Why Dictators Should Fear Big Cities

From Cairo to Tehran to Moscow, we've seen plenty of examples of dramatic confrontations between autocratic governments and their people in the world's major metropolises in recent years. But is there a measurable relationship between urbanization and anti-authoritarian politics? Ohio State University's Jeremy Wallace argues that there is in a paper for the Journal of Politics:

For the 237 regimes with urban concentration levels above the mean level in the data, the mean duration is 8.6 years and the annual regime death rate is 9.2%. For the 198 regimes characterized by low levels of urban concentration, the incidence rate is only 5.6% and the mean duration is 12.4 years. Regimes with capital cities that dominate the urban landscape fail nearly four years sooner and face 60% greater death rates.

Cities are problems for authoritarian control, the traditional narrative goes, because by concentrating large masses of people, they improve communication networks, allowing anti-establishment sentiment to spread. In physical terms, dense neighborhoods are also ideal centers of resistance, easily blocked by barricades and featuring plenty of hiding places. To counter this, the wide boulevards of capitals like Washington, Paris, and Beijing have a practical as well as aesthetic purpose: allowing easy movement of police or the military in times of civil disturbance. 

Wallace argues that authoritarian regimes typically lavish attention on big cities over rural areas, knowing that this is where resistance is likely to take root first. But, he says, these efforts can ultimately be counterproductive as they encourage more people to move to cities. This is likely to be an issue in coming years for rapidly urbanizing China, which has typically taken the thousands of rural village protests that take place each year far less seriously than any sign of organized opposition in major cities.  

Wallace's research also holds that dictatorships in countries with large centralized capital cities are more vulnerable than those with multiple large urban centers: for instance, Egypt as opposed to Syria. This seems somewhat related to Filipe Campante and Quoc-Anh Do's research on the relationship between capital city isolation and corruption. 

I also wonder whether there's a related connection between urban concentration and protest effectiveness in democracies. The Occupy Wall Street protests of 2011 and 2012 may have spread to dozens of cities throughout the United States, but would they have had a greater impact if U.S. political and economic power were concentrated in one city, as in Paris or Athens? Purely anecdotally, countries like France and Greece do seem to have more of a tradition of street protests directly influencing legislation than the United States has or, for that matter, Brazil.