Because of its early positive assessments, participatory budgeting (PB) has been praised by international policymakers and organizations. The Brazilian model has become an institutional blueprint, a “best practice” reproduced all around the world with more or less success (Ganuza and Baiocchi Reference Ganuza and Baiocchi2012; Goldfrank Reference Goldfrank2012; Oliveira Reference Oliveira2017; Wampler et al. Reference Wampler, McNulty and Touchton2018). In fact, more than 250 Brazilian municipalities adopted or ran one or another version of PB between 1989 and 2015. PB has also been diffused widely throughout Latin America and the rest of the world over the past decades (Porto de Oliverira Reference Oliveira2017), counting between three thousand and eighty thousand PB programs in the world, depending on the defining criteria (Sintomer et al. Reference Sintomer, Röcke and Herzberg2016; Wampler et al. Reference Wampler, McNulty and Touchton2018).
While scholars now wonder about the consequences of taking PB out of its original context and meanings through these diffusion processes (Baiocchi and Ganuza Reference Baiocchi and Ganuza2014), the evolution of the model itself and its survival (or not) in those Brazilian municipalities with a long tradition of PB remain largely overlooked in the literature. However, understanding how these programs evolved in their original context is crucial, as it informs how they operate in democratic governance processes over time.
In many cases across Brazil, PB has disappeared as those political leaders who supported it have left office. Recent data from the most comprehensive Brazilian PB census reveal that indeed, only 70 programs were still in operation in 2016, and it is very likely even fewer remain as of 2018 (Spada Reference Spada2014). In some cases, however, PB seems to have taken root and to have been institutionalized as a pillar of participatory architecture beyond partisan politics. Over the past 25 years, many PB programs have survived incoming political leaders because they have been institutionalized from below, through practice. This article therefore asks, in cases in which PB “stuck” as an institution of citizen participation in the municipal governance process, how and why has the institution adapted to or survived changes in the political and economic environment? How have these gradual changes affected PB’s practice over time?
Looking at the evolution of PB from a historical and institutional perspective, this article not only empirically addresses these questions but also contributes to the broader debates opened by historical institutionalism on mechanisms of gradual institutional change from within (Hacker et al. Reference Hacker, Pierson, Thelen, Mahoney and Thelen2015; Mahoney and Thelen Reference Mahoney, Thelen, Mahoney and Thelen2010). Building on these debates, the case of Belo Horizonte (Minas Gerais), which has had PB since 1993 and under several municipal administrations (PT or not), helps to understand better the transformation mechanisms operating behind PB’s stability. It shows that while political will is vital to PB’s survival in Brazil, that survival also depends on PB’s capacity to shift the balance of power among state and society actors in a way that makes PB endure and serve both actors’ interests in contexts of changing political and economic environments. In fact, the flagship case of Belo Horizonte illustrates that while the rhetoric of PB has remained central to the discourse of political elites, gradual and often hidden (at least in plain sight) changes have occurred from within the institution itself over time.
As this analysis will show, while seemingly stable, PB has adapted to the changing political and economic environments from within through combined processes of “hidden” institutional change: displacement, policy drift, conversion, and layering. While PB remained, in theory, an important institution in the participatory architecture of the city until recently, such changes also gradually transformed the PB institution from within, a process that has led to its unexpected deinstitutionalization.Footnote 1
The empirical analysis rests on a series of interviews conducted with PB participants, neighborhood activists, local bureaucrats in charge of the implementation of PB, and politicians who have been involved in the adoption and reform of PB over the past 20 years. These interviews were conducted at two distinct moments: first in 2008 and then in 2014. Several participants thereby were interviewed twice and provided a comparative assessment of PB’s transformations that greatly enriched the analysis. In order to get a better sense of where PB was in the most recent period, however, the research also used data from a survey conducted in Belo Horizonte between November and December 2014 with 467 participants during the assemblies of the 2015–16 PB cycle, segunda rodada.Footnote 2
A Gradual Theory of Institutional Change
In the past 30 years, Brazil has been often been described as a participatory democracy lab, with the creation and consolidation of a variety of innovative participatory mechanisms—policy councils, national conferences, and others—in various policy domains and at all levels of government.Footnote 3 Among these innovations figures participatory budgeting, one of the most praised vehicles for direct citizen participation in budgetary decisionmaking processes, which has now been around for more than 25 years. First adopted in the municipality of Porto Alegre in 1989 under Olivio Dutra’s Workers’ Party (Partido de los Trabalhadores, PT) government and closely associated with the election of leftist governments led by the PT (Goldfrank Reference Goldfrank and Baiocchi2003) in its early development phases, PB has since been adopted across Brazil.Footnote 4
In Pierson’s terms, PB is a public policy that can be defined as an institution, as it embodies rules that frame social interactions and create new organizations with state-backed decision or enforcement power for state authorities (Pierson Reference Pierson2003). PB indeed allows ordinary citizens and their delegates to deliberate, decide on the distribution of a portion of municipal resources, and collectively define policy priorities in various public spaces, such as assemblies, councils, and forums. It redefines the rules framing state-society interactions. The structure of the institution and the percentage of the municipal budget that is subject to citizen deliberation differ slightly from city to city, but some common characteristics are shared by PB experiments across the board.Footnote 5
While the institution is praised as a “best practice” around the world, recent panel data show that PB has had a varying rate of survival across Brazil, with a large number of municipalities abandoning the program over the last few years (Spada Reference Spada2014). It is indeed commonly understood that PB is less institutionalized than other constitutionally entrenched mechansims for participation (policy councils, for example), which makes them more prone to being politicized (Montambeault Reference Montambeault2018). This reality is generally overlooked by international appraisals of PB, and is often poorly explained by factors like electoral alternation and new incumbents’ lack of political will.
