THE METROPOLITAN DIMENSION OF HOUSING POLICY

The last decade (2000-2010) has been characterized by the resumption of the protagonism of housing policy in Brazil, with the consolidation of instruments and policies that have formed a consistent framework to address the issue in the country. However, despite the concentration of the problem in metropolitan areas, the policy and the national housing plan have essentially adopted municipal strategies. Given this context, how can the obstacles and limits to the regional articulation of municipal housing policies in metropolitan regions be discussed? We start from the assumption that, while regional articulation is inevitable, the planning and management arrangements and instruments available have structural limitations that make the integration of municipal policies unfeasible. To discuss this issue, a previous question must be answered: how to investigate the factors that condition the integration of housing policies in metropolitan regions? This is the question that this article seeks to answer, resulting in the proposition of an analytical investigative structure that underlies future studies in this field.


INTRODUCTION
Housing policy in Brazil has taken on a prominent role in the last decade, both in academia and public management, and is characterized by a great paradox.At the same time as the progressive revaluing of the subject -with the approval of the City Statute (2001), the National Policy ( 2004), the National System (2005) and the National Housing Plan (2008), there is a weakening of this structure in favor of a policy of accelerated economic growth.Indeed, while on the one hand the quantitative results of the My Home My Life Program (Programa Minha Casa Minha Vida) suggest a positive outlook, the qualitative results clarify an obscure side.When delegating the protagonism of the housing policy to the market, a pattern of social-territorial segregation and urban and environmental precariousness was reproduced that is similar to, or even worse than, the BNH (BONDUKI, 2008;ROLNIK, KLINK, 2011;SHIMBO, 2010).
Recent studies have demonstrated these limitations, with a recurrent lack of compatibility with land policy and the lack of integration between municipal policies in metropolitan areas (ROYER, 2009;CARDOSO, 2013;DENALDI, KLINK, SOUZA, 2010;GONÇALVES, 2011;MARICA-TO, 2014).In this context, when more than half of the housing deficit (FJP, 2013) and about 80% of the country's irregular occupations (IBGE, 2010a) are concentrated in metropolitan areas, it is imperative to question the reasons why the articulation between municipal housing policies have not received attention.
Several authors point to the need for federal articulation in the implementation of housing policy.In a context of asymmetries between metropolitan municipalities, with a concentration of economic dynamics and socio-spatial segregation, one can frequently observe the "export" of the housing deficit from the core municipalities (with high land prices) to the poorest municipalities, where the price of the land is lower and the capacity to control the use and occupation of the land is reduced, creating the ideal conditions for a peripheral urbanization characterized by illegality and the lack of infrastructure (FIX, PEREIRA, 2013;GONÇALVES, 2011;ROYER, 2013, MA-RICATO, 2011;BONDUKI, 2013).
Municipalities with lower revenues and significant growth rates are forced to invest heavily not only in the areas of housing and infrastructure but also in urban mobility, collective transportation, sanitation, solid waste management and social facilities.The panorama becomes more complex with each variable added: for example, when considering the distance between these settlements and employment, mostly located in the core municipality, the impact of the commuter movements on the road network creates an additional challenge.
Housing promotion has a close relationship with land use and occupation, which in turn is directly related to the directing of urban growth and the environmental preservation of certain areas, such as water sources.The interdependence between municipalities and between their sectorial policies is clear, demanding intergovernmental coordination and federal cooperation arrangements.
In order to discuss this issue, one must also bring to the fore the debate on "the (non)governability of metropolises" (RIBEIRO, SANTOS JR., 2010).Despite its economic, political and social importance, there is a picture of institutional fragmentation and political disinterest.As a symbol of the institutionalization of integrated planning and management, metropolitan regions were emptied (largely by post-88 municipalism) and have lost character due to the indiscriminate inclusion of municipalities that have no relationship with the metropolitan fact (BALBIM et al., 2011;FIRKOWSKI, 2013, KLINK, 2013).Ribeiro (2004aRibeiro ( , 2004b) contributes to the above, pointing to the nonexistence of effective policies aimed at the development of metropolitan areas.In the words of the author, "urban policies are now strongly intra-urban, sectoral and local" (RIBEIRO, 2004a, p.11), while the metropolises "are at the same time a fundamental scale of the Brazilian social question and orphans of political interest " (RIBEIRO, 2004b, page 23, author's italics).In short, there is a gap between the func-www.mercator.ufc.br The Metropolitan Dimension of Housing Policy tional territory (metropolitan space) and the institutional territory (arrangements for metropolitan / intermunicipal management).
