Research Traditions and the State of the Discipline in Brazil 1

The purpose of this article, that introduces the dossier titled “Policy Transfer, Diffusion and Circulation”, is, firstly, to present a simple and synthetic map of the field of studies of international policy diffusion, in order to situate the reader from Brazil and Latin America, as well as others interested from different regions. In its second section, the article presents a survey of the state of research on policy diffusion that has been produced in Brazil.

http://dx .doi.org / 10.25091/ S0101-3300201700010001 Nowadays, international policy diffusion processes are taking place in an increasingly visible and tangible manner.The displacement of urban policies, governmental programs, instruments of public action, administrative arrangements, norms, and state structures occurs very frequently.Policies are travelling rapidly from a government to another. 2 In the international literature on political science and public policy analysis, a large and diversified body of studies is readily available, displaying a variety of approaches, questions and strategies for analysis.In Latin America, Brazil included, this is a research field that is still under construction.The purpose of this article, that introduces the dossier titled "Policy Transfer, Diffusion and Circulation", is, firstly, to present a simple and synthetic map of the field of studies of international policy diffusion, in order to situate the reader from Brazil and Latin America, as well as others interested from different regions. 3In its second section, the article 4 presents a survey of the state of research on policy diffusion that has been produced in Brazil.
The history of policy diffusion scholarship is long and the first studies in the field date back to "at least 1889". 5The notion that policy ideas and instruments are transferred from a government to another has already been a current one in the mainstream debate in the areas of public policy analysis and comparative politics.Referential authors such as John Kingdon, Baumgartner and Jones, Peter Hall, John Ikenberry, Samuel Huntington and others had placed questions on transfers and diffusion -each one in her/his own manner -insisting on the relevance of processes such as stream of ideas, spillover effect, social learning, bandwagoning, and waves, using in certain cases metaphors and stylized narratives in order to explain the phenomenon.The discussion became more and more specialized and the field of policy diffusion gained strength progressively.As Benson and Jordan stated, "research into policy transfer underwent an exponential growth between the late 1990s and mid-2000s, but is now arguably in a more mature phase". 6Nevertheless, political scientists are not alone in their interest in policy diffusion, as sociologists, economists and psychologists are also producing a "growing body of work". 7The phenomenon is also calling the attention of geographers, 8 urbanists, 9 and historians, who have developed, since the 1980s, a subfield named "global history", "world history" or "transnational history".
The specialization of the field inevitably meant the development of new analytical tools for understanding how policy travels across governments, which improved the study of this political phenomenon and contributed to the unveiling of new research directions.In the following paragraphs we briefly present these new directions and the main recent advances in the field.The first is the fact that it was possible (and necessary) to overcome the trap of "methodologi-port, and Iasmim Ferreira Lomeu and Thales Leonardo de Carvalho, for their research assistance.
[4] This article and the present dossier are some of the results from the International Seminar on Policy Diffusion, held in 2016 at Centro Brasileiro de Análise e Planejamento (Cebrap) (Brazilian Center for Analysis and Planning), in São Paulo.For further information, please visit the website: www.policydiffusion.com.cal nationalism", 10 which allows us to think about public policy and the process of producing institutional instruments, programs and models taking into consideration inputs from beyond state borders.In other words, public policy is not only determined by domestic idiosyncrasies, characterized by cultural aspects, and framed by economic and political struggles among domestic interest groups. 11his means, in short, that the international and transnational dimensions also count, eventually being central to national/domestic policy making processes.In fact, there are actors that circulate from one state to another promoting policy models and best practices, as well as international organizations and transnational think tanks, inter alia, which have been directly engaged on public policy projects in different countries.Dynamics of institutional mimicry and lesson-drawing, for instance, usually emerge among neighboring countries. 12We should also recall that the analysis of policy diffusion brought together two subdisciplines of political science that, in spite of being complementary in different aspects, had followed each one their own path and development, namely: public policy analysis and international relations. 13he second advance was to move beyond the idea of comparative politics as the observation of similarities and differences between unities, circumscribed within the frontiers of a territory, bounded by certain jurisdictions (district, municipality, state, province, country etc.).The study of policy transfer, diffusion and circulation unavoidably insisted on the relevance of the transnational dimension of public policy. 14This means that more attention was brought to the fact that the production of public policies can be the result of processes that also occur in deterritorialized spaces, such as international meetings, forums, and summits. 15In these spaces there are frequent and relevant interactions among policymakers, as well as other actors.Recent summits such as Rio+20 or un Habitat iii are interesting examples of spaces where policy agendas are formed, cooperation agreements begin, funding for public policy projects is negotiated and conceded and policy models are divulgated and elected to be followed by specific governments.Transnational spaces where policymakers interact have constituted a locus of production, circulation and legitimation of ideas, models and instruments for public action.Policy transfer studies reinforce the idea that it is not sufficient to compare similarities and differences between units, but it is also important to trace the genealogy and trajectory of models and to understand how two (or more) countries are embedded in dynamics of mutual learning, competition and reciprocal influence.
