Commentary & Analysis

Two-Thirds of Democratic Voters Support Multiple Candidates—But Current Polls Won’t Tell You That

Compare the packed auditoriums full of energetic fans that you see on TV supporting candidates like Elizabeth Warren or Amy Klobuchar to their performance in the primaries and you might be left scratching your head. 

How can it seem like so many people support these candidates, yet they perform so poorly on Election Day?

According to a recent nationwide poll conducted by The Center for Election Science and Change Research, it’s thanks to the large slate of Democratic candidates and our terrible choose-one voting method. 

The poll of 821 likely 2020 Democratic primary voters shows that going into Super Tuesday, 67% of Democratic voters would be happy with multiple candidates. But as they head to the polls this week, voters will only be allowed to cast a vote for one candidate.

The result? Candidates are splitting the vote, making it appear that they have less support from Democratic voters than they actually do.

When voters were restricted to choosing only one candidate, they were conflicted. While a plurality chose Sanders (40%), the rest of the electorate is split between seven other candidates. This makes it appear that those candidates aren’t well-supported.

But, when voters were allowed to use approval voting, we see that true levels of support for each candidate are much higher. In approval voting, people are allowed to vote for all the candidates they support, and the candidate with the most votes wins.

While Sanders still leads the pack under approval voting, he does so with much more support, with 60% of voters approving of him. All other candidates saw a significant bump in support. Those candidates more than doubled their support compared to the more restrictive, choose-one poll. 

Contrast this with another popular voting reform, ranked-choice voting (RCV). Although it chose the same winner as choose-one and approval voting, RCV did not show an accurate level of support for each candidate. That’s because RCV produces a winner by transferring votes from candidates as they are eliminated until one candidate has more than half of the remaining, active ballots.  

It’s not just candidates who disappear. Issues can be overlooked by choose-one voting as well.

When respondents were asked to select only the single most important issue to them when voting, issues like education and income inequality appeared to be only minor concerns.

But Democratic voters actually have great consensus on the issues. When asked to select all the issues that matter to them when voting, 93% of respondents said that healthcare was a major issue for them. Other issues such as climate change (89%), education (79%), and income inequality (74%) were also top priorities.

The poll’s results, ironically, highlight the need for another issue that was important to Democratic voters—election reform. In nearly all elections across the United States, voters are limited to choosing only one candidate—a method that political scientists and mathematicians have said is the worst voting method there is. 

With choose-one voting, candidates can serve as “spoilers” for more mainstream candidates, candidates with similar ideologies can split the vote, and third-parties and independents have a hard time breaking through.

However, there are fairer, more representative ways of voting and in recent years, the movement for this type of reform has gained momentum. In 2016, the state of Maine became the first state to adopt ranked-choice voting and in 2018, Fargo, North Dakota became the first city to adopt approval voting with St. Louis, MO looking to do the same in 2020. 

A correction has been made in the article as Pete Buttigieg and Joe Biden’s plurality (choose-one) vote were accidentally switched.


The Center for Election Science is a nonpartisan nonprofit dedicated to studying and advocating for better voting methods to create fairer, more representative elections. In 2018, we helped Fargo, ND become the first city in the US to enact a method called approval voting.

Methodology – Polling was conducted online from February 25-27, 2020. Using its Dynamic Online Sampling technology to attain a sample reflective of likely Democratic primary voters, Change Research polled 821 nationally. All voters say they identify as a Democrat or Independent and have a 50-50 likelihood or greater of participating in their state’s primary or caucus. Post-stratification weights were made on age, gender, region, race, county density, ideology, and 2016 primary vote.