Portside

I’m glad we’re once again debating the antiquated Electoral College, inspired by Elizabeth Warren’s recent call to abolish the vintage 1787 Rube Goldberg-style invention for picking a president at a time when popular democracy was highly suspect and slave states were fearful of losing their power based on population alone.

Abolishing the Electoral College is a modern-democratic-principles slam dunk: Every vote should count the same, no matter where you live. Absent a baroque defense of hyper-federalization or a nakedly partisan argument dressed in the breeches and waistcoat of blind traditionalism, there is no good case for the Electoral College. Rather, the more interesting question is what to replace it with.

The most popular replacement proposal is a direct popular vote: Whoever gets the most votes nationwide wins. Certainly, this is an improvement over the Electoral College. But it still employs the worst form of voting — simple plurality. I say worst form because simple plurality means a candidate does not have to win a majority of votes to win the election, and leaves open the possibility that the least-preferred candidate wins because two other candidates split the remaining vote. Or, as more commonly happens, plurality elections marginalize third parties as spoilers, and campaigns become simply lesser-of-two-evils contests.

The two-round system, which most presidential democracies use, is an obvious improvement over the single-round plurality system. In the second round, the winning candidate mathematically has to appeal to a majority. This generally produces more moderate winnersthan simple plurality elections, which generate more extreme winners and more divisive politics. Eric Holder, among others, has proposed the two-round system.

France as a case study in the two-round presidential election

To understand how a two-round system might work, consider the 2017 French presidential election, which can illustrate both the benefits and the dangers of a two-round system. In the first-round election, centrist Emmanuel Macron got 24 percent of the vote and far-right candidate Marine Le Pen got 21 percent, narrowly coming in second.

In the second round, Macron won handily, defeating Le Pen 66 percent to 33 percent. In a head-to-head contest, unclouded by long-standing partisan affiliations, the French people solidly rejected the far-right populist, and France wound up with the candidate that most citizens could ultimately agree on.

The adage in France goes like this: On the first ballot, the voters select. On the second ballot, they elect. Or, put another way, on the first ballot voters choose; on the second they eliminate.

The 2017 French election shows a benefit of the two-round system: It created space for a shake-up of the existing ossified party structure. France’s two long-standing major political parties, the center-left Socialists and the center-right Republicans, had lost considerable support. In the two-round system, it was possible for a party to emerge without being dismissed as a spoiler.

But things could have easily gone differently. The fourth-place finisher was Jean-Luc Mélenchon, a far-left populist, who won just short of 20 percent of the vote. But Mélenchon might have done a little better, Macron a little worse, and the French people might have been left to choose between a far-left populist and a far-right populist.

A two-round system encourages more candidates but then requires a high level of strategic voting for voters, who might have complicated preferences about who they think can win, versus who they’d most like to win.

Moreover, Macron may have won with 66 percent of the general election vote, but the large margin was not so much a reflection of his broad-based popularity as it was Le Pen’s unpopularity. On his own, his popularity has fallen after he tried to claim a broader mandate than his support deserved.

Why ranked-choice voting is superior to the two-round system

A better alternative would be ranked-choice voting for president. As the name suggests, ranked-choice voting lets voters rank their choices. Voters mark their first choice candidate first, their second choice candidate second, then their third-choice candidate third, and so on. Easy as 1-2-3. Think of it as a political listicle: Candidates in this election, ranked.

The votes are then tallied as follows: If one candidate has an outright majority of first-place votes, that candidate wins. But if no candidate has a majority in the first round, second-choice preferences come into play. The candidate with the fewest number of first-choice votes is eliminated, and voters who had ranked that candidate first have their votes transferred to the candidate they ranked second. This continues until a single candidate gathers a majority, with subsequent preferences transferring as candidates get eliminated from the bottom up.

Ranked-choice voting improves on the two-round runoff elections in the following ways. First, it is more likely to produce a broadly acceptable winner. Ranked-choice voting rewards candidates who can appeal most broadly, because candidates compete to be voters’ second and third choices as well as their first. By contrast, in a two-round system, it’s quite possiblefor the two candidates with the most dedicated but not necessarily broadest support to advance to the final round, recreating the lesser-of-two-evils problem.

Second, a two-round system demands complex strategic calculations for voters, who face a trade-off between who they want to win and who they think can make it to the second round. This can lead to very different outcomes based on small variations in polling. California’s top-two primary displayed this problem last year, when Democratic voters struggled to coordinate around the candidate most likely to advance to the second round and avoid splitting a crowded field and allowing two Republicans to advance to the general election.

The complexities of these strategic calculations can also invite political operatives to support “loser” or “dummy” candidates designed to fracture the field, confuse voters, and pull away crucial support from frontrunners in crowded fields (as also happened in California). The ranked-choice voting renders this strategy pointless. There are no spoilers in ranked-choice voting.

Third, the two-round election can also lead to unintended vote-splitting among candidates competing for similar voters, denying voters the ability to find agreement through ranking. In ranked-choice voting, voters can vote sincerely. They don’t need to worry if they are wasting their votes. And nobody has an incentive to run loser candidates.

Fourth, an instant runoff is both less costly for governments (one election costs less than two) and for voters (one visit to the polls takes less time than two). Voter turnout often declines in a second round, either because voters’ preferred candidate didn’t advance or because voters simply lose interest. The voters most likely to drop off in the second round are the least educated.

Additionally, if we’re going to incorporate ranked-choice voting into how we select presidents, we should also use it for the primaries.

All else equal, a popular vote is better than the Electoral College (for obvious reasons). And a two-round popular vote is better than a single-round plurality popular vote, because by requiring the winner to earn a true majority, it avoids the problem of extreme minority candidates.

But ranked-choice voting is superior to the two-round system because it simultaneously maximizes voter participation and broad-based coalition-building. It also allows voters to vote sincerely and is most likely to avoid perverse outcomes.

We know more today about elections than we did in 1787. Our institutions should reflect that.

The Framers were political scientists with no existing model for electing a president and tremendous ambivalence over what they wanted out of an executive. They were practical politicians trying to work out an acceptable compromise.

We can forgive them for coming up with a sub-optimal approach given the urgency and limited time and political constraints. But with 230 years of experience in modern democracy, we now have plenty of data on what works best.

No electoral system is perfect. But some are demonstrably better than others. Rather than limit ourselves to an antiquated and inferior tradition, we should strive for the best possible. National popular ranked-choice voting is the best possible way of electing a president.

Lee Drutman is a senior fellow in the Political Reform program at New America. He is the author of The Business of America is Lobbying: How Corporations Became Politicized and Politics Became More Corporate.