Menu Close
What if this was our choice on Election Day? AP Photos/Gary Landers and Paul Sancya

How majority voting betrayed voters again in 2016

The system for electing the U.S. president went woefully wrong from the very beginning of 2016.

First, the two most disliked candidates ever nominated – Hillary Rodham Clinton and Donald J. Trump – emerged victors from their parties’ primaries, but shouldn’t have. Second, the increasingly controversial Electoral College system will formally elect Trump on December 19 despite Clinton’s lead of over two million in the popular vote.

The system is “rigged” all right, not for a candidate but against the voter. It fails to elect candidates the voters really want. Why? And what should be done about it?

Years of work in developing fair methods of representation and systems for electing candidates that truly respond to the opinions of the electorate have convinced me that the real culprit is majority voting and not the Electoral College. I will give my reasons.

Majority voting’s failures

Majority voting (MV) is an extremely crude approximation of the opinion of the electorate that has often elected a candidate counter to the popular will.

Walter Lippmann – claimed by many to be the most influential American journalist of the 20th century – realized this in 1925:

“But what in fact is an election? We call it an expression of the popular will. But is it? We go into a polling booth and mark a cross on a piece of paper for one of two, or perhaps three or four names. Have we expressed our thoughts … ? Presumably we have a number of thoughts on this and that with many buts and ifs and ors. Surely the cross on a piece of paper does not express them … [C]alling a vote the expression of our mind is an empty fiction.”

There have been 57 presidential elections. By my count, 12 of them elected candidates that were almost certainly not the true choices of the electorate, the last three occurring in 1912, 1992 and 2000.

Woodrow Wilson was elected in 1912 (with 41.8 percent of the popular vote) against incumbent Republican President William Howard Taft (23.2 percent) because of the Bull Moose candidacy of the former Republican President Teddy Roosevelt (27.4 percent): Either of them would most likely have won head-to-head against Wilson.

A similar scenario occurred in 1992 with Bill Clinton (43.0 percent) winning against George H. W. Bush (37.4 percent) because of the candidacy of Ross Perot (18.9 percent): Bush (father) would almost surely have beaten Clinton head-to-head.

And in 2000 George W. Bush (47.9 percent) won with a bare majority of 271 Electoral College votes against Al Gore (48.4 percent) because of the candidacy of Ralph Nader. Bush’s lead of a mere 537 (out of nearly 6 million) votes in Florida would have easily been erased if the 97,000 who voted for Nader could have expressed their preference for Gore.

Why does this happen? Because, as Lippmann suggested, MV does not permit voters to express their opinions fully.

In 1912 it was impossible for a Roosevelt voter to express a preference for Taft over Wilson, or a Taft voter to express a preference for Roosevelt over Wilson. Similarly, it was impossible for voters to express their preference for Bush (father) and Perot over Clinton in 1992, or for Nader voters in Florida to express their preference for Gore rather than Bush (son) in 2000. Had they been able to express their opinions of the candidates more accurately, the outcomes would have been different.

MV, as old as the hills, is merely a mechanism that has been accepted by force of habit. As Thomas Paine wrote in 1776 in “Common Sense” – “the most incendiary and popular pamphlet of the entire revolutionary era”:

“A long habit of not thinking a thing wrong gives it a superficial appearance of being right.”

Majority voting is such a thing. It is thought to be democratic, but isn’t, as these examples (and many others) show.

Don Lamb, an employee of La Scala Restaurant in Little Rock, Arkansas, cleans the front window of the restaurant, Saturday, Oct. 31, 1992. AP Photo/Danny Johnston

Ranked voting’s failures

Some reformers advocate another mechanism, “ranked voting” (RV). Instead of choosing one among the candidates the voter lists them all from their most to their least preferred. This 18th-century idea (from the French mathematician and political scientist Jean-Charles de Borda) is a better scheme for voters to express themselves – and so it must have seemed to the narrow majority of 51.99 percent of Maine’s voters who adopted one version of the possible methods based on RV, Ranked Choice Voting, in a statewide vote on November 8.

However, I argue that they were sold a bill of goods: RV’s drawbacks completely disqualify it.

First and foremost, RV is far from permitting an adequate expression of the voters’ opinions. A voter cannot reject all candidates, cannot consider two candidates equally good and cannot express strong versus lukewarm support (or rejection).

