The primary runoff alternative that could have changed Oklahoma elections
One of the peculiarities of the election process in Oklahoma is that regular primaries will have a runoff if no candidate receives a majority, but special elections do not have a runoff. The recent special election for Oklahoma County Sheriff, for instance, was won by P. D. Taylor with barely more than one-third of the votes. Considering that Taylor benefited from a spate of negative stories in the news about Mike Christian published just days before the election, including some from the paper that endorsed Taylor, it would not be at all unlikely that a runoff with the top two would have been won by second place finisher Brett Macy. This happens frequently in runoffs, such as last year when primary runner-up David Hooten beat incumbent Carolyn Caudill for County Clerk in the runoff. In fact, there very reason we have the runoff at all is because without it a candidate the majority opposes can win if enough other candidates split the vote.
A runoff ensures that candidate must win a majority at some point in the primary process. But there are a couple of problems with the runoff. One is that it’s expensive; in 2014 it cost Oklahoma about $800,000. Another is that the turnout is usually far lower. For instance, in the runoff for the Democrat nomination for Congress in Oklahoma’s fifth district in 2016, Al McAffrey ‘won’ with 20% fewer votes than the 37% he received in the primary.
There is a way to avoid the costs of a runoff while at the same time ensure that the winner has the support of a majority of voters. It’s called ranked choice voting, or instant runoff. The idea is simple. Instead of just marking the name of one candidate, voters would rank their first, second, and third choices.
If a candidate gets over 50% of the first choice votes then they win. If nobody gets enough to win, then the candidate with the least number of first choice votes is eliminated and the folks who cast those votes get their second choice counted. If there’s still no one with 50%, then the candidate with the least support is again eliminated and those votes migrate to their next preference. So how would this work in practice? Let’s look at a few Oklahoma elections and speculate!
In this hypothetical example Darrell Sorrels was eliminated due to having the least first choice votes and those who supported him had their second choice tabulated. Some folks weren’t interested in any other candidate. Others liked Sorrels but weren’t as negative about Taylor having been Undersheriff for John Whetsel. But most of those who voted for Sorrels wanted someone who wasn’t associated with Whetsel and split between Macy and Christian. After the new total was added up Mike Christian was eliminated. Now the people who voted for him as their first choice as well as those for whom he was their second choice behind Sorrels had their ballots recalculated. Again, some folks didn’t list a second or third choice and so did not add to the total of either remaining candidate, but of those that did a large majority supported Brett Macy because he was not associated with the recently resigned John Whetsel. In this scenario, Brett Macy wins the primary, but only by a small margin. Obviously if we actually held our elections this way the result could be far different.
A key feature of ranked choice voting is that if it were to be applied to primary elections there is no reason to not apply it to general elections as well. With the Libertarians continuing to be on the ballot in Oklahoma as well as the possibility of Independent candidates in any partisan contest, any general election race could have more than two choices and just as in a primary an instant runoff system would ensure the winner is supported by a majority. A few more examples might be of interest.
Disgraced former state Senator Ralph Shortey was first elected in 2010. In the primary he came in second at 38% while James Davenport received 49.94%, just three votes away from winning outright. Because Davenport fell just short of that magically 50% he faced Shortey in a runoff. He lost by 42% to Shortey’s 58%. With a ranked choice system it would be inconceivable that Davenport would not have been able to pick up enough second choice votes to be victorious. In this case the primary runoff put a man in office who would later be forced to resign over child prostitution charges.
Another election where ranked choice voting would clearly have led to a different result was the 2002 governor’s race. With five candidates after the Democrat’s nomination, Vince Orza was way out in front with over 44% in the primary. In second but way back with 28.51% was Brad Henry. There’s no way that a ranked choice system allows Henry to come from that far behind to win. But in the lower turnout runoff a month later Orza saw his number of votes drop while Henry was able to increase his vote total by over 35,000 and narrowly win with 52%. This put Henry in the general election with Republican nominee Steve Largent as well as Independent Gary Richardson. With enough wealth to fund a vigorous campaign, the populist former Republican Richardson probably drew more votes from Largent, although Largent made some missteps. Regardless, with ranked choice voting it’s difficult to see how Largent could have failed to make up the 6,866 gap to pass Henry if Richardson’s 146,000 supporters could have listed a secondary preference.
In short, ranked choice voting eliminates costly runoffs, ensures that winning an election means winning a majority, and eliminates the repeated example of low turnout runoffs essentially overturning the verdict of the initial primary. Another factor to consider is that all those second and third preferences can tell a lot more about what people are thinking when they step in the voting booth, and candidates will be more reluctant to go negative in campaigns when they may need some second or third choice votes in order to win.