The Coalitions Aren’t Lining Up

Young voters are for Bernie, older voters are for Mayor Pete, LGBTQ voters are for Warren, black voters are for Biden. Surprised? You shouldn’t be.

Buttigieg in the background, facing potential voters.
Democratic presidential candidate Pete Buttigieg at a Sunday morning service at Greenleaf Christian Church in Goldsboro, North Carolina, on Dec. 1, 2019. Logan Cyrus/AFP via Getty Images

In the lead-up to the 2016 presidential election, there was much talk about whether female voters would support the female candidate. Gloria Steinem suggested that young women preferred Bernie Sanders over Hillary Clinton only because they sought proximity to Sanders’ male supporters. Madeleine Albright told those same young women that there was “a special place in hell” for them if they didn’t help other women (presumably Clinton). Susan Sarandon, who supported Sanders in the primary and the Green Party’s Jill Stein in the general election, told the BBC that she wasn’t moved by the idea of electing the first female president because she didn’t vote with her vagina. On The Daily Show, Jessica Williams thrusted her pelvis over an imaginary voting-booth lever to demonstrate how that might work.

If this discourse seems simplistic or crude, it’s partly because, at the time, there weren’t very many examples of how women would behave when given the chance to vote for one of their own for president. White women have trended Republican for a long time, but many activists and strategists were still surprised when more than half of them voted for Donald Trump over a fellow white woman in 2016. More broadly, a great many U.S. voters have never, or only very rarely and recently, had the opportunity to support a viable presidential candidate who shares their race, religion, sexual orientation, or gender. We’re still figuring out how much shared identity influences voters.

Today, most activists would at least pay lip service to the idea that other considerations besides identity-based solidarity should come into play when assessing political candidates. But the idea that people will generally fall in line behind the person who looks or loves or believes like them still informs a lot of assumptions in political campaigns and punditry. Kamala Harris, for one, centered her campaign strategy on winning the South Carolina primary, hoping that the state’s majority-black Democratic electorate, and black women in particular, would deliver her a necessary win. When she wasn’t able to secure a lead or even maintain double-digit support in the state, she refocused on the earlier caucus in Iowa. Two months later, having made no progress, she was out of the race.

Harris isn’t the only current example of a candidate failing to earn the support of what would seem to be a natural affinity group. Black voters are overwhelmingly supporting Joe Biden, who is white. A December Quinnipiac poll of self-identified registered voters found Cory Booker, a black man, dead even at 2 percent with Pete Buttigieg, a white man who has been widely criticized for his record and rhetoric on race. (Booker ended his campaign on Monday after failing to qualify for a second consecutive debate.) Multiple polls show a plurality of young voters supporting Bernie Sanders, the oldest candidate on the Democratic slate and recent recipient of the youth-led Sunrise Movement’s endorsement. Buttigieg is the youngest, but his support from young voters is hovering around 2 percent, while senior citizens prefer Buttigieg over everyone except Biden. Women favor Biden by a sizable margin too, despite the fact that there are multiple women in the race, including Biden’s fellow top-tier candidate Elizabeth Warren. Julián Castro was the only Latino candidate in the race before he dropped out in early January, but in the course of his yearlong campaign, he was never the favorite for Latino voters. The most diverse presidential slate in history might be the one to demonstrate that identity solidarity doesn’t play a determinative role in voter behavior.

Why doesn’t it? One reason is that when a candidate is a member of an underrepresented demographic, her relationship with like-identified voters is often more fraught than a glass ceiling narrative would have us believe. These voters can be hardest on candidates who are like them, or react negatively to those candidates’ political performance in identity-specific ways. Political consultant Chris Jahnke told me in August that older white women can be some of the most sexist evaluators of older white female candidates; in these leaders, they see people who look like them enacting the very qualities—overconfidence, ambition, competitive drive—they’ve been admonished for. At Jewish Currents, Joshua Leifer writes of the age gap in Sanders’ Jewish base, “What for young Jews are sources of identification and pride—Sanders’s proletarian Jewishness, his agonistic politics—are, for many of their senior-citizen counterparts, reminders of political disappointments and quashed revolutionary hopes.” Many gay people and young people have a, shall we say, complicated relationship with Buttigieg’s politics and worldview. Personally identifying with a candidate can provoke a wide range of complex, and sometimes contradictory, reactions.

