Replication Data for: How Durable Are Experimental Increases in Women’s Representation?

Do field experimental interventions produce durable changes in gender representation? This article examines whether the experimental treatment effects in Karpowitz, Monson and Preece’s (2017) study “How to Elect More Women: Gender and Candidate Success in a Field Experiment” persist over time. The initial experiment showed that a single letter from Republican Party leaders could significantly increase women’s election to state delegate positions from neighborhood precincts. Two years later, differences between treatment and control conditions evaporated, though not because of regression to the mean among treated precincts. Instead, treated precincts largely retained earlier experimental gains, but the treatment effect size was smaller because of increases in women’s representation in the control condition. We examine four possible explanations for both the change in the control condition and the enduring patterns in treated precincts, including spillover, attitude change among caucus-goers, rates of women’s incumbency advantage across conditions, and the effects of larger trends in political participation.