The importance of ideas in institutional change and political processes

A importância das ideias na mudança institucional e nos processos políticos

La importancia de las ideas en el cambio institucional y los procesos políticos

L'importance des idées dans les changements institutionnels et les processus politiques

This article aims to show the crucial role that ideas play in institutional change, the formation of public policies, and actors’ grouping and orientation. These scenarios involve processes of creation of ideas and discourses, of disputes over the solution to problems and the legitimation or not of decisions. The role of ideas depends on our conception of their relationships with institutions, interests, and actors, and of the determinants of the changes we appreciate. The assumption promoted by rational choice that they are instrumental means of interests limits the understanding of the complexity of political processes. To vindicate the role of ideas, the article analyzes: various schools and authors based on the conceptual tension between interest and institutions; the relationship between ideas, political changes, and coalitions; and the role of discourse concerning them.

ideas; discourse; institutions; social learning; coalitions

Resumo

O objetivo do artigo é mostrar o papel crucial que as ideias desempenham na mudança institucional, na formação de políticas públicas, e no agrupamento e orientação dos atores. Todos estes cenários envolvem processos de criação de ideias e discursos, de disputas sobre a solução de problemas e a legitimação ou não de decisões. O papel das ideias depende da nossa concepção da sua relação com as instituições, interesses e atores. E dos determinantes que apreciamos nas mudanças. O pressuposto promovido pela escolha racional de que são um meio instrumental de interesses limita a compreensão da complexidade dos processos políticos. Para justificar o seu papel, o artigo analisa várias escolas e autores em termos da tensão conceitual entre interesse e instituições; a relação entre ideias, mudança política e coligações; e o papel do discurso em relação a elas.

ideias; discurso; instituições; coligações; aprendizagem social

Resumen

El objetivo del artículo es mostrar el papel crucial que las ideas juegan en el cambio institucional, en la formación de políticas públicas, la agrupación y orientación de actores. Todos esos escenarios implican procesos de creación de ideas y discursos, de disputas por la solución a problemas y la legitimación o no de las decisiones. El papel de las ideas depende de la concepción que tenemos sobre sus relaciones con las instituciones, los intereses y los actores. Y de los determinantes que apreciamos en los cambios. El supuesto promovido por el rational choice de que son un medio instrumental de los intereses limita el entendimiento de la complejidad de los procesos políticos. Para revindicar su papel, el trabajo analiza varias escuelas y autores en función de la tensión conceptual entre interés e instituciones; la relación entre ideas, cambios políticos y coaliciones; y el papel del discurso con relación a ellas.

ideas; discurso; instituciones; coaliciones; aprendizaje social

Résumé

L'objectif de cet article est de montrer le rôle crucial que jouent les idées dans le changement institutionnel, dans la formation des politiques publiques, ainsi que dans le regroupement et l'orientation des acteurs. Tous ces scénarios impliquent des processus de création d'idées et de discours, des conflits sur la solution des problèmes et la légitimation ou non des décisions. Le rôle des idées dépend de notre conception de leur relation avec les institutions, les intérêts et les acteurs. Et sur les déterminants que nous apprécions dans les changements. L'hypothèse promue par le choix rationnel selon laquelle ils sont un moyen instrumental d'intérêts limite la compréhension de la complexité des processus politiques. Pour justifier leur rôle, l'article analyse diverses écoles et auteurs en termes de tension conceptuelle entre intérêt et institutions, de relation entre les idées, le changement politique et les coalitions, et de rôle du discours par rapport à ceux-ci.

idées; discours; institutions; coalitions; apprentissage social

Introduction

This article is about the role of ideas in political processes and institutional change. Sikkink (1991) believed that it was a paradox that academics, engaged in the production of ideas, gave them so little importance. This relative indifference has been grounded in strong and widespread theoretical traditions. A meeting point between rational choice (RC) theory and classical structural conceptions is the marginal place they have accorded to ideas in social analysis, political processes, and institutional change. Both schools embrace a “strategic realism” that underestimates the complex link between actors, interests, and ideas and the fluid dynamics to which change processes give rise (Campbell and Pedersen, 2001; Birkland, 2016).

For decades, an increasingly refined field of research on the importance of ideas has developed under multiple perspectives of constructivist affiliation (Hay, 2006). They agree that rationality or structural imperatives do not sufficiently define how interests are translated into policies, authoritative discourses, and institutional changes. In dialogue with what they call International Political Economy (IPE), Abdelal, Blyth, and Parsons (2015) make this point. Rationalist and materialist models view incentives derived from economics as informing decision-making and determining governance decisions. They assume that political action varies “not with different interpretations of the world” (pos. 99), but with the mere presence of the resources and power shares that actors hold in different settings (pos. 137). They profess that, under increasing analytical sophistication, they “objectively” express “reality”.

Their assumptions form a vast “gap” that the systematic incorporation of constructivism can address. The central insight of constructivism – as Wendt (1999) argues – is that collectively held ideas shape the social, economic, and political world in which we live. The meaning of the social world, of its objects – how collective ideas generate different interpretations and solutions in the face of the environment or contexts of uncertainty – changes and limits the postulate of a fixed correspondence between actors and material structures. This dislocation implies a serious postulation of the weight of ideas.

Abdelal, Blyth, and Parsons (2015) consider that constructivism is aligned on four axes of research. In the construction of meaning (pos. 237-70) the processes of interpretation are central; they make it possible to orient oneself by narrowing the range of choices in open events and vary according to identities, norms, and conventions that value actions and authorities. As the authors put it, “A strong version of meaning-oriented constructivism holds that societies and policymakers rarely, if ever, interpret the world around them in purely material terms. Rather, they endow the economies in which they are embedded with social purposes” (pos. 249). Second, the cognition axis emphasizes the relationship between “shortcuts” and information. Action is a social construct not because agents need meaning to guide it, but because they depend cognitively on stabilizing frames to organize and analyze information. Shortcuts filter information in a non-neutral way.

Those who mainly introduce the study of ideas in institutional change (Blyth, 2002; Kristensen and Zeitlin, 2005; Hay, 2006; Schmidt, 2008) place themselves on the axis of uncertainty. The strong version of the axis distinguishes between a situation of risk where interests can be achieved under specific probabilities and uncertainty that does not admit that possibility because the outcome is unique and radically unpredictable (pos. 306). Postmodern or poststructuralist perspectives are placed on the axis of subjectivity. Like the other axes, they assert that interests are social constructs, but discourse and identity are central elements of the social context in which they are realized. The position of subjects in a discursive field defines their identity and interest. For them, norms not only regulate behaviors, but are expressions of power that exclude or include specific actions and beliefs (pos. 344).

According to several scholars, Abdelal, Blyth, and Parsons (2015) consider that constructivism is only a parameter of analysis. Furthermore, it does not represent a theory of everything or a fundamental epistemological presupposition. In this framework, the authors emphasize uncertainty as a key concept and diminish the potential of discourse. Due to its characteristics, discourse delimits identities, facilitates interpretations, sustains frames and narratives, and guides the ordering of information (Fischer, 2003; Hay, 2006; Schmidt, 2010). In other words, it addresses many of the central points described in the four research axes mentioned above. It supports ideas because, as Risse-Kappen (1994) has said, they do not float in the air. Fischer (2003, 2012) and Schmidt (2012) have presented their perspectives under a position that they generically call the return of ideas and discourse.

Schmidt (2008, p. 317-21) has developed discursive institutionalism (DI), which postulates overcoming the limitations of the other neo-institutionalism: rationalist (IR), historical (HI), and sociological (SI). While the latter agree that institutions modulate the formation of actors' expectations and strategies, discursive institutionalism distances itself for several reasons: it rejects the exogenous conception of change (IR; IH); it views interests as a social construction rather than as an instrumental-objective dimension (IR, IH); it understands uncertainty not only as an information challenge (IR) but above all as a cognitive problem; it considers norms as dynamic social constructions and not as static structures that weigh on agents (IS).

As members of this “fourth neo-institutionalism” (Schmidt, 2010, p. 2), a wide body of perspectives are considered: “ideational turn” (Blyth, 2002), discursive institutionalism (Campbell and Pedersen, 2001), ideational institutionalism, and constructivist institutionalism (Hay, 2006). All of them coincide with the critique of neo-institutionalism and exercise a variety of methods. This group takes ideas and discourse seriously within the processes of change; recognizes neo-institutionalism as background information; places discourse and ideas in contexts of meaning while associating them with “logics of communication”; and commits to a markedly dynamic vision of institutional change. Because of the variety of topics it brings together, this field is today one of the most fertile in analyzing ideas.

Notwithstanding this conjunction, Schmidt's DI professes her distinction by focusing less on the substantive content of the ideas and much more on the interactive processes to which the latter gives rise in political processes (2008, 2011). For her discourse is not only “text (what is said) but also context (where, when, how, and why it was said). The term refers not only to structure (what is said or where and how) but also to agency (who said what to whom)” (2008, p. 3). The interactive view entails the central question of what determines the success of ideas in political processes and institutional change. The forms of coalition of agents, the character of ideas, and social learning are at the core of that discussion (Hall, 1993; Fischer, 2003; Radaelli, 2008; Béland and Cox, 2016; Weible and Sabatier, 2018).

The analysis of ideas has identified many problems; we have outlined just a few. There is no single route for describing and reflecting on this body of problems, nor is there a perspective that enjoys all adherents. Every route passes through some places and avoids others: we are interested in the relationship between interests, actors, and institutions and, therefore, in the discussion between the role of ideas and neo-institutionalism. In this framework, ours is closer to the interactive view of discourse and, therefore, without reducing the interpretative and even normative weight of ideas, we also assume that they are a source of power and are linked to various stakeholders. We aim to show the vital role that ideas play in orienting actors and political processes or institutional changes. We do this through the analysis of schools or authors who have dealt with the subject and from various reflection points. It privileges scholars who are a traditional reference on the subject but recovers other more contemporary ones.

