There are extensive qualitative arguments in the corruption literature that relate the prevalence of value-dense natural resources to the pervasiveness of corruption. However, this raises the question of whether natural resources themselves are associated with corruption when other variables are held constant. If so, then natural resources may cause corruption, or there may be a systemic bias against resource rich countries across corruption indices. If the first possibility is true, then future development strategies need to incorporate incentives that discourage the theft of national resources for personal gain. If the second possibility is true, then an adjustment of corruption indices may be necessary.
I conducted an empirical investigation relating the prevalence of natural resources to levels of perceived corruption. Data was collected on variables such as real GDP growth, international trade exposure, and HDI from reputable sources such as World Bank, UN Comtrade, and World Penn Table. I consolidated these data sets and created new variables, such as international trade exposure. Data was collected from over one hundred twenty different countries. Analyses reveal that natural resources are associated with corruption across countries. The results are robust because they survived multiple non-linear transformations and suffer from relatively minor multicollinearilty and heteroskedasticity problems. This project further reveals that international data, particularly normative and distributional data from developing countries, sorely needs updating. Further research in this field requires time series modeling and a comprehensive “societal norms” variable in order to establish causation.
Geoffrey Ketchum, ’12
Bainbridge Island, WA
Majors: International Relations, Economics and Business
Sponsors: Todd Knoop and David Yamanishi