A twenty-five year old challenge and perspective: The study of the policy change in Advocacy Coalition Framework

CAPELLA, Ana Cláudia Niedhardt SOARES, Alessandra Guimarães ALVES, Renan do Prado A twenty-five year old challenge and perspective: The study of the policy change in Advocacy Coalition Framework Panel J Networks, complexity and innovation J105 Institutional change theory as a theoretical framework to study public management XIX IRSPM Birminghan 2015

Abstract The study of the policy process involves a mix of complex elements: actors, preferences, interests, perceptions, strategies, institutional design, specificities of the policy area, the amount of projects that constitute a policy, long periods of time to better understand the governmental action, among others (Sabatier, 1999). According to Thomas Dye (2008) these analytical models help the researcher to make an abstraction of the real world in an attempt to simplify the representation of political life, in order to provide better understanding of public policies. In the past decades, some efforts were make to build comprehensive frameworks, which could provide better explanations of public policy process (JOHN, 2013). This article aims to analyze one of these models: the "Advocacy Coalition Framework (ACF). The ACF was presented in its original version in 1988 and since then it has been improved by its authors based on the results of empirical applications of the model to several case studies worldwide (Sabatier and Smith, 1993). Our papers seek to understand how ACF can be used as helpful device in explaining policy change, specifically by mapping the extensions and revisions of the framework. Key-Words Policy change; Advocacy Coalition Framework; Public Policy 2

Introduction Analyzing public policy is a challenge for political science and public administration scholars and requires appropriate methodological support for the research has value and relevance, both in academic sense and as in political sense. To address this demand, new theoretical models have been developed in the field of public policy in order to explain the complex process of policymaking. The first theoretical explanations of public policy studies begin in the 1950s, based on rational thought. Among the theorists who thought public policy as something quantifiable and qualifiable, it should be noted the work of Harold Lasswell (1951) as one of the pioneers. The author considers that the main objective of the policy sciences is to "increase rationality in decision flow" (Lasswell, 1951 apud Andrew, 2005, p. 18). Lasswell proposes to study and analyze the policy process by dividing it into seven phases: information, promotion, prescription, invocation, application, completion and evaluation. The rational perspective is questioned by Herbert Simon (1957) and criticized by Charles Lindblom (1959). In his writings, Simon introduces the concept of bounded rationality, which limits the policymaker s process of decision. Criticizing the emphasis of rational decision, Lindblom shows that there were other important variables challenging the analyst of public policies, a process far more complex and organic than the rational perspective assumed. Another theorist who contributes to this discussion is David Easton (1965). His greatest contribution comes from the application of systems theory concepts to policymaking analysis. Easton explains public policies in terms of a relationship between the formulation (inputs) and results (outputs). The systemic perspective endorses another very popular explanation of the policy process: the policy cycle, which currently is divided into five phases: agenda-setting, formulation, decision making, implementation and evaluation. The policy cycle perspective has often been useful to allow the researcher to isolate each stage and focus in the one or another step of policymaking process (Pressman and Wildavsky, 1973; Hogwood and Gunn, 1984; Howlett and Ramesh, 1993). With the advance of studies in the field of public policy, new explanations begin to emerge in the early 1980s and 1990s. In this scenario, three theoretical models have expanded the comprehension in the field: the multiple streams model, proposed by John 3

Kingdon (1984); the advocacy coalition framework developed by Paul Sabatier and Jenkins-Smith (1988); and the punctuated equilibrium theory, proposed by Frank Baumgartner and Bryan Jones (1993). These theoretical constructs are currently the most commonly used because they offer a more organic approach, and enable a better understanding of the complex field of public policies (Capella et. all., 2014, p. 04). The Multiple Streams Model, proposed by John Kingdon (1984), focuses on explaining the agenda-setting process. The author states that changes in agenda status are a result of interaction of three elements (or streams): problems, policies and politics. Under certain circumstances and with the participation of an important actor a policy entrepreneur - the streams converge and open a window of opportunity. In this moment, problems are defined and recognized by the government; solutions are technically and politically viable; and the political context is favorable to the policy changes. Punctuated Equilibrium Model, proposed by Baumgartner and Jones (1993), "allows us to understand why a political system can be understood both incrementally, that is, maintaining the status quo, and also changing more radically, causing rapidly changes in public policy" (Souza, 2003, p. 33). Just as Multiple Streams, Punctuated Equilibrium is a theoretical model designed to analyze agenda-setting process. The authors seek to offer, within the same theoretical model, explanations for both processes of stability and change in public policies. For them, policies are largely based on incrementalism and, as such, have long periods of stability. But these moments are punctuated with rapidly moments of intense changes. These changes take place when policy monopolies disrupt, allowing different ideas and understandings about policies, as well new actors and institutional rules. After the moment of change, the system moves into a new point of stability, and incrementalism dominates the logic of policymaking again. At the end of the 1980s, the Advocacy Coalition Framework (ACF) was presented as an integrated explanation of the policy process. In its original version the ACF made important contributions to the literature on the complex process of public policy, among them a need to take longer-term time perspectives to understand policy change; a need for a more complex view of subsystems to include both researchers and intergovernmental relations; a need for more attention to the role of science and policy analysis in public policy; and a need 4