In the first part, we discuss why a non-exogenous and more dynamic conception of change brings to light the criticisms that are made to neo-institutionalism from a vindication of ideas. In the second, we reflect on the link between interests and ideas, debunking the assumption that these are mere reflections of the former, and we analyze their essential role in the orientation of actors in contexts of profound crisis. In the third section, the complex relationship between ideas, discourse, and large-scale or paradigmatic political changes and the associated issue of so-called social learning are analyzed. The formation of coalitions of actors around knowledge and ideas is also introduced. The last section highlights the effect of ideas, discourse on public policy, and the problem of how discursive interactions are configured; it also addresses a question that lies behind this dynamic: why do some ideas succeed while others do not? What is the nature of the coalitions that are formed for this success?

Institutions facing change: the need for ideas

The relationship of institutions with ideas constitutes a field of reflection, recognition, and critique of neo-institutionalism in its different versions. The discussion has as its background the formation of preferences, institutional practices, and the character of their change. Suppose we postulate that institutions are the plain expression of the interests of actors who only maximize. In that case, we assume that there is a remarkable correspondence between them, their motivations, and the institutional frameworks. Where, then, does the need for change arise? What explains the formation of new institutions and the collective action that accompanies it? Under this correspondence, institutional changes can only be explained exogenously. Several problems remain unresolved in the response: how exogenous variables become new interests; how actors acquire clarity about them in the change; and whether these precede the formation of institutions or if because of the lack of information, the process is reversed.

This theoretical gap has repeatedly pointed to the rational choice (RC) since it is not possible to address the link between interests and the formation of institutions without specifying preferences, without defining routes of action and, therefore, without a body of ideas. For this reason, King (1973, p. 294) has suggested that, to explain the variability in the routes of change and public policy, attention should be paid not only to the interaction between interest groups or elites, but also to the institutional processes and ideas. The implications of the gap are complex, and neo-institutionalism has tried to find some answers by incorporating endogenous variables and accentuating institutional weight to make room for ideas (Hall and Taylor, 1996). Because of its conceptual nature, it is rational neo-institutionalism (RN) that has shown the greatest difficulty in this subject.

It is a common mistake to emphasize only the continuity between RC and RN without specifying their differences. Regardless of the context, for RC, individuals or actors are involved in strategic games whose rationality, oriented towards maximizing one's interest, makes cooperation unlikely as they promote free-rider logics (Hardin, 2006, p. 6). In following their strategies, their relationship with institutions is instrumental, and the norms and ideas are marginal. By relying on a disorderly flow of strategies per actor, social action appears as unstructured, marked by limited information, and highly contingent. How an actor can be rational in that context is a pertinent question. Rational neo-institutionalism sees the analytical need to give structure to social interaction, and by doing so, gives conceptual priority to institutions over interests (Goldstein, 1993; North, 1993; Ostrom, 2011). As Blyth (2002, p, pos. 542) has indicated, the importance of institutions allows us to find a “mechanism to explain the apparent anomaly of stability” that derives from the disordered and individual action of RC.

For North (1993, p. 13), institutions are rules that constrain actors to shape human interaction and form structured incentives so that exchanges of any kind can take place. They reduce uncertainty and are “guides” for interaction to take place expectedly. For Ostrom (2011, p. 3-13), they are regulations that shape structured situations of action and determine positions and decision ranges of actors and, consequently, the possible outcomes. For both, the institutions affect the eligible strategies and modulate the field of interests that can be socially processed. The structuring effect of the institutions allows us to understand stability but leaves intact the exogenous explanation of change and does not solve the problem of variation of preferences and institutions. An endogenous perspective of change is needed, and with it, “ideas become the focal point... for explaining institutional supply and stability” (Blyth, 2002, pos. 543).

To solve the endogenous dimension within the RN, Goldstein and Keohane (1993) have assigned a relative place to ideas. They matter only when the objective variables do not explain the behavior; when the routes of action available to the actor do not present clear benefits as they help “to choose a reasonable rational plan of action” (Fischer, 2003, pos. 35), and act as “road maps” in conditions of uncertainty to form institutions. Along these lines, North (1993) has considered them “mental models of behavior” in the task of replacing inadequate institutions.

Although it is an important attempt to value ideas and accentuate the endogenous logic of change, the adjustment of the RN does not work at all for several reasons: a) it becomes insufficient by relegating ideas to a secondary role and depending on contexts of uncertainty (John, 1998, p. 154) as ideas are not useful only in those scenarios (Mansbridge, 1980); b) they are placed as “road maps” or “focal points” but it is not clear what determines that selection in the face of other possible ones and given the weakness of interests (Gofas and Hay, 2010); c) if they count seriously, it would have to be admitted that institutions open or close possibilities for the realization of interests and determine preferences of actors and not the other way around (Goldstein and Keohane, 1993). In other words, if ideas are only an instrumental means for the creation of institutions, what prevents them, once embodied, from affecting and shaping the interests of agents? The logic of the RN resists considering “the fact that ideas often shape the interests themselves... the fact that ideas affect how actors come to see – and change – their interests” (Fischer, 2003, p. 36).

Sociological and historical neo-institutionalisms (SN and HN, respectively) tend to recognize the weight of ideas but register some analytical tensions. Unlike RN, action appears in both as highly structured and implies a difficulty in explaining agency and change (Hall and Taylor, 1996). Within the SN, March and Olsen (1984) considered that it was necessary to transcend institutional analysis that focused on results because it limited action to criteria of mere utility or efficiency. The contrast with that position implied a marked normative and cultural orientation that was expressed in their definition of institutions as an “enduring collection of rules and organized practices, embedded in structures of meaning and resources that are relatively invariant” (March and Olsen, 2006, pos. 550).

Embedded in these structures are “identities and belongings: common purposes and accounts that give direction and meaning to behavior, and explain, justify, and legitimate behavioral codes. These are structures of resources that create capabilities for acting” (March and Olsen, 2006, pos. 553). These institutional structures are condensed into a broad repertoire of rules, norms, and routines that determine both what is accepted as correct routes of action and the appropriate criteria for legitimation. The institutions are, consequently, cultural sedimentations of ideas and values that limit options for structuring social interaction and contain the predominance of the individual utilitarian interest. They provide “order to social relations, reduce flexibility and variability in behavior, and restrict the possibility of a one-sided pursuit of self-interest of drives” (March and Olsen, 2006, pos. 712). An important implication of this perspective is that it transfers the regularity generated by the institutions to a cultural or codified meaning level. Politics is not only organized according to instrumental patterns of production and distribution of resources, but also refers to the interpretation, direction, and meaning of social life. That is why institutions need to be legitimized. It must be admitted that “there is... no perfect positive correlation between political effectiveness and normative validity” (March and Olsen, 2006, pos. 699).

It is unquestionable that in the rules, norms, and practices, there are structures of meaning that are based on ideas (for example, individual equality before the law). The SN values them and simultaneously limits them. Understood as resources that are available to legitimize actions, behaviors, and practices, the structures of meaning reside at the cultural and ideational level. It would be possible to bring them closer to ideas and discourse to address flexibility among actors and institutional change. But the SN takes another route and gives institutions an over-determining weight over the actors; they limit their actions, model expectations, and make their experiences comprehensible. The actors, thus, develop within contexts and organizational structures that determine norms, scripts, narratives, and cognitive frameworks (Scott, 2013). The actors can orient themselves with a logic of “appropriation” of these structures, but they can hardly modify them. To the extent that institutions are considered to make ideas operable, without circularity, they do not have a substantial effect on their formation or the variability of actions. The result is an undynamic vision of social life and its mutability. Another “core assumption – according to March and Olsen (2006, pos. 574) – is that the translation of structures into political action and action into institutional continuity and change are generated by comprehensible and routine processes. These processes produce recurring models of action and organizational patterns”. The insistence on routine level and the formation of patterns of action has led to saying that we are dealing with “actions without actors” (Dowding, 1994, p. 111) or structures without agents.

HN takes premises similar to the other two institutionalisms and articulates them in terms of path-dependent processes. HN emphasizes the institutions resulting from the past as determinants of the choices of agents who, nevertheless, remain rational. On the one hand, the institutions are – roughly speaking – cultural and normative incarnations, which de facto implies valuing ideas; on the other hand, the actors move in a double logic: they are culturally determined, but their behavior obeys guidelines of rationality and calculation (Hall and Taylor, 1996). The weight of ideas is minimized. The HN is effectively an amalgam between the rational and the normative sociological. The amalgamation seeks to resolve the tension between maximizing actors and stability and the imbalances to which that tension gives rise.

The change is then paradoxically formulated and the place of ideas also because we find ourselves with an actor divided between his parameters of reference. The determination of the sequence of decisions that analytically imposes the path-dependent, limits the capacity of action beyond the current conditions and superimposes the weight of the institutions born from that sequence. Thus, it is not clear how the link between actors and ideas or norms affects institutional change despite the fact they embody them. In other words, “If institutions make ideas `actionable´, then one cannot appeal to ideas to create institutions” (Blyth, 2002, pos. 656). According to Hay (2006, pos. 1650) under that double logic of behavior, the HN is “unlikely to offer much analytical purchase on questions of complex post-formative institutional change... merely compounds and reinforces the incapacity of rational choice and normative/sociological institutionalism to deal with disequilibrium dynamics”.

SN and HN have a conceptual structure that is more open to ideas, but as with RN, they have a lack of endogenous vision for change. The more the contextual and external character of the norms is accentuated, the more the institutional mutation tends to be perceived as exogenous and to reduce the role of ideas. As Rothstein (2005) has said, institutions should not be thought of only as incentive structures without a subjective weight, they are bearers of ideas, of collective memories that grant them reliability or not. Institutions change in relative conformity with the views held about them.

Interests and ideas in contingent contexts

Blyth (2002) argues that the limited relevance given to ideas derives from the conceptual confusion about the relationships between them, interests, and institutions. They tend to be considered mutually exclusive, especially the first two. Interests are presented as the key and sole determinant of the choice of actors in every situation. Even those strongly imperfect regarding information tend to hold that individuals have an infallible capacity to choose the best strategy. Indeed, with this reasoning we only construct an ex-post and circular explanation: we find that the choice of actors crystallizes into a specific behavior and indicate that it involves interests since these are already conceptually implicit in the very notion of choice (Blyth, 2003).

