This article was downloaded by: [Universit du Qubec Montral]

On: 06 September 2014, At: 01:55

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Research Papers in Education

Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/rred20

Is it what works that matters?

Evaluation and evidencebased
policymaking
Ian Sanderson
Published online: 24 Jun 2010.

To cite this article: Ian Sanderson (2003) Is it what works that matters? Evaluation
and evidencebased policymaking, Research Papers in Education, 18:4, 331-345, DOI:
10.1080/0267152032000176846
To link to this article: http://dx.doi.org/10.1080/0267152032000176846

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the
Content) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to
the accuracy, completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content
should not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or
howsoever caused arising directly or indirectly in connection with, in relation to or arising
out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Conditions of access and use can be found at http://www.tandfonline.com/page/termsand-conditions

Research Papers in Education 18(4) December 2003, pp. 331345

Is it what works that matters?

Evaluation and evidence-based
policy-making

Downloaded by [Universit du Qubec Montral] at 01:55 06 September 2014

Ian Sanderson

ABSTRACT
The notion of evidence-based policy-making (EBP) has gained renewed currency in the UK
in the context of the current Labour Governments commitment to modernise government.
Thus, a key driver of modernisation is seen as evidence based policy-making and service
deliverywhat matters is what worksin the context of a performance management
strategy for regulation of public services. The aim of this paper is to critically examine the
assumptions underpinning EBP asking, in particular, the extent to which the increased
emphasis on the role of evidence in policy-making is indicative of instrumental rationality
which erodes the normative basis of policy-making and undermines the capacity for
appropriate practice. The potential for theory-based evaluation to deliver on its evidential
promise is critically examined and, based upon an expanded notion of practical reason, it is
argued that we need to extend the scope of our concern from what works to what is
appropriate in addressing complex and ambiguous social problems, embracing ethical-moral
concerns.
Keywords: evidence; policy; evaluation; theory-based; rationality

. . . (I)n the degree in which an active conception of knowledge prevails . . . (c)hange

becomes significant of new possibilities and ends to be attained; it becomes prophetic of
a better future. Change is associated with progress rather than with lapse and fall. Since
changes are going on anyway, the great thing is to learn enough about them so that we be
able to lay hold of them and turn them in the direction of our desires. (Dewey, 1957,
p. 116.)

Ian Sanderson is Director and Professor of Policy Analysis and Evaluation, Policy Research
Institute, Leeds Metropolitan University.

Research Papers in Education ISSN 0267-1522 print/ISSN 1470-1146 online 2003 Taylor & Francis Ltd
http://www.tandf.co.uk/journals
DOI: 10.1080/0267152032000176846

Downloaded by [Universit du Qubec Montral] at 01:55 06 September 2014

INTRODUCTION
The role of scientific knowledge in public policy-making is an issue of enduring concern to
both policy makers and researchers, a legacy of the Enlightenment in the seventeenth century
when the focus shifted to the potential for the application of knowledge to change and
improve the world. A key figure in what A. N. Whitehead called the century of genius was
Francis Bacon who, Zagorin (1998, p. 222) argues, . . . gave science an ethos and social
function, the investigation of nature for human betterment which, if never universally
accepted, continues to be very widely regarded up to the present day as its ultimate rationale.
John Dewey, in his quest for a philosophical basis for the capacity to turn change in the
direction of our desires, was a passionate advocate of Bacons systematic critical empiricism
founded upon careful experimentation to . . . force the apparent facts of nature . . . (to) . . .
tell the truth about themselves, as torture may compel an unwilling witness to reveal what he
(sic.) has been concealing (Dewey, 1957, p. 32).
In contrast to Deweys optimism about the role of reason in attaining the better and
averting the worse (Dewey, 1993), the pessimistic diagnosis of some postmodernists about
a world out of control and beyond the reach of redemptive politics (OSullivan, 1993)
appears rather bleak. Of course, there are two issues here. The first concerns our capacity to
understand the worldto wrest those facts from the uncooperative witness; the second
concerns our ability to apply those facts to influence the direction of social change in the
context of wider political drivers. Enough has been written about policy-making processes to
show that ideas rarely win the battles, although there remains a deep-seated optimism about
their capacity to help win the wars.
The position adopted in this paper is one of constructive scepticism in a critical appraisal
of the present Governments pre-occupation with evidence-based policy-making (EBP). Tony
Blair set the agenda shortly after being elected for his first term of office by outlining the
commitment to modernise the public sector and declaring that what counts is what works.
In other words, the old age of ideologically-driven politics was to be consigned to the dustbin
of history and a new age of modern policy-making would be driven by research evidence of
what was proven to be effective in addressing social problems and achieving the desired
outcomes. In key policy areas such as crime, education and welfare-to-work, we continually
hear the Prime Minister and other ministers talk about their commitment to finding out what
works.
Now, for an applied social researcher, this position raises a dilemma. It is music to the ears
of the applied researcher committed to the modernist belief in social progress informed by
reason and immersed in contracted research from government departments and agencies
specifically intended to contribute to the development and improvement of public policies.
However, the music is mixed with distant alarm bells in the ears of the academic social
scientist influenced by the work of, for example, Jurgen Habermas, John Dryzek and Frank
Fischer (Sanderson, 1998, 1999). Is this emphasis on knowledge and expertise in policymaking a form of instrumental rationality, focusing on deriving correct means to given ends
at the expense of consideration of the appropriateness of those ends? Does it signal the
devaluing of democratic debate about the ethical and moral issues raised by policy choices?
How much does what works matter? How much emphasis can and, indeed, should be
placed on scientific evidence of what works in decisions about policies to address social
problems? This paper seeks to address these questions by the following route. First, I will refer
briefly to the historical context of EBP and outline in a little more detail the present

332 Research Papers in Education Volume 18 Number 4

Governments stance. I then focus on the role of evaluation in providing the evidence
required, asking to what extent it can tell us what works. Based upon the conclusion that the
rationalist conception of EBP has some rather shaky foundations, I question just how much
what works really does matter in the context of a broader conception of policy-making. In
the concluding part, I elaborate briefly a proposition on what does matter in making public
policy choices.

