The Elections Performance Index, or EPI, tracks 17 distinct indicators of election administration effectiveness. A state's overall average is calculated from its performance on all indicators, relative to all states across comparable federal election cycles—either presidential or midterm. A state with an average of 100 percent in a presidential year, such as 2012, would have the best value of any state on every indicator across all presidential years (in this case, 2008 and 2012), and a state with an overall average of zero in any given presidential year would have the worst value of any state on every indicator across all presidential years. This also holds true for midterm years, which currently include 2010 and 2014. Because these averages are based on the performance of all states, even if a state had a 100 percent average there still is room for improvement in future elections.
Further, because the overall EPI average is an average of all indicators, improvement on any one indicator, relative to other states, will not change the number much. A state that performs well on one indicator but close to average in every other will not, overall, perform at a level that is significantly different from the national average. Consequently, for a state's overall average to increase from one year to another, that state must do better across many indicators, relative to other states.
Presidential and midterm elections are not comparable using this interactive report because of significant differences in electorates, available data, and baseline scales. Additionally, of the 17 indicators of election performance, data are not collected for two—voting wait time and residual vote rate—during midterm elections, leaving 15 metrics and points of comparison for 2010 or 2014. For more information, please see the methodology.