The immediate unity between decision and interest lies in the fact that the latter is mistakenly considered as a “singular concept”, a kind of concept that cannot be “associated or related” to anything. The postulate of the exclusive priority of interest “ignores the fact that the concept of interest presupposes unacknowledged but very important cognates of interest, such as wants, beliefs, and desires... these cognates are not analytically separate from interest and must be considered as part of the concept of interest itself” (Blyth, 2002, pos. 717). Interests are categorically composed of several elements, and their articulation can present complex dynamics. They are a kind of “cluster concept”. In it, ideas play a substantial role and, to explain it, Blyth resorts to scenarios of uncertainty that can put in doubt the stability of the cluster.

He considers it limited to homologate the risk of decisions to the daily uncertainty. In Knightian-type uncertainty, a situation is generated – as in a radical crisis – where the risk, consequences and products cannot be calculated. It is understandable that in such a condition there are no references to acting and the agents are disoriented before the infinity of alternatives since the actors, as he has indicated (Rydgren, 2009, p. 73), “are ‘meaning-seeking’ beings in the sense that they strive to obtain cognitive closure”. In that context, interests “cannot be given by assumption or structural location and can be defined only in terms of the ideas that agents themselves have about the cause of uncertainty. Without reference to such ideas, neither interests nor strategies would have meaning…” (Blyth, 2002, pos. 773). Consequently, the results produced in such a situation “will also be a function of those ideas” (Blyth, 2002, pos. 785).

Resorting to an uncertainty scenario has a strong implication: the behavior of the actors is not a direct expression of their interests, but of the perception actors have of them in a given situation. The ideas specify, clarify, and make actionable the interests in specific social contexts. The two are intimately related without ever merging (Blyth, 2003). The argument coincides with others that assume the importance of the social construction of interests (Wendt, 1999, p. 113-35; Barnett, 2008). The conceptual indistinction between actors and interests makes it extremely difficult to address institutional variation and to explain the products of political processes.

Concerning change, Blyth has proposed four theses on the role of ideas that are of analytical interest. The first can be formulated in terms of a cognitive function: ideas reduce uncertainty before the formation of institutions. As in high uncertainty conditions institutions are not a reliable parameter, ideas allow for its reduction by operating as an “interpretive framework” and by providing the actor with an interpretation of the causes that produce it. On that operation, uncertainty is reduced before institutional formation; otherwise, construction itself would be impossible. The reduction of uncertainty and the political provision of new institutions are, therefore, two consistent but separate events. The argument becomes more evident if one thinks about moving from Keynesian to monetarist policies in the UK (Campbell and Pedersen, 2001). This argument is central to Hall's work which we will see below.

The second thesis postulates that ideas generate resources by enabling collective action and fostering coalitions. By interpreting crisis, ideas promote collective action by changing the perception of the costs and benefits of particular courses of action. As the interpretation takes hold or as part of that process, “political entrepreneurs” emerge who build and disseminate analyses that tell other actors why the world is different now. They offer a “new programmatic menu” that encourages the aggregation of political actors and the redefinition of interests. In this context, “ideas make them collectively actionable” (Blyth, 2002, pos. 883).

According to the third thesis, in the struggle over the current institutional context, ideas are used as “weapons”. They are the main resource for delegitimizing it: ideas fix the causes of a problem as a matter of institutional malfunction or a subset of its fabric. Under this construction, institutions are part of the problem, and their function requires legitimacy. Implicitly, the thesis coincides with SN. However, the dynamics are thought of from the perspective of the actors and their capacity to act. The agreement lies in the fact that institutions must generate meaning for the actors based on the body of values that reside in their formation (March and Olsen, 2006) and from which some policies are derived and not others. De-legitimization implies “answering” precisely the ideas that give them foundation and forming others.

The fourth thesis identifies ideas as a “blueprint”. They are used by actors to rebuild institutions after a successful period of contestation and criticism. The central point is that the ideas would act as “blueprints” for institutional building. The argument implies the fact that the articulation of a discourse of institutional delegitimization already contains a set of ideas about the desired institutions. Therefore, ideas would “dictate the form and content of the institutions that agents should construct to resolve” crises situations (Blyth, 2002, pos. 915). In this line, the thesis identifies the role of ideas for stability. By crystallizing into new institutions, they facilitate the coordination of expectations about the expected future through the establishment of conventions on appropriate policies.

Blyth's arguments range from the height of the crisis to the presence of new institutions: ideas are key to change and stability. His theses have been implicitly recovered or are the expression of various institutional and discursive approaches. He has also been harshly criticized because he maintains a dualism between the material conception of interests and the weight of ideas (Hay, 2006, p. 1.703-1.831), a point of tension in the institutional theory of rational choice.

Institutional change, social learning, and coalitions

In a much-debated article, Hall (1993) analyzes the transition from the Keynesian model to the monetarist or neoliberal model as a change of political paradigm. Although he has been criticized or adjusted by other authors, his argument continues to be a crucial reference in the analysis of ideas. In it, Hall firmly situates their role in the processes of change and policy formation. These processes correspond to types of social learning (SL). Among the first to use it for political analysis, Heclo (1974) introduced it to solve the problem of whether the state can act autonomously and at the same time formulate sound policies since it does not submit to the influence of parties, interest groups, or elites. This question is particularly interesting in welfare contexts.

As is recognized, state-centric perspectives presuppose that autonomy is key to the formulation of policies of general interest. Against pluralism, they deny the positive character of social pressure and the postulate that the diversification of state instances places structural limits on interests. Although the “general interest” remains notably imprecise, it is the result of preferences among officials; a kind of “pluralism within the state” (Hall, 1993, p. 275). SL allows Heclo to correct this closed image without reducing the role of autonomy: the policies are not arbitrary but rather cognitive results of previous experiences. The government, he argues (1974, p. 306), not only exercises power but also becomes confused in contexts of uncertainty. In the logic of the trajectory, the policies of the previous moment determine the next via SL.

Hall argues that the legacy imposed by the orientation of policies reduces the role of social pressure and assumes that experts – from the state or at the interface with intellectuals – are the ones who generate SL and exclusively design them. The formulation of Heclo's SL reduces or cancels out both the role of politicians and the need for social dialogue in the policy formation process. It gives all the credit to public officials or technicians. Autonomy now rests on this assumption. The concept of SL implies a strong role for ideas, but its development becomes an internal and almost exclusive dimension of the state (Heclo, 1974; Freeman, 2006). In contrast, Hall (1993, p, p. 276-78) argues that SL refers to the willingness to stably modify behavior or orientation to affect policy formulation. It takes many forms depending on the level at which the change or adjustment of public policy occurs. Depending on the recording of these changes, more open SL dynamics occur that deny or nuance Heclo's view.

The formation of policies is carried out considering three levels: goals, means, or instrument designs and their specific establishment (Hall, 1986, 1993). If we wish to raise the quality of life of older people (goal) we could approve a law for the elderly (means) and define (instruments) that any person over 60 years of age will receive monetary compensation. These are changes of the first order; they only imply the adjustment or mutation in the instruments while the other levels remain stable. They are second-order when the means and instruments are altered, and only the goals remain the same. Third-order changes imply the change in the instruments, means, and the hierarchy of goals. They seldom occur and, in certain situations, are expressed in phenomena such as the passage from Keynesianism to neoliberalism.

In the relationship between SL and these orders, Hall succeeds in formulating the role of ideas and the extent of change incisively. The first and second orders fit more clearly into Heclo's SL concept and indicate routine patterns of adjustment. The third is more problematic, expressing a profound transformation “of policy discourse associated with a paradigm shift” (Hall, 1993, p. 279). Anderson (1978, p, p. 23) stated that: “the deliberation of public policy takes place within a realm of discourse... policies are made within some system of ideas and standards which is comprehensible and plausible to the actor involved”. On that basis, Hall (1993, p, p. 279) confirms that policymakers work within a framework that “is embedded in the terminology through which policymakers communicate about their work, and it is influential precisely because so much of it is taken for granted and unamenable to scrutiny as whole. I am going to call this interpretative framework a policy paradigm”. The paradigm is the framework through which policymakers view the world, prescribes how the economy and society should be considered; it determines not only what policies are possible and with what instruments, but the very “nature” of the problem being addressed or what others call the definition of the problem (Stone, 1988). In this sense, “policy paradigms can be seen as one feature of the overall terms of political discourse” (Hall, 1993, p, p. 290).

To the extent that political paradigms establish new evaluation parameters, they are not commensurable in technical terms. For the defenders of the new ones, it isn’t easy to agree with their detractors on a set of data to establish judgments as technically valid for both. Behind this difficulty, substantial implications persist. The choice between paradigms is never scientific, although experts play a role. It is a battle of political judgments whose outcome depends not only on arguments but also on positions within a framework. The change of paradigm is either predetermined or accompanied by a shift in the locus of authority over policy formation. In this shift, politicians, in turn, must consider which experts or technicians they believe to be authorities, depending on whether the new paradigm appears to be robust and coherent. Authority only becomes entrenched (and routine) when new policymakers have recognized authority and can adjust procedures without the criticism of detractors altering the strength and coherence of the paradigm.

The implications of third-order change are diverse. The political mutation is not the result of the autonomous action of the state; it is a response to the presence of different groups and media in the debate of society. This dynamic breaks the closed process of policy formation, in which something similar to “policy networks” or “issue networks” are generated that expand the market of ideas. Within this framework, parties, interest groups, and other political actors not only hesitate, but also exercise influence or power. But doubt and power go together; they do not exclude each other; neither do ideas and interest. The actors require, we would say, ideas and interests to resolve the doubt and guide their influence. If we start from a rigid distinction between grassroots interests and ideas, we do not fully understand policy formation. It is not accurately described if it is characterized only as the pressure exerted by interest groups or parties: “The state is also linked to society by a flow of ideas between the two spheres” (Hall, 1993, p. 289). Competition for power is itself a vehicle for SL.

