To calculate the efficiency gap for legislative elections in 2014 and 2016, Oklahoma Watch compiled the total votes cast for each district’s Republican and Democratic candidates, plus the total combined vote.
For uncontested races, a total vote number was assigned based on the average total votes cast in contested races. The party that didn’t field a candidate in the race was then given 25 percent of the votes and the party with the uncontested candidate received the remaining 75 percent.
The creators of the efficiency-gap method note that accounting for uncontested races is one of the limitations of the formula.
Ideally, they suggest developing modeling to “use variables that have been shown in the past to predict vote share” to assign more accurate figures. But using the 25-75 percent split is among the “defendable” strategies they argue can be used.
Their modeling, along with Oklahoma Watch’s calculation, also excluded votes for third-party candidates. Unless there is a scenario where three-candidate contests are the norm, they maintain that just calculating votes for Democratic and Republican candidates is statistically valid. No third-party or independent candidates in Oklahoma won a significant share of votes in 2014 or 2016 House or Senate races.
Oklahoma Watch then calculated “wasted” votes, which are votes that play no role in determining the final outcome. For winning candidates, wasted votes are those above 50 percent plus one vote. For losing candidates, all votes count as wasted.
To compute the efficiency gap score, the total number of Democratic candidates’ wasted votes was subtracted from all Republicans’ wasted votes. The result – the difference between the two – was divided by the number of total votes to get a percentage.
Republicans won the most seats in the House and Senate in both years. The gap was 12 percent in 2014 and 10.5 percent in 2016 in the House, and 10.5 percent in 2014 and 14.4 percent in 2016 in the Senate.