OPINION: What went wrong June 5 — and what we can learn

The strategy of running two progressive candidates for mayor didn't work -- and the left needs to get better about vote-by-mail turnout

Editor’s note: Progressives who supported Mark Leno or Sup. Jane Kim are looking at the results of the election and trying to figure out what happened and what can be learned. Last week, we ran a piece by Democratic Party Chair David Campos with his analysis. This week, we add a perspective from longtime activist Calvin Welch.

London Breed’s final 2,546 vote (1.1%) victory over Mark Leno was the slimmest margin of any mayor since run-off elections were established in 1975.Her 37% of the first-place votes and 23% of the counted Ranked Choice Vote starkly reveal the depth of popular opposition to her policies.

The fact that 63% of those who voted for mayor chose someone else and that 77% of those casting second-place ranked-choice votes for Jane Kim preferred Leno over Breed exposes the challenges she will face in governing, but also reveal the absolute folly of “progressive leadership” following a politic that split their vote with two candidates for mayor.

Mark Leno and Jane Kim appear at a joint press event

The true cost of that folly will of course not be paid by those most responsible, but instead by San Francisco voters who turned out in stronger than expected numbers to advance programs and candidates able to resist the continued policy of displacement and hyper-gentrification at the very heart of the Lee administration.

That a cost will be paid is certain, with Breed’s post-election boast that her administration will “build more housing, build more housing, build more housing” dropping all pretense of any need to address the reality that not one in 20 current San Franciscan can afford these new houses and that tens of thousands of us face eviction and displacement in order to make room for them in a city that has no vacant land, is the second most dense major city in the nation and has no plan to pay for the needed infrastructure, including Muni, to support this mindless development.

This article will explore the nature of the vote, what lessons might be learned from it and, given the outcome, what needs to be done to win more than a simple majority of the Board of Supervisors, which will be necessary if any meaningful change benefiting current residents can be achieved in the November, 2018 election.

The structure of the vote for mayor

The media and local political pundits denoted four “major candidates” in the race: two “progressive/liberals,” Kim and Leno, and two “moderate/conservatives,” Breed and Alioto. In addition, there were two other moderate/conservative “minor candidates” in the race: Republican Richie Greenberg and Ellen Lee Zhou who, together, polled some 16,000 votes, close to Alioto’s 17,000 vote total. Greenberg and Zhou votes were centered in Districts 2 and 7 (Greenberg) and 4,10, and 11 (Zhou).

The combined Kim/Leno vote was a majority OF ALL VOTES CAST in Districts 1,3,5, 6, 8, and 9, while the Breed/Alioto vote was a majority in only District 2. In Districts 4,10, and 11 the Zhou/Greenburg vote cut into the Breed/Alioto vote, denying them the majority.

Both Breed and Leno won a plurality of the vote in the supervisor district they represented: Breed in D5 and Leno in D8. Kim, surprisingly, did not win a plurality in D6. Indeed, she beat Leno by merely 60 votes, coming in second to Breed. She actually lost to both Leno and Breed in the South of Market portion of D6.

Breed got 39% of the D5 vote, down from 52% in her re-election bid in 2016. Her vote drop was the largest in the Western Addition where she polled some10,300 votes in 2016 but only 5,700 votes for mayor. She got 10,000 fewer votes in D5for mayor than for supervisor.

But the impact of the progressive split vote can be seen in D8, Leno’s former district. D8 had the highest turnout of any district, with some 36,0000 votes cast for a 64% turnout. Rafael Mandelman won with nearly 60% of the vote receiving some 20,000 votes. Leno was an early endorser of Mandelman, when the outcome was in great doubt. Yet, Leno got only 12,700 votes for mayor — 7,200 fewer votes than Mandelman — nosing out Breed by a mere 600 votes.

Kim finished a distant third, with 4,500 fewer votes than Leno, but still polled 8,200 votes, clearly most of which would have gone to Leno. Had Leno got the same votes as Mandelman in D8 he would be mayor today. That simple, that clear: Kim/Leno split the overwhelming progressive/liberal vote in D8 and it was decisive.