These findings indicate the urgent need to articulate the different spheres of government and sectoral areas, in order to combat the fragmentation of policies and make performance feasible on the metropolitan scale.Given the above, how is it possible to think about the integration of public housing policies in metropolitan regions?
The adjective bleak could be used to describe this scenario, giving a dramatic tone to the contextualization.Thus, this would frame the reflection in the hegemonic format of current academic production, delineating a metropolitan leviathan for which no intervention attempt seems to succeed.In fact, it is necessary to recognize the multiplicity and complexity of metropolitan problems and it is not possible to evade discussing them.
However, starting with Ultramari's (2013) challenge, it is necessary to get rid of the vice of approaching this question as an insoluble enigma.For the author, the cumulative urban tragedy experienced by Brazil throughout the 1970s, 1980s and 1990s has familiarized us with a reality of uncontrolled urban growth, a widespread lack of infrastructure and the peripherization of poverty, "in which no optimism seems possible," overshadowing the perception that "new things are happening" (ibid, 167).
Despite the immense challenges, there should be a focus on policy developments as well as on recent experiences.In relation to the first, it is imperative to highlight the Statute of the Metropolis, Federal Law no.13.809 / 2015, which brings legal advances on several fronts, especially in relation to the formation of new metropolitan regions, the elaboration of Integrated Development Plans compatible with municipal master plans and the implementation of inter-federative governance.Still in the legal sphere, it is important to highlight the relevance of the Federal Supreme Court's Judgment regarding sanitation services in Rio de Janeiro's MR, opening promising prospects for the effective recognition of the importance of integrated management in metropolitan spaces.
Regarding the experiences underway, recent cases have indicated possible alternatives for the management of metropolitan spaces, based on the coexistence of compulsory and voluntary arrangements, where public agents' agreements are the result of multiple instances.
Particularly since the mid-2000s, there has been a proliferation of institutional arrangements for metropolitan governance, articulating different actors and scales, usually from a monosectorial approach.These arrangements, with a greater emphasis on inter-municipal consortia, are a mechanism of ascendant regional organization, allowing for cooperation based on the principle of juridical equality, in which the characteristic constraint of traditional metropolitan arrangements is avoided (CRUZ, 2002;DAVIDOVICH, 2004;LEAL, 2007;KLINK, 2010a;AZEVEDO, GUIA, 2004).In addressing the recent debates surrounding this issue, Klink (2009) inserts the public consortia into an approach defined as regionalized municipalism, in which the consortium represents "an embryo of a new institutional model for metropolitan governance [...] with broader arrangements of inter--federative collaboration, always driven by the autonomous will of the municipalities" (ibid, 424).
However, although voluntary arrangements emerge as an alternative management of the metropolitan territory, its constitution and functioning are not free of challenges and impasses.Major obstacles to such cooperation include: the absence of regional identity, hindering the construction of a "metropolitan consciousness" of the territory's management; the game of political interests, which hampers the formation of a commitment to build a metropolitan project of inter-municipal cooperation; the current legal and institutional structure, responsible for setting obstacles to metropolitan management and establishing a war between the municipalities to attract investments; and the absence of resources for intermunicipal integration (selective incentives), a key element in promoting articulation between the municipalities (MACHADO, 2009;GARSON, 2009;KLINK, 2009;LEAL, 2007;NASCIMENTO NETO, 2013;REZENDE, 2010;LINHARES, 2011;ROCHA, FARIA, 2004;ROLNIK, SOMEKH, 2004;ROJAS, 2010;RIBEIRO, 2004;BALBIM et al, 2011).The latter aspect, in particular, plays a central role in this discussion.Named by Abrucio, Sano and Sydow (2010) as the design of public policies, this strategy involves a federative model of articulation that presupposes a coordinating, inducing and financing role for the federal government, without the autonomy of subnational governments being withdrawn, both in the implementation and in the production of consensus on the policy.Thus, institutional design can "influence not only the greater articulation between levels of government, but also the consortium of federative entities" (ibid, 204), providing fundamental incentives for the formation of a collaborative environment and for stimulating cooperative action (GARSON, 2009;MAGALHÃES, 2010;SOUZA, 2004a).
