How can the federal government can improve high performance sport in Canada

PUBLIC SECTOR SPORT POLICY Developing a Framework for Analysis1

Barrie Houlihan Loughborough University, UK

Abstract Despite the increasing involvement of governments in sport, and the high level of academic interest in sports-related public policy issues such as equity, doping, harassment and violence, there is remarkably little analysis of sport policy that utilizes the major models and frame- works for analysis widely adopted in other policy areas. Following the development of a set of criteria for assessing the adequacy of analytic frameworks, four major meso-level analytic frame- works are examined: the stages model, institutional analysis, multiple streams model, and the advocacy coalition framework. These are assessed for their internal coherence and applicability to the study of sport policy. None of the four frameworks reviewed is considered to be sufficiently per- suasive and, consequently, a modified version of the advocacy coalition framework is developed and suggested as the most promising point of departure for the analysis of sport policy.

Key words • advocacy coalitions • institutionalism • multiple streams • policy analysis • sport

Introduction

Over the last 30 years or so there has been a rapid growth in academic interest in the public policy process and particularly how issues get on to the agenda of government, the factors that shape policy choice, and the impact of policy. Policy analysis has proved fertile territory for conceptual innovation, model-building, and analytical and normative theorizing. However, while policy areas such as the environment and social welfare have been subject to extensive analysis and have been the focus for the testing and refinement of early frameworks and models of policy analysis, sport has remained on the margins. This marginalization con- trasts with the recognition of sport by many governments as an increasingly important area of policy. Whereas few governments in the 1960s gave any explicit budgetary or ministerial recognition to sport, by the mid-1990s sport was an established feature in the machinery of government of most economically developed countries. For example, of 39 members of the Council of Europe sport was a clearly identifiable governmental responsibility in at least 26 countries. Of those 26, 15 had a ministerial post and department which included responsibility for sport, 10 had one or more national level agencies with a major responsibility for sport and, 20 identified sport as a significant responsibility of a sub-national tier of government.

INTERNATIONAL REVIEW FOR THE SOCIOLOGY OF SPORT 40/2(2005) 163–185 163

© Copyright ISSA and SAGE Publications (London, Thousand Oaks, CA, New Delhi) www.sagepublications.com 10.1177/1012690205057193

01_IRS articles 40_2 9/6/05 1:27 PM Page 163

A review of the nature of contemporary government involvement in sport shows it to be broad and often extensive. For example, for some countries, such as France and Italy, sport is a focus for a growing volume of state regulatory activity including licensing of coaches, control of doping, control of the sale of broadcasting rights for sports events, certification of stadium safety, and the licensing of sports clubs. In many countries, such as the UK, Canada and Australia, there has been a steady increase, since 1990, in the allocation of public funds to sports activities. In the UK, although local authority expenditure has remained steady, National Lottery funding has averaged around £250m per year and central government expenditure has risen from about £50m in 1999–2000 to just under £100m in 2003–4. In Canada, increased federal invest- ment followed the disappointing performance at the Athens Olympic Games, while in Australia an additional A$135m was allocated to elite sport in the three years prior to the Sydney Olympic Games. Various governments have sought, in recent years, to use sport as an instrument of social engineering to reduce juvenile offending, reduce obesity among the young, and achieve greater social inclusion of marginalized groups. There has also been a steady rise, since 1984, in the number of cities bidding to host the summer Olympic Games and, with the exception of the Atlanta Games, the explicit willingness of their governments to underwrite the risk. Finally, almost 200 governments have expressed their support for the World Anti-Doping Code published by the World Anti-Doping Agency in 2003.

Unfortunately, the increase in governmental interest in sport has not been matched by an equivalent increase in academic interest in the analysis of public policy for sport. In marked contrast to other areas of public policy, sport is notable for having no academic journal that concerns itself explicitly with issues of policy and the role of government. Policy journals (normally more than one) exist not only in the established policy fields such as transport,2 health,3 educa- tion,4 economic management,5 and social welfare,6 but also in the more recent areas of government concern such as the environment7 and culture.8 A sport policy journal is conspicuous by its absence. However, the lack of journals that deal with aspects of sport policy analysis might not be a matter of concern if there were a significant number of articles in the social science of sport journals written with an explicit concern with policy analysis. Unfortunately, a survey of nine major English language journals from January 2001 to September 2003 found that only 3 percent of the articles utilized the extensive array of concepts, analytical frameworks and theories developed in mainstream policy analysis to aid understanding of sport policy-making and the role of government. This was despite the fact that 16 percent of articles were focused on issues that are the subject of governmental concern in many countries (for example gender equity, sports participation levels, environmental impact of sports facilities, spectator violence, and doping). The point of concern here is not that the field of study is dominated by sociological, cultural studies, and economic perspectives, but to argue that the insights derived from these perspectives would be enhanced and complemented by a more explicit and theoretically informed understanding of the process of sport policy-making, and of the role and significance of government and state agencies.

164 INTERNATIONAL REVIEW FOR THE SOCIOLOGY OF SPORT 40(2)

01_IRS articles 40_2 9/6/05 1:27 PM Page 164

The aims of this article are: first, to provide an assessment of the potential value of concepts and frameworks, drawn from the policy analysis literature, for the study of sport policy through a review of four major meso-level theoretical frameworks, namely, the stages model, institutional analysis, multiple streams and the advocacy coalition framework; and second, to suggest a framework for analysis of sport policy which draws on the strengths of existing frameworks. The fulfilment of these aims may encourage further debate on the most promising ways of generating a richer understanding of the sport policy process, the articulation of interests, and the ways in which power affects policy outputs and outcomes. The context for the article is primarily the policy process in the UK although the conclusions drawn and the framework proposed have an applicabil- ity beyond the boundaries of the UK political system.

The next sections consider the impact on theory development of changes in the political and academic contexts of policy analysis, and the criteria against which policy analysis frameworks can be evaluated. An outline and assessment of the four policy frameworks leads to the conclusion that a modified version of the advocacy coalition framework is the most promising vehicle for the analysis of sport policy.

Approaches and Frameworks for the Analysis of Sport Policy

The focus for the following is public policy, defined as policies that originate within, or are dependent upon the resources of, the state. Furthermore, this review is concerned with policy analysis at the meso-level, that is, the level of national organizations (ministries, national sports organizations and interest groups). This is not to deny the significant insights that have been generated from neo-Marxist, corporatist or neo-pluralist macro- (or societal)-level theorizing or from micro- level studies of particular decisions, organizations, or local cases, but simply to indicate the initial focus for analysis and to acknowledge that it is at the meso- level that much recent theory-building and conceptual innovation has taken place (see for example, John, 1998; Kickert et al., 1997; Marsh, 1998; Sabatier, 1999).

The development of policy analysis over the last 30 years owes as much to shifting priorities of government as it does to academic debate. In particular, policy analysis as it developed in the 1970s was seen as largely facilitating demo- cratic government at a time when many capitalist democracies were witnessing a growth in the interventionist state. The focus thus tended to be on problem iden- tification, agenda setting and policy formulation. Such a focus drew heavily on a methodological framework derived substantially from a neo-positivist/rationalist epistemology reflected in an uncritical assumption of the separation of fact from value, the privileging of quantitative methods, and the search for generalizable results (Williams, 1998). When it became clear that major financial investment by a number of governments in programmes intended to reduce poverty, increase gender equity in employment, increase access to education by lower income groups, and, even to achieve greater equity of participation in sport across differ- ent socio-economic groups had substantially failed to deliver the expected results, the attention of policy analysts turned from problem identification to a concern to

HOULIHAN: PUBLIC SECTOR SPORT POLICY 165

01_IRS articles 40_2 9/6/05 1:27 PM Page 165

better understand the implementation process and to evaluate more rigorously the impact of publicly funded programmes.