for a more realistic model of the individual rooted more deeply in psychology rather than microeconomics. (Weible et al, 2011:349) The model also helps to answer many questions about the policymaking process: how a wide range of actors interact in the political subsystem? What features such actors use to transform their beliefs and ideas into policy? Is there some kind of learning between allies and adversaries in the political subsystem? What is the role of scientific and technical knowledge in policymaking? What factors can influence policy changes and explain whether they are large or incremental? (Weible et al, 2011). According Weible et al (2009), ACF was designed as a framework capable of simplifying the complex public policy process. The goal is to guide the research agendas, allowing the exchange of information among scholars and professionals and also assist in developing effective strategies for decision making. For this, the framework provides a structure, which remains the same since its inception: policy subsystem as the main unit of analysis; long-term perspective for understanding the subsystem; aggregation of the actors in coalitions and the design of the policy as translations of a coalition specific belief. These are assumptions that allow scholars to understand the political process, in particular the formation and maintenance of coalitions and policy change. In fact, one of the assumptions of this model is to analyze public policies in dynamic environments in order to explain the process of change. Working within the subsystems, policy actors relate with others in a long-term relationship in which are guided by a system of beliefs and priorities, seeking to translate it into policies (Sabatier, 1988). Revisions and extensions were made in the model and it allowed the authors to provide elements that contribute to the understanding of a crucial issue in policy analysis: how to explain the process by which a policy changes? According to Mahoney and Thelen (2010), when it comes to explaining the change, often the attention is driven to the "critical juncture", generally understood as moments that present opportunities to historical agents alter the institutional trajectory. One cannot deny that these episodes of institutional "turbulence" are fundamental to explain institutional change over time. However, this kind of explanation is useful to the most common changes and there are others - as important as that, and occurring more frequently in the incremental way - that can lead to significant changes in policy. Another criticism of Mahoney and Thelen (2010) is that the endogenous sources of change are forgotten, because explanations of institutional change involves almost always the "collapse" of a set of policies and their 5

replacement by another one. In the light of these considerations, our papers seeks to understand how ACF can be used as a helpful device in explaining policy change, specifically by mapping the extensions and revisions of the framework. The Advocacy Coalition Framework (ACF) Before making a broad description of the model, we need to take a closer look at definitions in order to clarify discussions on the framework and its theoretical scope. Hence, we use the definition of "coalitions" given by Sabatier (1988: 139) in his early studies, in which he states that coalitions are: a set of people from a variety of positions (elected and agency officials, interest group leaders, researchers) who share a particular belief system... and who show a non-trivial degree of coordinated activity over time. Conceived by Paul Sabatier in the late 1980s (and counting with collaborations during the following decades) the ACF is an analytical model that seeks to avoid the popular systematization of the public policy cycle. In this sense, it differs from both the Punctuated Equilibrium Theory and the Multiple Streams Model, as these, even bringing a more dynamic approach to policy analysis, focus on a specific phase of the cycle: the agenda-setting. This idiosyncrasy of the model has its origins in the militancy of the author against systemic approach. Although the author refute the idea of the public policy cycle, this model emerged from the Sabatier early experience with studies focusing on the implementation literature 1 and from his interest and Jenkins-Smith in understanding the role that technical information play in the process policy (Sabatier, 1986; Smith, 1990; Sabatier and Smith, 1988). Thus, in addition to creating an theoretical alternative to the policy analysis based on stages and formulate a analytical model based on a holistic approach, the main debates about the top-down and bottom-up approaches are present in the ACF. Following this understanding, the ACF model encompasses all phases of the policy cycle and focuses on policy changes. Sabatier (1988), using studies of Heclo (1974), says that political change occurs according to two variables: (i) changes in the social scale, economic and political and (ii) the strategic interaction of people within a policy community, a situation in which we find the occurrence of power and efforts to develop more ways to address public policy problems. That means that the author 6