In the field of urban policies, the sanitation sector is significantly relevant, particularly in the area of solid waste management, the focus of intense regulatory restructuring in recent years.The formulation of the sector's legal frameworks and the formatting of federal policies to foster consortium has led to a substantial increase in the number of inter-municipal solid waste management arrangements (NASCIMENTO NETO, 2013;PICININ, FORTINI, 2009).
At the same time, although it is a recent phenomenon, the emergence of inter-municipal arrangements in the transportation and urban mobility sector has also been observed (BORGES, DELGADO, 2011, BEST, 2011, URBS, 2012; PEDROSO, NETO, 2013).Initiatives were boosted by the approval of the National Urban Mobility Policy (Law 12587/2012), which foresees the articulation of federal entities to integrate public transport networks in its guidelines, delegating the responsibility for stimulating coordinated actions for common urban mobility policies in metropolitan regions to the Union.
Building on the thematic division adopted by the Ministry of Cities -Secretariats of (i) Transportation and Mobility, (ii) Environmental Sanitation and (iii) Housing 1 -it could be assumed that there is a similar effort by the federal government in the integration of housing policies.On the contrary, the opposite is the case, with incipient governmental action in the integration of housing policies and a small number of municipalities (75) involved in consortium actions, about 1% of the total (ARRETCHE et al, 2012).
Even though the National Housing System is decentralized and the executive role is delegated to the state and municipal governments, the federal government plays a fundamental role in the coordination, regulation and supervision of subnational policies, so that the priorities of state and municipal governments are deeply conditioned by federal policies (ARRETCHE et al, 2012;GONÇALVES, 2011).
Despite the importance of the above observation, it is observed that the main instruments of housing policy in Brazil disregard metropolitan integration within its scope: the National Housing Policy only has occasional recommendations to stimulate the creation and strengthening of regional and metropolitan institutional structures.The National Housing Plan limits the strategies dealing with the housing problem to the local scale and the Minha Casa Minha Vida Program adopts the municipality as the only planning unit.
Contributing to the discussion, Costa et al (2010, p.102) argue that due to the deep social--environmental interdependence between metropolitan municipalities and the transversal character of the housing issue, housing and land-use planning policies "should prioritize the integration of the metropolitan agenda, for policies related to the road and transportation system, environmental policies, and economic development and employment policies", making it essential to create mechanisms for integrating the guidelines contemplated by municipal housing policies, avoiding the fragmented action that may also become a competitor in attracting funds from other spheres of government.
In addition, although PlanHab (MCidades, 2009) recognizes the absence of regional instances of articulation of the actions of states and municipalities in metropolitan areas, it only recommends the (1) For analytical purposes, a similar approach was adopted to that of Soraggi and Soares (2011), considering the National Secretariat of Urban Programs as an intersectoral effort that, in general lines, is closely linked to the theme of housing.

www.mercator.ufc.br
The Metropolitan Dimension of Housing Policy integrated dimensioning of the land stock and housing deficit if the inability of each municipality to meet their needs in isolation is verified.Recent studies highlight this "municipalization" of housing policy, which exempts or relegates the metropolitan dimension to second place (BONDUKI, 2013;ROYER, 2013;SOUZA, 2009;COSTA et al, 2010;COSTA, MENDONÇA, 2010) reproducing the pattern of exclusionary urbanization and the peripherization of poverty characteristic of Latin American countries (CARDOSO, ARAÚJO, ARAGÃO, 2011).
