abstract: This dissertation assesses the impact of revenue diversification on state revenue growth and volatility and then, the economic, political and institutional factors that predict diversification. Previous studies, taking advice from modern portfolio theory, argue that diversifying a revenue portfolio can stabilize volatility and even lead to faster rates of growth over time. However, levels of diversification are not assigned randomly. Rather, differences among states in diversification might be a consequence of differences in states such as electoral cycles and the presence and strictness of tax limitations. Thus, the research question is: Whether or to what extent has diversification increased revenue growth and decreased volatility when the endogeneity of diversification is considered? Using two-stage least squares and fixed-effects regression models with the data of the 50 states from 1980 to 2011, I examined the impact of diversification, reflecting a state's own political and institutional characteristics (i.e., endogeneity), on growth and volatility. I found diversification was positively related to growth, but a diversified portfolio does not smooth volatility. Furthermore, I found that the level of revenue diversification increased in each year of legislators' terms and decreased in every year of governors' terms. These findings imply that legislators and governors have different preferences for diversification, perhaps due to different opportunities to enhance their reelection prospects. I then investigated the relationship between political leaders' year of the terms and changes in specific revenue sources, the biggest set of reelection opportunities. Selective sales and income taxes were negatively related to every year of legislators' terms. General sales taxes, corporate income taxes, and charges are positively related to every year of governors' terms. The results suggest that legislators focus on their districts or specific interest groups, closely associated with selective sales taxes. In contrast, governors' constituency-driven preferences lead them to be responsible for broader issues such as balancing the state budget, thereby using general sales taxes and charges as methods to do so. As a consequence of these political factors, levels of diversification will change, thereby influencing revenue growth and volatility. / Dissertation/Thesis / Ph.D. Public Administration 2013

The U.S. is experiencing the worst recession since the Great Depression. All levels of government have been hit really hard, this is especially apparent at the local level since services provided at the local level are woven into people’s daily life. Thus, how to “do more with less” is more urgent than ever before. The use of privatization came to surface as a sound solution for deficit-plagued governments as it is thought to be more cost effective and outperform the public sector in most cases. This dissertation contains two empirical chapters that examine determinants of privatization and specify the conditions under which it is optimal to buy and under which it is optimal to produce in-house. Chapter two explores determinants that contribute to the use of privatization at the local level in the U.S. This chapter incorporates spatial technique to perform the analysis, which is a different approach from much of the literature. Empirical results indicate that a local government’s sourcing decision is affected by its nearby local jurisdictions. External stakeholders’ involvement contributes to the use of outsourcing, whereas having a limited supply of service providers impedes it. Chapter three applies a transaction cost economics (TCE) framework complemented with a revenue volatility measure to disentangle the mechanisms that drive public services’ outsourcing decisions. Results suggest that, in general, services with higher asset specificity and higher contract management difficulty are less likely to be outsourced, and a robust and competitive market facilitates the use of outsourcing.