Internet Explorer 11 is not supported

For optimal browsing, we recommend Chrome, Firefox or Safari browsers.

Should Finishing First Guarantee a Candidate Victory?

North Carolina is trying a radical new vote count system. It’s an instant runoff — one that doesn’t require a second trip to the polls.

By Josh Goodman, Stateline Staff Writer

For a single seat on the state’s Court of Appeals, a group known as Equality North Carolina has endorsed Cressie Thigpen. It’s also endorsed Stan Hammer. And it’s endorsed Pamela Vesper, too. The gay rights group isn’t being indecisive. Instead, it’s acting in the spirit of the unusual election North Carolina is holding in November, one in which voters will get to register a first choice and a second and a third. The contest this fall for North Carolina’s Court of Appeals will be the first statewide general election in American history to use a system known as instant-runoff voting.

IRV, as it’s called, is just one of a menagerie of alternatives to the standard American “plurality voting” system — alternatives based on the notion that asking voters only for their first choice is a flawed measure of the public will. Under IRV, whichever candidate receives the fewest first-place votes is eliminated and his voters’ second choice is added to the tally. This process continues — with second and third and perhaps even fourth and fifth picks being counted — until someone breaks 50 percent and is declared the winner. In this way, IRV mimics a traditional runoff election, without the need for voters to return to the polls a second time.

IRV is starting to catch on. It’s used in local elections in several major American cities. It’s used for elections to the Australian House of Representatives. It’s even used to pick college football’s Heisman trophy winner and, as of last year, the Oscar for best picture.

Still, American state elections represent an almost entirely new frontier. A few states used IRV and similar voting systems for primaries many decades ago, but all had abandoned them by the 1930s. Today, a few states permit military and overseas voters to rank candidates as a way for them to participate in runoffs without having to submit a second ballot. But North Carolina is the first state to apply IRV to all voters for a general election.

Supporters hope that the North Carolina vote will help hasten a tipping point in which everyone from voting machine manufacturers to legislators to the electorate itself embraces IRV. But that will only happen if this year’s North Carolina experiment is deemed a success. To make IRV work, state election administrators will have to explain to the public that they’re to use a new method of voting in an election they may not know much about. Then, the administrators themselves will have to figure out how exactly to count the votes.

Coalition or confusion

The event that prompted North Carolina to try IRV was a special election for the state Supreme Court in 2004. A vacancy occurred that August, leaving the state without time to hold a primary. In the eight-candidate race, the winner, Paul Martin Newby, received less than 23 percent of the vote. More than three-quarters of the electorate preferred someone else. In the nominally nonpartisan contest, Newby’s support from the North Carolina Republican Party helped him earn the relatively small plurality of voters he needed to win, even though most voters might have wanted a less overtly partisan candidate.

IRV is designed to prevent results like that. Proponents make the case that their system tends to elect candidates who are broadly acceptable to the public. They also say that IRV can lead to more congenial campaigns, since candidates are wary of alienating opponents’ supporters. And, IRV offers voters an opportunity to support third-party candidates without fearing that they’ll be throwing their votes away.

Those arguments carried enough weight in the North Carolina Legislature that lawmakers in 2006 authorized a pilot project in which local governments would be able to use the system for municipal elections. The cities of Cary and Hendersonville have already tried it. Legislators also decreed that future special elections for state judges would be conducted with IRV, if the calendar didn’t allow for a primary. That’s what happened this year, when the U.S. Senate confirmed a judge on North Carolina Court of Appeals for the federal bench, creating an opening.

When you move below the state level, you find clear evidence of IRV gaining momentum. The cities of Minneapolis, St. Paul, San Francisco, Oakland and Memphis all have chosen to adopt it in the last decade. The biggest reason for the shift was the 2000 presidential election in Florida, which likely would have gone to Al Gore under an IRV system, had Ralph Nader voters listed the Democrat as their second choice. The reviews have been generally favorable, but not universally so. Burlington, Vermont, for example, abandoned IRV earlier this year after a controversial mayoral election.

North Carolina is a test on a much larger scale and one that could come with larger consequences. North Carolina is one of 10 states that still hold primary runoff elections. These runoffs are costly and typically feature low turnout. IRV is a potential, if unproven, alternative. “North Carolina is doing something new,” says Gary Bartlett, executive director of the North Carolina Board of Elections, “and I just hope that voters accept it on election day.”

Bottom of the ballot

In that regard, Bartlett has two major challenges. The first is educating voters. On this score, no one is especially optimistic. Court of Appeals races never draw much attention in North Carolina, a state that has largely avoided the costly, divisive judicial elections that have become common elsewhere. The contests appear far down the ballot, leading many voters to skip them entirely. “It’s going to be after the dogcatcher race, I think,” says Chris Dillon, one of the candidates for Appeals Judge. There are thirteen candidates, but they haven’t tried to take advantage of the new system by running as teams or offering up cross-endorsements which might have drawn attention to IRV — although some groups, like Equality North Carolina, have.

The State Board of Elections has circulated a voter guide with information on the system, but doesn’t have money for a TV or radio blitz. Officials say their focus will be on providing information at the polls and making sure poll workers know the procedure, so that voters can learn about it there. Of course, even if voters don’t realize they’re participating in an IRV election, it likely won’t be a disaster. They’ll still be able to vote for one candidate as their top choice, just as they would in a conventional election.

The second challenge is the more pressing one: counting the votes. The truth is that instant-runoff voting in North Carolina won’t be instantaneous at all. The state can’t start counting votes in a runoff until it knows which candidates finish in the top two. It can’t confirm which candidates are in the top two until the election is certified ten days after ballots are cast. Critics contend that this delay will call into question the integrity of the election.

Even after that, the process won’t be quick. North Carolina counties that rely on electronic voting machines lack any certified software for counting votes in an instant-runoff campaign. State election officials are still weighing whether to print readouts of the ballots, which would have to be counted by hand, or to use an electronic spreadsheet system that isn’t currently certified. A hand count would be cumbersome, even though the state is allowing voters to rank only their top three picks. In its purest form, IRV allows voters to rank as many candidates as they want. “My feeling on instant-runoff voting is that the General Assembly was premature to mandate it,” says George Gilbert, Guilford County’s director of elections. “We didn’t have the software to do it. They might have consulted with somebody.”

Even if North Carolina’s election goes poorly — if voters are confused or vote-counting is a long slog — that may not reflect inherent drawbacks of IRV. Over time, voters would be likely to get a handle on the system. Election machines could be designed and software written to count IRV votes quickly and safely. Still, this is something unprecedented in statewide politics. If things go wrong there may not be a second chance to get them right.

Andy Kim is a former GOVERNING staff writer.
Special Projects