Contacts

How Central Governments Affect Local Elections

, by Gunes Gokmen
An analysis by Tommaso Nannicini with Fernanda Brollo shows that when local elections get near the Brazilian government transfers less funds to nonaligned mayors, especially when they have been elected by a thin margin

Can the pre-election federal transfers to municipal governments be lopsided in favor of those associated with the coalition or the party of the head of government? Such a claim has long been debated and it resurfaces every election period with the opposition and the critics accusing the incumbent and the ruling party denying such allegations. Tommaso Nannicini (Department of Economics and IGIER) and Fernanda Brollo (University of Warwick) show, in their Tying Your Enemy's Hands in Close Races: The Politics of Federal Transfers in Brazil, published in the American Political Science Review (doi: 10.1017/S0003055412000433), that in Brazil, in pre-election years, those municipalities with a mayor affiliated to the president receive one-third larger discretionary transfers for infrastructure. This is because the federal government punishes the opposition party mayors that won by a narrow margin, so that their hands are tied for the upcoming elections.

The authors start with a theoretical framework in which the central government can allocate transfers to local governments in order to please voters directly and to increase the chances of winning of aligned candidates. The main idea is that the central government is in a position to use transfers to favour political friends or to punish political enemies at the local level. The model predicts that transfers are positively related to political alignment. Moreover, margin of victory in the last elections plays a role in that among the aligned municipalities those who won by a small margin receive more transfers, and among the unaligned municipalities those who won by a small margin receive less transfers.

Subsequently, by using a regression discontinuity approach, the authors overcome the endogeneity problems due to the fact that both alignment and transfers can be driven by other socio-economic factors. Based on Brazilian data, the authors identify the impact of political alignment between the local mayor and the president on the amount of federal transfers allocated to each municipality by comparing places where the candidate aligned with the president barely won with places where the candidate aligned with the president barely lost. The evidence suggests that in close elections municipalities with a mayor affiliated with the president receive one-third larger transfers for infrastructure during the last two years of the mayoral term. And such an effect of political alignment on transfers in close races is primarily driven by a punishment cut in transfers to unaligned municipalities.

In sum, the authors identify a causal effect running from political alignments to federal transfers by the central government. Regression discontinuity estimates, which accommodate for the presence of both time-invariant and time-varying confounding factors, indicate that mayors politically aligned with the Brazilian president receive larger federal transfers in close races by an amount in the range of 26% to 41%.