SL definitely induces an emphasis on ideas. But ideas are articulated through discourse:

Politicians, officials, the spokesmen for social interests, and policy experts all operate within the terms of political discourse... at a given time, and the terms of political discourse generally have a specific configuration that lends representative legitimacy to some social interests more than others... defines the context in which many issues will be understood (Hall, 1993, p. 289).

These actors exercise power and acquire it while being able to influence political discourse. In doing so, they, in turn, influence policy formation without using the typical routes of influence. Although Hall does not make it explicit, political discourse is an arena of action – a sphere in which ideas are contested and influence is generated so that formal or typical conceptions of the political system and its actors do not register. Within the discourse framework, the media and its association with experts or parties can be incorporated as actors in the processes of formation and legitimization of public decisions. And, of course, coalitions play a central role in this process.

From this set of influences, it is clear that “ideas are an important dimension of the process in which policy is made”. The articulation of policy around discourse helps to understand how policy formation is “structured by a particular set of ideas, just as it can be structured by a set of institutions” (Hall, 1993). Ideas, then, are also institutions, but in a particular way. On the one hand, both are mutually reinforceable because institutional routines confirm bodies of ideas about what is possible and desirable. On the other, “ideas embodied in a policy paradigm have a status somewhat independent of institutions that can be used, as in the case of monetarism, to bolster or induce changes in institutional routines” (Hall, 1993, p. 290). Ideas are a means of institutional change and operate under the conformation of SL processes, frameworks, and political discourses that express the influence of different actors.

The Advocacy Coalition Framework (ACF) developed by Sabatier and Jenkins-Smith (1999) is another recognized perspective that considers ideas central to political processes and links them to coalitions and learning patterns. One of the reasons for its broad influence lies in the fact that it proposes an open delimitation of the political subsystem. The formation of decisions and policies is not reduced, as Heclo (1974, p. 105) indicated, “to a small group of insiders” of the government. It includes a repertoire of actors that transcends the classic delimitation that focuses on legislators, officials, parties, or leaders of interest groups, and also incorporates specialized researchers and journalists as well as judicial officials (King, 1973; Sabatier, 1998; Sabatier and Jenkins-Smith, 1999). It results in more complex dynamics that do not fit with the dynamics supported by the policy network analyses either. While these analyses consider the resources that structure the relationship among actors as the key piece in policy formation, the ACF emphasizes belief systems (BS) and learning orientation (Sabatier, 1987; Smith, 2000). For Cairney (2018, p. 203-4) the difference matters because network analysis is often used to “isolate” privileged relationships between actors and government. In contrast, the ACF considers diverse levels of government and the presence of multiple, rationally-limited actors: decision makers in contexts of uncertainty and dispute within processes that can be very slow in producing results.

In terms of guidance, SL compensates for the limited rationality of the actors. Without denying rational patterns of action, it conceptually allows drawing a dynamic but stable political interaction. As we have seen, if it were based on pure interest, it would be highly disordered or lacking in structure. The ACF elaborates a theory of policy formation in which the actors influence them by transferring their beliefs rather than their mere material interests (Cairney, 2018; Weible, 2018). As in Heclo, between actors and policies, there is more than just power and interest. As in Hall, actors acquire influence because they can articulate a body of ideas – in this case, more technical ones – and not only because they are in a position to exert influence (Weible and Carter, 2017). More precisely, it follows from the relevance of BS that “many actors may be influential because they share a set of beliefs with a large number of others; translating those beliefs into policy decisions and outcomes is a common project” (Cairney, 2018, p. 3). It is precisely the beliefs that are the “glue” that brings together certain individuals and that sustain their coalitions (Weible and Sabatier, 2018).

The political subsystem is the basic unit where political processes occur. From the production of results, this subsystem is shaped by the actors involved in the effort to achieve them and by the issues around which they are grouped. This blurred image does not presuppose that anyone can effectively participate in policy formation; not all individuals are involved (Sabatier and Jenkins-Smith, 1999; Jenkins-Smith et al., 2018). Coalitions are the means and implicitly constitute limits to the openness of the political subsystem. They involve a wide range of “people from a variety of positions (elected and agency officials, interest group leaders, researchers, etc.). Individuals share a particular belief system (i.e., a set of basic values, causal assumptions, and problem perceptions) and show a nontrivial degree of coordinated activity over time” (Sabatier, 1998, p. 139).

Political processes are spaces – in some sense thematic – within which one coalition, with a specific BS competes with another to dominate areas of the political subsystem or influence its decisions. Coalitions are founded on common belief systems, but they also share resources and in that sense are “profitable” because they facilitate collective action by reducing transaction costs. They also institute strategic forms of coordination (Sabatier and Brasher, 1993; Henry, 2011). In the coalitions, there are weak forms of coordination carried out by “auxiliaries” who share information and commit themselves intermittently. Strong coordination also persists, exercised by leaders committed to the political system and its issues (Weible and Sabatier, 2018).

As a result of the political system, programs and public policies are configured as an implicit or explicit migration of the beliefs of one or several coalitions. They embody theoretical and causal pre-positions about problems (Mazmanian and Sabatier, 1983) and achieve solutions as a product of negotiations between coalitions or the expression of the weight of one of them. In any case, they are imports of the belief system that has crystallized into goals, rules, and incentives on a variety of topics (Jenkins-Smith et al., 2018, p. 486). The “import” explains the intense promotion or defense of specific policies by some sectors or coalitions and, conversely, their perception as threats by others. For this reason, the ACF analyzes policy formation as a dynamic of struggles, negotiations, learning, and political change.

Beliefs play a central role in such dynamics and, from the actor's perspective, are organized in three levels or systems (Sabatier, 1998, p. 194-95). Deep core beliefs are the fundamental normative values, even of ontological order, about human nature. They represent a body of rigid precepts, of unquestionable truths in the first order. Based on them, they justify the orientation and organization of society, the general role of government and the market, and the conception of welfare. These justifications crystallize into solid cultural forms (Jenkins-Smith et al., 2018). They are, in some sense, paradigmatic and address the very essence of a society's identity. They cover most areas of politics and are unlikely to change in the face of any evidence (Sabatier, 1998, p, p. 130).

Policy core beliefs link actors to the political subsystem precisely by defining topics and goals. They determine the fundamental political positions and therefore reflect not only normative orientations, but also more empirical aspects, particularly in the formation of decisions and policies. It is assumed that those who participate in this process have experience capable “of applying certain deep core beliefs to develop policy core beliefs in that subsystem” (Sabatier, 1987, p. 195). In any case, there is no one-to-one correspondence between the two belief systems. Therefore, a certain flexibility for change at the political level is generated without diluting any resistance. Secondary beliefs refer to the instruments for the achievement of objectives implicit in the other belief systems and, therefore, can acquire a more empirical dimension. The three belief systems maintain a hierarchy of operation and flexibility among themselves.

Since changes are mutations in politically-oriented beliefs, they can be classified concerning the belief system they affect. Those registered in the core (general and political) are the most important and imply changes in the goals of the system, something similar to Hall's paradigmatic idea. They are extremely challenging because they affect normative dimensions in politics; they become almost impossible if the coalition that holds them remains in power (Sabatier and Jenkins-Smith, 1999, p. 147-48). Because they are secondary, the other changes are more viable or, in line with the principle, more probable.

There are two conditions or routes of change: internal and external events. The disruptive effect of some external shocks depends on the minority coalition managing to establish them as confirmation of its belief system and to mobilize resources accordingly. The internal ones, through crises, serious political errors, or scandals, can affect the composition of coalitions and disrupt the belief system (Weible and Sabatier, 2018, p. 198-205). The conflict between coalitions depends on the intensity of the perception with which one of them sees its core beliefs threatened (Jenkins-Smith et al., 2018). In conditions of change, a central avenue for the ACF is one that is achieved through a political orientation of learning at the internal level. The guidance refers to a lasting alteration of intentional thoughts and behaviors as a result of past experiences or new information, and with an effect on attention or policy revision (Sabatier and Jenkins-Smith, 1999, p. 123; Freeman, 2006). Social learning is thus a mechanism of attention and regulation of cognitive conflicts that appeals to the flexibility or modification of beliefs of coalitions. In this process, science and technical information are central in the definition of problems, in the attribution of causalities and solutions. Hence, experts (university researchers and scientists, political analysts, consultants, etc.) play a crucial role in the formation of policies and the shaping of the belief system (Sabatier and Jenkins-Smith, 1999, p. 192).

For May (1992), the ACF's attempt to “technify” the entire learning process does not negate the strategic political use of knowledge. The intensity of conflict (“cross-coalition learning”) substantially affects the willingness to learn. In low-intensity conflicts, it is not motivated because coalitions can find non-rival subsystems of action. In acute or polarized conflicts, it is also not encouraged because coalitions maintain immovable positions. Willingness to learn flows best in medium-intensity conflicts (Sabatier and Jenkins-Smith, 1999, p. 206-7). For this reason, and regardless of the intensity of the conflict, the ACF introduces into the repertoire of actors of the political processes the figure of the “brokers” whose function is to minimize them, promote viable compromises between coalitions and reinforce the government authority for decision making and policy implementation (Cairney, 2018, p, p. 3). Management and mediation are required to break the resistance of coalitions.

The vindication of ideas and the reference to SL does not unify perspectives. Behind an open conception of the political subsystem, the ACF reaffirms a conventional concern of politics based on pressure or negotiation. In large part, this is because the interaction between coalitions is non-existent. Only the SL agrees. It is an internal mechanism for change. Hall (1993) limits the role of ideas too narrowly to the paradigmatic level but postulates a more open concept of SL by indicating a relevant role for political discourse.

In the background of the discussion lies the permanent tension between technical knowledge and politics. In any case, in a more refined way, this tension is repeated in other more contemporary analytical developments. The issue of SL finds a fertile field in its connection with political and governance processes. SL is closely linked to coalition building, epistemic communities, and knowledge-based actors (Scherfter, 2010). There is a debate on how science and experts influence the formation and change of public policies, and whether the instruments they promote and the institutional designs they postulate effectively stimulate learning and its good performance.

