Change,
Bonn.
UNIDO, 2009. Changing Patterns in Industrial Performance A UNIDO Competitive
Industrial Performance Perspective. Implications for Industrial Development. Research
and Statistics Branch Working Paper 05/2009. United Nations Industrial Development
Organisation, Vienna.
Universidad de Chile and Universidad Técnica Federico Santa María, 2008. Aporte
potencial de Energías Renovables No Convencionales y Eficiencia Energética a la Matriz
Eléctrica, 2008 – 2025. Programa de Estudios e Investigaciones en Energía del Instituto
de Asuntos Públicos de la Universidad de Chile and Núcleo Milenio de Electrónica
Industrial y Mecatrónica, Centro de Innovación en Energía de la Universidad Técnica
Federico Santa María.
Universidad de Chile, 2010. “Proyecto Centro de Innovación del Litio” presentation by
the Energy Centre, Universidad de Chile, Santiago de Chile.
EIA, 2011. International Energy Statistics. http://tonto.eia.doe.gov. US Energy
Information Administration, Washington DC.
Verdolini, E. and Galeotti, G., 2011- At home and abroad: An empirical analysis of
innovation and diffusion in energy technologies. Journal of Environmental Economics
and Management, 61:119–134
Wang, B., 2010. Can CDM bring technology transfer to China?-An empirical study of
technology transfer in China's CDM projects. Energy Policy 38(5): 2572-2585.
Wara, M., 2009. The Clean Development Mechanism’s early performance. In:
Schneider, S. Rosencrantz, A., Manstrandrea, M. (Eds), Climate Change Science and
Policy. Island press, Washington.
Wei, L., 1995. International technology transfer and development of technological
capabilities: A theoretical framework. Technology in Society 17(1): 103-120.
280
Woodside, A.G. and E.J. Wilson, 2003. Case study research methods for theory
building. Journal of Business and Industrial Marketing 18 (6/7):493-508.
Wooldridge, J.F., 2006. Introductory Econometrics, A modern approach. 3rd Edition.
Thomson Higher Education.
World Bank, 2009. World Development Report 2010 - Development and Climate
Change. The World Bank, Washington DC.
World Bank, 2011. Doing Business 2012: Doing Business in a More Transparent World.
The World Bank, Washington DC.
WBCSD, 2010. Enabling frameworks for technology diffusion: a business perspective.
World Business Council for Sustainable Development, Geneva.
WBCSD, Climate Focus and Ecofys, 2010. Private Sector and the UNFCCC: Options for
Institutional Engagement. Competitiveness and Innovation Framework Programme.
Report for the European Commission, Brussels.
World Economic Forum, 2011. Global Enabling Trade Report 2010. Geneva.
http://gcr.weforum.org/getr2010
WRI, 2009, Summary of UNFCCC submissions. World Resources Institute, Washington
DC.
Xu, B. and Wang, J. M., 1999, Capital goods trade and R&D spillovers in the OECD.
Canadian Journal of Economics-Revue Canadienne Economique 32(5): 1258-1274.
Xu, B., 2000, Multinational enterprises, technology diffusion, and host country
productivity growth. Journal of Development Economics 62(2): 477-493.
ZAP, 2008, “ZAP: Electric Car Incentives Approved in Uruguay”, 28 November 2008,
www.zapworld.com
Zhang, X. L., Chang, S. Y., Huo, M. L. and Wang, R. S., 2009, China's wind industry:
policy lessons for domestic government interventions and international Support.
Climate
281
8 Annexes
282
8.1 Chapter 2: Tables
TABLE 49- CHANNELS FOR PRIVATE SECTOR CLIMATE CHANGE TECHNOLOGY TRANSFER
TT mechanisms Factors contributing to the choice of Integration Interaction and

mechanism internal effort
Trade • Low import duties Low Low-
• Product compatibility Conventional
• Standards and certification
• Availability of after-sales service and
training
• Distributor capabilities
• Degree of system integration required
before use by final user
• Insurance and product liabilities
Turnkey contracts • Low domestic technological capabilities High Low-
• International competitive bidding Conventional
• Import duties
• Buyer training
• Corruption
Wholly owned • Acceptable financial risks High Low-
subsidiaries • Liberal foreign investment policies of Conventional
host government
• Large expected size of domestic market
• High import duties
• High costs of innovation and weak IPR
protection encourage FDI to protect
intangible assets (Maskus 2000)
(UNCTAD 2003)
Joint ventures in • Acceptable financial risks Medium Medium-
the local market • Strong IPR regimes (Maskus 2000) Conventional
• Large expected size of domestic market
• Need to adapt products to local
markets
• Large skills base of host country
• Strong local firms
• Export-oriented economy
• Localisation requirements of host
government
• Liberal FDI policies
Licensing • Strong IPR regimes (Maskus 2000) Low Medium-
agreements • Future domestic market and strategic Conventional
interests of MNC
• Acceptable financial risk
Twinning, • Ability to attend conferences Low High-
conferences, • Availability of counterpart resources Non
symposia and • Access to information and conventional
other person-to- communication means
person pathways • Intellectual property protection
Source: (IPCC, 2000; Lema and Lema, 2011; Maskus, 2000; UNCTAD, 2003; author contributions)
283
Table 50- Knowledge transfer metrics in Hoi et al. (2008)
Knowledge transfer Measures of quantity Measures of quality

activities
Networks • Number of people met at events • % of events held which led to
which led to other knowledge other knowledge transfer activities
transfer activities
Continuing • Income from courses • % of repeat business
Professional • Number of courses held • Customer feedback
Development • Number of people and companies
that attend
Consultancy • Number and value/income of • % of repeat business
contracts • Customer feedback
• % income relative to total research • Quality of client company
income • Importance of client relative to
• Market share their company
• Number of client companies
• Length of client relationship
Collaborative • Number and value/income of • % of repeat business
research contracts • Customer feedback
• % income relative to total research • Number of products successfully
income created from the research
• Market share
Contract research • Number and value/income of • % of repeat business
contracts • Customer feedback
• % income relative to total research • Number of products successfully
income created from the research
• Market share
Licensing • Number of licences • Customer feedback
• Income generated from licences • Quality of licensee company
• Number of products that arose from • % of licences generating income
licences
Spin-outs • Number of spin-outs formed • Survival rate
• Revenues generated • Quality of investors
• External investments raised • Investor/customer satisfaction
• Market value at exit (IPO or trade • Growth rate
sale)
Teaching • Graduation rate of students • Student satisfaction (after
• Rate at which students get hired (in employment)
industry) • Employer satisfaction of student
Other measures • Physical migration of students to •
industry
• Publications as a measure of
research output
284
TABLE 51- KNOWLEDGE TRANSFER METRICS IN JENSEN ET AL (2009)
Knowledge Measures of activity Measures of effect

transfer
activities
Networks • Number of attendances/ • Number of collaborative research
presentations at conference/ projects as a result of knowledge
seminar with industry participants exchange or networking activities
• Number of PhD exchanges with • Number of contract research projects as
industry a result of knowledge exchange or
networking activities
Continuing • Number of courses held • Customer feedback
Professional • Number of attendees
Development • Number of university-industry lab
researcher exchanges
• Number of other scientific and
research training schemes for
industry
Consultancy • Number of consultancy contracts • Value of consultancy contracts
• Number of collaborative research
projects generated by consultancies
Collaborative • Number and value of collaborative • Number of
research research agreements products/processes/publications
• Number and value of joint successfully created from collaborative
ventures research
Contract • Number of contract research • Value of contract research projects
research projects • Number of contract research projects
• Length of client relationship which led to other flow-on knowledge
transfer activities such as collaborative
research, licensing, industry sponsored
conferences, etc.
Licensing • Number of invention disclosures • Number of licences
• Number of complete standard • Income stream from licences
patent applications • Long term relationships created
• Number of patents granted following licensing
• Number of plant variety rights
• Value of copyright licences
Spin-outs • Number of spin-outs formed • Survival rate
• Revenues generated • Growth rate
• External investments raised
• Market value at flotation (IPO or
trade sale)
• Market value at exit
Teaching • Graduation rate of students • Student satisfaction (after employment)
• Rate at which students get hired • Employer satisfaction with graduate
(in industry)
Other measures • Number of research student • Research publications
placements in industry • Citation received
• Number of industry funded • Joint University-Industry publications
postgraduate positions/scholarhips and inventions
• Number of staff working on
commercialisation activity in
dedicated and support roles
• Provision of training in research
commercialisation
285
TABLE 52-UNFCCC PERFORMANCE INDICATORS ON TECHNOLOGY TRANSFER
Performance indicator Relevance Data Input– Value

availability Output- or
Effect Count
Technology needs and needs assessments
Amount of financial resources provided for the High Yes Input Value
TNA process
Number of programmes/projects for capacity- Medium Yes Input Count
building on TNAs in non-Annex I parties per
Annex II party, per IGO and in total
Number of targeted non-Annex I parties per High Yes Input Count
Annex II Party and per IGO
Number of published TNAs completed or High Yes Input Count
updated by non-Annex I parties
Synthesis reports made available by the High Yes Input Count
secretariat and discussed at the subsidiary
bodies
Number of technologies from TNAs High Yes but Output Count
implememented by non-Annex I parties with gaps
Technology information
Number of training programmes and workshops Medium Yes but Input Count
for building capacity with gaps
Number of national communications with High Yes Input Count
information on technology transfer activities
Number of reports with information on High Yes Input Count
maintaining, updating and developing TT:CLEAR,
addressing gaps and user needs
Number of technology information centres and High Yes Input Count
networks connected to TT:CLEAR
Number of users of TT:CLEAR from developing High Yes but Output Count
countries with gaps
Enabling environments
Performance on each of the six World Bank´s Not No
governance indicators relevant
Total volume (number and dollar value) of joint High Yes but Input Value
R&D opportunities for environmentally sound with gaps and
technologies (EST) by governments Count
Presence of clear policy guidelines to the Low Yes but Input Count
recipients of public funding on how to move with gaps
from R&D to commercialisation
Number of bilateral or multilateral programmes Medium Yes but Input Count
that have helped developing countries to with gaps
develop and implement regulations promoting
the use, transfer of and access to ESTs.
Presence of tax preferences and incentives on Medium Yes but Explanatory
imports/exports for ESTs with gaps
Volume (in US dollars) of export credits to High Yes but Explanatory
encourage the transfer of ESTs with gaps
Presence of EST transfer in national sustainable Medium Yes Input Count
development strategies
Rating of investment climate according to World High Yes Explanatory
Bank´s World Business Environment Survey
Percentage of government procurement budget High Yes but Explanatory
spent on ESTs with gaps
Degree of disclosure and transparency regarding Medium Yes but Explanatory
286
Effect Count
the approval processes with gaps
Counting and rating of the studies carried out Low Yes but Input Count
that explore barriers, good practices and with gaps
recommendations for developing enhanced
enabling environments
Percentage of participation of developing Medium Yes Input Count
country parties in partnerships
Capacity-building
Amount of financial resources provided for the High No Input Value
Capacity Building for development and transfer
of technology
Reported needs and agreed priorities for Medium Yes with Input Count
capacity-building for development and transfer gaps
of technology
Number of participants/experts in training High Yes with Input Count
programmes on the development and transfer of gaps
technologies, in particular on the development
of standards and regulations
Number of new and existing national and High Yes with Input Count
regional institutions operating as centres of gaps
excellence on the development and transfer of
technologies
Mechanisms for technology transfer
Number and volume of reported innovative High Yes with Input Count
public-private financing mechanisms and gaps
instruments in total and by Party
Report on the possible ways to enhance High Yes Input Count
cooperation between the UNFCCC and other
multilateral environmental agreements (MEAs)
Report on the references to objectives of other High Yes with Input Count
MEAs in national communications gaps
Number of reported barriers and good High Yes with Explanatory Count
experiences in total and by non-Annex I Party gaps
Report with guidance for technology needs High Yes Input Count
assessment reporting on joint R&D needs
Finance
Total annual global investment and financial High Yes but Output Value
flows in climate change mitigation technologies with gaps
Total annual global investment and financial High Yes but Output Value
flows in climate change technologies for with gaps
adaptation
Total annual investment and financial flows in Medium Yes but Output Value
climate change technologies-Convention with gaps
financial mechanisms
climate change technologies-Kyoto flexibility with gaps
mechanisms
climate change technologies-bilateral sources with gaps
climate change technologies-national sources with gaps
climate change technologies-multilateral sources with gaps
287
Effect Count
climate change technologies-private sources with gaps
288
8.1 Chapter 3: Tables
TABLE 53- UNCTAD INNOVATION CAPABILITY INDEX
Source: UNCTAD, 2005, World Investment Report
289
8.2 Chapter 4.
8.2.1 Annex 1: Questionnaires for case study research
290
291
292
293
294
295
296
8.2.2 Annex 2: Technology background
The selected case studies describe TT processes for six different kinds of technologies. Three
case studies refer to wind technologies, three to different types of biomass technology, the
remaining cases analyse transfers of Concentrated Solar Power (CSP), hydro-electricity,
Lithium-ion batteries for Electric Vehicles (EV) and landfillsbiogas. This chapter provides
some background about the status of these technologies in terms of their cost, global
demand and supply, stage of technological maturity and main technological gaps.
8.2.2.1 Cost
To compare the electricity production costs of the generation technologies included in the
case studies, we have used the SETIS Energy Cost Calculator51. The Strategic Energy
Technologies Information System (SETIS) is a tool for planning, communicating and pooling
resources for international cooperation in the field of energy technologies relevant to the
European Commission´s Strategic Energy Technology Plan (SET-Plan). In our analysis we
consider both the renewable energies covered by our case studies and fossil fuel
technologies, to compare the cost of incumbent and alternative energy generation
technologies. Data is based in EU activities. Costs in Chile could be higher than in the EU for
technologies with no previous implementation experience. In some other technologies, like
biomass or hydro, costs could also be lower due to the abundant resources in Chile.
Although no directly applicable to Chile, these figures are indicative of the expected costs of
clean energy generation, in three scenarios: 2007, 2020 and 2030.
The following technology types are included in the costs estimation:
• Natural Gas: GTCC - Gas Turbine Combined Cycle w/o CCS
• Natural Gas: GT-L - Open Cycle Gas Turbine - Large
• Coal: CFBC - Circulating Fluidised Bed Combustion w/o CCS
• Oil: Internal Combustion Engine
• Biomass Combustion large-scale (> 10 MWe)
• Biomass Combustion small-scale (< 10 MWe)
• Biogas - Landfill (with a gas engine)
• On-shore Wind (in a farm configuration)
• Hydropower Small-scale II(<= 10 MWe)
51
https://odin.jrc.ec.europa.eu/SETIS/SETIS1.html#
297
• Photovoltaics (based on crystalline silicon panels)
• Concentrated Solar thermal Power (CSTP) - parabolic trough collector with storage
and natural gas backup systems
SETIS calculations are based in a set of assumptions:
Assumptions:
• Inflation rate: 0%
• Nominal discount rate: 10%
• Interest during construction: 10%
• Sinking fund interest rate: 2.5%
• OM escalation rate: 0%
• OM learning effect: -0.5%
• Cost of CO2 allowances is not included
Technology assumptions can be consulted in the SETIS calculator: Net capacity, net
efficiency, load factor, technical lifetime, construction time, capital costs, operational
parameters, fuel price and learning rate for the capital costs.
Cost estimations, illustrated in Figure 69, show that onshore wind generation, landfill biogas
and large biomass combustion are already competitive with diesel based electricity
generation. However, they are more expensive than coal and natural gas fuelled generation.
All the other types of renewable energy generation are still more expensive than any fossil
fuelled alternative. Solar PV and CSP show particularly high costs as compared to fossil fuel
incumbents. In 2020 and 2030, renewable energies increase their competitiveness as a
result of higher fossil fuel costs and technological learning rates. Onshore wind becomes
cheaper than natural gas and diesel based generation and reaches similar costs to coal
based generation in 2030. By 2030 all the considered options of renewable generation
except small-scale biomass and solar PV and CSP, can compete with diesel based
generation. Landfill gas can also compete with natural gas based generation. Coal remains
cheaper than all the other options when the cost of CO2 is not internalised. Data show that
a carbon price is necessary for solar technologies, small-scale hydro and biomass to take off.
However, onshore wind, biogas from landfills and large-scale biomass projects could
compete with fossil fuel technologies in the long term even without carbon pricing policies.
298
FIGURE 69- ELECTRICITY PRODUCTION COSTS IN 2007, 2020 AND 2030
Electricity production costs 2007 (€/MWh)

