Essays of Political and Cultural Institutions



Journal Title

Journal ISSN

Volume Title



This dissertation comprises two essays. The first examines the role of a bureaucratic institution—the civil service system—in dampening political budget cycles across U.S. state governments. Using new data on state-specific civil service reforms throughout the twentieth century enables me to account for cross-state institutional heterogeneity, as well as to exploit within-state variation in the precise timing of reforms. I uncover sizable electoral cycles before a state replaces its patronage system with a system of merit-based recruitment and civil service protections for state government employees. Once the civil service system is introduced, the cycles diminish significantly or disappear completely. Reform-induced heterogeneity in budget cycles is especially pronounced in expenditure categories with high voter appeal. For example, states with the patronage system spend, on average, 27.5 dollars more per capita on roads and highways (an increase of 8.5 percent relative to the sample mean) during the two years prior to gubernatorial elections compared to the two years following the elections, whereas states with the civil service system do not exhibit such cycles. The evidence suggests that reformed bureaucracies exert considerable influence on politicians’ spending decisions.

The second essay examines the role of a cultural trait—individualism-collectivism—in shaping microeconomic preferences at the individual level. I adopt the “priming” method from social psychology to exogenously vary the salience of individualism and collectivism in laboratory settings. The findings indicate that subjects primed on collectivism make less risky and more patient financial choices, and report lower self-confidence than subjects primed on individualism. Findings from a supplemental experiment indicate that making collectivism salient does indeed strengthen subjects’ collectivistic perceptions, lending support to the interpretation of the main results as the effects of collectivism. Finally, I find that Hispanics and blacks are especially sensitive to being primed on collectivism, while Asians and whites are the least sensitive. This result is consistent with the evidence that Asian and white subjects in the sample are the most and the least group-oriented, respectively.



Political budget cycles, Politics, Civil Service, State Goverments, Individualism-Collectivism, Culture, Priming, Risk and Time Preferences, Confidence