| This study contributes to an understanding of corruption control by putting forth a causal explanation for successes and failures in curbing corruption in non-democratic countries. The main argument is that regime type shapes the anti-corruption strategies employed, and leads to different results. In addition, this study uses a threefold method of comparing corruption across countries and across time, proposes a general categorization of strategies and a new typology of regimes that govern non-democratic countries.; This study focuses on government corruption, defined as the abuse of government office for private gain. According to this definition, there are three categories of strategies: the checks and balances on power to control the abuse of government office, the regulation of power to reduce corruption opportunities, and the moral bounds of power to constrain the desire for private gain. These strategies develop into different forms and effectiveness under three types of non-democratic regimes: fully legalized regimes, semi-legalized regimes, and non-legalized regimes.; A fully legalized regime, characterized by separation of powers and highly institutionalized bureaucracies, tends to have autonomous checks and balances, two-level statutory regulations (general conduct and working procedures), and persuasive moral constraints. A semi-legalized regime, characterized by a legalized paramount leader and highly institutionalized bureaucracies, tends to have semi-autonomous checks and balances, three-level statutory regulations (general conducts, working procedures, and personal behavior), and coercive moral bounds. A non-legalized regime, characterized by a non-legalistic paramount leader and personalized bureaucracies, tends to have non-autonomous checks and balances, arbitrary regulations, and a leadership role model approach to moral bounds.; The thesis applies this causal mechanism to explain the successes and failures of corruption control in postwar Hong Kong, Singapore, and Taiwan (prior to 1988), drawing evidence from field research in the three entities. The thesis concludes with a discussion of the limits of this research and the questions remained to be further explored. |