700.0
600.0
500.0
400.0
300.0
200.0
100.0
0.0
Biogas Biomass: Biomass: Coal CFBC Hydro: Nat Gas Nat Gas Oil IC Solar: CSP Solar: PV Wind:
Landfill Solid Solid Small II GT-L GTCC Onshore
(large) (small)
Fixed production costs [€/MWhel] Variable production costs excluding carbon [€/MWhel]
Cost of electricity excluding carbon exp [€/MWhel]

350
300
250
200
150
100
50
0
(large) (small)
299
300
250
200
150
100
50
0
(large) (small)
Source: Energy Cost Calculator, SETIS. https://odin.jrc.ec.europa.eu/SETIS/SETIS1.html#
To allow comparability with Chile’s electricity prices data, the values in the graphs above
have been converted to US$ using the latest exchange rate of the European Central Bank at
17th May 2011, at 1.4171 USD/EUR.
TABLE 54- COST OF ELECTRICITY PRODUCTION WITH DIFFERENT TECHNOLOGIES
Cost of Cost of Cost of

electricity electricity electricity
Plant type
[US$/MWh] [US$/MWh] [US$/MWh]
2007 2020 2030
Biogas Landfill 126.9 119.5 114.3
Biomass: Solid (large) 127.5 136.5 141.1
Biomass: Solid (small) 236.4 259.2 271.4
Coal CFBC 74.3 83.2 83.9
Hydro: Small II 163.0 161.4 159.3
Nat Gas GT-L 98.6 143.7 151.4
Nat Gas GTCC 79.2 110.0 118.6
Oil IC 154.7 206.3 232.6
Solar: CSP 480.0 351.4 356.0
Solar: PV 863.0 447.2 286.4
Wind: Onshore 119.8 96.8 90.6
Source: Energy Cost Calculator, SETIS. https://odin.jrc.ec.europa.eu/SETIS/SETIS1.html# and ECB for exchange
rates
As regards lithium-ion batteries for EVs, their high cost is currently a critical concern for the
massive deployment of EVs. In 2009 they were valued at US$650/kWh and are expected to
go down to US$325/kWh by 2020 as a result of economies of scale (Lowe et al, 2010). The
cost breakdown of a 25-kWh battery pack, using 180-Wh Nickel/Manganese/Cobalt cells,
valued at 16,598 US$, is presented in Table 55. Cells account for 64% of the cost, electronics
for 32% and the final pack assembly and warranty for the remaining 4%.
300
TABLE 55- LITHIUM-ION BATTERY COST BREAKDOWN
Components $/battery % cost

Cell components Cathode 1663 14%
Anode 477 4%
Electrolyte 447 4%
Copper foil 184 2%
Separator 608 5%
Can header and terminals 1050 9%
Other materials 375 3%
Total material 4804 41%
Labour Labor for cell manufacturing 2586 22%
Total cell 7390 64%
Electronics Mechanical components 2053 18%
Electrical components 299 3%
Electronics (battery mng system) 1381 12%
Total electronics 3733 32%
Packs Labor for pack manufacturing 268 2%
Total packs 11391 98%
Warranty 228 2%
Total cost of batteries 11619 100%
Gross profit (30%) 4979
Total price of batteries 16598
Source: Lowe et al. (2010)
8.2.2.2 Global demand
According to the Renewables Global Status Report 2010 (REN21, 2010), renewable energies
supplied 19% of global final energy consumption in 2008 including mature sources like
traditional biomass and large hydropower. Traditional biomass, used primarily for cooking
and heating accounts for 13% of the global consumption and is declining. Large hydropower
represents 3.2% and the new renewables (small hydro, modern biomass, wind, solar,
geothermal and biofuels) represent 2.6% and are growing rapidly in developed countries
and in some developing countries
FIGURE 70 - RENEWABLE ENERGY SHARE OF GLOBAL FINAL ENERGY CONSUMPTION, 2008
Source: REN21, 2010
Renewable electricity reached 1,230 GW in 2009, increasing by 7 percent from 2008. This
represents 25% of global generation capacity and 18% of global electricity production.
301
Renewable capacity is only 305GW when large hydropower is not included. Wind has
experienced the largest growth in renewable capacity in the last years, with some additional
35 GW in 2009 followed by hydro, with 30 additional GW. Solar PV added more than 7 GW
in 2009.
FIGURE 71 - SHARE OF GLOBAL ELECTRICITY FROM RENEWABLE ENERGY, 2008
Source: REN21, 2010
Wind
Global wind installed capacity reached 159 GW in 2009, with cumulative capacity doubling
in less than three years. China was the top installer in 2009, representing more than one
third of the world market in this year, with a cumulative capacity of 25.8 GW. The United
States is the world leader in installed capacity with a total of 35 GW and Germany leads in
Europe with a capacity just under 25.8 GW, followed by Spain, with under 20GW. Some of
the most vibrant wind power markets are in Latin America and Africa, with significant
growth rates but relatively low capacity levels. Hundreds of additional MW are under
construction in Argentina, Peru and Uruguay and China plans 120 GW of new wind capacity.
Several countries now meet a significant share of their electricity generation with wind, such
as Denmark (20 percent), Spain (14.3 percent), Portugal (11.4 percent), Ireland (9.3 percent)
and Germany (6.5 percent).
FIGURE 72- WIND POWER CAPACITY, TOP 10 COUNTRIES
Source: REN21, 2010
302
Biomass and Biogas
Biomass can derive from forestry, agriculture or municipal waste as well as from crops
grown specifically as fuels. Biomass is available in solid, liquid and gaseous (biogas) forms. It
is commonly used to generate both heat and power and can also be converted to biofuels
for transport. A diversity of technologies exist to harness the energy potential of the
different types of biomass.
Global biomass power capacity was estimated at 54 GW in 2009. The United States are the
largest biomass power producers with around 8.5 GW of installed capacity, followed by
Japan and Germany. Many US coal and gas fired plants are being partially or totally
converted for co-firing biomass with fossil fuels. Co-firing is also popular in Germany and the
United Kingdom. Europe has an estimated 7 GW of solid biomass power generation
capacity, with the largest scale and number of plants in the forested Scandinavia and
Germany. Germany, Finland and Sweden represented over half of the total electricity from
solid biomass in Europe in 2008.
Biomass power generation has also grown significantly in developing countries including
Brazil, Costa Rica, India, Mexico, Tanzania, Thailand and Uruguay. China has a capacity of 3.2
GW and plans to install up to 30 GW by 2020. India has over 835 MW of agricultural waste
biomass capacity and more than 1.5GW of bagasse cogeneration plants. Brazil has over
4.8GW of biomass cogeneration plants at sugar mills, which feed into the grid almost half of
their electricity generation.
Biogas is used for electricity generation mainly in OECD countries, with Germany as the
largest producer with a capacity of 1.7 GW, followed by the United States. Some Asian
countries like Thailand and Malaysia have significantly increased their biogas power
capacities.
CSP
Investments in commercial-scale CSP started in 2005 after a period of stagnation. Global
capacity increased by 70% between 2005 and 2009 to reach 610 MW at the end of 2009. All
the capacity is concentrated in the United States, with 65% of the installations, and Spain,
which has driven most of the growth in the last years. Worldwide, 2.4GW of capacity were
being built or were under contract by early 2010, with Spain accounting for the vast majority
of this additional capacity. CSP is also entering developing country markets, with plants
planned or under construction in Abu Dhabi (100 MW), Algeria, Egypt and Morocco (20 MW
each). The Moroccan government has announced plans to build 2 GW of CSP by 2020 and a
similar capacity has been agreed in China to be ready by 2020. Also, in late 2009, financing
was approved for 1GW of capacity and transmission infrastructure in the North of Africa by
2020
Most CSP plants in operation use parabolic trough technology, but nearly half of the
capacity in construction or under contract will use linear Fresnel, dish engine or power-
tower technology. Storage technologies allowing electricity generation when the sun is not
shining are also advancing. The Spanish plants Andasol I and II began operations in 2009
with up to seven hours of thermal energy storage.
303
Hydropower
Global hydropower capacity was estimated at 980 GW by the end of 2009, of which 60 GW
were small hydro. It supplies 15 percent of global electricity needs. The largest market for
hydropower technologies is China, with 197 GW by the end of 2009, followed by the United
States, with 81 GW. Brazil had approximately 76 GW of capacity and Canada some more
than 75 GW. In un-electrified rural areas of developing countries small hydro is often used
to replace diesel generators. Significant increases in hydropower capacity were expected for
2011, mostly concentrated in Brazil, China, India, Malaysia, Turkey and Vietnam. Most new
capacity will come from large-scale projects, but a dramatic increase is also expected in
small-scale projects.
Solar PV
Solar PV has experienced an exponential growth between 2004 and 2009 and is now
present in more than 100 countries. Total installed capacity in the world is estimated at
21GW, with 7GW added only in 2009. The solar PV industry heavily depends on the German
market that holds 47% of the existing global capacity. The Spanish market for solar PV
suffered a precipitous decline in 2008.
FIGURE 73- SOLAR PV EXISTING CAPACITY, TOP SIX COUNTRIES, 2009
Source: REN21, 2010
Cumulative global PV installations were in 2009 nearly six times what they were at the end
of 2004 and analysts expect higher growth in the next four to five years. In 2009 thin film
represented 19% of the market and modules 22% of the market. There is a global trend
toward large-scale (greater than 200kW) solar PV plants, which total 5.8GW of capacity.
Large-scale plants are mainly operated in Spain, Germany and the United States but
increasingly introduced in developing countries, like the Philippines, Rwanda and Saudi
Arabia.
Lithium-ion batteries for EV
Demand for EV batteries is currently small but is expected to grow very quickly. The global
market for EV batteries in 2009 was estimated at US$1.3 billion in 2009 and is expected to
reach US$25 million by 2020 (Lowe, 2010). This would represent approximately 12% of the
market for EV in that date.
304
A global forecast by international management consultants PRTM estimates that future
sales of electric vehicles will reach 40 million vehicles by 2020, or roughly half of the global
vehicles market. Several types of batteries will provide energy for EV, with a growing
participation of lithium-ion batteries. A more conservative forecast by Total Battery
Consulting (TBC) estimates that future sales will be limited to 200,000 vehicles in 2015 and 1
million in 2020. An outlook for lithium-ion battery demand by Deutsche Bank (2009)
estimates 4 million vehicles sold in 2015, with roughly half of them needing lithium-ion
batteries. By 2017 EV demand would increase to over 7 million vehicles, with over 5 million
using lithium-ion batteries. The IEA (2011) forecasts EV sales over 5 million by 2020,
reaching 10% of vehicle stocks by 2030 and 50% by 2050.
Some countries have announced EV penetration targets. For example, Japan announced
that by 2020, EV sales would represent 15-20% of total Light Duty Vehicles (LDV) sales. The
Japanese Government announced a per-vehicle subsidy to cover half of the price difference
with conventional vehicles. China has also announced a stock target of 5 million EVs by
2020, which would be supported with subsidies per vehicle. Denmark exempts EVs from
vehicle purchase tasks and annual registration fees and provides subsidies for businesses
and municipal fleet operators. Other countries announcing stock targets are France (2
million by 2020), Spain (250,000 by 2014 and 2.5 million by 2020), Sweden (600,000 by
2020) and the United States (1 million by 2015).
8.2.2.3 Global supply
Wind
In 2009 European and Chinese firms dominated the wind turbine manufacturing sector, with
Danish Vestas in the first position, followed by GE Wind of the United States. China has
emerged as a global manufacturer of wind turbines, with three firms Sinovel, Goldwind and
Dongfang, among the top 10 manufacturers in 2003 (REN21, 2010). The growth of a local
industry has built on a strong domestic market for wind power but now Chinese companies
are exporting significant amounts of technology abroad and are planning to open
manufacturing plants abroad, particularly in the US. The EU still has the largest number of
leading companies among the top 10 and is driving the development of offshore wind
turbine technologies. The average utility scale wind turbine installed in 2009 was 1.6 MW
and the largest turbine installed was the 7.5MW Enercon gearless turbine.
305
FIGURE 74- MARKET SHARES OF TOP 10 WIND TURBINE MANUFACTURERS, 2009
Source: REN21, 2010
Wind turbines consist of several components and require construction and transportation
services, as detailed below:
• Rotors- They are typically made of three blades made of laminated materials with
high strength to weight ratios, such as composites, carbon fibre or fiberglass. The
world´s largest wind turbine blade supplier is LM Glasfiber. Usually, turbine
manufacturers also produce blades. The hub is one of the turbine´s heaviest
components and is usually made of cast iron.
• Generator. It is the heart of the wind turbine and is found inside the nacelle. It must
be robust enough to handle the frequent changes in the wind speed. The wind
turbine is fitted with a number of sensors to read the wind speed and direction, the
rotor speed, the temperature of the lubricants, the electrical power generation and
other variables.
• Tower. A high tower supports the nacelle and generator. They are typically made of
several steel sections and are about 70 meters tall.
• Construction. Civil engineering works are required to fit the tower to the ground and
to connect each turbine to the power grid, by constructing access roads, laying
electrical cables and installing electrical substations.
• Transportation. Wind turbine components are sensitive, valuable and heavy. They
can require high transportation costs. For this reason manufacturing plans often
locate close to high demand hubs.
Biomass power and heat
Solid biomass technology is quite mature, with forestry and paper industries accumulating a
long experience in using their residues as a source of energy.
In Europe, the solid the solid biomass industry provided 5.6TWh of electricity in 2008. The
decline of shipping costs in 2009 caused a significant growth in the wood pellet market in
Europe. Shipping costs have a strong influence in this market as they represent around half
306
of the total cost of pellets. The low cost of shipping and the increased demand of co-firing
for coal plants has caused increased the number of solid biomass generation projects under
construction.
The interest of co-firing has grown considerably at international level. The key challenges for
these projects are sourcing biomass, transportation, storage and handling of feedstock.
Solar PV
The solar PV value chain includes the industrial process of production of PV modules, the
assembly of integrated PV systems, their installation and the generation of electricity. The
industrial production process includes four stages: the purification of silica found in quartz
salts, ingot and wafer manufacturing, cell production, and PV modules assembly.
FIGURE 75- PV SUPPLY CHAIN
Source: de la Tour et al, 2010
Module prices have experienced a steep decline, with 50-60percent price reductions
between 2008 and 2009. Competition is particularly intense in the downstream segments of
PV cells production and modules assembly, in which profit margins are low. China is
particularly strong in these segments, in which technological barriers of entry are relatively
low (de la Tour et al, 2010).
The top 15 solar cell manufacturers produced 65percent of the 10.7GW of cells
manufactured in 2009. China and Taiwan produced nearly half of the total, followed by
Europe (18%), Japan (14%) and the United States (6%) (REN21, 2010).
307
FIGURE 76- MARKET SHARES OF TOP 15 SOLAR PV MANUFACTURERS, 2009
Source: REN21, 2010
CSP
CSP plants produce electricity by converting concentrated solar irradiation into energy. Solar
irradiation is concentrated and converted into usable thermal energy by concentrating solar
collectors. There are two main types of collectors: line-focusing systems such as the
parabolic trough collector and linear Fresnel collector, and point-focussing systems such as
solar towers or solar dishes. The conversion of heat into electricity is realized by a
conventional steam turbine.
The CSP core value chain consists of six main phases: project development, materials,
components, plant engineering and construction, operation and distribution.
FIGURE 77- BASIC STRUCTURE OF THE CSP CORE VALUE CHAIN
Source: Ernst&Young and Fraunhoffer, 2011
CSP manufacturers and developers have been predominantly focused on opportunities in

