Rally Around the Winner--A Two-Wave Panel Survey on the Impact of the U.S. Election on Foreign Policy Stances

Rally Around the Winner--A Two-Wave Panel Survey on the Impact of the U.S. Election on Foreign Policy Stances (DOI: https://doi.org/10.1093/ijpor/edaf040) was written by Eyal Rubinson and Gal Bitton-Alayof in 2025. It was published in International Journal of Public Opinion Research (vol. 37, no. 3).

A common model of voting behavior is that voters select a candidate based on policy positions or salient issues. An alternative, constructivist model introduces a potential causal link from election results to policy preferences. The authors call this a 'winner effect'. Possible explanations include:

• "media coverage and elite messaging often portray the winning candidate as more legitimate or competent, hence increasing the credibility of their policy positions, particularly on lower salience issues such as foreign policy (Banducci & Karp, 2003)"

• "outcome as a signal of public consensus" leading to voters changing their views "consistent with patterns of informational conformity (Cialdini & Goldstein, 2004)"

• "motivated reasoning": "individuals adjust their attitudes to reduce dissonance between earlier views and the new political reality (Redlawsk, 2002)"
• NOT a bandwagon effect; strictly after realization of the outcome

The authors use a two wave survey experiment, making use of the 2024 U.S. general election.

• Prolific web panel with incentives
  • First wave = Nov. 1-2
  • Second wave = Nov. 7-9
• Anticipated 20% attrition; targeted 900 responses for first wave so that second wave would have 720
• Actual RR:
  • First wave = 812 responses
  • Second wave = 692
• Questions about candidates' positions on...

  • relations with China

  • relations with Middle East, mainly Israel and Iran

  • approach to Russia-Ukraine War

  • relations with NATO

  • tax proposals
  • These questions skew towards foreign relations because the authors expect that this effect will be more present in this issue space.
  • Candidates' platforms were reproduced and then the survey asked for agreement on a 5-point scale.

The authors estimate attrition bias using t tests and chi-squared tests of demographic indicators (age, gender, ethnicity, and political affiliation) between second wave respondents and non-respondents. (These characteristics were measured in the first wave.) Age was found to be predictive of attrition.

The authors evaluate the agreement scores for validity and internal consistency by calculating an average and Cronbach's alpha for each candidate. Note that the sample is split by the three principal party affiliations (Democrat, Republican, and Independent). These statistics are consistent between waves. Between waves, Democrats' agreement with Harris' platform fell from mean of 4.18 to 4.10, while agreement with Trump's platform rose from 2.27 to 2.39. Republicans and Independents did not feature any such divergence; that is, Independents became more supportive of all platforms after the election.

The authors examine the further change after elections using ANOVA. Note that again party affiliation is part of the design. The election did have an effect on agreement, but with small magnitude. Party affiliation is predictive; Republicans had highest support for platforms (mean of 3.75), followed by Independents (3.42), and finally by Democrats (3.24). Lastly, a significant interaction between candidate and time, indicating that the election had a different effect on agreement with Harris vs. Trump.

Repeating this analysis with only foreign relations questions yielded equivalent interpretations, suggesting no difference between foreign relations and other issue spaces.

CategoryRicottone CategoryReadingNotes