RCV

Nov 012018
 

By Pedro Hernandez

San Leandro didn’t make a name for itself by sticking with the status quo.

The city of progress embraced innovation at every turn, from the iconic downtown pedestrian path to its fiber optic loop connecting businesses and residents to free wifi. It’s no surprise that San Leandro was at the forefront of election reform, joining its Bay Area neighbors to bring ranked choice voting (RCV) to city elections.

It was a logical fix for San Leandro, easily and effectively eliminating the costly top-two runoff while preserving the democratic principle that candidates need majority support to win elections.

The easy, 1-2-3 system empowers voters to rank their top three preferences on the ballot. If no candidate has a majority of first choice votes after the initial tally, the candidate with the fewest votes is eliminated. Those who ranked that candidate as their first choice have their ballots counted for their next choice. This instant-runoff continues until a candidate wins with more than 50 percent of the votes.

The 2010 debut in San Leandro – the same year as Oakland and Berkeley began using RCV – proved an early success for the more fair and democratic voting system.

Voters took to the polls in droves – turnout in the mayoral race jumped 11 percent compared to the November 2006 runoff, and a whopping 54 percent more than the June 2006 primary.

The ballots cast showed voters understood and appreciated the new system right away – 75 percent ranked at least two of the five mayoral contenders.

And the exceptionally close nature of the six-way contest – just 74 votes separated incumbent Mayor Tony Santos from challenger Stephen Cassidy in the first round – made voters’ choices all the more impactful, with more than 85 percent of ballots still counted through the sixth and final elimination round.

Though Santos carried a narrow lead in the first round, Cassidy’s win was the most fair outcome. Cassady appealed to more voters, allowing him to pick up the second and third choices needed to secure a majority.

The path of progress continued with successes in San Leandro’s subsequent council and mayoral RCV elections, which engaged high numbers of voters to cast accurate ballots that gave rise to a group of elected officials who not only received the broadest possible support of their constituents, but better reflected the city’s increasingly diverse population.

The four races decided by in 2014 were held in one election with RCV instead of two. It gave San Leandro its first city council that was majority-female and people of color, including the first elected representative of the Latino community, Councilwoman Corina Lopez in District 5.

The more reflective city hall is part of a larger trend in cities across the Bay Area that have seen significant increases in representation for women and people of color since adopting ranked choice voting, as several studies have shown.

And voters continued to give top marks to ranked choice voting, indicating they understood it and wanted to keep it, including previously disenfranchised groups such as young people, people of color and low-income voters, according to a 2014 survey.  Ranked choice voting rewards candidates that seek support from beyond their base, which in turn, leads to more civil and issue-driven campaigns.

The upcoming ranked choice voting elections on November 6 give San Leandro another chance to remind the rest of the country that we lead the way toward a better future, including in our government elections. Show up, rank your ballot and stand for the progress, fairness and diversity.

 

Sep 252014
 

2014 Edition

Here is the scoop on how to vote and how the votes are counted.

San Leandro instituted ranked choice voting  (RCV) in 2010.  This will be our third election using it.  I am a fan of RCV  for a couple of reasons.  First, it saves the city money to only have to conduct one election for Mayor/City Council – rather than an election and then a runoff.  Second, it gives people more of a choice and allows voters to cast “protest” votes without fearing that this will help the candidate they like the least.  Third, it costs less for a candidate to run one campaign rather than two (a regular one and a runoff).  The cheaper the campaign, the less the candidate is indebted to his contributors

RCV has its detractors.  Candidates that have a strong but discreet base (e.g. ethnic or other affinity groups) may prefer a plurality system, as they may be able to win without having to appeal to the majority of voters.  Candidates that have a lot of money may prefer runoff elections, as the primary election may exhaust their opponent’s funds, making it easier to defeat them in the general.  Special interest groups dislike RCV both because it makes it more difficult to predict who will win a race and therefore whom they should back, and because it dilutes the power of their endorsement and financial contributions.

How to Vote

Voting in a ranked choice election is relatively easy.  The ballot shows three columns marked “first choice”, “second choice” and “third choice”.  Below them are the names of the candidates.  You vote by completing the line next to the name of the candidate you prefer for that choice.  Note that you must only mark one candidate under each choice, as otherwise your vote will not count. Importantly, ranking more than one candidate makes it more likely your vote will keep counting until the end of the count, and ranking a lesser choice never hurts your first choice.

