Whitepaper: What Boston’s Preliminary Elections Show Us About Deep Democracy*
Introduction
What is Deep Democracy?
“Deep democracy is the inherent belief that those at the margins should be at center and that the inclusion of all voices allows for a more complete view of the system. This inclusion takes the radical redistribution of power and privilege. It takes resisting cultural hegemony, taking over the ballot box, and winning governance power. This takes developing the political application of deep democracy.”
At Rivera Consulting, Inc. Deep Democracy is a political giving and research framework, focused on identifying specific communities or geographic areas with high concentrations of communities of color, other historically oppressed communities, and college educated voters. We develop a mixed method research approach that tracks early and vertically political contributions to candidates and movement building organizations in deep democracy locations across an election cycle. We also assess campaign infrastructure and strategy of the organizing ecosystem to determine the impact of these factors in real time.
It means that movement power can boost candidates up and down the ballot, delivering wins for progressive candidates of color and holding them accountable to communities when they are in office. Deep Democracy dynamics are apparent in Boston, but Deep Democracy is fundamentally a broader statewide vision; our work so far this year has examined Deep Democracy dynamics from Springfield to Lowell to Lawrence. While Boston’s preliminary elections brought us some disappointments — voter turnout, for instance — they also affirmed key lessons for the promise and potential of deep democracy at home: Kendra Hick’s D6 victory is a clear example of how you can buck overall trends and flip the playbook to win.
What do we analyze in this whitepaper?
This memo includes four key components. The first section is a recap of the results of the preliminary mayoral and city council elections, including interactive precinct maps for each. The next three sections each illustrate aspects of Deep Democracy present in the race, starting with an analysis of how Michelle Wu and Annissa Essaibi George built their citywide coalitions through a decade of at-large races. The third section focuses on the role of Black voters in the prelim. In the final section, we take a deep dive into Deep Democracy in District 6, showing turnout relationships between support for Kendra Hicks and support for Julia Mejia in the At-Large race and illustrating the district’s uncommonly strong turnout.
Part 1: The Results
This analysis is based on precinct-level mayoral and city council results and on block-level 2020 census data aggregated to precinct boundaries.**
In the mayor’s race, the four top candidates showed four different paths to city-wide success in Boston; while only Councilors Michelle Wu and Annissa Essaibi George advanced, each candidate’s performance illustrates an approach to city-wide politics that can form the basis for a winning coalition, and they are each worth exploring in detail.
Two candidates — Acting Mayor Kim Janey and Essaibi George — relied heavily on neighborhood bases.
Each turned in dominant performances in their base areas — Janey cracked 50% of the vote in 25 precincts across Roxbury, Dorchester, Mattapan, and Hyde Park, while Essaibi George broke 50% in 17, mostly in Dorchester, Southie, and West Roxbury — but were weak in other parts of the city. In 97 precincts (38%), Essaibi George earned less than 10% of the vote, while Janey earned less than 10% in 87 precincts (34%).
By contrast, Councilor Andrea Campbell had strong citywide appeal, allowing her to place third ahead of Janey even without a strong geographic base (she only broke 50% in one precinct).
Janey and Essaibi George lost big in their worst areas, while Campbell broke 10% in nearly every precinct, missing the mark in just 22 precincts (9%).
Councilor Campbell narrowly missed the final, but her ability to put together a city wide coalition speaks volumes about her strength as a candidate. Her results demonstrate that like Essaibi George and Wu, Campbell politicians can knit together a city-wide coalition in Boston, and bode well for her in future elections in Boston.
Councilor Michelle Wu won by combining the strengths of each approach — strong base results and salvaging votes in weaker areas.
Like Janey and Essaibi George, Wu broke 50% in many of her base precincts, reaching this mark in 32 precincts concentrated in Jamaica Plain, Allston, and Brighton.
Like Campbell, she rarely dipped under 10%, missing this mark in just 3 precincts, including the famed Ward 1, Precinct 15, where the single registered voter sat out the election.
Together, these approaches netted Wu 33% and a first-place finish in the preliminary, and they will be assets as she takes on Essaibi George in the general.
