In poor countries, the discovery of natural resources can have two deleterious effects. First, the traditional economic resource curse can occur, where foreign demand for the resource (oil, gold, etc.) increases the real exchange rate, leading to less investment in the non-resource sector, leading to deindustrialization. Second, recent authors have hypothesized a political resource curse, where the discovery raises the benefit of being a politician, increasing corruption and diverting resources from the real economy into rent-seeking games. Vicente considers the discovery of oil in the late 90s in Sao Tome and Principe as a natural experiment of this theory. Surveying residents both in Sao Tome, and in the neighboring islands of Cape Verde – both possessing similar histories, politics, and colonial institutions at the time of the oil discovery – he attempts to tease out precisely what sort of corruption might increase. Through a difference-in-differences approach, he finds clear evidence in Sao Tome that “status” goods (like government scholarships) and agencies closely linked to the central government saw large increases in perceived corruption, whereas agencies without such direct links (such as hospitals) saw no such gain.
http://www.pedrovicente.org/oil.pdf (Final WP version; a published version is in JDE 92 (2010))