The narrative review considered peer-reviewed journal articles and books/book chapters. The data was collected based on simple searches through Google Scholar and academic institution libraries, the latter of which allowed for specific search terms to be entered. No restrictions on time-frames were set, “gray” literature was not included, and articles with an overt focus on elite sport were excluded. The key search terms applied were bottom-up policy, community sport, grassroots sport, meso-level, policy frameworks, sport, sport policy, and top-down policy.

In the first instance, this paper does not offer a comprehensive analysis. As a narrative review, the purpose is to provide a descriptive, not systematic, outline of the situation. Narrative reviews are not designed to be conclusive but instead provide a platform for further research (39). Unlike systematic reviews, for example, there are fewer guidelines governing narrative reviews (253). Researchers choose what should be the focus of the review and what will be included in order to offer useful research (255). Indeed, such literature reviews are almost always selective, non-exhaustive, and involve less demanding search methods (39).

Historically, analyses of sport policy have addressed a range of matters including political issues such as foreign affairs, developments relating to governance, and concerns relating to equality, such as people with disabilities (43) and, more recently, issues relating to physical fitness and health, participation in sport, and general welfare (26). The scope of policy analysis includes analysing data on government policy intentions and the efficacy of government initiatives (4, 19, 44). Such analysis has consistently stressed the fact that government investment in sport is inconsistent (45–47), yet given government policy on health and participation, sound policy processes and evaluation might be considered essential to sport policy architects (6, 12, 48). Consequently, although government intervention in sport policy, including defining strategies and subsequent evaluation, has been the subject of academic enquiry, few analyses have addressed sport policy effectively (14, 49). In that context, a greater awareness of the stages of the policy-making process and the nature of policy subsystems—defined by the diversity and number of stakeholders and their policy focus (50–52)—can help policymakers leverage sport's inextricable role in society (12, 53, 54) and to address issues pertaining to macro-political agenda-setting (37).

From the perspective of understanding sport's relationship to society—its social role, especially at grassroots level (54–56)—it is useful to recognize the utility of mainstream analytical frameworks in the policy process, particularly to appreciate why the theoretical constructs of sport policy are based on the importance of meso-level assessment, as opposed to macro or micro-level approaches. An initial consideration is that sport policy serves aspects of more general public policy, particularly concerning community policy matters, where meso-level frameworks are commonly applied (57–59). Labelled as “policies that originate within, or are dependent upon the resources of, the state” (4), public policy significantly shapes sport policy (37, 60). Informed by macro-level theorising and micro-level evaluation, the direction for robust analysis is linked to policy analysis at the meso-level (61, 62). All meso-level frameworks incorporate macro-level assumptions, and when considering an analytical framework, any structure also needs to adequately cater for matters of stability and change (62, 63), no less so than in the sport policy process (64). Indeed, the scope for expounding the process of change, and a drive toward stability, is of particular importance in any analysis because there is evidence of regular and rapid change in sport policies and subsequent objectives globally (4), and in Australia (47, 65–67).

In identifying and illuminating numerous variables that make up the creation of policy, several analytical frameworks have an explicit focus on key stages of the process but are lacking when investigating the connection, and often interdependency, between key stakeholders (4). Further, there is consideration of contemporary policy-making and (a) assessing the nature of application of meso-level frameworks across various disciplines (4) and (b) that the frameworks have been subject to germane level of examination of their value as models for sound analysis of the policy process (62, 68). Sabatier (68) and Houlihan (4) provided a benchmark to gauge the merit of a meso-level approach to policy analysis, as per Table 1.

Sabatier's (68) model supports generic policy analysis, whereas Houlihan's (4) version provides detail relating to sport; there is an emphasis on the conditions of stability and change in sports-related policy (64) due in part to “constant reviews of sports policy in many developed countries” (4). Such a scenario is in keeping with policy review processes in non-sporting disciplines in general which are very much subject to regular evaluation (69, 70). Indeed, policy processes and systems are complex (68, 71), challenged by a range of variables that disrupt the policy process. Such variables include (i) numerous and diverse stakeholders; (ii) timeframes for policy creation, implementation, and evaluation; (iii) involvement of government and the impact of strategies and objectives; (iv) conflict relating to issues and effect of policy change, and (v) coercive relationships directing the policy process (68). Due to the challenging nature of policy-making (14), such wide-ranging variables make it extremely difficult for the analysis to fully address every aspect of the process. In the following section, IA, the MSF, PNT, and the ACF, are outlined for understanding and addressing policy with a particular focus on policy implementation in sport and the impact of political, social, and cultural structures on the policy process (14).

