UC Berkeley

This dissertation studies local public finance in Brazil, with a focus on how federal and state government policies affect local spending and revenue generation.

In Chapter I, I provide a descriptive analysis of local public finance in Brazil, with a focus on local revenue generation. In addition to describing the system of public finance in Brazil from a municipal perspective, I discuss the major challenges facing local governments in Brazil and in other low and middle income countries and present the reader with a descriptive picture of local revenue generation in Brazilian municipalities using publicly available data on municipal accounting records, municipal population and GDP, and municipal revenue generation infrastructure and administration.

In Chapter II, I study the relationship between intergovernmental transfers and local revenue generation in Brazil. A major consideration in designing an intergovernmental transfer system is the concern that government transfers may "crowd out" local revenue generation. While traditional public finance theory suggests that this will be the case, empirical findings often refute this theoretical prediction in favor of the so-called Flypaper Effect. I take advantage of exogenous changes to state formulas determining the municipalities' share of value-added tax resources as an instrument for endogenous transfers to estimate local revenue generation responses to transfers. I find no evidence that government transfers reduce local per capita revenue generation in the context of two states in Northeastern Brazil, contributing to the limited body of rigorous evidence finding that intergovernmental transfers do not necessarily reduce incentives for local revenue generation and that the Flypaper Effect can indeed exist in certain contexts.

In Chapter III, I measure the impact of Brazil's personnel expenditure limits imposed as a component of its Law of Fiscal Responsibility, implemented in 2001. Personnel expenditure limitations require that municipalities spend no more than 60% of liquid current receipts on personnel. I measure the expenditure limit's impacts on a series of public finance outcomes by comparing changes in outcomes among those bound by the limits (those that were initially above the limit) to those just below the limit using a difference-in-difference regression framework with municipality fixed effects and controlling for state-specific flexible time trends. Chapter III adds to the large body of literature examining the impacts of tax and expenditure limits (TELs) but is one of the few to do so outside of the US context. It is also one of the few studies of expenditure limitations rather than tax limitations, as most of the US TELs are focused on limiting taxation rather than spending. While Brazil's personnel expenditure limits are generally successful in reducing personnel expenditures, municipalities seem to be substituting non-personnel expenditures for personnel expenditures rather than slowing the growth of total expenditures. At the same time, the expenditure limits seem to encourage municipalities to ease their spending constraints by increasing revenues where they can. The net effect is a small but significant reduction in deficit spending.