Center for Institutional Reform and the Informal Sector

Imagine an electoral system based upon open-list proportional representation, large multimember districts, candidate selection at the level of politically significant subnational units, and no limits to reelection. What would be the nature of constituencies? How would votes distribute spatially? What would determine campaign strategies? Scholars have undertaken little research on the strategic behavior of politicians under these rules. This paper explores spatial strategies by analyzing results from four Brazilian legislative elections between 1978 and 1990. I begin by characterizing vote distributions along two spatial dimensions: dominance and concentration. I then develop a theory of campaign strategy based on the nature of open-list PR and on the realities of Brazilian politics. The theory of strategy is evaluated with Poisson models predicting where deputies will offer budgetary amendments benefitting particular municipalities. The actual electoral outcomes of strategic behavior are then evaluated with ordinary least-squares regressions. The results demonstrate that deputies seek secure bailiwicks, search for vulnerable municipalities, and strive to overcome their own electoral weakness by delivering pork. Candidates’ tactics vary, partly because political backgrounds differ and partly because the differing demographic and ecpnomic contexts of Brazilian states reward some tactics and penalize others. The argument reported here has implications both for the Brazilian case and for other newly democratizing nations. I conclude, therefore, by assessing the importance of these findings for future research on pork-barrel politics, legislative behavior and electoral reform..b&h in Brazil and in other electoral systems. Latin America in the mid-l 990s is a region of optimism. Ffedgfing democracies are surviving; economies are stabilizing and growing. But Brazil, the region’s largest nation, remains an enigma. While prices rise 30% per month and the distribution of income deteriorates, political scandals rock the Congress, and unprecedented corruption led to the removal the nation’s first popularly electedPresident in thirty years. Increasingly, observers blame Brazil’s political institutions. Why arc political structures so ineffectual? Consider the party system and the legislature. Even by Latin American standards, Brazil’s party system is weak (Mainwaring and Scully, 1992). Few parties have genuine roots in society. Party vote shares are volatile over time and between presidential and legislative elections. In the Congress, party leaders exert little control over their delegations. Many, if not most, deputies spend the bulk of their time arranging jobs and pork-barrel projects for their constituents. Though electorally successful parties fall all across the ideological spectrum, some of the largest “center” parties are really just shells for deputies with no policy interests at all. Few Brazilian parties organize around national-level questions; the Congress, as a result, seldom grapples with serious social and economic issues. Brazil’s presidents benefit little from the weakness of the Congress. With only minimal chances to obtain stable legislative support, executives face politically independent governors, an electoral calendar imposing elections in three of every five years, municipalities -depending for their very survival on federal largesse, and a substantial core of deputies caring about their personal incomes first, reelection second, and public policy a distant third. Presidents govern by forming coalitions based upon cabinet appointments. Because these appointments must satisfy the demands of both party and regional forces, cabinets tend to very inclusive. The porkbarrel programs required to maintain them are costly, and policy innovation is extremely difficult. At the core of Brazil’s institutional crisis is the electoral system. A unique set of rules, usually referred to as “open-list proportional representation,” governs legislative elections. Politicaf scientists have explored Brazil’s version of open-list PR (De Souza and Lamounier, 1992; Fleischer, 1973, 1976, 1977; Kinzo, 1987, Lima Junior, 1991; Mainwaring, 19931, but the absence of appropriate data has limited research both in scope and breadth. Formal theories of voting, theories that have been enormously useful in understanding plurality systems like that of the U.S., have had little to say about multicandidate PR settings, except to note the absence of incentives for candidates to move toward the center (Cox, 1990). In sum, the consolidation of democracy in Brazil may well depend on our understanding of the relationship between institutional structures, especially the