Despite the existence of studies evidencing the limitations of public housing policies, at different levels of federalism, in promoting or establishing integration among metropolitan municipalities, Brazilian scientific production lacks research on the factors that generate obstacles and constraints to the inter-municipal articulation in housing (NASCIMENTO NETO, MOREIRA, 2014).Arretche et al (2012) corroborate the above, stating that, Little is known about how Brazilian municipalities cooperate in the provision of housing services.It is plausible to assume that, given the costs of the housing commodity, few Brazilian municipalities would have sufficient resources to implement large-scale housing programs and therefore depend on higher levels of government to do so.However, to date, these relationships and their importance to housing supply have been examined very little (ARRETCHE et al, 2012, page 132).
Given this context, one question emerges: what are the real obstacles and limits to the regional articulation of municipal housing policies in metropolitan spaces?We believe that, while there is an inevitability of regional articulation, the planning and management arrangements and instruments available have structural limitations that make the integration of municipal policies unfeasible.In other words, between the discourse and the practice of integration in housing policy, there is an insurmountable gap within the existing legal and institutional frameworks.
This finding requires in-depth studies to corroborate it as a solid hypothesis.To discuss this issue, a previous question must be answered: how to investigate the factors that condition the integration of housing policies in metropolitan regions?
In previous research, we have already analyzed the panorama of housing production in the contemporary academic setting (NASCIMENTO NETO, MOREIRA, 2014) and discussed the importance of theoretical modeling for public housing policy analysis (NASCIMENTO NETO et al, 2015).From these studies it was possible to identify the lack of research that identifies the obstacles to inter-municipal cooperation in the field of social interest housing.
It is this question that seeks to be discussed here.
In an exploratory way -and based on the extensive theoretical framework developed in previous works (see, for example, NASCIMENTO NETO, 2015) -this article proposes the construction of an analytical research framework.Its purpose is to act as a "pair of lenses", making field research possible, guiding the initial elements to be researched, the interviews to be carried out and the questions to be asked.This analytical research framework, far from embedding research within a statistical model of variables (in positivist terms), seeks to act as a starting point.Transporting the epistemological discussion carried out by Thiollent (2003) in relation to action-research, it is a scientific assumption that guides the investigation without disregarding the need to analyze the data within the context in which the phenomenon is found.
In the area of policy analysis, these theoretical structures are a fundamental element, allowing the representation of possible explanatory links that guide empirical research (FREY, 2000).Dye (2009Dye ( , 2011) ) corroborates the above, conceptualizing it as an abstraction of the real world in order to synthesize it in its essence, seeking to understand what is effectively significant in a public policy, relating constructs and suggesting relationships.
In this context, the efforts of Carlile and Christensen ( 2005) are noteworthy in systematizing the process of the construction of theories, structuring it within a scheme of interactive cycles.For the authors, the (cumulative) process of construction and validation of theoretical knowledge occurs www.mercator.ufc.brNASCIMENTO NETO, P.; MOREIRA, T.
in three stages -observation, categorization and association -and in deductive and deductive cycles that complement each other (Figure 1).Thus, the proposed analytical structure is the result of the development of the first turn around the pyramid, where the theoretical-conceptual reflections, gathered from the theoretical framework, allows analytical constructs to be established and the organization of their relationships in a referential frame.In future works, this analytical framework will be used as a "lens" to investigate the obstacles to inter-municipal cooperation in housing in targeted case studies.This is the second turn around the pyramid, verifying the adherence of the constructs to the reality and relevance of these variables in the integration of public housing policies in metropolitan regions.

BASIC THEORETICAL FRAMEWORKS: A BRIEF EXPLANATORY DESCRIPTION
In order to subsidize the development of the analytical research framework and to construct a plausible pattern of interrelationship between the proposed constructs, references were sought in widely recognized theoretical models of public policy analysis.These explanatory frameworks, designed to provide an understanding of the public policy formulation process (SOUZA, 2006), provide the researcher with clues as to why public decisions and their results remain stable, change, vary between different areas and impact their target audiences differently (JOHN, 2003).