A third advance consists on the replacement of stylized narratives, which have frequently used metaphors to portray those processes, al- [11] Howlett, 2000.
lowing the phenomena' s explanation to be more precise.The recent literature has consolidated a set of terms, approaches and theories to clarify specific processes that enhance or inhibit policy diffusion.For example, instead of arguing that ideas flow as if they were in a "primeval soup", there are different causal mechanisms that explain how and why they move from a government to another.Moreover diffusion analysis questions the pertinence and applicability of certain policy models in specific contexts, such as "global prescriptions" produced by international organizations.These models might not be the best alternatives, neither the instrument that will necessarily lead to policy success.Finally, the study of policy diffusion also allowed to highlight an emerging phenomenon, at least in Southern countries, such as Brazil: it is sometimes the case that it is not enough anymore only to produce successful policies, but it is seen as necessary to "market" them to the world, in a movement of policy legitimation (in the domestic arena) and exportation that may be connected to foreign policy strategies. 16n the next section we will present, very briefly, what may be understood as the three main research "traditions" or approaches that have been developed in the field, i.e.: policy transfer, diffusion and circulation.One should bear in mind, however: a) as some of these approaches have been developed rather recently, the term "tradition" must be employed cautiously; b) there are no purisms, in the sense that authors sometimes work in the frontier of different traditions, for example using both transfer and circulation approaches; and c) the three approaches are not presented in any chronological order, as policy diffusion studies are much older than policy transfer research.What explains our order of presentation is the scale of policy displacement: policy transfer deals with rather restricted processes involving a few political units and their interactions, while policy diffusion is seen as a process that encompass several states, eventually from distinct continents.Policy circulation is a term that has been employed to frame rather diffuse and multidirectional processes, as we will see ahead.

Policy transfer
The policy transfer approach condensed a set of studies that were already in development in public policy analysis along the late 1980s and the 1990s.These studies took into consideration issues of learning as well as international convergence, both emerging at that time on the debate.Originally it was basically an Anglo-Saxon proposal, which had as its central exponents David Dolowitz and David Marsh.Policy transfer is defined as "a process in which knowledge about policies, administrative arrangements, institutions and ideas in one political setting (past or present) is used in the development of policies, administrative arrangements, institutions and ideas in another political setting". 17The transfer movement often refers to a displacement of a policy from a government to another, being a process that is regarded, at least for analytical reasons, as a limited phenomenon.
Richard Rose was one of the precursors of the subfield, as he insisted on the fact that different countries -while facing similar public problems -draw lessons from other places, finding inspiration in actions previously undertaken by its pairs (local or national governments).They can seek for ready responses for public issues, even in a neighbors "garbage can". 18According to Rose, 19 lessondrawing can assume five different forms: copy (when a whole program is adopted); emulation (when adoption involves adaptation); hybrids (a combination of policy elements from two different places); synthesis (which combines policy elements from three or more different places); inspiration (when elements from policy programs in different contexts are used to produce a new model).It is evident, therefore, that the learning process does not necessarily produce exact copies from original models.As Rose stresses, policies are transformed when they arrive in new lands, frequently combining elements from distinguished places/experiences in order to produce adaptations of models regarded as innovative.
The first systematization of the subfield was done by Dolowitz and Marsh, 20 who wrapped up the debate, adding new insights and developing a broader framework for policy transfer analysis.Their proposal was built as a set of questions that could guide policy transfer research: Why do actors engage in policy transfer?Who are the key actors involved in the policy transfer process?What is transferred?From where lessons are drawn?What are the different degrees of transfer?What restricts or facilitates the policy transfer process?How is the process of policy transfer related to policy "success" or policy "failure"? 21oreover, for the authors transfer processes should be located in a continuum with two opposite poles.On one side is placed voluntary policy adoption and on the other extreme one will find coercive transfer, which, e.g., can occur after foreign imposition.
The proposal of Dolowitz and Marsh was important to consolidate policy transfer studies and has been widely used to guide numerous researches.In the words of Hadjiisky, Pal and Walker, they were the "first cartographers of a new continent of policy transfer". 22However, even if their questions are still fundamental for the definition of a research design and if their framework focused on some of the most important aspects of the phenomenon, such as the motivations for transfers, actors engaged in the process, the origin and elements of transfer, as well as the effects of policy transfer, in recent days there is a sort of consensus that their proposition is no longer sufficient.The framework portrays transfer processes as essentially linear, usually taking place from one government to another.Recent empirical evidences, nevertheless, are showing that these are complex movements, which involve different directions, as well as actions for translating or adapting the underlying meanings of public policies and the importance of their instruments.The so-called "sociology of public action" has been important to fulfill these questions.Besides that, authors such as Dolowitz and Marsh 23 have been reviewing and updating the heuristic contributions for the understanding of policy transfer, in order to bring us even further on the debate, as the reader will also see in Dolowitz' s article ("Transfer and Learning: One Coin Two Elements") published in this Dossier.