analyzes the conversion of coalitions formed by the actors involved in the process with community policies paying attention to the analysis of socio-economic and political changes that occur in these conditions. The ACF was presented in its original version in 1988 and since then has been improved by the authors by analyzing the empirical application of the model in several case studies and also critical considerations made by researchers that have used the theoretical framework (Sabatier and Smith, 1999). The concern of the authors in mapping studies that made use of the model in different parts of the world and in various policy sectors allowed not only the improvement of the model over time, but also an empirical basis that can provide information for scholars from various fields, sectors and different political contexts analyze the public policy process (Kübler, 2001; Nohrstedt, 2009; Albright, 2011; Ingold, 2011; Oliveira, 2011). In accordance with Sabatier and Weible (2007), the ACF is grounded on the following basis: (1) macro level, where most of the policies occurs among specialists within a policy subsystem, but their behavior is affected by larger socioeconomic factors and the political system; (2) meso level, where the solution to deal with the multiplicity of actors in a subsystem is to aggregate them in advocacy coalitions; (3) at the micro level (individual model) in which the conviction comes from the social psychology. These basic assumptions affect the dependent variables of the model, beliefs and political changes through a process of guided learning (policy-oriented learning) and by external disturbances. In the latest revision of the model these authors added two more features to explain political change: internal shock and the negotiated agreement. These factors, along with the original assumptions of the framework, help to explain how constraints and opportunities of subsystems affect actors over time and help to explain policy change. As already specified, the main ACF unit of analysis is the subsystem. This is because political systems involves many policy subjects and a large geographic area, which hinders the analysis of the political process, especially process of change. Subsystems operate in a broader political environment that is defined "by relatively stable parameters and external events, and constrained by long-term coalition opportunity structures, short-term constraints and resources of subsystem actors, and other policy subsystem events" (Weible et al, 2009: 123). 7

According to the authors, public policies are structured into subsystems consisting of a set of individual or collective actors, public and / or private organizations that deal with a certain policy area or problem and negotiate with each other, trying to influence the decisions taken (Sabatier, 1993; Sabatier; Jenkins-Smith, 1999). Within each subsystem we can find from two to five coalitions, formed by the most diverse positions actors (civil servants, entrepreneurs, elected representatives, researchers, civil society, etc.), who share beliefs, values, ideas, similar political objectives and demonstrate some kind coordinated activity over time. The ACF assumes that participants have strong political beliefs and seek effective means to turn their beliefs into actual policies. In this sense, the technical and scientific information play an important role in modifying beliefs of the political participants as well as researchers (university scientists, policy analysts, consultants, etc.), are seen as key players in the policy process (Zafonte; Sabatier 2004; Weible, 2005). The ACF also assumes that the beliefs of political actors tend to remain stable over time. Thus major policy changes are difficult to happen (Sabatier; Jenkins-Smith, 1999). To map these beliefs, which are the "glue" that holds the actors together in coalitions, the authors created a tripartite belief system structure, considering the degree of resistance to change. This approach seeks to facilitate the analysis of the influence of the actors, over time, in the policy. They are: i.deep core beliefs consists on beliefs deeply rooted in policy subsystems. There are general normative assumptions and ontological issues about human nature, the relative priority of fundamental values such as freedom and equality, etc. In other words, these are the deep beliefs, the individual axioms. ii.policy core beliefs - they are the fundamental policy beliefs. Generally speaking, they are in the center of the entire subsystem and consist in the most fundamental level to coalitions, as they set the alliances and coordinate activities among the members of the subsystem; iii.the instrumental aspects (secondary aspects) consists on secondary beliefs. Here we find instrumental deliberation and technical data required to implement the policy core, that is, the instrumental aspects relating to the pragmatical policymaking. They are practical questions that cause only minor effects on policy (Pedroso, 2011). 8