This refocusing also involved a period of critical reflection on epistemology and methodology. Studies of implementation highlighted the role of politics, interests and values in shaping policy through a continuous process of negotia- tion and interpretation (Pressman and Wildavsky, 1973). Attention was drawn to the ‘bottom-up’ nature of policy-making and, in particular, to the role of ‘street- level bureaucrats’ (Lipsky, 1980), such as teachers, coaches, or sports develop- ment officers, in ‘making’ policy in the everyday context within which they operated. However, while studies of implementation led, generally speaking, to a broadening of epistemological debates, much of the theorizing of evaluation research became preoccupied with strengthening neo-positivist approaches by attempting to identify ever more precise quantitative measures of programme impact and prescriptions for effective management of the implementation process.

Over the last 30 years the field of policy analysis has broadened in scope, methodology, and associated epistemological debates, not just in response to the shifting priorities of government and the perceived success or failure of govern- ment programmes, but also in response to broader debates in the social sciences. The structure/agency debate is replicated in policy analysis with discussions of the importance of the central government departmental location of sport, the extent to which gender inequity is structured into the fabric of sport as a social institution, and the role of ‘policy entrepreneurs’ in sport (e.g. Juan Antonio Samaranch, Primo Nebiolo). Policy analysis is also part of the debate, largely within political science, about the role and significance of the state as opposed to institutions of civil society, such as interest groups, in policy-making. Whether sport policy is substantially the outcome of governmental deliberation, or the product of a process of governance which implies a significant diffusion of power along neo-pluralist lines, is at the heart of many issues including the autonomy of ‘not-for-profit sports governing bodies’, the instrumental manipulation of inter- national sport for governmental purposes, and the control of the sale of sports broadcasting rights.

Contemporary policy analysis is also an arena for debate over the relative weight to be given to interests as opposed to ideas. On one side of the debate are those who, often arguing from a neo-Marxist position, perceive ideas to be little more than the rationalization of entrenched political interests. Thus notions about the importance of competitive sport in general (and, in the UK, in schools in particular) are seen as a part of a process whereby the capitalist class legitimize and secure their interests. On the other side of the debate are those who argue that ideas are independent variables in the policy process, and that grand ideas such as respect for human rights, religious fundamentalism, and Olympism cannot be subordinated to imperatives of instrumental rationality, the pursuit of rational self-interest, or even strategic action by interest groups.

This rich cross-disciplinary debate around questions such as structure and agency, the relationship between the state and civil society, and the significance of interests and ideas has generated considerable theory development. Sabatier (1999) identifies 11 potential frameworks including Baumgartner and Jones’s

166 INTERNATIONAL REVIEW FOR THE SOCIOLOGY OF SPORT 40(2)

01_IRS articles 40_2 9/6/05 1:27 PM Page 166

(1993), punctuated equilibrium framework, Berry and Berry’s (1992) policy diffusion framework and the four frameworks discussed in more detail below. These meso-level frameworks were selected not only because they illustrate both the relationship of policy analysis to changing government preoccupations and the debates in the broader social sciences, but also because they are the most fully developed, have already stimulated empirical application, and are the most intellectually robust having been subject to sustained critical evaluation. However, before embarking on the review it is important to identify a set of criteria by which explanatory adequacy may be evaluated.

Criteria for Evaluating Frameworks for Policy Analysis

Four criteria have been identified against which the frameworks are evaluated, the first and most important of which is a capacity to explain both policy stabil- ity and change (John, 1998; Sabatier, 1999). Much political analysis, influenced by structuralism and most clearly evident in varieties of systems theory, assumes that change is confined to the margins and that the context of public policy is essentially static. Even actor-centred theories of policy selection, such as rational choice theory, have a strong structuralist underpinning since ‘individual action is assumed to be an optimal adaptation to an institutional environment . . . therefore, the prevailing institutions (rules of the game) determine the behaviour of the actors’ (Tsebelis, 1990: 4). A capacity to explain change is especially important because in many, if not most, industrial countries change in sport policy has been rapid. Current explanations of policy change tend to be expressed in terms of the differing weight given to structure, agency and ideas. Explanations emphasizing structural factors exist in two main forms. The first, exemplified by some forms of Marxism, stresses the significance of (economic) forces that determine social development. The second draws attention to the constraints on individual action resulting from natural or social constraints. In relation to sport policy, natural constraints on policy change include climate, geography, topography and pollu- tion, while social constraints include patriarchy, racism, wealth distribution and homophobia. Explanations stressing the importance of agency not only draw attention to the role of agents in establishing and maintaining social structures through the routinization of behaviour, but also give greater weight to personal capacity to influence policy and to attributes such as leadership, policy broker- ing/entrepreneurship, charisma, and motivation.9 As suggested below, in public policy areas such as sport where values, norms and practices are more recently established there is perhaps greater scope for agency to affect policy. In addition to explanations drawn from the structure-agency debate there are those policy analysts (e.g. Fischer, 2003; Haas, 1992) who argue that the role of ideas and epistemic communities is undervalued. For Haas (1992) knowledge-based (epistemic) communities10 have the capacity, especially in relation to complex policy issues, to articulate cause and effect relationships, help states identify their interests, frame issues for collective debate, and propose specific policies.

The second criterion is the capacity to illuminate a range of aspects of the policy process. Too many frameworks concentrate attention and analysis on

HOULIHAN: PUBLIC SECTOR SPORT POLICY 167

01_IRS articles 40_2 9/6/05 1:27 PM Page 167

discrete aspects of the policy process, such as agenda-setting, interest articula- tion, policy impact or the role of the state. They lack a capacity to consider the policy process in a holistic manner which examines the inter-relationship between actors, aspects or stages. The third criterion is applicability across a range of policy areas. Any framework would benefit from having applicability beyond the sport policy area. The application of analytic frameworks in policy areas beyond sport should result in a greater sensitivity to the distinctive features of the sport policy area. Comparison of policy areas allows greater insight into the significance of systemic factors such as bureaucratic processes, organiza- tional culture, and political party ideology. In addition, the applicability of a framework across policy areas allows the examination of the policy process in policy fields, such as health and education, which have a substantial potential ‘spillover’ into the sports area. The final criterion is that the framework should facilitate a medium term (five to 10 year) historical analysis of policy change on the grounds that any period of shorter duration merely produces a snapshot of policy. Only with a period of at least five years is it possible first, to distinguish minor fluctuations in policy direction from sustained change and second, to identify significant explanatory factors. In the following section each of the four frameworks is outlined and evaluated in terms of the above criteria.

The Stages Model

Much of the stimulus for innovation in policy analysis has been generated by a growing dissatisfaction with the stages model, which dominated policy analysis between 1970 and the late 1980s, and which is based on a division of the policy process into a series of discrete stages following the rational actor model. Hogwood and Gunn (1984), for example, identified nine stages: deciding to decide (agenda setting), issue filtration (deciding how to decide), issue definition, forecasting, setting objectives and priorities, options analysis, policy implementa- tion, evaluation and review, and policy maintenance, succession or termination. While the stages model, in its many variants, can be used to conceptualize and analyse the policy process as a whole, it has more commonly been used to analyse particular elements or stages in the process and has resulted in a number of important studies such as those focused on agenda setting (e.g. Kingdon, 1995), implementation (e.g. Pressman and Wildavsky, 1973), and evaluation (e.g. Guba and Lincoln, 1987).

Within the area of sport policy one of the few examples of the use of the stages model is Houlihan’s studies of the public policy response to football hooliganism in England, and doping in Britain (1990, 1991). Houlihan uses both Hogwood and Gunn’s model of the ‘policy cycle’ and Down’s description of the ‘issue attention cycle’, arguing that they ‘draw attention to the interaction between the policy process and the context within which [they] take place [and that they] capture the dynamic interaction between politics and administration’ (1991: 159). The analysis was strongest in explaining how issues reached the governmental policy agenda, but weaker in explaining the process by which policy was made and explaining the impact of policy. In addition, as Giulianotti

168 INTERNATIONAL REVIEW FOR THE SOCIOLOGY OF SPORT 40(2)

01_IRS articles 40_2 9/6/05 1:27 PM Page 168

observed, the model was unable to capture the detail of the policy process and the ideological imperatives that led to shifts in the definition of the problem over time (1994: 12). Giulianotti also dismissed Down’s model as ‘excessively reductive and qualitatively unevaluative . . . [and] technicist’ (1994: 29).

