Hillary Clinton and Donald Trump are the two most disliked presidential candidates since the advent of modern polling in the 1950s, so it’s no surprise that recent polls have placed Support for the top third-party candidate in the presidential race, Libertarian Gary Johnson, at as much as 15%, with Green candidate Jill Stein receiving as much as 5% support. However, the results of a new experiment indicate that such results are dramatically overstating support for these candidates, with Johnson’s actual support coming in at less than 7% and Stein showing negligible levels of support from voters. Polls are creating an illusion of significant support when very little actually exists. While this election may be the best opportunity that third-party candidates have had in a long time, they’re still nowhere close to mounting an effective challenge to the two-party system.
This has to be especially disappointing for minor parties, as 2016 looked like it could be a breakthrough year for minor parties. Johnson has held major elective office as governor of New Mexico; Stein seemed ready to appeal to liberals disappointed by Clinton’s centrist policies. Combined with historically high unfavorability ratings for the major-party candidates, it wasn’t unreasonable to think that Johnson or Stein could hit 15% in the polls, gain access to the debates, and make a run from there. When Johnson’s support neared 10% in many national polls, it seemed like it might really be happening, that the major-party grip on American politics might finally be fading — but there was a problem. The polls, as it turns out, were wrong.
When polls choose to include Johnson and Stein along with Clinton and Trump, the results for the minor-party candidates include two different kinds of voters: those who support those candidates in particular and those who will voice support for anyone who isn’t Clinton or Trump. When the results of polls that include Johnson and Stein by name are compared with those that ask about support for “someone else,” something weird happens: There are way more Johnson and Stein supporters than there are people who say that they want to vote for “someone else” or say that they “don’t know.” This means that there’s a substantial group who say they’ll support Johnson or Stein when they’re listed and who voice support for Clinton or Trump when they’re not. These voters aren’t actually supporters of the minor parties — they’re just saying a name other than Clinton or Trump when that option is offered. Of course, any legitimate survey allows respondents to indicate that they’re going to support someone other than the major-party candidates, but relatively few voters choose the “other” option. This could be because they can’t name a candidate other than those listed, don’t want to look uninformed, or don’t want to go through the hassle of the inevitable follow-up question.
A lot of this has to do with how people answer survey questions. We like to think that survey respondents are revealing considered opinions when they answer questions, but many respondents, especially those who don’t think about politics very much, are just answering questions quickly. They haven’t thought much about who they’re going to vote for, and they may not even have thought much about which party they associate with until the pollster asks them. So it shouldn’t be a surprise when their responses shift based on the options given in the question.
At first glance, this doesn’t seem like that big a problem: Perhaps those voters will do the same thing on Election Day and actually vote for one of the minor parties. Aside from the fact that voters who aren’t tied to any candidate are unlikely to vote at all, there’s the fact that Johnson and Stein are far from the only minor-party candidates running and their names are not especially high on many ballots. In New Jersey, for instance, there are seven minor-party candidates listed, with Johnson listed fifth overall and Stein listed ninth. In Davenport, Iowa, Stein is listed fourth and Johnson is sixth. If voters are looking to cast a ballot for “someone other than Trump or Clinton,” there’s no reason to believe that Johnson and Stein will be the only recipients of the votes.
To figure out how many voters are actually supporting the minor-party candidates, an experiment was embedded in the most recent PublicMind poll of likely voters nationally. In the study, respondents were randomly assigned to receive one of two versions of the ballot question. In one version they were asked to choose between Clinton, Trump, Johnson, and Stein. In the other version of the question, they were asked to choose between Clinton, Trump, and two other minor-party candidates: James Hedges (nominee of the Prohibition Party) and Monica Moorehead (nominee of the Workers World Party). When Johnson and Stein — the third-party candidates most commonly included in polls — were given as choice, Johnson received 11% of the support of likely voters and Stein got 3% support, figures in line with what they are receiving in other polls.
However, when Hedges and Moorehead were included in the question instead, Hedges received 4% support and Moorehead got 3% support. Support for Clinton and Trump are unchanged by the choice of third-party candidates selected, while the proportion of voters saying they don’t know or are undecided goes from 4% to 9%. On Election Day it’s expected that Moorehead and Hedges will receive close to zero support; in 2008 and 2012 third parties other than the Green and Libertarian parties received less than 0.3% of the vote combined.
The results indicate that about 7% of voters are expressing support for anyone who isn’t Clinton or Trump. Johnson seems to have actual support of 5%–7%, as indicated by the increase in the proportion of voters who say they’re undecided or don’t know when Johnson is not included, and the difference in support between Hedges and Johnson. The news is worse for Stein. The fact that she does no better than Moorehead or Hedges indicates that her support in polls is an artifact of the surveys rather than reflecting any real support; anyone named in the polls would do just as well.
The structure of the American presidential election system means that minor parties have a huge uphill climb if they want to be competitive. Maybe someday, voters will be fed up enough with the major parties to support them — but this is not that day.