Dynamics of Economic Corruption

Abstract

Corruption, especially in developing countries, has been shown to have large persistent effects on inequality and economic growth. Efforts to eradicate or suppress corruption over the last two decades have floundered. This dissertation argues that urban population density is a major underlying factor of economic corruption that previous studies have never considered before. Increased population density communicate bribe success information rapidly and resistance to it becomes muted as more success is observed within a population. Urban density has not been considered before because while data for urban populations exist, data for urban extents did not. This dissertation uses satellite night-time lights data to get an estimate of urban extent and uses it to calculate urban population density. This research shows that an increase in urban density by 100 people per square kilometer has the same effect on corruption as a decrease in the per capita GDP by $172 (2005 dollars). Using spatial effect models we show why a small, smooth change in behavior parameters lead to large scale, sudden cascading effects in corruption across the population. Also using network theory models we show that most real worlds are small in the number of links required to traverse a population and it gets easier to find successful links between a briber and the bribed as urban density increases.