Saturday, May 29, 2010

Instant Runoff Voting, Election Integrity and FairVote

Updated June 19
Some critics of instant runoff voting suggest that its backers, including FairVote, should take election integrity concerns like verified voting, manual audits and transparent elections more seriously.

Local activists for IRV often have been leaders in local election integrity efforts - people like Rick Lass, who led the effort to win instant runoff voting in Santa Fe (NM) in 2008 and Anthony Lorenzo who led the effort to win an IRV ballot measure in Sarasota (FL) in 2007. As to FairVote, we've proposed procedures for auditing ranked choice voting elections and periodicaly highlight our views in communications to our members, like in this November 2009 Innovative Analysis. Here also is a link to our statement on election security and audits overall.

More broadly, FairVote was the first national group to propose establishing an affirmative right to vote in the Constitution, and in so doing highlighting a full range of federal, state and local laws and practices undermining suffrage rights. For years, we have also been leaders in the call for public interest voting equipment, including open source software and removal of profiteering from elections -- for instance, see this excerpt from a commentary in 2004:

"Public Interest" voting equipment. Currently voting equipment is suspect, undermining confidence in our elections. The proprietary software and hardware are created by shadowy companies with partisan ties who sell equipment by wining and dining election administrators with little knowledge of voting technology. The government should oversee the development of publicly-owned software and hardware, contracting with the sharpest minds in the private sector. And then that open-source voting equipment should be deployed throughout the nation to ensure that every county -- and every voter -- is using the best equipment.

Getting issues of election integrity right are of essential importance in running fair elections, however, and we look forward to ongoing communication with election integrity advocates who have constructive suggestions on how best to implement ranked voting in elections that are secure and can be audited.

Tuesday, May 25, 2010

Fairvote MN statement on Minneapolis RCV Election Cost Report

Minneapolis Introduction of RCV Pegged at $365,000

1/3 of amount attributed to one-time start-up costs

Further cost efficiencies expected in future elections as RCV-capable machines become available

A new report by Minneapolis Interim Elections Director Ginny Gelms concludes that costs associated with the city’s 2009 switch to Ranked Choice Voting were approximately $365,000. Of the total cost, slightly more than a third encompassed one-time expenditures that will not be required in subsequent elections. The report also noted that additional cost-efficiencies are expected as voting equipment is put in place and voters and election judges become increasingly familiar with the system. “Process improvements implemented from lessons learned in 2009 will likely make for a more efficient process . . . which will impact the overall cost,” Gelms wrote.

The first RCV election in Minneapolis last November proved highly successful, with 95 percent of voters polled calling it easy to use. Former Minneapolis Interim Elections Director Patrick O’Connor who oversaw the implementation of the new system, says “we proved that it could be well administered, quickly and accurately counted, and that voters had little problem with the concept.”

The April 26 report noted that Minneapolis’ 2009 municipal election cost $1.47 million, an increase over the $1.13 million spent in 2005 (adjusting for inflation). The hand count in the 2009 election represented the largest portion of RCV-specific expenses.

Should a hand count be needed in the next election, however, the city can consider other options for cost savings such as removing the requirement to record the names of all write-ins (as former Minneapolis Elections Director Pat O’Connor has recommended).

The report also indicated that if RCV-capable voting equipment was available in the next election to tally the ballots, costs would be reduced by more than half. Gelms has said that such equipment may be available within the next three years; the city is working closely with Hennepin County to have RCV-ready voting machines in place by the 2013 election. Such machines are currently used in San Francisco; Cambridge, Massachusetts; and will be used in upcoming November elections in Berkeley, Oakland and San Leandro, California.

The highly effective voter education effort leading up to the 2009 election was the other major RCV-related outlay, accounting for 30 percent of RCV costs. The effectiveness of the effort can be seen in the facts that 95 percent of voters found the new election process easy to use and that the entire election produced just one defective ballot among 45,968 cast. These results illustrate the importance of voter outreach and education, efforts which council member Robert Lilligren says should be a priority – to improve voter familiarity with city elections and promote turnout – regardless of whether RCV is used.

The city’s 2013 projections assume a continued strong investment in voter outreach, which FairVote Minnesota supports. Even so, RCV experiences in other cities, such as San Francisco, suggest that “earned media” coverage in newspaper, radio, and online news outlets can help defray future costs here as well.

It’s worth noting that the report only looked ahead one election cycle and as such didn’t address the potential savings achievable over the long run through the elimination of the primary in combination with the use of machines and reduced voter educational costs.

The move to RCV was led by FairVote Minnesota, a nonprofit organization working to enhance democracy through advocacy and public education.