If changes in political environments are certainly important in understanding the decline of PB in Brazil, it is quite interesting to note that in similar circumstances, certain PB programs survive. Belo Horizonte is a good example of PB stability, as the municipality still had PB programs as of 2015, after more than 20 years and multiple political alternations (Wampler Reference Wampler2015).Footnote 6 This means that under certain conditions, PB can stick as an institution. In those cases, PB programs have contributed to redistributing power resources toward politically salient groups, and have thus reconfigured the balance of powers in competitive electoral settings in favor of social and political actors who benefit from their participation in such institutions.
This reconfiguration, when it happens, makes the program resilient to drastic political change and can ensure its survival. At the same time, PB is also an important space for political competition (Goldfrank and Schneider Reference Goldfrank and Schneider2006), and its survival also rests on its ability to adapt to new and changing political environments. However, this does not mean that the institution’s functioning and practices have not changed as new political and social actors have come to power and as new ideas about participation have emerged in the political arena. This is what the historically grounded analysis of Belo Horizonte uncovers, highlighting the intertwined and often hidden mechanisms at play that gradually change the institution, its practices, and its meaning in democratic governance processes and in citizens’ views over time.
When PB Sticks: Theories of Institutional Change
There is a long tradition in neoinstitutionalism that looks at institutions’ development as path-dependent. From this perspective, institutions are a set of rules (formal and informal) shaping social action and actors’ interests and thereby creating the conditions of their own reproduction over time. While such a process unfolds in distinct ways for historical, sociological, and rational choice neoinstitutionalists, they all tend to overlook the mechanisms by which institutions change. Institutions are indeed largely understood as things that “stick” over time, following a historically grounded path punctuated by episodes of change, which generally occur at “critical junctures,” during moments of history when actors face choices that may alter the institutional path.
This approach to institutions thus emphasizes the importance of “external shocks” of various kinds to prompt institutional change and, in most cases, radical institutional reconfigurations (Mahoney and Thelen Reference Mahoney, Thelen, Mahoney and Thelen2010, 1–2; Thelen Reference Thelen2009). Given this conception, recent cases of PB policy replacement or abandonment in Brazil are best explained by an external shock. In effect, the relatively scarce literature that has explored PB’s survival rate in Brazil and elsewhere is consistent with this premise, as it generally shows that electoral loss (and in particular, the loss of PB’s initial promoter, the PT) becomes a critical juncture for institutional change, and is therefore the main factor explaining PB’s abandonment in Brazil (Spada Reference Spada2014).
This is indeed what happened in São Paulo, when PT incumbent Marta Suplicy lost the mayoral election in 2004 (Hernández-Medina Reference Hernández-Medina2010); or in Recife in 2012, when the incoming Brazilian Socialist Party (PSB) mayor, Geraldo Julio, terminated the PT-created program and replaced it with the PSB’s (or his) own participatory policy, called Recife Participa. Because PB’s rules and budget are not regulated by municipal or federal legislation, its adoption, success, and survival are often thought to rely heavily on the support and goodwill of the executive (the Prefeitura) and of the bureaucrats who manage the program on the executive’s behalf (de Sousa Reference Sousa2011).
This explanation matches the reality of Brazilian PB to the extent that it includes only those cases in which a radical change has indeed occurred, like Recife or São Paulo. PB’s empirical reality in Brazil is, however, not black or white, and careful observations of cases of institutional continuity in contexts of environmental change reveal a more complex process of change (Thelen Reference Thelen2009). Strong political support for PB from leaders in power is indeed vital to the success of participatory democracy institutions over time (Goldfrank Reference Goldfrank2011). However, incoming political actors’ will to maintain the institution depends in large part on the perceived support it receives from opinion leaders and the electorate and on the consequent cost of radically changing the institution.
As a public policy, PB’s initial goal was exactly this: to shift the balance of power from a strong executive and its friends toward the traditionally excluded by empowering civil society organizations and movements (Abers Reference Abers2000; Baiocchi Reference Baiocchi2005; Wampler Reference Wampler2012). PB has, nevertheless, unevenly transformed state-society relationships across the country, as many comparative studies have revealed (Baiocchi et al. Reference Baiocchi, Heller and Silva2011; Lüchmann Reference Lüchmann2014; Montambeault Reference Montambeault2015; Wampler Reference Wampler2007). Consequently, political elites’ willingness to maintain participatory institutions implemented by a previous administration cannot be defined without considering the way the institution has reconfigured (or reaffirmed) power relationships in the municipality over time. In fact, the empowerment of an autonomous and collectively organized civil society able to formulate and press demands and to mobilize votes and popular support like that in Belo Horizonte is a good predictor of institutional stability beyond changes in actors and dominant ideas in the political arena. In such cases, reforming or replacing PB becomes more and more costly for political elites as the mechanism institutionalizes through its social foundations: empowered societal groups would lose too much from the disappearance of a mechanism in which they have invested and through which they have evolved.
More specifically, this complex reality calls for a more comprehensive approach to institutional change, one that accounts for the dynamic and mutually transformative character of both institutions and actors and for the idea that institutional stability can also hide significant changes. In fact, institutional stability can hide another type of transformation: one that is more gradual and incremental, but that can also leave deep marks on the institution, its practices, and its actors. Belo Horizonte PB’s stability over time provides excellent empirical evidence to build on emerging theories of institutional change and contribute to the development of a better understanding of how “institutions evolve” following the so-called critical juncture (Hacker Reference Hacker2005; Mahoney and Thelen Reference Mahoney, Thelen, Mahoney and Thelen2010; Streeck and Thelen Reference Streeck and Thelen2005). The historical analysis of BH’s PB trajectory indeed allows us to uncover alternative and yet often hidden processes of institutional transformation behind apparent stability over time.
