Skip to main content
ABC News
Most Democrats Have Picked A Candidate, But They Could Still Change Their Minds

There are now fewer than 11 weeks left until the Iowa caucuses, and if you were to just look at the top line numbers in the polls, it would seem like most Democratic voters have picked a candidate to vote for. But the low number of undecided voters in many polls obscures the fact that plenty of Democrats still aren’t completely set on their choices. A national CNN poll from mid-October found that only 6 percent of Democratic or Democratic-leaning voters had no opinion about who should be the party’s nominee, but 53 percent of those who currently have a preference for president might change their mind about which candidate they support. And in September, Quinnipiac University found that only 13 percent didn’t know whom they were supporting, but 63 percent of those who did know might change their mind.

On one hand, those numbers seem high — and at the very least, they do demonstrate that the race is still in flux, which may help explain why some last-minute entries felt emboldened to run (or at least consider running). After all, former Vice President Joe Biden could cough up his lead to Sen. Elizabeth Warren, or a dark-horse candidate like South Bend, Indiana, Mayor Pete Buttigieg could surge further and attract more voters to his side. But on the other hand, these numbers aren’t actually that different from past election cycles. Voters consistently seem to be at least somewhat undecided at this point in the primary calendar. Understanding what can actually cause voters to change their minds is difficult, however, as some of these primaries were drastically shuffled around before voting season, while others largely remained the same despite voters being open to other alternatives.

For example, at this point in the 2004 Democratic primary, former Vermont Gov. Howard Dean led in the polls, but that support would shortly leave him for then-Sen. John Kerry, the eventual nominee. And indeed, polls around that time indicated a pretty unsettled race. According to Gallup/CNN/USA Today, 68 percent of Democrats or Democratic leaners who had a favorite candidate reported that they might still change their mind. And an ABC News/Washington Post poll found that 76 percent said there was at least some chance they would change their mind (53 percent said there was a “good chance” they’d change their mind, and 23 percent said there was a chance they’d change their mind but it was “unlikely”).

In the 2008 Democratic primary, then-Sen. Barack Obama came from behind to defeat an initial polling front-runner, then-Sen. Hillary Clinton, but voters didn’t report being as nearly unsettled as they were in 2004. In fact, the environment looked a lot like today’s, with roughly the same number of primary voters saying they’d consider changing whom they were planning to support. In early November 2007, per CNN, 60 percent of registered Democrats who were backing a candidate said it was possible they would change their mind — 7 points higher than CNN found this year. A Quinnipiac poll from late October 2007 put the number at 52 percent, 11 points lower than the pollster found in September 2019, though the 2007 and 2019 questions were asked differently.1 An Associated Press/Yahoo poll also pegged the number of voters who might change their mind at 52 percent. So while fewer voters expressed a willingness to change their mind than had said the same in 2004, the race obviously still shifted quite a bit, hinting that a come-from-behind win is a possibility for 2020 as well.

On the other hand, the 2016 Republican primary was also pretty similar to 2020 in terms of the number of voters who said they might change their minds — and that race shifted far less. A Quinnipiac poll from late October and early November 2015 found that 63 percent of Republican or Republican-leaning voters nationwide who had a candidate preference thought they might change their mind before the primary (as you’ll recall, that’s the same share as Quinnipiac’s most recent poll found for 2020). At this point in that primary, Donald Trump was fighting with Ben Carson for front-runner status, but Trump moved solidly into the lead not long after. This provides a second blueprint for how 2020 might go: The current Biden vs. Warren battle for front-runner status might resolve itself sooner rather than later if none of the lower-polling candidates leapfrog to the fore.

So while these polls can’t tell us just how competitive the primary will be or how much we can expect voters’ preferences to change, they can tell us one thing: 2020 doesn’t look like it will be as steady as the 2016 Democratic primary. That year, of course, Clinton was the dominant early front-runner, and while Sen. Bernie Sanders put up a good fight, Clinton remained the favorite right up until the point she clinched the nomination. Probably not by coincidence, this is the primary when voters appeared most set in their candidate choices. The aforementioned fall 2015 Quinnipiac poll found that only 44 percent of Democratic primary voters with a candidate preference were open to changing their mind. And only 50 percent told CBS News/The New York Times in November 2015 that it was “too early to say for sure” if they were set on their candidate.

But because the number of voters who might change their mind in 2020 is neither historically high nor historically low, it’s hard to know how fluid the race truly is. And in a race that has been pretty steady so far (the Warren upslope aside, candidates with good early polling still have good polling and candidates with bad early polling still have bad polling), perhaps the biggest takeaway is that things could be more fluid than we might think.

Derek Shan contributed research.


  1. The 2007 poll asked people how likely it was that respondents would change their mind and allowed them to grade the odds on a four-point scale from “not likely at all” to “very likely.” The 2019 poll asked voters if their mind was made up or if they might change it before the primary.

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