A Top-3 Presidential Election System

This way of voting would enable Americans to elect a third-party or independent candidate whom a majority prefer to either of the two major-party nominees.

Joe Manchin was correct to abandon his flirtation with a third-party presidential bid. As he properly recognized, the only role that a third-party candidate can play in the nation’s current electoral system is that of “spoiler” (siphoning away votes from one of the two major-party nominees who loses with the third-party candidate in the race but wins if that third candidate doesn’t run).

To his great credit, Manchin put an end to all the speculation about whether he would offer himself as a more centrist alternative to Trump and Biden by firmly declaring on Friday that he would not be a “spoiler” and thus “will not be seeking a third-party run.”

Manchin, however, is not wrong that the nation needs a more moderate third party in between the Republicans, now dominated by Trump and his MAGA brand of authoritarian populism, and Democrats, who remain a coalition of liberals and progressives.

Americans are hungering for a third major party to compete effectively against the increasingly polarized—even tribal—Republicans and Democrats. Gallup’s most recent poll on the topic, in October, showed 63% saying “a major third party is needed”—the highest percentage Gallup has ever received on this question.

The problem is the existing electoral system. The only way to have the kind of competitive third party that almost two-thirds of Americans want is to replace the existing two-party system with a new three-party system.

Like Manchin, Nikki Haley is learning this the hard way. As I’ve observed previously, she is effectively trying to run a third-party campaign in a two-party system. But it won’t work. She can’t win the Republican nomination because of Trump’s domination of the party after its MAGA makeover. And if she were to run as an independent candidate against both Trump and Biden in November, she would play the same spoiler role that Manchin rejected.

What’s so exasperating about the current situation is that either Manchin or Haley—or indeed any credible candidate from the middle of American politics—would beat either Trump or Biden if given the opportunity to run against each of them one-on-one. Neither Trump nor Biden have the support of a majority of American voters, and because of the tribalism of the country’s current hyperpolarized two-party conflict, each is despised by partisans on the opposite side of the divide.

If Manchin or Haley faced only Trump on the ballot in November, Manchin or Haley would easily win a majority, combining all the anti-Trump votes of Democrats, independents, and remaining non-MAGA Republicans. Conversely, if Manchin or Haley faced only Biden in November, either would easily win that one-on-one contest as well, combining votes of all Republicans and independents disaffected with Biden’s performance as president and/or concerned about his capacity to serve a second term.

There is an electoral system that would permit Machin or Haley as a third-party candidate to run against Trump and Biden separately, one-on-one, rather than having to run against both Trump and Biden together. It’s the “top-3” system that Eric Maskin and I described in our recent Project Syndicate column and which I elaborated upon in a previous Common Ground Democracy essay.

Both of these pieces explained the concept and mechanics of this top-3 system in the context of congressional, not presidential, elections. The same pressing need for a third party, because of the hyperpolarized tribal rivalry of the existing two major parties, exists with respect to congressional as well as presidential elections. Because the Republican party has become dominated by Trump and his MAGA movement, in many Senate and House races there is a missing moderate in between the MAGA candidate and the Democrat on the November ballot—a non-MAGA Republican who couldn’t win the GOP primary but whom November’s voters would prefer to either of their two alternatives. As detailed in my previous Common Ground Democracy essay, we see the same phenomenon over and over in the 2022 midterms, and it’s a pattern that’s repeating itself in this year’s congressional races.

The same top-3 system that would solve the problem of polarization and dysfunction in Congress would also work in presidential elections to enable a majority of the nation’s voters choose a third-party or independent candidate whom they prefer over each of the two major-party nominees. The way it would work is that states would hold a round of voting in September, after the parties hold their nominating conventions, and the September ballot would include each party’s nominee along with any additional third-party or independent candidates who qualified for the ballot by collecting enough signatures nationwide. Manchin or Haley, or other would-be moderate alternatives, could qualify for the September ballot either as independents or as the nominee of a new third party.

The September ballot would ask voters to identify the single candidate among those listed whom voters would like to advance to the November general election. The three candidates receiving the most votes would be the ones to advance. Presumably, Trump and Biden as the major-party nominees would be two of the three candidates to advance. So too would either Manchin or Haley—let’s assume Haley for sake of illustration, since she’s still running whereas Manchin isn’t—if she consolidated enough support to beat out other qualifying independent and minor-party candidates (like Robert Kennedy, Jr., Jill Stein, and Cornel West).

Then, in November, voters would cast ballots expressing their choices between each pair of the three candidates who received the most votes in September. If those three were indeed Trump, Biden, and Haley, then voters would express their preferences between (1) Trump and Biden, (2) Trump and Haley, and (3) Biden and Haley. Whichever candidate beat each opponent in these head-to-head matchups would be the election’s winner. For reasons already discussed, it almost certainly would be Haley.

This top-3 system would be a better way to elect presidents not just this year but always. It would incentivize all candidates—Republican, Democrat, third party, independent—to appeal to enough voters to prevail over each opponent. Sometimes the Democrat would be able to do that, sometimes the Republican, and sometimes—if the two major parties strayed too far from the middle or otherwise nominated unappealing candidates—a third alternative.

As explained in the two previous pieces, rarely but occasionally a three-way tie might occur in which each candidate beats one opponent but loses to the other. (If this system were in place for this year’s election, that situation almost certainly would not occur because of Haley’s capacity to build broad enough support to prevail over each opponent.) The simplest tiebreaker—and one most philosophically aligned with the goal of electing the candidate with the broadest support among voters—would be to elect the candidate whose single loss has the smallest margin of defeat.