The deep core beliefs emerge from childhood socialization and therefore are very difficult to modify. The policy core beliefs are resistant to change, but they are more malleable than the deep core beliefs. As for the instrumental aspects, these are regarded as the belief system component most likely to change over time as a result of new information (that arise in the subsystem) and learning processes. It is in this last level that are taken measures and decisions necessary to implement the policy core; is also at this level that are used strategies to achieve the values of deep core within the subsystem (Sabatier; Jenkins Smith, 1999). Sabatier and Weible (2007) believe that the operation of two or three of these beliefs is sufficient to identify at least two advocacy coalitions. However, they warn about the importance of operationalizing the largest possible number of beliefs to map existing coalitions and to identify the subdivisions within the coalitions that can generate subcoalitions (that are bounded together by certain fundamental policy beliefs, but are divided internally in other core policy beliefs or secondary beliefs). The ACF also presents a set of hypotheses that, over the years and with the revision of the model, have been expanded and are extremely important for understanding this theory. Currently there are fifteen cases that fall into three categories: (i) advocacy coalitions; (Ii) political change and (iii) learning-oriented policies (Sabatier, 2007; Weible et al, 2009). Below we highlight the assumptions related to advocacy coalitions. ADVOCACY COALITION HYPOTHESIS On major controversies within the policy subsystem when policy core beliefs are in dispute, the lineup of allies and opponents tends to be rather stable over periods of a decade or so. Actors within an advocacy coalition will show substantial consensus on issues pertaining to the policy core, although less so on secondary aspects. An actor (or coalition) will give up secondary aspects of his (its) belief system before acknowledging weaknesses in the policy core. Elites of purpose groups are more constrained in their expression of beliefs and policy positions than elites from materials groups. Within the coalition, administrative agencies will advocate usually more moderate positions than their interest-group allies. 9

Coalitions are more likely to persist if (i) the major beneficiaries of the benefits that a coalition produces are clearly identified and are members of the coalition, (ii) the benefits received by coalition members are related to the maintenance costs of each member, and (iii) coalition members monitor each others actions to ensure compliance. Actors who share policy core beliefs are more likely to engage in short-term coordination if they view their opponents as (i) very powerful, and (ii) very likely to impose substantial costs upon them if victorious. Actors who share (policy core) beliefs are more likely to engage in short-term coordination if they: (i) interact repeatedly; (ii) experience relatively low information costs; and (iii) believe that there are policies that, while not affecting each actor in similar ways, at least treat each fairly. (p.129) Weible et al (2009) highlight the care that researchers must take to assume that there is a uniformity of beliefs and behavior among the actors of coalitions, as many members (as the sub-coalitions idea show) might disagree about some beliefs at different levels, but by no means leave their home coalitions. The idea of these authors is that scholars should focus main actors and their auxiliaries in the coalition, as this facilitates explanations for the formation of sub-coalitions, waivers of members over time and dissolution of coalitions. As most policies occur within subsystems, one of the first challenges for application of the model is the definition of its scope, since there are overlapping and embedded subsystems 2. In order to improve the understanding of its terms, Sabatier advocates the expansion of the concept of subsystems, stating that our conception of policy subsystems should be broadened from traditional notions of 'iron triangles' - limited to administrative agencies, legislative committees, and interest groups at a single level of government - to include actors at various levels of government active in polícy formulation and implementation, as well as journalists, researchers, and policy analysts who play important roles in the generation, dissemination, and evaluation of policy ideas. (Sabatier, 1988: 131)" According to the authors, the fundamental rule to identify a subsystem is to look for material (e.g., water policy), territorial (a geographic area as Rio São Francisco, in Brazil, for example), or the political-administrative level, such as federal, state and municipal levels (Sabatier; Jenkins-Smith, 1993; Zafonte; Sabatier, 1998; Sabatier; Weible, 2007). 10

Policy change in the ACF The strength of the model, however, is the provision of a clear criterion to distinguish greater policy changes (followed by changes in the policy core beliefs) and smaller ones (followed by changes in secondary beliefs) (Sabatier; Jenkins-Smith, 1999). For Sabatier and Weible (2007) the behavior of actors within the subsystem can be affected by two sets of exogenous factors: a stable and the other dynamic. Stable factors includes topics such as the basic attributes of the problem; the basic distribution of natural resources; fundamental socio-cultural values and social structure and the basic constitutional structure (rules). In turn, the dynamic factors encompass changes in socioeconomic conditions; changes in the coalition government; changes of public opinion; policy decisions impacting other subsystems. The latter are seen as more permeable to change, even in a period of a decade or more. External shocks may change agendas, focus public attention on a specific issue and attract the attention of key decision makers, for example, the Executive. However, one of the most important effects of external shocks is the process of allowing the redistribution of resources between the coalitions. This redistribution can lead to the replacement of the previously ruling coalition by a minority coalition (Sabatier and Weible, 2007). To Weible et al (2009), one of the model challenges lies in the tools available to map the interdependence between subsystems, since a large number of policies are interconnected, and not limited to a single subsystem. In other words, the political process would be better understood if the researcher could analyze the subsystem structures and their interdependencies with other subsystems. A good example is human rights policies, like policies directed for juvenile delinquents, which are close intertwined with drug policy, education, etc., which means that a decision on drug policy subsystem can affect the policies for youth offenders. In this sense, the development of ACF theory about the overlap of political subsystems has gained ground among scholars of the model. Regarding to the stable factors, they rarely provide the impetus for behavioral change or policy change within a subsystem, so they are more easily changed, even in the period of a decade. However, Sabatier and Weible (2007) understand that they are important in establishing the resources and constraints of the actors within the subsystems. On the other hand, Weible et al (2009) points out that the effects of external 11