Downloaded by [Universit du Qubec Montral] at 01:55 06 September 2014

WHAT WORKS: THE NEW HOLY GRAIL OF POLICY-MAKING?

Faith in the power of reason to guide human affairs, of course, has long been regarded as a
central feature of modern society, gaining particular salience during the period of the
Enlightenment. In the early seventeenth century, Francis Bacon was influential in challenging
authority and tradition and emphasising the importance of experimentation and inductive
reasoning to derive scientific knowledge capable of providing the foundation for efforts to
improve human welfare (Zagorin, 1998). The philosophical basis of this challenge was
strengthened by Kants notion of . . . a pure reason, held universally by all rational beings . . .
employed to bring improvement to human society (Trigg, 2001, p. 222). Newtons laws of
motion provided an exemplar of the power of the new science, promising control over nature
and providing the basis for optimism about social and political progress (Bronk, 1998).
In spite of many challenges, most recently from post-modernists, this basic optimism about
the role of scientific knowledge remains embedded in Western liberal democratic political
systems. In the 1960s, the social sciences became an institutionalised component of
government policy-making, being seen as a key basis for more effective policy (Wagenaar,
1982). Indeed, Martin Bulmer (1987, p. 349) has argued that there was considerable overoptimism at this time about the potential role of social science based upon . . . an almost
euphoric sense that social science really could change the world. This optimism was shaken
in the 1970s and 1980s by a growing scepticism based upon a lack of evidence that social
science research was actually influencing policy decisions. This scepticism was reinforced by
the Thatcher Governments stance, founded upon Sir Keith Josephs famous antipathy to social
science, characterised by Bulmer (1987) as . . . a general hostility to the social sciences and
social research, with tinges of philistinism . . ..
Nevertheless, it is doubtful whether this scepticism about the social sciences inflicted any
lasting damage on the underlying faith in the scientific enterprise in the broader sense. Social
scientists were meanwhile busy seeking to rescue the notion of evidence-based policy through
work that indicated that the influence of research was much more complex than previously
supposed. Carol Weisss work in this area has been highly influential in highlighting the
limitations of instrumental notions of research use and emphasising its broader enlightenment function, influencing the conceptualisation of issues, the range of options considered
and challenging taken-for-granted assumptions about appropriate goals and activities (Weiss,
1982). Based upon this work social scientists could proceed, comfortable in the knowledge
that while their research might not have an immediate direct impact on policy decisions, it
would nevertheless percolate into the policy arena through a range of informal routes and
have a longer-term influence.
In the resurgence of enthusiasm for EBP that has accompanied the accession to power of
New Labour, such notions of enlightenment and conceptual functions of research have been

Is it what works that matters? 333

Downloaded by [Universit du Qubec Montral] at 01:55 06 September 2014

overtaken by a renewed optimism about achieving more direct and instrumental use of
research in policy-making processes. These were ushered in by the Prime Ministerial
declaration that what counts is what works (Powell, 1999, p. 23). In this context, what
works can be taken as referring essentially to types of government intervention that are
effective in addressing the problems at which they are directed and achieving their intended
outcomes and effects. The Governments preoccupation with EBP has developed in the
context of a model of modern, professional policy making that has been propounded by the
Cabinet Office:
Good quality policy-making depends on high quality information and evidence. Modern
policy-making calls for the need to improve Departments capacity to make best use of
evidence, and the need to improve the accessibility of evidence available to policy-makers
(Bullock et al., 2001, p. 25).
Modern policy-making lies at the heart of the modernising government agenda, which
is seeking to make government more responsive and effective in achieving results (Cabinet
Office, 1999; Sanderson, 2001, 2002). The emphasis is very much on results, expressed in the
form of measurable targets in government departments Public Service Agreements (PSAs)
with the Treasury (HM Treasury, 2000). In a strong performance management regime,
departments are accountable for achievement against their targets and considerable emphasis
is given to evaluation . . . showing what worked well in improving public services and why,
and considering what further practical steps were needed to enhance service delivery and
improve effectiveness (National Audit Office, 2001). This is indeed now a familiar plea
indicating the predominance of an instrumental view of evaluation in EBP.
The underpinning rationale of the Governments position on evidence-based policymaking was articulated a couple of years ago by David Blunkett, then Secretary of State for
Education and Employment, in a much-quoted lecture to the Economic and Social Research
Council (ESRC) (DfEE, 2000). He argued that . . . rational thought is impossible without
good evidence . . . social science research is central to the development and evaluation of
policy (DfEE, 2000, p. 24); emphasised the Governments . . . clear commitment that we will
be guided not by dogma but by an open-minded approach to understanding what works and
why . . . (DfEE, 2000, p. 2); and expressed his passionate belief that . . . having ready access
to the lessons from high quality research can and must vastly improve the quality and
sensitivity of the complex and often constrained decisions we, as politicians, have to make
(DfEE, 2000, p.4).
In this rationalist vision, there is little sympathy for the broader enlightenment function
of social research, much of which is seen as . . . inward looking, too piecemeal . . . too
supplier-driven rather than focusing on the key issues of concern to policy makers . . .
(DfEE, 2000, p.8). Blunkett voiced concerns about the focus, relevance and timeliness of
research and indicated a strong desire for more research that is directly accessible, intelligible
and relevant to users in the policy community. The emphasis, therefore, is on enhancing
instrumental use. Whilst acknowledging the place for fundamental blue-skies research, he
placed the major emphasis upon research with practical applications. He highlighted the
need for research that . . . leads to a coherent picture of how society works: what are the
main forces at work and which of these can be influenced by government . . .; the need
. . . to be able to measure the size of the effect of A on B . . .; and the . . . huge potential
for quantitative analysis. . . (DfEE, p.22).

