Skip to main content
ABC News
Which 2020 Democrats Already Have A Fan Base — And Which Don’t

Early horse-race polls of the presidential primary are sometimes accurate, sometimes not. So at this point in the election calendar, looking at how voters feel about each candidate can be more helpful. Measures like favorable and unfavorable ratings can reveal which candidates are polling poorly because they’re not popular and which are simply not well-known. This is important because most eventual nominees are one of two things early in the campaign: (1) household names who are already well-liked by members of their party or (2) relative unknowns.1

That’s also true of most major2 candidates (and potential candidates) in the 2020 Democratic field — but there is enough variation to be interesting, plus one notable outlier. Below, I’ve plotted the share of Democrats who can form an opinion of each major candidate (favorable rating plus unfavorable rating) against the candidate’s net favorability rating (favorable rating minus unfavorable rating) among Democrats,3 according to an average of national polls taken between Jan. 1 and Feb. 5 of this year.

On the high-recognition end of the spectrum (the right side of the chart above), we have Joe Biden, Bernie Sanders and Elizabeth Warren. Among the three, however, Biden has the highest net favorability rating (+69 percentage points), which is higher than one would expect. By contrast, Sanders’s net favorability rating is a bit lower than what one would expect, at +52 points. Meanwhile, Warren’s +44 net favorability rating is a bit better than one would expect.

“What one would expect” is based on the overall trendline that’s plotted on the chart above, which is our best guess at what the relationship is between the share of Democrats with an opinion of the candidate and her net favorability rating among Democrats. The line projects what the net favorability rating “should” be for each candidate based on how well-known he or she is. Here’s a look at how the candidates’ actual net favorability ratings compare with those projections (in other words, the distance between their head and the line in the chart above):

Biden and Bloomberg break the favorability pattern

2020 Democratic candidates’ actual and projected net favorability ratings among Democratic respondents, based on the share of Democrats with an opinion of the candidate, according to an average of national polls conducted Jan. 1 through Feb. 5

Net Favorability
Candidate Dems With Opinion Actual Projected Diff
Joe Biden 89% +69 +57 12
Kamala Harris 58 +38 +33 5
Beto O’Rourke 50 +32 +26 5
Julian Castro 41 +23 +19 4
Pete Buttigieg 20 +6 +3 3
Amy Klobuchar 32 +15 +12 2
Terry McAuliffe 27 +10 +8 2
Sherrod Brown 31 +13 +11 2
Elizabeth Warren 71 +44 +43 2
Cory Booker 56 +33 +31 1
Jay Inslee 19 +3 +2 1
John Delaney 22 +5 +4 1
John Hickenlooper 24 +6 +6 0
Kirsten Gillibrand 46 +21 +23 -2
Bernie Sanders 88 +52 +56 -3
Eric Holder 49 +21 +25 -4
Tulsi Gabbard 31 +4 +12 -8
Michael Bloomberg 62 +11 +35 -24

Source: Polls

As you can see, Biden is unusually beloved, even for a politician as well-known as he is. The 12-point difference between his actual net favorability rating and his projected net favorability rating places him at the top of the table. But the biggest outlier is Michael Bloomberg, who says he is still deciding whether to get into the race. With 62 percent of Democrats able to form an opinion of him, Bloomberg is one of the better-known contenders. But he has a net favorability rating of only +11 (24 points less than expected). Why is Bloomberg so relatively unpopular? Maybe Democrats are aware of his moderate views on economic issues; for example, he compared Warren’s proposal for a “wealth tax” with “non-capitalistic” Venezuela. Or, maybe, in the era of President Trump, Democrats are turned off by the idea of nominating a businessman for president.

Either way, it’s a problem for Bloomberg, should he decide to run. According to FiveThirtyEight’s research, only one high-recognition but low-favorability candidate early in the campaign has overcome that handicap to win a nomination since at least 1980: Trump.4 The president may not get enough credit for his extraordinary comeback in the primary campaign5 — he turned a -42 net favorability rating among Republicans (!) in a May 2015 poll into a +28 net favorability rating that September. Bloomberg isn’t starting from such a low place, but if he wants to improve his popularity among primary voters, he may have a harder time than Trump did. Although he has the advantage of wealth (like Trump), Bloomberg is unlikely to get the massive quantity of free-media coverage that Trump did, and he doesn’t appear to scratch an unsatisfied itch for Democrats the way Trump has for the GOP.

As Trump showed, net favorability isn’t the be-all and end-all of primary prognostication. But it has proved instructive in other years. As always, we continue to entertain many hypotheses about how the presidential primary works.

Derek Shan contributed research.

FiveThirtyEight’s 2020 draft: Episode 2


  1. As opposed to (3) household names who are not particularly well-liked. There’s one big (and recent) exception to this, but we’ll come to that later.

  2. Granted, this is my subjective read on the field. The analysis below excludes candidates without prior experience getting elected to office, as well as candidates who are considered unlikely to run. I was also forced to exclude some major candidates (e.g., Oregon Sen. Jeff Merkley) because there wasn’t data on them.

  3. Including voters who lean toward Democrats, where provided.

  4. A similar study we did found that, among candidates who trail in early horse-race polls, well-known candidates have worse primary track records than poorly known candidates.

  5. He probably gets too much credit, however, as a political tactician for his general election win.

Nathaniel Rakich is a senior editor and senior elections analyst at FiveThirtyEight.