Three essays on corruption
2014-02-21T10:26:23Z (GMT) by
This thesis is a study on corruption from an economics perspective. Corruption is defined as the misuse of power in public office for private gain. Unfortunately, its obscure nature makes it difficult to identify and fight it. The thesis’ aim is to increase our understanding of its mechanics and help toward this direction. In particular, it consists of three different studies, which focus on the causes of corruption, its consequences and measurement. The first section investigates whether political leaders are empirically associated with governance, one of its constituent elements being corruption. The results show that leaders do matter for bureaucratic quality and the rule of law, especially in autocratic regimes. Corruption does not seem to be affected by leaders though, which is an outcome probably driven by the data inadequacies and the inherent features of malfeasance. The second part elaborates on corruption as a driving force behind a salient feature of demographic transition. This is the dilemma that parents face whether to improve their children’s education or increase the size of their families. Corruption can affect their decision since it impedes the provision of public services important for the development of human capital. The model shows that higher levels of corruption increase fertility, diminish human capital and lower growth. Additionally, the model offers an explanation for the empirical observation of volatile fertility rates. Finally, the last study explores the measurement of malfeasance. Initially, a new dataset is built including measures of perceived corruption along with survey questions. A sample of 10 measures is chosen taking into consideration their nature and reliability. An advanced statistical model is applied to this sample in order to construct a new index of corruption, whose scores are accompanied by their measurement errors as well. The strength of the new index is its ability to combine the knowledge of numerous scholars and analysts who examine corruption. In addition, the model is equipped with tools that make feasible the comparison of the constituent measures’ different levels and reliability on the same underlying scale.