Sage Journals: Discover world-class research

Abstract

When politicians are swayed by scientific evidence, are they met with admiration or skepticism? U.S. college students and Prolific panelists (N = 1,213) rated six vignettes in a 2 (politician’s action: maintain original position vs. change position) × 3 (alignment with participant’s position: aligned vs. misaligned vs. control) within-subjects design. Politicians who changed their position were judged as more flip-floppy, but also warmer and more competent compared to those who maintained their stance. In addition, politicians who changed their stance elicited stronger voting intentions than those who maintained their stance. When the final stance of the politician was misaligned (vs. aligned) with the participant’s own position, these advantages attenuated or reversed. These results shed light on ways that heeding scientific evidence can impact how politicians are viewed.

Keywords

political psychology impression formation belief change partisanship

Statement of Relevance

Throughout the COVID-19 pandemic, former Chief Medical Advisor Dr. Anthony Fauci provided health and safety recommendations to the public, often updating advice to reflect the newest scientific evidence. Many praised Fauci’s guidance, with Time Magazine naming him 2020s Guardian of the Year (Kluger & Park, 2020). Others, however, dubbed him “Flip-flop Fauci” and “notable liar” suggesting that inconsistency undermined his credibility (Bruce, 2021). These two responses raise an important question: when are political figures rewarded, and when are they punished, for changing their position?

The current results suggest that the answer is not straightforward. Overall, politicians who changed their stance due to evidence were perceived more positively than those who didn’t. Nevertheless, politicians weren’t punished for ignoring evidence if it meant staying on the participant’s side. These findings raise questions about the role of partisanship compared to empiricism in shaping evaluations.

People are often reluctant to publicly admit when they might be wrong. Several factors contribute to this aversion to changing one’s position, including a perceived risk of being seen as less intelligent or less trustworthy. In fact, Fetterman and Sassenberg (2015) write that the main reason people are afraid to publicly change their views is fear of being viewed as incompetent. Politicians, in particular, may fear that such perceptions could translate to fewer votes. Here, we will test whether such fears are justified, examining how people respond to politicians who change their positions in response to high-quality scientific evidence.

What Does It Mean to Admit Wrongness?

We begin with a definition of the central phenomenon of interest: admitting wrongness. We borrow Fetterman et al.’s (2019) definition of wrongness admission: “… a public disclosure that one has been wrong about a belief or attitude and has subsequently changed that attitude or belief” (p. 193). A key feature of this definition, and what distinguishes it from an apology, is that there is no emotional guilt or shame expressed. In an apology, there is remorse shown (e.g., “I’m sorry for…”) suggesting that a morally wrong behavior has transpired. Stated in another way: admitting wrongness means publicly saying that one held an incorrect belief and is shifting to a belief that is more justified (e.g., Schumann, 2018).

Warmth and Competence

The current investigation explores what impressions are made when a politician changes their mind and thus, is informed by work on impression formation. Asch (1946) was one of the first researchers to investigate how people form impressions and found that people generally do so along two general dimensions: warmth and competence. Social psychologists have several names for this general idea, such as instrumentality and expressiveness, or superhero and saint traits (Parsons & Bales, 1955; Paulhus & John, 1998). We adopt the labels “warmth” and “competence” used in the Stereotype Content Model (Fiske et al., 2002, 2007). Abele and Wojciszke’s (2013) investigation demonstrated that the alternative labels capture very similar constructs. Warmth-based adjectives tend to be related to relationship maintenance and social functioning (e.g., kindness and likeability) and competence-based adjectives relate to task functioning and goal achievement (e.g., agency and intelligence). Warmth is related to the judgments of a person’s intent (e.g., to block vs. support another person’s goal), while competence is related to the evaluation of their capacity to act on those intentions.

Evidence That Changing One’s Mind Can Hurt Perceptions of Warmth and Competence

People are often reluctant to admit being wrong, particularly in public contexts, due to concerns about reputational costs (Fetterman & Sassenberg, 2015; Gilovich et al., 2002) . It has been argued that one key reason that people are reluctant to publicly change their mind is the fear of being viewed as incompetent (e.g., Fetterman & Sassenberg, 2015). Empirical evidence provides some support for the idea that people who change their positions are seen as less competent. For example, in some cases, people who changed their stance were perceived as less intelligent and more hypocritical, particularly for moral positions (Braver et al., 1977; Cialdini et al., 1974; Laurent & Clark, 2019). In line with these results, a meta-analysis showed that those who changed their moral attitudes were consistently labeled as ineffective leaders and hypocrites (Kreps et al., 2017).

Evidence That Changing One’s Mind Can Boost Perceptions of Warmth and Competence

In contrast, other work paints a more positive picture. People—including researchers themselves—largely overestimate the reputational consequences of changing their views (Chancellor & Lyubomirsky, 2013; Fetterman et al., 2019; Gilovich et al., 2002). Not only are negative evaluations overestimated, but positive social consequences are underestimated (Fetterman & Sassenberg, 2015). Individuals in varied contexts—strangers who argue on social media, professors at universities, and workplace supervisors—are rated as possessing greater warmth and competence after publicly changing their positions (e.g., Fetterman et al., 2022). Reputational benefits, such as increased trust and perceived effectiveness have also been observed when individuals are transparent about their thought process, sharing the reason they changed their view or being precise about their new stance (e.g., Norman et al., 2010). Recent findings suggest that entrepreneurs and researchers may even face negative reputational consequences if they refuse to change their original view despite high quality evidence (e.g., Altenmüller et al., 2021; John et al., 2019).

What If We Changed Our Minds With Evidence?

Perhaps one factor that influences how a person is perceived when they change their mind is whether they are responding to high quality evidence. Scant research has been conducted on this specific topic, but contemporary work may help serve as a guide. John et al. (2019) found that there are negative reputational consequences when entrepreneurs refuse to change their original belief despite contradictory evidence. In addition, work by Altenmüller et al. (2021) suggests that researchers who self-correct based on evidence are deemed more trustworthy and credible than those who do not.