Intertwined Changes: From Conversion to Deinstitutionalization
The set of theoretical tools developed by recent historical institutionalism allows for a better understanding of the mechanisms by which institutions incrementally change from within in relation to changes in the political, economic, or ideational spheres (Béland Reference Béland2007). More important, it uncovers processes of institutional change that are often hidden or at least camouflaged by apparent continuity but that deeply alter their purposes and effects (Hacker et al. Reference Hacker, Pierson, Thelen, Mahoney and Thelen2015) and might, in the long run, lead to their gradual deinstitutionalization, as in the case of PB.
As Streek and Thelen (Reference Streeck and Thelen2005) and later Mahoney and Thelen (Reference Mahoney, Thelen, Mahoney and Thelen2010) assert, institutions have internal properties that allow change to occur gradually and allow actors with sometimes opposing interests to undertake actions that will foster change through a variety of mechanisms. As a resource distribution mechanism, institutions not only shape interests, behaviors, and the definition of political will, but they also have a longer-term effect on the distribution of power among actors who can affect both the institution’s formal rules and the informal ways those rules are enacted and interpreted in practice. This is the case for PB, which combines two inherent characteristics allowing for hidden mechanisms of change to operate.
First, civil society actors involved in and empowered through PB have a strong tie to the institution and have a tendency to block change, increasing the costs of institutional change for politicians who have weak veto possibilities toward PB. Second, because PB is regulated at the level of the executive, the institution allows political actors and bureaucrats a great degree of discretion in the interpretation, implementation, and enforcement of PB rules. These two characteristics make PB institutions’ apparent formal stability very likely in contexts where, as in the case of Belo Horizonte, PB has strong popular support. At the same time, however, such characteristics make the institution very permeable to electoral and ideational changes. In fact, because PB’s rules are interpreted and enforced by the executive and its administration, political actors in power remain central to framing the way the institution works in practice and is integrated into the larger municipal governance system.
In Belo Horizonte, for example, local administrations have been able to leave their mark on PB and gradually transform the institution from within through a constant reinterpretation of the institution’s rules. For each PB cycle, the methodology and budget priorities that are included in the deliberative process are redefined and rediscussed by the administration, according to its own political and economic priorities, interests, and agenda.
How does change operate, then? Scholars of historical institutionalism have identified four main mechanisms that can induce gradual change to seemingly stable institutions: layering, displacement, conversion, and policy drift (Mahoney and Thelen Reference Mahoney, Thelen, Mahoney and Thelen2010). Layering is a process by which new elements are deliberately added to an existing institution, changing its structure from within, without replacing it. Displacement, on the other hand, is a situation in which a new model emerges and calls into question the legitimacy of the old one.
If gradual but rather visible processes of institutional change like layering and displacement have been quite thoroughly documented, Hacker et al. have recently called attention to what they call hidden transformative mechanisms for studying institutional stability: conversion and policy drift. Drift is a change that results from the inaction of actors as the context in which the institution evolves changes, while conversion occurs when political actors deliberately redirect the goals of a policy or an institution. As Hacker et al. rightfully note, these often hidden, incremental transformative processes are very consequential for the institution and its policy outcomes because they “provide a means by which political actors can alter governance without bearing responsibility with the broader public” (Hacker et al. Reference Hacker2013, abstract).
Conversion and policy drift are especially salient when looking at cases of institutional stability in which a direct action toward institutional change (displacement, layering) could, at least initially, be electorally costly for political elites facing intense competition. While rarely studied together by scholars of historical institutionalism, layering, displacement, conversion, and policy drift are not necessarily mutually exclusive mechanisms. They can be intertwined and can operate in different but sometimes complementary ways that allow political actors facing resistance to change and strong veto players to gradually induce profound and otherwise politically costly changes to a seemingly stable institution.
Analyzed through the lens of historical institutionalism and building on existing theories of gradual institutional change, the case of Belo Horizonte illustrates that conversion, policy drift, layering, and displacement are closely intertwined, reflecting the longer-term changes in the local political environment (the agents of change). This case also illustrates the importance of ideational environments (ideas) to understanding the direction of change, uncovering the intersections between PB’s institutional changes and the wider Brazilian dynamics of institutional innovation in the field of participatory democracy. In fact, this case study shows how institutions and ideas are mutually constitutive and transformative. Institutions constitute shared frames that guide action, but they are also shaped by the way actors interpret those rules, and by their ideas; that is, the set of beliefs that guide public action and shape political outcomes (Béland Reference Béland2007).
Over time, combined forces transform the institution from within, which can unexpectedly hollow it out. In fact—and this is what the case of PB in Belo Horizonte illustrates—these intertwined mechanisms significantly affected PB’s perceived legitimacy, and ultimately its political salience, until it was no longer a key site for citizen participation in the city. This case, therefore, is worth studying not only because it allows us to see the interplay between different incremental change mechanisms, but also because it promotes a better understanding of the path to deinstitutionalization from within, the result of incremental change that is largely overlooked and undertheorized in the literature on institutional change in comparative politics.
PB Institutional Change in Brazil: The Case of Belo Horizonte
Belo Horizonte, the capital of the state of Minas Gerais, is a very interesting test case for institutional change theories, as it is a case of PB survival over time and beyond electoral change (Luiz Lara Reference Luiz Lara2010). First adopted by a PT-led coalition in 1993, PB celebrated its 21st year of existence in 2014. The 2015–16 cycle was conducted by Marcio Lacerda’s administration, led by a coalition of the PSB, of which the PT has not been a part since the 2012 election.Footnote 7
While PB remains central to political discourse, a transformation has been observed over the years, starting even before the PT lost the election in 2012. Because PB’s legitimacy was established independently from the PT, it enjoyed a high level of support among autonomously organized civil society actors, and political and ideational change did not lead to radical institutional change. However, as we will see, this institutional stability hides an important shift that was generated endogenously and then unfolded gradually. In effect, a historical look at the institution, its meaning, and its practices shows that PB has gone through a gradual process of intertwined transformations that changed its purpose and outcomes. From a radical democratic reform that was at the core of the model of local governance, it has become a policy instrument among others and is gradually losing both political and social salience in Belo Horizonte.