More generally, the stages model was criticized for: 1) being descriptive rather than providing a causal explanation; 2) being inaccurate in its description insofar as it failed to capture the ‘messiness of policy-making’ (John, 1998: 25) and implying a false degree of rationality in the policy process; 3) having a top- down bias and often focusing on legislation rather than other forms of policy- making; 4) theorizing parts of the policy process rather than the process as a whole; and 5) being simplistic in its image of policy-making as a neatly sequen- tial and linear or cyclical process (Sabatier, 1999). Such was the weight of criticism that Sabatier felt able to conclude that ‘the stages heuristic has outlived its usefulness and needs to be replaced’ (1999: 7).

In terms of the criteria of adequacy, the profile of the stages model is mixed. The model is clearly weak in meeting the requirement that it contribute to an explanation of stability and change. Where change is acknowledged the model is hampered by its neo-positivist/rationalistic assumptions, raising the concern that what is being described and analysed is the presentation of policy rather than the exercise of power that underpins the process. While the stages model can illuminate a range of aspects of the policy process, its assumption of a broadly sequential relationship between them implicitly, at least, limits the researcher’s field of vision. On the third criterion, applicability across a range of policy areas, there is certainly an extensive body of literature in many policy areas, all of which are subject to the same strong criticism noted here. The stages model is also weak in relation to the final criterion as it is more effective in capturing particular moments in the policy process than in identifying patterns of influence and out- comes over a sustained period.

Even if one were to argue that the conceptual language of discrete stages is still useful, there is a clear need for a more theoretically sophisticated and holistic framework for sport policy analysis. However, as De Leon (1999: 29) cautions, ‘before we discard a useful friend . . . we need to make sure . . . that we have a better, more robust framework on which to rely’.

Institutional Analysis

An alternative, and arguably more robust, analytic framework is provided by institutional analysis. Interest in institutions was stimulated by growing dis- satisfaction with the behavioural emphasis in much of American political science of the 1960s that, it was argued, obscured deeply entrenched socio-cultural and political structures that shaped behaviour. For Thelen and Steinmo, institutions ‘shape how political actors define their interests and . . . structure their relations of power to other groups’ (1992: 2) and are seen as significant constraints and mediating factors in politics, which ‘leave their own imprint’ (1992: 8).

Although the concept of an institution is defined in a wide variety of ways there are two broad orientations in the literature, one emphasizing the signifi-

HOULIHAN: PUBLIC SECTOR SPORT POLICY 169

01_IRS articles 40_2 9/6/05 1:27 PM Page 169

cance of institutions as organizational entities (agencies, departments, parlia- ments, etc.), and the other, cultural institutionalism, which highlights shared values, norms and beliefs. Both variants have a historical dimension which emphasizes the ‘relative autonomy of political institutions from the society in which they exist; . . . and the unique patterns of historical development and the constraints they impose on future choices’ (Howlett and Ramesh, 1995: 27). Institutions constrain choice through their capacity to shape actors’ perceptions of both problems and acceptable solutions. As such, the emphasis on institutions is a valuable corrective to the tendency of much pluralist theory to treat organi- zations (the IOC, WADA board, and ministries of sport, for example) as arenas in which politics takes place rather than as independent or intervening variables in the process. Cultural institutionalism, with its emphasis on values, norms and beliefs, emphasizes the social construction of meaning and ‘how interest groups, politicians, and administrators decide their policy preferences’ (Fischer, 2003: 29). For Ostrom (1999), policy is made within ‘action arenas’ that develop operational rules which are nested within other more deeply rooted rules, name- ly, collective choice rules and, at a deeper level, constitutional choice rules which are more resistant to change.

The relevance of institutionalism within sport policy analysis is clear. Green (2003), Henry (2001), Houlihan and White (2002), Pickup (1996) and Roche (1993), all identify the organizational infrastructure of UK sport as a significant variable in shaping policy; Kraus (1990) and Wilson (1994) draw similar conclusions with regard to the United States, as do Macintosh (1991) and Macintosh and Whitson (1990) in relation to Canada. Allocation of functional responsibility for sport, federalism, the use of ‘arms length’ agencies, and the presence of a minister for sport are all seen as having a noticeable impact on sport policy and its implementation. Similar claims for the significance of cultural institutions are also widespread. Beliefs, norms and values associated with social class (Birley, 1996), gender (Hargreaves, 1994), disability (Thomas, 2003), and ethnicity (Carrington and McDonald, 2000) have all been demonstrated to have had, and indeed to continue to have, a marked impact on the character of UK sport policy.

One of the key strengths of institutionalism is that it directs attention to both the behaviour of actors and the structures within which they operate (Hall, 1986). A further strength is that institutionalism is a powerful corrective to those who are too ready to ignore the significance of state institutions in the policy process. Despite these strengths it is difficult to treat institutionalism as a fully articulated analytic framework and it is best seen as an analytic orientation or sensitizing concept. Among the weaknesses of institutionalism is its treatment of policy dynamics. At best, this is indistinct, and at worst it lapses into an implicit rational actor model which depends on a vague ideational formulation, or which collapses into a crude form of institutional determinism. A further related weak- ness concerns the treatment of interests – the tendency to substitute ideas for interests, and the assumption that institutions strongly influence interests (Pontusson, 1995). Attempts to suggest sources of institutional dynamics – such as previously latent institutions becoming more prominent, exogenous changes in the wider socio-economic environment, or the breakdown, replacement and

170 INTERNATIONAL REVIEW FOR THE SOCIOLOGY OF SPORT 40(2)

01_IRS articles 40_2 9/6/05 1:27 PM Page 170

creation of new institutions – simply moves the problem one step back and still leaves unspecified the factors or forces that caused changes in contiguous institutions or the wider institutional environment. Thus, in terms of the first criterion for evaluating analytical frameworks, the capacity to explain stability and change, institutional analysis is severely limited. Institutional analysis is also limited in relation to the second criterion, the capacity to illuminate a range of aspects of the policy process, due to its privileging of structure over agency. However, more positively, institutional analysis is clearly applicable over a range of policy areas and is useful in directing attention to the significance of both organizational and cultural structures for sport policy. Institutionalism is also strong in relation to the final criterion as it emphasizes the value of placing institutions in their historical context. However, at the end of a review of institu- tionalism one is left with the fairly self-evident conclusion that institutions matter, without explaining the extent to which they matter and the circumstances in which they would matter more or less. At best the framework is under- theorized, and at worst it privileges institutions on the basis of weak evidence.

The Multiple Streams Framework

The multiple streams framework is primarily concerned with the process of agenda setting and, as Kingdon notes, ‘is as far away from the sequential model of policy-making as can be imagined’ (1984: 175). It takes as its starting point the ‘garbage can’ metaphor of Cohen et al. (1972) that emphasizes the anarchical character of organizations and the policy process. Policy choice is a ‘garbage can’ where ‘various kinds of problems and solutions are dumped by participants as they are generated’ (Cohen et al., 1972: 2). Kingdon emphasizes the ambiguity, complexity, and degree of residual randomness in policy-making in marked con- trast to the dominant assumptions of actor rationality. He identifies three distinct streams. The problem stream comprises those issues which government policy- makers have identified as requiring action (as opposed to those they have chosen to ignore) prompted by focusing events (crises, for example), indicators of the changing scale of a problem (e.g. routine collection of statistics on juvenile crime, obesity, or cardio-vascular disease), and feedback on the performance of current policies. The policy stream is conceptualized as a ‘primeval soup’ within which ideas, sponsored by particular policy communities, float around and occasionally combine and rise to the top of the agenda, or are adopted by policy entrepreneurs who promote particular ideas and mobilize opinion. Ideas might include the belief that physical education is a crucial part of the school curriculum, or that privatization improves the quality of sports services. These ideas reach the top of the government’s agenda if they fulfil a number of criteria, including that they should be technically feasible and compatible with the dominant values of the community. Finally, the political stream, which is independent of the other two streams, comprises a number of elements including the national mood, organized political forces (political parties and interest groups, for example), and govern- ment.