Applied extensively in the policy realm, IA represents a robust meso-level framework that examines institutional pressure on stakeholder conduct by way of a range of regulations and behavioral norms (72–74). Institutionalisation embodies a phenomenon through which stakeholders consent to a collective view leading to legitimisation, which occurs “whenever there is reciprocal typification of habitual actions by types of actors [stakeholders]” (75). From the perspective of research into the policy process, institutions form a central aspect of the arena within which policy-making takes place (63). IA provides estimable understanding of the role of institutional guidance—and coercion—and the resultant behavior of stakeholders (76); indeed, a key attribute of IA is that it facilitates the concurrent evaluation of stakeholder conduct and the establishment in which they operate (77, 78). The formation of an institutional collective relies upon the (a) resources brought to a situation, (b) value assigned to actions, (c) way knowledge and information are acquired and used, and (d) processes and justification for particular courses of action (76). In turn, the realisation of a collective accelerates the concept of an institution (79–81), particularly when considering the institutional nature of sport, both globally (82–87) and in Australia (66, 88–90).

In light of the multiple variables involved, a sound approach for the interpretation and role of institutions in the policy process is necessary (4). Institutions can be defined as (a) organisational entities such as government agencies (e.g., the Australian Sports Commission) and (b) cultural constructs with values and beliefs amongst stakeholders (4). Further, institutions occupy an official place from a legislative or procedural perspective, and are also seated more informally in relation to behavior or customs (76). Within that context, IA focuses on adapting the institutional development process, alongside the reciprocal stakeholder values and beliefs, to resolve policy issues (14). Individually, or as an element of collective, official settings, IA provides an avenue for connecting two or more individuals to align behavior and subsequent action (91, 92), and can help “elucidate the role institutions play in the determination of social and political outcomes” (93).

In terms of the application in a sport setting, IA has been utilized in macro-level analyses, reflected in sport policy literature (94), where the centrality of institutions to the policy-making process is noted (63), and where the significance of the function of institutions in sport policy processes is documented (4, 95–97). O'Brien (98) incorporated IA in a discussion analysing stakeholders working together across unequal governance levels in sport. The imbalance stemmed from the status of government involvement in public policy-making and the attempt to foster a culture of evidence-based policy considerations (98). A noted aspect of the importance of IA in sport policy analysis is based on the reliance many sport organisations have on government resources (14).

Some examples of the strong IA link between government sport policy objectives include national sporting organisations in the United Kingdom (UK), such as Sport England and UK Sport, being heavily influenced by the government and its focus on “modernisation” of sport (99), alongside Scandinavian examples where competitive sport poses an issue from an organisational perspective, and where structural and institutional relationships present challenges for change because they are based on strong “egalitarian” values (100). Other examples of the application of IA include the evaluation of sport development processes and the contribution from political circles, inclusive of local government, and not-for-profit groups (101, 102), along with more recent examples of the conceptualisation of institutional change from an international perspective, and the increasingly topical evaluation policies relating to concussion legislation (103, 104).

At grassroots level, there have been changes in the philosophical foundation of policy together with a strategic shift resulting in the prioritisation of one sport policy objective (e.g., elite sport) over another (e.g., grassroots sport) and the subsequent variation in levels of commitment from different institutions (23, 76). IA offers the potential to commence a process for identifying the key issues confronting grassroots sport (64, 105–108) and within that setting, there is merit in taking grassroots sport clubs as the point of departure in the policy process to build on the sociological perspective of IA (16). Finally, the outcomes emanating from IA are useful in “explaining policy related [sic] actions, interactions and outcomes” (109). Based on such observations, institutions represent independent and intervening variables in the policy-making process (4), a delineation that impacts policy implementation in relation to the differing priorities of implementers vs. government. Indeed, centrally designed policy influences localized implementation, potentially catering for sometimes contradictory but not mutually exclusive perspectives, and offers a potential solution for analysing the implementation process (16).