To illustrate how RCV works, let’s say we have an election with four candidates: Smith, Chan, Jones and Garcia.  Jones and Garcia are the two leading candidates, Chan is the middling candidate and Smith is the protest candidate – with a strong message but no chance to win.  You support Smith’s message, but you want to make sure Jones does not get elected.

On your ballot, under “first choice” you fill the line next to Smith.  For your second and third choices, you select Chan and Garcia.  While the order is not important for your goal of making sure that Jones doesn’t get elected, if you mark Garcia as your second choice, chances are that your third choice, Chan, will not be counted, as Chan is likely to be eliminated before Garcia. If you prefer Chan to Garcia, you should put Chan as your second choice.

No matter how many candidates run, there are only 3 choices you can list.  In a race with four candidates, this is not an issue, you just vote for the three candidates you prefer.  In a race with 5 or more candidates, you may want to make sure that your third choice goes to a “safe” leading candidate, the least-bad option of the candidates likely to win.

How Votes are Counted

The process for counting votes is somewhat complicated and not in the least intuitive. Votes are counted in a rounds.  An algorithm is used to determine how many votes each candidate has at the completion of the process.  Here is how it works:

First Round: First Choices only

In the first round, the registrar will only count the number of valid first choice markings each candidate has received.  Second or third choices will not be considered at all.

If a voter neglected to mark a first choice on her ballot, but marked a second (or third, if she didn’t mark a second either), then that second/third choices will be counted as a first choice.  On the other hand, if a voter marked more than one candidate as his first choice (an overvote), then the ballot will be considered invalid and will be discarded, as the intention of the voter is not clear.  This ballot will not count to the total of votes cast on that or subsequent rounds.

If any candidate gets 50% +1 of first-choices, that candidate is elected.  If no candidate gets that many, then we go to the second round.

Let’s assume that 10,000 valid votes were cast in our imaginary election, with the first choices distributed in this manner:

Jones: 4,000
Garcia: 3500
Chan: 2000
Smith: 500

None of the candidates has 50% +1 of the first-choices  – they would need 5001 – so Smith, having the least votes,  is eliminated and the count goes into the second round.

Second Round: First Choices of top vote-getters + 2nd choices of lowest vote-getter

In the second round the registrar looks at the second-choices on the ballots that had marked the eliminated candidate as the first choice (Smith in our example).  The second-choices are added to the count of the remaining candidates’ votes.  Note that the registrar does not look at the second-choices for the ballots that have any of the remaining candidates marked as their first choice. That’s because every voter has one and only one vote, and your vote never counts for more than one candidate at a time.

Once again, if a voter has not marked a second choice, but has marked a third choice, then the third choice will be treated as the second choice.  If the voter marked the same candidate as both his first and second choice, and that candidate is eliminated, then the registrar will look at the third choice and treat it as a second choice.  If there are no 2nd or 3rd choices marked, or if more than one candidate is marked as a 2nd choice, that ballot will be considered exhausted/invalid and discarded and won’t count towards the total votes in that round.

In this and subsequent rounds, the registrar will count the number of ballots that remain valid, and calculate the percentage of the vote based on that number.

In our example, Smith got the least amount of first-choices so he’s out.  Now we look at the 2nd-choices of the 500 people who selected Smith as their first-choice.

2nd choice on 500 ballots that marked Smith as first choice

– Jones: 150
– Garcia: 200
– Chan: 100
– No or Invalid votes: 50

Totals after 2nd round

Total votes: 9950

– Jones: 4000 + 150 = 4150 votes (42%)
– Garcia: 3500 + 200 = 3700 votes (37%)
– Chan: 2000 + 100 = 2100 votes (21%)

None of the candidates have gotten the required 50% + 1 of the vote, so the candidate with the least amount of votes is eliminated (Chan in this case) and we go to the third round.

Third Round: First Choices of top vote-getters + 2nd or 3rd choices of 2 lowest vote-getters

In this round, the registrar will count the second choices on the ballots that had marked the most recently eliminated candidate as first choice (Chan in our example), and the third choices on ballots that had marked the two eliminated candidates as their first and second choices (Smith and Chan in our example).

