In Brazil, most voters do not identify with a political party. In spite of this relatively low level of partisanship, Brazilian elections, at least at the presidential level, have settled down into contests between the Partido dos Trabalhadores (PT) and the Partido da Social Democracia Brasileira (PSDB). The PSDB, led by Fernando Henrique Cardoso, won the first two post-stabilization elections (1994 and 1998), but the PT, headed by Luiz Inácio Lula da Silva in 2002 and 2006 and Dilma Rousseff in 2010, has won the last three contests.
In the national legislature, by contrast, the party of victorious presidents has never held a majority of seats. Still presidents have been able to build coalitions that make effective governing majorities.
Scholarly debates about Brazilian parties and elections mostly focus on presidential elections and on the roles of partisanship and ideology, with retrospective assessments of the economy and administration. These debates are important, but they suffer from one problem: presidential elections are single events held in unique circumstances.
Como o eleitor escolhe seu prefeito: campanha e voto nas eleições municipais (How Voters Choose Mayors: Campaigning and Voting in Municipal Elections), a collection of 13 articles organized and edited by political scientists Antonio Lavareda and Helcimara Telles, changes the level of analysis in an important and very useful way. Lavareda, president of MCI Estratégia, and Telles, a professor at Universidade Federal de Minas Gerais, focus on the 2008 municipal elections in the capital cities of key states. The chapters analyze campaigns and the bases of electoral choice in the cities of São Paulo, Rio de Janeiro, Belo Horizonte, Salvador, Fortaleza, Curitiba, Recife, Porto Alegre, Manaus, Belém, Goiânia, and Florianópolis.
The introduction, written by Antonio Lavareda, establishes the economic background to the 2008 elections. During the four-year period leading up to the vote, including the election year, growth was high and inflation remained low. The importance of public satisfaction with Lula’s administration was not lost on local politicians. While few voters actually cast votes based on a local candidate’s ties to Lula, his popularity negated campaign attacks on the national administration. Candidates who could link themselves to the president tried to nationalize their local elections by focusing on economic growth and the expansion of social benefits through the family stipend program; candidates of the opposition tried to “municipalize” the contests by focusing on local issues.
Lavareda also stresses political continuity. Thanks to a constitutional amendment in 1997 allowing executive officeholders to run for a second term, mayoral reelection rates have climbed steadily, from 58 percent in all municipalities in 2000 and 2004 to 67 percent in 2008. In municipalities that are state capitals, reelection rates in 2010 reached 95 percent.
While partisanship may arguably be less of a factor in national contests, Lavareda argues, it matters at the local level. Party fragmentation, which appears very high in the national legislature, is more manageable locally.
Typically, three or four parties capture most of the municipal vote, and smaller parties ally with parties that at least are not ideologically distant. In many states, particularly those in which the capital city does not sit on a rich resource base in relation to the state as a whole, state governors control vital resources and play central roles in municipal elections.
Partisanship also matters organizationally. Brazil restricts private spending on political campaigns, but in the final months before an election, candidates are entitled to airtime during the Free Period of Electoral Advertising (Horário Gratuito de Propaganda Eleitoral—HGPE). The number of daily minutes each candidate receives is a function of the legislative seats held by the parties supporting that candidate. Not surprisingly, major parties form coalitions with smaller parties to augment their airtime.
The importance of the HGPE is documented in the chapter on São Paulo by Jairo Pimental Jr. of Universidade de São Paulo and Claudio Luis de Camargo Penteado of the Universidade Federal do ABC. The authors provide a nuanced account of how Gilberto Kassab, the mayoral candidate of the Democratas party, won re-election with skillfully constructed alliances that gave him more time on the HGPE and “greater visibility and space for his campaign.” The result: voter opinion of his administration changed from negative to positive.
