New Data Support Use Of Instant Run-Off Voting
For Immediate Release
New data collected as part of a North Carolina State University study during the 2009 municipal election in Hendersonville, N.C., show that voters prefer instant run-off voting (IRV) to traditional voting – a finding that may build support for IRV. The use of IRV precludes the need for a second run-off election, saving voters time and providing election results more promptly.
According to an analysis of exit polling data from Hendersonville’s November election, where voters used IRV for the second time, most voters said it was easy to understand and that they preferred ranking candidates to choosing a single candidate. The exit poll was designed by Dr. Michael Cobb, an associate professor of political science at NC State, and was funded and administered by the North Carolina State Board of Elections.
IRV is a method where voters rank candidates in order of their preferences rather than choosing a single most preferred candidate. Ranking is used to avoid holding a second, run-off election between the top two vote getters at a later date. Hendersonville is currently the only city in North Carolina testing IRV.
All but 9 percent of voters who were interviewed said that IRV was either somewhat or very easy to understand, and 53 percent preferred ranking candidates while 37 percent preferred voting for only one candidate (another 10 percent had no opinion or liked both equally). “Voters who knew in advance that ranking would be used, who found it easy to use, and reported ranking the mayoral candidates were significantly more likely to prefer ranking,” Cobb says.
IRV is also being used in other metropolitan areas, such as San Francisco, but exit polls in these cities have sometimes found non-white and lower-income voters less supportive of using IRV. For example, some voters have reported that they find it more difficult to use. According to Cobb, there were no differences across groups of voters in Hendersonville, such as gender, age or income, although race could not be reliably compared because more than 96 percent of respondents were white.
Cobb also cautions that, while voters support IRV, their support is not strong. When voters were presented with a list of reasons cited by advocates to defend IRV, a majority of voters polled did not personally agree with any of them. “Voters seemed indifferent to these traditional justifications for using IRV and simply liked or disliked the experience of ranking for their own reasons,” Cobb says.
For example, just 27 percent of voters surveyed who preferred ranking also agreed that they would be upset if they would not be allowed to rank candidates in the future. On the other hand, 51 percent said that voters should be allowed to rank candidates for statewide offices, while just 20 percent disagreed.
The exit poll had a sample size of 322 voters and a margin of sampling error of +/-5 percent. Respondents were recruited by randomly asking voters to fill out an exit poll after casting their ballots at precincts during the day of the election and during early voting. Complete poll results can be downloaded at http://faculty.chass.ncsu.edu/cobb/.