Gilardi and Radaelli (2012) identify 4 types of SL in governance processes. Their distinction is important because the types diverge in their micro-foundations and have different normative consequences (p. 170-71). Although they do not consider it, the four types could be arranged on an axis whose extremes range from technical knowledge, under rational criteria, to politics and its symbols. The first type is identified with instrumental learning (Radaelli, 2009; Gilardi, 2010). It postulates theories of a bureaucracy governed by criteria of rationality and efficiency, not of normative order. Evidence on policies informs its decisions and agenda. It is characterized by process of knowledge updating under a Bayesian logic.

In this rationality, the best decision is estimated due to the analysis between prior beliefs and evidence, which in combination produces the so-called “posterior beliefs”. Prior beliefs are contrasted with new information and generate a “Bayesian update”. Thus, policymakers update their beliefs based on the effectiveness of the technique. They are all exposed to the same evidence, but their subsequent beliefs vary according to their prior beliefs. Only more solid evidence makes this variation commensurable and updates beliefs. Solid means that there are many points of ascertainment (analysis, evidence) and that the variability between them is considered low. If there are few and the variability is high, updating beliefs is not relevant (Gilardi and Radaelli, 2012, p. 168). This rationality implies that the disparity of interpretation tends to be neutralized, by the force of evidence, from one moment to the next but without ever reaching homogeneity. In one of its variants, agents do not evaluate policies according to statistical evidence but following “cognitive shortcuts” based on the representation and availability of data (Kahneman and Tversky, 2013). For example, using “success stories” that are assumed to be objectively so.

The second type would be reflective learning. Its field of analysis encompasses paradigmatic changes in societies and, therefore, institutional behavior and forms of social interaction. This level of analysis is close to that of Hall (1993). Campbell (1998) introduced the distinction between programs, paradigms, frames, and public sentiments as areas in which reflection is about ideas. However, reflective learning is today linked to forms of governance, especially the “new” ones. As networks support the latter, they comprise rational or non-hierarchical communication processes (Héritier and Rhodes, 2011). International coordination areas on public policy exemplify this type of governance: they are based on disseminating information through social and political networks with influence in different governance spaces. “Its instruments are benchmarking, peer review, common indicators, and iterative appraisal of plans and achievements of member states” (Gilardi and Radaelli, 2012, p. 168). In theory, this increases the capacity for innovation and learning, but this “open method” tends to generate asymmetry: there is more learning “from the top than from society” (Radaelli, 2008). Moreover, while making a reflexive use of knowledge, this type of governance introduces dynamics for “creating a pressure to converge” in a context in which expert elites capture and use knowledge better. Bovens, Hart, and Kuipers (2006) have shown that public policy evaluation policy patterns are reproduced in different countries. In reality, design and evaluation criteria are socialized as neutral and insuperable parameters.

Policy learning emphasizes that policy change is based on the authority of elected politicians and on the accumulated learning of the bureaucracy that defines a range of possible actions. For May (1992) it crystallizes in strategies that favor controlling the political dimension of issues, attenuating costs and responsibilities. Its effect is that the technical criteria of judgment decrease, organizations become politicized and reduce their efforts to search for better public policies (Brunsson, 1989). This dichotomy between rationality and politics is reformulated by Gilardi and Radaelli (2012, p. 170) by establishing that policy learning has different expressions of use: strategic – increases political control (Boswell, 2008); substantive – a resource to support a previous or decided position; and symbolic – communicates through signals and distributes blame. The fourth type of learning, the appropriately symbolic, helps to increase legitimacy, but not necessarily the government's good performance and its policies.

Instrumental learning projects an image in which good governance is guided by rational criteria and know-how internal to government structures. It fits well with Heclo's (1974) parameter, and if we take into account Radaelli's (2009, 2008) warning, the reflexive also concentrates on learning in the higher spheres. In that case, both learning moves under the assumption that efficient governance requires politics or political authority to give way to technical knowledge. It is not understood what authority would occupy; it would seem to be imagined as a necessary hindrance without any positive function. Without fitting in entirely, political learning fits better with Hall (1986, 1993), 1993) and above all with Blyth's (2013) reading of it, as we shall see shortly. Symbolism maximizes legitimacy as a criterion of governance. While it would seem easy to dismiss the symbolic and limit the political usefulness, the first two also have their difficulties.

Gilardi and Radaelli (2012, p. 171) point out some of them: the first is that the level of analysis of evaluation and policy change has not specified a convenient methodology that addresses the aggregation and connection between macro and micro levels, so that the effects may not be well evaluated. There is no standard, accepted measure of learning: without it, there is no distinction between what is and what is not learning (p. 172). This imprecision stimulates the “mantra” that learning has taken place in the face of crisis events as a mechanism to restore control and legitimacy to decision-makers (p. 173); symbolic learning is used. Several studies have shown that the recognition of specific public solutions, proclaimed as evidence-based or based on technical research, is articulated by stakeholders, epistemic communities, coalitions of different forms, or knowledge networks that give them veracity (Hajer, 1995; Dunlop and James, 2007).

This reflexive dynamic of FS has its contrasts: networks do not always encourage independent regulators to adjust to new learning patterns: they merely “reinforce the autonomy of regulators and their insulation from democratic processes” with the responsibility that this entails (Gilardi and Radaelli, 2012, p. 175). Börzel and Heard-Lauréote (2009, p. 142) argue that these knowledge networks primarily fulfill a socialization function, a standard definition of problems and solutions. In other words, they play a constructivist role. As Freeman (2006, p. 373) has indicated, “Learning begins in uncertainty: if there were no uncertainty, there would be no need for puzzling. This uncertainty is in part a function of inadequate information”; it is also a primary function of interaction and communication between agents.

On the bridge between information and interaction, Hall's position is interesting. Contrary to him, in examining economic policy in Great Britain, Oliver and Pemberton (2004) emphasize that policy learning does not always lead to paradigm change because it depends on the institutional capacity to drive the impulses for change. A paradigm that fails is not necessarily wholly replaced. Blyth (2013, p, p. 11-13) agrees that even when empirically the conditions are given, change is not generated, among other reasons, because the agents and networks that manage the knowledge related to the current paradigm choose to validate it by reproducing interpretation criteria and routes. This reading reinforces the constructivist explanation.

Blyth (2013, p. 2-7) argues that a paradox governs Hall's position on the paradigm shift: on the one hand, his causality refers to the accumulation of anomalies and failures, which implies a Bayesian logic in terms of rationality; on the other, the causality appeals to a discursive logic where ideas are central (i.e., of a constructivist type). We would say it attempts to link an instrumental SL with those that register a political dimension. The paradox is not a weakness; it is productive: its strength lies in the solution given by Hall. What prevents these causalities from being mutually exclusive is that paradigmatic changes acquire, by their scope, autonomy from Bayesian logic: they do not respond mechanically to the empirical accumulation of the deficiencies of the model in force. They are not commensurable in technical or scientific terms, but at the same time, the impulse reflects the failures of the previous paradigm. The change is autonomous because it is sociological rather than scientific: it needs to be self-founding by generating its patterns of legitimacy and authority. This need arises from the recognition that, in political terms, the veracity between paradigms does not rest on the criteria of one, and these are, without further ado, recognized by the others.

In the paradigm shift, authority and incommensurability of knowledge are the keys to SL. However, they have different roles. It is authority that allows linking the two types of SL and dissolving the paradox: “Although learning is most certainly about cognition, politics is not just who thinks, it is also about recognition: who gets to (authoritatively) speak. This is why authority is the real conceptual”. And Blyth (2013, p. 15) adds: “The struggle...is a struggle over the meaning of anomalies, not their existence. Incommensurability ensures that meaning is always contestable... authority, in such instances, matters perhaps most of all...”. The richness of Hall's paradox does not lie about apparently competing levels of causality. It introduces us to the analysis of the tension between change and incommensurability in Kuhn's route, but above all, in recognizing that this tension is resolved – until the next event – in a process where ideas are central to the dynamics that found discursive authority.

Discourse, policies, and discursive coalitions

The Argumentative Turn (Fischer, 2003; Fischer and Gottweis, 2012) and the discursive neo-institutionalism (Schmidt, 2008, 2012) have recognized the contributions of those who give a significant place to ideas in political analysis. Still, they revalue them in a different conceptual framework: the discourse and the interaction between actors to which it gives rise. Both perspectives assume essential elements of discourse and neo-institutional theories. They are based on a constructivist conception, but the first one is more aligned with the post-structural schools, the argumentative and deliberative perspectives, while the second one emphasizes the strategic of the discursive interaction.

Given its post-temporal feature, the argumentative turn (AT) radically refutes the position that the analysis of processes and policies can be reduced to a technical-scientific dimension. It recognizes that processes are social constructions, fields of discursive battle, results of communicative and argumentative interactions between actors (Fischer and Gottweis, 2012, pos. 70). Public policies have an empirical dimension, but we could say they are also institutionalized discourses. They are texts expressing orientations or preferences presented under a specific frame and narrated in a certain way (Fischer, 2003, pos. 176-85). It is in this sense that AT seeks to understand how the empirical and the normative are interwoven in the discursive processes of policy construction. Consequently, it emphasizes argumentative frames, narratives, or storylines and the access to and use of discursive practices as basic units of analysis (Fischer, 2003, pos. 185-90). Discourse is the reference category of these units. In this framework, human action is not limited to rational parameters or immediate empirical variables; the actors are “culturally shaped, communicatively based, socially motivated, and emotionally grounded” (Fischer and Gottweis, 2012, pos. 7). Under this variety of impulses, discourse allows us to observe how their interactions and dynamics are structured.

Discourse is more than discussion, talks, or limited linguistic interactions. It is an “instrument” that corresponds to the fact that all reality must be meant contextually. Discourse refers “to historically specific systems of meaning which form identities of subjects and objects” (Howarth, 2000, p. 9). It is unquestionable that objects, practices and actions of all kinds are socially charged with meaning and that this is conditioned by social and political struggles of the context. This process is executed through discourse. Along these lines, Hajer and Versteeg (2005, p. 175) defines it as “a specific ensemble of ideas, concepts, and categorizations that are produced, reproduced, and transformed to give meaning to physical and social relations”.