334 Research Papers in Education Volume 18 Number 4

Downloaded by [Universit du Qubec Montral] at 01:55 06 September 2014

It is easy to be cynical about politicians advocating the use of evidence in decision-making

when so much policy remains driven by the exigencies of the political process. However, if
we accept his arguments at face value, he can be seen as tapping into a rich seam of received
wisdom; as Robert Walker (2001, p. 307) argues, . . . few would quibble with Blunketts
aspiration to make social science research evidence central to the development and evaluation
of policy . . . Thus, on one level, it is self-evident that ideologically-driven policies potentially
represent a waste of public resources and that there is indeed a strong ethical case for ensuring
that scarce resources are directed into those policies that the evidence suggests are likely to be
most effective in addressing social problems (Judge and Bauld, 2001).
However, there are those who argue that the relationship between research evidence and
better policy decisions is not so self-evident. There are two main grounds for criticism. The
first questions the feasibility of the projectto what extent can evidence provide the basis for
better policy? The second questions its desirabilityto what extent should policy be guided by
evidence? In the next section I will address the feasibility issue in a discussion of some aspects
of the role of evaluation in EBP; in the following section I will turn to the desirability issue
in a discussion of how much what works does matter.

HOW MUCH CAN EVALUATION TELL US ABOUT WHAT WORKS?

It is clear, then that evaluation is seen as playing a major role in providing the required robust
evidence of what works. To what extent can evaluation deliver on this requirement? Of
course, this question raises fundamental issues in the philosophy and methodology of social
science to which I cannot do justice here. We could quickly become immersed in the
paradigm wars between positivists, social constructionists and realists and in this context we
cannot go there. I propose more modestly to focus on the current emphasis in the policy
evaluation community on theory-based evaluation as a means of deriving knowledge of causal
relationships which are seen as the key to understanding how and why policy initiatives work
in producing desired or intended social change (Davies et al., 1999; Walker, 2001).
It would appear that theory-based evaluation (TBE) has gained considerably increased
currency through its perceived contribution to EBP. The role of evaluation generally in this
context has been considered recently by the National Audit Office (2001, p. 25), which argues
that the appropriate response to increasing complexity in the policy making environment is
. . . to apply more powerful tools and draw on more specialist knowledge to enhance . . .
capacity to design and implement successful policies. Reliable and comprehensive
information and sound analysis is crucial to the understanding of problems and the need for
policy intervention and . . . helps to establish what works and to identify optimum
opportunities for intervention (National Audit Office, 2001, p. 67, 26). Considerable
emphasis is given in this model of modern policy making to the role of evaluation in learning
lessons:
Evaluation is important for determining the extent to which a policy has met or is meeting
its objectives and that those intended to benefit have done so. Evaluation can also help
departments learn lessons and share good practice in policy design and implementation.
For long-term policies, evaluation can identify ways in which the policy can be improved
or developed to increase its impact (National Audit Office, 2001, p. 14).

Is it what works that matters? 335

Downloaded by [Universit du Qubec Montral] at 01:55 06 September 2014

In their review of policy-making in various government departments, the NAO

highlighted a need for evaluation to be more practical, . . . showing what worked well in
improving public services and why, and considering what further practical steps were needed
to enhance service delivery and improve effectiveness (National Audit Office, 2001). This is
indeed now a familiar plea indicating the predominance of an instrumental view of
evaluation.
In the context of this agenda, considerable attention is now being focused on the promise
of TBE to deliver knowledge of the causal relationships that are seen as the key to
understanding what works. Recent work has emphasised the importance of improving our
theoretical knowledge of how and why different approaches to policy intervention work in
producing desired or intended social change (Davies et al., 2000; Walker, 2001). Pawson and
Tilley (1997) established the new mantra of evaluatorswhat works, for whom, in what
circumstances, and whywhich derives from the realist concern to understand how policy
mechanisms operate in different contextual circumstances to change outcomes. Of course,
this resonates well with the Governments vision of EBP. In this context there has been
increasing criticism of the instrumental rationality of goal-oriented and method-oriented
evaluation, which de-contextualises evaluation, restricts the analysis of potential effects, and
fails to provide an understanding of how interventions produce such effects (Chen, 1990).
Therefore, TBE is seen as a primary means of generating the causal knowledge required for
more effective policy design and implementation by testing the validity of the causal
assumptions upon which policy initiatives and programmes are based (Weiss, 1995). However,
while undoubtedly holding out considerable promise for the development of knowledge to
underpin more effective policy-making, TBE presents significant challenges in terms of
articulating theoretical assumptions and hypotheses, measuring changes and effects,
developing appropriate tests of assumptions and hypotheses, and in terms of the
generalisability of results obtained in particular contexts (Weiss, 1995). It is not appropriate to
discuss the technical arguments here (see Connell et al., 1995; Fulbright-Anderson et al.,
1998; Rogers et al., 2000) but I do want to say something about causal knowledge since this
is central to the promise of TBE.
Even if we set aside the potential difficulties that can arise due to the specification and
empirical measurement of theoretical constructs, substantial problems surround the process of
theory testing such that even in relatively simple policy interventions it is extremely difficult
to derive robust causal attribution. Even such simple interventions can involve a large number
of links in the theoretical chaininterlinked assumptions about how they will achieve their
intended effectsand in practice evaluations usually either have to be highly selective in the
analysis of causal links or have to parcel them up (Weiss, 2000). Scriven (1998), makes the
point that evaluators are not usually in the business of seeking deep explanations. In practice,
the causal linkages between the observed intervention and the observed effect will remain
something of a black box and our capacity as evaluators to open this up and identify how the
intervention works in terms of causal mechanisms is quite limited, especially when
interventions become more complex (Granger, 1998). Many of our key social problems
require such complex, cross-cutting interventions.
Realists argue that it is possible to derive causal understanding through direct analysis of
how theorised mechanisms work in particular contexts to produce outcomes (Pawson and
Tilley, 1997). Experimentalists counter that this approach does not provide a valid
counterfactual, which is required for robust inference that the outcomes are genuine
products of the intervention and would not have occurred in its absence (Cook, 2000).