Among the diverse unfoldings of policy transfer approach stands out the work of Diane Stone, 24 who interprets such processes as much wider and more complex phenomena.Stone insists on the fact that transfers do not occur simply through the bilateral relations between two states, but they are processes that usually involve multiple agents, such as non-state actors, including international organizations, think tanks, ngos and private organisms.The most recent contribution for the debate is the collective piece from Hadjiisky, Pal and Walker, 25 that brings together a set of case studies that highlights the necessity of combining analysis that takes into account the way macroprocesses are related to microdynamics in the production of policy transfers.This work is also innovative as it is not restricted to Northern countries transfers, but also includes case studies from the Global South (in such a respect, see also Evans). 26nother research stream that is close to policy transfer studies considers the phenomena of policy "convergence" and the "europeanization" of public policies.Convergence studies were ongoing in the debate, especially along the 1990s, referring to a broader, regional or global process that points out the social tendency that governments eventually have to adopt similar institutional structures even in different contexts. 27Studies on the European Union transfers produced a distinguishable body of literature dedicated to analyze the process of "europeanization" of public policy in the regional integration context. 28In fact, transfers in this context are diverse and involve different government levels, several coordination and imposition dynamics leading to the so-called "harmonization" of norms and public administrative instruments.An emerging issue in this research agenda will now be to frame "transfer resistance", such as occurred with the Brexit contention. 29Both convergence and "europeanization" are processes [24] Stone, 2001; 2004.[25] Hadjiisky; Pal; Walker, 2017.
[28] Saurugger; Surel, 2006.[29] We owe this comment to Leslie Pal, who shared with us a manuscript on the issue of transfer resistance that will be published in the forthcoming book that is another result of the International Seminar on Policy Diffusion we have mentioned in a previous footnote.
that involve multiple transfers and a large number of countries.That is why for this kind of analysis the notion of diffusion may to be more appropriate, as it will be presented in the next section.

Policy diffusion
The use of the term "policy diffusion" is more frequent in the literature produced in the United States, which in general has a more specific definition of the phenomenon than the one we have adopted in this article.It is accepted by different authors that international diffusion of public policies "occurs when government policy decisions in a given country are systematically conditioned by prior choices made in other countries". 30In terms of scale, diffusion can be understood as the adoption of policies by a group of countries or governments, while transfer studies are mainly concerned with supposedly unidirectional movements that involve the displacement of policies towards one or a small set of governments.In different studies diffusion is understood as the adoption of a policy by a cluster of countries, that can be identified by their geographic proximity, such as Latin American states, or by a shared historical background, that can be political, institutional and/or ideological, such as the post-communist countries, or even by socio-economic similarities, as Organization for Economic Co-operation and Development' s (oecd) members or the brics (Brazil, Russia, India, China and South Africa) countries.Finally, policy diffusion processes can take the shape of "waves" that involve floats of policy adoptions and reflux of abandonment, as exemplified by democratization processes in countries that later eventually returned to authoritarian regimes. 31tudies on policy diffusion were influenced by the analytical approaches developed to understand the spread of innovations, an area whose main exponent is Everett Rogers. 32 For the author the diffusion of innovations is a "special type of communication, in that the messages are concerned with new ideas". 33In political science, an important share of the diffusion research is dedicated to the study of the impact of the American federalism on policy diffusion, with a strong tradition of quantitative methodologies.Jack Walker 34 is considered a pioneer in the area.Inspired by studies on innovations diffusion, Walker investigated the velocity and patterns of adoption of services and programs across municipalities in the United States.These studies are interested in understanding policy diffusion inside the borders of the State, at the subnational levels.Moreover, in public administration studies -that aim not only to produce research findings, but also to impact the public policesthe diffusion analysis often involves also the identification of "best practices" and their dissemination. 350] Simmons; Dobbin; Garrett,  2008, p. 7.   [31] Huntington, 1993; Weyland,  2014.[32] Rogers, 2003.
[35] The term dissemination can refer to a diffusion that is directed and managed by certain actors, while diffusion can also be the result of a spontaneous process (Rogers, 2003, p. 6).
International policy diffusion, however, has been analyzed by the means of different methodological approaches, qualitative, 36 quantitative 37 and mixed, 38 which can combine both qualitative and quantitative, as well as include experiments.Most recently, the literature about policy diffusion has been concerned with the micro-foundations of the phenomenon, which brought the subfield closer to transfer studies.