Similar geographic dynamics played out in the preliminary election for at-large city council, where four progressive candidates of color, three moderate or conservative white candidates, and former At-Large councilor Althea Garrison advanced to the general election.
The progressive candidates broadly succeeded with maps that resembled Councilor Wu’s mayoral map — base support in Jamaica Plain and strong numbers throughout the city — though Carla Monteiro’s map shades more toward Andrea Campbell’s, with base support in Dorchester and stronger city-wide results without a JP base.
While coming back from a preliminary deficit to win in a general election is rare in Boston, this illustrates one potential path for her — if the JP progressives who showed up for Julia Mejia, Ruthzee Louijeune, and David Halbert in the prelim add Monteiro to their ballots, she can narrow the gap she faces with 4th-place finisher Erin Murphy.
For Murphy and for two other more moderate white candidates — Councilor Michael Flaherty and Bridget Nee-Walsh — the maps strongly resemble Annissa Essaibi George’s in the mayor’s race.
Part 2: In the Mayoral Race, a Decade of Trust-Building Paid Dividends
November 2nd, 2021 will be the first time two women face off in a Mayoral general election in Boston, but it is far from the first time Wu and Essaibi George have gone head to head. In fact, this year’s mayoral ballot will be the 10th citywide ballot on which the two have appeared together since they first sought at-large seats in 2013. We dove into the results from four of these elections — the last four general elections for Councilor At-Large — to show how each councilor has built trust and expanded their citywide coalition since then.
Both councilors ended 2019 well-positioned for mayoral runs, and their citywide operations helped boost them into the two spots in the final this year. Still, however, the preliminary results show us that not all citywide operations are created equal; Michelle Wu’s 2019 support was a much more reliable predictor of her 2021 performance than Annissa Essaibi George’s.
Councilor Wu’s 2019 performance explains more than half of the variation in her 2021 preliminary results, and each 1-point increase in support for Wu in 2019 was associated with .7 points of additional support in 2021, implying that a high share of Wu’s 2019 voters also supported her in this year’s mayoral race.
By contrast, there was no relationship between support for Annissa Essaibi George in 2019 and her 2021 mayoral vote share.
To us, this difference is all about Deep Democracy. It is one thing to earn a vote in an election and quite another to earn lasting trust that transcends election cycles. To do so, candidates and politicians must make communities true and authentic governing partners, not just rely on their support at elections. Deep Democracy can win at the ballot box — in 2018, Ayanna Pressley’s coalition delivered her to Congress after 9 years of trust-building on the city council, and this year, Michelle Wu’s coalition is poised to elect her Mayor — but its true test comes when candidates win election to high office. If she becomes our next Mayor, Michelle Wu will need to continue building and expanding her coalition to make every Bostonian in every neighborhood a true partner in government.
In 2013, there were 3 precincts in which at least half of all voters cast ballots for Annissa Essaibi George; in 2019, there were 144. This transformation — from a factional candidate with a low ceiling outside of Dorchester to a citywide force with strength in every neighborhood — proved vital both for Essaibi George’s first at-large win in 2015 and for her strong performances in 2017 and 2019, and helped her advance into the final this year.
In this year’s preliminary election, however, her map shaded more toward her 2013 performance than her 2019 one, as she rarely cracked 20% outside of her base in Dorchester and the largely white enclaves of South Boston, West Roxbury, and Charlestown. As we wrote in our election preview, these are a key political region of Boston and a key part of winning coalitions for moderate and conservative candidates in the city, but they do not provide enough votes to win without support elsewhere. Essaibi George’s 2019 results show that she can earn votes anywhere in the city, and she will need to do so to make the general election competitive.
While Annissa Essaibi George’s 2013 run set the stage for successful bids down the road, Michelle Wu broke into Boston politics with a second-place finish as a first-time candidate that year, finishing just 1,058 votes behind Ayanna Pressley at the top of the ticket. Her progressive base in Jamaica Plain and competitive city-wide operation — she appeared on at least a quarter of ballots in 239 of Boston’s 255 precincts that year — were enough to deliver a seat in 2013, but since then, she has emphasized city-wide outreach and trust-building and build a base in nearly every neighborhood. In 2019, 62.2% of Boston voters included her on their at-large ballots, and she cracked 50% on this metric in 204 of 255 precincts.
