It's Time for a New Primary Electoral System

It's Time for a New Primary Electoral System
Mark Felix/The Flint via AP
Story Stream
recent articles

When Donald Trump complained that the “system” was rigged against him, he was wrong. The system was not rigged at all, though the electoral system in which he contested for the Republican nomination actually helped his candidacy.

That system, the Anglo-American “first-past-the-post” or plurality method of election, declares a candidate victorious if he or she wins the largest number of votes, even if that number is less—sometimes far less—than a majority. Such a system works very well if there are two candidates or two major candidates and a few minor candidates.

It served the Democrats well in 2016, and it has generally served the country well in general elections by building a two-party system. But it is a very poor electoral method with which to run a primary election with 17 candidates, the number the GOP had at the start of the primary season. That is because it allows a candidate in a large field (or even one with just several candidates) to win with a relatively low plurality of votes. And if this result follows in state after state, it can produce a victorious candidate who has only minority support in his own party, because most voters have voted in favor of someone else.

Worse yet, if the victorious candidate has intense support among that minority but very thin support among other voters—in other words, if he is the second or third choice of relatively few voters—he will have great difficulty winning the general election. In a significant way, such a candidate will not be the most popular candidate who contested in that primary, because in a large field with only one winner, the most popular candidate—the one with widespread support—can only be discerned by identifying second and third, and maybe even fourth, preferences among the voters.

That is the Republican problem today. Do a little thought experiment: Suppose the standard for election were not a mere plurality but a real majority of more than 50 percent, and suppose there were a way of counting all voters’ second and third choices. Would Trump have been the winner? Probably not, because we know that he has intense minority support but great opposition among voters who cast ballots for other candidates.

There is an electoral system that surfaces second and third (and lower-order) preferences that has been used successfully in Australia for a century. It is called the Alternative Vote. It uses a ballot that asks voters to mark first, second, third, etc., preferences. Then, if there are more than two candidates, the one with the fewest first preferences is dropped out, and his or her voters’ second preferences are reallocated as if they were first preferences. The process is repeated until one candidate receives a majority. (There are several ways of doing the actual counting, but in each case the aim is to find a winner with real majority support among all the voters.)

Generally, the leader in first preferences wins even when all preferences have been counted, so alternative voting is not a radical departure from first-past-the-post, but it can produce a different result when most voters strongly prefer someone other than the first-preference leader or find that leader unacceptable.

Variants of this system have been used increasingly in American state elections, where there are often multiple candidates at the municipal level. In the U.S., it is usually called ranked-choice or instant-runoff voting. The U.S. has a long history of runoff elections, precisely to avert the possibility of choosing a candidate who wins on a small plurality. The instant runoff, however, has major advantages over a second-stage, later runoff. It does not have to be limited to the top two candidates, and it saves the expense of another election, often with poor turnout, and averts the suspense of deciding on a winner.

Another advantage is that all candidates in a crowded field must think from the very beginning not just about their most fervent supporters; they must also not offend the supporters of other candidates or risk losing their second-preference votes. Run-off elections make for fewer insults among candidates and more coalition building within parties at the primary-election stage, so the party goes into the general election more unified than it might otherwise be.

Now, it’s true that instant runoff requires a preferential ballot, and it may be that some states will not opt for that process, despite its advantages. Those states should then consider what some states do already—allocate to candidates a share of delegates corresponding to the candidates’ share of votes in the state primary.

Proportional representation would be a terrible electoral system for general elections in the U.S. for many reasons, but it is a good second-best in a field of several candidates in a primary election. It does not identify and count second and third choices as instant runoff voting does, but in a crowded field it acknowledges and counts the multiple first choices of voters, as plurality elections do not. 

If delegates are awarded proportionately and no candidate has a majority, it becomes obvious to voters that there is no clear winner, and nobody can make Trump-like claims that there is one.  Candidates will have an incentive not to drop out of the race at an early stage, as Republicans did this year.  In some close presidential primary contests, where no candidate wins a majority of delegates nationwide, proportional allocation could leave the ultimate nominating decision to convention delegates.

That is not a bad outcome.  The primary results will have given the public and the delegates a good indication of which candidates are popular without conferring the mantle of leadership on a candidate who is highly popular with some segments of the public but deeply unpopular with others. Convention delegates can then do what they should be doing all along: assess the electability of all the candidates and then give the party a candidate with a good chance of winning.  If both parties did this, it would be good for them and for voters who prefer not to hold their nose as they pull the lever in November.

Donald L. Horowitz is the James B. Duke Professor of Law and Political Science Emeritus at Duke University. 

Show commentsHide Comments