The chance of an issue getting on to the policy agenda is not a simple func-

HOULIHAN: PUBLIC SECTOR SPORT POLICY 171

01_IRS articles 40_2 9/6/05 1:27 PM Page 171

tion of the power of affected interests; rather, it is the result of the coincidence of the three streams to provide a ‘launch window’ where ‘a problem is recognized, a solution is developed and available in the policy community, a political change makes the time right for policy change, and potential constraints are not severe’ (Kingdon, 1984: 174). Kingdon’s framework is a powerful critique of rational models of decision-making, and also challenges the assumption of deeply entrenched institutionalized interests that distort the political system. Ideas are seen as an important ingredient in the policy process, which emerge from a ‘contingent and often contradictory selection process’ (John, 1998: 175). Successful ideas, when incorporated into policies, can have a ripple effect through the political system by spilling over into other policy fields.

There is much in the multiple streams framework that is attractive in addition to its challenge to the rationalist assumptions of other frameworks. Of particular value is that it can be integrated with other concepts, such as that of the ‘policy community’, and it is explicitly concerned with explaining change. Kingdon’s concept of ‘spillover’ of policy implications from stronger to weaker policy areas is also of potential value given the vulnerability of sport policy to manipulation by diplomatic, health and educational interests (Dery, 1999; Houlihan, 2000).

Kingdon’s concept of the ‘policy entrepreneur’ is also interesting, not only because it offers an opportunity to consider explicitly the role of agency in the policy process, but also because a number of recent studies of sport policy draw attention to the lack of systemic embeddedness of sport in national policy systems (Houlihan and White, 2002; Roche, 1993). A further strength of the multiple streams framework is that it has been applied successfully in a number of policy areas. Within sport policy analysis its application has been limited although Chalip (1996) used the framework as an element of his analysis of New Zealand sport policy and, more recently, Bergsgard (2000) drew upon the related analytic framework of Cohen et al. (1972) to analyse decision-making in Norwegian sport.

However, the weaknesses and limitations of the model become apparent when judged against the four criteria. The model offers, at best, a partial analysis of stability and change because, while it draws attention to the role of chance and the actions of policy entrepreneurs, it does so through a relative myopia towards structural factors and institutionalized power. The weakness in the framework’s underlying theory of power is especially important due to the prominence given to ideas in the policy process, where the significance of ideas in relation to inter- ests and power is under-theorized. Second, the preoccupation with agenda setting results in the relative neglect of other stages of the policy process, especially implementation. Third, and more positively, the multiple streams model is clearly applicable across a range of policy areas including sport, although it is arguable that it is less easily transferable across political systems and would be less illuminating in more centralized political systems such as France and the UK. Finally, and also more positively, the model allows analysis over the medium term although it is not a specific requirement of application.

172 INTERNATIONAL REVIEW FOR THE SOCIOLOGY OF SPORT 40(2)

01_IRS articles 40_2 9/6/05 1:27 PM Page 172

The Advocacy Coalition Framework

The advocacy coalition framework (ACF) has emerged as a highly regarded basis for policy analysis, both in its own right and in combination with other frame- works. The ACF is based on five assumptions:

• a time perspective of at least 10 years is required for the analysis of policy change;

• a focus on policy sub-systems/policy communities; • sub-systems involve actors from different levels of government and increas-

ingly from international organizations and other countries; • the possession and use of technical information is important; and • public policy incorporates implicit ‘sets of value priorities and causal

assumptions about how to realize them’ (Jenkins-Smith and Sabatier, 1994: 178).

Policy sub-systems normally comprise between two and four coalitions competing for influence, although one might be the dominant coalition. Belief systems provide the source of cohesion within coalitions, with beliefs being disaggregated into three levels: first, ‘deep core’ beliefs which cover basic values regarding, for example, gender relations and individual property rights; second, ‘policy core’ beliefs which are the basic normative commitments within the sub- system; and third, secondary policy core beliefs which are narrower beliefs regarding, for example, the seriousness of particular issues and the details of resource allocation within the sub-system.

Conflict between coalitions, often mediated by a ‘policy broker’, is a source of policy outputs and policy change although change can occur as a result of medium- to long-term ‘policy-oriented learning’ (Sabatier, 1998: 104). Policy- oriented learning describes relatively long-term changes in beliefs that result from ‘experience and/or new information’ (Sabatier, 1998: 104). Although Sabatier accepts that coalitions will resist information that challenges core beliefs, he argues for the essential rationality of coalition behaviour. Further sources of policy change include exogenous developments such as economic recession and war, and also changes in personnel, both of which constitute examples of ‘non- cognitive source[s] of change that can substantially alter the political resources of various coalitions and thus policy decisions’ (Sabatier, 1998: 105).

The ACF has been applied in over 30 case studies across a range of policy areas mainly in the United States, but also in Canada, the UK, Italy, and Poland. This led Sabatier to conclude that ‘the existence of advocacy coalitions . . . is now largely beyond dispute’ (Sabatier, 1998: 108). In many respects the ACF meets the criteria established previously. Stability is explained in terms of dominant coalitions and the persistence of deep core and policy core beliefs. The explana- tion of change is more complex and, in general, less convincing, relying as it does on a combination of instrumental rationality, exogenous events, and policy learn- ing. In addition, the membership of coalitions is only vaguely specified, as is the relationship between members. Furthermore, the implication that membership of coalitions is open requires more empirical support than is currently available. The framework is also weak in its theorization of power. The concept of power

HOULIHAN: PUBLIC SECTOR SPORT POLICY 173

01_IRS articles 40_2 9/6/05 1:27 PM Page 173

receives very little attention because of the underlying rationalist assumption that, in the medium term, evidence from policy learning will result in policy change even if it challenges policy core beliefs. Power then is a property of ideas rather than the outcome of resource control and the pursuit of interests. The failure of actors to act in accordance with evidence is due to factors such as limited time, computational constraints, or cognitive dissonance rather than the manipulation of the policy agenda through the mobilization of bias. However, the ACF has a broader focus than many of its rivals and has the potential to illuminate aspects of the policy process beyond a preoccupation with agenda setting. Finally, the ACF has, as Sabatier has pointed out, been widely applied across a range of different policy sectors in a number of countries, and has, more importantly, been subject to substantial debate and refinement as a result of application in the field.

The application of the ACF in the sports policy area has been limited. Green and Houlihan’s (2005) analysis of elite sport development policy in Australia, Canada and the UK, found some evidence of the existence of advocacy coali- tions, particularly in the UK policy system, and also confirmed the broad utility of the framework. Parrish (2003) applied the ACF to sports regulatory policy in the European Union, and Houlihan and White (2002) developed a modified version of the ACF in their study of UK sport development policy, reaching similar conclusions to Green and Houlihan (2005). All three studies found that the ACF was a valuable starting point for the development of analytic frame- works capable of illuminating the sport policy area. In summary, when assessing the ACF against the four criteria it is clear that the framework attempts to address explicitly both stability and change although, as mentioned, the reliance on the combination of instrumental rationality and exogenous events is only partially satisfactory. However, with regard to the second and third criteria, the ACF is successful in offering a holistic analysis of the policy process and has also been applied across a range of policy areas, although often selectively rather than comprehensively. Finally, the framework is explicit in requiring a medium-term period of analysis before conclusions can be drawn about the nature of the policy process.

