San Francisco’s Experience with Instant Runoff Voting
A Q/A with Gerard Gleason, member of the San Francisco Election Commission Gerard Gleason is a current member and past president of the San Francisco Elections Commission ( ). He addressed some questions about instant runoff voting (now generally called ranked choice voting – or “RCV” -- in San Francisco) in an email exchange with FairVote’s Rob Richie in May and June of 2009. His responses are shared with his permission. They are his personal thoughts and observations and are not part of any official policy or action involving the Elections Commission.

Some people have claimed that ranked choice voting (RCV) is costing San Francisco money, or at least not saving money because the city’s overall annual election budget has not gone down. What is your assessment of costs compared to cost savings?

Criticizing RCV in San Francisco on the basis of it costing money is a nonstarter. The facts are simple. The rule of thumb in San Francisco each citywide election costs about $3 million. A smaller voter pamphlet in the runoff might reduce the costs to $2.3 million, but that’s still far greater than any costs associated with using RCV.

Looking at annual budgets is misleading for trying to measure the costs of RCV because he San Francisco Department of Elections budget varies depending on how many elections we run. In the fiscal year for 2007/2008, for example, we conducted several elections, only one of which was an RCV election:
November 6, 2007 Municipal Election (3 RCV races Mayor, DA & Sheriff) February 5, 2008 Presidential Primary (No RCV election) April 8, 2008 Special Election for Congressional 12th District (No RCV election) June 3, 2008 Consolidated Statewide Primary (No RCV election)
San Francisco did have significant upfront costs in contracts with voting machine vendors to provide RCV capable tabulation software and related hardware modifications. But by avoiding at least one citywide runoff in 2005 and numerous runoffs in Board of Supervisor races in 2004, 2006 and 2008, the initial costs are being recouped.

I assume there are costs associated with San Francisco’s system requiring a separate ballot card for RCV elections. Could that cost be avoided?

Currently San Francisco provides a separate ballot card to voters in order to accommodate the RCV elections. I agree that allowing RCV and non-RCV offices and measures to appear on the same card would be the most cost-effective solution and believe the vendors could provide a solution. Frankly, the vendors should be the ones providing that solution. In fact our current vendor Sequoia Voting Systems had an interesting, different RCV sample ballot when they publicly demonstrated voting equipment that was RCV capable during the last contract RFP (request for proposal) period in San Francisco. But right now Sequoia instead is using a form of my original design.

Back before I was on the Elections Commission, I designed an early RCV demonstration ballot just to get past the hurdle of “it can’t be done.” I produced the ballot designs for the Department of Elections, when I was a Department employee from 1999 to 2002 -- the time frame when San Francisco switched from punch card to optical scan ballots. Because of my experience, RCV advocates asked me for a simple example of what the design could look like given the city’s equipment. Rather than invest any research and development time into doing its own design, the vendor at the time (ES&S) decided to just go with the design I had quickly created. I suspect it’s the same kind of attitude that led them to not yet come up with a solution on putting RCV and non-RCV races on the same ballot cards.

Now that you’ve had several RCV elections, what does San Francisco currently spend on voter education and pollworker training about RCV?

Costs for education of voters and training for pollworkers is significantly less now that RCV has been established. Early on, San Francisco conducted extensive training and public outreach regarding RCV when it was introduced in November 2004 for partial district elections for County Board of Supervisors, and again in November 2005 in citywide elections for some offices (Assessor, Treasurer & City Attorney). This education included public demonstrations at community forums, postcard mailings to voters, videos & PSAs.

Since 2005 the RCV outreach and education has been less and less as the public appears to have a firm grasp of how RCV works. The main voter outreach today is done in the form of pages in the voter guide. The Voter Information Pamphlet sent to voters now has only two (2) extensive pages on how to mark an RCV ballot. Here is a link to the online Voter Pamphlet for last November 4, 2008, where you can see that only pages 16 & 17 of the more than 260 total pages in the Pamphlet are devoted to RCV. [See ]

I believe the Department estimates the entire Voter Pamphlet production and distribution costs at about $4,000 per page…so RCV specific outreach costs in the Voter Pamphlet may be about $8,000. The Voter Pamphlet devotes about an equal amount of space to information on such topics as “how to know where to go to vote”, “accessible voting issues and machine operations (DRE)” and “provisional voting.” In short I do not think RCV is a particularly onerous problem to impart to voters any more than other information voters might need. The Department also has an online demonstration of RCV and does mention RCV at community events and forums, but the recent outreach activities appear to devote much more time and attention to accessible voting issues and machines (DRE demonstrations) to meet HAVA requirements than it does with RCV demonstrations.