In their systematization of the field, Simmons, Graham, and Schipan 39 identified four main causal mechanisms that drive policy diffusion, namely: a) coercion, which emphasizes the impositionmore or less forced -for a government to adopt a policy; b) learning, when governments draw lessons from other places; c) competition, that is associated to the race for innovations that are supposedly able to make certain territories more attractive (e.g. for the private sector or tourism); and d) emulation or construction, that relates to the processes of socialization of public policy, to the legitimation of programs or even to the objective of producing good images of policies that help to foster diffusion in a policy community.These four mechanisms, however, are not the only operating along diffusion processes.Empirical evidences of cases from the South, e.g.participatory budgeting, have showed a more complex reality, one whose comprehension needs the inclusion of other mechanisms in order to access the process in its integrity. 40bserving the phenomenon through the lenses of policy diffusion scholarship means that the adoption of similar policies in different contexts is not necessarily a coordinated process, and that it does not imply only rational decisions taken with complete information.Recent studies insist on three other aspects that could also drive diffusion. 41Transnational collective action can influence policy adoption in certain countries.Policymakers, from their part, frequently adopt cognitive shortcuts in order to select the policies to be adopted, usually finding inspiration in neighboring countries.Besides that, there are also internal constrains that may definitely matter in the policymaking processes, such as the eventual non-acceptance of international models by voters.
For instance, the process of political transformation and democratization in post-communist countries, during the 1990s, involved the adoption of new rules and electoral models that eventually replaced traditional political agents and norms.Valerie Bunce and Sharon Wolchik 42 highlight an important aspect of policy diffusion in their investigation of transfer agents in different post-communist countries.For the authors, transnational democracy promoters, including members of local opposition, international activists and democracy promoters from the United States, were fundamental to the diffusion of electoral models in the cases they analyzed.Their study shows that there is an intensive engagement of individual and collective actors from different countries and institutions.These transfer agents were organized in transnational networks that promoted democracy as a supposedly shared common cause, and its concretization would occur by the implementation of preconceived models.
From another perspective, Kurt Weyland 43 observes the diffusion of social policies and sectional reforms in Latin American countries from a post-positivist perspective, stressing the abstract dimension of policy diffusion, more precisely the cognitive heuristics of the phenomenon.Weyland points out the fact that policymakers tend to be "attracted to certain foreign experiences for more 'accidental', logically arbitrary reasons, including geographic and temporary proximity". 44uccessful models developed in other countries call more attention and usually end up being adopted despite any rational cost and benefits calculations that could be made by policymakers.That is due to the fact that successful models frequently seem to be "simple" solutions that supposedly resolve different problems nearly at a glance.Cognitive shortcuts are operated in this process.Policymakers "reform projects under tremendous time pressure and cannot afford a comprehensive, proactive search for relevant information" and instead of analyzing all the costs and benefits, "they rely on inferential shortcuts to learn about models they can emulate, gain a sense of their performance, and quickly translate them into domestic reform proposals". 45nalyzing the diffusion of certain legislation in the sectors of health, family and employment in oecd countries, Katerina Linos 46 wanted to go beyond explanations that emphasized the determinant role of international bureaucrats and technocrats working in international organizations.Linos looks for responses for the domestic adoption of foreign laws in the internal fundaments of democracy, in particular on how voters can influence the "import" of international models.The argument is that "international norms and democracy are mutual reinforcing". 47The approval of voters is fundamental.According to this research, voters can collect information from laws in vigor in other countries and pressure governments for adoption.From their part, politicians have to shape laws so that they will be more attractive and appealing for voters, using foreign models that are familiar to citizens or that are in harmony with the international mainstream.
If the diffusion approaches are important to analyze waves or groups of policy adoptions, it is not uncommon that some analysts will consider the process as a linear movement in which policies spread throughout different countries in a sort of mechanical process.The idea of circulation is broader and allows overcoming this flaw, as we will detail in the next section.

Policy circulation
The French approach to public policy is singular in its fineness and complexity, which in part comes from its proximity to the field of sociology.In fact, in France public policy analysis is frequently associated with the "sociologie de l'action publique" (sociology of public action).A set of early contributions emerged in the French debate with a focus on institutional mimicry and on the adoption or rejection of foreign models, 48 as well as on the import and export of state institutions that occurred along African decolonization process, where in different countries institutions were built mirroring Western state structures. 49In contemporary debate, the term "circulation of public action instruments" is a current one. 50Influences of authors from sociology of sciences (e.g.Bruno Latour), elites (e.g.Pierre Bourdieu), as well as from philosophy (e.g.Michel Foucault), are readily perceived in policy circulation studies, which is a field that has been growing progressively in France.