PB’s Institutionalization and Stability in Belo Horizonte
It was in 1993 that the PT first won election in Belo Horizonte, a victory made possible by an alliance of PT candidate Patrus Ananias with the PSB. During his campaign, and inspired by the successes of Porto Alegre, Ananias proposed the adoption of PB in Belo Horizonte, for governance to become more “radically inclusive” and “democratic.” His administration conceived a PB model adapted to the realities of the city that was meant to “answer the repressed demands of the population as well as allow a better control over municipal finances” (Ananias Reference Ananias, de Azevedo and Fernandes2005, 40).
In the early 1990s, PB became a marker of the PT’s modo petista de governar (PT way of governing), which meant, among other things, direct democracy, ample popular participation, and social justice (inversão de prioridades). With this idea, the party made a formal electoral commitment to invest the institutional space with participatory practices: the state had the responsibility of guaranteeing the participation of the population in the decisionmaking processes, including those involving the budget (Bittar Reference Bittar1992). Institutionalized citizen participation was, therefore, an intrinsic part of the democratic petista project of governance.
PB soon became the party’s flagship participatory institution. While it created formal rules, it followed an inclusive and transformative conception of citizen participation that generally respected the spirit of PT’s original democratic and social justice principles (de Paiva Bezerra Reference Paiva Bezerra2015). As Patrus Ananias explained in an interview,
[Participation was] [o]ur preoccupation. … We had a very clear priority, more generally, that was the struggle against poverty. I mean our idea was to govern for everyone. … And we figured that one way to fight against poverty was to engage people, including the poor. … The empowerment of people, of the families was fundamental … [and] had a pedagogical dimension of training citizens, of constructing an effective citizenship. (Ananias Reference Ananias2014)
While a few participatory public policy councils already had been formed following the adoption of the constitution in 1988, PB was the main axis for encouraging and valuing citizen participation and democratic engagement among traditionally marginalized communities and beyond. With the limits and problems inherent in the construction of an emerging and innovative institution, PB was conceived as the backbone of the PT’s participatory program.
Overall, and unlike many other experiences in Brazil, Belo Horizonte’s PB is considered to have been a relatively positive and transformative democratic experience, especially in relational terms. If it has its inherent limits, PB has not only contributed to inverting redistribution patterns toward the traditionally marginalized (Touchton and Wampler Reference Touchton and Wampler2014), but has also mobilized an important number of participants over the years (see table 1). Even though it has not necessarily led to the emergence of new civil society associations (Nylen Reference Nylen2002) or of citizenship as agency (Montambeault Reference Montambeault2016), PB is argued to have contributed to transforming the nature of state-society relationships from clientelistic to democratic and empowering (Montambeault Reference Montambeault2015; Wampler Reference Wampler2015).
Table 1 Number of Participants in Regional PB Cycles per Year, Belo Horizonte, 1994–2012
Notes: Beginning in 1999–2000, PB was organized over a two-year cycle.
Source: Created by the author with data from the Prefeitura Municipal 2014.
As I have shown elsewhere (Montambeault Reference Montambeault2015), a model of democratic cooperation emerged and institutionalized over the years, as PB has indeed become a privileged space for social forces to emerge, organize, and mobilize collectively and autonomously from the political sphere. The institution and its practices have redistributed power among social and political actors in Belo Horizonte: from unequal and dependent (under clientelism), power relationships have become pluralistic and codependent as new actors have become empowered and legitimized through PB (under democratic cooperation).
While this dynamic can be seen as a positive outcome of the program, it is also important to understand the reasons for the institution’s stability and changes or renewal over time, especially in terms of changing political and ideational contexts. In fact, pro-PB groups, which have emerged from their PB actuation, have become strong veto players. They systematically try to block attempts to make formal changes to the PB program through pressures and voter mobilization against too overtly reforming candidates during electoral campaigns. As a member of the PB administration team explained in an interview, PB is a government program, but it has been appropriated by belohorizontinos and has become independent of any political party, even though originally associated with the PT.
If a mayor could come and have the courage to … I think he would create a very big conflict in the city of Belo Horizonte. Can you imagine, you have here involved from five hundred to six hundred people who meet monthly to discuss the city. So it’s not that simple. … It’s not a simple challenge for a mayor to abandon PB in Belo Horizonte. It would create so much confrontation, so many conflicts, that it’s not viable. (PB Administrator 1, 2014)
Afonso, PB delegate for many years and president of his neighborhood association, confirms this perception:
PB. … Here in Belo Horizonte, in Porto Alegre, in other cities of the ABC Paulista. … They already had governments administrated by both the PT and other parties and that supported PB. It’s consolidated already. And the candidate— in any new election—always loses [by] speaking of the end of PB. (Afonso 2014)
These PB-empowered groups and opinion leaders indeed have the capacity to mobilize against change, as representatives of larger groups of people who have built their leadership position in their communities. This is why, in the eyes of many municipal administrators, it would be extremely costly politically to terminate PB. Mônica explained: “they can’t terminate [PB] because politically, no one has the courage to terminate. Because it was a gain of the popular sector, actually. A lot of people made many gains through PB” (Mônica 2014). Speaking to this, another member of the PB administrative team added:
And these people [the mobilized] delegates have many more people behind them. Because they represent. They go to their communities, and their community asks, “so in the next PB, what are we going to ask for?” So they represent people beyond themselves, there are other people and leaders. (PB Administrator 2, 2014)
As opinion leaders and mobilization agents, community leaders empowered by PB have the capacity to challenge the legitimacy and authority of the municipal government and to erode electoral support for mayoral candidates who threaten the survival of “their” institution. Thus, in Belo Horizonte, PB has institutionalized from below, through pressures from the actors that most benefit from the institution and who want to maintain their position through their participation in PB. However, this does not mean that the institution has not changed over the years, through intertwined mechanisms of transformation that operated from within by an executive that retains a significant share of the power in the working of the institution.