2nd-choices on 2000 ballots that marked Chan as first choice

– Jones: 650
– Garcia: 750
– Smith: 200
– No or invalid votes: 400
– Total votes so far: 9550

3rd choices on the 200 ballots that listed Chan as first choice and Smith as second choice

– Jones: 100
– Garcia: 50
– No or invalid: 50
– Total votes so far: 9500

3rd choices in the 100 ballots that listed Smith as a first choice and Chan as a second choice

– Jones: 20
– Garcia: 60
– No or invalid: 20

The total number of valid votes cast after the third round is: 9480

By adding the third-round votes to candidate’s totals we get:

– Jones: 4150 + 650 + 100 + 20 = 4920
– Garcia: 3700 + 750 + 50 + 60 = 4560

Jones got 51.9% of the vote, and thus he wins.

Gaming RCV

After much thought, analysis and reading, I am confident that it’s impossible for voters to “game” ranked choice voting.  With four or less candidates in a race, you really should vote for candidates in your order of preference.  There are a couple of caveats:

1)  If you want to cast a protest vote, mark that candidate first, otherwise your vote may not be counted for her, and nobody would know about your protest.

2) If you are a candidate,  you can become a second choice  by teaming up with another candidate in your race.  However, this can be a dangerous strategy.  If your opponent/collaborator is a strong candidate, your help may make her win.  If, on the other hand, she is a weak candidate, associating with her may hurt your standing with voters.  A better strategy is to ask voters who have already committed to one of your opponents to mark you as their second choice.

3) If there are more than four candidates, the  only change in strategy is to try make sure one of your  three ranked candidates is likely to be one of the two strongest candidates in the final round. In such a scenario, your first choice is  for the  candidate you like the most, your  second choice is for your 2nd favorite, and your  third  choice is for a candidate you can live with.

Note that while I speak about what the registrar will do, in reality the actually calculations are done by a computer using a preset algorithm.

Please let me know in the comments below if you have any questions.


This article was written with information provided by the Alameda County Registrar of Voters and Fair Vote.  It has been substantially revised for the 2014 election.  The omment below reference the original article written in 2012.  My gratitude to Rob Richie from Fair Vote for his invaluable help with this article.

Sep 172014
 

In the last week there have been two poorly-advertised and poorly-attended Mayoral and City Council candidate fora in San Leandro.  Mike Katz-Lacabe tweeted from the Mayoral fora.  He’s running for City Council himself, so he couldn’t report on that part of the fora, though he did note some of the “lightening questions” from the first forum.

Update: See also responses to the APA Caucus questionnaire below.

City Council Candidates

All San Leandro City Council candidates favored a marina with small boats – in other words, no support for paying to dredge the channel.

All San Leandro City Council candidates said that they did not support surveillance cameras throughout city. Leah Hall was late so no answer

San Leandro City Council candidate Deborah Cox said she supports marijuana dispensary but spoke against it at June 18, 2012, City Council meeting

All San Leandro City Council candidates say they support marijuana dispensary except Lee Thomas.

All San Leandro City Council candidates support ranked choice voting except Dist. 1 candidates David Anderson & Deborah Cox.

Mayoral Candidates

Dan Dillman says San Leandro’s pressing problem is perception. It’s a beautiful city.

Pauline Cutter says San Leandro’s most pressing problem is economics.

Diana Souza says San Leandro’s most pressing problem are the streets.  (The street conditions decreased every year she’s been a Councilmember).

Mayoral candidates on Marina: Cutter: exciting new development planned. Dillman: what voters want. Souza: new restaurants, hotel, conference center

San Leandro mayor candidate Souza asks for other candidates’ views on rent stabilization. Cutter: we need to consider. Dillman: what voters want.

San Leandro mayoral candidates on city staffing: Souza & Cutter: more cops. Dillman: use police from CHP, BART, Sheriff, Parks.

San Leandro mayoral candidate Diana Souza says working poor can be helped by recreational programs for youth, seniors and adults.

Breaking news: All San Leandro mayoral candidates support transparency at City Hall. Cutter & Dillman mention improving meeting minutes.