The policy circulation approach emphasizes the abstract dimension of those processes, the role of ideas and the underlying meaning of policies.The notion of "policy instruments" is frequently evoked.The need to understand model transmission as a transnational movement is another central issue in the discussion, as well as the action of individuals or groups, that adapt and recode political information along the diffusion processes.Circulation can be seen as a vast and continuous movement of production of models, emission, appropriation and translation of their contents by multiple actors (individuals or collective, governmental or non-governmental), which have different power resources.This movement is perpetuated by the means of new emissions of the reframed public action instruments to other parts of the world.It is usually a circular process that also involves mutual learning and can go forwards and backwards from a place to another, in a sort of long spiral.
French studies attempt to produce a deconstruction of the mechanical, linear and rational view of policy diffusion and convergence processes. 51They are seen as fluid movements that do not follow a fragmented sequence of stages and that are not always unidirectional, from one country to another or one place to another.For Hassenteufel, 52 it is necessary to produce a sociology of the transnationalization of public policy, in a way that we can operate a deconstruction of a policy in order to establish a proper semantic genealogy.This is operated, e.g., by tracking the path followed by key terms of a policy and their multiple meanings, which will appear in a specific lexicon and political context.
The individual action is particularly important for the authors that are close to the sociology of elites, as Yves Dezalay and Bryant Garth. 53 In their work it was observed the circulation of Latin American elites [49] Badie, 1992.
[50] It's worth to highlight that the debate uses the term "transfer" as well (e.g. the special issue of Critique Internationale organized by Dumoulin and Saurugger, 2010).
[53] Dezalay; Garth, 2002.between their home countries and important universities in the United States.The analysis highlights the role of Latin American elites, such as the so-called "Chicago Boys", in the inclusion of United States models in the center of their home institutions, Chile being a wellknown example.According to Hassenteufel, 54 transfer operators can combine different power resources (knowledge, legitimacy, personal relations, money, etc.).Investigating the circulation of "participatory engineering devices", Mazeaud, Nonjon and Parizet, 55 for example, stress the role played by experts and specialized knowledge that area professionals detain in their understanding of the circulation of participatory democracy mechanisms.Moreover, another important feature of such a recent analytical approach is its emphasis on spaces where transfer operators act, such as the international and transnational "courts" that forge the backstage for policy circulation.These spaces constitute the transfer agents own habitat, with modalities of interaction and games that are characteristic to each group.
The relevance of so-called "policy instruments" has been noticed in the last years, after the contributions of Lascoumes and Le Galès. 56nspired mostly by classics such as Weber and Foucault, their proposal is to put into question the choice of "instruments of public action", often considered as a technical and neutral process, and their effects in the relationship between civil and political societies.The socalled "instrumentalization of public action" is defined by the "set of choices and the use of tools (technique, means of operation, devices), that allows to materialize and operationalize governmental action". 57he instruments of public action carry values, establish specific social relations, including power relations, between government and the citizens that will benefit from the public policy.The effects produced by the choice of an instrument or another, or even by a specific combination of instruments, can be very different.Instruments are embedded in specific movements of global circulation and recommendation by international institutions.The notion of instruments allows us to split up a public policy into different small parts.This concept is important to the understanding of circulation processes, inasmuch as transfer scholars argue that policies do not circulate as a compact brick.Different policy elements flow from a place to another, which can condense in various combinations, transforming the original model and producing different outcomes.
Finally, the idea of translation derives from the work of Bruno Latour, 58 who, by analyzing the production dynamics and the disputes behind scientific and technological innovations, calls the attention to the process of interpretation, by scientific communities, of the arguments and facts, according to their own interests, while shaping the construction of a science field.The "translation" of the different policy elements serves to mobilize instruments of public action in order to achieve specific results.The "best practices" divulgated by international organizations, for instance, are generally policies removed from specific contexts, assuming a more technical shape and translated as global prescriptions, so that they can be adapted to other countries, with different political cultures.Frequently, this process involves a transformation of the ideological content underlying a public policy. 59The abstract dimension of the object in circulation is infused in a constant process of translation, 60 that even if subtle can lead to very different political paths and public policies effects.
If the circulation approach has advanced on the microsociological dimension of policy diffusion, it also needs to be better connected to global trends, so that we can link macroprocesses (e.g.globalization) to microdynamics (e.g.elite networking) in the analysis of policy transfers.