Discussion

The four frameworks for policy analysis outlined here differ substantially in the aspects of the policy process they illuminate. Differences exist, for example in the relative significance granted to structure and agency, the emphasis on explaining policy stability in contrast to change, and the weight given to ideas. While the stages model is strongly criticized it is surprising how many frameworks use its concepts, particularly agenda setting and implementation. This suggests that, while the model as a whole has substantial weaknesses, its component concepts are capable of extraction and adaptation, if only as prompts to enquiry or when incorporated into a more comprehensive framework.

Despite the variety of approaches under the umbrella of institutionalism there are a number of common concepts and concerns, the most significant of which is the requirement to take greater account of the institutional context, both organi-

174 INTERNATIONAL REVIEW FOR THE SOCIOLOGY OF SPORT 40(2)

01_IRS articles 40_2 9/6/05 1:27 PM Page 174

zational and cultural, within which public officials work. Among institutionalists there is a particular concern to recognize state agencies as policy actors and, frequently, lead policy actors. Since sport services are routinely delivered through a combination of departments and agencies at national and sub-national levels, the importance of taking institutional history, context and resources into account is clear. Sport policies, like other public policies, are not debated, shaped and delivered by an abstract entity, ‘the state’, but by particular states with distinctive administrative structures, coalitions, codes, values and histories, all of which affect the behaviour of particular public officials when faced with specific issues. Beyond this admittedly important, but one-dimensional admonition, institution- alism appears limited.

While acknowledging the insights offered by the stages model and institu- tionalism, both have weaknesses that make them unsatisfactory as singular frameworks for the analysis of sport policy. By contrast, both the multiple streams framework and the advocacy coalition framework offer richer and more elaborated tools for analysis. At first sight the multiple streams framework appears to be too closely dependent on the United States context to be of value elsewhere. However, while the process of policy-making that Kingdon described is, in part, attributable to the particular structure of the United States political system, and the multiple access points that the institutional arrangements in Congress, the executive and at the state level allow, it is also in part a conse- quence of the increase in the quantity and complexity of issues with which government is involved. The increasing burden on the public policy agenda, and the greater complexity of those issues is a feature of most, if not all, political systems in economically advanced countries. While applications of the multiple streams framework within sport are limited, there is sufficient evidence of its potential utility to encourage its future adoption. The linking of sport policy to non-sport issues and objectives such as nation-building, health, character development, and urban regeneration, is routine. Also the ways in which sports policies, such as sport development, evolve arguably provides as much evidence of the role of chance and opportunism as it reflects strategy and planning (Sibeon, 1997). The framework is also a valuable corrective to rationalist assumptions inherent in much policy analysis. Despite the need to be wary of ‘great man’ explanations that rely too heavily on contributions of particular individuals, the concept of the ‘policy entrepreneur’ might fit well in a policy area where the institutionalization and influence of sports lobbying is variable. Overall, the multiple streams framework has a degree of subtlety and sophistication that suggests its potential value beyond the peculiarities of the United States political system.

Of equal interest is the advocacy coalition framework with its emphasis on the study of long-term policy change, the importance of a hierarchy of beliefs, and the significance of the ideology/belief systems of policy-makers. The centrality of the concept of the policy sub-system or policy community, a feature it shares with the multiple streams framework, anchors the ACF in the main- stream of policy analysis on both sides of the Atlantic, but allows for the incor- poration of the valuable concept of competing coalitions within the same policy sub-system. Finally, the concept of the ‘policy broker’, which is close to

HOULIHAN: PUBLIC SECTOR SPORT POLICY 175

01_IRS articles 40_2 9/6/05 1:27 PM Page 175

Kingdon’s idea of the ‘policy entrepreneur’, is also one that has a particular resonance in the area of sport policy, where the level of organizational complex- ity alone suggests the scope for such a role.

In summary, the major current frameworks for analysis of policy show not only the richness of theory building and concept development, but also have significant weaknesses or lacunae which make them inadequate on their own as tools of analysis. This review suggests that an adequate analytic framework should be broadly consistent with the following characteristics. First, that the framework should allow an investigation of the interplay between structure and agency. Second, that considerations of structure should incorporate the adminis- trative infrastructure of the state, the structure of ideas/beliefs, norms and values in a society, the pattern of non-state organized interests that represent sports interests, and also the interaction between ideas and interests. Third, that the framework should be capable of explaining both stability and change. And finally, that the framework should enable a medium-term (five to 10 year) historical analysis of policy change.

A Framework for Analysing Sport Policy

Given the large number of existing meso-level models/frameworks for policy analysis it is important to avoid adding to the proliferation unnecessarily. With this in mind, the most fruitful strategy for framework construction is to take the ACF as the starting point, due to its ambitions to be comprehensive, the extent of application in the field of policy analysis and its particular concern with policy change over the medium term. At the heart of the proposed framework is a recognition that values, beliefs and norms have a capacity to constrain policy choice, but that their significance depends on the level at which they operate within the political system and the interests with which they are linked. Sensitivity to the potential importance of beliefs and values acknowledges that while organizational and interest group power is important, policy change is the product of a more complex range of influences. Table 1 provides a definition of the key variables, and illustrations of their interpretation.

To understand the development of sport policy and its current salience it is important to appreciate the environment in which it emerged and continues to operate. At a fundamental level it is necessary to identify deep seated policy predispositions or ‘storylines’ (Fischer, 2003), but it is also important to acknow- ledge that while the ideological context of policy will be significant, so too will factors such as the structure of the machinery of government and the pattern of interest group activity. Benson (1982) suggested, similarly to Sabatier, that it was possible to analyse policy in terms of the metaphor of levels where each level is partially autonomous, but embedded in a deeper level which sets limits on the degree of autonomy. At the shallowest, but still significant, level is the pattern of administrative arrangements for a service that refers to the organizational location of a function, the extent of division of labour between units, and the form of control exercised over subordinate units. Within the sport policy area in the UK, initial reliance on local authorities has, over the years, been augmented by a range

176 INTERNATIONAL REVIEW FOR THE SOCIOLOGY OF SPORT 40(2)

01_IRS articles 40_2 9/6/05 1:27 PM Page 176

of national agencies linked to a variety of central departments. The overall result has been a steady growth in organizational complexity and specialization. While institutional/administrative arrangements are located at the shallowest level they do become embedded over time (Granovetter, 1985) and are linked structurally to the deeper layers of the social formation, ‘nested within an ever-ascending hierarchy of yet more fundamental, yet more authoritative rules and regimes, and practices and procedures’ (Goodin and Klingemann, 1996: 18). Administrative units therefore affect policy as a consequence of their tendency, over time, to develop relatively stable preferences for policy tools, perceptions of problems, and modes of working which constrain the response to new issues.

In the UK it is a recurring complaint from sports ministers that making sport policy normally involves liaison and negotiation with a large number of other departments who have a secondary interest in the area. At times this administra- tive dispersal has been referred to pejoratively as fragmentation and the margin- alization of sport within large ministries responsible for health, education, and community safety for example. At other times the fact that sport impinges on the portfolio of ministers for education and health for example, has been interpreted more positively (Houlihan and Green, forthcoming) as it is seen as creating an opportunity to promote the instrumental value of sport and thus access additional public resources. Allocation of function is also important because of the tendency of administrative units to develop a distinctive culture and set of embedded practices that frame the response to particular issues. For example, attempts by a recent UK Home Secretary, David Blunkett, to incorporate a concern for community development (in which sport has a significant role to play) into a department that traditionally has focused more narrowly on policing and imprisonment have proved extremely frustrating. Similarly, concerns have been expressed recently about the appropriateness of UK Sport, with its dominant culture of elite achievement, having primary responsibility for the country’s anti- doping efforts.