San Leandro mayoral candidates on red light cameras: Cutter and Dillman oppose. Souza supports. Thinks they save lives.

San Leandro mayor candidates on SLPD acquisition of armored personnel carrier: Dillman opposed, Cutter researching, Souza supports.

San Leandro mayor candidates on Measure HH: (sales tax increase for 30 years) Dillman opposed to length. Cutter & Souza support HH.

San Leandro Mayoral candidates on whether they support marijuana dispensary: Cutter and Dillman: yes; Souza: No.

San Leandro Mayoral candidates on whether to keep ranked choice voting: Cutter says yes, Souza says no & Dillman says “what voters want.”

San Leandro Mayoral candidates on flying the flag of other countries: Cutter says no, Souza says yes, and Dillman says: whatever voters want.

Note: During the interviews for the Democratic Party endorsement, Souza and Cutter clarified that they are in favor of surveillance cameras, just not throughout the city.

APA Caucus Questionnaires

While many organizations ask candidates to fill out questionnaires, very few actually make the answers public.  The Asian Pacific American Democratic Caucus of Alameda County is the exception.  Here are the answers from San Leandro Candidates to APA Caucus questionnaires:

San Leandro, Mayor

San Leandro, City Council

District 1

District 3

District 5

Aug 272011
 

Tony Santos is at it again.  Since being defeated by Stephen Cassidy in the November 2010 San Leandro mayoral election, Santos has made no secret of his bitterness and resentment.  He has blamed anyone and everyone (excluding himself) for his defeat and has committed his forced retirement to the cause of defeating “ranked choice voting” throughout the country.  He’s written letters, had himself interviewed, and in his latest stunt, he testified in Sacramento against making it easier for California cities to institute ranked choice voting.  He is sure, he will tell you, that ranked choice voting was the ultimate culprit for his loss.  And he may be right.

It is difficult, even for insiders, to get a good grasp of the 2010 mayoral campaign in San Leandro.  Santos  had been in and out of City Hall for decades.  He was running for re-election as Mayor with the support of Labor, fire fighters, and, at the 11th hour, the police.  He was confident, even cocky, that he would easily win.  After all, no elected-mayor had ever lost  re-election in San Leandro (one appointed mayor, who’d promised not to run prior to his appointment, did lose).  Incumbents have many advantages in city races: they enjoy name recognition, can easily gather up endorsements and have a much easier time raising money – nobody wants to say no to a sitting Mayor or City Council member.  In addition, voters who don’t pay particular attention to city politics and are not extremely dissatisfied with city services, tend to vote for incumbents.  For this reason, sitting San Leandro mayors seldom face serious competition; it was almost incomprehensible to Santos that he would face not just one, but two credible opponents.

Council woman Joyce Starosciak seemed like an unlikely Mayoral contender.  In 2010, she was half-way through her second term in office, and could list few accomplishments and fewer long-term plans for San Leandro.  If her intentions were to actually run the city, it’d have made the most sense for her to wait until 2014 to run, when Santos would be termed out.  Starosciak, however, was in a hurry. As Assemblywoman Mary Hayashi’s heiress apparent, Starosciak was posed to run for Assembly in 2012. She’d be better positioned to win if she did so as Mayor of San Leandro, than as a mere City Council member.  She had good reason to believe she could win: a top campaign manager, lots of money (including some from a Hayashi-associated PAC), and support from the police (her husband is a sheriff deputy with Alameda County).   Most importantly, Starosciak was a woman.  After Ellen Corbett was elected to the California Senate in 2006, defeating better funded male opponents, the conventional wisdom within political circles was that women had a natural advantage in democratic districts (and San Leandro is heavily democratic). Starosciak’s main message during her campaign was indeed “I’m a woman and a native of San Leandro.”  It was a message that Santos feared.