336 Research Papers in Education Volume 18 Number 4

Downloaded by [Universit du Qubec Montral] at 01:55 06 September 2014

However, while experiments can provide a basis for causal inference in the limited
circumstances where a uniform, standardised intervention is provided for a clearly defined
target group under tightly-controlled conditions, this is based solely on the successionist
assumption that effect must follow the intervention if all other possible causes are controlled
out. They do not provide an understanding of how the effects are produced thus creating
problems of external validitythe ability to generalise from the result to other situations.
Although there remain matters of philosophical dispute, I would line up behind advocates
of the need to use the full range of methodological approaches to provide the best hope of
capturing various facets of policy interventions and piecing together a picture of how they
produce change in particular contextual circumstances (Sanderson, 2000). This is a stance of
very modest expectations in contrast to the bullish claims of some proponents of TBE made
in the context of rationalist conceptions of evidence-based policy-making. Even Carol Weiss
(1997) refers to the aggressively rationalistic stance of TBE but her overall position represents
a cautious and pragmatic response seeking the middle way between the horns of the Cartesian
dilemma. It is worth quoting her at length:
In my most optimistic moments, I succumb to the notion that evaluations may be able to
pin down which links in which theories are generally supported by evidence and that
program designers can make use of such understanding in modifying current programs and
planning new ones . . . Such hopes are no doubt too sunny. Given the astronomical variety
of implementations of even one basic program model, the variety of staffs, clients,
organizational contexts, social and political environments, and funding levels, any hope for
deriving generalizable findings is romantic. Nevertheless, theory-based evaluation can add
to knowledge. Even relatively small increments of knowledge about how and why
programmes work or fail cannot help but improve program effectiveness. And that is what
program evaluation is all about (Weiss, 2000, p. 44).
However, not all reactions to this problem are so pragmatic. I think we can legitimately
question whether the use of public money in resource-intensive theory-based evaluations is
appropriate if they are producing relatively small increments of knowledge. Improving the
knowledge base for policy intervention is clearly important but has to be set against other calls
on scarce public resources in terms of the value added for society. It will be interesting to see
how this issue plays out as we get more experience with TBE and as it becomes clearer just
how much it can deliver robust evidence of what works.
Of course, in the extreme, postmodernists will reject this whole agenda of attempting to
enhance the role of reason in the guidance of human affairs but many who acknowledge
the need to strive for social improvement reject such a nihilistic position (Trigg, 2001;
Oakley, 2000). Nevertheless, some critics take issue with the technical-rationalist
orientation represented by TBE. This stance is represented by Thomas Schwandt, who
regards evaluation in its dominant guise essentially as modernist project designed . . . to
tame the unruly social world, to bring order to our way of thinking about what does and
does not work for improving social life (Schwandt, 2000a). In this project, rationality is a
matter of correct procedure or method in a context where . . . policymakers seek to
manage economic and social affairs rationally in an apolitical, scientized manner such that
social policy is more or less an exercise in social technology (Schwandt, 1997, p. 74). He
sees the problem in the domination of technical rationality and expertise, which subsumes
issues of moral-political judgement. What he finds objectionable is the belief that practice

Is it what works that matters? 337

Downloaded by [Universit du Qubec Montral] at 01:55 06 September 2014

will somehow be less ambiguous or more rational if we can only find the right ways to
generate and apply evaluation knowledge (Schwandt, 2000a). He argues that the application
of scientific method to contemporary life can lead to the deformation of praxis, and
maintains that social scientific knowledge (which is general and theoretical) cannot provide
the primary basis for evaluation judgements under conditions of plurality, uncertainty and
difference (Schwandt, 2000b).
While acknowledging the role to be played by such knowledge, Schwandt argues that
evaluators must recognise the need for practical knowledge or wisdom to help practitioners
understand their practice better and to make wise moral-political judgements:
While of course we would like to have at our disposal the best general, scientific
knowledge we can acquire, the corrigibility, ambiguity, and circumstantiality of everyday
evaluative judgement cannot be eliminated, replaced or refined by relying on scientific
method and its associated rationality (Schwandt, 2000b).
Central to Schwandts perspective on evaluation is the notion of critical intelligence:
Critical intelligence . . . is the ability to question whether the . . . (end) . . . is worth
getting to. It requires not simply knowledge of effects, strategies, procedures and the like
but the willingness and capacity to debate the value of various ends of a practice. It requires
acknowledging and understanding the force of tradition (prejudices) in shaping the
conceptualisation of those ends, the means used to frame those ends, and the practices
employed to assess their effects and a simultaneous effort to transform that knowledge in
the process of coming to that understanding. This is fundamentally an exercise in practicalmoral reasoning. (Schwandt, 1997, p. 79.)
From this perspective, what matters is not so much what works as what is appropriate in
particular circumstances and evaluation is not merely technique involving robust objective
analysis but rather more craft activity involving reasoned judgement of various forms of
knowledge and normative implications. This leads us, therefore, to the issue of the desirability
of basing policy decisions on research evidenceto question just how much what works
does matter.

HOW MUCH DOES WHAT WORKS MATTER?

Schwandts conception of evaluation is consistent with analyses of the broader policy-making
process as more a communicative process based upon dialogue and argumentation than a
technical process based upon evidence. A key work in this tradition is by Giandomenico
Majone (1989) who takes issue with the . . . rationalist fallacy of believing that theories and
ideas alone are powerful enough to determine the course of events, and of interpreting policymaking as a purely intellectual exercise . . . (Majone, 1989, p.146) and criticises decisionism
as neglecting the role of social processes and rules. He presents an alternative conception of
policy making as a process of deliberation and argumentation, drawing on Aristotle for his
view of policy analysis as craft work, and emphasising the role of argument. . . a complex
blend of factual statements, interpretations, opinions, and evaluations (Majone, 1989, p.63).
In this perspective, the moral/ethical dimension figures prominently:

338 Research Papers in Education Volume 18 Number 4

Downloaded by [Universit du Qubec Montral] at 01:55 06 September 2014

. . . the choice of policy instrument is not a technical problem that can be safely delegated
to experts. It raises institutional, social and moral issues that must be clarified through a
process of public deliberation and resolved by political means (Majone, 1989, p. 143).
In Majones view, the notion of EBP places too much emphasis on the potential role of
causal knowledge in improving policy effectiveness and insufficient emphasis on the
normative, institutional and organisational context in which decisions and choices are made
and action is taken.
Alasdair MacIntyre (1984) also draws on Aristotle in his critique of the notion that rational
action is an essentially technical matter of applying morally neutral expertise as the means for
achieving authoritative ends. He argues that this notion neglects the intrinsic virtuessuch
standards as fairness, truthfulness, trust and honestythat are embodied in human practices
and which give such practices an inherently moral and ethical character. He defines a practice
as a cooperative human activity that has its own standards of excellence. In striving to achieve
these standards, practitioners realise internal goods, which are satisfactions intrinsic to the
activity derived from doing it well. A virtue, then, is . . . an acquired human quality the
possession and exercise of which tends to enable us to achieve those goods which are internal
to practices . . . (MacIntyre, 1984, p. 191). The implication of this position is that the model
of rationality underpinning evaluation should not be based upon an instrumental notion of
the effectiveness of means to given ends but rather upon a practical notion of the appropriateness
of action from a broader ethical-moral standpoint.
The philosophical basis for such a position is traceable back to Aristotle and the preEnlightenment acceptance that scientific reasoning and expertise was but one amongst many
legitimate bases for belief and actiona situation that Stephen Toulmin (2001, p. 29) terms
the Balance of Reason. Toulmin argues that this balance was disrupted in the seventeenth
century as, under the influence of Galileo and Descartes, exact sciences susceptible to
mathematical methods, theoretical abstraction and logical deduction gained ascendancy as
apparent means of overcoming the uncertainties and ambiguities that had previously been
accepted. Thus, began what John Dewey called the quest for certainty, which culminated in
Newtonian physics being taken as showing that the Solar System was an exemplar of a
rationally intelligible system demonstrating . . . regularity, uniformity, and above all stability
(Toulmin, 2001, p. 48). Toulmin argues that the hegemony of Newtonian dynamics, bolstered
by its . . . intellectual coherence with a respectable picture of Gods Material Creation . . .
(Toulmin, 2001, p.79), provided social sciences with an unrealistic and, indeed, distorting
exemplar of Serious Science, which has blinded them to matters of political and moral
concern and practical relevance:
. . . (A) traditional reliance on Euclidian and Newtonian models of theory continues to
focus attention on doing your sums right and conceals the equally important task of
making sure that you are doing the right sums; in other words, doing calculations that are
directly relevant to the practical situation in question (Toulmin, 2001, p. 66).
Toulmin calls for the restoration of the Balance of Reason by acknowledging the validity
and role of practical wisdom in assessing what is reasonable or appropriate in dealing with
human and social problems. Practical wisdom is a translation of Aristotles concept of phronesis
which, in contrast to episteme (theoretical knowledge) and techne (instrumental knowledge),
involves, according to Dryzek (1990, p. 9) . . . persuasion, reflection upon values, prudential

Is it what works that matters? 339

Downloaded by [Universit du Qubec Montral] at 01:55 06 September 2014

judgement, and free disclosure of ones ideas. Aristotles discussion of phronesis referred to
medical practice and helmsmanship, realms of skilled and experienced practice that rely
heavily on tacit knowledge. Such knowledge is grounded in experience but largely taken for
granted and embedded in routines that go without saying; it is not easily conveyed through
written reports and papers. A closely related concept is that of metis, which translates from
Greek as knack, wit, or cunning. Toulmin argues for the restoration of these forms of tacit
knowledge in a Balance of Reason, which recognises that . . . all scientific knowledge is a
balance of the theoretical with the practical, the verbal with the non-verbal (Toulmin, 2001,
p. 183).
I would argue that this position provides a more robust basis for dealing with the major
human and social problems that we face than instrumental forms of rationality that have
become embedded in Western liberal democratic political systems. A major challenge derives
from the increasing complexity of late modern society under the influence of increasingly
globalised processes of economic and technological change (Giddens, 1990). According to
Smart (1999, p. 63) We find ourselves abroad in a world in which social theory and analysis
is no longer able, with any credibility, to provide a warrant for political practice and ethical
decision-making. However, the dilemma is that . . . questions concerning political
responsibility and ethical decision-making, the difficulties of adjudicating between the
expression and pursuit of self-interest and the promotion and adequate provision of the public
domain, as well as the problems encountered in everyday social life of making a choice or
taking a stand, have if anything become analytically more significant . . . (Smart, 1999).
Dryzek (1990) argues that dominant responses to the challenge of complexity tend to
reflect the preoccupations of instrumental rationality, emphasising forms of analysis and
technical aides to decision making that are more sophisticated, in order to maintain control
over a problematical environment. We have seen aspects of just such responses in the
Governments approach to EBP in the context of the performance management controlstrategy. From the perspective of practical reason, such responses will have limited success for
two main reasons. First, by focusing on formal scientific and technical knowledge, they
neglect the key role played in problem solving by practical wisdom and informal tacit
knowledge. Second, by conceiving of rationality in terms of means to given ends, they neglect
the ethical-moral dimension of problem solving. Indeed, these two aspects should be seen as
inter-related in that practical knowledge or wisdom is necessarily applied in practice within
a normative framework.
We can consider this further by looking briefly at the organisational context in which
professional practice occurs in relation to policy-making and public service delivery. Recent
research on organisations has emphasised the importance of social relations and informal
processes founded upon tacit knowledge (Hatch, 1997; Moingeon and Edmondson, 1996).
New institutionalism also highlights the role of the informal normative order defined in
terms of norms, routines and conventions, which are largely tacit and implicit and deeply
ingrained in organisational life (March and Olsen, 1989; Lowndes, 1997). March and Olsen
(1989) argue that the focus should therefore be on the appropriateness of action and
behaviour, defined in relation to the normative order of obligation and necessity, rather than
on the rational order of preference and calculation and consequence.
Given this conception of the organisational and institutional context in which policymakers and practitioners encounter evidence, make judgements and take decisions, rationaldecisionistic models appear highly simplistic and distorting. For example, with reference to
evidence-based practice in social work, Stephen Webb (2001) argues that professionals

