Political parties organize Brazil’s political life before and after the elections and play a key role in relationships municipalities have with the states and the federal government. Public policies are decided during maneuvers that involve the state legislative assemblies and Congress, in other words, by the legislators themselves. Although members of the Chamber of Deputies may begin their careers on the local level, they gradually adopt a strategy of dispersing their voter support within a specific region. This forces them to respond to demands from their municipal bases while, at the same time, trying to expand them. Contrary to what people generally believe, there is no game of “give-and-take,” but rather a sophisticated mechanism that connects the different levels of power.
Those conclusions are the findings from the research study entitled Political Institutions and Public Spending: a Comparative Study of Brazilian States, which was conducted from 2009 to 2013 under the leadership of political scientist George Avelino Filho, professor at the Getúlio Vargas Foundation (FGV). Supported through FAPESP’s thematic project mechanism, the study, which attempts to elucidate how the political system functions in Brazil, is now focusing on a second stage, entitled Subnational political institutions: a comparative study of Brazilian states. “Very few studies have been carried out to understand how politics works in the states. Our intention was to apply what we already knew about the political system as a whole to politics at the state level,” says Avelino.
According to the researcher, there is a debate in Brazilian political science circles that centers on the weakness of the parties, the possible existence of “informal districts,” and how this influences the activities of the legislators. Having being elected by votes that were concentrated in a single region, in the so-called “informal districts,” deputies favor more fragmented public policies. The phenomenon, known as “pork barrel politics” implies a policy of bestowing economic benefits or services on a geographically circumscribed area.
The first step in Avelino’s study was to expand the focus to include the municipalities, since Brazilian federalism, unlike most federations elsewhere in the world, involves three levels of government. Since redemocratization, and to a greater extent since the current Constitution was adopted in 1988, mayors are responsible for implementing important public policies, such as health and education. “We want to look at the states as aggregates of municipalities but in addition, as voting precincts,” Avelino observes. And, by including a detailed investigation of votes at the precinct level, the study permits a more detailed analysis of election results and increases the understanding of the way in which deputies are being elected.
One of the researchers involved, economist Ciro Biderman, also an FGV professor, proposed that the “G index,” an indicator that measures the degree of geographic concentration of productive sectors and is widely used in studies of regional economies, be adapted to study the campaigns waged by deputies. And so, considering an election in the state of Amapá, for example, it would be mathematically expected that a candidate would obtain 60% of his votes in the state capital of Macapá, since that is where 60% of the state’s voters live. However, if that politician adopts the strategy of concentrating his campaign on a specific region, it can be expected that the number of votes won there would be proportionately higher than the percentage distribution of the electorate. This would result in a higher “G index.”
In the study coordinated by Avelino, the G index was first applied to elections in São Paulo State and then extended to all Brazilian states, covering the period 1996-2010. What was discovered was that the profiles of elected deputies tend to be concentrated in municipal terms, but somewhat dispersed regionally. “It is not hard to see why that happens. A deputy needs to seek support beyond the places where he is best known. He needs to travel, but that is expensive and the itinerary cannot be chosen at random. The most important party operatives are the mayors, and it is in the party organization that a candidate finds more efficiency for his campaign,” Avelino says. And so, a local leader starts by getting elected in his native city, then sets out to expand his action regionally and, finally, having become known as a legislator, can diversify his effort to win votes.
This deconcentration of votes, something equivalent to focusing his efforts as candidate more broadly, making the campaign state-based, is part of a risk diversification strategy that, in come cases, gains altitude in the state majority-based elections (for the senate and presidency), but may also bring an end to a politician’s career. “This is a moment of vulnerability, i.e., candidates leave their comfort zone in order to completely diversify their electoral assets, thus running the risk of not being competitive in any venue, which greatly increases the probability of losing an election,” Avelino explains.
His study proved that there does indeed exist an inter-party network in which mayors and federal deputies maneuver and that it operates within the logic of a search for federal funding for local governments and the subsequent reaping of electoral dividends. By cross referencing the data from the most recent elections, collected from a universe of 5,221 municipalities with fewer than 200,000 residents, FGV researchers discovered that mayors are responsible for an increase of about 20% in the votes that are cast, two years later, for candidates for federal deputy from their party. Mayors need funds in order to govern while legislators, once elected, will have shortened lives unless they have the local support needed to maintain and expand their electoral base.
Parties with a high degree of capillarity, like the Democratic Movement Party (PMDB) which, in the 2008 elections elected about 1,200 mayors within the universe of the 5,221 municipalities studied, preserve their clout—and their bargaining power—in the municipal elections. The PMDB mayors have been the key party operatives in campaigns for Congress. “We want to see how public policy decisions are related to the political party issue,” says Biderman. One hypothesis to be tested is whether the influence wielded by elected individuals, translated into votes for mayors and deputies, brings concrete dividends to the localities. “Decisions on public policy take votes into consideration, but it is also true that it is the kind of vote that will determine the public policy to be adopted,” he says.