In an interview in the Chronicle after the election, Breed claimed that her support for State Senator Scott Wiener’s SB827, which would have over-ridden all local zoning and public hearing policies in order to build high density, high rise housing along most streets in San Francisco (Wieners SB827), most notably in western neighborhoods, was “something people appreciated, because we have to build more housing.”

The fact that she never, in fact, campaigned on SB827 and failed to defeat a resolution of the Board of Supervisors that formally opposed it was not mentioned by the Chronicle. Indeed, popular opposition to SB827, especially in western San Francisco (both District 7 Supervisor Yee and District 4 Supervisor Tang voted to support the resolution opposing SB827was also missed by the Chronicle) was extensive and deeply held.

Breed’s position supporting and Kim and Leno’s opposition to SB827 was not missed by D4 and D7 voters. Both Kim and Leno aggressively campaigned against it in the west. Kim held a special event presser on West Portal Ave. announcing her adamant opposition. It paid off.

In D4, Kim came in second to Breed with Kim/Leno outpolling Breed 9,500 votes to 7,100. Adding in the D4 Alioto  vote – Alioto also campaigned against SB827 — the anti SB827 candidates out polled Breed 11,500 to 7,100. In D7, Leno came in second and Leno/Kim outpolled Breed narrowly 10,900 to 10,100. Adding in Alioto’s D7 vote widened the anti-SB827 margin to 13,500 to 10,100.

While these traditionally “moderate/conservative” voters in D4 and D7 rejected Prop. C (free childcare) and Prop. F (legal counsel for tenants facing evictions) they did not, in fact, “appreciate” Breed’s position on SB827. More about the potentially changing politics in D4 and its significance for the D4 Supervisor race in November below.

The vote-buy-mail factor

A continuing structural feature of the San Francisco vote is the rise in votes being cast by mailed-in ballots (known as “Vote by Mail,” VBM). It is now routine that the majority of votes cast are VBM ballots. But not all VBM voters are the same. There are at least two sets of VBM voters, classified by when they turn in their ballots: early VBM voters and “late” VBM voters. Early VBM voters tend to vote more “conservatively” than election day voters, late VBM voters, folks who drop off their ballots at the polling place or show up at City Hall during the last week before the elections tend to fall somewhat between election day voters and early VBM voters.

This greatly complicates a campaigns effort to “get out the vote” as there is now an extended period, 30 days, during which votes are cast and thus an extended period during which votes must be “got out.” As always, money makes a difference. If you have the money to mail all VBM voters, then track new requests for mailed ballots and then both money and dedicated staff to “chase” or track VBM ballots as they come in, you are ahead of the game. But only the most well-funded campaigns can do this. It is very difficult for a campaign that relies upon a volunteer “field campaign” to both do the necessary door-to-door voter ID of the usual field campaign AND to effectively “chase” VBM voters for 30 days. Mail and increasingly ads on social media and cable TV are used to chase these votes.  Some observers claim that younger voters tend to fall into the late VBM vote mode more than elderly voters, who tend to fall into the early VBM voter mode. Working people still seem to vote on election day after work and are able to be contacted by more traditional GOTV efforts.

Progressive campaigns have been slow to adapt to these changing patterns of voting, mostly due to their inability to raise money early enough to organize and fund a VBM effort and a bias to spend what early money they have on a field campaign.

Jane Kim had no chance to win this election after the early VBM votes were cast and counted. Like other recent progressive campaigns for mayor — Gonzalez, and Avalos — early VBM votes placed her so far behind that she could never catch up. While she made strong headway in the election day vote and with late VBM voters, she simply could not overcome the 13,000 votes she trailed Breed or even the nearly 7,000 votes she trailed Leno in the early VBM vote; for her the election was lost before any of the later votes were counted.

Leno, however, succeeded in overcoming the progressive curse of death by early VBM vote. He came in second to Breed, only about 7,000 votes behind. He did so by spending money buying a wall-to-wall cable TV presence about 30 days out, hitting early VBM voters as they were making up their minds. By the final election day count (Report 4) Breeds first place vote lead was shrinking and the RCV count gave the lead, narrowly to Leno.

But then something that has never happened before happened.