Closely related to the pattern of administrative arrangements and also nested within the more shallow structures of society, is the pattern of inter- organizational resource dependencies which is concerned with questions of the distribution among organizations of resources such as expertise, finance, facili- ties, potential elite athletes, authority, and administrative capacity. A recent example of resource dependence in the area of sports policy would be when schools link with local sports clubs with the former providing access to its students (potential club recruits) and the latter providing coaching expertise. Similarly, the British Sports Council had, during the 1970s, provided ‘pump- priming’ funding for local authority facility development projects where the latter provided revenue funding and administrative resources. More significantly, the introduction of new funding sources such as the National Lottery has increased markedly the influence of central government and its agencies. For example, UK Sport has used its resource control to force national sports organi- zations (NSOs) to ‘modernize’, which has generally been interpreted as requiring NSOs to professionalize their management, introduce more effective and transparent target-setting, and improve their talent identification and develop- ment processes. The complex pattern of resource dependency between NSOs,

HOULIHAN: PUBLIC SECTOR SPORT POLICY 177

01_IRS articles 40_2 9/6/05 1:27 PM Page 177

178 IN

TERN ATIO

N AL REVIEW

FO R TH

E SO CIO

LO G Y O

F SPO RT 40(2)

Table 1 Levels of Policy Analysis: Illustrations Drawn from the UK

Variables Illustration of variables Examples of the significance of variables

Shallow Administrative Changes in the distribution of functions between Transfer of responsibility for sport, in 1990, from the Department of the and more arrangements central government departments, between levels Environment to the Department of Education and Science. The effect was vulnerable of government, or between government and semi- to weaken links with local government, but put the Minister with to change independent agencies/ voluntary organizations. responsibility for sport closer to the important debate on the National

Curriculum and the place of Physical Education in schools. Two years later responsibility for sport was transferred again to the new Department of National Heritage which gave sport a much higher profile in government as the Minister had a seat in the Cabinet.

Pattern of inter- Changes in funding primarily, but also changes Introduction of the National Lottery, in 1994, increased funding organizational resources in access to other resources such as expertise and influence of central government at the expense of the Sports Councils dependencies political legitimacy. and local government, especially as local government expenditure was

under increasing pressure during this time.

More stable Interaction between with change structural interest over the groups: medium term

demand groups The consumers of most public services, such as Still of marginal importance with the exception of some elite athletes who tenants, patients, and students, are poorly organized have influence through players’ unions and bodies such as the BOA and often neglected during policy discussions. Athletes Commission. Users of public sector facilities are very weak.

provider groups The PE Association of the UK (PEAUK), the PEAUK was prompted into action in the late 1980s due to the threat to National Association for Sports Development, and exclude PE from the compulsory element of the National Curriculum. The the Institute of Sport and Recreation Management Association has remained an active lobbyist on behalf of school sport represent the interests of PE teachers, sports and PE. development officers and sports centre managers respectively.

0 1 _ I R S a r t i c l e s 4 0 _ 2 9 / 6 / 0 5 1 : 2 7 P M P a g e 1 7 8

H O U LIH

AN : PU

BLIC SECTO R SPO

RT PO LICY

179

direct support groups On issues related to school sport and PE bodies such The YST, CCPR, and national governing bodies of sport provide as the Youth Sport Trust (YST), the National intensive lobbying around the content of the PE national curriculum and Governing Bodies of sport, and the Central Council the pattern of sport offered both in school time and during ‘out of hours’ of Physical Recreation (CCPR) would fall into this activities. During recent years periodic surveys by the National Playing category. Fields Association highlighted the loss of school playing fields to

commercial developers and pressured government to limit the sale of school playing fields

indirect support groups On issues such as school sport and PE examples Periodic surveys by the SHA, in the 1990s, highlighted the gradual would include the Secondary Heads Association erosion of curriculum time for PE in secondary schools. The BLF (SHA) and the Big Lottery Fund (BLF, a funding recently provided £750m to help schools improve the quality of their stream of the National Lottery). sports facilities.

Service specific policy ‘Sport For All’ as a unifying policy objective in the Increasing selectivity, based on medal-winning potential and effectiveness paradigm 1970s gave way in the 1990s to a preoccupation with of youth development, has left many sports very short of resources.

elite success and school sport. Public funding for national governing bodies has become increasingly selective.

Dominant core policy Privatization, and greater emphasis on cost recovery Dominance, during the 1980s, of neo-liberal economics reflected in paradigm and the concept of ‘user pays’ were the dominant privatization and managerialism. Compulsory Competitive Tendering

policies of the 1980s, but were replaced, partially, introduced for leisure services and, in many local authorities, after 1997 by social inclusion, modernization and undermined Sport For All policies. Current emphasis on social joined-up government. inclusion has redirected some local authority policy away from sports

(skills) development to sports (social skills) development.

Deeply Embedded structural Patriarchal attitudes to participation in sport. Particularly from 1997, the mythology of British amateurism has been entrenched values fundamental to Deference. Class based pattern of participation dispelled and replaced by an acknowledgement of the need for greater

the social formation reflected, for example, in the persistence of the professionalism in the preparation of elite athletes. From the early 1990s distinction between amateur and professional the concept of equity has gradually become established in the sports participation. Voluntarism preferred to state policy discourse. intervention.

Source: Houlihan and White (2002); framework adapted from Benson (1982) and Sabatier (1999).

0 1 _ I R S a r t i c l e s 4 0 _ 2 9 / 6 / 0 5 1 : 2 7 P M P a g e 1 7 9

voluntary sports clubs and public agencies has substantial implications for the process of policy-making (due to the increase in the number of clearance points in decision-making) and for the role and influence of the state. While the greater complexity of resource dependency relationships is a symptom of the emergence of a ‘congested state’, it is also an indication of the extended state which has steadily retreated from direct service delivery but has extended its capacity to steer policy sub-systems through the strategic control over resources.

Operating within the matrix created by administrative structures is a set of structural interest groups. Adapting Benson’s classification there are four types of structural interest group that are relevant to the study of sports policy within a UK context: namely, demand groups (consumers of policy outputs such as facility users, clubs, and elite sportsmen and women), provider groups (the deliverers of services such as leisure services managers, PE teachers, coaches, youth workers, sports development officers, sports clubs, and commercial clubs/facility providers), direct support groups (those groups upon which organi- zations depend for systems support, such as national governing bodies of sport, commercial sponsors, local authorities and schools), and indirect support groups (related local authority services including land use planning, community devel- opment, non-Sports Council sources of funding such as the Foundation for Sports and the Arts, and commercial sponsors). It is the interaction between these inter- est groups that provides the policy process with an important dynamic. However, it also needs to be noted that analysing the interaction between these groups and the machinery of government is also significant for policy analysis. In particular it is important to acknowledge that the network of groups is both a source of demands but also, in the UK at least, the government’s infrastructure for imple- mentation. Consequently, there are many examples of government intervening, often under the label of ‘modernization’, to reshape the pattern of interest groups. In the UK, the Minister for Sport recently told three organizations that represent facility managers and sports development staff that he wanted them to merge as he would only talk to one organization in future. Similar ministerial pressure to merge was also placed on PE teachers’ organizations.

It is at the level of structural interest groups that the distribution of costs and benefits by governments and the impact of exogenous factors such as economic recession, inflation or economic growth have a direct impact and where indi- vidual and collective interests (whether class, sport specific, spatial, community, age, gender, etc.) experience and interpret policy outcomes and consequently generate demands and supports within the policy area. The number, degree of specialization and level of activity of structural interest groups has grown steadily within the UK sport policy area. While the capacity of structural interest groups to shape proactively the policy agenda is still open to question it is more plausible to argue, as do Houlihan and White (2002), that structural interest groups have a significant defensive potential to protect recent resource gains. What prevents this neo-pluralist model of the policy process drifting towards pluralist utopianism is an acknowledgement of the significance of the facilitative and constraining impact of the dominant policy paradigm, the service specific core policy paradigm, and what Benson refers to as the rules of structure forma- tion.

