Mitigating climate change necessitates substantial shifts towards low-carbon behaviors in affluent societies. Despite extensive research identifying influential factors and potential interventions, widespread voluntary behavior change remains elusive. Governments, fearing unpopularity, often prioritize technological solutions over policies restricting high-carbon behaviors, creating a "governance trap" where responsibility is mutually attributed between the government and the public. This study proposes a novel approach: high-profile leaders visibly adopting high-impact low-carbon behaviors to influence social norms. Three key reasons underpin this focus on leaders: their high status allows them to shape social discourse and norms; they often have significantly larger carbon footprints than the average citizen, raising equity concerns; and they possess greater responsibility and power to guide society's climate response. The study investigates the impact of politicians and celebrities "walking the talk" with specific high-impact behaviors (flying less, eating less meat, driving electric cars, improving home energy efficiency, and active travel), which are known to significantly reduce individual carbon footprints. Existing research suggests such leader actions enhance credibility and encourage similar behavior among the public. The study defines "leaders" as individuals intentionally influencing groups to achieve common goals, applying this to both politicians (responsible for societal climate responses) and celebrities (influencing attitudes and behaviors). Many politicians and celebrities have considerable carbon footprints, making their behavioral choices highly impactful.

The paper reviews existing literature on behavior change and climate action. It notes the challenges in achieving large-scale behavior change, and the tendency for governments to favor technological solutions over behavioral interventions. The literature highlights the role of social norms and influence of leaders in shaping behaviors. Specifically, the concept of embodied leadership, where a leader's actions embody their beliefs and influence observers, is introduced along with credibility enhancing displays (CREDs), which signal commitment through effortful actions. The research cites existing work suggesting that pro-social behavior, like adoption of low-carbon behaviors, can be viewed negatively due to suspicion of ulterior motives or perceived harm, leading to phenomena such as do-gooder derogation. However, the paper also presents counterarguments indicating that positive perceptions of pro-environmental behaviors can lead to emulation, especially if demonstrated by trusted opinion leaders. Existing studies on CREDs in the context of pro-environmental behavior are discussed, highlighting the impact of visible commitment on behavior adoption. The literature also notes the public's desire for government leadership in climate action. The authors introduce the concept of a "disembodied" framing of behavior change, which ignores the actions of high-status individuals and focuses on generalized behavior change, contrasting this with the proposed embodied leadership approach.

A pre-registered survey experiment was conducted using a nationally representative UK sample (n=1267) recruited via the Prolific platform. The experiment employed a between-subjects design with three conditions: (1) Leading by Example (leaders visibly adopting specified high-impact low-carbon behaviors for the past two years); (2) Not Leading by Example (leaders advocating for climate action without adopting the behaviors); and (3) Disembodied (information about climate action without mentioning a leader). Two leader types were included: politicians and celebrities. Participants were randomly allocated to one of the five conditions (2 leader types x 2 leader actions + 1 disembodied condition). The survey utilized Qualtrics software. Participants gave informed consent and read vignettes describing a fictional interview with a politician or celebrity (or a report on climate change in the disembodied condition). The vignettes described climate change and proposed solutions (technology, international agreements, behavior change), differing only in the leader's stated behavior. The study measured various dependent variables including willingness to act (using seven survey questions on adopting various low-carbon behaviors), perceptions of leaders (commitment to climate, effectiveness, warmth/competence, reactance, increased approval), and perspectives on climate change (leaders' efficacy, others' willingness to act, moral salience, pro-environmental identity, support for climate action, concern/risk perception, appetite for leadership, generalized trust). Factor analysis and component analysis were used to assess scale reliability and components. Confirmatory and exploratory analyses were conducted (ANOVAs, MANOVAs). Demographic data was also collected to check for the effectiveness of random allocation.

Analysis revealed no significant differences across conditions in terms of demographics, ensuring the random allocation was successful. A strong "appetite for leadership" was demonstrated: over 86% of respondents agreed that leaders should set an example by making lifestyle changes. A two-way omnibus MANOVA showed a significant main effect of Leader Action on most dependent variables, indicating that leading by example significantly influenced responses. A smaller, significant main effect of Leader Type was also found, although the differences between politicians and celebrities were largely insignificant. There was no significant interaction between Leader Action and Leader Type. Specifically, respondents in the Leading by Example condition reported significantly higher willingness to adopt low-carbon behaviors compared to the Not Leading by Example condition (small effect size). However, there was no significant difference between the Leading by Example and the Disembodied conditions. Analysis of individual Willingness to Act items revealed significant differences between conditions for making significant lifestyle changes, flying less, and using public transport more often. There was significantly higher positive perceptions of leaders in the Leading by Example condition (Cares/Believes, Warmth/Competence, Effectiveness, Knowledgeable, Increased Approval), supporting the CREDs hypothesis. There was no significant difference in perceptions of whether the leader exaggerated climate risk. Interestingly, the Not Leading by Example condition elicited significantly higher reactance compared to both the Disembodied and Leading by Example conditions. No significant differences were found between conditions for others' willingness to act, leaders' efficacy, moral salience, pro-environmental identity, support for climate action, or concern/risk perception. Finally, political orientation did not significantly influence responses to leading by example.

The study demonstrates that leaders visibly adopting high-impact low-carbon behaviors significantly increases public willingness to adopt similar behaviors, while a lack of visible action decreases this willingness. There was no difference in willingness between leading by example and simply providing information on the need for climate action. The findings support the CREDs theory, as leading by example increased perceptions of leader credibility and trustworthiness. Importantly, leading by example did not lead to increased reactance; instead, it was the absence of such actions that caused negative reactions. The consistent results across politicians and celebrities indicate the importance of symbolic leadership, especially for politicians responsible for societal responses to crises. The small effect size of willingness to act in this study might underestimate the actual potential of leading by example due to the use of a single fictional leader in the experiment. The study partially supports the embodied leadership theory, highlighting the importance of actions as a signal of commitment. Notably, the political leanings of respondents did not significantly impact the responses, with those on the right even reacting positively to the actions.

This research indicates that high-status individuals visibly adopting impactful low-carbon behaviors significantly increases public willingness to act and enhances perceptions of leader credibility. This supports theories of CREDs and embodied leadership, suggesting that actions are powerful signals of commitment. Leading by example may be a crucial factor in overcoming the governance trap. Future research should explore the impact of multiple leaders, different leader types, and the effects of leading by example on various types of behavioral changes in different contexts and countries.

The study's reliance on self-reported willingness to act might not accurately reflect actual behavior change. The use of fictional vignettes may limit the realism and generalizability of the results. The focus on a single leader and limited leader types also restricts generalizability. Further research is needed to assess the long-term impact and effectiveness in real-world scenarios.