A model is a representation of political life, an abstraction of the real world with a view to simplifying and understanding the fundamental elements of a public policy (DYE, 2009;SABATIER, 1999).As John (2003) points out, what distinguishes the different models is the basic assumption that each adopts in the analysis of the social processes involved.The result is a set of different analytical systems of human action, from which it is possible to identify, even within social complexity, relationships arising from a specific phenomenon.
The public policy frameworks used as a theoretical basis for consolidating the analytical framework of research are structured within a non-positivist view of the political process.In contrast to linear models of decision making -for which certain factors generate certain decisions -these models advocate a less structured and deterministic process.This does not imply the abandonment of technical-rational planning, but only the recognition of the preponderant role of the limited rationality of public managers and their political skills in increasing government performance.Based on an extensive bibliographical survey, three models were used as a basis for reflection, widely addressed in the field: (I) Multiple Streams, (II) Punctuated Equilibrium and (III) Advocacy Coalition At the same time, the impact of the game of political interests is inherent to this subsystem, which may hinder the formation of a commitment to construct a metropolitan project of inter--municipal cooperation.Particularly in the Brazilian case, where the federal system is characterized by the existence of three autonomous levels, government relationships are characterized by a high complexity (ROLHA, 2009;ROCHA, FARIA, 2004;ROLNIK, 2009).
As political authorities are elected at all three levels of government, these relationships are of a singular complexity, since they also involve a political-electoral logic that often ends up guiding the various possible combinations of vertical-horizontal relationships.It is important to note that, with the process of re-democratization, mayors gain a not inconsiderable power in the Federation, having fundamental support from the political trajectory of important actors, interested in both legislative positions and positions in other levels of the Executive, since mayors have privileged access to agents of the local electorate (ROCHA, FARIA, 2004, pp. 79-80).
Finally, it is important to emphasize the dialectical interaction between the policy construct and the construct of the recognition of the problem.The inclusion of certain themes in the political agenda can originate from their broad social recognition, while the actors involved in the political process can influence the media and society to turn their attention to a particular theme.
The interrelationship of these constructs results from the conjunction of elements of the three theoretical frameworks of policy analysis addressed in item 2. In all these frameworks, the concept of limited rationality is a basic presupposition, arguing that public managers are subject to imperfection and a restricted capacity of information processing, a short time for decision making and the influence of their own interests and of those groups in which they are involved.
From the Advocacy Coalition Framework (SABATIER, JENKINS-SMITH, 2007), the proposed structure incorporates the conception of the structuring of the political process in interrelated subsystems that generate restrictions and incentives for the implementation of a given public policy.In this context, the reflections from the Punctuated Equilibrium Theory (BAUMGARTNER, JONES, 2012) are complementary, adopting the understanding that public policies tend to remain stable for long periods, interrupted by short periods of instability when significant changes take place.Finally, based on the Multiple Streams framework (KINGDON, 1995), the proposed analytical framework adopts the concept of the three streams -problems, policy and political dynamics -, advocating that opportunities for significant changes in the political agenda tend to occur when there is a convergence between them, creating a policy window.
In summary, the proposed analytical research framework relates the issues that are believed to be preponderant to regional integration in housing policy among metropolitan municipalities (Figure 2).Another assumption adopted is that these five constructs interact in interdependent streams, which relate to each other. 4In this context, changes would tend to take place at times of convergence between streams, where: (a) in the problem stream, there is a clear recognition of the demand (in this case, the housing problem in the metropolitan territory); (b) in the policy stream, there is an institutional legal framework that allows the integration of public policies and inter-municipal cooperation, reinforced by the presence of a cooperative environment between the municipalities and the existence of selective incentives; (c) in the political stream, there is a political environment conducive to cooperation.