As we have presented in the paragraphs, the field of studies that analyzes the international displacement of public policies is becoming progressively more specialized, acquiring sophistication and different analytical strategies.Recent and ongoing researches have been presenting meticulous details of the phenomenon.However, there is still lack of dialogue and coordination, generating a profusion of terminologies and approaches that are frequently overlapping, as transfer, diffusion and circulation "traditions" clearly show.The proposal of this section was to present the three most important ways of understanding the phenomenon.The argument is that analysis can be improved and overcome cleavages if diffusion is observed, on the one hand, in terms of the scale of the movement, and, on the other hand, combining different layers of analysis.These layers regard the macro-, meso-, and microdynamics of the processes under investigation.Therefore, more than concurrent, the three "traditional" approaches can be seen as intrinsically complementary, and they can be combined to produce new research, especially in Brazil, where the field is still taking its first steps, as it will be showed in the next section.

POLICY DIFFUSION RESEARCH IN BRAZIL: SCIENTOMETRIC EVIDENCES OF THE INCIPIENT CONSTITUTION OF A MULTIDISCIPLINARY SCHOLARSHIP
It is no wonder that policy diffusion research has become a growth industry, particularly in political science.In their bibliometric study, Graham, Shipan and Volden found out a "dramatic surge in interest in diffusion", 61 as more than half of the nearly 800 articles about policy diffusion that appeared in the top political science journals between 1958 and 2008 were published in the last decade of that period.The fact that these authors have stated that, despite that "dramatic surge", [59] Porto de Oliveira, 2016.
[61] Graham; Shipan; Volden, 2013,  p. 673."we are nowhere near having a systematic, general understanding of how diffusion works" 62 should not concern us here, even if it indicates that there is still a lot of analytical work to do.
According to Graham,Shipan and Volden,63 in the last decades all subfields of political science in the United States (American politics, comparative politics, international relations, and others) showed a steady increase in their in policy diffusion, "with only a handful of articles published in the 1960s, an average of two articles per subfield per year throughout the 1970s, a steady increase throughout the 1980s and 1990s […] and a dramatic increase in the 2000s, when the rate of publication across all subfields more than doubled, as compared to the 1990s". 64n Brazil, as we will show in this section, policy diffusion research is not yet a grown industry, but there are strong evidences of the increasing interest of scholars from different fields.In the next paragraphs we will not try to explain such a delay, which obviously does not mean that it is a less important issue in the country.As we intend only to briefly portray the state of the discipline in Brazil, we will present a survey of the related academic articles, dissertations and theses and the upsurge of the question in scholarly fora.Before we present our survey, however, it is probably relevant to recall that, in Latin American scholarship, Brazil, where the field is still so embryonic, is not an exception, as it can easily be proven by a quick consultation of Redalyc website (Red de Revistas Científicas de América Latina y el Caribe, España y Portugal) and Google Scholar.
If most Brazilian scholarly work on policy diffusion dates from the second decade of xxi century, as we will detail ahead, the pioneering work is arguably Hochman' s 1988 65 article dedicated to the impact of international policy diffusion and organizational learning in the constitution of social insurance in the country, from the 1940s to the 1960s.However, the first and only books on the topic came to be published in Brazil almost thirty years later.It was just in 2016 that Porto de Oliveira' s thesis was published as a book, entitled Embaixadores da participação: a difusão internacional do orçamento participativo, 66 and that the volume edited by Faria, Coêlho and Silva, 67 entitled Difusão de políticas públicas, came to the light.
When it comes to gauging the research results published as academic articles, scielo (Scientific Electronic Library Online) website is a useful and reliable source, even if it is not an exhaustive one, as not all Brazilian academic journals are included.Note, however, that scielo journals usually are the most important in their fields and that it encompass all areas of knowledge.The search results presented in Table 1 are revealing of the fragility of the policy diffusion field in Brazil.
If 32 articles is definitely not an impressive figure, the total number of articles on policy diffusion becomes much smaller when we exclude the repeated ones and the articles that were published in foreign journals that are also available in scielo.Making these exclusions, we end up with only thirteen articles, including two that were published by foreign authors in Brazilian journals.Eleven of these thirteen articles were published between 2011 and 2016, which shows how recent the Brazilian scholarship on policy diffusion is.Nevertheless, a classification of the journals that published those thirteen articles by subject reveals the multidisciplinary character of this small academic production.Those thirteen articles were published by journals from the following areas: social sciences (4); administration (3); international relations (2); health sciences (2); urbanism (1); and information science (1).One should also notice that only four of them seem to be dedicated to intrafederation policy diffusion.
Capes is an agency of the Brazilian federal government, linked to the Ministry of Education, that is responsible for promoting, regulating and evaluating all Brazilian post-graduate programs.Table 2 shows the results of the search made using the same terms that we used to survey the published academic articles.
The figures shown in Table 2 are not impressive either.If 88 theses and dissertations (t&ds) must be considered a low number, given the social and political relevance of policy diffusion and its pervasiveness, the total number of t&ds on policy diffusion becomes much smaller when the repeated ones are excluded: 48.Thirty-eight of these 48 t&ds were presented between 2011 and 2015, which again shows how incipient the field still is in Brazil.Notice also no t&d on the subject was presented before 2000, while just one was finished between 2001 and 2005, and nine were presented in the period 2006-2010.