Featured Quote: Former city elections director Patrick O'Connor, who oversaw implementation of IRV in Minneapolis in 2009: "I have had the great fortune to be a small part of what could easily be considered the most significant civic exercise in the history of Minnesota government: the implementation of the first Ranked Choice Voting election in Minneapolis and in Minnesota. We proved that it could be well administered, quickly and accurately counted, and that voters had little problem with the concept."

Monday, May 24, 2010

The distorted "backlash" against IRV -- The Sunnyvale Example

Updated June 19, 2010

IRV opponent Joyce McCloy is behind an anti-instant runoff voting blog with a simple modus operandi: highlight links to webpages that might discredit instant runoff voting and its advocates. This "evidence" tpically is incomplete, deceptive or just not true. But those seeking to seed doubt about change only need for it to appear possibly true.

Today's screed about a new alleged "repeal" of IRV was a good example. As Ms. McCloy knows, IRV in recent decades has been repealed in two and only two governmental jurisdictions in the United States: Pierce County (WA) and Burlington (VT). Both of these repeals were disappointing to advocates, of course, but there was a context to them (see more on Pierce County and on Burlington).

And that's it. No other city has repealed IRV even as more cities, organizations and colleges keep using IRV every year. Ms. McCloy knows that Aspen (CO) hasn't repealed IRV, but she says it has because voters narrowly failed to pass a non-binding advisory question to keep it. She knows that Georgetown University uses IRV, but she says it's been repealed because students didn't use it for a single election in early 2009 (and never mind IRV's ongoing and growing use in nearly 60 colleges and universities, including adoptions in the past year year at colleges like Cornell, Brown and Brandeis). She knows that Cary (NC) didn't repeal IRV (rather it had used IRV in a one-time pilot in 2007), but she says it did. And so on.

Today she's excited about Sunnyvale, California. This year, based on a rule adopted by a prior council, the Sunnyvale might city council would have used instant runoff voting within if at least three members of its seven-member city council had run mayor. Such small-scale uses of IRV can be interesting to consider, but with at least three candidates (all councilors) and only seven councilors with votes (including the three candidates), things can get tricky - tie votes, say, and efforts to outsmart colleagues.

So in what seems like a very sensible decision that IRV advocates would be quick to support, Sunnyvale's city council decided to change its rule to use IRV for these internal elections of the council choosing the mayor in the future. But does this mean Sunnyvale has "repealed" IRV and that supporters of the change think IRV's too confusing as a system? Joyce McCloy has added Sunnyvale to her litany of "repeals" in communications, but consider this quote from a thoughtful blogpost by one of the backers of the change, Sunnyvale city councilor Jim Griffith:

First off, while ranked-choice (or instant run-off) voting is terrific for general elections or when you have a lot of candidates to choose from, it doesn’t work well with a small number of voters and a small number of candidates (in our case, 7 voters, 3 candidates). There’s a lot of opportunity for game-playing, plus a good chance of an end result that many members of the public simply won’t understand. Neither of those serve the public good, so I wanted to get rid of the ranked-choice option

Joyce McCloy in her post bolds a quote from Councilmember Griffith as if he is anti-IRV when in fact he thinks IRV is "terrific" for general elections. Sadly, however, don't hold your breath waiting for corrections to her blog. A website URL can't be wrong or misleading, right?

Addendum, May 25: I received an email from Sunnyvale city councilor Jim Griffith. He expressed appreciation for my post and clarified that in fact the Sunnyvale city council never used IRV. While there might have been three councilors seeking to be mayor this year, only two ended up doing so. But the potential choice among three councilors led to the council sensibly deciding that a binding IRV vote with only seven voters choosing among three or more candidates (all of whom also would be among the seven voters) wasn't appropriate. Griffith also reiterated his support for IRV for general elections.

Wednesday, May 19, 2010

North Carolina FAQ on Instant Runoff Voting April 2010

The following are questions frequently asked by North Carolina municipalities considering implementation of instant runoff voting (IRV) and answers based on responses from the State Board of Elections, the University of North Carolina School of Government, and scientific exit poll surveys conducted in North Carolina.

1. What is instant runoff voting (IRV)?

IRV is a majority voting system that combines a regular election and a runoff by giving voters the option of ranking candidates in order of choice. The two candidates with the most first choices advance to the runoff. In the runoff, each ballot is counted for whichever runoff candidate is ranked higher. As with a traditional runoff election, the winner is the candidate with the majority of votes in the runoff round.