A total of 98,629 votes were counted between the day after the election and June 15th. Some 73,000 were absentee ballots dropped off at polling places on election day, another 10,000 or so were absentee ballots received in the mail with post marks on or before election day and thus qualified to be counted, and some 14,000 provisional votes were counted. A provisional vote is a vote cast at a polling place in which the voter is not listed on the precinct roll and thus must be verified as being from a registered voter after all other votes have been counted.

In terms of the mayor’s race these “late VBM votes” were the decisive votes and its critically important for us to try to understand them.

Of those votes, 41% were cast from progressive voting districts 5,6,8, and 9, while 31% were cast from swing districts 1,3,10, and 11and 28% from conservative voting Districts 2,4, and 7. That a majority of these votes were from progressive districts can be seen in the votes for the four key ballot measures. Props. C and F got a higher percentage of yes votes and D and H a higher percentage of no votes from these voters than from those counted on election day.

Like election day voters, the majority of the votes cast for mayor were from the progressive Kim/Leno side with Kim/Leno getting 49% of them and Breed but 39%. The difference was that they broke Breed/ Kim, not Kim/Leno — as did the election day vote — giving Breed her slim margin of victory.

The Breed/Kim vote was unusual: Breed and Kim have been on the opposite sides of most of the key issues. But for a significant number of voters, the national narrative that encourages the election of women and women of color was more important than those differences.

The progressives opposed to Breed’s policies had a hard time countering that narrative, which was promoted by one-time progressive stalwarts like former state Sen. Carole Migden and longtime activist Debra Walker, who much of the progressive establishment supported for D6 supervisor over Jane Kim in 2010. The Leno campaign had challenges with that narrative, too.

The Kim campaign did a late ad buy on cable news when cable news was featuring this same narrative at the national level. Breed had enough money that she continued ad buys over the entire 30-day period. Leno had less money than Breed and spent it early.

Lessons learned

What lessons for the future can be learned from this election and how might we apply them to future elections?

First, don’t split your vote. In the face of an opponent who commands both money and media access, progressives must marshal our limited resources and concentrate on a single viable candidate. Ranked Choice Voting and having three votes will not overcome political folly.

Our side has a serious problem of leadership when it comes to candidates. We have no institution nor paramount personality that has the legitimacy to adjudicate between ambitious politicians seeking higher office. In the past the Democratic Party, organized labor or simply an old-time boss was the traffic cop. The last one in San Francisco, Phil Burton, died in 1983.We have been making it up as we go along ever since.

Don’t think Willie Brown plays this role for the other side; he simply acts like he does. He is totally dependent on business money and thus on candidates acceptable to them. He is about personal access not moving a wide agenda. The last Democrat to actually play that role was Phil Burton who was independent of business money and had access to both labor and party money and was about moving a liberal political agenda.

Not that the old-time boss model was perfect – in fact, a lot of women, LGBT people, and people of color felt left out in that era.

Progressives love to complain about “the machine” but there is no “machine” — just oodles of business money. We need a democratic, bottom up, grass roots “machine” able to not only devise progressive policy but do the hard but essential political work of saying yes and no to candidates and making it stick.

Breed did not have to play a fundamentally defensive ranked-choice game. Business interests are disciplined enough to not only pick a candidate but also tell others not to run and failing that, make sure they get few votes. Alioto, Greenberg and Zhou combined vote was less than one third of Breed’s and did not hurt her, while Kim and Leno split 122,000 votes — more than the four moderate/conservative candidates combined — guaranteeing that both lost. Shame on us.

Second, learn how to wage a 45-day GOTV campaign without throwing money at the problem. Don’t get me wrong, money would work, I just don’t believe our side will every have enough to do it. One way to think about the dominance of VBM is to view the late VBM vote as a 30-day GOTV exercise. For early VBM, we can mail or drop them before the 30 day window on voting is opened so we must devise techniques to track who has applied for the mail vote and who have returned them.

We must verify that late VBM voters tend to be progressive and early VBM voters not so much but still a critical vote to get, as equally critical as election day precinct voters. We must simply approach them differently. The Prop. F campaign did targeted walk lists by precincts for VBM voters and it worked for them.

