Isolated and Poor: the cost of remoteness from the capital city

This paper investigates whether areas isolated from the capital city are less de- veloped economically in Sub-Saharan Africa. We apply a boundary-discontinuity design using national borders that divide pre-colonial ethnic homelands to obtain quasi-experimental variation in distance to the national capital city. Based on night- lights and geocoded surveys, we find that a one percent increase in distance to the capital city causes a decrease in the probability of detecting nightlights by 3 percent- age points and a reduction in household wealth corresponding to 3.5 percentiles of the national wealth distribution. Our results suggest that a lower provision of public goods in isolated areas is a key link between remoteness and economic performance. Despite receiving worse services, people who are isolated exhibit a higher level of trust in their political leaders. We interpret this as pointing towards dysfunctional accountability mechanisms that reduce the incentives of state executives to invest into isolated areas.