Everyone is talking about electability —and that’s totally understandable, considering how badly Democratic voters want to beat President Trump. But it’s not totally clear from the polls which candidates can win a general election and, perhaps more importantly, which can’t. One set of polls released last week found former Vice President Joe Biden would only narrowly defeat Trump in six key swing states while the other top-tier candidates would be underdogs. Other polling finds all the top-polling Democrats — Biden, Mayor Pete Buttigieg, Sen. Kamala Harris, Sen. Elizabeth Warren and Sen. Bernie Sanders — would be very likely to defeat Trump.
So we’re left to mostly speculate. I’d expect electability talk to continue to be one of the defining features of this primary. But in my view, that conversation is lacking both nuance — at times veering too close to “Candidate X obviously can’t win” — and sophistication. Based on the coverage, you would think Medicare For All is the only policy position that really matters in terms of a candidate’s electability.
So I’m hoping, with less than 100 days before the Iowa caucuses, for a somewhat better discussion of electability. What might that sound like? Here are my three suggestions:
We should acknowledge and emphasize the uncertainty around electoral outcomes
Put simply, the amount we do know about the likelihood of different Democratic candidates defeating Trump may be greatly outweighed by what we don’t know.
Start with the polls, which have sparked much of the recent electability conversations. Surveys testing the eventual nominees conducted about a year before Election Day missed the actual margin by a huge amount, on average, in elections from 1944 through 2012.1 Year-out polls in 2016 and 2012 were closer to the mark, likely because partisan polarization has made presidential voting more predictable. Polls are inexact enough this far out that I just don’t think it’s worth making much of one candidate doing a few percentage points better than another against Trump in this or that poll.
That polarization, meanwhile, also probably limits the electability advantages or disadvantages of particular candidates. To put this bluntly, many voters are looking to back either a Democrat or a Republican — and that person’s party matters way more than their gender, race, sexual orientation or other individual factors.
“Electability is certainly still a thing, but the effect size has probably shrunk,” said Robert Griffin, an expert on voter demographics and the research director of the Democracy Fund Voter Study Group. He estimated that various presidential nominees from the same party would earn the same vote share, plus or minus 2 to 3 points, in a general election — although he emphasized that there is limited research on this question.
As Griffin’s estimate shows, there are surely some candidate effects on electability. Political science studies, particularly at the level of congressional elections, for example, have found that candidates closer to the political center do better in general elections. But even if that is true, I suspect that a fairly extreme candidate can still win in 2020 just by rallying co-partisans. Take 2016. It’s likely that a less controversial Republican, like former Ohio Gov. John Kasich or Sen. Marco Rubio, would have done better than Trump in the general election. But Trump still won.
Indeed, given Trump’s fairly low approval ratings and the high level of partisan polarization, Trump’s vote share likely has both a clear floor and ceiling. The 2020 Democratic nominee will probably enter the general election limited by a similarly narrow bound. Given this floor, Sanders and Warren are electable, even if they might be marginally less electable than Biden.
“It might make sense to talk about things in terms of a threshold, that someone is ‘electable enough,’” said Hans Noel, a Georgetown University professor and expert on party politics.
I’m not arguing that voters should disregard arguments about electability. I get why it matters. But I worry that some of the coverage in the press reads close to, “Candidate X can’t beat Trump.” Indeed, polling shows lots of Democratic voters have come to think about electability in those kinds of absolute terms. We need the coverage to hew more towards, “We don’t have much certainty about electability. Lots of candidates could win or lose. That said, Candidate X may have a slightly better chance against Trump than Candidate Y.”
It’s worth contextualizing electability
In addition to recognizing uncertainty, the electability conversation could use a lot more specifics, too. Democratic candidates could be “more electable” in different ways. Here are some common ones, all of which stem from areas where Hillary Clinton fell short against Trump in 2016:
- Winning more white, working-class voters in the Midwest (the much-discussed Obama-Trump voters, especially in Wisconsin, Michigan and Pennsylvania, the states Clinton narrowly lost) by taking more centrist policy positions.
- Winning more white, working-class voters in the Midwest by being more populist.
- Winning 2012 Obama voters, particularly blacks, who stayed home or voted for a third-party candidate in 2016 in the Midwest and in Florida in particular. I don’t think this strategy necessarily has an ideological bent — these voters might be reached by policy positions and perhaps just a candidate they connect with more, like Obama.
- Winning states with big Latino (Arizona, Florida, Texas) or black (Georgia, North Carolina) populations by maximizing turnout among people of color.
