Climate change beliefs and their correlates in Latin America

Prior work in advanced democracies (e.g., U.S., U.K., Germany, Australia) links climate skepticism to socio-demographics and political ideology, but little is known about these relationships in the developing world. In Latin America, scholarship has focused more on concern and risk perception than on distinct belief dimensions (existence, causes, consequences). Existing regional studies have either examined only one belief (e.g., anthropogenic origins) or relied on general surveys lacking dedicated climate modules, omitting key individual-level predictors. This is problematic given the role of public beliefs in enabling climate action and the region’s vulnerability to climate change. The study tests individual-level correlates of beliefs in existence, anthropogenic causes, and severe consequences across seven Latin American countries representing over 80% of regional emissions. It compares political ideology and socio-demographics to psychological factors (individualism/egalitarianism worldviews, NEP values, objective/subjective knowledge, perceived scientific consensus, trust in scientists, and personal experience with extreme weather) to assess which variables most strongly relate to each belief dimension.

The paper situates its contribution against literature showing that in advanced democracies political ideology and demographics often shape climate beliefs. However, Latin American evidence is sparse: one study examined anthropogenic beliefs but not existence or consequences; others emphasize country-level determinants (prosperity, democracy) rather than individual-level factors, often using multipurpose surveys without detailed climate modules. Prior global work highlights the potential roles of cultural worldviews (individualism/egalitarianism), NEP values, knowledge (objective vs. subjective), perceived scientific consensus, trust in scientists, and personal experience with extreme weather. Whether these patterns generalize to Latin America remained unclear, motivating the present multi-country, individual-level analysis.

Participants: Nationally diverse online samples from Argentina, Brazil, Chile, Colombia, Ecuador, Peru, and Mexico were recruited by Netquest (ISO 26362, ESOMAR compliant) using quotas for sex, age, and education, fielded Oct–Nov 2021. The final sample comprised 5400 adults (~830 per country; Ecuador n=421 due to panel constraints). Models used sample weights; demographic distributions approximated censuses, with some overrepresentation of higher education in Ecuador and Peru. Analyses often rely on complete cases (e.g., ~2883–2887 per model). Measures: Three dependent variables captured distinct belief dimensions. (1) Existence: binary question on whether climate is changing, followed by certainty (0–3), recoded into an 8-point scale (0=extremely sure not happening; 4=don’t know; 8=extremely sure happening). (2) Anthropogenic causes: among those assuming climate change is happening, coded 1 if “caused mostly by human activities,” else 0. (3) Consequences: perceived global impact on an 11-point scale (0=extremely bad to 10=extremely good), analyzed as probability of believing impacts will be negative. Key independent variables included: subjective knowledge (self-assessed 1–4), objective knowledge (six-item cause identification score, 0–6), perceived scientific consensus (1 if “most scientists think global warming is happening”), trust in scientists (1–4), NEP values (3-item mean; α≈0.53 after excluding a low-reliability item), individualism worldview (five-item mean; α=0.67), egalitarianism worldview (five-item mean; α=0.72), and personal experience with extreme weather (dichotomized). Political ideology was measured on left–right (1–10) and progressive–conservative (1–5) scales. Socio-demographics included sex, age, education (primary or less; secondary; undergraduate+), religion (with atheist as reference), race (including “Mestizo” where applicable), and income (minimum-wage brackets; 0–1 as reference). Statistical analysis: Ordinary least squares models were estimated for each outcome for the pooled sample and by country, with heteroskedasticity-robust standard errors and weights. For the existence scale, coefficients indicate changes on the 0–8 scale per unit increase; for anthropogenic and consequences (modeled as linear probability), coefficients multiplied by 100 represent percentage-point changes. Robustness: multilevel random-intercept models by country, country fixed-effects models, ordinal logistic (existence), and binary logistic (anthropogenic, consequences) yielded substantively similar results. Additional checks showed low multicollinearity (all VIF < 1.5), joint significance of predictors (F-tests), and stability under Benjamini–Hochberg multiple-testing adjustments.

- Prevalence of beliefs: Across all seven countries, over 90% believe climate change is happening; on average 93% attribute it primarily to human activity. Beliefs about consequences are more divided: about 65% believe impacts will be negative. - Existence of climate change: Psychological factors are strongest correlates regionally. Perceived scientific consensus associates with a +0.49 increase on the 0–8 existence scale (p<0.01). NEP (β=0.305, p<0.01) and personal experience with extreme weather (β=0.232, p<0.01) also correlate positively, though NEP reliability is low (α≈0.53). Subjective knowledge (β=0.140, p<0.01) and objective knowledge (β=0.165, p<0.01) are positive overall but only significant in subsets of countries. Trust in scientists is weakly positive at the 10% level overall. Egalitarianism is small/weakly positive (β=0.067, p<0.05). Individualism shows no significant correlation with existence. Political ideology exhibits weak associations overall: progressivism relates to slightly less skepticism (β≈0.102, p<0.05), significant mainly in Brazil and Chile; left–right is not significant. Socio-demographics are largely not associated (education, age, income, sex non-significant); a race effect (greater skepticism among Black respondents) appears only in Argentina. - Anthropogenic causes: Individualistic worldviews are the only robust negative psychological correlate: a one-unit increase in individualism corresponds to a 3.8 percentage-point reduction in believing climate change is mainly human-caused (p<0.01), significant in most countries except Ecuador and Mexico. Objective knowledge (β=0.054, p<0.01), perceived scientific consensus (β=0.070, p<0.01), trust in scientists (β=0.056, p<0.05), and NEP (p=0.045; p<0.01) are positively associated overall, though country consistency varies and NEP reliability cautions apply. Political ideology is weak: left–right not significant overall (β=0.015), with positives in Ecuador and Mexico only; progressivism relates to less skepticism (β=−0.025, p<0.05) but is not significant in most countries. Education is the only socio-demographic consistently associated with lower skepticism; sex, religion, income, age, and race are not significant overall. - Consequences: Individualism is a strong, consistent negative correlate: a one-unit increase reduces the probability of believing impacts will be negative by 11.5 percentage points (p<0.01), significant and similar in magnitude across all countries. Objective knowledge is positively associated (β=0.067, p<0.01; significant in 4 of 7 countries). Perceived scientific consensus (β=0.062, p<0.05) and NEP (β=0.081, p<0.01) are positive overall, with limited country-level consistency and NEP reliability caveat. Subjective knowledge, trust in scientists, personal experience, and egalitarianism are not significant overall. Politically, being on the left is associated with perceiving more negative impacts (β=0.085, p<0.01) but is significant in only 3 countries; conservatism vs. progressivism is not significant. Education (especially undergraduate+) is the most consistent socio-demographic correlate of perceiving negative impacts; age, sex, race, income, and religion show weak or no associations, with some denominations linked to minimizing impacts in a minority of countries. - Robustness: Results are stable across alternative model specifications, low multicollinearity, joint significance tests, and multiple-testing adjustments.

The study demonstrates that in Latin America skepticism about the existence and anthropogenic causes of climate change is limited, but skepticism about severe consequences is more prevalent. Psychological variables—especially individualistic worldviews—are the most influential correlates of skepticism, overshadowing political ideology and socio-demographics. Individualism strongly predicts lower acceptance of anthropogenic causes and less negative views of consequences, but not disbelief in existence. This pattern aligns with theories that acknowledging human causes and serious impacts implies support for mitigation/adaptation and government intervention, which conflicts with individualistic preferences. The findings suggest that perceived scientific consensus particularly bolsters belief in existence, perhaps because consensus provides tangible reassurance in low-trust contexts, though it appears less effective as a ‘gateway’ to beliefs about causes and consequences. Objective knowledge positively relates to all belief dimensions (though variably across countries), supporting information-focused communication strategies. The weak and inconsistent roles of subjective knowledge, personal experience, political ideology, and most socio-demographics indicate that climate belief formation in Latin America follows dynamics distinct from many advanced democracies, offering opportunities for broad, cross-ideological pro-climate coalitions. Strategically, communication could align with individualistic motivations (e.g., emphasizing economic co-benefits, pro-market solutions) and leverage trusted in-group messengers, while also deploying accessible media to convey scientific consensus and actionable steps.

This multi-country study provides one of the first comprehensive mappings of climate change beliefs and their individual-level correlates in Latin America. It shows widespread acceptance of climate change and its anthropogenic origins but greater division over consequences. Psychological factors—chiefly individualistic values—are the dominant correlates of skepticism, whereas political ideology and socio-demographics generally have limited influence. These insights can inform tailored, value-congruent climate communication strategies that emphasize co-benefits and leverage in-group messengers, and they support using objective knowledge and consensus messaging to strengthen beliefs. Future research should establish causal pathways via experimental designs, broaden conceptualizations and measurements of individualism and related value constructs, further examine the role of ideological liberalism, and test how social trust interacts with individualism to shape beliefs. Continued within-region comparisons can clarify contextual moderators and improve the design of interventions across the Global South.

- Cross-sectional design precludes causal inference; potential endogeneity cannot be ruled out. Experimental studies are needed to test causality. - Narrow operationalization of individualism (attitudes toward government) may not capture the broader construct; other facets (e.g., prioritizing personal over collective interests) should be examined. - The NEP scale exhibited low internal reliability (α≈0.53 after item removal), warranting caution in interpreting NEP-related results. - Personal experience with extreme weather was measured without explicitly attributing events to climate change, which may weaken associations; future work should test attribution-sensitive measures. - Online panel samples, while quota-matched and weighted, may overrepresent higher education in some countries; however, robustness checks and alignment with representative surveys on related constructs mitigate this concern.