In sum, there is initial evidence that both entrepreneurs and scientists are rewarded when changing their mind based on evidence and are punished when they do not do so. But what of the political realm? Politicians who claim to have changed their mind are believed to suffer harsh reputational consequences and are accused of “flip-flopping” (also referred to as “waffling” (e.g., Carlson & Dolan, 1985; Hoffman & Carver, 1984)). The current investigation seeks to understand if politicians who take a new position on an issue (i.e., admit wrongness), in response to high quality evidence, are still evaluated harshly (i.e., as incompetent and cold).

Politicians Who Change Their Position

Politicians have a vested interest in being evaluated positively by their constituents in a democratic government because of the possibility of being punished in the voting booth. They are often advised to maintain their political position and flip-flopping is thought to be a sign of hypocrisy (e.g., Agranov, 2016; Andreottola, 2021). Sigelman and Sigelman (1986) explored public responses to U.S. presidents who acted “out of character.” They provided vignettes of presidents who were originally peace-oriented or war-oriented followed by their most recent actions which were either in line with their original platform or out of character. Their investigation had mixed results, not clearly demonstrating a consistent pattern in how people evaluate politicians who change their mind. Stated simply, they were puzzled that their participants did not clearly punish those who flip-flopped.

Tomz and Van Houweling (2008) conducted a review of political science theories on how a candidate’s position predicts people’s voting intentions. Specifically, there are three types of theories: proximity theory, discounting theory, and directional theory. Proximity theory predicts that people vote rationally (e.g., Downs, 1957; Grofman, 2004). Specifically, if a voter supports a certain position (e.g., 15% tax reduction), they should have higher voting intentions for people who hold a view closer to versus further from that position (e.g., 20% vs. 30%). Discounting theory predicts that voting intentions usually ignore, or discount, the idealistic campaign promises candidates make during elections. Instead, voters focus on what actions can concretely be accomplished given the current government administration. This theory would predict a voter would be more likely to vote for someone who can realistically achieve their goals rather than someone more idealistic, even when the latter is more aligned with the voter’s views. Lastly, directional theory describes voters as taking a dichotomous approach to issues, focusing on “my side” versus “their side.” Directional theory predicts that candidates are more likely to receive votes from people who perceive themselves and the candidate to be on the same side (see Rabinowitz & Macdonald, 1989). If there are two candidates on the voter’s side, the voter will generally prefer the politician holding the more extreme view.

Although there is some psychological research to suggest that politicians who are willing to admit wrongness face little social cost (Leary et al., 2017), political scientists have suggested that politicians are indeed punished for this behavior (Tomz & Van Houweling, 2010). Inconsistent politicians were operationalized via vignettes that described a politician who voted with one group 2 years ago but voted with a different group at the time of the survey (e.g.., voted liberal 2 years ago, voted moderate or conservative most recently). Participants reported lower intentions to vote for inconsistent politicians compared to consistent ones and they judged those who repositioned as worse leaders and more dishonest, but also more open-minded. Thus, the evidence suggests repositioning is mostly punished, while having a very slight benefit.

Political scientists have also investigated factors that influence when political repositioning is costly versus beneficial to a political candidate. Doherty et al. (2016) explored several of these factors, including the passage of time and the alignment between the politician’s belief and that of the voter. The researchers found that when less time has passed since the repositioning occurred, there was a more pronounced negative impact on voter evaluations. The researchers also present evidence that when a politician adopts a final political position that aligns with the participant’s personal belief, a positive impact on evaluations occurs that outweighs the negative of repositioning without justification. This alignment effect has been noted in other recent work (e.g., Croco, 2016). The current study is, to our knowledge, the first to test how people evaluate politicians who change their mind when doing so is a response to high-quality scientific evidence.

Real-World Examples of Politicians Accused of Flopping

When investigating politically relevant data, some researchers have argued for a bottom-up approach that considers actual language used in the political sphere (Dunmire, 2012). On both sides of the political aisle the “flip-flopper” label is quickly, and frequently, thrown around by opponents. This accusation became so prevalent that in 2008 the fact checking website PolitiFact, created a rating system dubbed the “flip-o-meter” (Adair, 2008). This flip-o meter labels a politician’s consistency on a position with one of three ratings: “no flip” (no significant change in a position), a “half-flip” (a partial change in position), and a “full flop” (a complete change in position). People are often skeptical and critical when politicians change their beliefs.

For example, former United States Secretary of State Hillary Clinton was said to have committed a full flop when she changed her position on same-sex marriage during the 2016 presidential campaign (Sherman, 2015). Although Clinton explained that her repositioning was due to receiving new information, the public was critical and suspicious of the motivation behind her newfound beliefs (e.g., Shafer, 2015). Similarly, with the Republican President Donald J. Trump receiving a full flop when he stated he was not considering pausing tariff as on options on April 7th, 2025, only to do so on April 9th (Sherman, 2025). Full-flop examples occur across the political aisle, and accusations of flip-flopping can encroach upon the realm of science.

An illustrative example can come from the recent global pandemic. Throughout the COVID-19 pandemic, former Chief Medical Advisor Dr. Anthony Fauci provided health and safety recommendations to the public, often updating advice to reflect the newest scientific evidence. Many praised Fauci’s guidance, with Time Magazine naming him 2020s Guardian of the Year (Kluger & Park, 2020). Others, however, dubbed him “Flip-flop Fauci” and “notable liar” suggesting that inconsistency undermined his credibility (Bruce, 2021). These accusations of being a flip-flopper have persisted years after the pandemic, with Dr. Fauci justifying his decisions by highlighting the uncertainty built into the nature of science:

As things evolve and change and you get more information, it is important that you use the scientific process to gain that information and perhaps change the way you think of things, change your guidelines and change your recommendation (Shutt, 2024).

The Current Study

This study sought to examine how politicians would be evaluated after publicly changing their position in response to high-quality scientific evidence. Participants read vignettes about politicians who learned about new evidence on a variety of issues (e.g., the death penalty, euthanasia). We examined three research questions (RQs) and tested four hypotheses (Hs):

Pre-Registered Research Questions and Hypotheses

RQ1.