I do not claim to have all the answers, but I do have the questions: how can we avoid being beaten by the early VBM vote, how do we chase (identify and track) later VBM voters. How do we integrate election day GOTV with VBM GOTV?

Third, the combined strength of the Kim/Leno vote in D4 and D7 clearly demonstrate that western neighborhoods are moving ever so slightly in the same direction as eastern neighborhoods on displacement and development issues. D1 is already there now, voting not only for progressive candidates but also for progressive issues like C and F and against D. How do we maximize that emerging convergence? What set of issues, able to be placed on the ballot, might assist that coming together?

There is a growing awareness that city policies governing demolitions of residential uses is a key high ground to stake out protecting residents and buildings in existing neighborhoods both east and west. Breed will never be able to “build more housing, build more housing, build more housing” without demolitions of existing homes and the displacement of existing residents. This could be made a key issue in the D4 campaign by progressive candidates running this November.

Funding Muni to meet existing and future transit needs is also a key development issue that stands a fair chance of appealing to the transit-poor western neighborhoods as well as transit reliant eastern neighborhoods.

Fourth, we have failed to articulate a convincing narrative about a left -progressive future for San Francisco. Such a narrative should be based on the challenges of not only resistance to Trump policies but also a vision of the future of this city that is both inclusive and socially just.

This November will most probably see a major ballot measure aimed at taxing the gross receipts revenue of the most profitable businesses in San Francisco to build new housing and provide social and health services for homeless people or people at risk of becoming homeless. While the campaign will be an uphill fight, we on the left in this city have a chance to use it to frame the debate about urban poverty and homelessness in San Francisco that differs from the Chronicle. Creating that new narrative is an essential part of articulating a new “San Francisco story” not based on fancy food and tech titans.

The Chronicle has become an active campaign resource used by the pro-big-business side. Its collusion with Breed campaign operatives in planting an attack on Kim as a news story, its repeated use of Breed campaign poll numbers to show Leno running a distant third and its attempt to claim that tech titans had no real interest in local politics when both Kim and Leno were campaigning against their undue influence in Room 200 are examples of “fake news” not aimed at Trump but us, residents of San Francisco. The ability of the Chronicle to shape news coverage by other media, to actually define deeply conservative even reactionary politics and policies as “moderate” has grown as other daily news sources disappear. The fact that it has a tiny and shrinking circulation and is a major revenue drain for the Hearst Corporation except for its downtown real estate holdings is simply another set of “facts” that never keep it from spinning its narrative of our city.

There is some good news on this front. As newspapers disappear the web is picking up some of the slack. Mission Local, a vibrant and compelling digital newspaper, has just added Joe Eskenazi as managing editor and the quality and extent of coverage of local politics is sure to increase; San Francisco Public Press does San Francisco stories with a depth and reach not found anywhere else; Fogtown blog covers housing and land use issues with accuracy and humor; Found SF offers rare insights to San Francisco history including a unique set of reprints of neighborhood newspapers; and finally here, 48 Hills, which has become the electronic newspaper of political record here in San Francisco as its readers expand beyond the Bay Area.

But more, much more must be done to reclaim for residents of the city the ability to tell their history and the history of their communities and what a compelling narrative that would be.

Finally, I am going to have to disagree with former supervisor, current Democratic County Committee Chair and my campaign ally, David Campos, and call for the repeal of ranked choice voting. In the name of full disclosure, I opposed and campaigned against RCV when it was initially proposed. But this election outcome should raise serious question in anyone mind about the efficacy of RCV.

Let me start by repeating the simple and stark facts: London Breed got 37% of the first-place votes and 23% of the RCV vote and she is now Mayor. How is she a “majority elected” mayor by either common sense or any other system but our current RCV system?

All of the pre- election blather that the election would turn on second place votes was simply wrong. It turned on first place votes, not second place votes. Kim and Leno split their first-place votes and Breed kept all of hers and was elected. The RCV strategy of maximizing second-place votes could not have been worked better than was done by both the Kim and Leno campaigns. Yet, it could not overcome the fact that both got fewer first place votes than did Breed.