There is, however, a view that is problematic for IA which is that there is a lack of clarity about what an institution is and what institutionalism means (81). Ostrom's (76) reference to institutional pressures is validated by Thelen's (110) earlier argument that institutions meld the interests of stakeholders based on their relations of power with other stakeholders, which represents constraints and mediating factors in the political arena. Fischer (111) stated aspects of IA pointed to attempts to discreetly manipulate power within institutional structures, and Ostrom (76) recognized numerous organisations were governed by regulations which dictate patterns of interaction for organisations. Cairney (81) proposed that institutions are both the physical structures in which policy is made but also the rules that dictate behavior and how policy is made, with IA providing for a more robust emphasis on regulations, norms, and actions than on conventional arrangements.

The imprecision of IA is in part based on the variations in interpretation, whereas some uncertainties could be dismissed if IA was excluded as a “unified body of thought” (93). From a policy framework perspective, further problems arise because each new framework is perceived as a novel approach to the concept of institutionalism (93). In addition, ensuing deliberations create problems for the theoretical contribution that IA offers because its strength may lie in its multi-theoretical character (112). Thus, rather than strict distinction or disagreement between each version, IA may allow for the assessment of competing propositions drawn from different theories (81, 113). Houlihan (4) points to an historical context to IA, noting the autonomous nature of institutions, which contrasts markedly with resultant constraints enforced on society (114), viz., institutions can restrict options by influencing stakeholder perceptions and interpretations of problems and solutions (4, 115). Further, IA offers a mechanism for the control stakeholder behavior and expectations by addressing issues with generally appropriate methods (116).

Overall, the main disadvantage of IA is the explicit focus on various connected aspects of organisational structures, resulting in a reduced focus on institutional factors (94). In addition, IA generally only facilitates macro-level analysis (14), which is problematic if adopting a emphasis on all stages of policy creation and implementation, particularly the views of stakeholders implementing policy. Further, from a sport policy perspective, IA has been applied to analyse institutional conditions as opposed to technical settings (94) where, for example, the structure of sport organisations and the reliance on government resources results in a scenario where stakeholder behavior is influenced by more than institutional regulations and norms (14). Finally, the presumption that an institution might shape stakeholder opinions, behavior, and interests is potentially inaccurate for policy implementers involved with grassroots sport, where there is a lack of sophistication in terms of the process of implementation (117–119).

Established on a “garbage can” model of organisational choice (120, 121), Kingdon and Thurber's (122) MSF is theoretical paradigm applied in sport policy analysis (14). The concept of policy choice linked to the MSF leads to the garbage can analogy (123), where an array of policy issues, along with potential answers, are discarded (121). Although discussion and reason are fundamental to policy formation (124), the process comes with a degree of complexity where decision-making is characterized as a form of anarchy (125). The MSF emphasizes the chaotic character of organisations involved in policy creation, and political manipulation forms a significant part of the process (4, 51, 126). Indeed, there is a degree of ambiguity and accompanying attempts to regulate such uncertainty, in turn leading to governmental influence that “is more than just persuasion and identity construction” (125). Noting the manipulative and complex nature of the policy-making process (127), the MSF is made up of three distinct streams, namely, problems, policies and politics (51). Further, three additional conventions direct the MSF policy process: (i) individual processing is methodical and concurrent; (ii) there are substantial time limitations; and (iii) the three streams are separate (125).

Based on the key assumptions, the MSF provides an avenue to uncover the connections between individuals that manipulate policy processes, known as policy entrepreneurs, and individuals who are controlled as part of the policy process, the policy-makers (125). Policy entrepreneurs are “more than mere advocates of particular solutions; they are actors [stakeholders] that are power brokers and manipulators of problematic preferences and unclear technology” (125) who try to link the three MSF streams, which is contrary to their ordinarily level of independence (4, 51, 68, 99, 125). Further, policy entrepreneurs address disordered choices that have been created in an imprecise fashion, viz., a chaotic assortment of theories and concepts without any guidelines for implementation (121). Policymakers are those involved in making policies and policy decisions, often members of a government department, legislature, or head of an organisation responsible for making new rules or laws. Policy-makers can also be institutions, as opposed to individuals (128, 129). Often described as experts who consider their role to serve a cause greater than themselves, policy-makers face constant pressure on their cognition and emotion, and need to gather information quickly and effectively (130–133).