Furthermore, when RV has actually been used by juries in such competitions as figure skating, gymnastics and diving, its results have sometimes been so wildly peculiar that increasingly it has been abandoned in favor of methods that ask judges to evaluate competitors instead of ranking them. Figure skating juries’ rules, for example, made the change in response to the 2002 winter Olympic scandal in pairs figure skating.

Majority judgment

My colleague, Rida Laraki, and I have developed a new method of voting, majority judgment (MJ), which avoids the drawbacks of MV and RV.

MJ asks voters a simple and natural question such as that recently posed by the Pew Research Center: “What kind of president do you think each of the following would be – a great, good, average, poor or terrible president?” In its last national survey of registered voters (Oct. 20-25) Pew reported the following results (here adjusted to sum to 100 percent):

All one needs to do is look at the evaluations of the two candidates in the table above to conclude that Clinton is better evaluated than Trump.

But what exactly is the majority opinion?

Clinton would be an Average President because in a majority vote between Average and any other “grade,” it wins. This is most easily seen by noting that a majority of 8%+27%+20%=55% believes she would be at least Average – so Average defeats any lower grade – and a majority of 20%+11%+34%=65% that she would be at most Average – so Average defeats any higher grade. It suffices to start from each end of the spectrum adding percentages until a majority is reached; in practice the sums from both directions will always reach a majority at the same grade.

Similarly, a majority believes Trump would be a Poor President because 54 percent believes he would be at least Poor and 57 percent that he would be at most Poor. With these evaluations majority judgment elects Clinton since the majority evaluates her above Trump.

MJ simply uses the majority principle – the idea that the majority can represent the whole – to deduce the electorate’s evaluation of every candidate, called their majority-grades, instead of using it to compare the number of votes each candidate receives.

No system is perfect. But majority judgment is far superior to any other known system. Here’s why:

  • It is easier and more natural for voters since grading is familiar since school days;
  • It obtains more information from voters and puts more confidence in them by permitting them to express their opinions accurately;
  • It gives more information about the standing of candidates in the eyes of the public – had Clinton won she would have known her standing: Average;
  • Most importantly, it elects the candidate highest in the esteem of the electorate.
John and Colleen Kramer, of Stockton, Missouri, vote at the Caplinger Mills Trading Post on Tuesday, Nov. 8, 2016, in Caplinger Mills, Missouri. (AP Photo/Charlie Riedel)

What happened this year?

Pew Research – without realizing that their question serves as the basis of a method of voting – posed exactly the same question this year in January, March and August as well as late October.

In every case the majority evaluated Clinton an Average President and Trump a Poor President; moreover, their respective grades remained remarkably similar over all four polls, suggesting that despite all the hoopla – emails, sexism, racism, walls, FBI, secret speeches, jail and so much more – the electorate’s opinions concerning the two candidates remained very much the same throughout the year.

And yet Trump beat Clinton. Why? MV denied voters the right to express their opinions adequately in the state face-to-face encounters.

U.S. voters were in revolt, determined to show their exasperation with politicians. But how, with the majority vote, could they express this disgust other than by voting for Trump?

With majority judgment some of them would surely have rated Clinton as Poor or Terrible to make the point, but Trump as Poor or Terrible as well, exactly as the Pew survey shows.

This could well have been the case in each of several states where their total votes were close such as Florida (a difference of 1.3 percent in their vote totals), Michigan (a difference of 0.3 percent), Wisconsin (a difference of 0.8 percent) and Pennsylvania (a difference of 1.1 percent). With MJ the result would then have been much closer to a true expression of voters’ opinions and so of the popular will: 307 Electoral College votes for Clinton, 231 for Trump.

Well before the vote on Nov. 8 something else went wrong. Trump and Clinton should not have been the victors in the Republican and Democratic primaries – they are, after all, generally considered to be the least popular candidates of recent history. But the primaries were decided by majority vote as well. Had the primaries used majority judgment, the general election would have pitted Bernie Sanders against John Kasich.

Imagine how different the country and the world would feel today – and be tomorrow – had they been the candidates!

The time has come to replace the obviously undemocratic mechanism of the majority vote by a method that captures the true will of the electorate: majority judgment.

Editor’s note: this article was updated to make clear that Ranked Voting has different versions, including Ranked Choice Voting.

Want to write?

Write an article and join a growing community of more than 180,400 academics and researchers from 4,911 institutions.

Register now