One should also pay attention to the fact that no t&d and no published article seem to employ the terms "circulação de políticas TABLE 2 Brazilian theses and dissertations on policy diffusion (All areas -Capes Database -1987-2016)   Terms used in the Search All indexes ("Todos os índices") Number of theses and dissertations "Difusão de políticas públicas" (public policy diffusion) 14 "Difusão de políticas" (policy diffusion) 23 "Transferência de políticas públicas" (public policy transfer) 4 "Transferência de políticas" (policy transfer) 11 "Circulação de políticas públicas" (public policy circulation) 0 "Circulação de políticas" (policy circulation) 0 "Disseminação de políticas públicas" (public policy dissemination) 6 "Disseminação de políticas" (policy dissemination) 11 públicas" (public policy circulation) or "circulação de políticas" (policy circulation), which probably indicates the small influence of such a French research "tradition" in Brazil.The multidisciplinary character of the field in the country is once again evident when we observe the knowledge areas of the postgraduate programs in question.If here the diversity is greater when compared to the one that came out of our survey of the articles and journals (nine areas versus six), the most common subject areas do not change much.Those 48 t&ds were developed in post-graduate programs dedicated to the following areas: social sciences (14, including public policy programs); administration (11, including public management); health sciences (10); international relations (3); urbanism (3); education (3); economy (2); history (1); and regional development (1).
With regard to the theses and dissertations, as our n is a bit bigger, it may be worth noting a few patterns related to the use of the distinct terms or concepts.Tabulations that are not presented here, due to lack of space, have shown that: a) t&ds from the social sciences tend to employ mainly the term "policy diffusion"; b) those from administration and international relations use with roughly the same frequency three of the four terms used in the search ("policy diffusion", "transfer", and "dissemination"."Policy circulation" appeared in no search, as we have already noticed); and c) t&ds from the health sciences most frequently use the term "policy transfer".Notice that those figures should be understood not only as the product of the various research "traditions" and the way they frame policy diffusion phenomena, but also as probably induced by the pioneering studies in the different areas.
When it comes to the way research on policy diffusion appears in Brazilian academic fora, our survey was inevitably narrower, as we were able to study only part, though a very important one, of the larger social science field, that here includes international relations.Besides that, our survey was also restricted to verifying frequencies, and not approaches or contents of any kind.What we will present below results from the analysis of the official programs of the meetings that are regularly held by the following Brazilian academic associations, which are the most relevant in their fields: 1) Associação Nacional de Pós-Graduação e Pesquisa em Ciências Sociais (Anpocs) (National Association of Graduate Studies and Research in Social Sciences); 2) Associação Brasileira de Ciência Política (abcp) (Brazilian Political Science Association); and 3) Associação Brasileira de Relações Internacionais (Abri) (Brazilian International Relations Association). 68Figure 1 presents the number of expositions of any kind related to policy diffusion that were made during the meetings of the three professional associations.
[68] The printed or the online versions of the meetings' programs were thoroughly searched and all kinds of presentation related to policy diffusion were counted.We have scrutinized: 1) all Anpocs annual meetings from 2003 to 2016, i.e., from the 27th to the 40th; 2) all ABCP biannual meetings since the firs , which was held in 1998 (with the exception of the 5th, that happened in 2006); 3 a) all fiv Abri biannual meetings since the firs one, that happened in 2007; and finally, 3 b) all three regional meetings, called "seminars", that Abri organized biannually, since the firs , starting in 2012.The classifiction was made only on the basis of the presentation title, which is certainly a fragility.Contrary to what happened in the case of the published articles and of the theses and dissertations, when the above specifie terms were searched using the available automatic mechanisms, in the case of the academic meetings the authors of the present paper had to classify the presentations themselves.As the criteria adopted were not particularly strict, it is quite probable that some of the selected presentations were in fact more related to policy internationalization than to policy diffusion.
Three general patterns may be observed in Figure 1: a) with the exception of 2014 abcp meeting, the number of presentations on policy diffusion has not been particularly remarkable; b) a great internal variation in all three associations; and c) an overall pattern of increasing presence of policy diffusion as a research question.