Stephen Cassidy’s reasons for running in 2010 were murkier.  He had been elected to the San Leandro School Board in 2004, after defeating a well-connected incumbent, but after four tumultuous years in the Board he’d chosen not to seek re-election (he had, however, engineered the election of two friendly candidates in his stead).  He had long been rumored to have further political ambitions so it wasn’t a complete surprise when he decided to run for Mayor.  He did so basically with a message of competency: “elect me and get someone who can actually read a budget and think long term.”  Cassidy’s political strategy was to appeal to distinct voter groups with slightly different messages: to conservatives and the elderly, with a message of fiscal responsibility; to parents, with promises of better relations between the cities and the schools; to Spanish and Chinese speaking voters, with fliers in their own languages.  Cassidy’s underdog status meant that he was unable to get big donations from corporations and PACs, but he made up for it by building an impressive grass-root apparatus and tapping into the professional experience of his supporters.  Cassidy recruited a former evangelist for Apple Computer, an expert on messaging, as his campaign manager.  A local graphic designer put tens of hours of work into his campaign literature and his website  while a top computer programmer provided back end and mailing list support. Another techie was in charge of his voter data analysis.  The other candidates had to pay tens of thousands of dollars for similar services.  Meanwhile, an army of volunteers organized fundraisers and house coffees, walked, telephoned, wrote letters and passed on fliers for his campaign (full disclosure, I helped Cassidy with his campaign in the initial months).  Cassidy himself spent over a year walking the streets of San Leandro and introducing himself to voters.  His was a grass-root campaign at its best.

The November 2010 San Leandro city election was the first time in which ranked choice voting (RCV) , also known as “instant runoff voting,” would be used in San Leandro.  The city had amended its charter to permit ranked choice voting early in 2000, but had only implemented it a few months before.  The appeal of RCV to city staff was that it allowed for both general elections and runoffs to be conducted at one time, saving staff time and the city money.  Progressive members of the City Council liked that RCV allowed elections to be held in November (rather than June), which meant they’d have greater voter participation.  However, RCV was not popular with all City Council members.  Starosciak, in particular, wanted to change San Leandro’s elections back to a plurality system, where the candidate that got the most votes would win, even if s/he did not reach 50% of the votes.  Given the assumption that female candidates had a natural advantage with the electorate, Starosciak figured a plurality system would virtually guarantee her winning.  Her plan never got much traction with the other members of the Council, however.

As discussions for implementing ranked choice voting got more serious in early 2010, then-Mayor Tony Santos became its biggest advocate.  He was seduced by the rhetoric of Fair Vote, an organization that lobbies for RCV, and convinced by his friends in Labor to push the City into implementing RCV for the 2010 elections.  He went as far as suspending a tied vote on RCV, so that a City Council member that had been absent for the vote could break the tie and have the ordinance pass.  Santos later said that supported RCV because he had been told RCV mostly benefited incumbents.

As the 2010 Mayoral race developed, none of the campaigns had a firm grasp on who the top contenders were and how RCV would play into the equation.  The Santos campaign conducted an informal telephone poll at the end of the summer which showed Santos in the lead, with Starosciak a distant second.  Though the poll was methodologically flawed, Santos took it as a sign that he would easily win and that Cassidy wasn’t a threat of any kind.  While he delighted in attacking Cassidy through a friendly blogger, he didn’t feel compelled to put much effort into his campaign.  He rarely walked, never fliered, and his first mailer did not even arrive until several days after absentee voters had received their mail-in ballots (ideally, you want to have your mailer arrive at the same time as the ballots, so your name is fresh in the mind of those voters who like to vote right away).  It wasn’t until mid-October that Santos’ campaign realized that Cassidy posed a considerable threat: more and more people  they encountered were openly supporting Cassidy, the city was blanketed with his lawn signs, and letters to the papers were overwhelmingly in Cassidy’s favor.  Santos’ campaign quickly issued a mailer answering Cassidy’s charges against him and put forth robo-calls from, among others, Senator Ellen Corbett, urging voters to vote for him.  This helped him enormously with poll-day voters, but it was too late to impact those who had already mailed their ballots.

Starosciak’s campaign was in trouble from the moment Cassidy entered the race.  As most challengers, Starosciak needed to run on a message of  “change,” but as a sitting member of the City Council there was just so far she could carry this message without convincing voters that they should vote for Cassidy, the outsider, instead.  Cassidy’s criticism of City Council’s actions – like spending all the city’s reserves – while primarily aimed at Santos, also applied to her.  She couldn’t answer the charges, however, without helping Santos’ own campaign.  Her strategy, then, seemed to be to keep out of the fray and run on her gender, personal charm and San Leandro native status.   While she ran a more energetic campaign than Santos, and hired people to walk and make phone calls for her campaign, she never really acquired much momentum.  By the fall, even the police had abandoned her and co-endorsed Santos.