2. How is IRV an improvement over a city’s current method of election?

IRV can improve each method of election used in our cities:

· If a city has a plurality election method, IRV would eliminate the possibility of a winner who was strongly opposed by a majority of voters, but won because votes for opposing candidates were divided among them.
· If a city has a partisan primary method, IRV would eliminate the problem of a majority of voters having limited choices that have been filtered by political parties and save the cost of an extra election.
· If a city has an automatic nonpartisan runoff, IRV would ensure a majority winner without the guaranteed cost of holding two separate elections.
· If a city has a conditional runoff, IRV would prevent the need for a separate costly low-turnout runoff.

3. What does an IRV ballot look like and how does a voter mark it?

A sample IRV ballot used in Cary can be found here. A voter simply fills in the first bubble next to his or her favorite candidate, the second bubble next to the voter’s second choice candidate, and the third bubble for their third choice. A voter may rank only one or two candidates if they so choose. Ranking more candidates does not count against your first choice, but increases the chances that you will elect a preferred candidate or prevent the election of a disliked candidate.

4. How are IRV elections counted on North Carolina voting machines?

The State Board of Elections has developed secure, inexpensive means to count IRV ballots on all voting equipment used by North Carolina cities. Hendersonville’s Ivotronic direct recording electronic system allows a final result on the night of the election. Cary’s M100 optical scan system allows a determination of first choices on the night of the election counted at the polls, with a final result one or two days later at a central location where all counting will be done on machine.

5. Is IRV legal in North Carolina?

In 2006, North Carolina entered into a pilot program by adopting law allowing up to ten jurisdictions (each year) to try IRV for their local elections, including school board elections. In the summer of 2008, North Carolina extended the program for three years with passage of Elections Amendment Act (S-1263), authorizing continued use of the instant runoff voting for willing counties and municipalities.

6. Where has IRV been utilized in North Carolina?

Both Cary and Hendersonville utilized Instant Runoff Voting in 2007. Hendersonville used it again in 2009. Outside North Carolina, cities recently adopting IRV include Memphis (TN), Oakland (CA), and Minneapolis (MN). It is a common method for elections of organizations, including the president of American Political Science Association and of student leaders at nearly 60 universities, including North Carolina State and Duke.

7. What expenses does IRV incur for a municipality that implements it for one year under the pilot program?

A jurisdiction participating in the IRV pilot program is responsible for the cost of educating voters, candidates, and election officials on the election method. These costs were minimal, however. With the use of a good tested ballot design, education costs can be pennies per registered voter. In addition, groups like the League of Women Voters can provide volunteer voter assistance. The State Board of Elections has developed inexpensive procedures for conducting the IRV count. Despite the minimal cost of voter education in Cary, the City saved $28,000 with IRV.

8. What did the voter education programs consist of in Cary and Hendersonville?

In both Cary and Hendersonville a simple ballot design was the key to successful voter education. Cary city officials sent sample ballots in utility bills and issued a media advisory about the ballot change. Local newspapers, in both cities, ran articles bout the new ballot design. Board of Elections staff visited civic organizations to inform them of the new method. Local radio stations ran 30-second Public Service Announcements. The North Carolina Center for Voter Education produced an informative video on IRV that ran on government access stations.

9. Did IRV save money in Cary and/or Hendersonville?

By avoiding a runoff in one of its districts, Cary saved $28,000 in 2007. A citywide runoff would have cost more than $100,000. Hendersonville did not need a runoff in 2007 or 2009, but any runoff would have cost much more than IRV.

10. Were any surveys conducted of voters in Cary and/or Hendersonville and did voters understand IRV?

Dr. Michael Cobb, assistant professor of political science and survey researcher at North Carolina State University (NCSU), in 2007 designed and analyzed the results of an exit poll survey on how voters in both Cary and Hendersonville felt about IRV. Karen Brinson of the NCSU Board of Elections managed the poll. The survey revealed that 68 percent of Cary voters preferred IRV, and 81% found it “very easy to understand.” It revealed that 71% of Hendersonville voters preferred IRV, and 86% found it “at least somewhat easy to understand.” Cary’s own survey in 2008 found similar findings, with even more dramatic differences among the more than 60% voters who gave IRV a top rating of 7-9 compared the 6% of voters who gave it a rating of 1-3.

In 2009, Dr. Cobb again did a Hendersonville survey. Voters overwhelmingly found IRV easy to use, and only 20% opposed its use for statewide elections.