the United States and Spain in 2009. In the United States, renewable portfolio standards
have spurred new project development opportunities (REN21, 2010). Some of the leading
308
firms in CSP include Brightsource, eSolar, Siemens, Schott, SolarMilenium, Abengoa Solar,
Infinity and Acciona. Molten salt thermal storage technologies, in which Chile is involved,
are provided by several actors in collaboration. The engineering company Sener is currently
the most experienced player in thermal storage for CSP plants. The components used for
solar thermal storage are standard in chemical and energy plants, except in the molten salts,
which are mainly provided by SQM in Chile. Recent salt price increases might have been
caused by increasing demand by the CSP industry. However, a German company is working
on a prototype to use concrete as an alternative storage material, which would damage
Chilean opportunities as a supplier of raw materials for CSP storage (Ernst&Young and
Fraunhoffer, 2011).
Hydro
Hydropower is the most mature of the renewables industries, given its long history and
large-scale. In developing countries such as China, Brazil, Ethiopia, India, Malaysia, Turkey,
Vietnam and Chile, utilities and developers are planning large new hydropower projects.
Leading equipment manufacturers include Voith, Alstom, Andritz, Impsa, BHEL, Hitachi and
Makamidi.
Lithium-ion batteries for EV
Lithium-ion batteries are the most suitable existing technology for EV because of their low
weight and high energy output per unit of battery mass. Lithium-ion batteries for EV are a
new market which only began in 2009. Currently the industry is mostly supplying the
electronics market.
The lithium-ion battery value chain consists mainly of four activities:
• Supply of key materials, including cathode precursors (lithium, cobalt, nickel,
manganese), anodes (graphite precursor or natural graphite) and electrolyte
materials (organic solution, lithium salt and polymer precursor for polymer batteries)
• Manufacture of cell components and electronics. Electronics include mechanical,
electrical and electronic components.
• Manufacture of integrated systems, including cell manufacture and final pack
assembly.
• Manufacture of vehicles using battery systems.
Error! Reference source not found. shows a detailed value chain and the major global
players for each stage. Global players with US manufacturing locations are indicated in black
font. The origin of the companies is indicated in brackets otherwise. As shown in the figure,
Chile is only involved at the lower end of the value chain as a provider of lithium
compounds.
309
FIGURE 78- MAJOR PLAYERS IN THE LITHIUM-ION BATTERY VALUE CHAIN
Source: Lowe et al, 2010
China, Japan, South Korea, France and the United States are currently the major lithium-ion
battery manufacturers for hybrid and EV. Error! Reference source not found. shows that
Japan is the global leader, with 57% market share, followed by South Korea (17%) and China
(13%), which are quickly increasing their shares (Lowe et al, 2010). The US only have one
company near the top.
FIGURE 79- GLOBAL LITHIUM-ION BATTERY MARKET SHARE, BY COUNTRY AND FIRM
It is important that batteries manufacture takes place near auto manufacturing, because
vehicle and batteries can share R&D and manufacturing facilities and most importantly,
310
automakers need agile and reliable suppliers nearby. For this reason, battery cell and pack
companies and automotive firms often enter in partnerships or joint ventures to develop
battery technology for vehicles. These alliances are particularly strong in Japan
8.2.2.4 Global innovation
Patent data for selected Clean Energy Technologies reveal a period of stagnation until the
mid-1990s followed by a fast rate of increase. Patenting activity in solar PV, wind and carbon
capture has been especially intense in the last ten years.
FIGURE 80- PATENTING GROWTH RATES FOR SELECTED CLEAN ENERGY TECHNOLOGIES
Source: UNEP, EPO and ICTSD, 2010
Japan shows the highest patenting activity in the technologies selected by a UNEP, EPO and
ICTSD (2010) study, being followed by the US and Germany. The Republic of Korea is in the
fourth position, showing a considerable increase in patenting. The UK and France complete
the top six patenting countries. Patenting of clean energy technologies is therefore clearly
dominated by OECD countries.
311
FIGURE 81- COUNTRIES LEADING PATENTING ACTIVITY IN CLEAN ENERGY TECHNOLOGIES
Source: UNEP, EPO and ICTSD, 2010
Japan also leads R&D and patenting related to EV and in particular to lithium-ion batteries, a
technology not covered by the UNEP, EPO and ICTSD (2010) study.
FIGURE 82- PATENTS AND RESEARCH PAPERS RELATED TO LITHIUM-ION BATTERIES 1998-2007, BY COUNTRY
China is the most active emerging economy in terms of clean energy technology patenting,
predominantly in the area of solar PV. However, in relative terms, the rates of patenting per
312
company are very low both for solar PV and wind technologies. This suggests that, while
Chinese companies are leading manufacturers, they do not hold a significant amount of
technology. The same can be said about India, with very low patent rates in solar and wind
technology. The main patenting areas for Brazil are hydro and biofuels, but patenting rates
are very low compared to China or India. This shows that leading ethanol producers in Brazil
are more focused on the production process than on developing new technologies for
biofuels.
Table 56 shows the main gaps and priorities in RD&D of the technologies covered by our
case studies, according to the IEA (2010). The larger gap is found in clean vehicle, where
among other things, battery cost reduction is required to make EV able to compete with
Internal Combustion Engine (ICE) vehicles.
TABLE 56- ESTIMATED GLOBAL GAPS IN PUBLIC RD&D SPENDING AND MAIN PRIORITIES FOR TECHNOLOGIES
COVERED IN CASE STUDIES
Technology Gap (billion Main RD&D priorities

USD)
• Vehicle efficiency RD&D, including lightweight materials, advanced thermal

Clean vehicles 21-43
combustion engine (ICE)-based power trains and on-board diagnostics
• Efficiency improvements through new Nano-materials
• Basic research
• Battery cost reduction to achieve EV break-even cost with ICEs
• Improving production processes
• Cost-effective conversion of cellulose-rich biomass to usable energy
Bioenergy 1-2
• Sustainable bioenergy cycles
• Methodologies and standards for long-term sustainable performance
• Stronger, lighter materials to enable larger rotors and improved tensile
Wind 2-3
strength
• Design of dedicated offshore wind turbines
• Advanced sub-surface structures
• Use of superconductor wires to reduce transmission losses
• Development of advanced wind forecasting models
• Improve efficiency for crystalline silicon PV technologies and automation of
Solar energy 1-3
manufacturing to reduce costs as companies scale up production.
• Thin film PV: increased improving device structure, large area deposition
techniques, interconnection and manufacturing
• System level: improve the product requirements for building integration and
minimize the environmental impacts of very large-scale PV deployment
• CSP: include increasing system efficiency through higher process temperatures,
reducing material consumption and automating operations
• Solar heating and cooling: development of compact seasonal heat storage,
innovation in collector design, heat storage, cooling devices and advanced
materials
Source: IEA, 2010
313
8.3 Chapter 5
8.3.1 Annex 1: Long list of indicators of climate change technology transfer