Are politicians judged more favorably when they change (vs. maintain) their public stance on a political issue in response to strong evidence?

H1.

We predicted that politicians who changed (vs. maintained) their public stance would i) be rated as more “flip-floppy”; ii) be rated as more competent; iii) be rated as warmer; and iv) elicit higher voting intentions.

RQ2.

Are politicians judged more favorably if their final position on a political issue is aligned (vs. misaligned) with the participant’s own position?

H2.

We predicted that politicians who ultimately aligned (vs. misaligned) with the participant’s own view would (i) be rated as less “flip-floppy”; (ii) be rated as more competent; (iii) be rated as warmer; and (iv) elicit greater voting intentions.

RQ3.

Does the main effect of a politician changing their public stance (H1) depend on whether the politician’s final position aligns with the participant’s own position?

H3a.

We predicted the expected effect of politician action would be (i) smaller in size for flip-floppiness, but (ii) larger for competence, (iii) warmth, and (iv) voting intention ratings when the politician’s final position was aligned (vs. misaligned) with the participant’s own position.

H3b.

We predicted a simple effect such that politicians in the Maintain and Align conditions (vs. Change and Misalign conditions) would (i) be rated as less “flip-floppy”; (ii) be rated as more competent; (iii) be rated as warmer; and (iv) elicit higher voting intentions.

Open Science Statement

The study design, data collection target, exclusion criteria, hypotheses, and analyses were pre-registered on AsPredicted. The pre-registration, data, primary materials, Supplemental Materials, and SPSS syntax used to run the analyses are posted on the Open Science Framework (https://osf.io/a3b47/?view_only=1d850a3043064272803eea54a0b5333a).

Method

Design

The study utilized a 2 (politician’s action: maintain position vs. change position) × 3 (final position alignment: aligned vs. misaligned vs. control) within-subjects design.

Vignette Details

We used two resources to choose topics about which participants would likely have strong opinions. The first was a report identifying issues with large partisan gaps (Pew Research Center, 2019). According to this report, the following were highly divisive topics: gun regulation, the climate and environment, immigration, and the ideal role of government. The second resource was a previous study conducted using a convenience sample of the human subject pool at the University of Alabama (UA)—the same subject pool used here (Whitt 2022). In this study, UA students generally expressed strong views about physician assisted suicide, the death penalty, and abortion.

Drawing from both these sources, we compiled a list of eight polarizing topics: (i) abortion, (ii) gun regulation, (iii) the death penalty, (iv) affirmative action, (v) loosening U.S. border regulations, (vi) physician assisted suicide, (vii) pursuing renewable energy, and (viii) government financial assistance programs. In parallel, we created a series of “control” vignettes featuring topics that were not politically charged: (i) controlling the pheasant population of a made-up county, (ii) regulating “Stalacium” (a made-up building material), (iii) insuring F9-Houses (a non-existent house type), and (iv) extending term lengths for an educational board in a made-up county. These vignettes allowed us to see how participants responded when politicians changed their stance on issues in which participants were unlikely to be invested.

All vignettes described a politician who ran on a policy position and new evidence relevant to that position. Participants were told that the evidence came from high-quality research conducted by scientists from both sides of the political aisle. The results of the research were always contrary to the politician’s initial stated position.

As an example, one vignette described a politician who ran on the platform of stricter immigration policy, believing that loosening border regulations would increase crime. The politician then responds to research that shows no relationship between crime rates and border policy. In the “maintain position” version, the politician acknowledges the study but maintains their original position. In the “change position” version, the politician changes their position on the relationship between border policy and crime rates. Importantly, the change in position was limited to the specific point addressed by the evidence. In this way, vignettes describe small position changes calibrated to the evidence, rather than dramatic changes, like a complete reversal on border policy, that might seem abrupt or unjustified.

For the polarizing vignettes, topic was counterbalanced with stance (pro vs. anti), politician action (change position vs. maintain position), and politician gender (male vs. female) resulting in a total of 64 (8 × 2 × 2 × 2) vignettes. For the control vignettes, topic was counterbalanced with politician action and politician gender, resulting in a total of 16 (4 × 2 × 2) vignettes. Stance was not manipulated for control vignettes as they were designed to be neutral. Manipulating politician gender was done to improve generalizability and allow us to explore gender effects.

Procedure

After providing informed consent, participants were asked to indicate their attitudes about the 12 topics featured in the vignettes using a 7-point Likert style scale (1 = strongly disagree, 4 = neither agree nor disagree, 7 = strongly agree). Based on their responses they were presented with six randomly ordered vignettes, with two addressing control topics on which they reported neutral stances (i.e., a response of 4), and four addressing polarizing topics on which they reported strong stances (i.e., responses of 1, 2, 6, or 7). Three of the vignettes featured a politician changing their position, while the other three featured a politician maintaining their stance. In addition, three vignettes featured a female politician while the other three featured a male politician. Finally, two of the polarizing vignettes featured a politician who ended up aligned with the participant’s views, whereas the other two polarizing vignettes featured a politician who ended up misaligned with the participant’s views.

For each politician, participants provided ratings of (i) perceived flip-floppiness; (ii) perceived warmth; (iii) perceived competence; and (iv) the participant’s voting intentions. For exploratory purposes, participants were asked to indicate the trustworthiness and believability of the studies in the vignettes and to complete measures of intellectual humility, willingness to admit wrongness, affective political polarization, and demographics.

Power Analysis

We used “Superpower” (Lakens & Caldwell, 2021) to conduct a power analysis based on the smallest cell differences of interest in accordance with the previously stated hypotheses (see Supplemental Materials for details). This analysis revealed that a sample of 1,200 would provide at 80% power to detect the predicted interaction effect (H3).

Participant Information and Exclusions

Participants

Participants were recruited from two independent sources. The first source was the human subjects research pool at the University of Alabama (UA), through which we compensated undergraduate students with course credit (n = 2,398). The second source was the data collection platform Prolific, through which we compensated Prolific users at a rate of $8/hour (n = 98). All participants had access to a Qualtrics survey link and were provided with a brief description of the study before choosing to participate.

