Mayor's Race By the Numbers: Turnout, Bases, and Trends

With less than a week before the June election, I know what you all are thinking. It’s either, “It can’t come soon enough!” or, “wait...what election?”

Last month, the Beacon hosted a “Primary Pick’em” event to help fellow amateur election watchers organize our thinking about how this election is playing out. For those who missed it, I’ll start with a quick primer.

Most election watchers like to start with the current polling, but to me that’s like going to the bathroom before you take your pants off. In order to figure out how people will vote, you first need to know how many people will vote, and where. The three biggest pieces of this puzzle are turnout, bases, and trends.

Voter Turnout

Here is a table of voter turnout in the last five elections and what was at the top of the ballot:

National elections excite more voters than local elections, and general elections excite more voters than primaries. What does it mean that that this upcoming election is both a local election and a primary? In general, it means turnout is going to be low. Perhaps the national environment will galvanize voters, but I’m skeptical. If history is any indication turnout will be somewhere between 40-50 percent.

However, not every part of San Francisco turns out equally. Next, we need to look at where these voters live.

San Francisco is conveniently divided into 11 supervisorial districts: Each district is supposed to contain approximately the same number of residents, but those residents aren’t all equally likely (or able!) to vote. 

This table-cum-heat map shows what proportion of the vote came from each district, and the stark nature of the disparate voter turnout. The voters in District 8 (Castro, Noe, Glen Park) routinely out-vote their neighbors to the south in District 11 (Ingleside, Excelsior, Crocker-Amazon) by a 2-to-1 margin.

In the last mayoral election, voters in Districts 2 (Marina, Cow Hollow, Pacific Heights), 7 (West of Twin Peaks), and 5 (The Haight, Western Addition) joined District 8 to comprise almost half of all voters, despite being a lot less than half of all residents.

District Uniformity 

Knowing where voters live tends to tell us a lot of a about what kinds of policies and candidates they prefer. However, there are diverse viewpoints within districts too.

Last year, I wrote about voter uniformity in the 2016 election—how much voters within a district vote agree with each other, and found that some districts are more cohesive than others. This amplifies the power of those voters, as policies and candidates that appeal to their preferences are likely to get a lot of votes.

The chart below shows how many votes each district contributed to their most preferred outcome less the votes of those who disagreed across the closest races in 2016. District 2 voters didn’t have the highest turnout, but they did have the highest degree of cohesion by a country mile. As a result, these voters got what they wanted more often than not.

Voter Histories

We already know a lot about voters as they head to the polls to vote for the next mayor. Voters like to vote for people they previously voted for, or people like those people. Jane Kim recently ran a city-wide campaign for State Senate, which she narrowly lost. The results of that election looked like this, with Kim in blue and Scott Wiener in red:

It may be helpful to compare this to the map of Ed Lee’s mayoral reelection (Lee also in blue) in 2015:

If you look for areas that are blue in both maps (Chinatown, parts of the Richmond, and much of district 11) you’ll find the areas where voters probably have a tough choice to make. Will they stick with Jane Kim, or vote for someone more aligned with the former mayor?

Two other high stakes areas are going to be hotly contested in this election. One is District 5, Supervisor London Breed’s home turf, where she narrowly won reelection just as Jane Kim narrowly carried the district.

Voters in the Western Addition and Inner Sunset will have to choose between two candidates that they previously preferred. As I wrote previously, Jane Kim appears to believe that she’s going to lose many of these voters, which is why she made a play for voters in West Portal.

Likewise voters in District 8 will have a choice to make. These voters are previously very strong Ed Lee/Scott Wiener voters, which would suggest an affinity to London Breed. However, this is former State Senator Mark Leno’s home turf, and it has provided him with a reliable voter base for every public office he’s held since serving as District 8 Supervisor roughly two decades ago.

It’s been a while since Leno has run a competitive election, but when he did, he relied on these voters to carry him to victory. Many District 2 and District 8 voters (two high-turnout districts) have chosen Mark Leno at every opportunity.

A lot has changed since Leno first ran for office, but Leno’s candidacy depends, in large part, the unchanged loyalty of voters in District 8 who could make up 14% of the voters in this election.

Undecideds Will Be Decisive

The most recent polling indicates that 15-20% of voters are undecided—and why shouldn’t voters be undecided? What’s the advantage of making a decision earlier than necessary? If you voted on the first possible day, you would have missed Angela Alioto take a page from Donald Trump’s playbook and decry sanctuary city policies.

The most recent polling indicates that Breed has 28% of the vote, Leno has 21%, and Kim has 17%—all far short of the 50% needed to win decisively. If undecideds vote overwhelmingly for one or two of the three leading candidates, it will be game over for the third.

What those remaining voters do could decide the election, and we have a one clue about who and how where those eyeballs seem to be looking: Google searches.