11. What jurisdictions opted to use IRV in the pilot program in 2009?

Based on its administrative ease and strong public support for IRV, Hendersonville’s city council approved implementing IRV in for its 2009 elections and has indicated unanimous interest for 2011 elections and beyond. Cary considered doing so as well At a March 12th public hearing, city supporters of IRV heavily out numbered opponents, while the city’s 2008 survey revealed that a s strong majority of Cary voters understood and preferred IRV – indeed voters were ten times more likely to give it their highest rating of 7 to 9 compared to the lowest rating of 1-3. The council did not pursue the pilot, however, in part because of some uncertainty over the counting method for optical scan machines. Now that the State Board of Elections has developed a sensible procedure for doing IRV on the state’s optical scan machines, more cities may use IRV in 2011.

12. Does IRV make it harder for racial minorities (groups who have been historically disenfranchised and denied equal access to formal education) to vote, dissuade them from voting, or dilute the impact of their collective votes?

African American endorsers of IRV include President Barack Obama, U.S. Rep. Jesse Jackson, Jr., and election law scholar and Department of Justice official Spencer Overton. Surveys in both Cary and Hendersonville (as well as those in more diverse cities throughout the U.S.) show that voters of all racial groups understand IRV equally well. In most cases, IRV protects the strength of racial minority groups by preventing the need for a separate runoff for which racial minority voters have traditionally been the least likely to return, and by avoiding the possibility that racial minority voters will split their votes between opposing racial minority candidates of choice.

Racial minorities have been strong backers of IRV when it has appeared on the ballot, including super-majority support in landslide victories in cities like Oakland, Minneapolis and Memphis (TN), all of which passed IRV by margins of two-to-one or more in recent years. Voters of color also have been strongly supportive of the system in exit polls, and in the City with the most IRV elections, San Francisco, the Board of Supervisors has become more diverse during the six years the system has been adopted.
In San Francisco turnout among racial minority voters increased under IRV in the decisive runoff round, and prevented the “splitting” of Asian votes among four strong candidates of choice in a district race for election to the Board of Supervisors in 2006.

13. What steps does a municipality take to enter the IRV pilot program for 2011?

Each local governing board participating in the pilot (after receiving written instructions from the State Board of Elections on implementation) must approve participation in the pilot program, and agree to cooperate with local board of elections in a voter education program. The local board of elections must also approve participation. If a jurisdiction is in more than one county, all county boards of election must approve it. For more information please see Instant Runoff Voting: Goals, Standards, and Criteria for Implementation and Evaluation (presented to the North Carolina State Board of Elections on December 11, 2008 by the University of North Carolina School of Government). The report can be found at

14. How are IRV ballots exactly tallied to determine a winner?

First, all ballots are counted toward their first choice. If a candidate receives a majority of first choice rankings, the candidate wins the election. If no candidate receives a majority of votes, then all candidates are eliminated except for the top two vote getters. In the runoff round, each ballot is counted for the runoff candidate ranked ahead of the other runoff candidate on the ballot. Any ballot listing one of those candidates first is counted for that candidate. Among the remaining ballots, any ballots listing one of those candidates second is counted for that candidate. Any ballots without a first or second choice ranking for a runoff candidate are counted for whichever runoff candidate is ranked third. The winner is the candidate with a majority of the total votes.

In all jurisdictions, the first round of tallying is conducted using the jurisdiction’s current tallying procedure. For subsequent rounds of counting (if necessary) in jurisdictions in counties using direct-record electronic (“DRE”) machines, vote information is transferred to an Excel spreadsheet, validated for transfer accuracy, and then sorted and counted, either through the regular Excel functions or by hand-to-eye from the Excel spreadsheet. For jurisdictions located in counties using optical scan (OS) machines, subsequent rounds of counting take place centrally, using a methodology developed by the State Board of Elections that allows the count to be conducted entirely on OS machines.

15. Are these tallying methods certified and constitutional?

All of the above methods are certified and constitutional.

16. Who endorses IRV?

Public interest organizations like Democracy NC, Common Sense Foundation, Common Cause, North Carolina League of Women Voters, FairVote, NC Fair Share, Southern Coalition for Social Justice, NC Public Interest Research Group, Martin Luther King Jr. Coalition, Independent Progressive Politics Network, Southerners on New Ground, NC Green Party, and Traction support IRV. Individuals backing the North Carolina IRV pilot include House Speaker Joe Hackney, House Minority Leader Paul Stam, and John Hood of the John Locke Institute. Nationally, President Barack Obama, former governor Howard Dean and Sen. John McCain are among those who have acted on their support for IRV.

For more information and videos on IRV please visit:

NC Votes 123 Coalition

Democracy North Carolina

FairVote's IRV Program (w/videos)

NC Ctr. for Voter Education video

2009 Hendersonville survey

Wake County review of Cary vote