Electoral Proximity and Issue-Specific Responsiveness

Abstract Do elections increase responsiveness of legislators to their constituents? Previous studies that examine the effect of electoral proximity have been unable to hold the roll-call agenda constant and control for differences in unobserved covariates between legislators. This paper utilizes a natural experiment in four state legislatures—Arkansas, Illinois, Florida, and Texas—where term length was randomly assigned. This design compares the responsiveness to constituency opinion of those randomly assigned to a two-year term to those assigned a four-year term on different issue areas, like the economy, environment, and crime. I find no evidence for an electoral proximity effect on responsiveness. In addition, in the Illinois State Senate, the causal effect of electoral proximity on responsiveness is measured on several individual roll-call votes, including the legalization of medical marijuana and gay marriage.