The political economy of government formation and local public goods

This thesis examines three questions: first, do national government coalitions favour local governments connected to them to receive local public goods? Secondly, does favouritism in the allocation of public goods imply large welfare losses? Finally, how national governments form, and what are the consequences of this for national policy making? These questions are answered in the particular context of Brazil, where rich data on national politics and local public good allocation is available. The first chapter of the thesis summarizes aspects of the Brazilian context that are relevant for the rest of the thesis – covering aspects of Brazilian national politics, and of the rules for allocation of funds for local public goods. The chapter also discusses the disaggregated data on the universe of matching grant transfers from the Brazilian national government to municipalities, used in the second and third chapters. The second chapter answers the following question: are regions connected to the national government favoured to receive funding for local public goods? While a broad literature shows that “politically connected” regions receive more funds from national governments, it is unclear whether this reflects favouritism, or simply connections allowing the national government to know better the needs of regions connected to them. The chapter finds evidence broadly consistent with favouritism. The third chapter then examines the welfare losses associated with favouritism. I build a model of grant requests by cities and approvals by the national government and provide estimates of the model’s parameters. Despite ample evidence of favouritism, if the only source of conflict between the national government and society is due to favouritism, the welfare losses for society due to favouritism are of the order of 0.24% of the budget for grants. The second and third chapters suggest large effects of the national coalition over local public good provision. The fourth and final chapter, instead, analyses how national coalitions interact with national policies. More precisely, do government coalitions form to include legislators ideologically close to the executive, or ideologically unattached legislators whose votes are “easier to buy”? Moreover, what are the consequences of this for policy making at the national level – in particular, for roll call votes in the chamber of deputies