Appropriating for the Team: Distributive Politics and Intra-Party Coordination in Developing Democracies

The conventional wisdom about elections in the Brazilian OLPR is that they are uber individualistic. Candidates are elected mainly on the personalistic connection they are able to build with voters. In order to do so, they target particularistic goods to their local constituencies, engaging in clientelism and pork barrel politics. All this flows from the electoral system. Intraparty competition in OLPR weakens party labels, which are useless in deciding among copartisans, and forces candidates to adopt personalistic strategies in order to distinguish themselves from their copartisans. Accordingly, parties play a minimal role. On the other hand, when it comes to developed countries with similar OLPR systems it is taken for granted that parties are central players, selecting candidates and coordinating campaigns. This disparity, among other factors, has led some scholars to argue that parties are in fact equally important in Brazil. Nevertheless, they haven’t yet been able to provide an account of the mechanisms through which parties play such roles. Clientelism and pork barrel, however, are not incompatible with intra-party coordination. Politicians may channel particularistic goods through intra-party networks. Preliminary qualitative evidence suggests that this might be the case in Brazil. Congressmen do engage in distributive politics, but privilege constituencies under the local political control of copartisan mayors. In order to test this hypothesis, we adapt the classic close-elections regression discontinuity design to analyze pairs of congressmen-mayors. We take advantage of the fact that Brazilian electoral rules allow for multi-party lists, which means that sometimes the last elected candidate and the first un-elected one within a list are from different parties. This allows us to test simultaneously whether copartisanship, local constituencies (measured by the candidates’ concentration of votes at the local level) or a combination of both increases the likelihood of a congressman appropriating resources for the municipality. This work contributes to a growing body of literature arguing that particularistic practices are not incompatible with strong parties in developing democracies. Contrary to what some analysts once prescribed to (then) new democracies, strengthening parties might not reduce clientelism and pork barrel politics or bolster accountability for universal policies.