180 INTERNATIONAL REVIEW FOR THE SOCIOLOGY OF SPORT 40(2)

01_IRS articles 40_2 9/6/05 1:27 PM Page 180

The dominant policy paradigm is the set of values and assumptions that influence policy choice and administrative practice across a range of services. Just as ‘privatization’ was a prime example of the dominant policy paradigm of the Thatcher years, the current UK Labour government’s dominant policy para- digm would include values such as social inclusion, modernization and ‘joined- up’ government. These Labour values have been introduced, admittedly with varying degrees of success, across all central government departments and have been particularly significant in shaping sport policy since 1997. Nested within the dominant policy paradigm will be a service-specific policy paradigm. In sport the paradigm has shifted between emphases on ‘sport for all’, social inclusion, elite development, and fitness and health. Recently, in the UK, there have been a number of subtle shifts in the service-specific policy paradigm, an example of which has been the steady retreat from a policy discourse centred on the value of ‘sport for all’ as an integral part of the process of elite development to one which emphasizes the demonstration effect of elite success as a catalyst for increasing participation. The former discourse assumed an asystematic approach to elite development according to which mass participation was important in providing the volume of participants from which the elite would emerge. In contrast, the current emphasis is on improving the sophistication of early talent identification, and concentrating that talent in a limited number of specialist training centres or development clubs (thus probably improving Olympic medal chances), but clearly insulating the elite development process from wider issues relating to participation.

It is at the levels of the dominant policy paradigm and the service-specific policy paradigm that ideas have their most obvious impact. Elite development, using sport to tackle youth disaffection, and strategies to promote physical activity are all subject to extensive policy transfer and policy-oriented learning between countries. However, while substantial policy-oriented learning is a key assumption of the ACF, this emphasis on the medium-term rationality of the policy process needs to be tempered with an acknowledgement that policy actors do not always refer to evidence and the weight of expert opinion. Very often actors turn to, what Fischer refers to as, ‘discursive storylines’ that are decision- making short cuts which ‘function to condense large amounts of factual informa- tion inter-mixed with normative assumptions and value orientations that assign meaning to them’ (2003: 87). Storylines might cover the role of men and women in sport, the assumption that participation in sport builds character, or the idea that sport should be self-regulating. Whether particular storylines are considered benign or malign depends, obviously, on one’s own normative position.

Establishing and maintaining dominant storylines is both a contemporary struggle between interests, and a reflection of deeply entrenched and longer established biases within the policy process. As such, they are part of the rules of structure formation that set the limits to policy action, both by defining activities that are acceptable and those that are not. Benson’s notion of ‘deep structure’ corresponds broadly to the ideas of the cultural institutionalists and relates to the biases inherent in the fabric of the political process. At this level are those beliefs and values that are both taken for granted and very slow to change, including beliefs about the appropriate/acceptable involvement in sport by men and

HOULIHAN: PUBLIC SECTOR SPORT POLICY 181

01_IRS articles 40_2 9/6/05 1:27 PM Page 181

women, attitudes towards amateurism in sport, and acceptable behaviour by sportsmen and women on the field of play.

The final element in this framework is the role of the policy broker. On the one hand, to promote the contribution to policy making of the individual actor challenges the assumptions implicit in much meso-level policy analysis which emphasizes the central role for organizations and groups and downplays the importance of the individual. On the other hand, in congested policy making systems there is greater scope for individual policy entrepreneurship, especially where, as in the case of sport, the organizations and interest groups have yet to become deeply embedded within the political/administrative infrastructure. In addition, there is evidence from studies in the UK that the role of policy broker is being fulfilled, and is important in shaping policy in some areas such as school sport and physical education (Houlihan and Green, forthcoming).

In conclusion it is acknowledged that this framework is not without its problems – for example, the identification of discrete levels of analysis, the articulation of the pattern of influence between levels, and the relative autonomy that each level enjoys. However, this modified version of the advocacy coalition framework is nonetheless valuable in that it highlights the importance of an understanding of the social and recent historical context of policy decisions, and encourages us to look beyond the surface phenomena of day-to-day politics. More specifically the modified ACF outlined above provides a framework that enables the researcher to acknowledge, and also systematically investigate, the complexity and ‘messiness’ of contemporary sports policy making. In so doing the framework helps to avoid the premature acceptance of unicausal explanations of policy change, whether they be in terms of class power or interest activity. Second, the framework recognizes the importance of both structures (as adminis- trative arrangements and as embedded practices and values) and agency. Finally, the framework, with various modifications, has the advantage of having been applied widely across different policy areas and in different countries.

At the start of this article the absence of a tradition of policy analysis of sport was highlighted as a significant gap in the social scientific study of the field. The increasingly prominent role of the state as variously promoter, regulator, resource provider, manipulator and exploiter of sport is beyond challenge. The old exhortation heard from many sports administrators and politicians ‘to keep politics out of sport’ has long been abandoned. While it would be an exaggera- tion to say that sports social scientists want to ‘keep policy analysis out of sport’ there does seem to be a marked reluctance to acknowledge the contribution that this discipline can make to our understanding of sport and society. This article contributes to the discussion of the benefits of widening the range of disciplines available for the study of sport.

Notes

1. The author gratefully acknowledges the thoughtful comments of the Editor, Prof. Jean Harvey, University of Ottawa and the two anonymous referees.

2. For example, the Journal of Transport Economics and Policy and Transport Policy. 3. For example, Health Policy and Health Education Policy.

182 INTERNATIONAL REVIEW FOR THE SOCIOLOGY OF SPORT 40(2)

01_IRS articles 40_2 9/6/05 1:27 PM Page 182

4. For example, the Journal of Education Policy. 5. For example, Contemporary Economic Policy and Oxford Review of Economic Policy. 6. For example, the Journal of European Social Policy and Critical Social Policy. 7. For example, Environmental Economics and Policy Studies and Environmental Engineering and

Policy. 8. For example, International Journal of Cultural Policy and Culture and Policy. 9. Between the structure and agency extremes, approaches such as the ‘new political economy’

theories may be seen as non-structuralist variants of Marxism that consider ‘collective agents’, such as social classes and social movements in their analyses of social change. These have not yet been incorporated directly into work on policy analysis.

10. For Haas, an epistemic community is ‘a network of professionals with recognised expertise and competence in a particular . . . issue area’ (1992: 3). He argues that ‘control over knowledge and information is an important dimension of power’ and suggests that the potential of epistemic communities to exercise influence increases with uncertainty and particularly the uncertainty found in areas of policy where states are strongly dependent on the policy choices of other actors.

References

Baumgartner, F. and Jones, B. (1993) Agendas and Instability in American Politics. Chicago, IL: University of Chicago Press.

Benson, K. (1982) ‘Networks and Policy Sector: A Framework for Extending Interorganizational Analysis’, in D. Rogers and D. Whetten (eds) Interorganisational Co-ordination, pp. 137–76. Iowa: Iowa State University.

Bergsgard, N.-A. (2000) ‘National Facilities for Ski Sport – “Yes Please – All Three”’, paper presented at Pre-Olympic Congress, Brisbane, September.

Berry, F.S. and Berry, W. (1992) ‘Tax Innovation in States: Capitalizing on Political Opportunity’, American Journal of Political Science 36: 715–42.

Birley, D. (1996) Playing the Game: Sport and British Society 1914–1945. Manchester: Manchester University Press.

Carrington, B. and McDonald, I. (2000) ‘Race’, Sport and British Society. London: Routledge. Chalip, L. (1996) ‘Critical Policy Analysis in Sport: The Illustrative Case of New Zealand Sport

Policy Development’, Journal of Sport Management 10: 310–24. Cohen, M.D., March, J.G. and Olsen, J.P. (1972) ‘A Garbage Can Model of Organisational Choice’,

Administrative Science Quarterly 17: 1–25. de Leon, P. (1999) ‘The Stages Approach to the Policy Process: What Has it Done? Where Did it Go

Wrong?’, in P. Sabatier (ed.) Theories of the Policy Process, pp. 19–32. Boulder, CO: Westview Press.