If I were Mark Leno, this chart would make me very nervous. If Google searches are any indication of interest, then he hasn’t attracted nearly as much attention as Breed or Kim over the last 30 days.

His recent ad buy on all the major broadcast networks (60 seconds at 7pm on May 24th) may help remedy that – however, it doesn’t help that he didn’t buy time on TNT, which was airing the Golden State Warriors game at the same time as his commercial.

Here’s the search trend for all the candidates during the day of the roadblock ad buy:

You can see that Mark Leno’s yellow line jumps up after the ad played at 7pm. However, so does London Breed’s—probably because he prominently attacked her by name in the ad. It’s an unfortunate tradeoff inherent in attack ads, boosting the opponent’s name recognition and potentially leading many undecided voters to Breed’s campaign materials.

Ranked-Choice, But In Reverse

It’s unlikely that any candidate will win over 50% of the first-choice vote, which means voters’ second and third choices on the ranked-choice ballot will ultimately determine our next mayor. Much has been made of Mark Leno and Jane Kim’s ranked-choice alliance. They’ve each endorsed the other for #2, and are encouraging their supporters to follow suit.

While this strategic alliance makes a certain amount of sense, given that they are both currently polling behind Breed, it ignores all the people who are voting for someone else. Angela Alioto was polling at 8%; Republican Richie Greenberg was polling at 6%, though Republicans often receive about 10% citywide; and the other candidates are polling in the low single digits.

There appears to be as many people voting for all other candidates as there are undecided voters. The #2 choice for undecided voters is incredibly important—potentially more important than the choice Kim or Leno’s voters make for #2, because these other votes will be allocated first.

If one of the three main contenders manages to woo the Alioto or Greenberg voters’ second choice, it could propel them far enough ahead to win the election.

Bases Loaded

If Mark Leno is running as strongly in District 8 as I suspect, than it means he’s pulling very few voters from other parts of the city. A candidate can get quite close to 20% on votes from District 8 alone. This means that Breed and Kim are pulling similar number of voters, but are appealing to a much wider geography. If Leno isn’t able to expand beyond this base or win over undecideds, then he will probably come in third place and his voters—who are also Scott Wiener voters—seem likely to pick London Breed as their 2nd choice.

Expanding beyond his base would require Leno to poach voters from Kim’s base, perhaps in the Mission, Bernal Heights or Western SOMA; or it could also involve securing the votes of some of his old District 2 voters (among the least “progressive” in the city). Doing both is a difficult needle to thread, since running a campaign to court Kim voters requires alienating folks in District 2, and vice versa.

If Leno is winning District 8 voters, Breed’s slight advantage in the polls should send chills down the spine of the Kim and Leno campaigns. It means she’s doing so without District 8 first-place votes. This group was absolutely crucial to Scott Wiener’s victory.

How to win

Breed: Run far enough ahead of the other two—probably in the high 20s to low 30s—and have enough crossover appeal with Kim or Leno voters to put her over the top. Breed probably needs to come in first place by several points and then do well among the initial ranked-choice rounds (Alioto and Greenberg) to win. If the initial vote is close, then Leno or Kim will probably win. 

Leno: Run strong enough in District 8 to keep Breed’s total low and come in above Kim. Leno could win even if he comes in a close second to Breed, as Kim voters are probably more likely to pick him than vice-versa. Watch the voter turnout: Leno’s voters come from high-turnout neighborhoods, which means he is likely to benefit from lower turnout.

Kim: Pick up as many currently undecided voters as possible, hope for high turnout, and that Leno voters reject the impulses that led them to vote for Scott Wiener last year and strongly choose her as their second choice.  Kim’s path to victory appears narrowest because her voters tend to come from low turnout neighborhoods and Leno voters will have an affinity for Breed.

This is our first serious city-wide race with more than two candidates, and without an incumbent. Voters have several well-funded, serious candidates to choose from, and three candidates are running neck-and-neck to the finish line. Regardless of the outcome, we’re going to learn a lot about what San Francisco voters want from a mayor. That will be a valuable lesson, given that the winner could likely shape city policy for the next decade.

Andy Mullan is the Beacon’s very own data guy. He analyzes data and their impact on local politics in our regular feature, SFByTheNumbers. Find more in-depth, obscure data analysis and adventures in new adulthood at his blog, Lord of the Fails.

Get a first-hand account on the decisions, strategies, and players that defined the June 2018 election, at our Election Debrief: Campaigns and the Candidates Who Ran Them! June 14th, 6PM - 8PM.

Grab a beer with the owners of acclaimed restaurants Cassava and Auntie April's as we tackle the big questions facing small business! Join us for Small Biz & Beer: Discussing the Urban Economy - June 12th, 6PM!