Indicator Source Included
1
FOREIGN CC-related imports UN Trade Statistics database (COMTRADE) Yes
INPUT Clean energy-related UNEP, Bloomberg New Energy Finance, SEPI: Global No
investments Trends in Sustainable Energy Investment 2010 (Data
only available for: China, India, Brazil, Taiwan, Pakistan,
Indonesia, Thailand, Vietnam, Sri Lanka, Philippines,
Mexico, Chile, Peru, Nicaragua, Panama, Costa Rica,
Egypt, South Africa, Ethiopia, Uganda, as countries with
less tan 0,1bn investment are omitted)
These data does not distinguish between local and
foreign investment, therefore it cannot be directly
used.
Climate change-related United Nations Conference on Trade and Development No
FDI. (UNCTAD), 2010. World Investment Report 2010:
Investing in a Low-carbon Economy, United Nations.
Provides info only on outward FDI for the top investors,
not on the recipient countries, therefore this
information is not sufficient.
Foreign workers in CC Not available No
technology sectors
CC-related licence Not available No
payments to foreign
companies
CC-related patents filed Data available only for top performers (China, Brazil, No
by foreign actors India), not for a valid sample of developing countries, in
papers about clean energy patenting
Foreign companies with Not available
R&D expenses in CC-
related sectors
OUTPUT Renewable energy US EIA. Data available on installed MW, but it is not Yes
capacity. MW installed by possible to know which of them used foreign
foreign owners or with technologies. However, we can use claims of
foreign technologies technology transfer in CDM projects to estimate the
installed capacity that has required foreign technology
transfer
kWh of kJ generated or US EIA. Data available on installed MW, but it is not No, highly
saved by foreign-owned possible to know which of them used foreign correlated
projects or with foreign technologies with
technologies capacity
tCO2e captured or US EIA. Data available on renewable generation and No, data
destroyed by foreign- biofuels production, but it is not possible to know not
owned projects or with which of them used foreign technologies available
Production of CC-related Export values (in billion USD) for renewable and energy Yes
technologies or products efficiency technologies were acquired from Trade
by foreign owners or with Statistics Database (COMTRADE). They do not
foreign technologies represent completely the total production of
A proxy could be exports renewable energy technologies by a country, but are a
proxy for the production of internationally competitive
technologies.
Revenues from local These data are not available No
patents developed with
314
Indicator Source Included
foreign support
Number of local scientific These data are not available No
publications in
collaboration with
foreign institutions
EFFECT tCO2e reductions Data for this exercise are not available for a significant No
explained by foreign CC- sample of developing countries
related inputs
Cost of clean energy Data for this exercise are not available for a significant No
explained by foreign CC- sample of developing countries
related inputs
TFP of CC-related Data for this exercise are not available for a significant No
industrial sectors sample of developing countries
explained by foreign CC-
related inputs
8.3.2 Annex 2: Long list of indicators of enabling factors for technology transfer
Variables shaded in grey are not considered in the analysis. The reasons for not including
them are detailed in the column “included”
Type of TT Variables Indicator Source Included
variable aspect
Economi Flows EDB Ease of Doing Business score, The World Bank Yes
c and of 2011 data. .
institutio foreign CPI score Corruption Perception Index Transparency Yes
nal inputs score, 2010 International
framewo IPR Intellectual Property Rights index, Property Rights Yes
rk 2010 Alliance
INCOMETAX Average income tax rate, 2011 Heritage Found. Yes
and Wall Street
Journal
CORPTAX Average corporate tax rates, Heritage Found. No- correlation
2011 and Wall Street with INCOMETAX
Journal
TAX Total tax rate (% of commercial World Bank No- correlation
profits) indicators with INCOMETAX
CRED Domestic credit to private sector World Bank Yes
as a percentage of GDP, 2009 Indicators
data
CREDEN Domestic credit for renewable - No- Data not
energy projects available
Demand Output GDP Gross Domestic Product in World Bank Yes
size current Million US$ at Indicators
purchasers' prices, 2009
GDPgrowth Average GDP Growth between World Bank Yes
2005 and 2009 Indicators
GDPpc GDP per capita in current US$ World Bank Yes
Indicators
ENUSEPC Energy use per capita in kg oil World Bank No- High
equivalent, 2008 Indicators correlation with
GDPPC and
CO2PC
ENUSE Energy use, total, in kg oil World Bank No- High
equivalent, 2008 Indicators correlation with
315
variable aspect
GDP
CO2 CO2 emissions in kt, 2007 World Bank No- High
Indicators correlation with
GDP
CO2pc CO2 emissions per capita in World Bank Yes
metric tons, 2007 Indicators
CO2GDP CO2 emissions per GDP Yes
ELEC Electric power consumption per World Bank No- High
capita in kWh, 2009 Indicators correlation with
GDP
Energy Output PDIES Pump price for diesel fuel (US$ World Bank Yes
prices per liter) 2010 Indicators
PGASOL Pump price for gasoline (US$ per World Bank No- High
liter) 2010 Indicators correlation with
PDIES
PELEC Electricity prices International No- Data only
Energy Agency available for 15
developing
countries
Industria Local HTECHEXP High-technology exports (current World Bank No- High
l inputs US$) indicators correlation with
develop Techno ISO9001 and
ment logy more missing
effect values
HTEXPC Title: High-technology exports as World Bank Yes
a percentage of manufactured indicators
exports
ISO 9001 Number of companies with ISO The ISO 9001 Yes
9001 certification, 2008 Survey
TFP 2005 Total Factor Productivity levels UNIDO Yes
relative to the US for 2005.
CIP Competitive Industrial UNIDO Yes
Performance score.
INDPOL Industrial policy - No- Data not
available except
for TARIFF
LINKIND Linked industrial policy - No- Data not
available
ELECCON Concentration of the electricity - Data not
generation sector systematically
collected
available for most
developing
countries
Technol Local RDEXP Research and development World Bank No- Large number
ogical input expenditure as a percentage of indicators of missing values
develop Techno GDP, 2007
ment logy RES Number of researchers per World Bank No- Large number
effect million people, 2007 indicators of missing values
ARTICLES Number of articles published in World Bank No- High
scientific and technical journals, indicators correlation with
as an average during the period PATLOC
2003-2007
Enrol3 Tertiary education school World Bank Yes
316
variable aspect
enrolment ratio, as a percentage indicators
of population
ENEDU Specific capabilities in clean - No- Data not
energy technologies systematically
collected in
developing
countries
INTEDU Workforce with intermediate - No- Data not
technical education systematically
collected in
developing
countries
ENGSCI Percentage of tertiary enrolment UNESCO No- Large number
in engineering, science, of missing values
manufacturing and construction
disciplines, 2008
PATLOC Stock of patents filed by local WIPO statistics Yes
inventors during the period 1883- database and
2009, calculated following the own calculation
perpetual inventory method with
a 10% discount rate
PATLOCCC Stock of patents filed by local WIPO statistics No- Data not
inventors specifically related to database available for most
climate change technologies developing
countries
Foreign Flows TARIFF Most Favoured Nation average World Trade Yes
trade of applied tariff rates applied for Statistics
and foreign non agricultural goods. A high
investm inputs rate represents high protection,
ent 2009
TRADEOP Trade openness calculated as World Trade Yes
imports plus exports divided by Statistics
GDP, 2009
ROYALTIES Payments to foreign parties for World Bank No- High
royalties and licences fees indicators correlation with
(current US$), 2009 PATFOR
INVESTFREE Index of investment freedom, Heritage Found. Yes
2011 and Wall Street
Journal
FDIOP Foreign Direct investment net Own calculation Yes
inflows as a percentage of GDP, using World Bank
2009 indicators
FDIk Foreign Direct Investment, net World Bank No- High
inflows in current thousand US$, Indicators correlation with
2009 GDP
PATFOR Total stock of patents filed by WIPO statistics Yes
foreign inventors during the database and
period 1883-2009, calculated own calculation
using the perpetual inventory
method with a 10% discount rate
PATFORCC Total stock of patents filed by WIPO statistics No- Data not
foreign investors specifically available for most
related to climate change developing
technologies countries
317
variable aspect
LOG Logistics performance index: World Bank Yes
Overall (1=low to 5=high) indicators
Access Local SOLAR Estimated renewable energies Buys et al, 2007, No- Variables not
to inputs WIND potential for different renewable country stakes in considered
renewab HYDRO energy sources, in Millions of toe. climate change individually, but
le GEOTH negotiations, aggregated
energy SUGAR World Bank paper
MANURE
SAVANNA
JATROPHA
REACNH
REACCESS Access to renewable energy Buys et al, 2007, Yes
sources. It adds up values for all and own
the previous sources. Variable calculation
also considered in per capita and
per GDP values.
WSHACC Access to wind, solar and hydro Buys et al, 2007 Yes
resources. Variable also
considered in per capita and per
GDP values.
Access Output FOSSIL Production of oil, coal and gas in US EIA, 2009 Yes
to fossil energy units. Variable also
fuels considered in per capita and per
GDP values.
Demand Output FIT Countries that have implemented IEA Policies and Yes
-pull Feed-in tariffs or that provide Measures
policies guaranteed premiums to database,2011
renewable electricity generation.
1=yes, and 0=no feed-in tariffs
Other IEA Policies and No- Other
policies Measures demand-pull
policies are
difficult to
compare
318
Case Processing Summary
Cases
Valid Missing Total
N Percent N Percent N Percent
EDB 120 44.9% 147 55.1% 267 100.0%
CPI score 106 39.7% 161 60.3% 267 100.0%
IPR 74 27.7% 193 72.3% 267 100.0%
CRED 111 41.6% 156 58.4% 267 100.0%
GDP 125 46.8% 142 53.2% 267 100.0%
GDPg 130 48.7% 137 51.3% 267 100.0%
GDPpc 125 46.8% 142 53.2% 267 100.0%
ENUSEPC 87 32.6% 180 67.4% 267 100.0%
ENUSE 87 32.6% 180 67.4% 267 100.0%
CO2 129 48.3% 138 51.7% 267 100.0%
CO2pc 129 48.3% 138 51.7% 267 100.0%
ELEC 86 32.2% 181 67.8% 267 100.0%
PDIES 112 41.9% 155 58.1% 267 100.0%
PGASOL 112 41.9% 155 58.1% 267 100.0%
HTECHEXP 87 32.6% 180 67.4% 267 100.0%
HTEXPC 87 32.6% 180 67.4% 267 100.0%
ISO 9001 Companies 129 48.3% 138 51.7% 267 100.0%
TFP 2000 81 30.3% 186 69.7% 267 100.0%
TFP 2005 80 30.0% 187 70.0% 267 100.0%
CIP 74 27.7% 193 72.3% 267 100.0%
RDEXP 38 14.2% 229 85.8% 267 100.0%
RES 20 7.5% 247 92.5% 267 100.0%
ARTICLES 130 48.7% 137 51.3% 267 100.0%
Enrol3 78 29.2% 189 70.8% 267 100.0%
ENGSCI 46 17.2% 221 82.8% 267 100.0%
PATSTOCKLOCAL 98 36.7% 169 63.3% 267 100.0%
PATSTOCKFOR 97 36.3% 170 63.7% 267 100.0%
MFNTARIFF 121 45.3% 146 54.7% 267 100.0%
TRADEOP 116 43.4% 151 56.6% 267 100.0%
Royalty payments 76 28.5% 191 71.5% 267 100.0%
INVESTFREE 124 46.4% 143 53.6% 267 100.0%
FDIOP 130 48.7% 137 51.3% 267 100.0%
FDI 123 46.1% 144 53.9% 267 100.0%
LOG 106 39.7% 161 60.3% 267 100.0%
INCOMETAX 123 46.1% 144 53.9% 267 100.0%
CORPTAX 123 46.1% 144 53.9% 267 100.0%
TAX 124 46.4% 143 53.6% 267 100.0%
REACNH 130 48.7% 137 51.3% 267 100.0%
REACCESSTOT 130 48.7% 137 51.3% 267 100.0%
Economic and institutional framework

Each of the variables represents different aspects of the institutional and economic
framework, as their correlation is not very high. Therefore all of them will be
considered in the analysis: Ease of Doing Business (EDB), Corruption Perception Index
(CPI), Intellectual Property Rights (IPR) and credit available (CRED) for the private
sector.
319
The different indicators of tax burden do not show a strong correlation. The tax
indicator provided by the World Bank is not related to any of the indicators by the
Heritage Foundation and the Wall Street journal. All of them have a similar number of
missing values. The INCOMETAX value will be the one selected, as it reflects incentives
for labour mobility which are not captured by other indicators.
Demand size
The three different measures considered for the absolute and per capita size of
demand for clean technologies are highly correlated. GDP, GDPpc and GDPg are
selected to represent variations in demand size and growth because they have the
lowest number of missing values and the most recent data.
Four measures of size per capita
320
Energy prices
Diesel and gasoline prices are highly correlated. Diesel prices will be selected to
represent fossil fuels prices.
Industrial development indicators

Among indicators of industrial development, only the value of high technology exports
and the number of ISO 9001 registered companies are highly correlated. The later is
selected to represent the quality of local suppliers as it has a lower number of missing
values. The variables CIP, TFP2005 and HTEXPC are kept for the analysis.
321
On the other hand, the variable ISO9000 is highly correlated with the size of the
economy, which may present problems later on.
Technological development
Many technological development variables have a very high number of missing values,
which suggests the need to reduce their number to avoid loss of representation in the
analysis. The stock of local knowledge (PATSTOCKLOCAL) is one of the variables with
the lowest number of missing values and is highly correlated with RDEXP and RES, with
many missing values. Therefore it will be used to represent the variations in these
other variables. It will also be used to represent the variable ARTICLES, as both show a
high correlation. The tertiary education school enrolment ratio (ENROL3) will be kept
as an indication of the general education in the country, but the percentage of
enrolment in engineering, science, manufacture and construction disciplines will be
rejected due to the high number of missing values.
322
Foreign flows of technology
The two variables representing the flow of knowledge through foreign patents,
ROYALTIES and PATSTOCKFOR, are highly correlated. The latter is kept as it has a lower
number of missing values.
Both trade openness (TRADEOP) and Foreign Investment openness (FDIOP) are kept as
they show different channels of foreign technology and their Pearson correlation
statistic is not higher than 0.9.
323
TRADEOP does not have a strong correlation with the level of tariffs (MFNTARIFF) or
the logistic performance of the economy (LOG). For this reason, all variables are kept
for the analysis.
The absolute value of foreign direct investment (FDI) is highly correlated with the size
of the economy (GDP). Therefore, the latter will be used to represent variations in FDI.
It is strange that the index of freedom of investment (INVESTFREE) and the level of
openness to FDI do not show a significant correlation. Both variables are kept in the
analysis as they may represent different aspects.
324
8.3.3 Annex 3: Country codes
Afghanistan AFG Georgia GEO Pakistan PAK
Algeria DZA Ghana GHA Panama PAN
Angola AGO Guatemala GTM Peru PER
Argentina ARG Guinea GIN Philippines PHL
Armenia ARM Guinea-Bissau GNB Qatar QAT
Azerbaijan AZE Guyana GUY Russian Fed RUS
Bahrain BHR Haiti HTI Rwanda RWA
Bangladesh BGD Honduras HND Samoa WSM
Belarus BLR India IND São Tomé and Principe STP
Belize BLZ Indonesia IDN Saudi Arabia SAU
Benin BEN Iran IRN Senegal SEN
Bhutan BTN Iraq IRQ Seychelles SYC
Bolivia BOL Jamaica JAM Sierra Leone SLE
Botswana BWA Jordan JOR Solomon Islands SLB
Brazil BRA Kazakhstan KAZ Somalia SOM
Burkina Faso BFA Kenya KEN South Africa ZAF
Burundi BDI Kiribati KIR Sri Lanka LKA
Cambodia KHM Korea, DR PRK Sudan SDN
Cameroon CMR Kyrgyz Rep KGZ Suriname SUR
Cape Verde CPV Lao PDR LAO Swaziland SWZ
Central African Rep CAF Lebanon LVA Syria SYR
Chad TCD Lesotho LSO Tajikistan TJK
Chile CHL Liberia LBR Tanzania TZA
China CHN Libya LBY Thailand THA
Colombia COL Madagascar MDG Timor-Leste TMP
Comoros COM Malawi MWI Togo TGO
Congo, DR ZAR Malaysia MYS Tonga TON
Congo, Rep. COG Mali MLI Tunisia TUN
Costa Rica CRI Mauritania MRT Turkey TUR
Côte d'Ivoire CIV Mauritius MUS Turkmenistan TKM
Cuba CUB Mexico MEX Uganda UGA
Djibouti DJI Micronesia FSM Ukraine UKR
Dominica DMA Moldova MDA United Arab Emir. ARE
Dominican Rep DOM Mongolia MNG Uruguay URY
Ecuador ECU Morocco MAR Uzbekistan UZB
Egypt. EGY Mozambique MOZ Vanuatu VUT
El Salvador SLV Myanmar MMR Venezuela, RB VEN
325
Equatorial Guinea GNQ Namibia NAM Vietnam VNM
Eritrea ERI Nepal NPL West Bank and Gaza WBG
Ethiopia ETH Nicaragua NIC Yemen, Rep. YEM
Fiji FJI Niger NER Zambia ZMB
Gabon GAB Nigeria NGA Zimbabwe ZWE
Gambia, The GMB Oman OMN
8.3.4 Annex 4: Descriptive graphs of dependent variables
8.3.4.1 Imports of renewable energy technologies
326
327
328
8.3.4.2 Exports of renewable energy technologies
329
REEXPPC can be approximated as a normal distribution N(3.55, 2.794)
330
REEXPGDPlog are approximated through a normal distribution N(8.82, 3.05)
331
8.3.4.3 Renewable energy capacity involving technology transfer
The histogram and box plot RECAPTTlog is shown below. It can be approximated by a
normal distribution N (5.88, 2.466)
332
Values per capita of the renewable energy capacity requiring foreign technology
transfer (RECAPTTPC) can be approximated through a normal distribution N (3.57,
1.745)
333
The distribution of relative values per GDP transformed with logs (RECAPTTGDPlog) is
shown below.
334
8.3.5 Annex 5: Regression equations
8.3.5.1 Explaining exports of renewable energy technologies per capita

