Institutions and Urban Concentration: The Role of State Capacity

Abstract

Previous research has argued that non-democratic institutions tend to facilitate high levels of urban concentration due to government policies that favor primate cities. However, the mechanisms by which governments favor their large cities are still unclear, and whether favoritism predominately takes the form of lower extraction rates or more public goods. Differences in mechanism are important because disparities in public goods have different consequences for country-level growth than differences in extraction rates. We present a simple model where non-democratic governments are incentivized by rents to provide public goods. In this model, government favoritism to a specific region will take the form of superior public goods, when government coercive capacity is generally low and lower extraction rates when coercive capacity is high. Consequently, the model predicts that cities which are small/less productive will be favored in low capacity countries, but these results will be attenuated as capacity increases. Results from a panel data analysis of urban concentration for most of the large cities in the world for the period 1955-2010 are consistent with the model’s predictions. A calibrated version of our model implies that the effect of favoritism on urban concentration is negative for most countries of the world due to most countries having low capacity, resulting in favoritism towards small cities. Our results imply that under non-democratic governments, primate capital cities tend to grow faster than other large cities due to an increase in the provision of public goods. Using city level estimates of developed land per capita as a measure of public goods, we find evidence consistent with this public goods channel.