OIL CURSE AND INSTITUTIONAL CHANGES: WHICH INSTITUTIONS ARE MOST VULNERABLE TO THE CURSE AND UNDER WHAT CIRCUMSTANCES?

摘要

This article extends recent analyses linking the alleged oil curse to a broader set of institutions (13 in number) than democracy, the institution that has received the most attention in the literature. It does so using panel data for over 100 countries between 1975 and 2005, wherever possible, and compares the effects obtained with several different measures of both the importance of oil and experience in the industry and of the interactions between them. Most importantly, instead of simply examining the effect of oil and experience in the industry on the contemporary levels of these various institutions, this study focuses on the effects on changes in the various institutional indicators from one decade to another. While not surprisingly our results reveal considerable sensitivity in the effects of oil resources, oil experience, and interactions across different specifications, they also suggest a number of important findings. The most robust of these are the significant negative effects of oil rents on bureaucratic quality and on socioeconomic conditions. We also find that the number of years since peak oil discovery has a positive effect on government stability, but a negative one on bureaucratic quality. When interactions are allowed for, still more negative effects on institutions are identified, at least partially re-enforcing several of the institutional links in the oil curse hypothesis.