In relation to the three streams—problems, policies and politics—the MSF focuses on problems where public stakeholders want solutions, or where policy-makers deem action is required, in contrast to problems that are deliberately disregarded (4, 51). There is also the requirement for consensus to address issues, thus creating a catalyst for ensuring problems become part of a policy agenda (134). The policy stream incorporates a grouping of interrelated stakeholders pursuing a matter of public policy that is important to them for instrumental reasons (135). Alternatively, policy entrepreneurs embrace a particular issue and seek consent for solutions (4), and maintain the salience of a policy issue on the political agenda (95). The political stream occupies a position completely independent of other streams and includes national sentiment, political parties and lobby groups, and government (4, 51, 125). By observing general reactions and opinion, governments are better able to enact specific policy agendas or, conversely, dim the prospects of others (136).

In a sporting context, the MSF has been widely used in sport policy (14) and it is further positioned as a useful platform for sport policy analysis in nations worldwide and with diverse political conditions (4). In terms of policy agendas in sport, limited success on the international stage, declining participation levels, or rising obesity levels, serve as examples of common policy issues, especially based on mainstream initiatives. The scale of specific circumstances, potential for change, and success or failure of existing strategies are focal considerations within the MSF (125). Within that context, the success or otherwise of sport policy “can have a ripple effect through the political system by spilling over into other policy fields” (4), including situations where stakeholders focused on industry-specific initiatives are absorbed into more universal policy directives (60). The enunciation of inclusivity concepts within the sport policy system, however, reveals that sport policy is susceptible to manipulation by political, health and educational interests (137, 138). The matter of rising obesity levels serves as a topical example, demonstrating a direct correlation between declining levels of sport and physical activity and the resultant increase in expenditure on issues related to public health (14). Indeed, the MSF offers insight into the links between physical activity and public policy and helps facilitate a greater understanding of the problem and potential solution (139). Other examples of the MSF evaluation of policy change in physical education policy can be subject to the turnover of government ministers and general bureaucracy (140).

Changes in government and national sentiment can exert a considerable effect on policy agendas (125), pointing to the concept of how events of a political nature can focus attention on the policy process and create opportunities to legitimize sport and shape associated policy (14). By way of example, the accomplishments of British athletes competing at the 2008 Olympic Games in Beijing led to higher levels of government investment in the lead up to the 2012 London Olympics, alongside a general policy shift resulting in significant government support for the staging of an increasing number of international sports events in the UK. In Australia, following the success of the 2000 Sydney Olympics, the Federal government supported the 2016 Commonwealth Games in the Gold Coast and committed to underwrite fifty per cent of the costs for Brisbane to host the 2032 Olympic Games (141). Within that context, the three streams of the MSF link together at critical moments in time, wherein policy choices are made (125). Forming a critical stage of the MSF, the combination of the three streams leads to policy problems achieving recognition by government, and thus greater potential for a solution. For example, in sport there are potential favorable circumstances that relate to funding cycles linked to new polices to, for example, increase participation in physical activity through sport over a set period. Other examples include that of the Brazilian Olympic Committee (BOC) expanded investment possibilities by implementing policies that stimulated preparation for the Olympic and Paralympic Games in 2016 (142). From an MSF perspective, the BOC approach demonstrated the efficacy of the activities of policy entrepreneurs in the high-performance sport policy domain (142). In Australia, policy entrepreneurs have featured in grassroots sport, albeit subject to a complex array of factors internal and external to grassroots sport clubs, in various policy initiatives (13, 143).

In spite of the MSF's application by policy specialists, however, there are criticisms that relate to the assumptions of the framework and the limited perspectives of aspects of the policy process (125). These criticisms include a lack of unambiguous, falsifiable hypotheses within the MSF (62), and that the original “garbage can” model does not elicit robust findings which instead emanate from the MSF assumptions (144). Conversely, outcomes stemming from the MSF were arrived at due to the variations in each of the three streams, linked deliberately by policy entrepreneurs, and where data from different policy areas across various nations has been observed (51, 145, 146). The appropriateness of conceptualising independent streams, however, is debatable and the MSF streams can be deemed mutually dependent, with one prompting variations and modifications in the other(s) (123, 144).