This top-3 system is easy to understand and administer. Voters can quickly comprehend the returns as they come in on Election Night. Unlike with Ranked Choice Voting, there is no complicated formula to determine the election’s winner. All a voter needs to know—and all the media needs to report—are the three head-to-head tallies between each pair of candidates. Whoever is ahead against each opponent, or whoever is behind by the narrowest margin if no one is ahead against both opponents, is the leader in the race as the ballots keep getting counted.

Is it a drawback of this system that it requires voters to cast an extra ballot in September, as well as the final November ballot—not to mention the ballot that the voter must cast to participate in one of the party’s pre-convention primary elections? Well, compared to the Survivor-style presidential election system that I described before, this top-3 system is much simpler and less demanding of voters.

Moreover, it is increasingly apparent that presidential primaries are not all that they are cracked up to be. This year, the Republican party’s primaries look like they will be effectively over after South Carolina and before Super Tuesday, if it’s not effectively over already. And the Democratic party never really had a pretense of a competitive primary: Biden wasn’t even on the ballot in New Hampshire, and he still won in a landslide as a write-in candidate. In any event, there is increasing talk about the Democrats needing to pick a new nominee at the party’s convention regardless of the votes in the primaries. Whether or not that happens, it would be reasonable for voters to think that the September ballot described above is a lot more meaningful as an exercise of democracy than voting in a partisan primary.

So, if a top-3 presidential election system is a good idea, how do we get it adopted? One possibility is a constitutional amendment, but that’s too difficult to achieve.

A better way to put a top-3 presidential election system in place is through a modified form of the National Popular Vote Interstate Compact (NPVIC) plan. NPVIC is an agreement among states to award their electoral votes to the winner of the national popular vote (rather than to the winner of the popular vote in the particular state). NPVIC doesn’t take effect until enough states sign on to control 270 electoral votes, the number constitutionally needed to win the presidential election. Right now, sixteen states (plus DC) have joined NPVIC, but together they have only 205 electoral votes—65 short of the necessary number. But if and when NPCIV reaches 270, then the compacting states can control the outcome of the presidential election regardless of how other states choose to appoint their presidential electors.

NPVIC, in its current form, does not create a top-3 presidential election. On the contrary, unfortunately, it replicates the basic problem of the existing two-party system. By awarding its electoral votes to the pluralitywinner of the national popular vote, NPVIC (if it were in effect this year) would elect either Trump or Biden, whichever one got more votes even if that number was below 50% and, most importantly, even though Haley (or Manchin) would beat either Trump or Biden nationwide one-on-one. Thus, in its current form, NPVIC fails to elect the candidate whom a majority of voters would most prefer among those running for the office.

The NPVIC concept, however, can be modified so that the states agree to award their electoral votes to whichever candidate wins according to the top-3 system as defined above. In other words, all the states joining the agreement would hold the September ballot among qualifying candidates, as described. Then, in November, all these states would provide their voters with a ballot that has the voters express their preferences between each pair of the three candidates who received the most votes in September. Whichever of these three candidates beats each other opponent in these head-to-head matchups (or whichever of the three candidates has the narrowest single defeat), would win all of the electoral votes of the states joining the agreement. As with NPVIC itself, as long as the states joining this agreement collectively have at least 270 electoral votes, then the winner according to the top-3 method would become president.

To be clear, unlike with NPVIC itself, the calculation of the top-3 winner would not necessarily be based on ballots cast in all fifty states (plus DC). Instead, the top-3 winner would be determined among all ballots cast collectively in the states joining the agreement. (This would be true for counting the September ballots to determine which three candidates advance to the November top-3 ballot in those states, as well as counting the November ballots in those states.) In other words, the states joining the agreement would be pooling all their votes into one agreement-wide election. Any states that did not join the agreement would not have their ballots as part of the pool. But if (as with NPVIC) the agreement doesn’t take effect unless and until enough states join to control 270 electoral votes, then once the agreement becomes operational, whichever candidate wins the top-3 election based on all the ballots that form the agreement’s pool is guaranteed to win the presidency regardless of how many votes that candidate gets in states that have not joined the agreement.

Moreover, precisely because the agreement controls the outcome of the presidential election in this way (as soon as enough states join), it will cause every other state to join the agreement immediately after it reaches the magic number of 270. No state will want to be shut out of determining who wins the presidency. Therefore, even if a state did not join the top-3 agreement initially, it would add its own ballots to the top-3 pool so that its citizens can participate in the process that will control the election’s outcome.

Unfortunately, it is almost certainly too late to get enough states to join this kind of agreement to make the top-3 system operational for this year’s election—as much as this year’s election in particular would benefit from this system. Let’s hope the frustration that Haley, Manchin, and others must have knowing that a majority of American voters would prefer them to either Biden or Trump, and yet the current system does not enable American voters to make this choice, will motivate Haley, Manchin, and others to organize an effort to adopt the top-3 system in time for the 2028 election.

As a country, we are skirting disaster because we don’t have the top-3 system in place already. Our failure, at least so far, to solve our border crisis and to provide funding for Ukraine (as a means to protect the NATO alliance from Russian aggression), are among the failures of governance attributable to our existing two-party electoral system under current conditions of hyperpolarization.

We wouldn’t be in the mess we are currently in if we already had top-3 voting. We must pray that we can manage to survive the current challenges until we are capable of realizing that this will solve our difficulties—and that it’s not already too late.