Exclusions

After pre-registered exclusion criteria were applied, N = 1,213 participants were included in analyses (50.6% retention; see Supplemental Materials for details).

Survey Measures

The order of measures, and items within measures, was randomized, with the exception that the attention checks were always at the end of each vignette block and demographic questions were always at the end of the whole survey. After completing all measures, participants were shown a page with debriefing information and given the option to withdraw their data.

Primary Dependent Measures

For multi-item measures, we report McDonald’s Omega (ω) as a measure of inter-item reliability (Hayes & Coutts, 2020; McNeish, 2018). The four main outcomes described in the following section were assessed for each of the six vignettes.

Perceived Flip-Floppiness

Participants rated each politician using a single item: “How flip-floppy do you believe [Politician’s name] to be?” (1 = not at all flip-floppy, 4 = unsure, 7 = very flip-floppy).

Measuring Judgments of Competence and Warmth

Descriptors for competence and warmth were chosen based on the recommendations of Halkias and Diamantopoulos (2020).

Perceived Competence

Participants rated each politician on the following four descriptors: (i) capable, (ii) competent, iii) efficient, and iv) skillful (1 = not at all [descriptor], 4 = unsure, 7 = very [descriptor]). These items were averaged to create a composite “Competence” score, McDonald’s ω = 0.92.

Perceived Warmth

Participants rated each politician on the following four descriptors: (i) friendly, (ii) kind, (iii) likeable, and (iv) nice (1 = not at all [descriptor], 4 = unsure, 7 = very [descriptor]). These items were averaged to create a composite “Warmth” score, McDonald’s ω = 0.92

Voting Intentions

To measure voting intentions, participants responded to a single item: “How likely would you be to vote for [Politician’s Name?]” (1 = very unlikely, 7 = very likely).

Results

A series of factorial within-subjects ANOVAs were performed to investigate the main effects of politician action (RQ1), position alignment (RQ2), and their interaction (RQ3) on (i) perceived flip-floppiness, (ii) perceived competence, (iii) perceived warmth, and (iv) participants’ intentions to vote for the politician.

RQ1: Main Effect of Politician’s Action

Politicians who changed (vs. maintained) their positions were rated as more flip-floppy, more competent, and warmer. Furthermore, participants indicated weaker intentions to vote for politicians who changed (vs. maintained) their positions (Table 1). These results support Hypothesis 1.

Table 1.

Main Effect of Politician Action (RQ1): Means, Standard Deviations, and Repeated Measures Analysis of Variance for Primary Outcomes.

Primary outcome	Maintain stance		Change stance		ANOVA(N = 1,213)
Primary outcome	M	SD	M	SD	Hedge’s g [95% CI]	F(1, 1,212)	η_p²
Perceived flip-floppiness	2.20	1.22	4.46	1.41	1.72 [1.61, 1.83]	1,693	.59
Perceived competence	3.98	0.92	4.41	0.90	0.47 [0.39, 0.56]	50.6	.04
Perceived warmth	3.94	0.7	4.44	1.55	0.65 [0.58, 0.72]	183.3	.13
Intentions to Vote	3.37	1.00	3.05	0.78	0.46 [0.38, 0.54]	34.6	.03

Note. All outcomes range from 1 to 7.

RQ2: Main Effect of Position Alignment

Politicians whose final position aligned (vs. misaligned) with that of the participant were rated as less flip-floppy, more competent, and warmer. Furthermore, participants indicated stronger intentions to vote for politicians whose final position aligned (vs. misaligned) with their own (Table 2). These results support Hypothesis 2.

Table 2.

Main Effect of Final Position Alignment (RQ2): Means, Standard Deviations, and Repeated Measures Analysis of Variance for Primary Outcomes.

Primary outcome	Aligned final position		Misaligned final position		ANOVA (N = 1,213)
Primary outcome	M	SD	M	SD	Hedge’s g [95% CI]	F(1, 1,212)	η_p²
Perceived flip-floppiness	3.22	1.15	3.46	1.14	0.21 [0.15, 0.27]	46.5	.04
Perceived competence	4.55	0.86	3.83	0.90	0.82 [0.74, 0.90]	531.6	.31
Perceived warmth	4.45	0.78	3.90	0.85	0.68 [0.60, 0.75]	384.9	.24
Intentions to vote	4.34	1.17	2.89	1.13	01.25 [1.17, 1.35]	1,022.9	.46

Note. All outcomes range from 1 to 7.

RQ3a: Interaction Between Politician Action and Position Alignment

H3ai: Investigating Perceived Flip-Floppiness

There was a modest but statistically significant interaction between politician action and final position alignment predicting flip-floppiness, F(1, 1,212) = 3.86, p = .05, η_p² = .003 (Figure 1). To examine this effect, planned contrasts were conducted. The average flip-floppiness rating for politicians was higher when they changed their stance and were aligned (M = 4.41, SD = 1.78) in contrast to politicians who maintained their stance and were aligned (M = 2.03, SD = 1.47), t(1,212) = 35.7, p < .001, Hedge’s g = 1.46, 95% CI [1.36, 1.56]. The average flip-floppiness rating for politicians was also higher when they change their stance and were misaligned (M = 4.72, SD = 1.76) in contrast to politicians who maintained their stance and were misaligned (M = 2.20, SD = 1.60), t(1,212) = 35.6, p < .001, Hedge’s g = 1.50, 95% CI [1.40, 1.60]. These results did not support Hypothesis 3ai.

Figure 1.

Ratings of primary outcomes for politicians who maintain (vs. change) their position, and whose final position align (vs. misalign) with those of the participant.

