Abstract: This paper studies how media shape social movements. In the U.S. Civil Rights Movement, I exploit quasi-random variation in television rollout from the 1948 FCC licensing freeze, corroborated by a terrain-based measure of reception. Early access raised the annual probability of protest by 6.9 percentage points, some 68 percent of the mean. In the short run, television coordinated protest around nationally salient events, and the response was larger where demonstrators and police clashed. In the long run, it doubled NAACP membership where a branch existed and shifted beliefs about others’ support for desegregation, not respondents’ own.

for information or a copy of the paper contact patrizia.pellizzari@unibocconi.it