| Abstract: | The relative importance of selection and incentives is essential for understanding how elections structure politicians’ behavior. I investigate the relative magnitudes of these two effects in the context of US House members’ constituency communication. Consistent with previous research, I find that there is a negative cross‐sectional relationship between electoral security and the intensity of constituency communication. The negative relationship holds in a panel‐data setting where only within‐legislator variation in electoral security is used to identify the effect of electoral security on legislator behavior. Due to the likely presence of myopic voters, the impact of electoral security increases as the election approaches. Point estimates suggest that the total effect is almost entirely driven by incentives, and I am able to reject the hypothesis that the incentive effect is zero at conventional levels of statistical significance. |
