Testing Theories of Government Growth in the Fifty States, 1945 to 1998.

Abstract

This dissertation explores the determinants of government growth in the American states. Understanding the causes of public sector growth is important as the administration of government programs is increasingly devolved to the states. Past research has typically been on national-level growth. However, the variety in state institutional structures, resource bases, and population needs makes for an ideal "comparative laboratory." With some exceptions, state-level studies have had a public finance approach that often exclude relevant political and demographic factors that can lead to increased public sector size. My analysis of state government growth tests thirteen models covering both traditional political explanations as well as explanations found in the public finance literature. I use a pooled cross-sectional time-series research design to determine the causes of growth in forty-nine states for the years 1946 to 1997. The study categorizes explanations of government growth as either responsive or excessive. Responsive explanations suggests that growth is a reflection of the needs and demands of the population. Excessive explanations are those that posit growth beyond that demanded by the citizenry. These explanations are tested in a combined model on both undeflated and deflated state government size. I find strong support for three responsive explanations: Political Needs, Party Control, and Political Culture. The analysis indicates weak support for Wagner's Law with confirmatory evidence confined to urbanization. Two components of Wagner's Law, industrialization and per capita income, are negatively related to government growth. Four excessive explanations gain considerable support: Intergovernmental Grants, Bureau-Voting, Divided Government, and Unfunded Mandates . The analysis indicates that only state government employees have a positive impact on public sector growth. The results reveal that the effect of divided government on state government growth is contingent on the lack of state supermajority requirements for tax increases. There is only weak evidence in favor of the Constituency Size theory. Overall, my study suggests that state government growth is best explained by theories that fall within the excessive category. Of the variables correctly signed, the effects of intergovernmental grants and state employees on government growth are the two strongest in both the undeflated and deflated models.