Cassidy, meanwhile, kept pushing his message of “competence” and “fiscal accountability,” accusing City Hall of mismanagement at every possible turn.  His big issue was “pension reform” for City employees.  This made him terribly unpopular with the employee unions and Labor but hit a nerve with local voters as even the Democratic candidates for Governor and State Treasurer started to talk about the need for pension reform at the state level.

Of the three campaigns, the only one that seemed to take RCV into account was Cassidy’s.  As he and his supporters walked and called voters, they specifically asked those who expressed support for one of the other candidates, to mark Cassidy as their second choice.  Neither of the other two campaigns seem to have done this.  Indeed, Santos’ campaign showed contempt for the whole RCV process by telling supporters to not mark second or third choices.

One of the biggest criticisms of ranked choice voting is that final election results were not available until weeks after election night.  This proved to be the case in San Leandro.  Election night results had Santos winning the first round by 62 votes (out of over 20,000 cast); two weeks later, when all second and third place votes had been counted, Cassidy was ahead by 232 votes and was declared the winner.  Santos could not accept the results, and to this day he has not conceded defeat.

At first glance, it would appear that Santos is right when he blames RCV for his defeat.  He did have 62 more 1st-place votes than Cassidy and in a plurality system like the one Starosciak advocated, Santos would have won.  However, San Leandro specifically got rid of plurality elections in 2000, after Ellen Corbett and Shelia Young became mayor with less than 50% of the total vote.  San Leandrans did not want a mayor which the majority of the people had not elected.

Had RCV not been implemented in San Leandro, the Mayoral elections would have taken place in June 2010, with the two top candidates facing a runoff in November. Given the lack of independent, accurate polling data (polls are too expensive to conduct for most local elections), it’s impossible to say what would have happened if that had been the case.  Starosciak only received about 20% of the first-choice vote in November, so it’s reasonable to believe she wouldn’t have been one of the two top vote-getters in June, but a longer campaign cycle together with the increased media attention to Cassidy’s pet issue, pension reform, probably benefited Cassidy considerably.  It definitely allowed him more time for both campaigning and fundraising.

But had Cassidy and Santos been the top vote-getters in June, it is by no means clear who would have won in November.  Santos would have surely benefitted both by knowing who his real opponent was and how popular he was.  If the election had been as close in June as the first round was in November, Santos would surely have seen the need to step up his campaign.  Given how well he did with poll-day voters once he took campaigning seriously, that may very well have given him a victory.

On the other hand, a June-November election cycle might also have benefited Cassidy. Cassidy’s main problem, as is the case with all underdog challengers, was raising enough funds to put forth a credible campaign. A good June placement would have demonstrated to political contributors that Cassidy was a very viable candidate and would have likely opened their wallets.  Such placement would also have re-energized Cassidy’s volunteers and brought additional supporters into the campaign.

An additional factor that cannot be accurately assessed is who Starosciak would have endorsed.  After her November defeat she was quite bitter with both Santos (who made a series of serious personal attacks against her during the campaign) and Cassidy (whom she sees as having stolen the election from her), so it’s hard to envision her supporting either.

In any case, we will never know. We also don’t know how the November 2010 San Leandro RCV election will affect future elections in San Leandro or elsewhere.  Future candidates will hopefully learn of the importance of seeking 2nd and 3rd choice votes.  Note that this does not mean that candidates will necessarily run “kinder,” less negative campaigns.  Candidates in races with clear front runners need to attack their records in order to take votes from them.  If anything, what an RCV election may do is focus political attacks on the front runner(s), but that is just as true in regular elections.

When all is said and done, Santos has nobody to blame for his defeat but himself: he antagonized voters by treating them callously during public fora and by taking unpopular positions (e.g. not supporting San Leandro Hospital, dismissing fears about crime and supporting a large affordable housing project), he put little effort into running his campaign and he did not take RCV into consideration in his campaign strategy.  Other incumbents can learn from his mistakes.