H3aii: Investigating Perceived Competence

There was a statistically significant interaction between politician action and position alignment predicting competence, F(1, 1,212) = 230.87, p < .001, η_p² = .16. To examine this effect, planned contrasts were conducted. The average competence rating for politicians were lower when they changed their stance and were aligned (M = 4.48, SD = 1.23) compared to politicians who maintained their stance and were aligned (M = 4.61, SD = 1.19), t(1,212) = 2.6, p = .008, Hedge’s g = 0.11 95% CI [0.03, 0.19]. The average competence rating for politicians was higher when they changed their stance and were misaligned (M = 4.19, SD = 1.22) compared to politicians who maintained their stance and were misaligned (M = 3.46, SD = 1.34); thus, t(1,212) = 13.9, p < .001, Hedge’s g = 0.57 95% CI [0.49, 0.65]. These results did not support Hypothesis 3aii.

H3aiii: Investigating Perceived Warmth

There was a statistically significant interaction between politician action and position alignment predicting warmth, F(1, 1,212) = 243.25, p < .001, η_p² = .17. To examine this effect, planned contrasts were conducted. The average warmth rating for politicians who changed their public stance and were aligned (M = 4.45, SD = 1.04) was not significantly different from that of politicians who maintained their public stance and were aligned (M = 4.46, SD = 1.05), t(1,212) = 0.4, p = .695, Hedge’s g = −0.02, 95% CI [−0.09, 0.06]. The average warmth ratings were higher for politicians changed their public stance and were misaligned (M = 4.32, SD = 1.03), compared to ratings of politicians who maintained their stance and were misaligned (M = 3.48, SD = 1.21), t(1,212) = 20.0, p < .001, Hedge’s g = 0.75, 95% CI [0.67, 0.83]. These results did not support Hypothesis 3aiii.

H3aiv: Investigating Participants’ Intentions to Vote for Politician

There was a statistically significant interaction between politician action and stance alignment predicting intentions to vote, F(1, 1,212) = 459.39, p < .001, η_p² = .28. To examine this effect, planned contrasts were conducted. The intention to vote for politicians was weaker when they were changed their stance and were aligned (M = 4.05, SD = 1.71) compared to politicians who maintained their stance and were aligned (M = 4.63, SD = 1.59); t(1,212) = 8.8, p < .001, Hedge’s g = −0.36, 95% CI [−0.44, −0.28]. The intention to vote for politicians was stronger when they changed their stance and were aligned (M = 3.47, SD = 1.65) compared to politicians who maintained their stance and were aligned (M = 2.31, SD = 1.46), t(1,212) = 18.9, p < .001, Hedge’s g = 0.75, 95% CI [0.67, 0.83]. These results did not support Hypothesis 3aiv.

RQ3b: Do We Prefer Those Who Align With Our Views and Who Ignored Scientific Evidence or Those Who Listened to Scientific Evidence But Now Misalign With Us?

Planned contrasts were conducted on ratings of politicians who ended up aligned with the participants’ own view while maintaining their public stance versus politicians who ended up misaligned with the participant after considering scientific evidence (i.e., maintain and align vs. change and misalign). Politicians in the maintain and align condition (vs. the change and misalign condition) were rated as less flip-floppy, more competent, and warmer. Furthermore, participants indicated stronger intentions to vote for politicians in the maintain and align condition (vs. the change and misalign condition; Table 3). These results support Hypothesis 3b.

Table 3.

Means, Standard Deviations, and Repeated Measures Analysis of Variance on Primary Vignette Outcomes Investigating the Interaction Effect of Action x Alignment.

Primary outcome	Maintain and align*		Maintain and misalign		Change and align		Change and misalign*		ANOVA (N = 1,213)
Primary outcome	M	SD	M	SD	M	SD	M	SD	F(1, 1,212)	η_p²
Perceived flip-floppiness	2.03	1.47	2.20	1.60	4.41	1.78	4.72	1.76	3.9	.003
Perceived competence	4.61	1.19	3.46	1.34	4.48	1.22	4.19	1.22	230.9	.16
Perceived warmth	4.46	1.05	3.48	1.21	4.45	1.04	4.32	1.03	243.2	.17
Intentions to vote	4.63	1.59	2.31	4.46	4.05	1.71	3.47	1.65	459.4	.28

Indicates two conditions that planned contrasts were conducted on (H3b). All outcomes range from 1 to 7.

Discussion

Within social psychology, Lovakov and Agadullina (2021) have provided empirically grounded benchmarks for small (g = 0.15), medium (g = 0.36) and large (g = 0.65) effects. A small social psychological effect has the same magnitude as the relationship between low-level lead exposure and childhood IQ. A medium effect is similar in size to the association between positive parenting behavior and child behavior problems. Finally, a large effect is similar in size to social conformity in the Asch line judgment task (Asch, 1946; Meyer et al., 2001). These may serve as a helpful reference for interpreting the effects observed in the current study.

Changing Your Position Means You’re Flip-Flopping

Will politicians be seen as flip-floppers if they publicly change their position? The data supports this with a large effect size (Hedge’s g = 1.72). However, this doesn’t seem to mean that politicians who change their stance will face universally negative consequences. Although politicians were perceived as more flip-floppy when they changed (vs. maintained) their public position, they were also seen as more competent (Hedge’s g = 0.47), warmer (Hedge’s g = 0.65), and elicited higher voting intentions (Hedge’s g = 0.46). Nevertheless, our data do not show that politicians will always be better off changing their view in line with scientific evidence.

Punishing “Them,” Not “Us”

Our results show that when a politician ends up disagreeing with the participant, they can garner more positive impressions when they change their position. When a politician ends up agreeing with the participant, ignoring evidence works in their favor. One possible interpretation is to assume that, in most cases, politicians who start out aligned with the participant are seen as sharing the same party affiliation (e.g., both the politician and the participant are Democrats). Viewed in this way, our data suggest that people evaluate party-members similarly, regardless of whether they change their policy stance. However, people evaluate non-party members harshly when they ignore evidence, but highly when they heed evidence that pushes them closer to the participants’ own views. Finally, when pitting effects of alignment and the politician’s action against one another, data suggests the benefit of alignment appears to overshadow the cost of ignoring evidence; politicians in the maintain and align condition received better ratings for competence, warmth, and voting intentions than those in the change and misalign condition. This aligns with research showing increases in affective polarization over the last 30 years (Westfall et al., 2015).