Democratic voters are also constantly wringing their hands over the question of electability, that amorphous, unknowable quality that would almost always have the party capitulate to America’s most bigoted inclinations. Black voters’ particular insight into American racism led some of them to worry that America was too racist to elect a black president. A majority didn’t swing Obama’s way until he proved he could win white voters in the 2008 Iowa caucuses. The concern over whether anyone besides a white, straight man is really “electable” still lingers in 2020, particularly in light of Clinton’s 2016 loss. It almost certainly helps explain why Joe Biden is really popular among black voters and the national front-runner right now. Biden is the familiar, establishment-favored candidate. In an election that has Democratic voters atypically more concerned with beating the opposing candidate than nominating a candidate who shares their values, an old, white, bumbling man with centrist politics might seem like the person who can best woo old, white, bumbling men away from Trump—and, indeed, that’s how Biden is pitching his candidacy. This assumption, buoyed by sexist and racist statements from “undecided” voters, points to another underdiscussed dynamic of identity politics: White men have identities too; they’re just rarely acknowledged as such.

At any rate, across-the-board support for Biden tends to drown out some telling dynamics deeper in the cross tabs. Drill down further (one helpful tool is Politico’s proportional support graphics, which disaggregate each candidate’s supporters by race, gender, age, ideology, education, and several other attributes, then illustrate how much the makeup of a candidate’s base differs from the makeup of the full Democratic electorate) and you can start to see some trends that complicate simplistic ideas about identity-based voter preferences. Sanders is disproportionately popular among black men, but not black women. He’s disproportionately popular among black and white voters who didn’t go to college, but not those who did. Warren is disproportionately popular among women, but if you add race to the equation, you’ll see that it’s entirely due to support from white women; she’s disproportionately unpopular with black women. With Biden, the opposite is true.

These nuances help clarify why identity-based predictions of voter behavior often get it wrong: People don’t only have one identity. Women and black voters and working-class voters aren’t distinct classes of people. They intersect. Each voter has multiple identities that influence her political worldview, her life experiences, and her impulses toward voting-booth solidarity. Sometimes, a woman wants to vote for a woman because she’s a woman. Sometimes, she wants to vote for a woman because that candidate is highly progressive (female voters tend to be more progressive than male ones, in part because of how each political party treats women). Sometimes, a female voter is also really wealthy, and she’d rather protect high-earners than put a billionaire-trolling woman in charge. Sometimes, she’s white and racist. Sometimes, she’s a service member who wants a veteran in the White House. Sometimes, maybe, she evaluates a candidate’s policy priorities and leadership potential, divorced from her own identity and interests, and votes accordingly. Or maybe she just picks the person all the folks in her church choir are voting for! Every voter is a mishmash unto herself.

Two recent polls of LGBTQ voters, a notoriously un- and underpolled demographic, demonstrate the fallacy of broad-brush assumptions about minority voting behavior. In one poll of 816 LGBTQ people who were likely to vote in the Democratic primary, nearly a third claimed Warren as their preferred candidate. Buttigieg, the first major openly gay presidential candidate, came in fourth place. (He got 14 percent of survey respondents, close behind Sanders and Biden.) Why might this be? Well, LGBTQ voters are far more progressive than the full Democratic electorate: Sixty-four percent of LGBTQ poll respondents said they’d prefer a candidate who’d “fight for big ideas” and “larger changes”; just 28 percent said they want someone willing to “compromise” on “incremental changes.” It makes sense that LGBTQ people would go for Elizabeth “big structural change” Warren over Pete “unity” Buttigieg, gay though he may be.

In another survey, only 30 percent of LGBTQ people who said they were planning to vote in the Democratic primary said they were even considering Buttigieg. There were also significant differences of opinion within the LGBTQ acronym: Forty-five percent of gay men said they were considering Buttigieg, compared with 29 percent of lesbians and 22 percent of bisexuals. Is that because gay men see themselves in, and want to vote for, a fellow gay man? Or because men are generally more conservative than women, who made up 100 percent of the lesbians and 75 percent of the bisexuals surveyed—both of whom preferred Sanders and Warren to Buttigieg? Why do lesbians and bisexuals prefer center-of-the-road Biden to Buttigieg? Is it the safety and electability thing? Are lesbians and bi people turned off by Buttigieg’s brand of gayness? Do they just not know him well enough yet? Or is it because the bi poll respondents were more than twice as likely to identify as black as the gay ones? Who can say! The only clear takeaways are that the LGBTQ community is not a political monolith, and that Buttigieg’s gay identity isn’t enough to win over the majority, or even a plurality, of his fellow LGBTQ Democrats.

There will always be some people who analyze candidates primarily on identity grounds. One lesbian told Politico that Buttigieg was all fine and good, “but I’m one of these women who thinks we are way overdue for having a woman in the White House. That’s a lens through which I’m going to filter my decision.” That article, which went on to claim that lesbians are “divided over which glass ceiling to break first,” was titled, “Why some lesbians don’t want Pete Buttigieg to be president.” The implied assumption here is that lesbians need more of a reason than straight voters to reject his candidacy. But there are a million reasons why lesbians might not support Buttigieg, especially in a primary slate that’s almost uniformly aligned in favor of LGBTQ rights and protections. (Are white men voting against Sanders expected to explain themselves?)

The early 2020 slate was just diverse enough to demonstrate basic lessons in uncharted territory. In 2007, in an attempt to explain why there was a 26-point gap between black women and black men who supported Clinton’s candidacy, one CNN analyst said, “Black women don’t just vote their black identity. They also vote their identity as women.” In 2020, it’s more obvious than ever that the identity explanation is just one part of the picture. Like all voters, black women support the candidates they like the best and those they believe can win: With multiple women and black people—including one black woman—to choose from in the Democratic primary, a near majority of black women voters still rallied behind a white man. As many observers have pointed out, Biden’s current popularity surely benefits from black voters’ continued reluctance to support a black candidate who hasn’t yet proved she or he can win white voters in a national election—yes, even after Obama. It’s complicated. And personal.

Too often, nervous Democrats concerned about “electability” take away oversimplified identity-based lessons from election failures. Instead of looking at Trump’s win and seeing voter suppression, the strategic failures of the Clinton campaign, and depressed voter turnout among people of color who said they didn’t like either candidate, they see evidence that American swing voters are too sexist to support a woman for president. In the Atlantic last week, Ibram X. Kendi argued for a different interpretation, reframing the narrative of the swing voter from the older white moderate who flips between parties to the young person of color who flips between voting Democrat and not voting at all or voting third-party. There are many ways to win or lose an election. But when voter identity is used to explain electoral losses, it’s rarely the younger person of color who gets the extra attention.

The electability issue is a strange one because perceptions of a candidate’s ability to win are both specious—based on biased assumptions about other people’s preferences—and self-fulfilling. Using Kendi’s framing, one could argue that the most electable Democrat is a black one, or one with a reasonably sophisticated understanding of racial justice—a candidate who might excite those black swing voters who came out for Obama in 2008 and 2012, then didn’t vote at all in 2016. But today’s black voters overwhelmingly prefer Biden; their support is essential to his current lead in the national polls, while no candidates of color have made it into the January debate. And yet: Biden is also disproportionately popular among white voters who exhibit high levels of “white consciousness” and “racial resentment,” which should raise questions about whether his messaging could truly resonate with the set of black voters Obama energized. Again, racial groups are not a monolith—Biden does better with older black voters than with the younger ones Kendi identifies as crucial swing voters. The candidate “black voters” prefer today might not be the one to capture the particular set of black voters who sway elections. And it’s worth wondering whether Biden would make more overtures to those racially resentful whites in a general election campaign, making him even less appealing to voters of color.

Identity-based arguments about electability yield all sorts of assumptions, partially because there still just isn’t much data to parse. The Democratic Party is still in the very early days of testing its own hunches about how the vast majority of its members—people who aren’t straight white men—will behave when given the opportunity to vote for a presidential candidate who shares their race, gender, or status as a sexual minority. The results are already a lot more complicated than the average armchair analyst could have predicted. At any rate, most voters aren’t looking at polling data or debriefs of previous elections to inform their assumptions about electability. They’re relying on their own worldviews. When I interviewed senior citizens about the Democratic candidates in September, several people said they thought Biden was too old for the job, and possibly too old to win. “We need someone who can get the young people excited,” one elderly woman told me. “Someone like Pete Buttigieg.”