The IRV Victory in San Francisco San Francisco became the first major American city to adopt IRV to elect its local officials by a comfortable 55 percent-45 percent margin on March 2002, overcoming negative editorials in the daily newspapers, attacks from Mayor Willie Brown, well-funded opposition from the downtown business community as well as from well-connected political consultants concerned about losing opportunities for making money from the eliminated runoff election. The CVD organized a grassroots effort that reached enough San Francisco voters through extensive literature drops and phone calls to head off the late and high profile attacks. The campaign also garnered endorsements from a range of leading civic players, including Assembly Leader Kevin Shelley – since elected California’s Secretary of State – leading 1999 mayoral candidate Tom Ammiano, and significant organizational backers.3 In the end more than 60 percent of Latinos and African Americans supported the change, and pro-IRV arguments seemed to find a footing in all major constituencies.
The result reversed a 56 percent-44 percent defeat for choice voting in 1996, which would have applied to the Board of Supervisors. Our case was a strong one. The City had held local elections every November (citywide offices in the odd years and members of the Board of Supervisors on even years, with both sets of offices having staggered terms), with a December runoff election if no candidate received a majority in November. This process generated clear problems. First, each runoff in San Francisco costs taxpayers at least $2 million. Second, runoffs put candidates under great pressure to raise money quickly, giving greater access for special interest contributors and undermining campaign finance reform. Third, voter turnout dropped: in 2000, it declined by 50 percent between the first and second round, and the December 2001 runoff for City Attorney – the City's second most important office – generated a paltry 13 percent turnout. By producing a majority winner in a single election, IRV maximizes turnout and saves taxpayers the cost of paying for two elections and candidates the need to raise extra cash quickly.
In comparing why IRV was successful here, but failed (by margins of nearly two-to-one) in Eugene, Oregon in 2001 and in the state of Alaska in 2002, we have learned some lessons. The first is that there must be a widely agreed-upon problem to fix. In the defeats of IRV in Eugene and Alaska and choice voting in San Francisco in 1996, there was no potential cost saving of millions of dollars resulting from reducing the number of elections. In contrast, IRV in San Francisco folded the December runoff into the November general election, allowing proponents to draw support from more moderate and conservative voters and to gain enough interest from the Department of Elections to keep it neutral in the campaign, while winning strong support from progressive voters who perceived holding just one election helping their candidates by boosting turnout and reducing the impact of campaign contributions. There were sensible arguments for Eugene to go to IRV, but not ones that could hold up to a negative campaign funded by opponents who wanted to use the initiative to mock the city council. In Alaska, there was a history of third party spoilers that hurt Republicans, but that problem sent a mixed message – Republicans were divided about whether it was good to allow third parties to potentially build support under IRV, while the Democrats were mobilized to oppose IRV as a power grab by Republicans. Election administrators were loudly opposed to IRV in both Eugene and Alaska because of new perceived burdens on them.
In addition, we learned to avoid focusing too much on process. In contrast to Eugene, the San Francisco campaign focused on the simply stated benefits of IRV rather than the mechanics of the electoral system – its website was called “improve the runoff,” and the stress was on saving money and making the election process more efficient. The name “instant runoff voting” wasn’t even used on key pieces of campaign literature, instead just saying “Vote Yes on Proposition A.” Once process is raised, a cautious voter will shift to a ‘no” vote unless they come to understand that processt. A related lesson was that a serious campaign required serious resources. The Eugene city campaign spent less than $5,000 and the Alaska statewide campaign spent some $50,000. Once confronted with negative attacks and fervent opposition, that amount was far too little to allow an adequate response. In San Francisco, advocates spent $70,000, and the Center devoted tens of thousands of additional dollars in staff time to the effort.
There is an important postscript to the San Francisco story: powerful reform opponents do not give up easily. IRV is still alive in San Francisco, but it was not implemented in 2003 on schedule. In August 2003, a Superior Court judge ruled against forcing the City to use IRV for the November mayoral elections despite finding that the San Francisco Department of Elections was breaking the law for failing to implement IRV, characterizing its efforts to implement IRV since passage of the charter amendment as "fumbling" and "haphazard." The judge gave the Department permission to postpone implementing IRV until 2004 because he feared that, with time running out before the November election and the pressures of the statewide gubernatorial recall election in October, the department could not be relied on to implement IRV fairly. It was a classic Catch-22. The various government agencies charged with fulfilling the law had dragged their feet to the point where, when finally they were sued for not implementing, the judge ruled that it was too late to follow the law in 2003.
Opponents had seized upon distrust of hand ballot-counts, instead requiring that IRV be implemented on election equipment - which had to be reconfigured. This meant that, suddenly, several new players entered the picture. The vendor had to modify its equipment and receive federal and state certification for the change. The Department of Elections had to develop a plan and timeline for the vendor that was adequate for getting IRV on time and negotiate a contract. The Board of Supervisors had to appropriate funds for the contract. The Secretary of State had to certify the new equipment. The opponents who spent more than $100,000 at the ballot box to block IRV likely spent even more on public relations and legal fees in interfering with each step in this process. With none of the players doing their job - despite steady pressure from IRV proponents - the opponents ultimately achieved their goal: a mayoral runoff at the end of 2003 won by the well-financed candidate backed by the business community.
Looking ahead for IRV, there are incremental gains poised to be won in several legislatures, including requirements that new voting equipment be required to support ranked choice systems, and at least two ballot measures in cities that have reasonable chances to pass in 2004. The presidential race may again raise the “spoiler” controversy, and the argument for IRV over runoffs is stronger than ever. New data from our Center shows that voter turnout decreased in the decisive round in 82 out of 84 runoffs in federal primaries from 1994 to 2002, by an average of more than a third, and campaign finance abuses are particularly pronounced when candidates have to raise money for a second round of election.