Support The Bulwark and subscribe today.
  Join Now

Want to Stop Extremists from Becoming President? Try Primary Runoffs

It’s a way to ensure that winners receive majority support—and that fringe candidates won't cruise to victory on pluralities alone.
March 9, 2023
Want to Stop Extremists from Becoming President? Try Primary Runoffs
A voter fills out a ballot at the Ames Public Library on primary Election Day on June 7, 2022 in Ames, Iowa. (Photo by Stephen Maturen/Getty Images)

Donald Trump was the first president to come to office with no prior experience in government or the military. He had never served as a state governor or lawmaker, as a member of the U.S. House of Representatives or Senate, as vice president, or as a decorated military leader—the training grounds for previous presidents. A central reason why Trump was able to bypass these political apprenticeships, which provide valuable experience in learning how to govern in complex settings, was the lack of runoffs in presidential primaries. The method of primary voting used in the United States allowed him to capture the GOP nomination with only a plurality of the vote—even though a majority of GOP primary voters supported other candidates.

Prior to the introduction of nomination primaries by the Progressive movement in the early 1900s, political bosses typically handpicked their favorite candidates—often toadies—in now-legendary smoke-filled (and bourbon-scented) back rooms at party headquarters. This approach changed dramatically with the use of primaries as a way of opening up leadership selection to rank-and-file voters.

Under the current version of the presidential primary, Democrats select most of the delegates for their national party conventions by a form of proportional representation based on the state primaries. By contrast, Republican primary elections in most states follow the winner-take-all rule, wherein a candidate who wins a plurality of votes in a crowded field wins most or all of the convention delegates at stake in that primary. It is this system that made it possible for Trump in 2016 to ride a series of plurality wins in multiple primaries to corral the GOP nomination.

Let’s revisit what happened during that cycle. Seventeen Republican candidates sought the party nomination, including several prominent officeholders. Among the competitors was the heir apparent of the Bush dynasty, Jeb Bush, a former governor of Florida. Other notables included sitting Senators Ted Cruz, Lindsey Graham, Rand Paul, and Marco Rubio, and former Senator Rick Santorum; the former speaker of the House, Newt Gingrich; John Kasich, the sitting governor of Ohio, and Chris Christie, a former governor of New Jersey.

Against this crowded field, Trump consistently won primary contests—yet his victories came overwhelmingly by way of pluralities, not majorities. In the four opening contests, Trump attracted only about a third of the vote but got 62 percent of the delegates. In South Carolina, for example, Trump gathered 32.5 percent of the GOP vote in a crowded field but was awarded all fifty state delegates to the national Republican presidential nominating convention. Senators Rubio and Cruz each received about 22 percent of the vote.

Subsequently, on Super Tuesday, Trump won seven of eleven states—all with pluralities. Indeed, he exceeded 40 percent of the vote in only two states: Alabama and Massachusetts. Nonetheless, he walked away with 43 percent of the delegates awarded that day. He went on to gain additional plurality margins of victory in subsequent primaries. Not until April 19 did Trump achieve an outright majority win, polling 59 percent in his native state of New York.


What would have happened if, instead of using the winner-take-all system, a top-two runoff system had been in place in 2016? Under a runoff rule, if no candidate receives more than 50 percent of the vote on election day, the top two contenders will face off in a second election held, say, one week later. This guarantees majority support for the winner.

Had Republicans used such a system in 2016, voters who preferred candidates other than Trump could have turned in the runoff toward whichever non-Trump candidate had survived the first round. So, in South Carolina, the runoff would presumably have been between Trump and Rubio (who slightly edged out Cruz). Most likely, many of the voters who had turned to Cruz and the other candidates would have gone for Rubio in the runoff, and Trump would have been defeated.

And the same pattern would likely have been seen in the other states. In all likelihood, one of Trump’s opponents with experience in high office might well have emerged on top had runoffs been in place. Instead, with no majority requirement, Trump managed to sweep the bulk of the Republican convention delegates, relying on plurality support. Even though Trump had been rejected by most Republican primary voters, he had managed to secure the GOP nomination.

While some states and hundreds of cities use runoff primaries, this method has never been part of the selection process for the nation’s highest office. As a result, the United States remains without an important guardrail to prevent amateurs, especially ones who have limited support, from slipping into positions of responsibility beyond their capacities. Adoption of runoff primaries in the 2024 elections would offer a means of voting to help ensure this nation’s future chief executive had attracted broad public support.