For Fischer (2003) showing how this process of construction of meaning arises and the importance it acquires for social life and its interactions is the task of discursive political analysis. At the level of interactions, the study implies understanding which social practices and which specific logics of power reproduce discourse. Discourse practices delimit the range of subjects and objects through which “people experience the world, specify the views that can be legitimately accepted as knowledge, and constitute the actors taken to be the agents of knowledge” (Fischer, 2003, pos. 86). Over time, these discursive delimitations are established as norms governing the understanding of social life, and the interpretations to which they give rise are little reflected by the actors (Shapiro, 1981, p. 130). They remain embedded in institutional practices and deliberations.

The political analysis then begins by recognizing that discourses are distributed across institutions. In the face of the dominant ones, there are always others who compete for recognition and power. Actors speak from specific conceptions and positions. It is essential in the analysis to carry out identification and validation through the institutions reproducing them. The place of an actor is not a solitary projection; it is constituted by a discourse that refers to another one, as when the opposition speaks to the government or the employer to the worker (MacDonell, 1986, p. 2-3; Fischer, 2003, pos. 86). A discourse is dominant when it determines the nodal points of the discussion, accentuating some and excluding others; when its concepts provide meaning to the interpretations on a variety of themes, problems, situations and actors; when it is socially or politically successful. In short, in the interpretative processes. Besides the capacity to structure the position of the actors, the discourse crystallizes into practices that support or modify power relations in communities or between groups. The discursive resources available in a specific social situation (the appeal to rights or technical aspects, for example) and how they are stratified according to those power relations, are another central variable of the analysis (Fischer, 2003, pos. 92-99).

Policy formation is an argumentative battle. The analyses must pay attention to the content of the discourse that supports or promotes them, observe the dynamics of actors as a discursive interaction, and identify how all this impacts on institutional practices. In this way, social interaction and the position of actors are not considered the result of a set of defined roles; they are understood as an exchange of competing arguments aimed at determining the meaning of the “realities” in dispute (MacDonell, 1986, p. 43-59). In addition to the arguments themselves, necessary in the content is the information that is assumed to be valid or refused as irrelevant; understanding the claim behind a discursive position and, therefore, examining the argumentative structure, style, and context that justifies specific actions and rejects others. It is necessary to identify how the assumed “knowledge” can promote or not agreements and to allow the reorientation of the interests of the actors (for example, when we moved from the concept of acid rain to that of sustainable development).

In line with Hajer (1995), Fischer (2003, pos. 94-99) assumes that in the interaction, the actors are constituted by the discourse. Thus, ideas that are not within the discursive interaction are irrelevant; the argument is only plausible concerning other positions. The actors are actively engaged in the elaboration of views, in choosing and adapting ideas, in shaping and disseminating them. They seek not only to convince their opponents, but to engage other actors in their problem definition. Actors attribute meanings to each other and to the positions their discourse assigns to others; from there, they attribute shame or responsibility. In the discursive interaction, a key point is how framing gives a topic the status of problematic while others appear as adequate according to certain predominant discourses.

One mechanism by which these framings are created or reproduced is storylines. They are a type of narrative that allows the actors to land different discursive categories to give meaning to the social order or specific problems (Hajer, 1995, p. 56). Its primary function is to provide unity in the bewildering variety of problem discourse components that otherwise have no clear or meaningful pattern of connections. It is a synthetic construction that works because people do not understand the world through complicated conceptual systems of a theoretical or analytical order. The narrative cancels out some aspects and accentuates others; it mixes information with normative orientations and reinforces their meaning. Storylines are social constructions that express the position of actors or institutional practices in the struggle for the sense of the world or its issues. After all, the social order depends on successfully reproducing the understanding of which actors and institutions are vital to that order (Fischer, 2003, pos. 99-101).

The uniqueness of the discursive analysis of the policies can be seen in its contrast with the ACF. In line with Hajer, Fischer (2003, pos. 94-100) perceives that advocacy coalitions, when based on core beliefs, are stably and uniquely shaped by a rigid delimitation. The associative game of the actors and the possible routes of action are thus reduced; the processes of change and the effects of innovation of SL become slow. The discursive analysis, in contrast, postulates the formation of discursive coalitions based on interpretations of themes, in storylines that guide preferences rather than technical aspects. These are aggregations of actors or individuals who do not necessarily meet or coordinate physically. They are oriented towards the definition of problems since it is built through discursive interaction. Advocacy coalitions, instead, have fixed preferences and seem to have an anticipated clarity in the way of conceptualizing the problem or issue of attention. Even under a debate format, for them, SL is a technical, cognitive process to which experts are central. For discursive coalitions, it is further a discursive and interactive process, and the same condition is fulfilled by the formation of policies.

Discursive institutionalism (DI) maintains a relationship of identity and distance with AT. DI is an “umbrella” concept that groups schools dealing with ideas, particularly those from the perspective of neo-institutionalism (Schmidt, 2008, 2010). For Schmidt (2012) this range of approaches is unified by the fact that they take ideas and discourse seriously; consider the institutional context as an essential framework of information; place ideas within a “context of meaning” and discourse within a logic of communication; and, finally, have an endogenous and more dynamic vision of change. But DI introduces a set of concerns different from ideational (Campbell and Pedersen, 2001) or constructivist institutionalism (Hay, 2006). It not only asks who is speaking and from what place something is being said, but also under what communicational dynamics. Along these lines, DI is distinguished because it sees in the discourse a more solid unity than the ideas to address institutional or policy change (Schmidt, 2010). For DI, ideas are essential, but those who focus exclusively on their generation, deliberation and legitimacy, relegate the implicit interactive processes of discourse.

As an organization of values, principles, and knowledge, ideas are located on three levels: specific policies, programs, and paradigms or philosophies. All of them contain cognitive and normative ideas (Schmidt, 2011, 2012). Cognitive ones are intended as a guide for action; they define and establish solutions to problems according to technical-scientific parameters. The normative ones legitimize policies by indicating that the solutions and the parameters of reference entail values that are identified with the ideals and interests of the general public. The classification, however, does not clarify why some ideas do succeed and rule or shape policies, programs, or philosophies while others do not. It does not specify what criteria alone make the adoption of a policy inevitable. The statement that “good ideas” (based on knowledge and appearing to be more appropriate) generates efficient and successful solutions is not entirely correct. In politics, many bad ideas succeed. If we resort to the paradigmatic level to answer about the success of ideas, it would be necessary to identify through which processes of discursive interaction it acquired validity because none of them seems to sustain itself. Consequently, “discourse is more versatile and overarching than ideas” (Schmidt, 2008, p. 7).

Connolly's (1983) conception of discourse alludes to institutionalized structures of meaning that orient political thought and action in a certain direction. They are structures of ideas that are represented. Schmidt (2008, 20, 2012) assumes this definition and identifies two dimensions of analysis for the success of ideas: their content or representation in discourse and their interaction. To consider the discourse as a unit of observation is to understand that it simultaneously refers to the ideas that represent us and to the interactive process through which they are transferred or exchanged. Without this discursive exchange, it is not possible to understand how ideas pass from the individual to the collective (Schmidt, 2008, 20, 2010). As an interactive process and above all as representation, discourse facilitates the deliberation and legitimization of our collective actions. Elaborating persuasive discourses helps the agents in the formation or change of institutions, and that capacity refers to their foreground discursive abilities that must be identified. This capacity expresses actors who need to make sense, who are emotional, and whose interests are not only, nor primarily, material (Schmidt, 2008, 20, 2010, 2011). The representative dimension of the discourse may involve the analysis of frames, narratives, memories, and other methodologies. As it considers to whom, how, and when a discourse is addressed, it can explain the success or failure of ideas.

The success of ideas also depends on the interaction dynamics testing the discourse as representation and process capable of their effective transmission and legitimation (Schmidt, 2008). The importance of discursive interaction lies in the connection it resolves. Even when agents are seen as bearers of ideas as in the ACF,

the connection between ideas and collective action remains unclear. The missing link is discourse not as representation but as interaction, and how ideas conveyed through discursive argumentation lead to action. But discourse also cannot be considered on its own, since it requires agents who articulate and communicate their ideas through discourse in exchanges that may involve discussion, deliberation, negotiation, and contestation (Schmidt, 2012, pos. 91).

Actors generate and deliberate ideas through discursive interactions that guide collective action within institutional contexts. Concerning these, individuals affiliate as “discourse coalitions, epistemic communities, and advocacy coalitions” (Schmidt, 2008, p. 302). All are forms of organization of actors articulated by ideas or beliefs. Discourse coalitions do not require the sharing of a whole body of ideas or visions; they compete based on the need to promote a program, some policy instruments, or specific ideas. ACF share a more narrowly defined set of ideas, are more cohesive, and have more direct access to policy formation. Epistemic communities identify with an idea or purpose but do not register any organized linkage (Schmidt, 2012, pos. 100-2).

Grouped under different forms, actors when entering the political sphere commit themselves to two types of tasks or dimensions. In the coordinative discourse, individuals are involved in the creation of political or programmatic ideas – in deliberating, negotiating, and reaching agreements among themselves given the variety of options and policy designs that are presented in a situation. It usually involves civil servants, politicians, experts, civil or interest organizations, and activists, among others. The communicative discourse may bring together some of these actors and incorporate others: “members of parties, the media, community leaders, social activists, public intellectuals, experts, think-tanks, organized interest, and social movements” (Schmidt, 2008, p. 310). Often these actors organize themselves as “political forums” or present themselves as “informed citizens” although they may include legislative figures. The communicative dimension of discourse involves the interaction between political actors and various audiences. It is oriented towards the deliberation, rejection, justification, or legitimation of ideas or political proposals before these publics (Schmidt, 2008, 2012).