340 Research Papers in Education Volume 18 Number 4

Downloaded by [Universit du Qubec Montral] at 01:55 06 September 2014

working in public sector organisations employ reasoning strategies which . . . consistently fail
to respect the canons of rationality assumed by the evidence-based approach (Webb, 2001, p.
64). People employ cognitive heuristics that are selective in relation to evidence and decisions
are influenced by factors such as the politics of inter-agency relations and internal
organisational interest groups and . . . are based upon the reflexive understanding of
contestable beliefs and meanings and not determinate judgements (Webb, 2001, p. 68).
The role of the normative context of social and political rules, conventions and structures
of power and authority in shaping professional practice and action is emphasised in John
Foresters (1993) analysis of planning practice as communicative action. Not only does such
practice draw on sources of knowledge other than explicit scientific evidence (in particular
on tacit and experiential practitioner knowledge) but it also necessarily addresses normative
considerations: . . . questions of norms and values are . . . necessarily influential and
constitutive of the very sense of action itself (Forrester, 1993, p. 72). In the context of
community care, Janet Lewis (2001) has recently highlighted the danger, in evidence-based
practice, of excluding both the practical wisdom of the experienced practitioner and the
experience of service users as forms of knowledge that do not constitute valid evidence.
However, the problem here is not limited to the restriction of the cognitive basis of practice.
The exclusion of these perspectives also excludes the value stances that go with them,
resulting in a de facto privileging of the normative commitments of academics, managers and
policy makers (cf. Beresford, 2001).
The argument can be applied to professionals working in other public service contexts.
Scientific evidence tends to be at a relatively high level of generality whereas professionals are
faced with decisions about dealing with particular problems in particular circumstances and
institutional contexts. Research evidence is notoriously slippery even in terms of informing
general policy guidelines to deal with such problems. When it comes to the specifics of
practice, it recedes even further into the background in decisions on appropriate action. To
take an example from the educational field, a schools policy for dealing with bullying should
certainly be informed by evidence of what is generally effective but responses to bullying
incidents will be dominated by practice wisdom, cautiously teasing out the most appropriate
course of action in the specific circumstances in a context of informal rules, heuristics, norms
and values. The question for teachers is not simply what is effective but rather, more broadly
it is, what is appropriate for these children in these circumstances.

CONCLUSION: WHAT DOES MATTER?

I would not want my argument to be misconstrued. I certainly do not line up with postmodernists who take a nihilistic view of a world beyond the reach of redemptive politics (to
use OSullivans phrase again). I have long believed in the need to apply the findings from
social scientific research more systematically in our collective efforts to address social problems
and improve the lives of those suffering poverty and disadvantage. The danger lies in the
instrumental conception of rationality that underpins what Frank Fischer (1990) terms
technocratic politics and what Thomas Schwandt sees as the reduction of policy-making to
an exercise in social technology.
In this context, I have outlined concerns over the current wave of optimism about EBP
and about the potential of theory-based evaluation. Although I would count myself as a
supporter of this approach to evaluation, I have doubts about the feasibility of its promise,

Is it what works that matters? 341

Downloaded by [Universit du Qubec Montral] at 01:55 06 September 2014

especially in relation to complex, crosscutting, multi-intervention programmes. Such contexts

present very severe challenges to evaluators in seeking to derive evidence of what works in
terms of an understanding of the causal mechanisms by which an intervention achieves
change and how the operation of such mechanisms is influenced by relevant contextual
factors. I suspect that much of the evidence produced by ostensibly theory-based evaluations
will continue to provide limited purchase on causal mechanisms and present serious problems
of external validity.
In part, this is due to the character of the evaluation enterprise. Evaluation is, by its nature,
part of a policy process conditioned by interests located in social and political processes
operating in particular organisational contexts. Thus, as an exercise in practical-moral
reasoning, evaluation takes place within an institutional framework, which differs in
significant ways from that which is applicable to the pursuit of pure social scientific
knowledge. For example, the policy client normally has a strong interest in seeing findings
that can be used to improve policy design or implementation within timescales dictated by the
political process. Evaluation funded by government will inevitably be focused to a significant
degree on a process of learning and improvement rooted in short-term practice rather than
one oriented to longer-term development of research evidence. Evaluators must satisfy the
needs and demands of multiple stakeholders, often within limited budgets, and this involves
balancing and compromising amongst various prioritiesa strategy not particularly
conducive to the kind of sustained, in-depth research that Weiss (1997) argues is necessary for
successful theory-based evaluation.
Ultimately, I think we have to be modest in our expectations about the kind of knowledge
that we can derive as a basis for policy-making. We have seen that in his concern to promote
knowledge for human good, Stephen Toulmin urges us to . . . recall the practical wisdom of
the sixteenth century humanists, who hoped to recapture the modesty that had made it
possible for them to live happily with uncertainty, ambiguity and pluralism (Toulmin, 2001,
p. 80). Dupre (2001) advocates a thorough-going pluralism that . . . goes all the way down
to the basic metaphysical issues of causality and what kind of things there are (Dupre, 2001,
p. 183). He argues that there is no one fundamental perspective that can explain human
behaviour and, indeed, that . . . it remains easier to say what does not work than what does
(Dupre, 2001, p. 184). We can achieve fragments of the whole picture from various
perspectives and we need to develop skills of integrating these fragments; . . . trying to make
one or even a few such fragments stand for the whole presents us with a deformed image of
ourselves (Dupre, 2001, p. 183).
Moreover, we have seen how practical wisdom must rely heavily on tacit knowledge
grounded in experience and how the application of such knowledge in professional practice
necessarily occurs within a normative framework of values, rules, heuristics and conventions
that can make decision-making and action as much a matter of obligation and necessity as a
matter of rational analysis. For Stephen Toulmin, practical wisdom helps us . . . in untying the
knots in which our lives enmesh us (Toulmin, 2001, p. 123) and in this task . . . intelligent
analysis of the factual soil from which our problems spring . . . must be . . . guided by ideals
that make rational assessments stepping stones to reasonable decisions (Toulmin, 2001, p.
213). This is in accordance with John Deweys notion of creative intelligence through which
we can transform the world for the better. Such intelligence has relevance to our ends and
values and . . . frees action from a mechanically instrumental character . . .; its function is . . .
to project new and more complex endsto free experience from routine . . . to liberate and
liberalize action (Dewey, 1993, p. 67). From this perspective, what matters is arriving at such