Part 3: Black Voters Strongly Supported Kim Janey, and Will Likely Decide the General
In a historic election cycle, three of the five major candidates for mayor were Black Bostonians, and two of the top four were Black women. In a narrowly decided election, though, both lost out to At-Large City Councilor Annissa Essaibi George by just over 3,000 votes. Importantly, this result did not reflect low turnout among Black voters; turnout in predominantly Black precincts largely mirrored the city-wide turnout rate of 24.7%.
Since the election, several analyses have accused Acting Mayor Janey and Councilor Campbell of “splitting the Black vote.” This fundamentally misconstrues two key facts. First, Black voters in Boston are not a monolith. Black candidates earn support from non-Black voters, and non-Black candidates earn support from Black voters.
Second, our analysis suggests that voters in many predominantly Black precincts did consolidate their support behind Mayor Janey this year. 83% of the variation in Kim Janey’s vote share was explained by the share of non-Hispanic Black voters in a precinct in the 2020 census, and for every 1-point increase in the share of Black voters, Janey’s vote share went up by .69 points. While it is impossible to discern individual votes from precinct-level data, these results are consistent with Janey winning around two thirds of votes cast by Black Bostonians this year.
These results mean that most Black voters did not see their preferred candidate advance to the general election; this reality presents a new opportunity for community accountability for Boston’s next Mayor.
Precincts with more Black voters cast far more ballots for candidates who did not advance, meaning that these precincts have many of the undecided voters who will decide the general election for Councilor Wu or Councilor Essaibi George.
In the precinct with the highest share of Black voters in Boston — Mattapan’s Ward 18, Precinct 3 — more than three quarters of voters cast ballots for a candidate who did not advance, and this rate is typical for many other predominantly Black neighborhoods in Roxbury, Dorchester, Mattapan, and Hyde Park.
While the next Mayor of Boston will not be Black, she cannot win without making Black voters part of her governing coalition, both on November 2nd and for the next four years. Black voters will continue to play a formidable role in shaping the future of the city and state.
Part 4: Deep Democracy in District 6
The preliminary election also gives us a clear success story for Deep Democracy: Kendra Hicks’s City Council campaign in District 6 and her role in shaping turnout in the at-large and mayoral races. Many political pundits see “coat-tails” as a one-way phenomenon, where races at the top of the ticket drive turnout for races further down the ballot. In Deep Democracy, however, campaigns for any office can transform the electorate by building the deep relationships needed to turn out new voters. We see Deep Democracy in three key aspects of this race: the Hicks campaign’s organizing approach, the elevated turnout relative to other districts, and the strong correlation in results between Kendra Hicks and Councilor Julia Mejia, who runs similarly community-driven campaigns prioritizing similar deep engagement.
The Campaign*** and Voter Turnout
Shortly after the preliminary election, Hicks campaign chair Mike Leyba reported that the campaign had knocked on over 25,000 doors and made 66,000 phone calls. This effort was powered by volunteers signing up for over 1,000 voter outreach shifts, and the campaign also mobilized 1,000 voters through relational organizing. These numbers would be impressive for many at-large campaigns, but in a District race where just over 21,000 voters showed up at the polls and just over 18,000 voted in the City Council race, they are astounding. While citywide turnout failed to break 25% this year, turnout in District 6 was dramatically higher, as 36.9% of registered voters cast ballots in the election. This rate is by far the highest of any of Boston’s nine city council districts.
Part of this disparity is based on the neighborhood composition of the district, as West Roxbury traditionally has among the highest turnout in municipal elections. Still, turnout in District 6 was far higher even than the district’s baseline. In 2017’s general election, when slightly more voters voted citywide than in this year’s prelim, District 6 cast 18,122 votes; this year, with slightly fewer voters participating citywide, the district cast 21,185 votes. Amid lower overall turnout, this increase implies that thousands of voters in the district turned out in 2021 based on a dynamic that did not exist in 2017: a contested district council election and an at-large election with a strong movement-driven candidate building her political base in District 6.