shocks cannot be understood in isolation, one must look at the internal affairs of the subsystem. Another form of policy change is explained by oriented learning, defined by the authors as "relatively enduring alternations of thought or behavioral intentions which result from experience and which are with concern the attainment (or revision) of policy objectives" (Sabatier; Jenkins-Smith, 1999 p.123). According to the authors, the secondary beliefs are more susceptible to policy learning, as they tend to be more flexible to information than the core beliefs of the policy. Sabatier and Weible (2007) believe that the scientific information and techniques play an important role. Even if the accumulation of technical information is not capable of changing the view of the opposing coalition, it can bring important impacts on public policy by changing the views of policy brokers or other governmental actors. According to Weible et al (2009), although studies that applied the model corroborate that policy learning often occurs at the secondary level of beliefs - especially when the issues are technical or when the conflict restricted to intermediate levels, and when there is a professional forum - there are still uncertainties regarding to the factors that facilitate cross-coalition learning, and even relating learning to policy change. These authors also assert that in situations of high conflict, cross-coalition learning tends to reinforce preexisting beliefs. That is, for these authors the conditions for learning among coalitions are necessary, but not sufficient for policy change to occur. According Weible et al (2009) these observations do not diminish the importance of policy learning within the model, mainly because of its connection with the hypothesis of political change. However, these authors recognize the empirical difficulties faced by scholars to assess learning and then relate it to policy change explanation. For Sabatier and Weible (2007) external disturbances can lead to rapid changes in the structure of the subsystem and even affect fundamental beliefs of individuals. But policy learning might take a decade or more to change the secondary beliefs. Explain policy change is a significant challenge for applying the model. Below are the five hypotheses raised by the authors on learning-oriented policy (SABATIER, 2007; Weible et al, 2009). HYPOTHESIS POLICY-ORIENTED LEARNING 12

Policy-oriented learning across belief systems is most likely when there is na intermediate level of informed conflict between the two coalitions. This requires that: (i) each have the technical resources to engage in such a debate; and that (ii) the conflict be between secondary aspects of one belief system and core elements of the other or, alternatively, between important secondary aspects of the two belief systems. Policy-oriented learning across belief systems is most likely when there exists a fórum which is: (i) prestigious enough to force professionals from different coalitions to participate; and (ii) dominated by professional norms. Problems for which accepted quantitative data and theory exist are more conducive to policy-oriented learning across belief systems than those in which data and theory are generally qualitative, quite subjective, or altogether lacking. Problems involving natural systems are more conducive to policy-oriented learning across belief systems than those involving purely social or political systems because in the former, many of the critical variables are not themselves active strategists and because controlled experimentation is more feasible. Even when the accumulation of technical information does not change the views of the opposing coalition, it can have important impacts on policy at least in the short run by altering the views of policy brokers. 3. Revision of the framework: contributions to policy change explanations The revision of the ACF made by Sabatier and Weible (2007), presented two new explanatory variables for political change: internal shocks and negotiated agreements. Regarding to internal shocks (as well the external ones), the authors highlight they are critical in redistributing political resources, since they place the spotlight of the public opinion about a problem within the policy subsystem and, therefore, internal shocks have the potential to attract new resources or provoke their redistribution (eg.: public support, financial support, etc.) 3. This shift in resources may tip the power structure of the policy subsystem from one dominant advocacy coalition and one or more minority coalitions to two or more competitive advocacy coalitions or, in a complete reversal, to a different dominant advocacy coalition with more than one different minority advocacy coalitions. (Sabatier; Weible, 2007:205) The ACF recognizes the importance of internal and external shocks as significant causes for political change and also shows distinctions between them. According to the authors: 13