Is pollworker training difficult with RCV? Do voters get it?

I have worked in precincts and as a pollworker for the past 18 elections in San Francisco. I go to pollworker training before each election. San Francisco by far has some of the best pollworker training in California, and I would say, the entire nation.

For the past couple of years, pollworker training classes have devoted no more than five minutes to RCV issues. This is during a class that is usually an hour and a half long. Frankly, the pollworkers I have worked with get RCV.

My observation has been that San Francisco’s Elections Department presents RCV as it was approved by voters in 2002. Its approach is simply: “This is how we conduct elections for city offices”. The “opportunity” is there for voters to make a ranking of which candidate they would like for a given office, but there is no special emphasis on filling out all your rankings. In competitive RCV races, many voters do rank three candidates, but I know some RCV advocates would like to see RCV explained in such a way as to encourage voters to mark all three choices. But, as with having a separate runoff election, all that can be offered is to give voters the “opportunity” to participate, be it RCV or a separate runoff election.

Because many voters in San Francisco obtain information in languages other than English, it would be important that someone other than myself give a you further feedback on voter understanding of RCV. I am not saying there is an issue with people who speak other languages. As an English-only speaker, I personally cannot say for certain. [Editor’s note: Analyses are encouraging on this front. See Instant Runoff Voting and Its Impact on Racial Minorities, a report by The New America Foundation and FairVote, June 2008, posted at:]

Do you think having RCV in San Francisco might lead the city to replace a paper-based system with touchscreen / Direct Recording Electronic (DRE) system?

The idea of blanket use of DRE machines is not even being considered. The State of California has considerable restrictions in place regarding the use of DRE voting machine. Additionally, the San Francisco Elections Commission has put in place a policy preferring marked paper ballots over other forms of voting.

Do you have other observations about the RCV process in San Francisco?

I’m glad to give you and anyone else my thoughts on the topic of elections. I think it is healthy for our democracy to have honest dialogue and consider what works best for the people. Because I am a member of the Elections Commission in San Francisco, however, I have to make clear these are just my personal thoughts and observations and are not part of any official policy or action involving the Elections Commission.

That said, voters in SF understand RCV and there is very little negative feedback from voters about it. One reason RCV is popular in San Francisco is the simple fact that it consolidates city elections. San Francisco voters like to participate in our democracy, but people in general lead busy and complicated lives that do not always allow them time to get to the polls or remember to complete and mail in a ballot.

I worked the polls this spring for the special California State election and many voters were surprisingly hostile about having to come vote for something extra. Our old system of having a follow-up runoff just a few weeks later is a burden for some people. That may be why some of the runoff elections have very low turnout. Someone should figure out the optimum frequency for public elections otherwise there might be a burnout factor. If RCV has a flaw, as we use it in San Francisco, it may be the ballot with the current voting equipment cannot tally more than three choices. Just as with the limitation of having to have a separate card for RCV races, the equipment vendors could improve this, and allowing a complete ranking of all candidates would be best. Although some voters might find that intimidating, here in San Francisco, voters are already familiar with RCV and appear to know what they want to do as far as making election choices.

Have you seen any changes in campaigns since the adoption of RCV?

As for RCV creating less hostile campaigns, for me as a casual observer, it may be hard to say. But let me share this from 1987, when we had a separate runoff system. There was a competitive election for Mayor of San Francisco in which the perceived frontrunner had a campaign slogan in the first round of voting along the lines of “Working Together.” Then, during the 6-week runoff campaign his slogan changed to “San Francisco is Worth Fighting For”. I remember how in the runoff “together” quickly turned to “fighting”.

I find that RCV offers several interesting advantages, cost savings being a big factor. Some think it might lead to less negative campaigning or cooperative alliance. Again, it’s hard for me to really weigh in on all that as I don’t pay to much attention to the campaigning. For me the question of RCV comes down to this: if we as a community decide we want a standard of elections being decided by 50% +1, then the next question is how to we achieve that? Personally, and again I stress this is my personal opinion, I don’t think making a choice between the two top finishers weeks after the main election is the best system – especially as it is usually done, served up by campaign consultants, many times to a shrill pitch. I think I am capable of and satisfied making my own ranked choices that don’t have to be just the lesser of two evils.