Accordingly, the MSF's aptness for sport policy analysis is brought into question, as evidenced by its limited application (60, 127). Although the framework offers some direction in the policy analysis, that is largely only from a macro perspective, thus impacting its appeal to sport policy research. As a result, the MSF provides limited interpretation of matters relating to policy stability and change (99), particularly due to the claim that the MSF does fully assess policy beyond the initial opportunity (147), instead focussing mainly on formation and disregards other steps, notably matters of implementation (4, 14). Finally, timeframes for change are not an explicit consideration, there is overt attention on the notion that institutional foci influence political agendas (4), and there is an absence of empirical evidence (121). Ultimately, the capacity of the MSF to formulate and assess the complete policy process in sport is constrained; the “model focuses on policy issues, political intervention and solution development” (14).

PNT provides an avenue to assess numerous variables involved in policy change and the policy-making process as part of policy communities. As an integral component of PNT, a policy community is based on restricted, durable, and interdependent relationships, based on common values and beliefs amongst stakeholders, that form a homogeneous structure with joint accountability for policy implementation (14, 148). In terms of key stakeholders, governments frequently represent the dominant party in a policy community, at the expense of other non-government stakeholders (135, 148). Additional literature relating to PNT suggests it is suitable for investigating the impact on network stability in sport policy processes (149–151), and has the potential to provide a mediatory function between government and non-government stakeholders (152). Although a significant methodology for the analysis of public policy-making (153, 154), however, PNT lacks a single, accepted definition (63, 154, 155). Further, PNT has evolved as a result of undue government influence in the policy process (156), a situation that neglects the views of other stakeholders and the relationships between them (135).

The PNT approach builds on several theoretical frameworks that offer explanatory power of policy networks, both on the level of strategic interaction processes as well as on the level of institutional relations. Policy networks rely on “dependency relationships that emerge between both organisations and individuals who are in frequent contact with each other in particular policy areas” (157). Accordingly, PNT assumes that stakeholders that form a policy community are reliant upon each other to combine resources to achieve objectives (14). Ultimately, this situation benefits each stakeholder in reaching their goals based on mutual understanding (158). Within such policy communities, interaction between non-governmental stakeholders is largely egalitarian (159) and based on “stable patterns of social relations existing between interdependent actors [stakeholders], which take shape around policy problems and/or programmes” (61). There are four key elements that form the basis for PNT: (i) rational decisions, (ii) individual relations, (iii) formal analysis, and (iv) stable structures (160–162).

Understanding the nature of relationships in PNT is impacted by the way power is used, particular in relation to influencing government decisions (135). From the perspective of managing the balance of power, the interaction between stakeholders has led to numerous considerations for influencing the policy process, including the aforementioned policy communities, the concept of iron triangles—constraints based on budget, scope and schedule (163)—and focused issue networks (135). Thus, incorporating the view of power, PNT caters for the incorporating the nature of membership, stages of integration, sharing of resources, and levels of control, in the development of network types (148). Such viewpoints have methodological flaws, including what potentially amounts to overclassification, with designations for one set of circumstances being interchanged and applied to describe completely different phenomena (164). To address the inconsistency, a two-dimensional approach was developed based on dual factors: stakeholders and their interrelations (164). In terms of explaining these factors, there are two key variables: (i) composition element—the characteristics of stakeholders; and (ii) structural element—the forms of stakeholder relationships (164). Alongside the categorisations, the policy area of a network—“the territorial and functional specificities” (164)—incorporate political considerations that impact the ratification and structure of a policy network.

In sport, policy networks are often focused on altering the balance of power in order to influence the policy process by, in part, separating policy discussions from government agendas (14). Such manipulative objectives are often a result of relationship instability amongst unsuitable stakeholder numbers, resulting in the limited application of PNT in sport policy (14). An initial example relates to attempts in Australia, Canada, and the UK to analyse and elite sport policy programs based on global developments (165) where countries “interpret and adapt external policy pressures to their particular national circumstances and history” (166). Further, examination of the status and potential homogenisation of international sports policy results in policies being introduced internally, as part of a process of incorporating sport policy into national objectives (166). Other examples include the application of PNT to understand partnerships and policy development processes in the UK relating to physical education and sport programs, where various partnerships impacted policy outcomes (150). Also, PNT was applied to develop a new form of governance in UK sport to address aspects of interference in education and sport policy communities (167).