TABLE 57- MODEL RESULTS: EXPORTS OF RENEWABLE ENERGY TECHNOLOGIES PER CAPITA
I II III IV V VI VII
Dependent REEXPPClo REEXPPClo REEXPPClo REEXPPClo REEXPPClo REEXPPClo REEXPPClo
variable g g g g g g g
N 54 54 54 53 54 54 42
Constant -7.227*** -5.982*** -6.555*** -5.415*** -7.262*** -8.929*** -7.222***
(1.322) (1.704) (1.628) (1.583) (1.381) (1.679) (1.533)
WSHACCPClo .889*** 0.856*** 0.894*** 0.836*** 0.898*** 0.919*** 0.883***
g (0.251) (0.252) (0.275) (0.246) (0.270) (0.248) (0.291)
CIP 17.283*** 16.403*** 20.440*** 18.158*** 17.388*** 19.280*** 17.150***
(3.085) (3.169) (3.079) (3.034) (3.288) (3.283) (3.589)
GDPlog .396*** 0.381*** 0.286** 0.402*** 0.595*** 0.356**
(0.127) (0.127) (0.135) (0.140) (0.176) (0.171)
IPR .589** 0.539** 0.604** 0.651*** 0.585** 0.590** 0.607**
(0.229) (0.233) (0.252) (0.225) (0.236) (0.226) (0.269)
EDB -0.006
(0.005)
GDPpclog 0.330
(0.227)
PDIES -1.269*
(0.647)
PATLOCPClog -0.013
(0.134)
PATFORlog -0.254
(0.158)
ENROL3log 0.125
(0.280)
R-Square 0.691 0.699 0.646 0.714 0.691 0.707
Adjusted R- 0.667 0.669 0.618 0.685 0.660 0.677
Square
F 27.986 22.801 22.825 24.436 21.947 23.607
Note: Standard errors in parentheses;*** p<0.01, ** p<0.05, * p<0.1
335
8.3.5.2 Explaining imports of renewable energy technologies per capita

TABLE 58- MODEL RESULTS: IMPORTS OF RENEWABLE ENERGY TECHNOLOGIES PER CAPITA
I II III IV V VI VII
Dependent REIMPPClo REIMPPClo REIMPPClo REIMPPClo REIMPPClo REIMPPClo REIMPPClo
variable g g g g g g g
N 71 71 71 64 71 68 67
Constant -1.108 -0.427 -0.845 -1.620* -0.909 -0.881 -1.730*
(0.756) (1.309) (0.945) (0.826) (0.817) (0.790) (0.938)
GDPpclog .668*** 0.639*** 0.656*** 0.629*** 0.619*** 0.645*** 0.619***
(0.109) (0.119) (0.113) (0.115) (0.133) (0.113) (0.120)
CREDlog .688*** 0.624*** 0.673*** 0.591*** 0.711*** 0.619*** 0.580**
(0.172) (0.200) (0.176) (0.185) (0.176) (0.183) (0.199)
EDB -0.002
(0.003)
TARIFF -0.012
(0.025)
IPR 0.249*
(0.131)
FOSSILpclo 0.026
g (0.039)
PATFORlog 0.066
(0.052)
LOG 0.514
(0.435)
R-Square 0.593 0.596 0.595 0.616 0.596 0.603 0.602
Adjusted R- 0.582 0.578 0.577 0.597 0.578 0.585 0.583
Square
F 50.340 33.407 33.252 32.620 33.429 32.927 32.272
336
8.3.5.3 Explaining renewable electricity generation capacity with technology

transfer per capita
TABLE 59- MODEL RESULTS: RENEWABLE GENERATION CAPACITY WITH TT PER CAPITA
I II III IV V
Dependent RECAPTTPClog RECAPTTPClog RECAPTTlog RECAPTTlog RECAPTTlog
variable
N 70 70 70 70 70
-5.336*** -2.667 -4.377* -6.271*** -5.802***
Constant
(1.866) (1.643) (2.426) (1.977) (2.101)
.814*** 0.912*** 0.756*** 0.784*** 0.830***
GDPpclog
(0.186) (0.175) (0.209) (0.186) (0.189)
-.679*** -0.574*** -0.682*** -0.666*** -0.707***
IPR
(0.191) (0.190) (0.192) (0.190) (0.201)
.284*** 0.159** 0.282*** 0.304*** 0.304***
REACPClog
(0.084) (0.77) (0.085) (0.085) (0.094)
1.581** 1.506** 1.982*** 1.745**
LOG
(0.626) (0.641) (0.689) (0.712)
-.132** -0.126** -0.138** -0.130**
FOSSILPClog
(0.062) (0.063) (0.061) (0.062)
PDIES 0.768
(0.542)
EDB -0.003
(0.004)
FIT -0.903
(0.664)
PATLOCpclog -0.51
(0.104)
R-Square 0.406 0.344 0.409 0.422 0.408
Adjusted R- 0.360 0.304 0.354 0.368 0.352
Square
337
I II III IV V
Dependent RECAPTTPClog RECAPTTPClog RECAPTTlog RECAPTTlog RECAPTTlog
variable
F 8.871 8.642 7.388 7.797 7.347
FIGURE 85- RESIDUALS VS PREDICTED VALUES PLOT OF THE SELECTED MODEL TO EXPLAIN
RECAPTTPCLOG
8.3.6 Annex 6: Second iteration of clustering analysis with variables not transformed with
logarithms and excluding outliers Saudi Arabia, Qatar and China
8.3.6.1 Ward’s method
The agglomeration schedule shows that a four cluster structure is appropriate.

TABLE 60- AGGLOMERATION SCHEDULE
Cluster Combined Stage Cluster First Appears

Stage Cluster 1 Cluster 2 Coefficients Cluster 1 Cluster 2 Next Stage
1 77 102 .028 0 0 20
2 59 81 .091 0 0 6
3 35 106 .188 0 0 13
4 30 85 .291 0 0 7
5 11 115 .405 0 0 10
6 13 59 .560 0 2 17
7 30 46 .727 4 0 30
8 19 67 .902 0 0 25
9 54 83 1.086 0 0 23
10 11 108 1.308 5 0 27
11 57 112 1.539 0 0 29
12 25 118 1.848 0 0 21
13 35 89 2.168 3 0 27
14 2 123 2.507 0 0 25
15 8 74 2.851 0 0 16
16 8 44 3.227 15 0 41
17 13 88 3.664 6 0 28
18 37 97 4.135 0 0 26
338
19 36 56 4.616 0 0 21
20 14 77 5.188 0 1 33
21 25 36 5.764 12 19 39
22 4 113 6.409 0 0 34
23 51 54 7.107 0 9 28
24 15 52 7.831 0 0 37
25 2 19 8.613 14 8 40
26 37 53 9.512 18 0 30
27 11 35 10.478 10 13 39
28 13 51 11.481 17 23 38
29 29 57 12.514 0 11 43
30 30 37 13.806 7 26 41
31 23 86 15.108 0 0 36
32 69 109 16.437 0 0 44
33 14 80 17.811 20 0 38
34 4 73 19.455 22 0 42
35 63 121 21.450 0 0 45
36 23 84 23.856 31 0 43
37 15 92 26.525 24 0 42
38 13 14 29.675 28 33 40
39 11 25 33.072 27 21 46
40 2 13 36.713 25 38 46
41 8 30 42.010 16 30 47
42 4 15 47.937 34 37 48
43 23 29 53.921 36 29 45
44 69 101 60.579 32 0 49
45 23 63 67.939 43 35 48
46 2 11 76.634 40 39 47
47 2 8 89.463 46 41 50
48 4 23 110.460 42 45 49
49 4 69 136.712 48 44 50
50 2 4 216.948 47 49 0
The dendrogram shows a different structure to the one in the cluster analysis with
variables transformed with logs. In a four cluster structure, we find a very large cluster
with 34 cases and three smaller groups with 8, 6 and 3 clusters.
339
340
The cluster formation is presented in the following table, indicating as well the
observed sub-clusters in each group.
TABLE 61- CLUSTER STRUCTURE SECOND ITERATION
Cluster
Members Subdivisions
number
1 2:DZA 59:KEN The dendrogram shows the following subdivisions in
the cluster:
8:BGD 67:MDG
• Cameroon, Madagascar, Algeria and Zambia
11:BEN 74:MDA
• Colombia, Uruguay, Egypt and Jamaica
13:BOL 77:MOZ • Benin, Uganda, Tanzania, Peru, Ecuador,
14:BWA 80:NPL Syria
• Honduras, Nigeria, Paraguay, Bolivia, Iran,
19:CMR 81:NIC Nicaragua, Kenya, Nepal, Botswana,
25:COL 83:NGA Mozambique, Sri Lanka
30:CIV 85:PAK
35:ECU 88:PRY
36:EGY 89:PER
37:SLV 97:SEN
44:GEO 102:LKA
46:GTM 106:SYR
51:HND 108:TZA
53:IDN 115:UGA
54:IRN 118:URY
56:JAM 123:ZMB
2 4:ARG 73:MEX The dendrogram shows the following subdivisions in

the cluster:
15:BRA 92:RUS
• Argentina, Turkey and Mexico
52:IND 113:TUR
• Brazil, India and Russia
3 23:CHL 84:OMN The following subdivisions can be noticed in the

dendrogram:
29:CRI 86:PAN
• Chile, Panama and Oman
57:JOR 112:TUN
• Costa Rica, Tunisia and Jordan
63:LVA 121:VNM • Lebanon and Vietnam
4 69:MYS The dendrogram shows the following subdivisions in

the cluster:
101:ZAF
• Malaysia and Thailand
109:THA
• South Africa stays on its own
The ANOVA table shows that neither access to renewable energy sources per capita,
nor production of fossil fuels per capita, are useful to distinguish between the different
clusters. All the other variables are significant to distinguish between clusters.
341
TABLE 62- ANOVA
Sum of Squares df Mean Square F Sig.
Zscore(IPR) Between Groups 13.536 3 4.512 6.953 .001

Total 44.034 50
Zscore(CIP) Between Groups 23.980 3 7.993 15.420 .000
Total 48.344 50
Zscore(LOG) Between Groups 22.809 3 7.603 30.047 .000
Total 34.702 50
Zscore(GDP) Between Groups 18.075 3 6.025 43.472 .000
Total 24.589 50
Zscore(GDPPC) Between Groups 3.373 3 1.124 11.170 .000
Total 8.104 50
Zscore(CRED) Between Groups 45.701 3 15.234 63.540 .000
Total 56.969 50
Zscore(REACPC) Between Groups .011 3 .004 .907 .445
Within Groups .190 47 .004
Total .201 50
Zscore(FOSSILPC) Between Groups .001 3 .000 1.374 .262
Within Groups .006 47 .000
Total .006 50
Cluster descriptives show the mean, standard deviation and standard error of each
cluster, as well as the confidence interval for each cluster´s mean. This allows us to
identify outliers in each of the clusters. Descriptives are presented in Appendix 9.5. The
Tukey post-hoc test, also presented in the Appendix indicates similarities and
dissimilarities between clusters for each variable and shows that:
• For GDPpc:
o Cluster 1 and 4 are not dissimilar
o Cluster 4 is not dissimilar to any of the other clusters
• For GDP
o Clusters 1, 3 and 4 are not dissimilar in terms of GDP
o Cluster 2 is dissimilar to all the other clusters in the size of GDP
• For CRED
342
• For IPR
o Cluster 3 is not dissimilar to any of the other clusters
• For CIP
• For LOG
• For REACCpc and FOSSILpc none of the clusters are dissimilar
The means plot allow us to visualise the characteristics of each cluster, which are
defined in more detail in Error! Reference source not found.. In the means plot we do
not consider the variables REACCpc and FOSSILpc as they are not significant to
differentiate between clusters.
FIGURE 86- MEANS PLOT
TABLE 63- CLUSTER CHARACTERISATION
Cluster
Members Characteristics
number
1 2:DZA 59:KEN
• Lowest GDP, not dissimilar to Cluster 3
8:BGD 67:MDG
• Lowest GDPpc
11:BEN 74:MDA
• Lowest CRED, not dissimilar to Cluster 2
13:BOL 77:MOZ
• Lowest IPR
14:BWA 80:NPL
343
19:CMR 81:NIC • Lowest CIP, not dissimilar to Cluster 3
25:COL 83:NGA • Lowest LOG
30:CIV 85:PAK
35:ECU 88:PRY
36:EGY 89:PER
37:SLV 97:SEN
44:GEO 102:LKA
46:GTM 106:SYR
51:HND 108:TZA
53:IDN 115:UGA
54:IRN 118:URY
56:JAM 123:ZMB
2 4:ARG 73:MEX • Highest GDP
15:BRA 92:RUS • Second highest GDP per capita, not
dissimilar to cluster 3
52:IND 113:TUR • Second lowest CRED, not dissimilar to
Cluster 1
• Medium IPR, not dissimilar to Cluster 3
• Second highest CIP, not dissimilar to
clusters 3 and 4
• Second highest LOG, not dissimilar to
cluster 3
3 23:CHL 84:OMN • Second lowest GDP, not dissimilar to

29:CRI 86:PAN Cluster 1
• Highest GDP per capita, not dissimilar to
57:JOR 112:TUN Cluster 2
63:LVA 121:VNM • Second highest CRED
• Medium IPR, not dissimilar to Cluster 2
• Second lowest CIP, not dissimilar to
Cluster 1
• Medium LOG, not dissimilar to Cluster 2
4 69:MYS • Low GDP, not dissimilar to Clusters 1 and