There are two things fundamentally wrong with Ranked Choice Voting.

  1. Fewer people vote in RCV mayoral elections than in actual run-off elections.

Run-off elections started in 1975. Eight elections were held between 1975 and 2003. All resulted in runoffs.  Average turnout in those 16 elections was 51%. In three — 1975, 1999 and 2003 — turnout in the run-off was greater than in the general election, averaging 52%.

Starting in the 2007 Mayor election RCV “instant run-offs” went into effect. Turnout in RCV elections up to this year averaged 41%. A full 10% below the actual run-off elections.

This June’s election turnout was 53%, the highest turnout under RCV for mayor ever, but it occurred under special circumstances: a statewide governor election and a special election in D8, the highest voting district in the city, both boosting voter interest. Fully 14% of the total citywide vote came from D8.

Turnout in D6 and D10, the districts with the most African American voters, in an election with an African American woman in a close race, turned out 43% and 46% respectively

  1. Under RCV, thousands of voters never get their votes counted for the winners.

In this election a total of108,310 votes were cast for the six candidates that did not finish first or second. Only 86,753 were “transferred” to Breed and Leno. The 21,459 votes that were not “transferred” were classified as “exhausted votes” almost all being from voters who only voted for one candidate and thus had no second and third votes to transfer. It was as if those 21,000 folks never showed up for the runoff.

Of the 86,753 votes transferred only 75,722 votes were added to either Breed or Leno’s totals.

So, of the 253,487 San Franciscans that went to the polls and cast a vote for mayor, only 229,045 total votes were counted for Breed and Leno.

RCV is 14 years old. Four mayoral elections have occurred under it. Tens of thousands of voters still cast only one vote. Thousands more have their votes simply not counted in the final “RCV Slate” due to the RCV algorithm that no one knows anything about because it is a private contractors private property.

What if there was a recount, that the vote was so close that it triggered one. How would anyone check the RCV algorithm to make sure is was both accurate and fair?

RCV replaces an actual democratic vote with a gimmick, hardly the historic mission of “progressives.” The Progressive Party of the late 19th and early 20thcentury was noted for EXPANDING the vote, not replacing it. In California and a handful of other states where Progressives were victorious state law was changed to allow referendum, initiative and recall three procedures which allowed voters at elections to override elected bodies and politicians by the power of direct vote. The great national victory of the Progressive movement was the 1914 passage of the 17th amendment to the US Constitution which required the direct election by popular vote at an actual election of US senators, banning forever the practice of state legislators and or governors selecting them in secret.

At a time when voting rights are under attack why would one of the most progressive localities in the nation accept a system that “disappears” thousands of votes?

But the real problem of RCV is its political impact. Actual run-off elations give folks a second chance to make a political decision about a candidate, something “instant run-offs” don’t allow.

Last week’s essay by Campos surprisingly cites the Gonzalez /Newsom run off of December 2003 (the last actual Mayor run-off) as an argument for RCV. I say surprisingly because the 2003 runoff had a far higher turnout (55%) than did the November general (46%) and higher turnout than for any RCV election for mayor. Moreover, Gonzalez got a whopping 79,000 more votes in December than he did November while Newsom only got 46,000 more — showing that Gonzalez picked up the overwhelming majority of votes cast for the other candidates in November.

While it might be true, as Campos recounted, that the other candidatesmight have had a hard time endorsing Gonzalez, those of us in his campaign knew that their voters had no such difficulty. Gonzalez lost because his campaign manager simply did not believe that the VBM voter was nearly as important as election day GOTV. He was wrong and even with RCV both Avalos and Kim’s campaign seemed to have shared the same mistaken priority with the same result.

This November’s election will be the final act giving definition to this month’s vote. If a left progressive can be elected in D6 against two powerful pro-market rate development advocates and if a left progressive can pull off a victory in D4, both will be seen to be a defeat of Breed and her backers exposing her actual weakness implied in her narrow win in this month. If a $300 million a year tax is voted in on the most profitable businesses in the city in order to fund very-low and low-income housing development and the provision of health and human services for the poorest of our poor, then the June election will have been but a blip on the screen of the biggest urban political soap opera now playing.