Internal shocks confirm policy core beliefs in the minority advocacy coalition(s) and increase doubt within the dominant coalition. Internal shocks that indicate monumental failures of the policies and behaviors of a dominant advocacy coalition also strongly affect the belief systems of policy participants. For the minority advocacy coalition members, internal shocks confirm their policy core beliefs (e.g., regarding the causes or seriousness of the problem in the policy subsystem). This galvanizes the membership of minority coalitions. For the dominant advocacy coalition, internal shocks increase doubt about their policy core beliefs and put into question the effectiveness of their policies (Sabatier; Weible, 2007: 205). In the original version of ACF, Sabatier and Jenkins-Smith (1988) brought the external shocks as a necessary cause for political change. According to them, the external shocks are beyond the control of subsystem actors, so that they are more likely to cause policy change. The internal shocks were not a concern at that time, because the authors understood that the devil shift created patterns of polarization between the coalitions that practically closed the channels of communication between members of rival coalitions. The devil shift occurs when the actors in a coalition perceive their opponents as the most powerful and evil than they really are. This happens because of the political losses within the subsystem that lead individuals to remember more of the defeats to their opponents, than eventual gains over them. That is, the devil shift stimulated a process of actors insulation within the subsystem and, without interaction, it was impossible that internal change occurred. An external shock capable of changing the vision and the perception of the coalition members about the subsystem, the policies and the members of the rival coalitions was extremely necessary. With the new revision of Sabatier and Weible model (2007), the authors have come to recognize the internal shocks as an alternative way for major changes in politics, because they have the potential to alter the distribution of power among coalitions (these authors assume that most political subsystems are dominated by a defensive coalition with one or more minority coalitions). According to the authors, external and internal shocks have the capacity to redistribute critical political resources and to attract new resources that can significantly change the power relations between the coalitions in the subsystem. However, unlike the external shocks, the internal ones cause instability between the forces of subsystem and they directly challenge public policy core beliefs. Thus, internal shocks tend to confirm the beliefs of minority coalitions, while that 14

cast doubt on the beliefs of the ruling coalitions. As domestic events occur within the subsystem, actors seek through them to highlight the flaws in current practices subsystem. The other way for change - negotiated agreements - serve as an explanatory variable in the absence of external or internal disturbances. That is, in certain critical situations (so-called hurting stalemate), the coalitions, formerly in conflict, come to a negotiated agreement that represents a substantial change in the status quo which translates into a large change. For Sabatier and Weible (2007) the basic condition for the success of the negotiations is a situation in which all coalitions involved in dispute envisage a continuation of the status quo as unacceptable, a situation of a clear political impasse. Below, we can observe that disputes can occur directly and indirectly, following some instruments that coalitions use for bargaining, thus influencing the political game. Weible et al (2009) present a number of important provisions that affect the change: the stalemate hurting, effective leadership, consensus based decision rules, diverse funding, duration of process, commitment of members, the focus on empirical issues, an emphasis on building trust, and lack of alternative venues. The model also present two hypothesis of political change (Sabatier, 2007; Weible et al, 2009): POLICY CHANGE The policy core attributes of a governmental program in a specific jurisdiction will not be significantly revised as long as the subsystem advocacy coalition that instituted the 15

program remains in power within that jurisdiction except when the change is imposed by a hierarchically superior jurisdiction. Significant perturbations external to the subsystem (e.g., changes in socioeconomic conditions, public opinion, system-wide governing coalitions, or policy outputs from other subsystems) are unnecessary, but not sufficient, cause of change in the policy core attributes of a governmental program. As we can see, the ACF in shares with policy network literature a focus on the interpersonal relationships to explain human behavior. According to Sabatier and Weible (2007), the ACF assumes that the beliefs and behavior of stakeholders are embedded within informal networks and policies are, in a way, structured by networks established between major political participants. The ACF argues that policy participants will seek allies with people who hold similar policy core beliefs among legislators, agency officials, interest group leaders, judges, researchers, and intellectuals from multiple levels of government. If they also engage in a nontrivial degree of coordination, they form na advocacy coalition. Coordination involves some degree of working together to achieve similar policy objectives. The ACF argues that advocacy coalitions provide the most useful tool for aggregating the behavior of the hundreds of organizations and individuals involved in a policy subsystem over periods of a decade or more (Sabatier; Weible, 2007, p.196). In summary the ACF model provides explanations to policy changes that occur in one or more decades, as this is the minimum time to accumulate bibliography and government documents in order to provide robust and solid analysis of public policy. Therefore the framework seeks to explain policy change through four ways: (1) policy learning; (2) external shocks (stable and dynamic parameters); (3) internal shocks and (4) negotiated agreements. Beyond these four elements of policy change, the fundamental principles of ACF are: individual model of social psychology; policy subsystem as unit of analysis; advocacy coalition as a key political actor; common beliefs as the "glue" that holds coalitions and "professional forums" to facilitate learning through coalitions. Below we present a diagram to better understand how the model may help explain major policy change (occurring at the core of policy) and smaller ones (occurring in secondary aspects). 16