Deep Democracy Alignment
Deep Democracy-aligned organizing clearly shaped the result of the District 6 election, but we are also impressed by the role District 6 played in shaping At-Large and Mayoral results. This nexus is clearest in the relationship between results for Kendra Hicks and those for Julia Mejia in the at-large race; while we are focused on Hicks’s campaign, the Mejia campaign also ran an impressive organizing operation, and it is likely that efforts from each campaign boosted the other. Hicks’s vote share was strikingly correlated with the share of ballots on which Mejia received votes. Variation in Hicks’s results explained 96% of the variation in Mejia’s results in the district’s 36 precincts, and for every additional point of vote share for Hicks, Mejia gained .79 points. This relationship holds both in areas where both did well and in those where both trailed their opponents, and implies that throughout the district, most voters who supported one candidate also backed the other. A similar but less pronounced result holds in the Mayor’s race, where Hicks’s vote share had a strong positive relationship with votes for Michelle Wu, weakly positive relationships with Kim Janey and Andrea Campbell, and a strong negative relationship with Annissa Essaibi George.
In November, turnout dynamics in District 6 will play a critical role in all three of these races. The mayoral matchup features two candidates who fared well in the district, with Wu drawing more from Jamaica Plain and Essaibi George more from West Roxbury. The at-large race features two progressive women of color — Julia Mejia and Ruthzee Louijeune — who relied on District 6 to power big wins in the prelim, and a third — Carla Monteiro — who will need more support in the district to overcome the deficit she faces against a more conservative candidate. And in the District 6 race, Kendra Hicks will need to maintain and expand the enthusiasm that powered her preliminary win. District City Council races often fly under the radar, but on the morning of November 3rd, we may look back at District 6 as a new engine of Deep Democracy in Boston.
Part 5: Data Limitations and Conclusions
A key limitation of this analysis is that election data cannot tell us how individual voters behave beyond precinct boundaries. For some analyses, Boston’s residential segregation and very strong voting patterns allow us to make inferences about individual voter behavior based on precinct-level data — for instance, we can be quite confident that most Black voters voted for Kim Janey, that most Kendra Hicks supporters in District 6 also voted for Julia Mejia, and that Michelle Wu retained more of her 2019 at-large city council supporters than Annissa Essaibi George. Precinct-level data is always inadequate for exploring individual voters’ behavior, though, and this is clear in discussing three key electorates in Boston. Latino voters, AAPI voters, and young voters (18-35).
Throughout the election, pundits speculated about how “the Latino vote” would break. This is not a question that we can answer through precinct-level data. Michelle Wu did well in all six of the precincts with the most Latino voters in Boston, winning between 35% and 42% of the vote in each, but all six of these precincts are in East Boston, and it is just as possible that they broke strongly for Wu based on her plans to address coastal flooding or her endorsement from District 1 City Councilor Lydia Edwards. In Dorchester’s Ward 14, Precinct 9, where 50% of residents identify as Latino, she won 14.6% of the vote, among her worst showings anywhere in the city.
Latino voters in Boston are as varied and diverse as the city as a whole, and while we cannot make any concrete claims about how Latino voters voted based on the precinct-level data we have, evidence from public opinion polling suggests that Latinos broke largely as the city did, giving a first-place finish to Michelle Wu and breaking roughly evenly between the other three major candidates for second. In a Suffolk/Boston Globe poll conducted just before the election, Michelle Wu led Latino respondents with 26% of the vote, followed by Kim Janey with 20% and Andrea Campbell and Annissa Essaibi George with 17% each, with 14% still undecided. Still, a single poll represents just a tiny fraction of the citywide Latino electorate; in total, only 65 people who responded to the poll identified as Hispanic or Latino.
This does not mean that it is impossible to analyze the voting patterns of Latino voters or other hard to reach communities in Massachusetts; on the contrary, at Rivera Consulting, Inc., we know this kind of research is possible. If we can do it for the Census count, we can do it for our democracy work. It means having a qualitative research approach that compliments traditional electoral research and most importantly, has community experts and organizations like Mass Voter Table, Mass Budget, MassAllance and community hubs across the state who organize their communities driving the work itself. Understanding political preference is key to understanding near to long term power building opportunities locally — a core tenet of Deep Democracy Massachusetts.