Dery, D. (1999) ‘Policy By the Way: When Policy is Incidental to Making other Policies’, Journal of Public Policy 18(2): 163–76.

Fischer, F. (2003) Reframing Public Policy: Discursive Politics and Deliberative Practices. New York: Oxford University Press.

Giulianotti, R. (1994) ‘Social Identity and Public Order: Political and Academic Discourses on Football Violence’, in R. Giulianotti, N. Bonney and M. Hepworth (eds) Football, Violence and Social Identity, pp. 10–36. London: Routledge.

Goodin, R.E. and Klingemann, H.-D. (1996) ‘Political Science: The Discipline’, in R.E. Goodin and H.-D. Klingemann (eds) A New Handbook of Political Science, pp. 3–49. Oxford: Oxford University Press.

Granovetter, M. (1985) ‘Economic Action and Social Structure: The Problem of Embeddedness’, American Journal of Sociology 91(3): 481–510.

Green, M. (2003) ‘An Analysis of Elite Sport Policy Change in Three Sports in Canada and the UK’, unpublished PhD thesis, Loughborough University, UK.

Green, M. and Houlihan, B. (2005) Elite Sport Development: Policy Learning and Political Priorities. London: Routledge.

Guba, E.G. and Lincoln, Y.S. (1987) ‘The Countenances of Fourth Generation Evaluation:

HOULIHAN: PUBLIC SECTOR SPORT POLICY 183

01_IRS articles 40_2 9/6/05 1:27 PM Page 183

Description, Judgement and Negotiation’, in D. Palumbo (ed.) The Politics of Program Evaluation, pp. 202–34. Newbury Park, CA: Sage.

Haas, P.M. (1992) ‘Introduction: Epistemic Communities and International Policy Co-ordination’, International Organization 46(1): 1–35.

Hall, P. (1986) Governing the Economy: The Politics of State Intervention in Britain and France. Oxford: Oxford University Press.

Hargreaves, J. (1994) Sporting Females: Critical Issues in the History and Sociology of Women’s Sport. London: Routledge.

Henry, I. (2001) The Politics of Leisure Policy, 2nd edn. London: Palgrave. Hogwood, B. and Gunn, L. (1984) Policy Analysis for the Real World. London: Oxford University

Press. Houlihan, B. (1990) ‘The Politics of Sport Policy in Britain: The Examples of Football Hooliganism

and Drug Abuse’, Leisure Studies 9: 55–69. Houlihan, B. (1991) The Government and Politics of Sport. London: Routledge. Houlihan, B. (2000) ‘Sporting Excellence, Schools and Sports Development: The Politics of Crowded

Policy Spaces’, European Physical Education Review 6(2): 171–93. Houlihan, B. and White, A. (2002) The Politics of Sport Development: Development of Sport or

Development through Sport? London: Routledge. Houlihan, B. and Green, M. (forthcoming) ‘The Changing Status of School Sport and Physical

Education: Explaining Policy Change’, Sport, Education and Society. Howlett, M. and Ramesh, R. (1995) Studying Public Policy: Policy Cycles and Policy Sub-systems.

Oxford: Oxford University Press. Jenkins-Smith, H. and Sabatier, P. (1994) ‘Evaluating the Advocacy Coalition Framework’, Journal

of Public Policy 14: 175–203. John, P. (1998) Analysing Public Policy. London: Pinter. Kickert, W., Klijn, E. and Koppenjan, J. (eds) (1997) Managing Complex Networks: Strategies for the

Public Sector. London: Sage. Kingdon, J. (1984) Agendas, Alternatives and Public Policy. Boston, MA: Little Brown. Kingdon, J. (1995) Agendas, Alternatives and Public Policy, 2nd edn. Boston, MA: Little Brown. Kraus, R. (1990) Recreation and Leisure in Modern Society, 4th edn. New York: Harper Collins. Lipsky, M. (1980) Street Level Bureaucracy: Dilemmas of the Individual in Public Service. New

York: Russell Sage Foundation. Macintosh, D. (1991) ‘Sport and the State: The Case of Canada’, in F. Landry et al (eds) Sport . . .

The Third Millennium, pp. 269–75. Sainte-Foy: Les Presse de Universitairé de Laval. Macintosh, D. and Whitson, D. (1990) The Game Planners: Transforming Canada’s Sports System.

Montreal: McGill-Queens University Press. Marsh, D. (1998) Comparing Policy Networks. Buckingham: Open University Press. Ostrom, E. (1999) ‘Institutional Rational Choice: An Assessment of the Institutional Analysis and

Development Framework’, in P. Sabatier (ed.) Theories of the Policy Process, pp. 35–71. Boulder, CO: Westview Press.

Parrish, R. (2003) ‘The Politics of Sport Regulation in the European Union’, Journal of European Public Policy 10(2): 246–62.

Pickup, D. (1996) Not Another Messiah: An Account of the Sports Council 1988–1993. Bishop Auckland: Pentland Press.

Pontusson, J. (1995) ‘From Comparative Public Policy to Political Economy: Putting Political Institutions in their Place and Taking Interests Seriously’, Comparative Political Studies 28(1): 117–47.

Pressman, J. and Wildavsky, A. (1973) Implementation. Berkeley: University of California Press. Roche, M. (1993) ‘Sport and Community: Rhetoric and Reality in the Development of British Sport

Policy’, in J.C. Binfield & J. Stevenson (eds) Sport, Culture and Politics. Sheffield: Sheffield Academic Press.

Sabatier, P. (1998) ‘The Advocacy Coalition Framework: Revisions and Relevance to Europe’, Journal of European Public Policy 5(1): 98–130.

Sabatier, P. (1999) ‘The Need for Better Theories’, in P. Sabatier (ed.) Theories of the Policy Process, pp. 3–17. Boulder, CO: Westview Press.

184 INTERNATIONAL REVIEW FOR THE SOCIOLOGY OF SPORT 40(2)

01_IRS articles 40_2 9/6/05 1:27 PM Page 184

Sibeon, R. (1997) Contemporary Sociology and Public Policy Analysis: The New Sociology of Public Policy. London: Kogan Page and Tudor.

Thelen, K. and Steinmo, S. (1992) ‘Historical Institutionalism in Comparative Politics’, in S. Steinmo, K. Thelen and F. Longstreth (eds) Structuring Politics: Historical Institutionalism in Comparative Analysis, pp. 1–32. Cambridge: Cambridge University Press.

Thomas, N. (2003) ‘Sport and Disability’, in B. Houlihan (ed.) Sport and Society: A Student Introduction, pp. 105–24. London: Sage.

Tsebelis, G. (1990) Nested Games: Rational Choice in Comparative Politics. Berkeley, CA: University of California Press.

Williams, W. (1998) Honest Numbers and Democracy: Social Policy Analysis in the White House, Congress and the Federal Agencies. Washington, DC: Georgetown University Press.

Wilson, J. (1994) Playing by the Rules: Sport, Society and the State. Detroit, MI: Wayne State University Press.

Barrie Houlihan is Professor of Sport Policy in the Institute of Sport and Leisure Policy, Loughborough University. His most recent books include, The Politics of Sports Development: Development of Sport or Development through Sport? (with Anita White; Routledge, 2002); Sport and Society: A Student Introduction (edited volume, Sage, 2003); and Elite Sport Development: Policy Learning and Political Priorities (with Michael Green; Routledge, 2005).

Address: Institute of Sport and Leisure Policy, School of Sport and Exercise Sciences, Loughborough University, Loughborough LE11 3TU, UK. Email: [email protected]

HOULIHAN: PUBLIC SECTOR SPORT POLICY 185

01_IRS articles 40_2 9/6/05 1:27 PM Page 185