101:ZAF 3
• Medium GDP per capita, which is not
109:THA dissimilar to any of the other clusters
• The highest CRED
• The highest IPR, not dissimilar to clusters
2 and 3
• The highest CIP, not dissimilar to Cluster
2
• The highest LOG, not dissimilar to
Clusters 2 and 3
344
We tested the sensitivity of the cluster formation to the elimination of variables
REACCpc and FOSSILpc which were not significant to differentiate between clusters.
Cluster analysis without these variables delivered an identical cluster formation.
We compare the results obtained using differentiating variables with and without
logarithms. The cluster formation using variables transformed with logs rendered 4
clusters with 10, 12, 20 and 10 cases each, including two relatively high performers
(Clusters 2 and 4) and 2 relatively low performers (clusters 1 and 3) with Clusters 2 and
3 being the highest and lowers performers overall. The cluster formation using not-
transformed variables renders 4 clusters with 34, 6, 8 and 3 cases each, in which
Cluster 1 has the highest number of cases and is the lowest performer and Cluster 4
has the lower number of cases and is the highest performer overall.
To analyse the stability of the different countries as regards their belonging to
different clusters in the iterations performed, Table 64 shows the results of
hierarchical cluster analysis in the different iterations. In broad terms, the cluster
formation without logs ( iterations 3 and 4) joins clusters 1 and 3 from iteration 1 with
logs in a single cluster. It also splits Cluster 2 from the formation with logs in two
clusters. The most interesting cases showing some instability in their clusters place are:
Botswana, El Salvador, Jamaica, Uruguay, Russia, Oman and Vietnam.
o Botswana, Jamaica and El Salvador belonged to Cluster 4 of high performers in
iteration 1 with logs , but are placed in Cluster 1 of low performers in iteration
2. These countries did not reach the lower bound of the mean of many
variables when they were placed in Cluster 4 of high performers in iteration 1.
Therefore, it is not strange that the new cluster structure places them among
low performers.
o Uruguay was also placed in Cluster 4 of high performers in iteration 1 but in
iteration 3 is placed among low performers. This country did not show a
consistently lower performance than its peers in Cluster 4.
o Russia was placed in Cluster 1 in iteration 1, among low performers, but in
iteration 3 is placed in Cluster 2 with high performers. Russia showed in the
first iteration that it exceeded the upper bound of the mean of many variables
and therefore was performing better than many of its peers.
o Oman has been moved from Cluster 2, in iteration 1, to Cluster 3 in iteration 3.
This may be because Saudi Arabia and Qatar have not been included in the
analysis and therefore Oman cannot join the highest performers group through
its proximity to these other countries. Oman has a lower GDP, CIP and LOG
than the highest performers and therefore does not belong to that group
o Vietnam had been placed in group 3 of low performers in iteration 1 but is now
placed in group 3 among medium-high performers. Vietnam exceeded the
performance of many of its peers in Cluster 2 of iteration 1, and therefore it can
belong to a group of better performers.
TABLE 64-CLUSTERS FORMATION IN HIERARCHICAL CLUSTER ANALYSIS
Case 4 clusters 4 clusters

with no logs with logs
It. 3 and 4 It. 1 and 2
2:DZA 1 1
8:BGD 1 3
345
11:BEN 1 3
13:BOL 1 1
14:BWA 1 4
19:CMR 1 3
25:COL 1 1
30:CIV 1 3
35:ECU 1 1
36:EGY 1 1
37:SLV 1 4
44:GEO 1 3
46:GTM 1 3
51:HND 1 3
53:IDN 1 1
54:IRN 1 1
56:JAM 1 4
59:KEN 1 3
67:MDG 1 3
74:MDA 1 3
77:MOZ 1 3
80:NPL 1 3
81:NIC 1
83:NGA 1 3
85:PAK 1 3
88:PRY 1 3
89:PER 1 1
97:SEN 1 3
102:LKA 1
106:SYR 1 1
108:TZA 1 3
115:UGA 1 3
118:URY 1 4
123:ZMB 1 3
4:ARG 2 2
15:BRA 2 2
52:IND 2 2
73:MEX 2 2
92:RUS 2 1
113:TUR 2 2
23:CHL 3 4
29:CRI 3 4
57:JOR 3 4
63:LVA 3 4
84:OMN 3 2
86:PAN 3 4
112:TUN 3 4
121:VNM 3 3
346
69:MYS 4 2
101:ZAF 4 2
109:THA 4 2
8.3.6.2 k-means method
As with the hierarchical cluster analysis, the clustering analysis using untransformed
variables delivers a more unbalanced cluster structure, that does not differentiate
among a large number of countries placed in a single country. Pre-selecting 4 clusters,
we would have clusters with 32, 9, 7 and 3 cases.
TABLE 65- NUMBER OF CASES IN EACH CLUSTER
Cluster 1 32
2 9
3 7
4 3
Valid 51
Missing 69
The ANOVA shows that in this case REACCpc and FOSSILpc are not useful to
differentiate variables across clusters.
TABLE 66- ANOVA
Cluster Error
Mean Square df Mean Square df F Sig.
Zscore(IPR) 4.572 3 .645 47 7.087 .001

Zscore(CIP) 9.054 3 .451 47 20.092 .000
Zscore(LOG) 7.439 3 .264 47 28.231 .000
Zscore(GDP) 4.770 3 .219 47 21.810 .000
Zscore(GDPPC) 1.069 3 .104 47 10.262 .000
Zscore(CRED) 15.445 3 .226 47 68.263 .000
Zscore(REACPC) .002 3 .004 47 .369 .776
Zscore(FOSSILPC) .000 3 .000 47 1.651 .190
The Clusters formation is presented in Table 67, as well as the characteristics of each
of the clusters, which are taken from the comparison of their means for each variable,
and the Tukey post-hoc test of dissimilarities between clusters.
TABLE 67-CLUSTER FORMATION WITH K-MEANS
Cluster Members Characteristics Outliers

1 DZA MDG • The lowest GDP, not dissimilar to • Uruguay overperforms
BGD MDA Clusters 3 and 4 in GDPpc, IPR, CIP and
BEN MOZ • The lowest GDPpc, not dissimilar to LOG
BOL NPL Cluster 4
BWA NIC • The lowest CRED, not dissimilar to
CMR NGA Cluster 2
CIV PAK • The lowest IPR, not dissimilar to
347
Cluster Members Characteristics Outliers
ECU PRY Cluster 3
EGY PER • The lowest CIP, not dissimilar to
SLV SEN Cluster 3
GEO LKA • The lowest LOG
GTM SYR
HND TZA
IRN UGA
JAM URY
KEN ZMB
2 ARG IDN • The highest GDP • Brazil overperforms its
BRA MEX • High GDPpc, not dissimilar to peers in GDP, CRED, IPR
COL RUS Clusters 3 and 4 and LOG
CRI TUR • Second lowest CRED, not dissimilar • Indonesia
IND to Cluster 1 underperforms in
• Medium IPR, not dissimilar to GDPpc, IPR and LOG
Clusters 3 and 4 • Russia underperforms in
• Second highest CIP CIP and LOG, but
• Medium LOG, not dissimilar to outperforms in GDPpc
Cluster 3
3 CHL PAN • The lowest GDP, not dissimilar to • No country consistently
JOR TUN clusters 1 and 4 under or overperforms
LVA VNM • The highest GDPpc, not dissimilar to its peers
OMN clusters 2 and 4
• Second highest CRED
• Medium IPR, not dissimilar to
Clusters 1, 2 and 4
• Second lowest CIP, not dissimilar to
Cluster 1
• Medium LOG, not dissimilar to
cluster 2
4 MYS • Second highest GDP, not dissimilar to • No outliers
ZAF Clusters 1 and 3
THA • High GDPpc, not dissimilar to
Clusters 1, 2 and 3
• Highest CRED
• Highest IPR, not dissimilar to Clusters
2 and 3
• Highest CIP
• Highest LOG
The means plot shows graphically the characteristics of each of the clusters.
348
3.5
2.5
2
Series1
1.5
Series2
Series3
1
Series4
0.5
0
GDPpc
LOG
GDP
CRED
IPR
CIP
-0.5
-1
The cluster formation with k-means using some variables transformed with logs
rendered 4 clusters with 4, 10, 18 and 20 members. The two smaller groups mainly
contain the best performers, with the first Cluster 1 including large countries, with
relatively high GDPpc, high production of Fossil fuels per capita and high access to
renewable energy sources per capita, but relatively bad performance in enabling
factors like access to credit, competitive industrial performance, IPR protection and
logistic performance. Cluster 2 shows better performance in these enabling factors,
but lower production of fossil fuels and access to renewable energies per capita. The
larger Clusters 3 and 4 include worse performers, with Cluster 4 performing better
than Cluster 3 in all variables except access to renewable energy sources.
When none of the variables was transformed with logs, the result was 4 clusters with
32, 9,7 and 3 members. The first, larger cluster 1 includes the worse performers in all
variables. Cluster 2 includes large countries, with low access to credit and medium
performance in IPR and CIP. Cluster 3 includes small countries with good access to
credit and good logistic performance but low competitive industrial performance. The
smallest cluster 4 includes relatively large countries with very good performance in
terms of access to credit, IPR, CIP and LOG.
To analyse the stability of the different countries as regards their belonging to
different clusters in the iterations performed, Table 68 shows the results of non-
hierarchical cluster analysis with k-means in the different iterations. In broad terms, in
the cluster formation without logs ( iteration 2):
• Members of Cluster 1 in the first iteration separate, with Argentina and Russia
joining Cluster 2 and Oman joining Cluster 3.
• Some members of cluster 2 in the first iteration remain in the same cluster
(Brazil, India, Mexico and Turkey), while Malaysia, South Africa and Thailand
349
separate to create their own cluster 4 of excellent performers. Chile changes
from Cluster 2 to Cluster 3, due to its small size and low industrial
performance.
• All members of Cluster 1 of low performers remain together in Cluster 3 in the
second iteration.
• Members of Cluster 4 in the first iteration separate according to their
performance. Colombia, Costa Rica and Indonesia move to Cluster 2. Jordan,
Lebanon and Panama join Cluster 3. The remaining countries move to Cluster 3
with the low performers.
TABLE 68- CLUSTERS FORMATION IN NON-HIERARCHICAL, K-MEANS CLUSTER ANALYSIS
Iteration Iteration
1 logs 2 no logs
Argentina 1 2
Oman 1 3
Russia 1 2
Saudi Arabia 1
Brazil 2 2
Chile 2 3
China 2
India 2 2
Malaysia 2 4
Mexico 2 2
Qatar 2
South Africa 2 4
Thailand 2 4
Turkey 2 2
Algeria 3 1
Bangladesh 3 1
Benin 3 1
Bolivia 3 1
Cameroon 3 1
Cote d'Ivoire 3 1
Georgia 3 1
Kenya 3 1
Madagascar 3 1
Moldova 3 1
Mozambique 3 1
Nepal 3 1
Pakistan 3 1
Paraguay 3 1
Senegal 3 1
Tanzania 3 1
Uganda 3 1
Zambia 3 1
Botswana 4 1
350
Colombia 4 2
Costa Rica 4 2
Ecuador 4 1
Egypt 4 1
El Salvador 4 1
Guatemala 4 1
Honduras 4 1
Indonesia 4 2
Iran 4 1
Jamaica 4 1
Jordan 4 3
Lebanon 4 3
Nigeria 4 1
Panama 4 3
Peru 4 1
Syria 4 1
Tunisia 4 3
Uruguay 4 1
Vietnam 4 3
8.3.6.3 Comparison of results of Ward’s and k-means methods with variables

without logs
We will compare as well the cluster formations achieved by both methods when taking
all variables untransformed with logs. In these cases, the differences among countries
are accentuated and the result is a much more uneven distribution of countries across
clusters. The hierarchical Ward’s method delivers four clusters with 34, 6, 8 and 3
members. The four clusters in the non-hierarchical k-means method have 32, 9, 7 and
3 members. The cluster´s characteristics are very similar in both method´s cluster
formations. The most populated Cluster 1 includes the lowest performers. The smallest
Cluster 4 includes the highest performers, but with relatively small sized economies.
Cluster 2 includes large economies with low credit for the private sector but higher
than average performance in CIP and LOG. Cluster 3 includes small economies with
good access to credit for the private sector but low CIP.
351
FIGURE 87- HIERARCHICAL (WARDS) CLUSTERING MEANS PLOT FOR SECOND ITERATION WITH NO LOGS
FIGURE 88- NON-HIERARCHICAL (K-MEANS) CLUSTERING MEANS PLOT FOR SECOND ITERATION WITH
NO LOGS
3.5
2.5
2
Series1
1.5
Series2
Series3
1
Series4
0.5
0
LOG
GDPpc
GDP
CRED
IPR
CIP
-0.5
-1
Table 69 shows the members of Cluster 1 in the clustering formations of both

methods. The cluster shows a high stability, with 32 of its members remaining in it in
both clustering methods. Colombia and Indonesia, however belong to Cluster 1 in the
Ward’s method but are placed in K-mean’s Cluster 2 among large economies with good
CIP and LOG. Colombia scores above the upper bound of Ward’s Cluster 1 mean for
352
GDP, GDPpc, LOG, IPR and CIP. Indonesia scores above the upper bound value of the
confidence level of the mean for GDP, LOG and CIP.
METHODS
Members of Cluster 1 in both Ward’s and k-means Members of Cluster 1 in

methods Ward’s and Cluster 2 in k-
means
Algeria Nepal Colombia
Bolivia Pakistan Indonesia
Ecuador Paraguay
Egypt Senegal
Iran Tanzania
Peru Uganda
Syria Zambia
Bangladesh Guatemala
Benin Honduras
Cameroon Nigeria
Cote d'Ivoire Botswana
Georgia El Salvador
Kenya Jamaica
Madagascar Uruguay
Moldova Nicaragua
Mozambique Sri Lanka
The Cluster 2 delivered by the k-means method is larger than that of the Ward’s
method. 6 countries remain stable in their belonging to this cluster. Colombia and
Indonesia are placed in Cluster 1 of lower performers by the Ward’s’ method and
Costa Rica is placed in Cluster 3. Costa Rica scores lower than the lower bound of
Ward’s’ Cluster 3 confidence level for the mean of CRED and scores higher than the
upper bound for CIP. Therefore it also is well placed as a member of Cluster 2.
METHODS
Members of Cluster 2 in Members of Cluster 2 in k- Members of Cluster 2 in k-

both Ward’s and k-means means and Cluster 1 in means and Cluster 3 in
methods Ward’s Ward’s
Russia Colombia Costa Rica
Argentina Indonesia
Brazil
India
Mexico
Turkey
Membership to Cluster 3 remains stable in both clustering methods, with only Costa
Rica changing from Ward’s Cluster 3 to k-means Cluster 2.
353
METHODS
Members of Cluster 3 in both Ward’s Members of Cluster 2 in k-means and

