Twitter Adoption in Congress

We study the early adoption of Twitter in the 111th House of Representatives. Our main objective is to determine whether successes of past adopters have the tendency to speed up Twitter adoption, where past success is defined as the average followers per Tweet — a common measure of “Twitter success” — among all prior adopters. The data suggests that accelerated adoption can be associated with favorable past outcomes: increasing the average number of followers per Tweet among past adopters by a standard deviation (of eight followers per Tweet) accelerates the adoption time by about 112 days. This acceleration effect is weaker for those who already have adopted Facebook and those who have access to information about a large number of past adopters. We later find a positive relationship between an adopter's own success and the success of adopters preceding him/her. Thus, there may exist benefits associated with adopting Twitter based on past successes of others. In general, the patterns we find are consistent with predictions generated by a simple model of adoption delay with learning.