Depending on the type of coalition that prevails and the balance of the coordinating and communicative dimensions, more or less accentuated forms of top-down or bottom-up discursive interaction are verified. In the former, the political elites generate and communicate ideas, and so there is a “master-discourse” which governs political vision and deliberation. In the second, political communication expands, and so does the generation of ideas. Institutional contexts influence the strength of one of the two types: for example, highly presidential regimes tend to favor top-down discursive interactions.

Public deliberation can be manipulated or subjected to effective rhetoric and does not in itself ensure better democratic outcomes (Schmidt, 2010, p. 18) because private interests are always present. However, in acceptably democratic contexts, public debates cannot be controlled by a single actor or a group of actors. The exercise of the two dimensions, coordination and communication, has a substantial influence on the success or failure of ideas and discourses. This exercise significantly affects the actors' capacity to influence because, among other reasons, it breaks with the conception that power is an exclusive function of the actors' position in specific scenarios.

The relationship between power and ideas is central to policy formation. However, it is not easy to understand if the former is taken as the latter's source or a mere expression of the former. At the same time, pure discursive interaction is insufficient to explain the influence of ideas or their success. Carstensen and Schmidt (2015) put the point well by considering ideational power (IP) as a specific category, distinguishable from other forms of power. In its broadest sense, IP refers to the ability of certain actors to influence the normative and cognitive beliefs of others through the use of ideational elements (discourse, practices, symbols, myths, narratives, collective memories, stories, frames, norms, grammars, and identities). It operates through persuasion, imposition, or directly through the influence of ideological contexts that define the range of possibilities of interpretation and beliefs of others (p. 5). There are three forms of IP.

The definition of power through ideas is similar to the IP because it has an enormous neo-institutional charge that inspires all three forms. The central point is that it is persuasive, relying on normative and cognitive reasoning and argumentation. The success of a cognitive argument depends on its ability to define the problems and solutions assumed to be adequate. Every argument, even the scientific one, requires its translation into politics a normative implication given the deliberative public arenas. This type of power emphasizes the capacity of actors to commit themselves to the ideas they hold. They are not internalized in the minds of the actors; they are resources in discursive battles and therefore require an exercise of critical and creative agency (p. 9). Power over ideas refers to the possibility of controlling the meaning of ideas, imposing or resisting new alternatives (p. 4-10). Typically, those actors who impose them have traditional power (coercive, structural, or institutional) and are in a position to promote their ideas to the exclusion of others; the others tend to accentuate the persuasive ideational character, but also in an instrumental way. Both actors may not listen as resistance (p. 11). The power in ideas refers to the authority that some of them have over others who are excluded. It implies the constitution of ideological background structures – knowledge systems, discursive practices, and reasoning parameters – determining which ideas have authority (p. 13). Once these structures are instituted, agents try to depoliticize ideas to the point they are taken for granted (p. 13).

In a different position, Béland and Cox (2016) argue that ideas are beliefs and do not have power by themselves (p. 4-6); they substantially influence political changes or modify power relations when they play the role of “coalition magnets”. These coalitions express the capacity of an idea to articulate a broad spectrum of interests or preferences of different groups or individuals and, therefore, it is feasible to be used strategically by policy entrepreneurs (p. 10). The capacity resides in the fact that the idea is ambiguous or polysemic (it admits different understandings), and its valence rests on a solid positive and emotional connotation (p. 6-14). Typically, these ideas are new in the political debate or acquire a new meaning that broadens their possibility of incorporating different perspectives. In short, they are manipulated by political entrepreneurs, promoted by key agents who give them authority and bring together interests that previously had no common platform. When these three features are fulfilled, coalition magnets are generated (p. 7).

According to the three types, IP acquires three features: first, it is exercised through the construction of intersubjective meaning structures that give meaning; second, it is top-down or the other way around; and it considers the discursive struggles in political arenas and the resistances to which it gives rise from below. That is to say, it claims agency and remains linked to logics of interaction and communication with certain instrumental features. Coalition magnets suspend the level of interaction and discourse but recognize the need for political entrepreneurs and even the instrumental use of ideas. On the other hand, they reject that these are a force in themselves: their potentiality lies in the possibility of a representation that crosses particular interests, brings them together, and gives them emotional and normative value. Among these nuances, today we find a fruitful field of debate on ideas.

Final considerations

1. Ideas do not replace or deny interests; but confining the behavior of individuals or social action to base-line-interests is limited and imprecise. Interests do not automatically translate into infallible routes of action regardless of the structures of meaning, ideas, and discourses that contextualize specific scenarios. The fusion between actors and interests that rational choice postulates as an unfailing condition for the orientation of action has been questioned in the field itself. The Bloomington school of rational neo-institutional inspiration disbelieves in such fusion. This approach recognizes that calculating individuals are “failing learners” (Ostrom, 2011, p. 11). For Aligica and Boettke (2011, p. 39) – members of that school – the “ideas – or correlated concepts such as learning or knowledge – frame and permeate choice. Ideas set into motion actions, ideas give solutions, but they also generate new problems and challenges... an account of human societies is fundamentally an account of the social avatars of ideas and knowledge, manifested through choices” (p. 39). Ideas, social knowledge, and decisions are relevant links in and beyond public policy.

2. Ideas are not only constructions that succeed or generate new problems, nor does SL lead (under a Bayesian logic) tendentially to an ever-better society. The thesis that SL is politically based at the general level of change and rational at more concrete levels is illustrative, but its dichotomous logic reduces the analysis. One of the most fertile warnings of Gilardi and Radaelli (2012) is their insistence on considering the macro and intermediate levels in which SL operates. There are thousands of intermediate social dimensions in which political knowledge is necessary, and the different forms of SL can act in a non-exclusive manner. The Covid-19 vaccination policy and the resistance movements have clarified this; the European community's forms of coordination are another example.

3. The observation of coalitions is a helpful tool for analyzing major changes, but there are undoubtedly intermediate dimensions in which SL operates, and ideas and new institutions are generated. Without renouncing this, the study of ideas must go beyond their role in the general framework of significant crises or paradigmatic changes. This is a significant insight of Schmidt (2008, 2010) vis-à-vis Hall (1993). The discursive interaction can delimit intermediate spaces of analysis in which forms of power are acting through ideational elements and without a crisis of major proportions being registered but decisions being taken. Moreover, the relatively general suspension of the parameters of reference did not occur only in great moments of crisis: specific conflicts or intermediate processes of social innovation register it. The new social processing of gender differences is just one example.

4. These innovations indicate that there are coalitions of different types or groups identified with them behind the ideas. The SL plays in many ways in social innovation, and Bayesian rationality cannot always fix routes (the interrupted clutch). The link between ideas and coalitions helps to explain their success, but also resistance to social innovation. Along these lines, we find the power of Béland and Cox's “coalition magnets”, which indicate that innovation requires aggregation of preferences of different actors. In social life, there are thousands of concepts or labels that articulate diversities. The concept of citizen allows for a unity of differences among diverse individuals; the search for non-binary pronouns articulates many individuals with diverse gender identities. The formation of these magnets and the institutions to which they give rise in contexts of patterned change are of significant analytical and research interest. So are the other considerations indicated.

