Refining relational climate conversations to promote collective action

The study addresses the attitude–behavior gap in climate engagement in the United States, where many people express concern about climate change but comparatively few participate in collective climate action. Despite evidence that talking about climate change can increase concern and promote discussion, most Americans rarely or never discuss it with friends and family, reinforcing a spiral of silence. Relational climate conversations—dialogue between people who know each other—are promoted as a strategy to move concerned individuals toward action, yet their mechanisms and effects on collective action are understudied. This study examines three structured conversations between US climate activists and non-activists they know to evaluate whether such conversations increase knowledge, concern, personal response efficacy, and intentions to act, and whether they translate into participation in collective climate action. It further explores barriers to action, how activists proximize action (make it concrete and close), and whether such proximization influences efficacy and action. Research questions: (1) Do conversations increase concern, knowledge, personal response efficacy, and intention to act, and do they influence participation in collective climate action? What barriers are reported? (2) How do activists proximize collective climate action? (3) Does proximization influence non-activists’ action? (4) Do efficacy beliefs mediate any relationship between conversations and collective action?

The background synthesizes work on psychological distance, construal level, and efficacy as they relate to climate engagement. Psychological distance—temporal, spatial, social, and hypothetical—can serve as a barrier by placing climate change impacts as far away or affecting others. Empirical findings on proximity and policy support/actions are mixed, with contextual moderators noted. The paper extends this lens to the psychological distance of collective climate action itself, proposing that action may be perceived as future, elsewhere, by dissimilar others, or unlikely. Social identification with environmental activists predicts activism and relates to collective efficacy, which influences action and policy support. Common conversation advice emphasizes finding common ground (social proximity), while less work has examined making action proximal along hypothetical, spatial, and temporal dimensions by discussing specific actions and local opportunities. Construal-level theory links abstraction to psychological distance; messages that move from abstract “why” to concrete “how” can promote near-term action. Efficacy—self, response, collective, and participative—predicts collective action and policy support, and climate discussions can increase self-efficacy to converse further. However, it remains unclear how climate conversations affect efficacy related to collective action. The study posits that concretizing (proximizing) collective action in conversations could raise perceived efficacy and participation.

Design: A mixed-methods study with one treatment group and two control groups (n=41 each) of US residents not currently involved in collective climate action but who knew at least one climate activist. Treatment participants engaged in a series of three relational climate conversations with an activist they knew; both took surveys across the study. Two control groups were used: a survey-only group (to test for survey effects) and a no-intervention group (baseline over the same period). IRB approval was obtained from the College of Saint Benedict and Saint John’s University; written informed consent was collected. Treatment protocol: Conversation 1 (≈60 minutes) covered (1) seriousness of the climate crisis, (2) existence of solutions, (3) climate justice (unequal responsibility/impacts and justice-centered solutions), and (4) opportunities for collective climate action, especially locally. Conversations 2 and 3 (≈30 minutes each), spaced three weeks apart, asked about further thoughts, any actions taken, and supports/barriers. Conversations occurred via Zoom or in person and were transcribed (Otter.ai). Activists were encouraged to ask questions, listen actively, find common ground, share personal stories, and invite partners to act. A focus group of activists informed the multi-conversation design. Measures and surveys: Activists and partners completed pre-study surveys and post-conversation surveys after each conversation; partners also completed a 3-month follow-up. The survey-only control completed identical surveys on the same timeline but without conversations; the no-intervention control reported on attitudes and actions over the prior five months at one timepoint. Partner attitudes were measured on five-point scales: knowledge (average across 7 domains: climate science; current local/global impacts; future impacts; root causes; climate justice; solutions), concern (average across current local/global impacts, future impacts, climate justice), personal response efficacy (“To what extent do you believe your actions can help counteract climate change,” 0–4), and intention to act (average likelihood to take collective actions; make lifestyle changes; talk to others; seek information). Actions were self-reported and aggregated into (a) collective action score (count of distinct collective action subtypes since baseline: joining an organization, attending a training, testifying publicly, contacting a representative, other collective actions) and (b) total action (collective actions plus seeking information, talking to others, lifestyle changes, donating). Recruitment and participants: Activists were recruited via outreach to 230 US organizations engaged in climate action, environmental science departments, and the AESS listserv; activists recruited known non-activists as partners and for control groups. Relationships were often close (friends 44%, family 29%, significant others 10%); 51% rated relationships as very close. Activists were mostly very progressive (69%); partners were very (46%) or somewhat progressive (34%). Most activists reported being impacted by climate change (89%); partners were split between unsure (51%) and impacted (37%). Age and financial situation differed significantly across groups (age: χ²(14, n=41)=41.1, p=.04; financial situation: χ²(8, n=41)=15.9, p<.001). Six treatment pairs did not complete all conversations; three partners missed some surveys. Qualitative dataset: 50 pairs; quantitative analyses: 41 pairs. Each control group: n=41. Analytic approach: Conversation transcripts were coded in ATLAS.ti using a priori code families (e.g., collective action, climate justice, solutions, barriers, active listening, storytelling) and inductive grounded theory to add subcodes and emergent themes. Statistical analyses were conducted in Python. For attitudes (RQ1a), mixed ANOVAs compared treatment vs survey-only across pre, interim, and post (with checks for normality, homoscedasticity, sphericity via Shapiro–Wilk, Levene’s, Mauchly’s tests). For actions (RQ1b), one-way ANOVAs compared treatment, survey-only, and no-intervention groups. Spearman correlations tested associations between activists’ proximization strategies (counts across three conversations) and partners’ outcomes (collective action; change in personal response efficacy).