Stability as Gradual Change: Conversion, Layering, Policy Drift, and Displacement in Belo Horizonte
As a public policy, participatory budgeting and the practices it engenders are highly permeable to politics and, more generally, to dominant ideas about citizen participation. Political leaders and parties in power changed partly as the result of the rise of the left in Belo Horizonte, a city with a strong conservative tradition. The electoral success of the PT—which already had promoted the implementation of PB— thus depended on its constant ability to negotiate political alliances with the PSB and other centrist political formations in order to become a catchall coalition able to get elected (Wampler Reference Wampler2007).
Even if the alliance between the PT and the PSB was never sealed as such by an official political pact, it remained relatively stable until 2012, when PSB mayor Marcio Lacerda rejected the traditional coalition before being elected for his second mandate. Despite these changes in the executive, citizen participation has remained a central feature of all mayoral discourse in Belo Horizonte. However, the way political leaders from both the PT and the PSB defined and emphasized the concept has changed over the years, affecting the way PB has been interpreted and enforced in practice, which has resulted in its gradual transformation over time.
At first, PB was implemented in Belo Horizonte as an annual process to which only a limited portion of the annual municipal budget was allocated. The deliberative bodies included the traditional regional assemblies, conducted in the nine regiões político-administrativas (RPAs, politico-administrative regions).Footnote 8 Since the first years of the policy implementation, however, PB has faced a significant challenge that affects the perception of compliance among the population as far as the institution is concerned: the completion of the approved projects and the delivery of the public works within a reasonable timeframe after the end of the PB cycle. Figure 1 shows that over the years, 1,536 projects were approved in PB assemblies throughout the entire city.Footnote 9 However, only 78 percent of these projects are complete today, and some of them have been lagging since 1999. The figure indeed demonstrates that many projects take a very long time to be executed and completed by the local administration. This discrepancy between the number of projects approved and those being concluded seems to have grown over the years, in part because the PB continues to approve about 100 to 125 projects per year while completing fewer of these projects over time.
Figure 1 Number of Projects Approved vs. Concluded, PB Belo Horizonte, 1994–2014
Presented as a way to avoid delays in project implementation, the first reinterpretation of PB rules happened in 1998 when the process went from annual to biennial. This reform was promoted by PBS candidate Célio de Castro, who implemented the reform on his election as mayor of Belo Horizonte (1997–2001). The regional process was thus reorganized over a two-year cycle that included two phases: project definition and project monitoring and implementation. The first phase is the one that includes the direct participation of citizens, organized in neighborhood groups or via the officially registered civil society organizations (CSOs) active in each community, especially the neighborhood associations. With these reforms, the municipal government hoped to be better able to achieve the general spirit of the program, which remained the same, explained ex-PT mayor Fernando Pimentel (2000–2008) in an interview: to facilitate the incorporation of citizen participation in the decisionmaking process over budget-related questions of urban development expenditures and priorities (Pimentel Reference Pimentel2008).
Presented by the incoming administration as a measure to increase PB’s efficiency and capacity to complete the works approved with fewer delays, this transformation was actually a form of conversion, which nonetheless raised doubts in the minds of many participants. During this period, the PB participation rate declined consistently, as participants were uncertain about the future of the policy (see table 1).
This reform indeed coincided with a change in the political environment: the leadership of the PT-PSB coalition in power passed to the PSB faction. However, the reform was not followed by a consequent increase in the total share of the municipal budget subject to citizen deliberations and votes. This aggravated the already existing problem of lack of financial resources, as less than 5 percent of the total budget was discussed in the PB process. For many early participants in PB, this reform was therefore the first observable manifestation of a deeper change in the practices of the institution that was contrary to the essence of PB at its origins: a change in the meaning of the concept of citizen participation. As Efigênia, a long-time activist and president of the neighborhood association in a vila of Centro-Sul, explained, speaking of the then-mayor’s decision to make PB biannual:
the municipality decided that they would be doing this plenary every two years, this was in 1998, and they just brought it to us, they approved it, brought it to us and presented it as if there were no other alternatives: “look, it will be like this and that’s it.” We did not get any way to discuss this with them. … The municipality approved this change the way they wanted it and did not respect people’s participation and opinions. Whatever change they think is appropriate they do, independently of the position [or] opinion of the people who constructed PB with them. (Efigênia 2014)
In reaction to this change, and starting in the early 2000s, community leaders from across the city joined their efforts and pressed the administration to create a citywide monitoring body that would work closely with the PB administration team to make sure the deadlines for public works delivery were respected and thus be the “eyes of the community” at the city level.
The return of a PT mayor to lead the PT-PSB coalition in 2002 coincided with another important event that changed the political environment for PB: the victory of PT’s founder, Luiz Ignacio Lula da Silva, in the presidential election. Mayor Pimentel thus came into office in the context of a changing PT, which was moving from a previous emphasis on empowered participation toward a more pragmatic and fragmented understanding of social participation (Baiocchi et al. Reference Baiocchi and Ganuza2013), seeking to build support beyond the poor and marginalized to include the rising middle class.Footnote 10 From the modo petista de governar, participation became a método de gestão (method of administration) during this period (Paiva Bezerra Reference Paiva Bezerra2015), an ideational change that has had important repercussions for the institutional evolution of PB in Belo Horizonte. In fact, the changes in the city’s demographics (the middle class) and in the party’s changing conception of participation at the national level contributed to the gradual decline of PB as the main participatory mechanism in the city.