Concluding remarks The ACF seeks to provide elements for understanding the complex process of public policymaking. However, as stated by Sabatier and Weible (2007), the framework is not intended to exhaust all analytical possibilities. The authors highlights there are still 17

many unanswered questions, which means that there are many opportunities to explore the application of the framework and its limitations. In the specific case of policy change, the explanations consist, even today, in a challenge for scholars seeking to use the framework as a theoretical tool in their studies. The challenges come from methodological issues that make possible, for example, connecting policy learning and policy change. Or, as the framework allows, explaining the policymaking process as a whole, not just to investigating one specific stage of the cycle. Although the ACF is a framework designed to explain the policy process, Weible et al (2009) point out that many researchers combine the heuristic stages with the model, while the authors understand the cycle as a part of the policy process. The proposal of the ACF is precisely the contrary: the framework aims to investigate the degree of coordination among the coalition members operating in different stages of the policy. Sabatier and Weible (2007) emphasize some questions that remain to be answered, such as: How the structures of political opportunity affect the beliefs of the coalition, resources, stability and strategies (Zafonte and Sabatier 2004; Kübler 2001)? After an external or internal shock, what are the causal processes that lead to political change (Nohrstedt 2005)? To what extent the ACF can be applied to global policy subsystems (Liftin 2000)? What are the instances of nonlearning (Weible et al 2011)? How to prove that the changes of beliefs are a direct result of the policy learning process? How much time is necessary in order to explain the changes in the actor s beliefs inside the coalition? Wouldn t be better to the researcher s understanding to consider belief changes by analyzing the institutions? These are some questions that still need to be answered, reflecting the challenges facing researchers who choose to work with this framework. But considering the public policy literature, where there is still a prevalence of analysis focusing in the policy cycle and few theories that address issues related to the policy change, the ACF stands for bringing methodological options to researchers and well-established guidelines to explain the complex process of policy change. Notes 1 In the book entitled Implementation and Public Policy (1983), and in partnership with Daniel Mazmanian the authors propose an implementation model that outlines five stages starting with formulation through the review and reformulation. For the analysis and defense of the model, the authors are used five different cases of implementation of public policies. 18

2 There are subsystems at local, national and even international level. The authors cite the example of housing policy, the local housing agency is part of a local subsystem housing, but overlaps the local land use and transport subsystem, which is within the federal and state subsystems housing. The issue may further complicate matters, when there are international treaties that add an international authority, which is quite limited to impose their will on national and sub-national entities (Sabatier and Weible, 2007). 3 Typology of coalition resources: formal legal authority to make policy Decisions; public opinion; information; mobilizable troops; financial resources; skillful leadership (Sabatier and Weible, 2007). References ALBRIGHT, E. A. 2011. Policy Change and Learning in Response to Extreme Flood Events in Hungary: An Advocacy Coalition Approach. Policy Studies Journal, p. 485 511. ANDREWS, C. W. "As policy sciences como 'ciência': método e reificação". Perspectiva, São Paulo, 13-37, 2005 CAPELLA, A. C.; ALESSANDRA, G. S; BRASIL. "Pesquisa em políticas públicas no Brasil: um mapeamento da aplicação de modelos internais recentes na literatura nacional ". IX Encontro da ABCP, 2014. DYE, T. Understanding Public Policy. 12th Ed. Upper Saddle River/NJ, Pearson/Prentice Hall, 2008. EASTON, D. A framework for political analysis. Publisher, Prentice-Hall, 1965. HECLO, H. (1974), Modern social politics in Britain and Sweden: from relief to income maintenance. New Haven, Yale University Press, 1974. HOGWOOD, B. W, GUNN, L. A. Policy Analysis for the Real World, Oxford University Press, 1984, Oxford HOWLETT, M., RAMESH, M. Patterns of policy instrument choice: policy styles, policy learning and the privatization experience, Review of Policy Research 12 p. 3-24, 1993 INGOLD, K. 2011. Network Structures within Policy Processes: Coalitions, Power, and Brokerage in Swiss Climate Policy. Policy Studies Journal, p. 435 59. JOHN, Peter. "New directions in public policy: theories of policy change and variation reconsidered". Paper presented at International Conference on Public Policy. Grenoble, 26-28 june, 2013. KÜBLER, Daniel. 2001. Understanding Policy Change with the Advocacy Coalition Framework: An Application to Swiss Drug Policy. Journal of European Public Policy, p. 623 41 19