The research data indicate that the connections are stronger between mayors and federal deputies than between mayors and state legislators. That difference can be explained by the decentralization of the political and administrative powers that reinforced the transfers of programs and constitutionally-mandated revenues to the municipalities.
Political science literature reports that the election of an American president ends up influencing the number of votes received by many congressmen from the same party. That effect, called the “coattail effect,” operates from top to bottom within a single election campaign. In other words, a well-regarded president tends to increase the chances that candidates for federal congressman from his party will get elected. By applying that method to Brazilian elections two decades ago, U.S. researcher Barry Ames found that the influence flowed from bottom to top, i.e., it had a reverse coattail effect. He justified the results by saying that there was what he called “informal districts” in Brazil.
Avelino’s research borrowed the term “reverse coattail effect,” but reached different results. The parties, not the deputies, are the ones who organize successful performance in the elections conducted in the states during which all the federal legislators are elected. “Why would Congress put so much effort into trying to do away with virtualization? the researcher asks. The answer is that since redemocratization, the parties have been worried about losing their freedom to stitch together different sources of support on the national and state levels as a means of improving their chances in the state legislative elections.
In partnership with the Brazilian Political Science Association, researchers obtained access to raw data gathered by the Superior Electoral Tribunal (TSE). This enabled them to obtain election data from the precinct level. In practical terms, the research resulted in the development of the Copespdata platform, a software application that enables the public to consult election data (www.fgv.br/cepesp/cepespdata). Freely accessible, its interface is more user-friendly than that of the TSE, and so can be helpful to researchers from any institution. Also, using data from the Brazilian Institute of Geography and Statistics (IBGE), it is possible to aggregate votes at the level of a microregion or mesoregion. In order words, one can obtain figures on groups of municipalities that identify niches within a state that are economically and socially better organized.
The platform has allowed researchers to make a series of unprecedented correlations, some with surprising results. So, in the 2010 elections, states like Rondônia, Paraná, Espírito Santo, and São Paulo presented higher than expected degrees of vote concentration, in contrast to the results from Sergipe, Rio Grande do Norte, Piauí, Acre, and Amapá. Of the nine Northeastern states, only Ceará and Alagoas recorded a higher than expected degree of concentration of votes for their federal deputies, given the fragmentation of parties. Those data contradict the popular view that the less prosperous regions are dominated by a more traditional politics based on electoral strongholds and a concentration of votes—putting to rest once and for all the suggestions as to the presence of “electoral corrals” in Brazil.
Along another line of investigation pursued by the project, researchers mapped the formation of the cabinet-level departments (secretariados) in 14 states during the period 1994-2010. The goal is to explore the diversity among Brazil’s states in order to better understand how those coalitions are composed and whether they can be related to the national coalition.
Such analyses need a local observation point, which is why researchers decided to form a federative research network composed of political scientists and other experts who have experience with subnational governments. The studies will continue in a second thematic project, approved as continuation of the first one. The idea is that the network would cover all 27 states and would conduct a case study for each state’s specific reality, feeding the information into the Copespdata database in order to be able to make comparisons among states.
The initial findings are promising in that they reveal something that defies generally held beliefs or conflicts with preconceived—usually negative—ideas about politics in Brazil. But, according to those responsible for the study, there is still a long road ahead. Avelino suggests some ad hoc reforms that could strengthen the political parties in their roles as mechanisms for representation, such as terminating, or at least regulating, the formation of coalitions for legislative elections, and strengthening party loyalty. “In the first case, we would see a reduction in the number of parties, making it easier for voters to understand who is running against whom. In the second, we would increase the costs of leaving—changing—parties, which would encourage politicians to invest in their current parties. We should not forget that political parties are still the best means of representing segments of the public and including them in the democratic political system. If we want a more egalitarian society, we must reinforce that important mechanism of inclusion,” he concludes.”
Political Institutions and Public Spending: a Comparative Study of Brazilian States (No. 2008/03595-7); Grant mechanism Thematic Project; Principal investigator George Avelino Filho (FGVSP); Investment R$293,504.60 (FAPESP).
AVELINO, G. et al. Articulações intrapartidárias e desempenho eleitoral no Brasil. Revista de Ciências Sociais. v. 55, n. 4, p. 987-1.013. 2012.
AVELINO, G. et al. A concentração eleitoral nas eleições paulistas: medidas e aplicações. Revista de Ciências Sociais. v. 54, n. 2, p. 319-47. 2011.