For many Republican candidates, an endorsement from former President Trump is seen as the key to electoral success. But how important are Trump’s endorsements in voters’ minds? Could other factors, such as policy stances be just as, or even more, important? Scott Blatte, Danielle Piccoli and Matt Zachem surveyed over 1,300 Americans to determine the effects of a Trump endorsement on support for candidates. They find that an endorsement from Trump decreases support for that candidate among Democrats but does not increase Republicans’ support. By contrast, when candidates were shown to support typical Republican policies, this decreased Democrats’ and increased Republicans’ support.

Since the end of his presidency in January 2021, former President Donald Trump has retained his place as the Republican Party’s de facto leader. In a stark departure from previous one-term presidents, Trump has continued to endorse his favored Republican candidates in their primary elections. These endorsements undoubtedly help consolidate Trump’s power, but do they serve the party’s interests?

Both The Economist and The Washington Post sought to answer this question following the 2022 midterms, and reached the same conclusion: Trump’s endorsement is associated with electoral underperformance for the Republican candidate who receives it. But what these analyses are unable to answer is whether Trump’s endorsement causes the underperformance, or Trump simply prefers weaker candidates who are destined to disappoint with or without an endorsement. In a recent experiment, we attempted to answer this question, while offering evidence on the relative importance of an endorsement.

Our research reveals three asymmetries associated with former President Trump’s endorsement. Our first two asymmetries are partisan. First, our findings suggest that a Trump endorsement causally decreases the likelihood that the average voter will vote for a Republican candidate. Although these decreases were significant among Democratic respondents, there were no significant changes for Republican respondents. Second, we found that Republicans value a candidate who supports their preferred policies more than a Trump-endorsed candidate, whereas Democrats do not show a preference. Our final asymmetry is directional. Unlike Trump’s endorsement, his disavowal of a candidate had no effect on voting behavior.

Testing the effects of a Trump endorsement on candidate support

In October 2022, just days before the 2022 midterms, we ran a survey experiment on over 1,300 American adults. We gave all respondents the following prompt: “Suppose you are deciding who to vote for in the General Election. Terry Mitchell is a 56-year-old white male. He is the Republican nominee for a congressional seat in your state.” All respondents were randomly assigned to view either “conventional” Republican positions (“unconventional” Democratic positions) or “unconventional” Republican positions (“conventional” Democratic positions). Respondents then received one of the following conditions: a Trump endorsement, a Trump anti-endorsement (i.e., President Trump urging voters not to vote for Mitchell), or no mention of Trump. We followed this with a question on how likely the respondent was to vote for “Terry Mitchell”.

Endorsements and policy matter

So, does a Trump endorsement matter? Suggestively, yes. As Figure 1 shows, we found that an endorsement from former President Trump leads to a decrease in support for that candidate among Democrats, but no significant increase in support among Republicans. Overall, this translated to a roughly four-point reduction in the overall likelihood of voting for a hypothetical Republican candidate, a marginally significant finding. In comparison, among all respondents, the effect of a Trump urging voters not to vote for the candidate, was close to a precise zero, indicating an unequal effect.

Figure 1 – Likelihood of voting for candidate by party affiliation

Our experiment also allowed us to examine the effect of aligned policy stances, meaning positions commonly associated with Republican politicians and voters. We find that both Democratic and Republican voters were significantly affected by policy stances. Democrats decreased their likelihood of voting for Mitchell by 14 points and Republicans increased their likelihood of voting for Mitchell by 16 points when presented with typical Republican policies. These findings broadly show that policy stances do matter to voters.

“Kari Lake & Donald Trump” (CC BY-SA 2.0) by Gage Skidmore

Which matters more, policy stances or endorsements?

Because we randomized policy stances and the endorsement simultaneously, we can directly compare the policy and endorsement effects. For Democrats, there was little evidence that the two effects differed from each other. However, Republicans’ likelihood of voting was significantly more sensitive to policy than it was to Trump’s endorsement. This is a surprising finding, with no explanation readily available from our experiment. However, we speculate that the average Republican expects that a Republican candidate will be pro-Trump. Mentioning an endorsement only confirms what Republican respondents already assume. On the other hand, Democrats might be less likely to assume a Trump affiliation until told otherwise.

Endorsements can directly counteract policy.

Our final question was whether one effect counteracts the other. First, we looked at Democrats. We found that even if our hypothetical candidate held “conventional” Democratic policy stances, the mention of a Trump endorsement significantly decreased Democrats’ likelihood of voting for the aligned candidate by 14 points. When we looked at the results for Republicans, we observed nearly identical trends, but in the opposite direction. A Trump endorsement paired with “unconventional” Republican stances significantly increased Republicans’ likelihood of supporting our hypothetical candidate by 11 points. In other words, we found evidence that a Trump endorsement could counteract the effect of those stances by resulting in a significant decrease in likelihood of voting for Democrats and a significant positive increase in likelihood of voting for Republicans.

Our findings provide insight into the decision-making process of voters. Trump’s endorsement negatively affects Democrats, an effect that is only weakly offset by a positive effect among Republicans. Policy, while still important, does not always reign supreme, as Democrats provide no evidence of valuing policy more than endorsements. However, an endorsement is powerful enough to decrease Democrats’ likelihood of voting even when the candidate has “conventional” Democratic policy stances and increase Republicans’ likelihood of voting even when the candidate has “unconventional” Republican policy stances. This finding shows that although policy plays a substantial role in voting decisions, it is not immune to elite signals.

Our final point concerns the (presumptive) “elephant in the room”: the implications for upcoming elections. Although Trump has been out of office for almost two years, he continues to impact voters. As the Republican nominee in the 2024 Presidential Election, we speculate that the Trump effect will only be amplified this November. Our findings indicate that Trump’s presence will be a boon for down-ballot Democrats and a bane for Republicans. 

• This article is based on the paper, ‘The Causal Effects of a Trump Endorsement on Voter Preferences in a General Election Scenario’, in PS: Political Science & Politics. 
• Please read our comments policy before commenting.   
• Note: This article gives the views of the author, and not the position of USAPP – American Politics and Policy, nor the London School of Economics. 
• Shortened URL for this post: https://wp.me/p3I2YF-dR2