It is essential to emphasize the interpretative nature of the proposed relational structure.As stated earlier in this chapter, its role is to be a "lens", a starting point to guide the analysis and investigation of the case study, with a view to identifying obstacles to the integration of MR housing policies.Thus, far from being a statistical correlation of variables (in positivist terms), the analytical research framework proposed in this chapter aims to synthesize the constructs and their possible relationships according to this researcher's understanding.
This initial analytical framework conforms to a scientific assumption, which guides the investigation without disregarding the need to analyze the data within the context in which the phenomenon is found.Returning to Brandão's (2004) broader discussion of the theoretical formulations of urban-regional development, we must now delve into the concrete and historical to capture the (4) Although the graphic representation expresses a watertight ordering, the fluid nature of these streams is defended, in which recurrent overlapping occurs.structures, dynamics and central relationships of the phenomenon, with the challenge of retaining the general determinations and differentiating them from contextual particularities.This is the purpose of our future work.

FINAL CONSIDERATIONS
The advances of social housing policy of the last decades are indisputable, as well as the counter-claims generated by a program that has been disconnected from its generating policy.Unlike the 1970s and 1980s, where every effort was made to tackle the exponential problems of the urban demographic explosion of the 1990s, where atomized alternatives for intervention were developed from the local scale, and from the 2000s, with the juxtaposition of great progress with the persistence of old mistakes, the second decade of the twenty-first century brings to the fore an agenda of new discussions for housing policy in Brazil.
Although the issues of this new research and public management agenda involve obstacles that historically have always been present, their relevance has changed.Issues such as financialization, the State-market relationship and the integration of housing policies in metropolitan areas have taken on a prominent role, either through the perception of the shortcomings of past experiences or the proposition of improvements in more recent experiences.
This finding finds support in Kuhn's theory of scientific revolutions (2011), which explains the progressive character of knowledge.After a certain scientific revolution, other problems are discussed, whose existence was not even considered by the previous generation.In fact, few of the problems discussed in contemporary times go back to the historical beginning of the phenomenon to which they are linked.As the author argues, "previous generations have dealt with their own problems, with their own instruments and canons of resolution.And it was not just the problems that changed, but the entire network of facts and theories " (KUHN, 2011, p.180).
In this context, the elements discussed at the beginning of this article demonstrate the indispensability of the integration of housing policies among metropolitan municipalities.It is argued here that this question does not refer simply to an improvement of the policy, but to the correction of a failure from which a whole effort of attainment of the housing policy can be compromised.Several studies demonstrate the impact of this lack of articulation in the implementation of the Minha Casa Minha Vida Program.
Given the above, how to investigate the factors that condition the integration of housing policies in Brazilian metropolitan regions?
It is on this issue that this article sought to advance, proposing an analytical research structure that could support future studies in the area.The theme is complex and multifaceted, requiring an analytical structure to contemplate the interpretative nature of political relationships, involving both public and private actors.According to Souza (2006a), in complex societies -such as contemporary societies -the state has only a relative "autonomy" to act, being influenced by external and internal pressures, as well as the formal and informal rules governing its institutions.This interaction between the various factors involved leads to the formation of certain capacities, which, in turn, create specific conditions for the implementation of public policies.
In this context, it is imperative to highlight the approval of the Statute of the Metropolis (Federal Law 13,089 / 2015), which, despite almost two years of existence (it was sanctioned in January 2015), has not so far provided clear perspectives of its impact on sectoral public policies.
Finally, it is necessary to point out that the empirical research carried out has restrictions due to its interpretative nature, that is, it is strongly linked to the researcher's evaluation, their perception of the phenomenon studied and, even if tangential, their worldview.However, it is understood www.mercator.ufc.brNASCIMENTO NETO, P.; MOREIRA, T.

Figure 1 -
Figure 1 -Theory Building Process Source: the author, based on Carlile and Christensen (2005).
www.mercator.ufc.brTheMetropolitan Dimension of Housing Policy

Figure 2 -
Figure 2 -Graphical representation of the analytical structure of the investigation of inter-municipal cooperation for housing policy in MRs www.mercator.ufc.brNASCIMENTO NETO, P.; MOREIRA, T.