While most states operate under a plurality rule in statewide elections, hundreds of cities use runoffs to determine winners when no candidate polls a majority in the initial vote. Experience has shown, though, that the runoff procedure is not without its drawbacks. One problem with relying on runoffs in the presidential nominating primaries is that the campaigns of the two leading candidates would have to ramp up activity in a state a second time, even as the electoral calendar inexorably pushed the selection process toward the next state contests. Candidates who had to keep an operation going in one or more states in which they were competing in a runoff—while also mounting efforts in new states—would face burdens not borne by candidates who had failed to make the cut in earlier states. Another downside is the added costs to a state in holding a second, conclusive primary election a week or so after the first one narrowed the field to just two candidates.

Yet the two-tiered runoff procedure has attractive features as well, including an important potential advantage. A second balloting might well prompt candidates to broaden their appeal by adopting broader stands on policy issues. This outcome would enhance the chances of selecting a candidate who enjoys wide support, instead of someone who limps to the finish line with backing from a small but dedicated portion of the electorate. This broadening effect is apt to yield mainstream office-seekers able to govern with a less extreme view of the general public’s needs and hopes. The time between the first and second vote would also allow voters an opportunity to digest new information free of the cacophony of the first round—a chance to take an uncluttered second look.


While the runoff system we described above involves a second vote, an alternative, one-tiered “instant runoff” procedure (also known as “ranked-choice voting,” or RCV) would avoid the necessity of voters making a return trip to the polls. It would also free campaigns from sanity-threatening efforts to continue competing in one or more states while simultaneously going full-bore in new states popping up on the calendar—not to mention the fact that eschewing a second round of balloting would save states and candidates millions of dollars.

A sizable challenge to using the instant runoff method in presidential primaries, though, would be the need to educate the electorate on this unfamiliar format. While a small but growing number of cities (including San Francisco, New York, and Minneapolis), as well as the states of Maine and Alaska, have implemented forms of instant runoffs, most Americans have no familiarity with ranking preferences on a ballot.

Here’s how instant runoffs work in RCV: Voters may place a “1” next to the name of their first choice, then “2” for their second, and down the list according to the total number candidates, or up to some specified maximum number. If one of the candidates is the first choice of half or more of the voters, he or she automatically wins the contest. When no one in the field is able to garner a majority of the vote, however, the candidate who comes in last is eliminated. The second-choice ballots for the eliminated candidate are then redistributed among the survivors, with this process reiterating until someone in the field registers more than half of the vote. Rather than a second casting of ballots, simple math is used to sort out the range of preferences on the original ballots in deciding the winner. Of course, as in the past, voters could, if they wish, continue to vote just for a single choice on the initial list of candidates.

Would American voters balk at the relative complexity of this method of redistributing ballots to eliminate the weaker candidates? Would this approach further fuel the misgivings many citizens have about election procedures and the legitimacy of results? In contrast to RCV, the method of a two-tiered runoff contest held a week apart are easy to understand and rest on the widely accepted tradition of a majority rule in American politics.

Naturally, individual state parties could decide which procedure to adopt: ranked choice voting or the runoff primary. On balance, the relative complexity of RCV when joined with a majority support threshold—both of which would be new for many voters—may make a simple runoff rule more acceptable. A sizable component of the American electorate already questions the legitimacy of election procedures. The novelty of RCV could baffle some voters and further fuel suspicions about election fairness and reliability, leading to a decline in electoral participation.

Still, either system—a two-tiered runoff primary or an instant runoff—could help guard against what happened in the 2016 presidential sweepstakes.

It’s too early to tell whether an embrace of either the runoff primary method or the ranked-order voting procedure would make a difference in 2024. The GOP field could again see a host of would-be presidents: Former South Carolina Governor Nikki Haley has already announced she is running, and additional possible contenders include (in alphabetical order) Ron DeSantis, Kristi Noem, Mike Pence, Mike Pompeo, Tim Scott, Chris Sununu, and Glenn Youngkin. But this time around, Trump is not an outsider, but a former president with enormous influence on the party and wide support in base. It’s too early for meaningful polling, but it is not hard to imagine that he could command majority victories in the primaries he only won by pluralities in 2016.

Still, whether relevant in this cycle or not, a runoff primary system would in the long term serve the country well. Successful candidates should have to demonstrate their ability to appeal across a broad spectrum of citizens in their party. A competition for majority support—as more clearly set up by the runoff procedure than in ranked-choice voting—would be likely to tamp down extremism in the GOP nomination contests for the nation’s highest office.

Charles S. Bullock III and Loch K. Johnson

Charles S. Bullock III is the Richard B. Russell Professor and Distinguished University Professor of Public and International Affairs at the University of Georgia. Loch K. Johnson is Regents Professor of Public and International Affairs Emeritus at the University of Georgia. They are co-authors of Runoff Elections in the United States (UNC Press, 1992), winner of the V.O. Key Prize from the Southern Political Science Association.