At its origin, PB was meant to include the poor and marginalized, to empower the invisible majority. After 2000, however, the preoccupation of Belo Horizonte’s PT leaders changed, in line with the party’s shift: citizen participation was important, but it was no longer an empowering participation, a school of citizenship. Instead, it became more pragmatic and electorally driven, the challenge being to reach the middle classes, who were not inclined to participate and organize to formulate demands in time-consuming PB assemblies.
During Pimentel’s mandates, PB Regional continued in Belo Horizonte. The question of lagging delivery was still vivid, however, weakening the efficiency and credibility of the process among citizens. At this time, however, Pimentel’s administration decided neither to address this problem nor to allocate more funds in view of closing the growing gap between approved and concluded projects. In fact, the total share of the budget invested in PB during the Pimentel years decreased every year, increasing the delivery problem. However, Pimentel did expand the concept of what PB was as a participatory institution. In fact, in 2006, Pimentel’s team proposed to add a virtual dimension to the traditional PB Regional, which constitutes a process of institutional displacement.
It was thus, with these political and ideational changes, that Pimentel launched the first edition of PB Digital, a participatory process in which citizens were invited to vote electronically for a large infrastructure project among four or five possible projects proposed by the municipal government. Speaking of this first experience in an interview he gave me in 2008, Pimentel said,
This first experience was very rich, positive to show us another way that will not substitute the traditional budget that we call presence-based, but that will complement it very adequately … the presence-based [process] is fundamental, but it cannot account for everything in a city of 20 million inhabitants, it doesn’t account for the complexity of the city, for this you need a process within which a large portion of the population feels represented … and the digital format is very good to mobilize and call attention. (Pimentel Reference Pimentel2008)
Many observers consider the implementation of PB Digital, which occurred in parallel with PB Regional, a very successful exercise (Coleman and Sampaio Reference Coleman and Sampaio2017; Sampaio et al. Reference Sampaio, Maia and Marques2011). With 172,938 votes cast electronically in 2006, 124,320 in 2008, and 92,728 in 2010, it can be argued that PB Digital indeed reached a much wider number of participants, especially among the middle classes (Prefeitura municipal de Belo Horizonte Reference Baiocchi, Braathen, Teixeira, Stokke and Törnquist2013). This shift, however, constituted an important change in focus for PB in the city. In fact, PB Digital was always presented as a complement to PB Regional, but its introduction changed the space occupied by the latter in the model of participatory governance.
Moreover, because PB Digital is an exercise in direct democracy—it is close to a referendum—the concept and practices of citizen participation on which it relies differ from those associated with PB Regional. In fact, contrary to the prevailing conception of participation that was at the core of the implementation of PB in 1993, here participation does not mean face-to-face interactions, deliberation, learning about the city’s preoccupations, and so on (Wampler Reference Wampler2015). Participation is limited and does not really involve much more than a vote from participants. It also removes the element of public discussion about what should be done for the city, about the vision that citizens want to develop for their city that precedes the vote in PB Regional, as the municipal government proposes the projects submitted to the vote. Thus, under Pimentel’s government, PB Regional remained a strong institution and was not questioned, but by including the digital format as PB, PB Digital displaced PB Regional and changed the meaning and practice of citizen participation, pushing it closer to participation as direct vote rather than as social inclusion and emancipation.
The third period in PB’s institutional evolution corresponds with the arrival of PSB candidate Marcio Lacerda to lead the PT-PSB coalition in power, and the later rupture that has left the PT without access to the executive since 2012. From the beginning, Lacerda has been a champion of citizen participation, promising in his 2012 electoral platform to widen even more social participation mechanisms. However, PB is slowly slipping under the radar in the political discourse, as well as the municipal government’s activities, to the benefit of a diversification of participatory tools as the political and ideational environment changes in the governing coalition. In fact, the notion of participation that is expressed by the main political actors and bureaucrats in charge of the interpretation of PB rules and enforcement has changed into something closer to “the government hears you,” rather than an emancipatory process aimed at transforming power structures.
In an interview, the head of city planning described participation as a matter of receiving citizens, listening to those who want to talk about their needs, their problems: “I think that this is it. With our flexibility, we make people participate” (Head of City Planning 2014). Without questioning their existence, participatory practices in the municipality are indeed rethought to become part of what Lacerda’s team calls shared governance (gestão compartilhada). This concept is somewhat removed from PB’s initial concept of emancipatory and deliberative participation, as explained by the head of planning: “we give the citizen a chance to know better what we are doing in the administration. We go there and listen, and give feedback.” Participation is, in fact, more about listening, consulting, and transparency.
This is the idea behind the modifications that were made to the structure and responsibilities of the municipal government secretariat, as well as the creation of an office dedicated to gestão compartilhada, directly attached to the secretary of government and centralizing all participatory programs (including PB) close to the mayor’s cabinet. Lacerda and his administration did not directly challenge the existence of PB, but they markedly changed the participatory architecture of the city through PB displacement. Not only did they introduce changes to the neighborhood boundaries, thereby disempowering local PB leaders’ support bases, but they also created a new participatory process at the municipal level called the Planejamento Participativa Regionalizado (PPR, Regional Participatory Planning). PPR, a citywide participatory forum aimed at generating long-term planning goals for the city, was launched in parallel with the PB process by the Lacerda administration in 2011. Officially promoted as complementary to PB, it became Lacerda’s flagship participatory program and the backbone of the gestão compartilhada system, displacing PB and changing its purpose as the main locus of citizen participation in the city (Author interviews, July–August 2014).
While PB is no longer the only way to talk about citizen participation in Belo Horizonte, and while the program has its share of difficulties with the delays in project delivery that date to 1999, Lacerda’s administration has never questioned the existence of the institution itself. Every two years, the PB cycle starts, and the population in local plenary assemblies approves more public works. However, data on project completion show inertia on the local government’s part during this period with respect to compliance with previous PB commitments. For example, between the 2011–12 and the 2013–14 PB cycles, no previous projects were completed; hence the stagnation at 1,195 observed in figure 1. During this period, however, 123 new projects were approved, increasing the steps the municipality must take to close this growing gap and achieve successful policy outcomes from the participatory process.
From Gradual Change to Deinstitutionalization from Within
Certain observers talk about PB’s transformations over the years as a renewal (Wampler Reference Wampler2015), but an institutional analysis shows that, in fact, the aforementioned changes are often hidden or implicit. This is an important distinction, because it makes it more difficult for us to grasp the amplitude of those changes, to assess their consequences for the institution itself, its practices and policy outcomes. In the case of Belo Horizonte, PB cycles continued to happen every two years, at least until 2014.Footnote 11 Nevertheless, the policy observed, coupled with more recent changes (including the diversification of participatory tools that institutionalizes new practices), has led to a gradual delegitimation of PB as an efficient and inclusive participatory mechanism among participants and citizens. The combination of layering, conversion, drift, and displacement concretely changes the way citizens think about their participation in the local democratic governance process. And this change in perception may lead to even more significant changes and the gradual deinstitutionalization of PB.
The gradual changes observed in Belo Horizonte over time have generated a significant loss of confidence in the institution on the part of citizens and some apathy among participants, all of which is expressed in different ways. Trust is, indeed, an important component of policy legitimacy, as Sampaio and Barros (Reference Sampaio and Barros2017) have shown in the case of PB Digital. Consequently, PB has lost its political salience among citizens over the past few years: it has become harder and harder to mobilize citizens to participate in PB, something that is clearly shown by the significant decrease in numbers of participants since 2009.
The growing and accumulated delays for project delivery in PB are due at least partly to some form of institutional inertia, something that participants are very aware of in Belo Horizonte. Figure 2 shows that during the 2015–16 cycle, held in 2014, 44.6 percent of the surveyed participants said they thought most public works approved in the past cycles were late, while 14 percent were under the impression that all the projects were late.
Figure 2 PB Participants’ Evaluation of Completion Rate for Public Works, Belo Horizonte, 2014
Moreover, participants felt mostly dissatisfied with the way the municipal government handled the situation. As figure 3 shows, 40 percent of the respondents said they were sometimes satisfied with the information on the status of previous projects given by the municipality, and 21.8 percent said they were never satisfied. This is a situation I also observed in my interviews and in the PB meetings I attended over the years. In particular, I remember sitting in on a meeting of the Municipal Comforça, the organization in charge of overseeing the execution process once the projects are approved.Footnote 12 The meeting was in August 2014, two weeks before the official municipal opening of the 2015–16 PB cycle by Mayor Lacerda. At this meeting, the tensions between the administration and the citizens elected by their neighbors as PB city delegates were palpable, as the only subject most people wanted to talk about was the delays and the ways the municipal administration was handling (or not handling) the situation. Delegates were seeking responses and solutions from the administration that never really came, leaving them very dissatisfied at meeting’s end.
Figure 3 Participants’ Level of Satisfaction with Municipal Government Information About PB, Belo Horizonte, 2014
Because the municipality mostly has not paid attention to it, the question of delays in execution and delivery of public works has become increasingly problematic and affects the way the institution, based on wide popular mobilization, deliberates investment priorities. It has not only turned the debates over future projects into an ongoing and unresolved discussion and conflict over the growing number of unconcluded projects, but it also has increased the popular feeling that PB is no longer an efficient mechanism to formulate demands and obtain access to public resources. Speaking about the difficulties his association has encountered in mobilizing participants in this context, activist participant Rogério explained that this situation indeed makes mobilization very complicated and weakens the credibility of the mechanism.
Today the process [of concluding the public works] is even more behind. So much so that there is a decrease in people’s participation. You don’t get to mobilize as you mobilized before. … Lindéia [neighborhood] was territory 5 and now is 3. It was always an example of mobilization. If I’m right, it did not reach two hundred organized participants in PB, because community leaders don’t want to wear them out in the process. … When you try to mobilize, what people do is: “but we will ask for others? And the ones that have been concluded until today? Well, I am not going in there today, no.” (Rogério 2014)
Efigênia, from Centro-Sul, says the same: “For us today, it has become more complicated because the municipality is very late with the conclusion of public works since 1999, we have here a huge number of achievements [conquistas] that were not delivered.” She explains that this situation has consequences for residents’ ability to mobilize and for the credibility of PB.