342 Research Papers in Education Volume 18 Number 4

Downloaded by [Universit du Qubec Montral] at 01:55 06 September 2014

reasonable decisions and ensuring that we take action that is appropriate to situations that
are both morally and factually ambiguous (cf. Harmon, 1995). The practical rationality of
this process requires inclusive, open and free debate, ensuring that all those with a stake in
decisions are able to bring to bear their knowledge (in whatever form) and their normative
commitments.
If we accept this position, and acknowledge the need to nurture the application of practical
wisdom in our efforts to improve the world, then we need to broaden the focus of evaluation
beyond the technical concerns of measuring effects, identifying causes and assessing what
works. It can help in the practical task of identifying what is appropriate or reasonable. To do
this, it must be acknowledged that the ethical and moral implications of policies and the values
and goods (and bads) that they promote are amenable to rational consideration and debate
(cf. Julnes et al., 1998). Broadening the focus of evaluation in this way also involves broadening
its methodologies beyond analytical techniques to include methods and accompanying
institutional frameworks to promote full, free and open normative debate among all those
with a stake in the policies concerned, including service users and citizens.
In this way, evaluation can strengthen the basis for making wise policy choices with
profound ethical and moral implications. It can strengthen our capacity to answer what
Zagorin (1998, p. 224) calls Tolstoys anguished questionwhat shall we do and how shall
we live? It is this question that tasked John Dewey, who was passionately committed to the
application of intelligence to the solution of human and social problems and transforming the
world for the better, advocating experimentation and an open-minded will to learn. His is
the appropriate final word in advocating . . .
. . . the necessity of a deliberate control of policies by the method of intelligence, an
intelligence which is not the faculty of intellect honored in text-books and neglected
elsewhere, but which is the sum-total of impulses, habits, emotions, records, and
discoveries which forecast what is desirable and undesirable in future possibilities, and
which contrive ingeniously on behalf of an imagined good (Dewey, 1993, p. 9).
REFERENCES
BERESFORD, P. (2001). Evidence-based care: developing the discussion, Managing Community Care,
9, 36.
BRONK, R. (1998). Progress and the Invisible Hand: The Philosophy and Economics of Human Advance.
London: Warner Books.
BULLOCK, H., MOUNTFORD, J. and STANLEY, R. (2001). Better Policy Making. Centre for
Management and Policy Making, London: Cabinet Office.
BULMER, M. (Ed.) (1987). Social Science Research and Government: Comparative Essays on Britain and the
United States. Cambridge: Cambridge University Press.
CABINET OFFICE (1999). Modernising Government. Cm. 4310. London: The Stationery Office.
CHEN, H-T. (1990). Theory-Driven Evaluations. Newbury Park: Sage Publications.
CONNELL, J. P., KUBISCH, A. C., SCHORR, L. B. and WEISS, C. H. (Eds) (1995). New Approaches
to Evaluating Community Initiatives, Volume 1: Concepts, Methods and Contexts. Washington, DC: Aspen
Institute.
COOK, T. D. (2000). The False Choice Between Theory-Based Evaluation and Experimentalism. In:
ROGERS, P. J., PETROSINO, A. J., HEUBNER, T. A. and HASCI, T. A. (Eds) Program Theory
Evaluation: Practice, Promise, and Problems, New Directions in Evaluation, No. 87, 2734. San Francisco:
Jossey Bass.

Is it what works that matters? 343

Downloaded by [Universit du Qubec Montral] at 01:55 06 September 2014

DAVIES, H. T. O., NUTLEY, S. M. and SMITH, P. C. (1999). Editorial: what works? The role of
evidence in public sector policy and practice, Public Money and Management, 19, 35.
DAVIES, H. T. O., NUTLEY, S. M. and SMITH, P. C. (Eds) (2000). What Works? Evidence-Based Policy
and Practice in Public Services. Bristol: Policy Press.
DEWEY, J. (1957). Reconstruction in Philosophy. Enlarged Edition. Boston, MA: Beacon Press.
DEWEY, J. (1993). The Need for a Recovery of Philosophy. In: MORRIS, D. and SHAPIRO, I.
(Eds) John Dewey: The Political Writings. Indianapolis, IN: Hackett Publishing Co.
DfEE. (2000). Influence or Irrelevance: Can Social Science Improve Government? Secretary of States ESRC
Lecture Speech, 2 February. London: Department for Education and Employment.
DRYZEK, J. S. (1990). Discursive Democracy: Politics, Policy and Political Science. Cambridge: Cambridge
University Press.
DUPRE, J. (2001). Human Nature and the Limits of Science. Oxford: Clarendon Press.
FISCHER, F. (1990). Technocracy and the Politics of Expertise. Newbury: Sage Publications.
FORESTER, J. (1993). Critical Theory, Public Policy and Planning Practice: Toward Critical Pragmatism. New
York: State University of New York Press.
FULBRIGHT-ANDERSON, K., KUBISCH, A. C. and CONNELL, J. P. (Eds) (1998). New
Approaches to Evaluating Community Initiatives, Volume 2: Theory, Measurement and Analysis. Washington,
DC: Aspen Institute.
GIDDENS, A. (1990). The Consequences of Modernity. Cambridge: Polity Press.
GRANGER, R. C. (1998). Establishing Causality in Evaluations of Comprehensive Community
Initiatives. In: FULBRIGHT-ANDERSON, K., KUBISCH, A. C. and CONNELL, J. P. (Eds)
New Approaches to Evaluating Community Initiatives, Volume 2: Theory, Measurement and Analysis.
Washington DC: Aspen Institute.
HATCH, M. J. (1997). Organization Theory: Modern, Symbolic, and Postmodern Perspectives. Oxford:
Oxford University Press.
HARMON, M. M. (1995). Responsibility as Paradox: A Critique of Rational Discourse on Government.
Thousand Oaks, CA: Sage.
HM TREASURY (2000). 2000 Spending Review: Public Service Agreements. Cm. 4808. London: HM
Treasury.
JUDGE, K. and BAULD, L. (2001). Strong theory, flexible methods: evaluating complex communitybased initiatives, Critical Public Health, 11, 1938.
JULNES, G., MARK, M. M. and HENRY, G. T. (1998). Promoting realism in evaluation: realistic
evaluation and the broader context, Evaluation, 4, 483504.
LEWIS, J. (2001). What works in community care?, Managing Community Care, 9, 36.
LOWNDES, V. (1997). Change in public service management: new institutions and new managerial
regimes, Local Government Studies, 23, 4266.
MACINTYRE, A. (1984). After Virtue: A Study in Moral Theory. Notre Dame: University of Notre
Dame, IN Press.
MAJONE, G. (1989). Evidence, Argument and Persuasion in the Policy Process. New Haven, CT: Yale
University Press.
MARCH, J. G. and OLSEN, J. P. (1989). Rediscovering Institutions: The Organizational Basis of Politics.
New York: The Free Press.
MOINGEON, B. and EDMONDSON, A. (Eds) (1996). Organisational Learning and Competitive
Advantage. London: Sage.
NATIONAL AUDIT OFFICE. (2001). Modern Policy-Making: Ensuring Policies Deliver Value for Money.
London: NAO.
OAKLEY, A. (2000). Experiments in Knowing: Gender and Method in the Social Sciences. Cambridge: Polity
Press.
OSULLIVAN, N. (1993). Political integration, the limited state, and the philosophy of postmodernism, Political Studies, XLI, 2142.
PAWSON, R. and TILLEY, N. (1997). Realistic Evaluation. London: Sage Publications.

