| Corrupt practices in public sector are observed in every economy at some extent but they are much more common and rampant, especially, in less-developed and developing countries. The aim of this dissertation is threefold. First, we explore if there is any technique that we can use to generate time series index for corruption. Second, we investigate the association between corruption and economic growth in both theoretical and empirical frameworks. Lastly, in a game theory framework, we analyze the issue whether corruption destroys allocative efficiency.; By utilizing MIMIC (multiple indicators, multiple causes) technique, we generate a time series index of corruption for 22 OECD countries for the period 1975–1993. The results indicate that the highest corrupt practices exist in Japan, Portugal, Spain, Greece, South Korea, and Turkey whereas the lowest ones are present in Netherlands, Norway, New Zealand, Denmark, Great Britain, Sweden, and Ireland. On the other hand, France, USA, Finland, and Austria experience moderate corruption level. Also, Canada, Belgium, Australia, Iceland, and Italy improve their rank standings over the estimation period.; We found a significant strong correlation between our estimated index and commercially available ICRG index. We also identified a statistically significant negative impact of our estimated corruption index on the investment in a panel regression but not on the economic growth.; In a Lucas (1988) type of growth model we show that economies with corruption experience lower economic growth than those without corruption. This theoretical implication is supported by empirical finding. In OLS and TSLS analyses, we found a negative association between corruption level and growth rate of real per capita GDP. This finding mostly remained valid when we used two different corruption indexes as well as other control variables.; We also consider a bribery game under a pre-existing environment of alleged nepotism where players initially suspect that the official has a favored firm. We show that even if there is no actual nepotism, observed nepotism in a bribery game may disrupt allocative efficiency. We show that this result holds in several different forms of a bribery game. |
