In summary

If California had used ranked-choice voting on Super Tuesday, the votes would have been counted a lot sooner and we would have a different winner.

By L. Sandy Maisel and David W. Brady, Special to CalMatters

L. Sandy Maisel is the Goldfarb Family Distinguished Professor American Government at Colby College; he is currently a visiting scholar at the Hoover Institution at Stanford University, David W. Brady is the Davies Family Senior Fellow at the Hoover Institution, They wrote this commentary for CalMatters.

If California voters had been allowed to rank multiple candidates on their ballot in order of their preference, as one state does, the real winner of the Super Tuesday primary held nearly two months would have been announced much sooner. And the preference of a majority of voters might surprise you. 

More than a third of the voters who opted for mail-in balloting chose candidates who dropped out before Super Tuesday, so their votes really did not count. But, under a system of “ranked-choice voting,” their second and third preferences would have been tallied in the final count. 

Our analysis shows that if votes were reallocated according to voter preference between the two remaining viable candidates, former Vice President Joe Biden would have prevailed over Sen. Bernie Sanders.

We all can agree that any election where ballots are cast early for candidates who drop out before Election Day — and the results are unknown for more than a month after voting ends — needs rethinking.  

Two years ago, voters in Maine opted to use ranked-choice voting in most of their elections. They have continued to use it in primary elections and for general elections for the U.S. House and Senate. Exit polls show that most voters prefer it over the plurality voting system they used in the past. 

In this type of election system, voters rank candidates in a multi-candidate field in order of their preference.  In a typical ranked-choice voting election, after the votes are cast, the candidate with fewest votes is eliminated and that candidate’s supporters’ votes are transferred to the voters’ second choice candidate.  The process is repeated until one candidate has a majority.

A ranked-choice voting system could be adapted to a primary election, in a multi-candidate field, in which low polling candidates are eliminated until all remaining candidates have at least 15% of the vote, the threshold stipulated by the Democratic Party for a candidate to receive any delegates.  

If a ranked-choice voting system were in use in California, or other states with significant early voting or mail-in voting such as Washington state which uses all-mail ballots, the problem of voter disenfranchisement — because a candidate drops out between the time a ballot is cast and when it is counted — would be eliminated.  We also would have a clearer sense of voter preference when all ballots are counted.

While we can never know for sure, because candidates would campaign differently under different rules and because voters might make different decisions, we do know that supporters of candidates who dropped out could still have their votes counted for an active candidate.  And we can draw some interesting conclusions from recently conducted polls.

After the Nevada caucuses and the South Carolina debate but before Biden’s victory in South Carolina, YouGov conducted a poll of 1,507 registered California voters for Stanford University’s Hoover Institution and the Bill Lane Center for the American West.  The poll has a sampling error of +/- 3.1 percentage points.

On that survey, we asked respondents how they would have ranked candidates had they been given that opportunity.  In analyzing the results, we accepted the Democratic Party rules holding that only candidates receiving at least 15% of the vote would receive any delegates.  Thus, we eliminated the candidate with the fewest votes, reallocating that candidate’s voters to their second choice candidate, and continued to do so (allocating to the highest remaining candidate) until all remaining candidates had at least 15% of the vote.

Sanders led Biden by approximately 7% of the votes tabulated in California in the days after the Super Tuesday vote.  Only those two candidates polled 15% statewide, thus they were the only candidates who received an allotment of the delegates awarded statewide.  Former New York City Mayor Michael Bloomberg and Sen. Elizabeth Warren were each at about 13%; each would receive some delegates at the district level.   None of the other candidates polled even 10%.

However, that is not the message that reflects California voters’ real preferences.  If Californians had ranked candidates, Sanders would still have led, but his margin over Biden would have been less than half of the total reported.   Warren and Bloomberg would have surpassed the 15% threshold.  What is clear is that the supporters of the candidates who dropped out before Super Tuesday — former Mayor Pete Buttigieg, Sen. Amy Klobuchar and California businessman Tom Steyer — would go more toward Biden and Bloomberg than toward Sanders and Warren.

Our analysis allowed the votes for Bloomberg and Warren to stand, as they were still active candidates when votes were cast.  If their votes are reallocated according to voter preference between the top two candidates, Biden would have won over Sanders.

We do not have similar polling data for Washington, whose votes were cast the week after California’s;  but the logic holds that the close race between Biden and Sanders — Biden won by 1.4% — would have swung even further toward Biden. Washington exit polls showed that of voters who decided in March, Biden had a 55-19 margin while voters who decided before then preferred Sanders by 40-21. 

Moreover, in our California poll and our national polls which allowed voters to rank candidates, voters whose first choice was Bloomberg, Buttigieg or Klobuchar most often selected Biden as their second choice; and these voters outnumbered those whose first choice was Warren, most of whom selected Sanders as their second preference.

The California and Washington results would have been different under ranked-choice voting.  In California, the ranked system would have given Biden the win and in Washington the likely result would have been a bigger Biden win.  

Interpreting voter true preferences when many candidates are running is more complex than simply declaring whoever gains a plurality of the votes as the winner.  Just as important, ranked-choice voting, if used in states with mail-in ballots or early voting, would let all voters’ preferences be reflected in the final tallies, even if a voter’s first choice preference had withdrawn between the time of voting and of counting ballots.  


L. Sandy Maisel is the Goldfarb Family Distinguished Professor American Government at Colby College; he is currently a visiting scholar at the Hoover Institution at Stanford University, David W. Brady is the Davies Family Senior Fellow at the Hoover Institution, They wrote this commentary for CalMatters.

We want to hear from you

Want to submit a guest commentary or reaction to an article we wrote? You can find our submission guidelines here. Please contact CalMatters with any commentary questions: