How natural disasters and environmental fears shape American climate attitudes across political orientation

The study addresses persistent polarization in U.S. public opinion on climate change, where conservatives typically exhibit lower concern, greater skepticism of climate science, and reduced risk perception relative to liberals and moderates. Recognizing that climate change is global, gradual, and difficult to assess subjectively, the authors explore how threat perception and fear shape climate attitudes across political orientations. The research tests three hypotheses: H1) moderates, and especially conservatives, express lower climate concern than liberals; H2) fear of both natural disasters (e.g., floods, droughts, wildfires, hurricanes) and anthropogenic environmental disasters (e.g., pollution, oil spills, industrial accidents) is associated with increased climate concern; H3) the positive association between fear of disasters and climate concern is more pronounced for moderates and conservatives relative to liberals. The purpose is to provide practical insights for climate communication and theoretical contributions to models linking threat, emotion, and ideology.

Prior research consistently finds conservatives in the U.S. and other polarized democracies perceive lower climate risk, show more skepticism, and distrust climate science relative to liberals and moderates. Ideological differences in threat prioritization suggest conservatives react more to local, intentional threats with malicious intent, whereas liberals prioritize global, unintentional or inaction-based threats. Affective science links fear and worry to greater environmental risk perceptions, voting importance, policy support, adaptation, and information seeking—even beyond political orientation for some outcomes. Disaster literature differentiates between natural disasters (uncontrollable, climate-exacerbated hazards) and anthropogenic environmental disasters (human-caused, controllable hazards that perpetuate climate change). People attribute more blame and perceive greater environmental impact from anthropogenic disasters and ascribe ethical concerns to human-involved events. Although classic theories often posit threats generally increase conservatism or intensify prior ideology, emerging work argues effects depend on threat context and perceived ideological competence to resolve the threat. Given issue ownership linking climate solutions to liberal parties, environmental threats may induce more liberal climate attitudes among those who feel threatened, particularly conservatives.

Data come from the 2023 SPEER Survey, an online Qualtrics-administered, opt-in panel of 2,188 U.S. adults (18+) with quota-based recruitment to approximate U.S. census benchmarks (age, gender, race/ethnicity, education, region). The study followed University of Oklahoma IRB protocols (Approval #15823) and obtained informed consent. Dependent variables (each recoded to 1–4 scales for approximate normality) measured four dimensions of climate concern adapted from prior surveys: climate belief (anthropogenic contributions), climate risk (perceived risk to people/environment), climate importance (personal importance), and climate worry. Independent variables included political orientation (liberal, moderate, conservative) and six self-reported fear domains based on Chapman University’s Survey of American Fears: fear of natural disasters, environmental (human-caused) disasters, strangers, crime, human events, and personal tragedy. Fear scales are linear self-placement measures with 10 categories (0–10). Interaction terms were constructed for political orientation with natural disaster fear and with environmental disaster fear. Models controlled for party affiliation, age, sex, socioeconomic status, family status, religious variables, and subjective weather perception. The primary analyses used OLS regressions for each dependent variable, estimating baseline and interaction models. Post-hoc marginal effects and simple slopes were examined across levels of fear scales to test moderation and convergence across political orientations.

Baseline results (Tables 1–2) show strong support for H1: relative to liberals, moderates and especially conservatives exhibit significantly lower climate concern across belief, risk, importance, and worry. Adjusted R^2 values range approximately 0.369–0.492 (belief and risk) and 0.423–0.497 (importance and worry). Supporting H2, fear of environmental (human-caused) disasters is positively associated with all four climate concern domains and is among the strongest fear predictors. Fear of natural disasters positively predicts climate risk, importance, and worry, but not belief in climate change. Fear of strangers and crime weakly and negatively predict climate risk and importance, while fear of human events and personal tragedy are generally unrelated. Interaction analyses (supporting H3 for natural disasters) indicate fear of natural disasters significantly increases climate concern for conservatives relative to liberals across outcomes, with limited effects on liberals and modest effects on moderates. Marginal effects show political differences are largest at low-to-moderate fear levels and diminish with higher fear: for natural disaster fear, at the 25th percentile, marginal estimates span (liberal to conservative) approximately: belief 3.16 vs 2.59; risk 3.06 vs 2.45; importance 2.86 vs 2.39; worry 2.68 vs 2.17. At the 75th percentile, these ranges narrow: belief 3.09–2.82; risk 3.04–2.70; importance 2.77–2.64; worry 2.76–2.44. Convergence thresholds appear around fear scale values of 8–10, with liberals and moderates or moderates and conservatives no longer differing significantly at high fear levels; liberals and conservatives converge last, often near the highest fear values. For environmental disaster fear, interactions provide mixed support for H3: conservatives (and to some extent moderates) show stronger gains in belief and risk with increasing fear relative to liberals, but differences in importance and worry are less consistent (no significant slope differences for liberals vs conservatives; some marginally higher importance for moderates). Marginal estimates at the 25th vs 75th percentiles of environmental fear also show convergence: belief 3.03–2.44 (25th) vs 3.35–3.02 (75th); risk 2.87–2.26 vs 3.37–2.95; importance 2.66–2.21 vs 3.20–2.83; worry 2.47–1.99 vs 3.07–2.66. Overall, higher levels of disaster-related fear reduce ideological gaps in climate concern, with the strongest and most consistent convergence observed for natural disaster fear.

Findings strongly support H1 and H2, reaffirming that conservatives exhibit lower climate concern and that fear—particularly of anthropogenic environmental and natural disasters—is positively associated with concern. Evidence for H3 is robust for natural disasters and mixed for environmental disasters: conservatives’ concern increases more steeply with natural disaster fear, and somewhat more with environmental fear in belief and risk domains, leading to convergence at high fear levels. Theoretically, results challenge universal threat-to-conservatism models and support context-dependent frameworks (e.g., ideology-affordance), wherein perceived capacity of an ideology to address a threat guides attitudinal shifts. Given that climate solutions are generally attributed to liberal actors, conservatives who feel threatened by environmental hazards adopt more liberal climate attitudes. The discussion also considers the role of discrete emotions—particularly anger accompanying human-caused disasters—which may elevate perceived risk independent of fear and help explain partial interactions for environmental fear. Practically, the study suggests that emphasizing tangible, proximal environmental threats and connecting them to climate change may increase climate concern among conservatives, potentially improving support for policies and adaptive behaviors—provided accessible actions are offered to prevent fatalism or inaction.

This study demonstrates that disaster-related fear meaningfully shapes climate concern across the political spectrum, with conservatives showing the largest increases and convergence toward liberals at high fear levels—especially for natural disaster fear. It advances theory by supporting context-sensitive models linking threat, emotion, and ideology, and provides applied guidance for communicators and policymakers seeking bipartisan support for climate action. Future research should: examine additional threat stimuli (both climate-related and beyond) that may evoke shifts toward liberal attitudes; directly measure discrete emotions (e.g., anger vs fear) to parse affective mechanisms; include behavioral measures alongside attitudes; and test interventions that connect local environmental threats to global climate change while providing clear, feasible behavioral pathways.

Methodologically, the online opt-in sample, while quota-matched on key demographics, overrepresents liberals and undercounts conservatives/Republicans, reflecting common online survey biases. Political affiliation measurement is limited by a large share of self-identified independents (37%) without lean measures, risking misclassification. Outcomes assess attitudes rather than actual behaviors. While interaction terms achieve statistical significance, they add little explained variance beyond baseline models; much variance is accounted for by direct effects. These issues may affect generalizability and the interpretation of moderation effects.