Much of the electability coverage in the news media revolves winning the white working class. It assumes that Biden, in particular, is best positioned to win the general election because his more moderate ideological positioning will better appeal to white Obama-Trump voters in Pennsylvania, Wisconsin and Michigan. That might be correct. But the evidence we have now doesn’t totally support that conclusion: Biden (+10) is doing best in head-to-head polls against Trump, followed by Sanders (+8), then Warren (+7), then Harris (+5) and Buttigieg (+5), according to averages of national polls from RealClearPolitics. That doesn’t line up with being more moderate on policy making you more electable, since the candidates are probably Biden-Buttigieg-Harris-Warren-Sanders, in terms of who is most centrist, at least on the policy positions they have proposed in this campaign.2 It should also give some pause about relying so much on early polls. Since Buttigieg in particular is less well-known than Biden, that probably hurts the mayor’s numbers right now.
And the electability discussion mostly ignores the other options: being more populist, or winning over more black and Latino voters. (It also ignores other, more speculative ways to gain an electoral advantage, like, say, a surge in youth turnout.) Take running a more populist campaign or appealing to black voters, respectively: Casting Warren and Sanders as less electable than Biden essentially assumes that the two most populist candidates in the primary won’t be able to appeal to white working-class voters, and that the two most liberal candidates can’t win over Jill Stein voters or inspire young blacks who might have stayed home over backing Clinton. Those assumptions might be true, but I’m not sure.
Even the conversation about winning the Midwestern white working class may be missing the mark. I think it’s entirely possible that Biden is perceived to be a better general election candidate because of factors unrelated to ideology and harder to talk about in public. He may be perceived as more “likable” or relatable than other candidates, a quality that might carry gender or racial undertones. Is that perception because of his actual policy stances, or because he is an older straight white man? (Also, even perceptions of his ideology might be affected by his race and gender. Research suggests that voters view women and people of color as more liberal than their actual issue positions.)
Here’s another way that electability and policy might be being linked too closely. The news coverage I read about Warren often casts her as a risky general election candidate largely because of her left-wing positions. But when I talk to Democratic voters, they often worry that her gender is the problem and that some voters won’t back her because of sexism. This is a challenging subject to talk about — but if a big part of the election is electability, it’s worth being more explicit and detailed. In assessing Warren, Democratic voters are weighing policy positions but also gender and sexism. Why is this important? Because if Warren doesn’t win the nomination and a big factor was Democratic voters felt that her Medicare for All position was a potential general election killer, that’s Warren’s fault. If we learn that voters were wary of nominating Warren because she is a woman, that’s not really Warren’s fault and speaks to more general societal challenges.
We should talk more about trade-offs
In the context of a Democratic primary, some types of electability may come at a price — a price many Democratic voters might not want to pay. That price should be more of a factor in the electability debate.
Democratic candidates trying to maximize electability, for example, may be more cautious on policy, particularly by taking stands on racial issues that prioritize America’s white majority over its people of color. In his 1992 campaign, for example, Bill Clinton went out of his way to criticize a black rapper and activist named Sister Souljah at an event put on by Jesse Jackson Sr.’s civil rights group. Black scholars bemoaned how little President Obama spoke about poverty or racial issues from 2009-12, as he positioned himself for reelection.
When Biden was asked about the “legacy of slavery” in a debate earlier this year and segued into talking about black parents needing to play records to their children, that was covered as a gaffe by the candidate. It was. It was also Biden invoking racial language that might be more appealing to Obama-Trump voters than Black Lives Matter activists.
In this campaign, several candidates are supporting a proposal to study the idea of reparations for black Americans as restitution for slavery and Jim Crow-era discrimination. Biden, though, has been noncommittal about the proposal. Reparations aren’t popular: Only 29 percent of Americans support the idea, according to a recent Associated Press-NORC poll. (So that explains Biden’s wariness.) But there’s a huge racial split on the issue — 74 percent of black Americans support reparations, compared to 15 percent of white Americans.
When we’re talking about electability, it’s worth explaining this trade-off (“Biden is wary of embracing a reparations study, likely to preserve his electability”), particularly since the 2020 campaign has focused a lot on the reverse version of this dynamic (essentially, “Sanders embraces Medicare For All, potentially reducing his electability”). In other words, to maximize electability, what liberal goals are considered expendable by Democratic candidates? Some Democrats want a female president. But is that desire outweighed by the desire to beat Trump?
To conclude, I think it’s natural that, because Democrats desperately want Trump out of office, the Democratic nomination battle has turned into a debate over electability. I just want that debate to be more appreciative of the complexity and uncertainty around all of this.