Bibliographical references

• Abdelal, R.; Blyth, M.; Parsons, C. (eds.). Constructing the international economy. Cornell: University Press, 2015.
• Aligica, P. D.; Boettke, J. P. “The two social philosophies of Ostroms’ Institutionalism”. The Policy Studies Journal, vol. 39, n° 1, p. 29-49, Aug. 2011.
• Anderson, C. The logic of public problems: evaluation on comparative policy research. In: Ashford, D. (ed.). Comparing public policies. Sage, 1978.
• Barnett, M. Social constructivism. In: Baylis, J.; Smith, S.; Owens, P. (eds.). The globalization of world politics: an introduction to international relations. Oxford: Oxford University Press, 2008.
• Béland, D.; Cox, R. H. “Ideas as coalition magnets: coalition building, policy entrepreneurs, and power relations”. Journal of European Public Policy, vol. 23, n° 3, p. 428-45, Dec. 2016.
• Birkland, T. An introduction to the policy process. New York: Routledge, 2016.
• Blyth, M. Great transformations: economic ideas and institutional change in the twentieth century. Cambridge: Cambridge University Press, 2002.
• Blyth, M. “Structures do not come with an instructions sheet: interest, ideas, and progress in political science”. Perspectives on Politics, vol. 2, n° 4, p. 695-706, Dec. 2003.
• Blyth, M. “Paradigms and paradox: the politics of economic ideas in two moments of crisis”. Governance: an International Journal of Policy, Administration, and Institutions, vol. 2, n° 2, p. 197-215, abr. 2013.
• Börzel, T. A.; Heard-Lauréote, K. “Networks in EU multi-level governance: concepts and contributions”. Journal of Public Policy, vol. 29, n° 2, p. 135-51, Jul. 2009.
• Boswell, C. “The political functions of expert knowledge and legitimation in European Union immigration policy”. Journal of European Public Policy, vol. 15, n° 4, p. 471-88, May 2008.
• Bovens, M.; Hart, P.; Kuipers, S. The politics of policy evaluation. In: Moran, M.; Rein, M.; Goodin, R. E. (eds.). The Oxford handbook of public policy. Oxford: Oxford University Press, 2006.
• Brunsson, N. The organization of hypocrisy: talk, decisions, and actions in organizations. Chichester and New York: John Wiley and Sons, 1989.
• Cairney, P. An advocacy coalition framework of policy change and the role of policy-oriented learning therein. In: Lodge, M.; Page, E. C.; Balla, S. J. (eds.). The Oxford handbook of classics in public policy and administration. Oxford: Oxford University, 2018.
• Campbell, J. L. “Institutional analysis and the role of ideas in political economy”. Theory and Society, vol. 27, n° 3, p. 377-409, Jun. 1998.
• Campbell, J. L.; Pedersen, O. K. (eds.). The rise of neoliberalism and institutional analysis. Princeton: Princeton University Press, 2001.
• Carstensen, M.; Schmidt, V. “Power through, over, and in ideas: conceptualizing ideational power in discursive institutionalism”. Journal of European Public Policy, vol. 3, n° 2, p. 318-337, Dec. 2015.
• Connolly W. E. The terms of political discourse. Princeton, NJ: Princeton University Press, 1983.
• Dowding, K. “The compatibility of behaviouralism, rational choice, and ‘new institutionalism’”. Journal of Theoretical Politics, vol. 6, n° 1, p. 105-17, May 1994.
• Dunlop, C.; James, O. “Principal-agent modelling and learning: the European commission, experts, and agricultural hormone growth promoters”. Public Policy and Administration, vol. 22, nº 4, p. 403-422, Oct. 2007.
• Fischer, F. Reframing public policy: discursive politics and deliberative practices. Oxford: Oxford University Press, 2003.
• Fischer, F.; Gottweis, H. (eds.). The argumentative turn revisited. London: Duke University Press, 2012.
• Freeman, R. Learning in public policy. In: Moran, M.; Rein, M.; Goodin, R. E. (eds.). The Oxford handbook of public policy. Oxford: Oxford University Press, 2006.
• Gilardi, F. “Who learns from what in policy diffusion processes?” American Journal of Political Science, vol. 54, n° 3, p. 650-66, Jul. 2010.
• Gilardi, F; Radaelli, C. Governance and learning. In: Levi-Faur, D. (ed.). The Oxford handbook of governance. Oxford: Oxford University Press, 2012.
• Gofas, A.; Hay, C. The ideas debate in political analysis: towards a cartography and critical assessment. In: Gofas, A.; Hay, C. (eds.). The role of ideas in political analysis: a portrait of contemporary debates. London: Routledge, 2010.
• Goldstein, J. Ideas, interests, and American trade policy. Ithaca, NY: Cornell University Press, 1993.
• Goldstein, J.; Keohane, R. (eds.). Ideas and foreign policy: beliefs, institutions, and political change. Ithaca, NY: Cornell University Press, 1993.
• Hajer, M. Discourse coalitions and the institutionalization of practice: the case of acid rain in Great Britain. In: Fischer, F; Forester, J. (eds.). The argumentative turn in policy analysis and planning. Durham, NC: Duke University Press, 1995.
• Hajer, M.; Versteeg, W. “A decade of discourse analysis of environmental politics: achievements, challenges, perspectives”. Journal of Environmental Policy & Planning, vol. 7, n° 3, p. 175-84, Jan. 2005.
• Hall, P. A. Governing the economy: the politics of state intervention in Britain and France. New York: Oxford University Press, 1986.
• Hall, P. A. “Policy paradigms, social learning, and state”. Comparative Policy, vol. 25, n°3, p. 275-96, Apr. 1993.
• Hall, P.; Taylor, R. “Political science and the three institutionalisms”. Political Studies, vol. 44, n° 5, p. 936-57, Dec. 1996.
• Hardin, R. La acción colectiva y el dilema del prisionero. In: Colomer, J. (ed.). Lecturas de teoría política positiva. Madrid: Instituto de Estudios Fiscales, 2006.
• Hay, C. Constructivist institutionalism. In: Rhodes, R. A. W.; Binder, S. A.; Rockman, B. A. (eds.). The Oxford handbook of political institutions. Oxford: Oxford University Press, 2006.
• Heclo, H. Modern social politics in Britain and Sweden. New Haven: Yale University Press, 1974.
• Henry, A. “Power, ideology, and policy network cohesion in regional planning”. Policy Studies Journal, vol. 39, n° 3, p. 361-83, Aug. 2011.
• Héritier, A.; Rhodes, M. (eds.). New modes of governance in Europe: governing in the shadow of hierarchy. Basingstoke: Palgrave Macmillan, 2011.
• Howarth, D. Discourse. Buckingham: Open University Press, 2000.
• Jenkins-Smith, H. D., et al. The advocacy coalition framework: an overview of the research program. In: Weible, C.; Sabatier, P. A. (eds.). Theories of the policy process. Avalon Publishing, 2018.
• John, P. Analyzing public policy. London: Pinter, 1998.
• Kahneman, D.; Tversky, A. Choices, values, and frames. In: MacLean, L. C.; Ziemba, W. T. (eds.). Handbook of the fundamentals of financial decision making: part I. Singapore: World Scientific Publishing Company, p. 269-78, 2013.
• King, A. “Ideas, institutions, and the politics of governments: a comparative analysis, I and II”. British Journal of Political Science, vol. 3, nº 3, p. 291-313, Jul. 1973.
• Kristensen, P. H.; Zeitlin, J. Local players in global games: the strategic constitution of multinational corporation. New York: Oxford University Press, 2005.
• MacDonell, D. Theories of discourse: an introduction. Oxford: Blackwell, 1986.
• Mansbridge, J. Beyond adversarial democracy. New York: Basic Books, 1980.
• March, J.; Olsen, J. “The new institutionalism: organizational factors in political life”. American Political Science Review, vol. 78, p. 732-49, Sep. 1984.
• March, J.; Olsen, J. Elaborating the ‘new institutionalism’. In: Rhodes, R. A. W.; Binder, S. A.; Rockman, B. A. (eds.). The Oxford handbook of political institutions. Oxford: Oxford University Press, 2006.
• May, P. J. “Policy learning and failure”. Journal of Public Policy, vol. 12, n° 4, p. 331-54, Dec. 1992.
• Mazmanian, D.; Sabatier, P. Implementation and public policy. Glenview: MD: Scott Foresman, 1983.
• North, D. Institutions, institutional change, and economic performance. Cambridge: Cambridge University Press, 1993.
• Oliver, M. J.; Pemberton, H. “Learning and change in 20th-century British Economic Policy”. Governance, vol. 17, n° 3, p. 415-41, Jul. 2004.
• Ostrom, E. “Background on the institutional analysis and development framework”. The Policy Studies Journal, vol. 39, nº 1, p. 7-21, Feb. 2011.
• Radaelli, C. M. “Europeanization, policy learning, and new modes of governance”. Journal of Comparative Policy Analysis, vol. 10, n° 3, p. 239-54, Sept. 2008.
• Radaelli, C. M. “Measuring policy learning: regulatory impact assessment in Europe”. Journal of European Public Policy, vol. 16, nº 8, p. 1.145-64, Nov. 2009.
• Risse-Kappen, T. “Ideas do not float freely: transnational coalitions, domestic structures, and the end of the cold war”. International Organization, vol. 48, n° 2, p. 185-214, May 1994.
• Rothstein, B. Social traps and the problem of trust. Cambridge: Cambridge University Press, 2005.
• Rydgren. Beliefs. In: Peter, H.; Bearman, P. (eds.). The Oxford handbook of analytical sociology. Oxford: Oxford University Press, 2009.
• Sabatier, P. “Knowledge, policy-oriented learning, and policy change: an advocacy coalition framework”. Knowledge: Creation, Diffusion, Utilization, vol. 8, n° 4, p. 649-92, Nov. 1987.
• Sabatier, P. “The advocacy coalition framework: revisions and relevance for Europe”. Journal of European Public Policy, vol. 5, n° 1, p. 98-130, Feb. 1998.
• Sabatier, P.; Brasher, A. M. From vague consensus to clearly differentiated coalitions: environmental policy at lake Tahoe, 1964-1985. In: Sabatier, P.; Jenkins-Smith, H. (eds.). Policy change and learning. Boulder, CO: Westview Press, 1993.
• Sabatier, P.; Jenkins-Smith, H. The advocacy coalition framework: an assessment. In: Sabatier, P.; Jenkins-Smith, H. (eds.). Theories of the policy process. Boulder, CO: Westview Press, 1999.
• Scherfter, L. “The use of scientific knowledge by independent regulatory agencies”. Governance, vol. 23, n° 2, p. 309-30, Mar. 2010.
• Schmidt, V. “Discursive institutionalism: the explanatory power of ideas and discourse”. The Annual Review of Political Science, vol. 11, p. 303-26, Jun. 2008.
• Schmidt, V. “Taking ideas and discourse seriously: explaining change through discursive institutionalism as the fourth ‘new institutionalism’”. European Political Science Review, vol. 2, n° 1, p. 1-25, Feb. 2010.
• Schmidt, V. “Speaking of change: why discourse is key to the dynamics of policy transformation”. Critical Policy Studies, vol. 5, n° 2, p. 106-26, Jun. 2011.
• Schmidt, V. Discursive institutionalism: scope, dynamics, and philosophical underpinnings. In: Fischer, F.; Gottweis, H. (eds.). The argumentative turn revisited: public policy as communicative practice. Duke University Press, 2012.
• Scott, R. Institutions and organizations: ideas, interests, and identities. Sage Publications, 2013.
• Shapiro, M. Language and political understanding: the politics of discursive practice. New Haven, CT: Yale University Press, 1981.
• Sikkink, K. Ideas and institutions: developmentalism in Argentina and Brazil. Ithaca: Cornell University Press, 1991.
• Smith, A. “Policy networks and advocacy coalitions: explaining policy change and stability in UK Industrial Pollution Policy”. Environment and Planning C: Government and Policy, vol. 18, n° 1, p. 95-114, Feb. 2000.
• Stone, D. A. Policy paradox and political reason. Glenview, IL: Scott, Foresman and Company, 1988.
• Weible, C. Introduction: the scope and focus of policy process research and theory. In: Weible, C.; Sabatier, P. (eds.). Theories of the policy process. Avalon Publishing, 2018.
• Weible, C.; Carter, D. “Advancing policy process research at its overlap with public management scholarship and nonprofit and voluntary action studies”. Policy Studies Journal, vol. 45, n° 1, p. 22-49, Feb. 2017.
• Weible, C.; Sabatier, P. (eds.). Theories of the policy process. Avalon Publishing, 2018.
• Wendt, A. Social theory of international relations. Cambridge: Cambridge University Press, 1999.

Publication Dates

• Publication in this collection
  08 June 2022
• Date of issue
  Jan-Apr 2022

History

• Received
  24 Nov 2020
• Accepted
  12 Jan 2022