We compare two widely publicized measures of state electoral integrity in the United States: the Electoral Integrity Project's 2016 U.S. Perceptions of Electoral Integrity Survey and the Pew 2014 Elections Performance Index. First, we review the theoretical and empirical differences between the two measures and find that they correlate at a surprisingly low level across the states. Second, given this low correlation, we examine the component parts of these indices and find that both are capturing multiple dimensions. Third, we examine how the components and the individual indicators that comprise each measure are linked to citizens' stated perceptions about electoral integrity. Throughout the article, we articulate a set of preemptive recommendations that urge researchers to be cautious and deliberate when choosing among measures of electoral integrity to use in future empirical studies.

American state politics scholars have generally relied on Ranney's measure of the partisan composition of state legislatures and governors' offices to evaluate competition between parties for control of state government, and Holbrook and Van Dunk's measure of the competitiveness of individual state legislative elections to evaluate the degree of electoral competition in a state. Both measure “competition” and were previously correlated with one another, so researchers might be tempted to consider them two measures of the same concept. This would be mistaken, however, because they are measuring two distinct concepts. We use new data on state legislative partisan balance and election returns to compute (and make publicly available) the two measures of competition from 1970 to 2003, a time span that is significantly longer than any previous study. We show that the relationship between the two measures has drastically changed over the last 30 years. Although the two measures were positively correlated in the 1970s and 1980s, they are now (as we might expect, given they are different concepts) negatively correlated. We investigate one possible explanation for this change and conclude by discussing a set of practical recommendations for scholars who plan to incorporate a measure of competition in future studies.