PollerCoaster 2020: The Electability Puzzle | Crooked Media
25% OFF NEW ANNUAL SUBSCRIPTIONS—Join Friends of the Pod Today! 25% OFF NEW ANNUAL SUBSCRIPTIONS—Join Friends of the Pod Today!

PollerCoaster 2020: The Electability Puzzle

Top Stories

Crooked Media has partnered with Change Research to conduct a series of Democratic primary polls. The first entry in this PollerCoaster series is now available. You can see all our crosstabs here, browse mobile-friendly toplines here, read open-ended responses here, and download the raw data here.


After the trauma of the 2016 election, it’s natural that prospective Democratic primary voters are fixated on nominating the presidential candidate they believe to be the most electable. But it is increasingly apparent that these voters have formed questionable impressions of what constitutes electability, and in many cases, thrown their support not to the candidates they prefer most in the abstract, but to ones they believe, for one reason or another, will fare better in the general election against President Trump.

That’s one of the key takeaways of Crooked Media’s survey of Democratic voters in Iowa, New Hampshire, and South Carolina.

We asked 935 voters from Monday, June 17 to Wednesday, June 19 not just for their candidate preferences, but for their second choices, and for information about how they reached their decisions.

Like many pollsters, we found four candidates at the top of the pack. Joe Biden leads with 29 percent, followed by Bernie Sanders with 20 percent, Elizabeth Warren with 19 percent, and Pete Buttigieg with 14 percent. No other candidate has double-digit support. 

These same four candidates dominate the second-choice rankings as well, but the ordering there looks much different. Warren tops the list of respondents’ second choices with 20 percent, followed by Biden with 18 percent, and Sanders and Buttigieg with 15 percent. Kamala Harris is the top choice of only five percent of respondents, but 11 percent say she is their second choice. 

These rankings mask the extent to which many voters have selected candidates on the basis of perceived electability. Yes, each candidate’s supporters tend to think their candidate would be the best president. But 22 percent of respondents said they prefer their top choice candidate over their second choice candidate because they believe their top choice “has a better chance of beating Trump.” Of those voters, a full 45 percent believe their second-choice candidates would fare worse against Trump because they worry that average Americans won’t like their second-choice candidates as much as their top choice, and this is especially true of people who chose Warren, Harris, and Buttigeg as their second choices. Additionally, large proportions of all candidates’ supporters feel that Biden is the likeliest to beat Trump: 50 percent of Buttigieg supporters, 28 percent of Warren supporters, and 20 percent of Sanders supporters believe this. 

What it all means is that a substantial number of primary voters are making decisions about whom to vote or caucus for next year through the prism of other people’s presumed biases. And that may be creating unnecessary distortions, because there’s little reason to suspect voters are equipped to gauge these biases or the nature of electability in general. It may also be creating a self-perpetuating feedback cycle, where voters tell pollsters they support the candidates that political commentators have deemed electable based largely on their strengths in the polls.

The question of electability has frustrated many Democratic activists—particularly those whose favorite candidates have suffered from negative perceptions of their electability. It is common for these activists to assert that the concept electability is essentially fiction. Most polls show Trump losing head-to-head matchups with all leading Democrats, so people should just support the candidates they like best. 

There is a compelling logic to this argument, but it is also too convenient.

If we could run the election two dozen times, one candidate would almost certainly outperform all the others—even if by slim margins. Whoever the most electable candidate is, that person would stand the best chance of beating Trump and would be likeliest to beat him by the largest margin. If Democrats can manage to nominate the most electable candidate, it might make the difference between Democrats winning or losing the election. It might even determine whether the next Democratic president will have to govern with Democratic or Republican Senate. These are profound stakes: Will America give Trump a second term? Will America be able to govern itself?

That means two things are true: An enormous amount hangs on Democratic voters getting the electability question right, and yet most voters are shooting in the dark for the right answer. 

The outcome of the 2016 election looms large in voters’ minds as they select their candidates today. In open-ended responses, voters repeatedly explained that they’d chosen male candidates over female alternatives because of their view that male candidates are more electable—that the country may be too sexist to elect a woman president. The 2018 elections apparently did little to allay the concern that women candidates won’t get a fair shake in 2020.

What do our survey results say about this concern? Admittedly not much directly, but there are some interesting clues.

1. None of the leading candidates has obvious vulnerabilities that the others lack. Every candidate has haters, which means some respondents said they’re likely to sit out the election or vote third party if their most disfavored Democrats win the nomination. But for Biden, Sanders, Warren, and Buttigieg (and indeed for most candidates) this number is remarkably stable. For Biden it’s 14, Sanders 16, Warren 15, and Buttigieg 16.

2. Most, but not all, leading candidates draw their support very unevenly within the Democratic coalition. Biden struggles with younger voters, but not older ones. Sanders struggles with older voters but not younger ones. Buttigieg struggles with black voters. Warren, by contrast draws her support fairly evenly across demographic groups.

3. Nearly all candidates underperform with black voters relative to their overall levels of support because Biden completely crushes all other candidates among black voters with 62 percent. This is both a source of immense strength for Biden, and also a point of vulnerability, should his support among black voters begin to slip for any reason.

4. As voters become more aware of the candidates in the race, their perceptions begin to change in ways that benefit Warren and Buttigieg and hurt Sanders and Biden. Among the most-aware voters, Warren leads with 25 percent, followed by Biden with 23, Buttigieg with 16, and Sanders with 14. Among all respondents, half see Biden as the most electable. But this overall perception of his electability correlates negatively with candidate awareness—as voters learn more about the field, they become less likely to see Biden as the most electable. This is also true of Sanders. By contrast, the more voters learn about the field, the more they come to view Warren and Buttigieg as electable.

This isn’t enough to say that the conventional wisdom about electability is wrong, but it is enough to say that we should all question our assumptions. And now it’s up to the field to convince the Democratic voters of Iowa, New Hampshire, and South Carolina that their perceptions are deceiving them.