Implications of the Present Study

Theoretical Implications

When we began this work, two bodies of research suggested different outcomes for politicians who change their position. On the one hand, political scientists provided evidence that politicians who change their stance face negative consequences. On the other hand, research on position change showed that changing one’s policy stance may elicit positive evaluations. Our results are consistent with this first body of work in that politicians who changed their position were viewed as more flip-floppy. Examined holistically, however, our data seems more aligned with the second body of work.

We believe this study advances the research on public perception and decision making in several ways that can inform researchers. Previous work suggests that political flip-flopping is universally damaging to credibility (e.g., Doherty et al., 2016; Tomz & Van Houweling, 2010). Our findings challenge this overly simplistic view with data suggesting that politicians who change their stance due to strong scientific evidence are sometimes perceived as more competent and warmer—a finding that is dependent on an individual’s own political views. A second theoretical contribution of our study is it underscoring the influence of partisanship in decision making (e.g., intentions on who to vote for). Our study finds evidence of motivated reasoning in perceptions of political figures. This provides empirical support for Directional Theory (Rabinowitz & Macdonald, 1989), and specifically adds to the literature showing that people overlook “scientific evidence” in favor of partisanship (i.e., are they on my side).

An avenue for future research is further investigation of how participants’ own views on flip flopping are affected by politically motivated reasoning. Although our data cannot adequately address this question, researchers can build on work by scholars such as science historians who provide real-world examples of when scientists interact with political institutions. One recommended starting point would be work done by Oreskes and Conway, 2011, who educate readers about how some politicians/corporations discourage responsiveness to scientific evidence (by fostering doubt) in order to fuel their own personal interests (e.g., in fossil fuels, cigarettes).

Real-World Implications

Our results challenge the simplistic notion that politicians should never admit they are wrong. The results suggest that changing one’s position when it is clearly justified elicits high perceptions of competence and warmth, even from voters who disagree with your position. All in all, changing one’s policy position in a justified way seems to have a positive impact, despite being accompanied by the “flip-flopper” label.

Our results also suggest voters are not particularly bothered when politicians ignore evidence provided the politician ends up agreeing with them. For many political issues, we would expect empirical data to inform solutions, and would therefore expect politicians to pay attention to those data. Unfortunately, our results suggest that people may be reluctant to punish candidates on their side. The implications are concerning if political tribalism wins out where empiricism should lead the way.

Our findings may provide practical insights to policy makers for political strategy and policy framing. Much work has been done on how politicians can use rhetorical strategies to persuade voters, and our study uniquely adds to that body of literature by examining how science justification can increase perceptions of competence and warmth—if only when voters agree with the conclusion. This matters for democratic decision making as policy makers must uniquely balance empirical evidence and weigh it against the cost of their constituents’ perceptions on the issue.

Simply citing “scientific evidence” does not guarantee increased support, particularly on controversial topics such as climate change or gun regulation. One critical factor in determining the public’s response may be how the evidence is communicated. Transparent explanations of why a position has changed, framing it as a process of learning or evidence-based decision making, may foster favorable perceptions (Altenmüller et al., 2021; Norman et al., 2010), especially among voters already inclined toward that conclusion (Fetterman et al., 2022). Policy makers can draw on bipartisan efforts and involve trusted community voices to reduce the perception of partisanship (Stekelenburg et al., 2022). Further, efforts to improve the public’s understanding of the scientific evidence (e.g., accessible summaries of cited key findings) may help citizens interpret policy shifts as a strength rather than inconsistency (Oreskes & Conway, 2011; Rudman et al., 2013).

Limitations

To Whom do the Results Generalize? These data likely generalize best to those most highly represented in our sample: young, white, female college students who live within the U.S. and are politically unengaged. A strength of the sample is its political diversity, so results likely speak to both Republicans and Democrats, from those who identify as very liberal to those who identify as very conservative across different domains (general, economic, social). The results may also generalize to white American college students as a whole. Finally, the results may bring more limited insight to populations of Americans who are non-white, older, do not pursue a higher education, or are politically active. Results may also bring some limited insight to non-U.S. samples.

Vignettes: High Control, Low Ecological Validity

One critique of our approach is that we assume participants can accurately imagine how they would respond to hypothetical vignettes. Perhaps participants believe they would commend a politician who changes their position, but when it occurs, they underestimate the negative feelings experienced. Additionally, this study assessed voting intentions not voting behavior. Many Americans may intend to vote but not follow through. The current study can provide understanding of the intention, not the action.

Conclusion

Our results provide optimism that when politicians listen to scientific evidence this can lead to reputational benefits. Nevertheless, they also revealed a potential concern: participants seemed unbothered by politicians who ignored scientific evidence and stayed on the participant’s “side.” When there is a strong scientific body of evidence informing a political issue, it seems critical to reflect on, and update, our beliefs as necessary. Given that we all hold wrong beliefs, it seems important to listen to strong evidence, reward those who do the same, and hold accountable those who do not, across party lines.

Supplemental Material

sj-docx-1-sgo-10.1177_21582440251377997 – Supplemental material for The Perks of Being a Flip-Flopper: Examining Evaluations of Politicians Who Change Their Position Due to Scientific Evidence

Supplemental material, sj-docx-1-sgo-10.1177_21582440251377997 for The Perks of Being a Flip-Flopper: Examining Evaluations of Politicians Who Change Their Position Due to Scientific Evidence by Jacob F. Miranda, Alexa M. Tullett and Joshua A. Baker in SAGE Open

Footnotes

ORCID iDs

Jacob F. Miranda

Alexa M. Tullett

Joshua A. Baker

Ethical Considerations

The study received ethical approval from The University of Alabama’s IRB (approval #22-07-5787) on August 1st, 2022. All procedures were in accordance with the ethical standards of the institutional review board and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Consent to Participate

Informed consent was obtained from all individual participants included in the study.