In Rio de Janeiro, Fábio Vasconcellos and Marcus Figueiredo of the Universidade do Estado do Rio de Janeiro look at how the vote was age- and class-linked. Partido Verde candidate Fernando Gabeira won the votes of richer, younger and more educated people. His opponent, Eduardo Paes of the Partido do Movimento Democrático Brasileiro (PMDB), did well in the first round among voters who held a positive view of outgoing Mayor César Maia (then a member of the Democratas party), only to see those same voters opt for Gabeira in the second round. While Gabeira based his campaign on clean government, Paes promised he could do more for the poor.
Party affiliation was the key ingredient for success in Salvador, according to Cloves Luiz Pereira Oliveira of Universidade Federal da Bahia and co-authors Dalmir Francisco and Tiago Prata L. Storni of Universidade Federal de Minas Gerais. There, incumbent mayor João Henrique faced very low approval ratings a year before the election.
When he changed parties, dropping the Partido Democrático Trabalhista and joining the PMDB, he gained the support of a member of President Lula’s cabinet, Geddel Vieira Lima, who controlled public works resources and the party organization. As an evangelical, João Henrique also drew support from that community and was able to position himself as a champion of the poor.
Exceptions, of course, don’t prove rules. But the case of Manaus, the only municipality in which an incumbent lost, is instructive, as noted by Luciana Fernandes Veiga and Sandra Avi dos Santos of Universidade Federal do Paraná and Malco Braga Camargos of Pontifícia Universidade Católica de Minas Gerais.
Serafim Corrêa, seeking reelection from the PSB, lost the contest in the capital city of Amazonas to Amazonino Mendes of the Partido Trabalhista do Brasil. Corrêa had won in 2004 due to scandalous personal behavior by his opposition.
But in 2008 the “machine” of ex-Governor Gilberto Mestrinho and Amazonino Mendes, which had ruled Amazonas since the 1950s, retook the municipality and relied on clientelistic exchanges with poor neighborhoods to do so. Corrêa was constantly attacked by the opposition-controlled media and, as a result, he began the campaign with low approval ratings and lost in the second round.
The book’s final chapter, by Lavareda and Telles, looks at these cases, along with those from other cities, to construct an overall model of Brazilian mayoral elections. In some states, especially poorer ones where mayors are dependent on state largesse, the connection between the governor and the mayor is crucial. In other cases, like the cities of Belo Horizonte and São Paulo, national elections affect local politics: it’s PT versus anti-PT. Ideology seems to matter in some cities, especially in the more economically developed areas such as São Paulo, Belo Horizonte and Florianópolis.
Lavareda and Telles emphasize that party organization and partisanship matter, but their importance depends on historical factors rather than on a state’s level of economic development. In Recife, a poor city in a poor state, a candidate without political strength won election simply because he was affiliated with an electorally strong party. In São Paulo it worked the other way: anti-PT sentiments helped elect a conservative candidate.
In Brazil, where partisanship is weak, campaigns matter much more.
Overall, this book marries qualitative and quantitative methodologies in a thorough but never boring treatment of local electoral processes. It stands as a testament to the sophistication of Brazilian social science.
Its major weakness is the reliance on cross-sectional rather than longitudinal survey data. Why is this important? Partisanship is hard to study in Brazil because it changes over the course of a political campaign. People reporting no partisan affiliations at the beginning of a campaign come to like a candidate and share that candidate’s party preference.
Thus Brazilian partisanship, unlike U.S. partisanship, can be endogenous. To analyze partisanship more rigorously, scholars need to interview the same respondents at various points during the campaign, measuring their changing partisan identification, their responses to campaign messages and their evolving evaluations of political candidates.
These authors cannot be faulted for relying on the data available to them, but a complete analysis of partisanship will require better (and more costly) databases. Still, by emphasizing the variety of local and state political processes, and by systematically linking these subnational processes to broader themes of national-level politics and partisanship, Antonio Lavareda, Helcimara Telles and their colleagues have made a major contribution to our understanding of Brazilian politics.