Even if the empirical study that we have just presented in this second section of the paper has not focused on the crucial issues of theory and methods, our results seem to be relevant.The evidence that we were able to gather and that was presented in the preceding pages clearly shows how incipient the policy diffusion field still is in Brazil, but it also shows its multidisciplinary constitution and that the question is gaining momentum among Brazilian scholars.The time is ripe for that increasing centrality, given the growing importance of policy diffusion in Brazilian governmental practices, both at the federal and the subnational levels, and the need of Brazilian policy sciences, that are now quite well established in the country, 69 to take one step further, in the direction of recognizing and fighting against their pronounced methodological nationalism.After all, as Weyland had already noted, "policy making in contemporary Latin America is not confined to a national scale, but commonly takes foreign inputs into consideration".[70] Weyland, 2004, p. 257.

FIGuRE 1 Number of presentations on policy diffusion in selected Brazilian academic associations
Source: own elaboration from officia information provided by the associations.fusion research", as stated by Graham, Shipan and Volden, 71 which certainly should concern Brazilian and Latin American political science and correlated disciplines.Nevertheless, even if methodological nationalism has been increasingly recognized and criticized everywhere, at the regional level, but not only, there are still some important shortcomings to overcome, if the intention is to develop such a research agenda in our countries.Scholars are still scattered and divided, and the lack of dialogue among authors is quite evident, as well as the terminological disrupt and the disputes among analytical schools.
The cleavages and competitions among ways to understand the phenomenon are not very productive when they are used to demarcate intellectual territory or reinforce small feuds of research, instead of the expected advance on the collective comprehension of policy diffusion.However, the research questions, methodological strategies and arguments overlap different research "traditions" on the field, as it was clear in the discussion presented previously, in our first section.More integration among authors and approaches is definitely needed.More than that, it is important to produce complementarity in researches, in a way that we can advance together an overall understanding of those complex and increasingly relevant processes.
As we have already seen, an important barrier to such an expected development regards the terminological diversity to refer to policy diffusion, which confuses analysts.In fact, sometimes terms such as policy "transfer", "diffusion" and "circulation" are used as synonyms, but many other times they are employed as hypernyms.They are used to indicate similar phenomena, which are not always identical. 72Elsewhere, we have argued that a form to surpass this barrier is to simplify the narrative on policy displacements and conceive their flows according to the scale of the movement. 73In this sense, our argument proposes that the term "transfer" can be used to talk about a specific displacement of a policy from one jurisdiction to another, while "diffusion" is referred to a collective adoption of a public policy, and "circulation" is a longer and vaster flow, in time and space, that can also imply back and forward policy movements.Such a usage is based, as we have already seen, on the different approaches developed in the field of policy diffusion, which are connected to regional research "traditions".
Among the most urging limits of the field today is the fact that the vast majority of the studies have been produced by authors from Northern institutions investigating policy diffusion within those countries or from them to Southern states.Despite the surge of policy innovation in countries from the South -such as Brazil -and their circulation towards both Southern and Northern countries, we still lack adequate concepts, analytical strategies and approaches capable of capturing and better explaining these phenomena.
[72] Elkins; Simmons (2005), p. 37, distinguished a set of terms associated to diffusion as a result (e.g.isomorphism, convergence, waves) and others connected to diffusion as a process (e.g.contagion, demonstration effect, mimicry, Galton' s Problem, cascades).In another study, Graham, Shipan  and Volden (2013) found a total of 104 terms in the literature that have been used to indicate how diffusion occurs and its outcomes.
The articles that were selected for publication in the present dossier, published by Novos Estudos, will certainly contribute to the overcoming of those deficiencies.David Dolowitz' s paper, after reviewing the state of transfer literature, questions the way policy transfer, learning and knowledge utilization have been understood and asks if a learning link can lead to a better policy transfer.Jennifer Constantine and Alex Shankland' s article deals with policy transfer in the realm of international development cooperation, stressing the new role played by developing countries and questioning the possibilities of mutual learning.Sergio Montero' s paper focus on a single process of intercity policy transfer and learning, between Bogotá, Colombia, and Guadalajara, Mexico, but highlights pretty relevant and innovative analytical questions, such as the role played by "persuasive practitioners" and the "art and dangers of narrative simplification".Finally, but not less important, Patrick Hassenteufel, Daniel Benamouzig, and Magali Robelet, advancing the concept of "policy translation", question the limits of the policy convergence literature in their comparison of the adoption of evidence-based health agencies in three European countries (United Kingdom, France, and Germany).
Osmany Porto de Oliveira is a professor and researcher at unifesp.He is the author of Embaixadores da participação: a difusão internacional do Orçamento Participativo (São Paulo, Annablume, 2016), among other books.Carlos Aurélio Pimenta de Faria is a professor and researcher at puc Minas.He has edited Difusão de políticas públicas (Santo André, Ed. ufabc, 2016), among other books.

FINAL
REMARKS: CHALLENGES FOR AN EMERGING FIELD IN BRAZILNowadays we are witnessing, particularly in Europe and in the United States of America, a sort of accelerated "diffusion of policy dif-[69] Marques;Faria, 2013.