Although it is suggested that PNT can assume a mediatory function between government and key stakeholders (152), the model is criticized for its complex approach by combining multiple policies as part of the analysis. While there is the potential to apply top-down and bottom-up processes to provide a more thorough analysis, there is a risk that individual policies are not adequately assessed (158). In addition, the meso-level policy PNT approach creates a need for greater consideration of potentially complementary theoretical application (152), but introducing or merging other macro-level paradigms does not guarantee a sound analysis. Further, such investigation focuses on the associations between stakeholders as opposed to having a focus on the implementation and the consequences of policy; PNT does not have the capacity to address the steps that follow policy implementation (14). There are other issues that relate to PNT that are sometimes characterized by the diverse range of stakeholder interests (148) and on occasion viewed as volatile, unstable, and associated with policy consultation rather than the development of policy (148, 162). Although the categorisation of networks emanating from PNT is often helpful when evaluating various types of stakeholder nexuses at different levels, such methods usually represent radical stakeholder connections, and the practical application does not allow for a comprehensive approach (168).

Emerging in the 1980s, the ACF has featured in studies of public policy in many developed nations—both Western and Eastern (68, 169–171)—and increasingly so in developing countries (172, 173). The ACF is a contemporary meso-level framework in the policy realm, intended to expand policy analysis by incorporating a greater range of processes than other frameworks (4, 50, 52, 174–178). The framework provides a method to examine the behavior and attitudes of numerous policy stakeholders by clustering them into advocacy coalitions, formed by a range of stakeholders with common values, beliefs, and perceptions of policy issues, and who establish levels of coordination over time (179).

As an instrument of collective public action designed to address a public problem (180), the ACF is viewed as holistic, with a broad, multi-level focus on the policy process (50, 52) and has been applied in various domains and in numerous regions globally (176, 181, 182). The influence of the ACF on public policy development, implementation and analysis has been significant and has changed and shaped thinking about public policy (183, 184) based on the examination of policy change, policy learning, and the structure and stability of coalitions (147, 172, 184, 185). According to its founder, Sabatier (186) posits that the ACF has at least four basic premises: (i) policy change can take ten years or more; (ii) policy change occurs through subsystems involving the interaction of stakeholders seeking to influence government; (iii) the subsystems involve all levels of government; (iv) policies are conceptualized in conjunction with beliefs (186).

The format of the ACF is based on three key foundations, as follows: (i) a macro-level understanding that policy-making is based on the stability of behavior of key stakeholders in a policy subsystem is; (ii) a meso-level assumption that analysing stakeholder behavior is best achieved by clustering them into advocacy coalitions; and (iii) a micro-level approach to the beliefs of each stakeholder in a policy subsystem (50). The make-up of an advocacy coalition relies principally upon the alignment of beliefs which in turn generate the enthusiasm necessary to stir stakeholders into action—such stakeholders often engage in political activity to translate beliefs into action (184)—and it is the fundamental beliefs of implementing stakeholders involved in policy that lead to understanding the policy implementation process. The ACF encourages a fusion of top-down and bottom-up elements to better understand key aspects of the policy process (187–190). The top-down perspective is based on the influence of dominant parties in policy creation process and the fulfilment by those stakeholders responsible for implementing policy. The bottom-up approach argues that a sound process relies upon effective policy implementation; thus, the policy process should commence at the bottom and influence policy creation from that position (191).

