FairVote Minnesota (January 2009) - Minnesota’s unresolved US Senate race highlights the increasingly competitive nature of partisan elections in the state and raises a number of questions about our voting system, the laws that guide our electoral process and the impact – if any – that Instant Runoff Voting (IRV) might have had in the 2008 Senate race.

Q. How would IRV have made a difference in Minnesota’s US Senate race?

A. It’s very likely that IRV would have had a significant impact on Minnesota’s 2008 US Senate race. It would have ensured that a winner would have received the affirmative support of a majority of the voters.  (IRV likely would have also been impactful in the 3^rd and 6^th Congressional Districts races and in the state legislative races in Districts 41A and 51A, all of which were decided by a minority of voters.) 

IRV likely would have encouraged the candidates to mount more issue-focused and positive campaigns in order to position themselves as appealing second choices.  This would have worked to eliminate some of the negative campaigning that became a hallmark of the race.

IRV would have eliminated the “spoiler” effect that occurs in plurality-take-all elections when there are three or more candidates on the ballot. IRV provides all candidates a meaningful opportunity to influence the tone and substance of the race and, in turn, gives voters the opportunity to cast a vote for their first choice without concern that they might throw their vote away or be helping elect their least preferred candidate.  

Q. Would IRV eliminate recounts?

A. No, IRV does not prevent ties and near-ties in close elections.  Accordingly, a recount could still be triggered when the margin of victory in an IRV election is 0.5 percent or less. 

However, IRV would make recounts less likely in the final round because, in effect, there are a greater number of votes cast in the final round for the top two candidates than in the initial round. According to a recent study on election recounts conducted by the national FairVote organization, the greater the number of votes cast, the less likelihood of a recount.

Q: How does IRV work?

A: In an IRV election, voters rank candidates in order of preference. If no candidate reaches the majority threshold in round one, an instant runoff occurs in which votes for less popular candidates who are eliminated are reapportioned to more popular candidates based on the voter’s ranked preferences.

Using the Franken-Coleman race as an example, the recount canvassing report shows that a total of 2,424,637 votes were cast for Coleman or Franken in the first round; each candidate received 42 percent of the vote. An additional 462,675 votes were cast for Dean Barkley and two other candidates. In an IRV election, the share of these ballots where the voter indicated Coleman or Franken as backup choices would have been counted for these top two candidates in the final round, increasing the likelihood of a decisive winner.

The real benefit of IRV is that it produces winners with the affirmative support of a majority of the voters and it does so on Election Day in November, when turnout is highest and most diverse. Thus, even if there is recount in an IRV election, the outcome will be supported by a majority of the voters.

Q. Is IRV better than a Georgia-style delayed runoff system?

A.  Yes, because IRV consolidates two-round runoffs into a single election.  This single-step election eliminates the need for a costly second election and brings a timely end to expensive campaigning. 

Two-step runoffs also drive down voter turnout.  This was seen in the 2008 Georgia Senate race, a state that requires a December runoff when no candidate reaches the 50 percent + 1 threshold on Election Day. Georgia’s second election was held on December 2^nd and just 41 percent of voters returned to the polls for the decisive election, compared to 76 percent of voters who came out on Election Day in November.

Q.  Wouldn’t an IRV ballot be confusing to voters and cause more ballot errors?

A.  Claims of voter confusion and folly under both the current system and IRV are grossly exaggerated. Certainly, there are some voters who mismark their ballots and this is true under any voting system. However, the spoiled ballot rate is very low in both current plurality elections and IRV elections where they are in use across the country. 

In the Senate race, there were fewer than 11,000 rejected ballots due to ballot error (after re-qualifying some 1,350 rejected absentee ballots during the recount), well below 1 percent of the 2.9 million ballots cast. And, many of these rejected ballots were disqualified due to signature error or administrative error and not to mistakes in filling in the ovals.   

Election data show that spoiled ballot rates under IRV are also low, below 1 percent and exit polls that survey first-time IRV voters show that the vast majority of voters understand how to use IRV. 

Good ballot design, user-friendly instructions for precinct and absentee voting and strong voter education are the ingredients of a successful election using any voting method.

Q.  Is the recount law necessary?

A. FairVote Minnesota supports Minnesota’s recount law and believes that the recent recount process validates the accuracy of the state’s voting equipment in counting ballots and demonstrates that our election laws and the administrative processes to carry out a recount are well-conceived and properly carried out.  Should a recount be needed under IRV procedures, clear procedures should be in place just as they are now for our current elections.

Hand-counted IRV elections are routine in democracies like Ireland; depending on the number of workers, manual IRV counts can be accomplished quickly. The 1.2 million votes in the 1997 presidential election in the Republic of Ireland, held using instant runoff voting, were counted in a single day.  Protracted recounts could occur not because of the IRV voting method, but because of factors such as disputed ballots, absentee voting procedures and administrative errors.

We have learned from the recount that improvements are needed in our electoral process, in particular absentee voting and voter registration procedures, to minimize voter error and the number of rejected ballots. We support efforts to make these election procedures work better.

We are also encouraged by the growing public understanding of the need for broader electoral reform such as Instant Runoff Voting to improve the democratic process.