Rank Choice Voting in the San Francisco Mayor's Race.
It is much tougher to anticipate how voters will rank candidates than some in the punditry would like us to believe.
Aaron Peskin’s entrance into the mayor’s race has thrown a left-handed monkey wrench into the plans of the tech-libertarian-populists movement in San Francisco. What looked like a three-way race between candidates who were all pretty amenable to that more conservative approach to governance, has been shaken up because of the presence in the race of a progressive candidate with an undeniable and deep understanding of what it takes to get things done in San Francisco.
There is already talk of Mayor London Breed, Mark Farrell and Daniel Lurie forming an anybody but Peskin movement. This would require them to urge their supporters to rank all three of them ahead of Peskin in the city’s rank choice voting (RCV) system to ensure that Peskin, who has a good chance of doing well in the first round, gets less support in the following rounds.
Tim Redmond at 48Hills wrote a brief, but persuasive, analysis of why this is very unlikelty to occur. He argued that Breed, Farrell and Lurie all want to be mayor, so at some point in the campaign anti-Peskin unity will give way to hardball campaign tactics among the three others. Daniel Lurie is not asking his family to invest so much money so that he can help make Mark Farrell mayor. Similarly, London Breed, a career politician, is not going to be nice to Lurie or Farrell if she believes one of them could defeat her in this race.
Redmond’s concludes:
“I see three candidates (Breed, Lurie and Farrell) who care more about themselves than about any vision or agenda for the city. I see candidates who would just as soon align with Peskin if they thought it would help them win.
I see a big tech-funded anti-Peskin effort, with the potential of many millions of dollars, because the tech industry cares about its bottom line and those folks don’t want a mayor who has a history of trying to regulate them.
But the other conservative candidates working together? Maybe—but only if they think it will benefit them.”
-Tim Redmond, 48Hills
Another reason why an anybody but Peskin movement is unlikely to succeed is due to do with some common misconceptions about how voters rank their choices in RCV systems. Insiders, journalists and pundits frequently overstate the extent to which voters bring ideological clarity to a rank choice system, and the extent to which ideology is an applicable framework for local elections.
This is particularly relevant in the mayor’s race because all the major candidates will claim they are progressive sometimes and at other times claim they are not. Other than Peskin, all the candidates will also at times attack each other for being too progressive. Additionally, because this election will occur alongside the high turnout national election, there will be many low information voters who are unlikely to be aware of where each candidate stands on an ambiguous left-right continuum.
Peskin-Safai-Breed-Lurie-Farrell is probably how the five major candidates land when ordered from left to right. However, while Peskin is the progressive in the race, the ideological differences between the other four sometimes seems quite small. The image below reflects this ordering and one way to think of the bell curve of the San Francisco electorate. I assume some readers will disagree with both my ordering and shape of the bell curve, but that is further evidence that it is difficult to predict how voters will see and rank the candidates in an RCV system.
Given this, one would expect a progressive voter to rank their choices in precisely that order, while the most conservative voters would do the reverse with Farrell first and Peskin fifth, or more likely left of the ballot entirely. We would also expect a Lurie supporter to rank either Breed or Farrell second with either Safai, Breed or Farrell third and fourth. A Breed voter would rank Lurie or Safai second.
Just reading the above paragraph should provide a sense of why these assumptions don’t hold up. For example, a Peskin supporter might have encountered Lurie on the campaign trail been impressed with him and rank him second. A Farrell voter might like his conservative approaches to drugs, crime and homelessness, so rank him first, but might choose Peskin second because of his opposition to wanton development and real estate speculation.
Some voters might put Peskin, Breed and Safir their top three choices simply because they don’t want to be told what to do by either of the two rich white guys from Pacific Heights, Lurie and Farrell. Others may see this as a four-person race and not rank Safai at all despite him being, from this narrow and flawed ideological perspective, a very viable candidate.
Additionally, a left-right continuum is just one of many ways to think of San Francisco politics-and one that is probably much clearer to journalists, politicians and pundits than to voters. One could also imagine an inside to outsider-more accurately slightly less inside-continuum that would go something like Breed-Peskin-Farrell-Safai-Lurie, as well as various other taxonomies for these five candidates.
There are 120 ways to rank the top five candidates and it is likely we will see ballots with all 120 combinations. That Peskin-Lurie-Safai-Farrell-Breed ballot might look very strange to some insiders, but that voter will have had her reasons.
There is certainly a block of voters who don’t want Peskin to be mayor and will not rank him on their ballots, but there are also voters who will want anybody but Breed and leave her off their ballots. Similarly, some voters will be put off by Farrell’s macho conservative rhetoric, or Daniel Lurie paucity of experience and leave one of them off their ballot. This is what competitive elections are about, but that is a long way from the emergence a cohesive and impactful anybody-but-Peskin (or Breed, Farrell or anybody else) movement.
The column fails to mention the so called "bullet" vote..Voting for only ONE candidate. That is a very effective vote.
Shared and forwarded.