and k-means methods Cluster 3 in Ward’s
Oman Costa Rica
Vietnam
Chile
Jordan
Lebanon
Panama
Tunisia
The smallest and best performer Cluster 4 keeps a stable membership, with Malaysia,
South Africa and Thailand.
METHODS
Members of Cluster 4 in both Ward’s

and k-means methods
Malaysia
South Africa
Thailand
8.3.6.4 Comparison of clustering results with results of PCA
Clustering structures without taking logs for some of the variables show a worse match
with the results of principal components analysis than clustering structures taking logs
for some variables.
354
8.3.7 Annex 7: Cluster analysis tables
TABLE 73- CLUSTER DESCRIPTIVES- HIERARCHICAL CLUSTER ANALYSIS FIRST ITERATION
95% Confidence
Interval for Mean
Std. Lower Upper
N Mean Deviation Std. Error Bound Bound Minimum Maximum
Zscore(G 1 10 1.050644 .595684 .1883720 .6245170 1.476771 -.01558 2.04636
DPlog) 2 4 3
2 12 1.541886 .619543 .1788467 1.1482479 1.935525 .45748 2.72249
9 1 8
3 20 .0973610 .491296 .1098571 -.1325727 .3272947 -.57940 1.09696
7
4 10 .2288109 .355004 .1122621 -.0251437 .4827655 -.20081 1.07013
5
Tot 52 .6393156 .794133 .1101264 .4182276 .8604037 -.57940 2.72249
5
Zscore(G 1 10 .4802810 .400691 .1267097 .1936436 .7669185 -.11359 1.21564
DPpclog) 9
2 12 1.124773 .809274 .2336175 .6105843 1.638961 -.43692 2.94943
0 3 8
3 20 - .492642 .1101582 -.8535647 -.3924371 -1.29083 .23123
.6230009 3
4 10 .8927090 .315341 .0997198 .6671271 1.118290 .44104 1.30290
1 9
Tot 52 .2839838 .920163 .1276036 .0278088 .5401589 -1.29083 2.94943
5
Zscore(C 1 10 .0247811 .410262 .1297364 -.2687031 .3182654 -.71992 .62635
REDlog) 7
2 12 .9070632 .944820 .2727460 .3067531 1.507373 -.95744 2.16980
9 3
3 20 - .784079 .1753255 -.4529664 .2809547 -1.18867 1.82112
.0860059 3
355
4 10 .7487883 .718794 .2273028 .2345935 1.262983 -.40493 1.63062
7 1
Tot 52 .3250065 .860990 .1193979 .0853051 .5647079 -1.18867 2.16980
6
Zscore(I 1 10 - .546492 .1728162 -.4553840 .3264910 -.74596 .85760
PR) .0644465 5
2 12 .9912245 .706525 .2039562 .5423197 1.440129 -.04440 2.76181
9 3
3 20 - .784377 .1753921 -1.0028115 -.2686115 -2.24928 .65715
.6357115 5
4 10 .5268629 .757469 .2395329 -.0149984 1.068724 -1.24707 1.25848
8 1
Tot 52 .0731659 .964510 .1337535 -.1953555 .3416874 -2.24928 2.76181
5
Zscore(C 1 10 - .736934 .2330389 -.8750581 .1792835 -1.56298 .98348
IP) .3478873 9
2 12 1.251886 1.179344 .3404474 .5025664 2.001206 -.98160 3.21599
3 8 1
3 20 - .585882 .1310073 -.6395303 -.0911274 -1.28391 .55326
.3653288 1
4 10 .2893090 .752283 .2378928 -.2488420 .8274600 -1.30717 1.49510
3
Tot 52 .1371207 1.032299 .1431542 -.1502734 .4245148 -1.56298 3.21599
3
Zscore(R 1 10 .1484772 .455195 .1439454 -.1771500 .4741045 -.43893 .91139
EACPClo 7
g) 2 12 .2811379 1.007594 .2908674 -.3590570 .9213327 -.69620 2.89471
0
3 20 .0027683 .498915 .1115608 -.2307311 .2362678 -1.12187 .80406
0
4 10 - .821086 .2596502 -.6070694 .5676698 -1.16860 1.32645
.0196998 0
Tot 52 .0907076 .694158 .0962625 -.1025475 .2839627 -1.16860 2.89471
2
Zscore(F 1 10 .9836529 .247976 .0784171 .8062609 1.161044 .52253 1.38156
OSSILP 9 9
Clog) 2 12 1.183454 .748957 .2162055 .7075897 1.659319 .44005 3.05691
8 1 9
3 20 - .692769 .1549078 -.7019012 -.0534493 -1.08235 .76638
.3776753 8
4 10 - .738475 .2335264 -.9673708 .0891765 -1.08235 .73743
.4390972 9
Tot 52 .2325675 .982207 .1362076 -.0408808 .5060158 -1.08235 3.05691
7
Zscore(L 1 10 .0717940 .398600 .1260485 -.2133476 .3569356 -.73489 .48905
OG) 3
2 12 1.597084 .562102 .1622650 1.2399421 1.954227 .60031 2.40841
9 0 8
3 20 - .544220 .1216913 -.2760954 .2333104 -1.17996 .93412
.0213925 2
4 10 .5474622 .808174 .2555671 -.0306708 1.125595 -.84616 1.99116
3 2
Tot 52 .4794179 .864613 .1199003 .2387080 .7201279 -1.17996 2.40841
3
TABLE 74- TUKEY POST-HOC TEST, HIERARCHICAL ANALYSIS, FIRST ITERATION
Mean 95% Confidence Interval

Dependent (I) Ward (J) Ward Difference (I- Lower Upper
Variable Method Method J) Std. Error Sig. Bound Bound
Zscore(GD 1 2 -.49124270 .22399201 .140 -1.0873688 .1048834
Plog) *
3 .95328319 .20260840 .000 .4140668 1.4924996
356
*
4 .82183327 .23395203 .005 .1991998 1.4444667
2 1 .49124270 .22399201 .140 -.1048834 1.0873688
*
3 1.44452589 .19102103 .000 .9361478 1.9529040
*
4 1.31307597 .22399201 .000 .7169498 1.9092021
*
3 1 -.95328319 .20260840 .000 -1.4924996 -.4140668
*
2 -1.44452589 .19102103 .000 -1.9529040 -.9361478
4 -.13144992 .20260840 .915 -.6706663 .4077665
*
4 1 -.82183327 .23395203 .005 -1.4444667 -.1991998
*
2 -1.31307597 .22399201 .000 -1.9092021 -.7169498
3 .13144992 .20260840 .915 -.4077665 .6706663
*
Zscore(GD 1 2 -.64449201 .23252020 .038 -1.2633148 -.0256692
*
Ppclog) 3 1.10328192 .21032243 .000 .5435356 1.6630282
4 -.41242796 .24285942 .336 -1.0587673 .2339114
*
2 1 .64449201 .23252020 .038 .0256692 1.2633148
*
3 1.74777393 .19829389 .000 1.2200401 2.2755078
4 .23206405 .23252020 .751 -.3867588 .8508869
*
3 1 -1.10328192 .21032243 .000 -1.6630282 -.5435356
*
2 -1.74777393 .19829389 .000 -2.2755078 -
1.2200401
*
4 -1.51570989 .21032243 .000 -2.0754562 -.9559636
4 1 .41242796 .24285942 .336 -.2339114 1.0587673
2 -.23206405 .23252020 .751 -.8508869 .3867588
*
3 1.51570989 .21032243 .000 .9559636 2.0754562
*
Zscore(CR 1 2 -.88228210 .32506264 .044 -1.7473948 -.0171694
EDlog) 3 .11078701 .29403022 .982 -.6717369 .8933109
4 -.72400713 .33951685 .158 -1.6275879 .1795736
*
2 1 .88228210 .32506264 .044 .0171694 1.7473948
*
3 .99306911 .27721435 .004 .2552985 1.7308397
4 .15827497 .32506264 .962 -.7068377 1.0233877
3 1 -.11078701 .29403022 .982 -.8933109 .6717369
*
2 -.99306911 .27721435 .004 -1.7308397 -.2552985
*
4 -.83479414 .29403022 .032 -1.6173180 -.0522702
4 1 .72400713 .33951685 .158 -.1795736 1.6275879
2 -.15827497 .32506264 .962 -1.0233877 .7068377
*
3 .83479414 .29403022 .032 .0522702 1.6173180
*
Zscore(IP 1 2 -1.05567101 .30920849 .007 -1.8785899 -.2327521
R) 3 .57126501 .27968960 .187 -.1730932 1.3156232
4 -.59130939 .32295773 .272 -1.4508202 .2682014
*
2 1 1.05567101 .30920849 .007 .2327521 1.8785899
*
3 1.62693601 .26369388 .000 .9251484 2.3287236
4 .46436161 .30920849 .444 -.3585573 1.2872805
3 1 -.57126501 .27968960 .187 -1.3156232 .1730932
*
2 -1.62693601 .26369388 .000 -2.3287236 -.9251484
*
4 -1.16257440 .27968960 .001 -1.9069326 -.4182162
4 1 .59130939 .32295773 .272 -.2682014 1.4508202
2 -.46436161 .30920849 .444 -1.2872805 .3585573
*
3 1.16257440 .27968960 .001 .4182162 1.9069326
*
Zscore(CI 1 2 -1.59977359 .34852142 .000 -2.5273188 -.6722283
P) 3 .01744150 .31524948 1.000 -.8215547 .8564378
4 -.63719631 .36401875 .310 -1.6059857 .3315931
*
2 1 1.59977359 .34852142 .000 .6722283 2.5273188
*
3 1.61721509 .29722006 .000 .8262018 2.4082283
*
4 .96257728 .34852142 .039 .0350320 1.8901225
3 1 -.01744150 .31524948 1.000 -.8564378 .8215547
*
2 -1.61721509 .29722006 .000 -2.4082283 -.8262018
4 -.65463782 .31524948 .175 -1.4936341 .1843584
4 1 .63719631 .36401875 .310 -.3315931 1.6059857
*
2 -.96257728 .34852142 .039 -1.8901225 -.0350320
3 .65463782 .31524948 .175 -.1843584 1.4936341
Zscore(RE 1 2 -.13266062 .30168820 .971 -.9355653 .6702440
357
ACPClog) 3 .14570892 .27288725 .950 -.5805456 .8719635
4 .16817704 .31510305 .950 -.6704295 1.0067836
2 1 .13266062 .30168820 .971 -.6702440 .9355653
3 .27836954 .25728056 .702 -.4063498 .9630889
4 .30083765 .30168820 .752 -.5020670 1.1037423
3 1 -.14570892 .27288725 .950 -.8719635 .5805456
2 -.27836954 .25728056 .702 -.9630889 .4063498
4 .02246811 .27288725 1.000 -.7037865 .7487227
4 1 -.16817704 .31510305 .950 -1.0067836 .6704295
2 -.30083765 .30168820 .752 -1.1037423 .5020670
3 -.02246811 .27288725 1.000 -.7487227 .7037865
Zscore(FO 1 2 -.19980191 .28152344 .893 -.9490406 .5494368
*
SSILPClog 3 1.36132813 .25464753 .000 .6836162 2.0390400
) *
4 1.42275003 .29404164 .000 .6401957 2.2053043
2 1 .19980191 .28152344 .893 -.5494368 .9490406
*
3 1.56113003 .24008400 .000 .9221771 2.2000829
*
4 1.62255194 .28152344 .000 .8733132 2.3717906
*
3 1 -1.36132813 .25464753 .000 -2.0390400 -.6836162
*
2 -1.56113003 .24008400 .000 -2.2000829 -.9221771
4 .06142191 .25464753 .995 -.6162900 .7391338
*
4 1 -1.42275003 .29404164 .000 -2.2053043 -.6401957
*
2 -1.62255194 .28152344 .000 -2.3717906 -.8733132
3 -.06142191 .25464753 .995 -.7391338 .6162900
*
Zscore(LO 1 2 -1.52529096 .25036259 .000 -2.1915990 -.8589829
G) 3 .09318647 .22646148 .976 -.5095119 .6958848
4 -.47566825 .26149520 .277 -1.1716043 .2202678
*
2 1 1.52529096 .25036259 .000 .8589829 2.1915990
*
3 1.61847743 .21350994 .000 1.0502480 2.1867069
*
4 1.04962272 .25036259 .001 .3833146 1.7159308
3 1 -.09318647 .22646148 .976 -.6958848 .5095119
*
2 -1.61847743 .21350994 .000 -2.1867069 -
1.0502480
4 -.56885471 .22646148 .071 -1.1715530 .0338436
4 1 .47566825 .26149520 .277 -.2202678 1.1716043
*
2 -1.04962272 .25036259 .001 -1.7159308 -.3833146
3 .56885471 .22646148 .071 -.0338436 1.1715530
*. The mean difference is significant at the 0.05 level
TABLE 75-CLUSTER DESCRIPTIVES WITH ANOVA, NON-HIERARCHICAL CLUSTER ANALYSIS, FIRST