Author Contributions

JFM: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Writing—original draft, Writing—review & editing, Supervision, Project administration, Funding acquisition. AMT: Conceptualization, Methodology, Resources, Supervision, Writing—original draft, Writing—review & editing. JAB: Software, Validation, Investigation, Data curation, Writing—original draft, Writing—review & editing.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported financially by The University of Alabama’s Graduate School and Psychology Department.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The study design, data collection target, exclusion criteria, hypotheses, and analyses were pre-registered on AsPredicted. The pre-registration, data, primary materials, supplemental materials, and SPSS syntax used to run the analyses are posted on the Open Science Framework ().

Supplemental Material

Supplemental material for this article is available online.

References

Abele

A. E.

Wojciszke

(2013). The Big Two in social judgment and behavior. Social Psychology, 44(2), 61–62. https://doi.org/10.1027/1864-9335/a000137

Adair

(2008). Introducing the flip-o-meter. Politifact. https://www.politifact.com/article/2008/aug/05/introducing-flip-o-meter/

Agranov

(2016). Flip-flopping, primary visibility, and the selection of candidates. American Economic Journal Microeconomics, 8(2), 61–85. https://doi.org/10.1257/mic.20130036

Altenmüller

M. S.

Nuding

Gollwitzer

(2021). No harm in being self-corrective: Self-criticism and reform intentions increase researchers’ epistemic trustworthiness and credibility in the eyes of the public. Public Understanding of Science, 30(8), 962–976. https://doi.org/10.1177/09636625211022181

Andreottola

(2021). Flip-flopping and electoral concerns. Journal of Politics, 83(4), 1669–1680. https://doi.org/10.1086/715070

Asch

S. E.

(1946). Forming impressions of personality. Journal of Abnormal and Social Psychology, 41, 258–290.

Braver

S. L.

Linder

D. E.

Corwin

T. T.

Cialdini

R. B.

(1977). Some conditions that affect admissions of attitude change. Journal of Experimental Social Psychology, 13(6), 565–576. https://doi.org/10.1016/0022-1031(77)90055-5

Bruce

(2021, October 6). Flip-flop Fauci: The purveyor of doom and gloom. The Washington Times. https://www.washingtontimes.com/news/2021/oct/6/flip-flop-fauci-the-purveyor-of-doom-and-gloom/

Carlson

J. M.

Dolan

(1985). The waffle phenomenon and candidates’ image. Psychological Reports, 57(3), 795–798.

10.

Chancellor

Lyubomirsky

(2013). Humble beginnings: Current trends, state perspectives, and hallmarks of humility. Social and Personality Psychology Compass, 7(11), 819–833. https://doi.org/10.1111/spc3.12069

11.

Cialdini

R. B.

Braver

S. L.

Lewis

S. K.

(1974). Attributional bias and the easily persuaded other. Journal of Personality and Social Psychology, 30(5), 631–637. https://doi.org/10.1037/h0037448

12.

Croco

S. E.

(2016). The flipside of flip-flopping: Leader inconsistency, citizen preferences, and the war in Iraq. Foreign Policy Analysis, 12(3), 237–257. https://doi.org/10.1093/fpa/orw006

13.

Doherty

Dowling

C. M.

Miller

M. G.

(2016). When is changing policy positions costly for politicians? Experimental evidence. Political Behavior, 38, 455–484. https://doi.org/10.1007/s11109-015-9321-9

14.

Downs

(1957). An economic theory of democracy. Harper and Row.

15.

Dunmire

P. L.

(2012). Political discourse analysis: Exploring the language of politics and the politics of language. Language and Linguistics Compass, 6(11), 735–751. https://doi.org/10.1002/lnc3.365

16.

Fetterman

A. K.

Curtis

Carre

Sassenberg

(2019). On the willingness to admit wrongness: Validation of a new measure and an exploration of its correlates. Personality and Individual Differences, 138, 193–202. https://doi.org/10.1016/j.paid.2018.10.002

17.

Fetterman

A. K.

Muscanell

Covarrubias

J. J.

Sassenberg

(2022). When you are wrong, just admit it: Wrongness admission leads to better interpersonal impressions on social media. Social Psychology, 53(1), 34–55. https://doi.org/10.1027/1864-9335/a000473

18.

Fetterman

A. K.

Sassenberg

(2015). The reputational consequences of failed replications and wrongness admission among scientists. PLoS One, 10(12), e0143723. https://doi.org/10.1371/journal.pone.0143723

19.

Fiske

S. T.

Cuddy

A. J.

Glick

(2007). Universal dimensions of social cognition: Warmth and competence. Trends in Cognitive Sciences, 11, 77–83. https://doi.org/10.1016/j.tics.2006.11.005

20.

Fiske

S. T.

Cuddy

A. J. C.

Glick

(2002). A model of (often mixed) stereotype content: Competence and warmth respectively follow from perceived status and competition. Journal of Personality and Social Psychology, 82, 878–902. https://doi.org/10.1037/0022-3514.82.6.878

21.

Gilovich

Kruger

Medvec

V. H.

(2002). The spotlight effect revisited: Overestimating the manifest variability of our actions and appearance. Journal of Experimental Social Psychology, 38(1), 93–99. https://doi.org/10.1006/jesp.2001.1490

22.

Grofman

(2004). Downs and two-party convergence. Annual Review of Political Science, 7, 25–46.

23.

Halkias

Diamantopoulos

(2020). Universal dimensions of individuals’ perception: Revisiting the operationalization of warmth and competence with a mixed-method approach. International Journal of Research in Marketing, 37(4), 714–736. https://doi.org/10.1016/j.ijresmar.2020.02.004

24.

Hayes

A. F.

Coutts

J. J.

(2020). Use omega rather than Cronbach’s alpha for estimating reliability. But…. Communication Methods and Measures, 14, 1–24. https://doi.org/10.1080/19312458.2020.1718629

25.

Hoffman

H. S.

Carver

C. S.

(1984). Political waffling: Its effects on the evaluations of observers. Journal of Applied Social Psychology, 14(4), 375–385. https://doi.org/10.1111/j.1559-1816.1984.tb02245.x

26.

John

L. K.

Jeong

Gino

Huang

(2019). The self-presentational consequences of upholding one’s stance in spite of the evidence. Organizational Behavior and Human Decision Processes, 154, 1–14. https://doi.org/10.1016/j.obhdp.2019.07.001

27.

