The Effect of the Internet on Political Mobilization
This project studies empirically the effect of the Internet on protests worldwide. We compile a novel panel dataset that combines geo-referenced data on Internet quality and weekly protests for over 18,907 subnational (ADM2) districts from 236 countries and the years 2006-2012.The Internet penetration data was constructed by combining over a trillion (1.5 X 10^12) IP activity(offline/online) observations to a commercially-available, IP-geolocation library. Our identification strategy exploits random weekly variation in global Internet latency to identify the causal effect of the Internet on local protests. According to our estimates, latency-adjusted Internet increases the occurrence of local protests. We show that most of the variation in the effect of the Internet on local protests comes from national differences in political institutions and local differences in Internet penetration.