ITERATION
95% Confidence
Interval for Mean
Std. Lower Upper Minimu Maximu
N Mean Deviation Std. Error Bound Bound m m
Zscore(GDPl 1 4 1.3376338 .65736927 .32868464 .2916126 2.3836550 .45748 2.04636
og) 2 10 1.6268597 .59289605 .18749019 1.2027274 2.0509920 .82362 2.72249
3 18 .0425706 .46816969 .11034865 -.1902447 .2753860 -.57940 1.06514
4 20 .5429505 .54667146 .12223945 .2871004 .7988006 -.20081 1.64772
Total 52 .6393156 .79413315 .11012645 .4182276 .8604037 -.57940 2.72249
Zscore(GDPp 1 4 1.4292980 .31379889 .15689945 .9299740 1.9286221 1.10750 1.73482

clog) 2 10 1.0298624 .84903458 .26848831 .4224997 1.6372252 -.43692 2.94943
3 18 -.6224004 .55343094 .13044492 -.8976151 -.3471856 -1.29083 .57641
4 20 .4977275 .49630528 .11097723 .2654494 .7300055 -.49372 1.28334
358
Total 52 .2839838 .92016304 .12760365 .0278088 .5401589 -1.29083 2.94943
Zscore(CRED 1 4 .3020548 .84301739 .42150870 -1.039374 1.6434836 -.95744 .80993

log) 2 10 1.1933506 .81691073 .25832985 .6089678 1.7777333 -.24319 2.16980
3 18 -.2762018 .63933899 .15069365 -.5941376 .0417340 -1.18867 .98244
4 20 .4365121 .66580026 .14887746 .1249080 .7481162 -.42470 1.82112
Total 52 .3250065 .86099091 .11939796 .0853051 .5647079 -1.18867 2.16980
Zscore(IPR) 1 4 .8325398 .45099869 .22549934 .1149002 1.5501793 .45671 1.35870
2 10 1.0279725 .75451725 .23859930 .4882234 1.5677217 -.04440 2.76181
3 18 -.6680052 .81822841 .19285829 -1.074900 -.2611098 -2.24928 .65715
4 20 .1109419 .68783671 .15380496 -.2109756 .4328594 -1.24707 1.25848
Total 52 .0731659 .96451056 .13375355 -.1953555 .3416874 -2.24928 2.76181
Zscore(CIP) 1 4 -.0019641 .70781700 .35390850 -1.128259 1.1243307 -.98160 .68116
2 10 1.5148654 1.0362922 .32770438 .7735466 2.2561842 .09978 3.21599
3 18 -.5429738 .64976550 .15315120 -.8660946 -.2198530 -1.56298 .55326
4 20 .0881503 .67159231 .15017261 -.2261646 .4024652 -1.30717 1.49510
Total 52 .1371207 1.0322998 .14315423 -.1502734 .4245148 -1.56298 3.21599
Zscore(REAC 1 4 1.4710883 .97049586 .48524793 -.0731872 3.0153638 .82312 2.89471

PClog) 2 10 -.1120165 .40936086 .12945127 -.4048556 .1808226 -.69620 .56742
3 18 .1727283 .50023317 .11790609 -.0760318 .4214884 -1.12187 .91139
4 20 -.1578251 .58927743 .13176644 -.4336154 .1179652 -1.16860 1.32645
Total 52 .0907076 .69415890 .09626252 -.1025475 .2839627 -1.16860 2.89471
Zscore(FOSS 1 4 1.7758859 .89904360 .44952180 .3453070 3.2064649 .98961 3.05691

ILPClog) 2 10 .8718981 .50565513 .15990219 .5101742 1.2336220 .23951 2.03862
3 18 -.3180003 .77998895 .18384516 -.7058797 .0698790 -1.08235 1.25502
4 20 .0997495 .87986103 .19674291 -.3120381 .5115372 -1.08235 1.22282
Total 52 .2325675 .98220747 .13620767 -.0408808 .5060158 -1.08235 3.05691
Zscore(LOG) 1 4 .8854370 .75819008 .37909504 -.3210126 2.0918866 -.03947 1.65735
2 10 1.6879533 .51766432 .16369983 1.3176386 2.0582681 .90630 2.40841
3 18 -.1538316 .53035909 .12500684 -.4175730 .1099097 -1.17996 .65595
4 20 .3638710 .58296760 .13035552 .0910337 .6367082 -.84616 1.99116
Total 52 .4794179 .86461355 .11990033 .2387080 .7201279 -1.17996 2.40841
TABLE 76- POST-HOC TUKEY TEST, NON HIERARCHICAL CLUSTER, FIRST ITERATION
Depende (I) Cluster (J) Cluster Mean 95% Confidence Interval

nt Number of Number of Difference (I- Upper
Variable Case Case J) Std. Error Sig. Lower Bound Bound
Zscore(G 1 2 -.28922588 .31795360 .800 -1.1354188 .5569670
DPlog) *
3 1.29506315 .29708096 .000 .5044201 2.0857062
*
4 .79468332 .29436783 .046 .0112609 1.5781057
2 1 .28922588 .31795360 .800 -.5569670 1.1354188
*
3 1.58428904 .21196907 .000 1.0201604 2.1484176
*
4 1.08390920 .20814949 .000 .5299459 1.6378725
*
3 1 -1.29506315 .29708096 .000 -2.0857062 -.5044201
359
*
2 -1.58428904 .21196907 .000 -2.1484176 -
1.0201604
*
4 -.50037984 .17461047 .030 -.9650833 -.0356764
*
4 1 -.79468332 .29436783 .046 -1.5781057 -.0112609
*
2 -1.08390920 .20814949 .000 -1.6378725 -.5299459
*
3 .50037984 .17461047 .030 .0356764 .9650833
Zscore(G 1 2 .39943558 .34864389 .663 -.5284356 1.3273068
*
DPpclog) 3 2.05169838 .32575652 .000 1.1847390 2.9186578
*
4 .93157056 .32278152 .029 .0725288 1.7906124
2 1 -.39943558 .34864389 .663 -1.3273068 .5284356
*
3 1.65226280 .23242926 .000 1.0336820 2.2708436
4 .53213498 .22824100 .105 -.0752993 1.1395693
*
3 1 -2.05169838 .32575652 .000 -2.9186578 -
1.1847390
*
2 -1.65226280 .23242926 .000 -2.2708436 -
1.0336820
*
4 -1.12012782 .19146464 .000 -1.6296865 -.6105691
*
4 1 -.93157056 .32278152 .029 -1.7906124 -.0725288
2 -.53213498 .22824100 .105 -1.1395693 .0752993
*
3 1.12012782 .19146464 .000 .6105691 1.6296865
Zscore(C 1 2 -.89129579 .41403276 .151 -1.9931911 .2105995
REDlog) 3 .57825654 .38685282 .449 -.4513029 1.6078159
4 -.13445735 .38331985 .985 -1.1546142 .8856995
2 1 .89129579 .41403276 .151 -.2105995 1.9931911
*
3 1.46955233 .27602184 .000 .7349555 2.2041492
*
4 .75683844 .27104806 .036 .0354786 1.4781983
3 1 -.57825654 .38685282 .449 -1.6078159 .4513029
*
2 -1.46955233 .27602184 .000 -2.2041492 -.7349555
*
4 -.71271389 .22737422 .015 -1.3178414 -.1075864
4 1 .13445735 .38331985 .985 -.8856995 1.1546142
*
2 -.75683844 .27104806 .036 -1.4781983 -.0354786
*
3 .71271389 .22737422 .015 .1075864 1.3178414
Zscore(IP 1 2 -.19543277 .43628661 .970 -1.3565539 .9656883
*
R) 3 1.50054502 .40764578 .003 .4156478 2.5854422
4 .72159790 .40392291 .292 -.3533914 1.7965872
2 1 .19543277 .43628661 .970 -.9656883 1.3565539
*
3 1.69597779 .29085774 .000 .9218970 2.4700585
*
4 .91703067 .28561663 .012 .1568985 1.6771629
*
3 1 -1.50054502 .40764578 .003 -2.5854422 -.4156478
*
2 -1.69597779 .29085774 .000 -2.4700585 -.9218970
*
4 -.77894712 .23959536 .011 -1.4165996 -.1412946
4 1 -.72159790 .40392291 .292 -1.7965872 .3533914
*
2 -.91703067 .28561663 .012 -1.6771629 -.1568985
*
3 .77894712 .23959536 .011 .1412946 1.4165996
*
Zscore(CI 1 2 -1.51682954 .44300813 .007 -2.6958391 -.3378200
P) 3 .54100964 .41392605 .563 -.5606017 1.6426210
4 -.09011444 .41014583 .996 -1.1816652 1.0014363
*
2 1 1.51682954 .44300813 .007 .3378200 2.6958391
*
3 2.05783919 .29533875 .000 1.2718328 2.8438456
*
4 1.42671510 .29001690 .000 .6548722 2.1985580
3 1 -.54100964 .41392605 .563 -1.6426210 .5606017
*
2 -2.05783919 .29533875 .000 -2.8438456 -
1.2718328
4 -.63112408 .24328662 .059 -1.2786004 .0163522
4 1 .09011444 .41014583 .996 -1.0014363 1.1816652
*
2 -1.42671510 .29001690 .000 -2.1985580 -.6548722
3 .63112408 .24328662 .059 -.0163522 1.2786004
*
Zscore(R 1 2 1.58310479 .33275701 .000 .6975145 2.4686951
*
EACPClo 3 1.29836004 .31091257 .001 .4709059 2.1258142
g) *
4 1.62891343 .30807313 .000 .8090161 2.4488107
*
2 1 -1.58310479 .33275701 .000 -2.4686951 -.6975145
360
3 -.28474475 .22183801 .578 -.8751383 .3056488
4 .04580864 .21784060 .997 -.5339463 .6255636
*
3 1 -1.29836004 .31091257 .001 -2.1258142 -.4709059
2 .28474475 .22183801 .578 -.3056488 .8751383
4 .33055339 .18274005 .282 -.1557859 .8168927
*
4 1 -1.62891343 .30807313 .000 -2.4488107 -.8090161
2 -.04580864 .21784060 .997 -.6255636 .5339463
3 -.33055339 .18274005 .282 -.8168927 .1557859
Zscore(F 1 2 .90398782 .46596939 .225 -.3361302 2.1441059
*
OSSILPC 3 2.09388628 .43537999 .000 .9351780 3.2525945
log) *
4 1.67613642 .43140383 .002 .5280102 2.8242626
2 1 -.90398782 .46596939 .225 -2.1441059 .3361302
*
3 1.18989846 .31064626 .002 .3631531 2.0166438
4 .77214861 .30504857 .068 -.0396992 1.5839964
*
3 1 -2.09388628 .43537999 .000 -3.2525945 -.9351780
*
2 -1.18989846 .31064626 .002 -2.0166438 -.3631531
4 -.41774986 .25589625 .371 -1.0987851 .2632853
*
4 1 -1.67613642 .43140383 .002 -2.8242626 -.5280102
2 -.77214861 .30504857 .068 -1.5839964 .0396992
3 .41774986 .25589625 .371 -.2632853 1.0987851
Zscore(L 1 2 -.80251631 .33483054 .091 -1.6936250 .0885924
OG) *
3 1.03926866 .31284998 .009 .2066584 1.8718789
4 .52156606 .30999284 .344 -.3034403 1.3465724
2 1 .80251631 .33483054 .091 -.0885924 1.6936250
*
3 1.84178497 .22322036 .000 1.2477125 2.4358575
*
4 1.32408237 .21919804 .000 .7407148 1.9074500
*
3 1 -1.03926866 .31284998 .009 -1.8718789 -.2066584
*
2 -1.84178497 .22322036 .000 -2.4358575 -
1.2477125
*
4 -.51770261 .18387877 .034 -1.0070725 -.0283327
4 1 -.52156606 .30999284 .344 -1.3465724 .3034403
*
2 -1.32408237 .21919804 .000 -1.9074500 -.7407148
*
3 .51770261 .18387877 .034 .0283327 1.0070725
*. The mean difference is significant at the 0.05 level.
361

Ana Pueyo Velasco

Cargado por

Información del documento

Derechos de autor

Formatos disponibles

Compartir este documento

Compartir o incrustar documentos

Opciones para compartir

¿Le pareció útil este documento?

¿Este contenido es inapropiado?

Copyright:

Formatos disponibles

Ana Pueyo Velasco

Cargado por

Copyright:

Formatos disponibles

Departamento de Ingeniería Química Industrial y Medio Ambiente

Escuela Técnica Superior de Ingenieros Industriales

Climate change technology transfer to developing

Ana Pueyo Velasco

Julio Lumbreras Martín Pedro Linares

Presidente: Dra. Encarnación Rodríguez Hurtado

Secretario: Dra. Mª Jesús Sánchez Naranjo

Vocal: Dr. José Ignacio Pérez Arriaga

Vocal: Dr. Jim Watson

Vocal: Dr. Xabier Labandeira

Suplente: Dr. Walter Leal

Suplente: Dr. Pablo Martínez de Anguita

Realizado el acto de defensa y lectura de la tesis el día 5 de marzo de 2012

EL PRESIDENTE LOS VOCALES

1.2 International policy context: technology transfer in the UNFCCC

1.2.1 Status of technology transfer in international climate change negotiations

1.2.2 The Effectiveness of existing UNFCCC technology transfer catalysts

Instrument Achievements Pitfalls

1.2.3 The Technology Mechanism

1.2.4 The Prospects of UNFCCC instruments in promoting climate change TT to developing

1.3 Objectives and research questions

2.1. Conceptualising technology transfer

2.2 Channels of technology transfer

2.3 Measuring technology transfer

2.3.1 Economics of technology diffusion

2.3.2 Vertical technology transfer: knowledge transfer metrics

2.3.3 Climate change technology transfer

2.3.4 Technology transfer metrics in the framework of the UNFCCC

2.3.5 Summary and identification of gaps

Type of Measure- Body of literature

Output Value Licence revenues

2.4.1 Economic and institutional frameworks promoting private investment

2.4.2 Technology demand factors and policies

2.4.3 Technology supply factors and policies

2.4.4 Industrial development factors

Source: OECD, 2001

2.4.5 Bilateral country characteristics

2.4.6 Summary and identification of gaps

Type of factor Enabling factor Indicator Expected

Economic and Ease of doing Ease of doing business index Positive

Protection of IPRs Index of IPRs strictness Positive

Open trade • Average tariffs Positive

Low corruption Corruption index Negative

Technology Demand size • Population Positive

Demand growth • Annual GDP growth Positive

Electricity prices Electricity prices Positive

Demand pull policies • Qualitative or dummies for: Positive

Technology Human capital • Enrolment in tertiary education Positive

R&D • General R&D spending Positive

Patent stock • Stock of patents of specific technology fields Positive

Technology • UNCTAD index of innovation capability Positive

Technology push • Government-sponsored R&D Positive

Industrial Industrial • Competitive industrial performance index Positive

Market power in • Profit levels of incumbent companies Negative

Cost of factors • Labour cost Negative

Industrial policy • Local content requirements Positive in

Bilateral Trade relationships • Trade importance (total dyad trade/sum of Positive

Geographical • Km between country pairs Negative

Language • Dummy having the same language Positive

Technological • Index about the characteristics of the Negative

3.2 Overview of Chilean energy system and GHG emissions

3.2.1 Energy consumption and GHG emissions in Chile

Total CO2 emissions from the consumption of energy (Mt )- 2009

Chile's Carbon Intensity using PPP