Kluger

Park

(2020, December 21). Guardians of the year: Dr. Anthony Fauci and frontline health workers. Time. https://time.com/guardians-of-the-year-2020-anthony-fauci-frontline-health-workers/

28.

Kreps

T. A.

Laurin

Merritt

A. C.

(2017). Hypocritical flip-flop, or courageous evolution? When leaders change their moral minds. Journal of Personality and Social Psychology, 113(5), 730–752. https://doi.org/10.1037/pspi0000103

29.

Lakens

Caldwell

A. R.

(2021). Simulation-based power analysis for factorial analysis of variance designs. Advances in Methods and Practices in Psychological Science, 4(1), 1–14. https://doi.org/10.1177/2515245920951503

30.

Laurent

S. M.

Clark

B. A. M.

(2019). What makes hypocrisy? Folk definitions, attitude/behavior combinations, attitude strength, and private/public distinctions. Basic and Applied Social Psychology, 41(2), 104–121. https://doi.org/10.1080/01973533.2018.1556160

31.

Leary

M. R.

Diebels

K. J.

Davisson

E. K.

Jongman-Sereno

K. P.

Isherwood

J. C.

Raimi

K. T.

Deffler

S. A.

Hoyle

R. H.

(2017). Cognitive and interpersonal features of intellectual humility. Personality and Social Psychology Bulletin, 43(6), 793–813. https://doi.org/10.1177/0146167217697695

32.

Lovakov

Agadullina

E. R.

(2021). Empirically derived guidelines for effect size interpretation in social psychology. European Journal of Social Psychology, 51(3), 485–504.

33.

McNeish

(2018). Thanks coefficient alpha, we’ll take it from here. Psychological Methods, 23, 412–433. https://doi.org/10.1037/met0000144

34.

Meyer

G. J.

Finn

S. E.

Eyde

L. D.

Kay

G. G.

Moreland

K. L.

Dies

R. R.

Eisman

E. J.

Kubiszyn

T. W.

Reed

G. M.

(2001). Psychological testing and psychological assessment: A review of evidence and issues. American Psychologist, 56(2), 128–165.

35.

Norman

S. M.

Avolio

Luthans

(2010). The impact of positivity and transparency on trust in leaders and their perceived effectiveness. Management Department Faculty Publications. 144. http://digitalcommons.unl.edu/managementfacpub/144

36.

Oreskes

Conway

E. M.

(2011). Merchants of doubt. Bloomsbury Publishing PLC.

37.

Parsons

Bales

R. F.

(1955). Family socialization and interaction processes. Free Press.

38.

Paulhus

D. L.

John

O. P.

(1998). Egoistic and moralistic biases in Self-Perception: The interplay of Self-Deceptive styles with basic traits and motives. Journal of Personality, 66, 1025–1060. https://doi.org/10.1111/1467-6494.00041

39.

Pew Research Center. (2019, December 17). In a politically polarized era, sharp divides in both Partisan coalitions. Pew Research Center. https://www.pewresearch.org/politics/2019/12/17/in-a-politically-polarized-era-sharp-divides-in-both-partisan-coalitions/

40.

Rabinowitz

Macdonald

S. E.

(1989). A directional theory of issue voting. American Political Science Review, 83, 93–121.

41.

Rudman

L. A.

McLean

M. C.

Bunzl

(2013). When truth is personally inconvenient, attitudes change: The impact of extreme weather on implicit support for green politicians and explicit climate-change beliefs. Psychological Science, 24(11), 2290–2296. https://doi.org/10.1177/0956797613492775

42.

Schumann

(2018). The psychology of offering an apology: Understanding the barriers to apologizing and how to overcome them. Current Directions ins Psychological Science, 27, 74–78.

43.

Shafer

(2015). The hole in Hillary’s flip-flop excuse. Politico Magazine. https://www.politico.com/magazine/story/2015/10/democratic-debate-hillary-clinton-flip-flop-213247/

44.

Sherman

(2015). Hillary Clinton’s changing position on same-sex marriage. Politifact. https://www.politifact.com/factchecks/2015/jun/17/hillary-clinton/hillary-clinton-change-position-same-sex-marriage/

45.

Sherman

(2025). Flip flop: Trump said he wasn’t looking at pausing tariffs. Two days later he did. Politifact. https://www.politifact.com/factchecks/2025/apr/09/donald-trump/flip-flop-trump-pauses-country-tariffs-90-days/

46.

Shutt

(2024). Fauci defends his work on COVID-19, says he has an ‘open mind’ on its origins. Iowa Capital Dispatch. https://iowacapitaldispatch.com/2024/06/03/fauci-defends-his-work-on-covid-19-says-he-has-an-open-mind-on-its-origins/

47.

Sigelman

C. K.

(1986). Shattered expectations: Public responses to ‘out-of-character’ presidential actions. Political Behavior, 8(3), 262–286. https://doi.org/10.1007/BF01002101

48.

Stekelenburg

Schaap

Veling

van’t Riet

Buijzen

(2022). Scientific-Consensus communication about contested science: A pre-registered meta-analysis. Psychological Science, 33(12), 1989–2008. https://doi.org/10.1177/09567976221083219

49.

Tomz

Van Houweling

(2010). Candidate repositioning. Unpublished manuscript. http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.387.5419&rep=rep1&type=pdf

50.

Tomz

Van Houweling

R. P.

(2008). Candidate positioning and voter choice. American Political Science Review, 102(3), 303–318. https://doi.org/10.1017/S0003055408080301

51.

Westfall

Van Boven

Chambers

J. R.

Judd

C. M.

(2015). Perceiving political polarization in the United States: Party identity strength and attitude extremity exacerbate the perceived partisan divide. Psychological Science, 10(2), 145–158. https://doi.org/10.1177/1745691615569849

52.

Whitt

C. M.

(2022). The effects of teaching about the replication crisis on undergraduates’ epistemic dependence [Unpublished doctoral dissertation, The University of Alabama].

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.00 MB