344 Research Papers in Education Volume 18 Number 4

Downloaded by [Universit du Qubec Montral] at 01:55 06 September 2014

POWELL, M. (Ed.) (1999). New Labour, New Welfare State? The Third Way in British Social Policy. Bristol:
The Policy Press.
ROGERS, P. J., PETROSINO, A., HEUBNER, T. A. and HASCI T. A. (Eds) (2000). Program Theory
Evaluation: Practice, Promise, and Problems, New Directions in Evaluation, No. 87. San Francisco, CA:
Jossey Bass.
SANDERSON, I. (1998). Beyond performance measurement? Assessing value in local government,
Local Government Studies, 24, 125.
SANDERSON, I. (1999). Participation and democratic renewal: from instrumental to communicative rationality?, Policy and Politics, 27, 325341.
SANDERSON, I. (2000). Evaluation in complex policy systems, Evaluation, 6, 433454.
SANDERSON, I. (2001). Performance management, evaluation and learning in modern local
government, Public Administration, 79, 297313.
SANDERSON, I. (2002). Evaluation, policy learning and evidence-based policy making, Public
Administration, 80, 122.
SCHWANDT, T. A. (1997). Evaluation as practical hermeneutics, Evaluation, 3, 6983.
SCHWANDT, T. A. (2000a). Meta-analysis and everyday life: the good, the bad and the ugly, American
Journal of Evaluation, 21, 213.
SCHWANDT, T. A. (2000b). Further diagnostic thoughts on what ails evaluation practice, American
Journal of Evaluation, 21, 225.
SCRIVEN, M. (1998). Minimalist theory: the least theory that practice requires, American Journal of
Evaluation, 19, 57.
SMART, B. (1999). Facing Modernity: Ambivalence, Reflexivity and Morality. London: Sage Publications.
TOULMIN, S. (2001). Return to Reason. Cambridge, MA: Harvard University Press.
TRIGG, R. (2001). Understanding Social Science: A Philosophical Introduction to the Social Sciences, 2nd Edn.
Oxford: Blackwell Publishers.
WAGENAAR, H. C. (1982). A Cloud of Unknowing: Social Science Research in a Political Context.
In: KALLEN, D. B. P., KOSSE, G. B., WAGENAAR, H. C., KLOPROGGE, J. J. J. and
VORBECK, M. (Eds) Social Science Research and Public Policy-Making. Foundation for Educational
Research in the Netherlands. Windsor: NFER-Nelson.
WALKER, R. (2001). Great expectations: can social science evaluate New Labours policies?,
Evaluation, 7, 305330.
WEBB, S. A. (2001). Some considerations on the validity of evidence-based practice in social work,
British Journal of Social Work, 31, 5779.
WEISS, C. H. (1982). Policy Research in the Context of Diffuse Decision-Making. In: KALLEN, D.
B. P., KOSSE, G. B., WAGENAAR, H. C., KLOPROGGE, J. J. J. and VORBECK, M. (Eds) Social
Science Research and Public Policy-Making. Foundation for Educational Research in the Netherlands.
Windsor: NFER-Nelson.
WEISS, C. H. (1995). Nothing as Practical as Good Theory: Exploring Theory-Based Evaluation for
Comprehensive Community Initiatives for Children and Families. In: CONNELL, J. P.,
KUBISCH, A. C., SCHORR, L. B. and WEISS, C. H. (Eds) New Approaches to Evaluating
Community Initiatives, Volume 1: Concepts, Methods and Contexts. Washington, DC: Aspen Institute.
WEISS, C. H. (1997). How can theory-based evaluation make greater headway?, Evaluation Review, 21,
501525.
WEISS, C. H. (2000). Which Links in Which Theories Shall We Evaluate?, In: ROGERS, P. J.,
PETROSINO, A., HEUBNER, T. A. and HASCI, T. A. (Eds) Program Theory Evaluation: Practice,
Promise, and Problems, New Directions in Evaluation, No. 87, pp.3545. San Francisco: Jossey Bass.
ZAGORIN, P. (1998). Francis Bacon. Princeton, MA: Princeton University Press.

CORRESPONDENCE
Professor Ian Sanderson, Policy Research Institute, Leeds Metropolitan University, Bronte
Hall, Beckett Park Campus, Leeds LS6 3QS, UK. E-mail: i.sanderson@leedsmet.ac.uk

Is it what works that matters? 345