Date of Award

Spring 2013

Document Type


Degree Name

Doctor of Philosophy (PhD)


Public Administration and Urban Policy

Committee Director

John R. Lombard

Committee Member

John C. Morris

Committee Member

David Selover


Since the 1980s, privatization has gained increasing acceptance among state governments. Yet, few empirical studies have investigated the factors influencing the level of state government privatization focusing on a multitude of programs and services aggregated across departments. Most prior state level empirical research has emphasized single cases or programs but has not addressed the aggregate level of privatization undertakings across the states. The paucity of empirical research that investigated the amount of state privatization and the drivers thereof created an important gap in the literature that this study attempted to fill. Drawing on historical and contemporary privatization literature, this research examined the influences of variables related to socioeconomic, economic, political, and ideological factors on the level of state government privatization.

This study employed ordinal logistic regression and tested fourteen hypotheses and four state comparative models (socioeconomic, economic, political, and ideological) and developed a fifth model of best fit. The bivariate results show that all but state pension spending and political culture variables were insignificant. The multivariate results indicate that in the socioeconomic model only state pension spending variable was significant in the expected direction and the hypothesis was supported. In the economic model, state per-capita spending, state fiscal capacity, and state deficits were significant in the opposite direction than expected and the hypotheses were not supported. With the exception of the political culture, all the variables in the political model were insignificant. The traditionalistic political culture was significant at both the bivariate and multivariate level, but in the opposite direction than expected and the hypothesis was not supported. The moralistic political culture was significant in the expected direction, but its significance disappeared in the model of best fit. All the variables in the ideology model failed to achieve statistical significance. In general, the analysis reveals that a large part of the variance in the dependent variable remained unexplained.

Overall, the findings of this study suggest that socioeconomic, political, and ideological factors are less likely to influence state government privatization. On the other hand, the findings do suggest that economic factors matter; although the influences of the significant variables in the economic model were in the opposite direction than expected, the findings nonetheless appear to provide tentative support to the argument in the literature that economic factors are more likely to influence the level of privatization by state governments.