Conventionally, a top-down approach to the policy process is based on a centralist model of policy creation and suggests that the likelihood of success at the implementation stage is based on the ultimate compliance of those implementing policy (192–194). In opposition to the top-down method, bottom-up theory focusses on the perceptions of grassroots stakeholders, or “street-level bureaucrats” (195, 196), the logic being that stakeholders responsible for implementation are also key to policy outcomes. Further, the bottom-up approach is based on the skills and abilities of stakeholders to arrive at a consensus, based on common values and beliefs, to constructively impact the environment and the associated implementation process. Supporters of bottom-up theory argue that it is a more realistic and progressive approach due to the focus on the perspectives and subsequent actions of street-level implementers (191). The decisions and routines of stakeholders responsible for policy implementation are determined by the burden of said implementation, a situation that is more challenging with a solely top-down policy process. Such conditions result in part from the paradox of the grassroots, street-level implementer: while the actions of policy implementers need to be orchestrated to realize policy aims based on the nature of policy creation as part of the political process, the effort of implementers also requires improvisation and responsiveness to individual circumstances (74).

In terms of examples in sport, the ACF has been operationalized by several researchers (175, 177, 197, 198) with numerous specific attempts to evaluate sport policy through the ACF (16, 199, 200). Significant research has been dedicated to the varying levels of commitment from the government to, for example, elite sport vs. grassroots sport (201). In their analysis of Canadian and UK examples, Green and Houlihan (177) found the ACF useful in understanding the process of change and policy making in elite sport. Further, the analysis of policy subsystems highlighted complex dealings between the respective national governments and sport organisations seeking to advance systems for elite sport, and it was noted that stakeholder beliefs were crucial in the process of change occurring over many years. The element of time, as per Sabatier (186), was noted by Houlihan and Green (99) where the changing status of sport policy in schools in the UK was analysed. In the European Union (EU), Parrish (202) assessed aspects of law relating to sport and identified two advocacy coalitions; the Single Market coalition, the dominant sport policy subsystem in the EU that, based on clear capitalist foundations, supported the commercialisation of sport, along with a socio-cultural coalition responsible for influencing sport policy generally (202). More recently, Fahlén and Skille (175) applied the ACF to assess policy processes impacting Sami sport, in Norway, and Yilmaz (198) examined the beliefs of coalition stakeholders in the EU and issues regarding aspects of integrity relating to agents in sport.

Despite the above examples, there are numerous challenges associated with the ACF in policy analysis. Firstly, the framework lacks sound consideration of the influence of institutions in the policy process (16), a potentially substantial oversight when considering institutional influence on stakeholders (203). Within that context, belief systems are considered more important than institutional affiliations (62), giving rise to a problematic view that belief systems do not guarantee a robust foundation for interaction within an advocacy coalition (14). Although a coalition may engage in activities in a synchronized fashion to achieve a common goal, they may not be not “jointly agreed upon” (204), such as monitoring other stakeholders' resources or strategies and shifting their behavior accordingly (205).

With issues linked to harmonising stakeholder beliefs within an advocacy coalition, it is difficult to identify a pathway for a coordinated approach to policy implementation at ground level. For example, beliefs stemming from grassroots sport issues would likely conflict with elite sport policy beliefs, thus impacting the potential for policy stability as part of a coalition structure. The ACF implies that stability relies upon “dominant coalitions and the persistence of deep core and policy core beliefs” (4). Such a platform, however, is critiqued for its hierarchical focus on the stability of belief systems in a coalition and the perception of consistency across all stakeholders. Schlager (206) argued that not all stakeholders in a coalition will have common beliefs; rather, it is potentially more realistic to assume that stakeholders may be able to coalesce, they can equally have divergent opinions. Further, the ACF's description of policy change relies upon a multifaceted interrelationship between cogent norms, external events, and policy learning, but fails to adequately account for the concept of power impacting coalition structure and relations (4).

Finally, while the ACF theoretically caters for a comprehensive analysis of policy-making, it does not properly outline multi-tier influence, viz., the impact of local, regional, and national involvement, particularly of government, on the policy-making process (4). Overall, there are issues with the ACF cultivating a comprehensive approach to the entire policy-making process and due to its meso-level application (50, 52, 207), policy input at some levels or from some stakeholders can be limited (4); thus, for example, clubs involved with grassroots sport—the stakeholders responsible for policy implementation—can be omitted from a potentially useful opportunity to influence the policy process (16).

Although there are numerous considerations, variables, and nuances when synthesizing the findings of this review, the scope of this paper is not to provide an exhaustive outline, rather to point to the overall influence of mainstream meso-level policy frameworks on the sport policy process. Accordingly, to provide contrast on key points, the review is broken down into a comparison between meso-level frameworks, as per Table 2.