The Honorable Nancy Skinner
P.O. Box 942849
Sacramento, CA 94249-0014
Re: Your concerns about instant runoff voting
Dear Assembly Member Skinner,
I am executive director of FairVote, a non-profit, nonpartisan organization that studies election laws with the goal of changes that better respect every vote and every voice. I would like to congratulate you on your service as an elected official and respond to the concerns that you expressed regarding instant runoff voting in your 5/22 statement on AB 1121 (the legislation authored by Assembly Member Davis that would provide 10 local jurisdictions in California with the option to use ranked voting systems).
Our organization (www.fairvote.org ) has been at the forefront of advocating electoral systems that better serve the needs of America’s diverse communities. We have extensive expertise in electoral systems design and areas such as voting rights and electoral competitiveness. Two of our board members reside in California, and we have a long history of education in your state. Given our considerable experience and expertise in this area, we believe we can contribute to addressing your stated concerns.
The two specific concerns that you expressed regarding instant runoff voting (IRV) were: 1) potential effects on voters who may not be well informed as to how the system works; and 2) the potential ability of independent expenditure committees and others to manipulate such a system.
First, in terms of voters understanding instant runoff voting, the results in the jurisdictions that have used ranked ballot methods have been subjected to vigorous testing and analysis. We have the benefit of seven exit polls, three conducted in San Francisco (CA) and four more conducted in Cary (NC), Hendersonville (NC), Burlington (VT), and Takoma Park (MD). Of the San Francisco surveys, two were conducted by San Francisco State University’s Public Research Institute, and a third was by the Asian Law Caucus, for the 2004, 2005 and 2006 elections, in which voters were asked their thoughts and opinions about instant runoff voting. These surveys have polled many thousands of voters on their level of understanding about instant runoff voting as well as their preference for it or their prior system — see http://www.instantrunoff.com/e…
These surveys show that voters overwhelmingly understood ranked voting, with responses for good understanding ranging from 86% to 95% of voters surveyed, depending on the jurisdiction (and I would suggest the remaining number of voters is smaller than those who would say they understand the Electoral College system and doesn’t interfere with their ability to cast an effective ballot). The Public Research Institute and Asian Law Caucus surveys also learned that voters used the system effectively, as measured by their usage of multiple rankings or the maximum number of rankings.
Additionally, voters also expressed a strong preference for using ranked voting over their old systems, with preferences ranging from 71% to 89% preferring IRV over their old system, depending on the jurisdiction. The San Francisco surveys polled on a wide range of questions, and the detailed findings support the fact that every single demographic in the city — defined by where they live and their race, age, gender, party and political philosophy — preferred IRV to the old runoff system. Additionally, voters two-to-one perceived the instant runoff voting system as more fair than the prior two-round runoff system. So from the standpoint of
voter understanding and acceptance, ranked voting has established an impressive track record.
In addition to the exit poll surveys of voters, the extremely low rates of invalid ballots in IRV elections suggests that voters understood the mechanics of using the system. In a recent election in Aspen, 100% of the ballots cast for Mayor were valid, i.e. no voter failures. In that same election, 99.1% of ballots cast for city council using a novel variation of ranked voting were valid. In two IRV elections in Burlington, 99.9% of the ballots were valid in the first election and 99.99% were valid in the most recent election after a recount. In San Francisco, 99.6% of ballots were valid, higher than for non-ranked elections.
All of these jurisdictions engaged in at least some voter education and outreach prior to their IRV election — sometimes quite inexpensively, as in the case of Burlington (VT) and Cary (NC). AB 1121 of course requires that any jurisdiction that uses ranked ballot methods must also organize an educational and outreach campaign, and that should help ensure that voters in these ten California jurisdictions have as positive a experience as voters in these other cities using IRV.
Second, you expressed a concern about the “potential ability of independent expenditure committees and others to manipulate such a system.” Again, we have the benefit of a great many election contests having been conducted using instant runoff voting that suggests that this concern has not manifested in actual elections. Not once in the many dozens of contests run using IRV has anyone ever suggested that there was manipulation of the system by independent expenditure committees. Indeed, in San Francisco’s 2008 elections for Board of Supervisors, there were four open seats and all four of those seats were won by candidates who were outspent by their opponents, particularly in independent expenditures. There is no evidence of which I am aware, either anecdotal or statistical, that independent expenditures have had a greater impact in IRV elections than in non-IRV elections — on the other hand, the San Francisco Ethics Commission found that independent expenditures soared in traditional runoffs before their adoption of IRV.
I note that you also expressed concern about other potential “unintended consequences” as a result of using instant runoff voting. As I stated earlier, the experience of using instant runoff voting in San Francisco and other jurisdictions has been generally quite positive. Voters appear to enjoy having the option of ranking several candidates rather than just picking one candidate. It also has a rapidly growing history in non-governmental elections, including many of California’s universities, including Stanford, UCLA and U-California-Chico.
Perhaps the best way of addressing the concern about unforeseen consequences is to emphasize that AB 1121 is a pilot bill with a very limited scope and mandate. AB 1121 only authorizes ten out of four-hundred general law jurisdictions to use ranked voting. Those ten jurisdictions would be required to study the impact of ranked ballot elections on their jurisdiction. The essence of structuring the bill as a pilot is to ascertain if there are any unforeseen consequences or if voters have any problems using the system. In that sense, the bill would seem perfectly constructed to address your concerns.
Should you have any need for further discussion, please do not hesitate to contact me at [email protected] or (301) 270-4616. I would be happy to go into more detail on any of this information provided here or seek to address any additional questions you have.