We won the streets, urbanization, sanitation, but the municipality is not investing, so we have been losing credibility among our community … you succeed in mobilizing an entire sector to get upgrades in the streets of this sector, but this upgrade never happens so we fall into discredit. … This is what happened here, with PB, the municipality was late, and in the following plenaries, participation decreased because people lost their faith, they lost their trust in the thing, so today we cannot mobilize as we used to. (Efigênia 2014)
This apathy toward PB and the difficulty in mobilizing participants seems to be exacerbated by the recent reforms undertaken by the Lacerda administration. For example, the timing of the 2015–16 cycle generated a feeling of institutional disuse among participants. In fact, this cycle was supposed to take place between April and June 2014. However, it was opened only at the end of August, with the assemblies taking place in late November and early December. The administration explained the timing in two ways. First, it was rethinking the participatory architecture to make it better and closer to the mayor’s program, by moving PB to the Secretaria de Governo. Second, the initial timing was complicated by the coming of the World Cup to the city in June and July. For many, however, this was a sign that the local administration was gradually abandoning the program. Rogério, a longtime PB delegate from Barreiro whom I knew as a participatory enthusiast back in 2007, now expressed this sentiment:
Before, PB was every year, at the beginning. They went for every two years. This year, actually, it will be three years. They say they will catch up after, but will they really have a PB in 2016? The 2014 PB will end in 2015. In this perspective, the bidding process for public works will only start in 2016, the year after. … Thus the participation will be worse. Because the citizen will not understand: “Look, this was last year, and the project will go for bids now.” So sometimes you think the following, that it ends up being a way to demotivate people even more and to weaken the credibility of the process. This is a very serious problem. (Rogério 2014)
Thus, gradual and not so visible processes of change to PB affect its policy outcomes, which generates dissatisfaction among the public, as the institution is considered no longer to be fulfilling its mandate. It generates very poor policy outcomes for people who invest time and mobilize for their community, as most of them do not see the results of their mobilization for many years.
Through this gradual change, political elites are also gradually deinstitutionalizing PB by layering, converting, and eventually displacing it, making it socially and politically irrelevant. Participation remained at the core of the left-to-center coalitions’ discourse over the years, but PB is no longer the backbone of its institutional articulation in Belo Horizonte. Instead, PB has shifted toward becoming one policy tool among others in the municipal participatory architecture, and now relies on an idea of participation that is moving further and further away from the idea of inclusion and emancipation of the poor that prefaced PB’s implementation in the city. PB is perceived as losing its political salience in the city. Roberto, president of a neighborhood association, explains that he indeed feels participation is no longer an important concept in Belo Horizonte and that PB seems to be merely a façade.
Today, I feel that it [participation] does not have any importance. My feeling is that this [PB] is done to “look good on paper.” And they [municipal officials] don’t have the courage to end it. This is my feeling. … Naturally, a mayor prioritizes one thing and another, one other thing. I think this is natural. I thought that this one [Lacerda] is not giving a lot of importance to PB. But sometimes, in numbers, in documents, in meetings, he can even say the opposite—this is a feeling. (Roberto 2014)
As the 2015–16 cycle unfolded in the fall of 2014, participants seemed to think Roberto was indeed right. During one month, plenary assemblies were organized in all the subregions. However, it was very hard to get the information about PB if you were not a regular participant or someone in the close circle of PB offices. The PB portal website was not actualized, and the Facebook page of the municipality never promoted the event, both signs that mobilizing for and publicizing PB were no longer a priority of the administration.
Conclusions
As historical institutionalists like Streek and Thelen (Reference Streeck and Thelen2005), Mahoney and Thelen (Reference Mahoney, Thelen, Mahoney and Thelen2010), or Hacker et al. (Reference Hacker, Pierson, Thelen, Mahoney and Thelen2015) suggest, institutions have internal properties that allow institutional change to occur gradually, through a variety of mechanisms used by actors driven by both political will and the inherent features of the institution. The historical analysis of Belo Horizonte’s PB institutional trajectory in a changing political environment shows that although they are often studied separately, the four mechanisms these scholars identify do not operate independently from one another. In fact, conversion, policy drift, layering, and displacement are institutional change mechanisms that are closely intertwined, reflecting the longer-term changes in the local political environment.
The analysis also highlights the importance of both the national and local dominant ideational environments, which is the way participatory democracy was defined and discussed in the public sphere, to understanding the direction of change in PB practices at the local level. Thus, and contrary to what has happened in other Brazilian municipalities, the empirical analysis of PB’s trajectory of apparent stability shows that, at least until 2014, PB was not abandoned in Belo Horizonte, even though the political and ideational environment changed tremendously in the city over its 20 years of existence. However, it has been reformed through a gradual process of reinterpretation and political inaction in the face of challenges to its efficiency.
If such changes sustain institutional adaptation to new contexts, they are not without implications, something historical institutionalists mostly have overlooked and have not yet theorized. In fact, change can imply a deeper transformation of the institution from within as its practices and policy outcomes also change in the process. The case of Belo Horizonte contributes to the literature by allowing us to uncover and to theoretically understand one of these unexpected implications for institutional stability. Gradual institutional change and the adaptation of the participation ideal did not directly question the formal existence of PB as an institution, but they did change its practice and its policy outcomes from within. This shift led to decreased legitimacy as a citizen participation tool and lower political salience, mostly because citizens valued PB less and less. In fact, as in the case of PB in Belo Horizonte, changes generated over time through the interplay between policy drift, conversion, layering, and displacement mechanisms may even lead to an institution’s gradual deinstitutionalization and abandonment over time. As the “father” of PB in Belo Horizonte, Patrus Ananias, told me, “there are many ways to abandon a public policy. You can do it formally, you can do it informally. … It is the practice that defines it” (Ananias Reference Ananias2014).
On a cautionary note, I would like to conclude by saying that this approach to looking at PB’s stability in Belo Horizonte was not meant to disqualify this transformation or to impose a normative judgment on the fate of PB in the city. The goal was instead to look at the dynamics of institutional adaptation from another theoretical lens, one that uncovers the reality that in certain circumstances, politicians can rely on a variety of mechanisms for institutional change in contexts of political and ideational environment change. Change, adaptation, and, in this case, deinstitutionalization from within are responses to new political, social, and economic realities that change the environment in which institutions evolve, and can even pave the way to further institutional innovation.
