Many governments in the United States hold elections on days other than national Election Day. Recent studies have argued that the low voter turnout that accompanies such off-cycle elections could create an advantage for interest groups. However, the endogeneity of election timing makes it difficult to estimate its causal effect on political outcomes. In this paper, I develop a theoretical framework that explains how changes to election timing affect the electoral fortunes of organized interest groups. I test the theory by examining the effects of a 2006 Texas law that forced approximately 20 percent of the state's school districts to move their elections to the same day as national elections. Using matching as well as district fixed effects regression, I estimate the causal effect of the switch to on-cycle election timing on district teacher salaries, since teachers and their unions tend to be the dominant interest group in school board elections. I find that school districts that were forced to switch to on-cycle elections responded by granting significantly lower salary raises to teachers, supporting the hypothesis that school trustees were less responsive to the dominant interest group after the switch.