UC Berkeley

This dissertation identifies the impact of various political institutions on public finance decisions in the United States. The particular institutions covered are the local voter initiative process, mayoral term limits, and political party allegiances; I utilize various econometric techniques to identify how each of these institutions influences the fiscal decisions at the state or municipal levels.

Chapters 2 and 3 provide two different approaches to identifying the effects of the local initiative. The initiative process influences policy directly as citizens are able to implement their desired policies; however, the threat of initiative also has the potential to influence the policies that elected officials will pass. In Chapter 2, I explain the history and utilization of the local initiative in California. In particular, I identify the factors that are associated with citizen initiatives, and thus a stronger threat of citizen action. Using data from local elections and initiaitives in California, I find that a strong political majority increases the probability of bringing a citizen initiative to the ballot; the presence of a majority increases the probability of success. Racial diversity leads to an increase in governance initiatives; these areas have more diverse preferences and are thus more likely to be dissatisfied with governance, on average. These cities cannot agree on the specific policies they prefer (they propose no more tax, land use, or initiatives of other categories), but they can agree to limit the powers of government. I also find that the use of the local initiative does not lead to significantly altered fiscal outcomes. The exception to this is that governance initiatives are associated with higher levels of outstanding debt and interest obligations, even when controlling for racial diversity and income.

Chapter 3 provides a direct test of whether the availability of the local voter initiative and recall processes and term limits influences municipal fiscal decisions in the United States. Using an adaptation of the Campbell-Mankiw (1990) model of forward-looking decision making, I find that the threats of voter initiative and recall do not significantly influence local fiscal policies. This suggests that voter preferences and local fiscal policy are aligned in the absence of these institutions. Term limits lead to increased short-sighted decision making, by removing officials' accountability to voters. This effect is roughly equivalent to removing almost all access to municipal borrowing. The analysis improves upon previous studies of US municipal fiscal policy by utilizing a panel of annual observations of government fiscal choices between 1970 and 2004 for more than 500 local jurisdictions. In doing so, the research design identifies the extent to which cities act according to a rational tax-smoothing model or are short-sighted in their fiscal decision making over time. In addition to term limits, large and diverse populations are associated with short-sighted fiscal policies.

Chapter 4 applies the probabilistic voting model to construct a measure of party allegiance for jurisdictions in the United States. This political score isolates underlying voter preferences from characteristics that uniquely influence each election and is used to test the core and swing voter hypotheses of intergovernmental transfers. I find that party ties are important determinants of state-to-local transfers in the North- eastern United States and in California. In particular, when Democrats gain seats in state legislatures, they tend to reward their core constituencies. In California, Democrats also reward swing constituencies. When Republicans gain seats in state legislatures, they reward swing constituencies.