- Attitudinal change (treatment vs survey-only, mixed ANOVAs): Treatment participants showed greater increases from pre to post in: • Knowledge: +19% vs +6% (p<0.001) • Intention to take action: +12% vs +2% (p=0.013) • Personal response efficacy: +11% vs +1% (p=0.021) • Concern: +3% vs −4% (p=0.042); small increase likely due to a high initial concern (mean 3.4/4). - Actions (one-way ANOVAs across treatment, survey-only, no-intervention): No significant differences in total actions or collective actions across groups. • Mean collective action subtypes per participant: 0.24 (treatment), 0.17 (survey-only), 0.23 (no-intervention) • Mean total actions per participant: 2.12 (treatment), 1.76 (survey-only), 1.88 (no-intervention) • Attribution: 100% of treatment participants attributed their actions to the study vs 58% in survey-only. • Common actions across all groups: talking to others (≈63–71%), seeking information (≈51–65%); less common: lifestyle changes (≈27–44%); collective action and advocacy (≈17–24%); donating (≈5–12%). - Reported barriers to collective action (qualitative counts among partners): lack of free time (22); low personal response efficacy (19) and/or self-efficacy (12); lack of knowledge about climate action (17) or climate issues (12); other competing concerns (17); isolation from action communities (11); psychological distance from impacts (9) and from action (8); despair (10); burnout/exhaustion (9); not hearing about opportunities (7); feeling overwhelmed (by climate change: 6; by climate action: 2). - Psychological distance and abstraction of action: Partners frequently discussed collective action abstractly and distanced it socially, temporally, and spatially. Counts of action distancing forms (n=50 pairs): abstraction (18), social (7), temporal (4), spatial (1). Many framed lifestyle changes as more practical/controllable than collective action, sometimes reducing hope when these changes seemed ineffective. - Activist proximization strategies (n=50 pairs): explaining types of action (28 activists; 113 instances), naming specific organizations (25; 83), sharing personal action experiences (25; 63), suggesting actions (18; 47), inviting partners to act together (15; 31), linking partners’ skills to action (14; 24). Partners accepted most invitations/suggestions (37/43), yet overall collective action remained low; intentions did not significantly correlate with total actions taken. - Correlational findings (Spearman): No significant correlation between total proximization counts and partners’ collective action participation. However, explaining action was positively correlated with increases in partners’ personal response efficacy from pre to post (r(37)=0.47, p=0.003). Discussing systemic solutions in the abstract (e.g., infrastructure change) was negatively correlated with efficacy change (r(37)=-0.336, p=0.036). These suggest that talking concretely about action (the “how”) may bolster perceived efficacy more than abstract solutions talk.

Relational climate conversations produced meaningful attitudinal gains—greater knowledge, higher personal response efficacy, and stronger intentions to act—supporting prior evidence that conversations can shift beliefs and motivation. Yet these shifts did not translate into significantly greater rates of collective action than control conditions, indicating a persistent intention–behavior gap. Qualitative analyses illuminate why: many participants construed collective action as abstract and psychologically distant (future, elsewhere, for dissimilar others), while lifestyle changes felt proximal and manageable, despite often being less impactful. This distancing, coupled with awareness that some lifestyle actions are marginal for mitigation, sometimes engendered helplessness. Activists deployed a spectrum of proximization strategies to counteract distance—explaining action types, sharing concrete experiences, naming local organizations, offering suggestions and invitations, and tailoring to partners’ skills. While such strategies did not directly correlate with increased collective action participation in this sample, explaining action was associated with higher growth in personal response efficacy, a known predictor of collective action. Conversely, focusing on abstract systemic solutions correlated with reduced efficacy gains. These findings suggest that bridging from high-level values and abstract solutions to concrete, proximal collective actions—who, what, where, when, and how—may be a promising pathway to mobilize concerned individuals, potentially by increasing perceived efficacy. However, constraints like limited time and competing priorities may still impede behavior change, indicating that conversations alone may need to be paired with organizational supports and clear, accessible on-ramps to action.

This study contributes evidence that relational climate conversations reliably enhance knowledge, personal response efficacy, and intentions among concerned non-activists, but by themselves rarely increase participation in collective climate action beyond survey effects. Applying psychological distance and construal-level theory clarifies that partners often perceive collective action as abstract and distant, while lifestyle changes feel concrete but can undermine hope when perceived as ineffective. Activist strategies that concretize and proximize collective action appear to strengthen personal response efficacy more than abstract solutions talk. Implications: For organizations aiming to mobilize collective action, conversations should emphasize concrete, local, near-term opportunities; link actions to participants’ skills and identities; and include specific invitations and co-participation pathways. Providing frameworks, resources, and trainings could help bridge intentions to action. Future research: Replicate with larger, more representative, and cross-national samples; experimentally manipulate conversation parameters (participant number/relationships, spacing, modality), content emphasis (concrete opportunities vs abstract solutions), and supports (follow-up, social onboarding). Assess multiple efficacy constructs (self-, response-, collective-, participative efficacy) and test causal pathways from action talk to efficacy to behavior.

- Sample representativeness: Treatment and control groups were not representative of US populations; participants skewed politically progressive and relatively insulated from direct climate impacts. The US context further limits global generalizability. - Group differences: Age and financial situation differed significantly across groups, which could confound outcomes. - Low base rate of collective action: Limited variance in collective action may have reduced power to detect correlations between discourse strategies and behavior. - Self-report and attribution: Actions and attitudes were self-reported; treatment participants may have over-attributed actions to the study due to demand characteristics. - Dropout and missing data: Some treatment pairs dropped out and some partners missed surveys, reducing the quantitative sample and potentially biasing results.