2016 Republican Presidential candidates Senator Ted Cruz, a Republican from Texas, from left, Donald Trump, president and chief executive of Trump Organization Inc., and Senator Marco Rubio, a Republican from Florida. (Daniel Acker/Bloomberg)

Hard though it may be to imagine now, a few months ago, some observers were making the case that Donald Trump’s presence in the Republican primary was good for “establishment” Republicans. Dana Milbank, for instance, argued that his outlandish behavior could conceivably make other Republican candidates appear more palatable by association. One of Jeb Bush’s most effective campaign moments came in a debate when he attempted to contrast what he saw as Trump’s fanciful policy goals with the more tangible, reasoned visions of traditional candidates.

But this logic has not been tested in a systematic way, so I attempted to find out if it holds up to scientific scrutiny: Does the presence of a particular candidate affect impressions people have of candidates around her?

To answer this question, I administered an online survey in December 2015 in which Republican voters were asked to read about and evaluate different hypothetical candidates running in a party primary. Individuals were randomly assigned to one of three scenarios:

  1. They evaluated an ideologically moderate candidate running in their party’s primary;
  2. They evaluated an ideologically conservative candidate running in their party’s primary; or
  3. They evaluated both the moderate and the extreme candidate side-by-side.

I then compared Republicans’ evaluations of the moderate candidate when he was considered in isolation (Scenario 1) with evaluations of him when he was considered alongside the conservative (Scenario 3).

The answer: Yes, ideologically extreme candidates do change how Republicans view other primary candidates. The moderate candidate was viewed as much more conservative when he was evaluated in isolation than when he was assessed alongside a more conservative candidate. Similarly, the conservative candidate was perceived to be somewhat more conservative when considered by himself relative to when he is contrasted with the moderate candidate (not shown).

Voters also express different levels of support for candidates based on who else is on their minds. Since Republicans are the more conservative party, it follows that among Republicans who evaluated both candidates (Scenario 3), more than two-thirds indicated that they liked the conservative candidate more than the moderate option.

It makes sense, then, that we see Republicans express significantly higher levels of support for an ideologically moderate candidate when that candidate is alone rather than next to a conservative option. Even though the moderate candidate is exactly the same in both scenarios, the presence of a more conservative option causes Republican voters to view that moderate candidate less favorably.

The technical stuff

These are preliminary results from an online survey experiment. While numerous steps were taken to ensure as valid a sample as possible, such samples are never perfectly representative of the overall population. Although individuals in this sample are very similar ideologically to respondents in other national datasets (like the 2014 General Social Survey and 2016 American National Election Survey Pilot Study), we don’t want to infer too much from samples like these. This is just the first foray into measuring these effects.

But given that, the results are provocative, suggesting that voters’ impressions of a candidate are shaped not merely by the candidate herself, but also by the candidates around her—not just in this race but potentially beyond.

This research suggests that a candidate can be seen in significantly different ways based on who else is running. This makes sense intuitively. But it underscores an important point: Candidates can only control so much. Jeb Bush was a front-runner last summer, enjoying the backing of a SuperPAC with well over $100 million in its campaign coffers and poised to take advantage of Florida’s recent switch to a winner-take-all primary. The electoral stars were aligned for Bush to compete for the nomination, yet he found himself out of the running before Florida even voted.

Had Trump not thrown his hat in the ring, Bush might well have remained in the top tier of candidates heading into the Florida primary on March 15. But Trump’s presence seems to have changed how some people evaluate “insider” candidates — even if those candidates themselves remained largely the same.

As the candidate fields narrow and we seek explanations for what went wrong for particular presidential hopefuls, we should keep in mind that it’s not just about the candidates. It’s about the other candidates.

Eric Loepp is an assistant professor at the University of Wisconsin, Whitewater.