Working Paper: NBER ID: w24495
Authors: Camilo Garcajimeno; Angel Iglesias; Pinar Yildirim
Abstract: How do social interactions shape collective action, and how are they mediated by the availability of networked information technologies? To answer these questions, we study the Temperance Crusade, one of the earliest instances of organized political mobilization by women in the U.S. This wave of protest activity against liquor dealers spread between the winter of 1873 and the summer of 1874, covering more than 800 towns in 29 states. We first provide causal evidence of social interactions driving the diffusion of the protest wave, and estimate the roles played by information traveling along railroad and telegraph networks. We do this by relying on exogenous variation in the rail network links generated by railroad worker strikes and railroad accidents. We also develop an event-study methodology to estimate the complementarity between rail and telegraph networks in driving the spread of the Crusade. We find that railroad and telegraph-mediated information about neighboring protest activity were main drivers of the diffusion of the protest movement. We also find strong complementarities between both networks. Using variation in the types of protest activities of neighboring towns and in the aggregate patterns of the diffusion process, we also find suggestive evidence of social learning as a key mechanism behind the effect of information on protest adoption.
Keywords: No keywords provided
JEL Codes: D71; D83; N11; N31; N71; O18; Z12
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| meetings in neighboring towns (Z00) | response strength of towns (R59) |
| one additional crusade event in a neighboring town linked by rail (N93) | probability of holding a crusade event in the following 10 days (C69) |
| information transmitted through the telegraph network (L96) | protest diffusion (D74) |
| telegraph-mediated information (L96) | effectiveness of telegraph-mediated information (L96) |
| presence of both rail and telegraph access (R49) | probability of holding a crusade event (C73) |