LINDBLOM, C. E. The Science of Muddling Through. Public Administration Review. V. 19, pp. 79-88, 1959. MAHONEY, J. & THELEN, K. A Theory of Gradual Institutional Change. In: Mahoney, J. & Thelen, K. (Eds): Explaining Institutional Change - Ambiguity, Agency and Power. New York: Cambridge University Press, 2010, pp. 1-37. NOHRSTEDT, D. 2005. External Shocks and Policy Change: Three Mile Island and Swedish Nuclear Energy Policy. Journal of European Public Policy 12 (6): 1041 1059 OLIVEIRA, A. P. M. A Prova Brasil como política de regulação da rede pública do Distrito. Federal. Brasília, 2011. PEDROSO, M. M. Inteligência Decisória e Análise de Políticas Públicas: o caso das Unidades de Pronto Atendimento (UPAs). 2011. 338f. Tese (Doutorado em Administração) - Programa de Pós-Graduação em Administração, Universidade Federal de Brasília, Brasília. 2013 PRESSMAN, J. L.; WILDAVSKY, A.. Implementation. Berkeley: University of California Press, 1973. SABATIER, P. An advocacy coalition framework of policy change and the role of policyoriente learning therein. California, 1988 SABATIER, P. A. The Need for Better Theories. In Sabatier (ed.). Theories of the Policy Process. Oxford, Westview Press, 1999. SABATIER, P, JENKINS-SMITH H. Special Issue: Policy Change and Policy-Oriented Learing: Exploring an Advocacy Coalition Framework," Policy Sciences 21:123-278, 1988 SABATIER, P. 1986. "Top-Down and Bottom-Up Models of Policy Implementation: A Critical Analysis and Suggested Synthesis," Journal of Public Policy 6 (January):21-48. SABATIER, P. A.; WEIBLE, Christopher M. The advocacy coalition framework: innovations and clarifications. In: SABATIER, P. A. (Ed.) Theories of the policy process. Boulder, CO: Westview Press, 2007, p. 189-220. SABATIER, Paul A.; JENKINS-SMITH, Hank C. The advocacy coalition framework: an assessment. In: SABATIER, P. (Ed.). Theories of the policy process. Boulder, CO: Westview Press, 1999. SABATIER, Paul A. Theories of the Policy Process. 2 ed. Boulder, CO: Westview Press, 2007. p. 189-220. SABATIER, P. A., and ZAFONTE, M,. 2004. Are Bureaucrats and Scientists Members of Advocacy Coalitions? Evidence from an Intergovernmental Water Policy Subsystem. 20

In P. A. Sabatier, ed., An Advocacy Coalition Lens on Environmental Policy,under review. SABATIER, P. A. & JENKINS-SMITH, H. C., (ed.). Policy Change and Learning. An Advocacy Coalition Approach. Boulder, San Francisco, Oxford: Westview Press. 1993. SIMON, H. Comportamento Administrativo. Rio de Janeiro: USAID. 1957. SOUZA, C.. Estado de campo da pesquisa em políticas públicas no Brasil. Revista Brasileira de Ciências Sociais, São Paulo, v. 18, n. 51, fev. 2003. WEIBLE, C. M, 2005. Beliefs and perceived influence in a natural resource conflict: An advocacy coalition framework approach to policy networks. Political Research Quarterly, 58:461 475. WEIBLE C. M et. al. Themes and Variations: taking stock of the advocacy coalition framework. The Policy Studies Journal, vol. 37, nº 01, 2009. WEIBLE C. M et.al. A Quarter century of the Advocacy Coalition Framework: an introduction to the special issue. The Policy Studies Journal, vol. 39, nº 03, 2011. 21