Fair, effective, and affordable: a recipe for globally popular climate policy

Technical Summary: Fair, Effective, and Affordable: A Recipe for Globally Popular Climate Policy

Problem Statement
Policymakers face a critical challenge in enacting and sustaining climate policies: public backlash often leads to policy reversals, as seen in the suspension of France's fuel tax, the UK's delay of its internal-combustion vehicle ban, and Canada's repeal of its carbon price. While prior research has extensively examined individual characteristics that predict support for climate action (e.g., political ideology, climate worry), there is a gap in understanding why specific policies are popular while others are not. This paper posits that the relative popularity of climate policies is not merely a function of individual demographics but is shaped by widely shared perceptions of policy consequences. The authors propose that if these perceptions are consistent across diverse societies, they can serve as a predictive tool for identifying policies likely to garner broad public support.

Methodology
The study employed a large-scale, cross-national survey design involving 75 societies across all inhabited continents, with a total sample of $n = 14,053$ respondents. The research utilized the Global Social Norms (GSN) network to recruit convenience samples of approximately 200 adults per society.

1. Policy Set: Respondents rated their support for 54 distinct climate policies, categorized into subsidies, requirements, limits, and taxes. Support was measured on a four-point scale (strongly oppose to strongly support).
2. Perceived Policy Consequences (PPCs): For each policy, respondents identified which of seven potential consequences they believed others would cite as reasons for supporting or opposing the policy. These dimensions were: effectiveness, fairness, low cost for households, low cost for government, benefit to nature, protection of freedom, and likelihood of promoting corruption. The primary PPC metric was the "global net balance" (proportion citing as support minus proportion citing as opposition).
3. Experimental Component: A pre-registered embedded experiment randomly assigned respondents to view specific policies accompanied by expert statements regarding consequences.
   • Negative Treatment: 15 policies were paired with negative expert statements (e.g., "Experts say this policy has little effect on greenhouse gas emissions").
   • Positive Treatment: 15 policies were paired with positive expert statements (e.g., "Experts say this policy significantly reduces greenhouse gas emissions").
   • Control: Policies were presented without expert statements.
4. Analytical Approach:
   • Observational: Ordinary Least Squares (OLS) regression was used to predict global policy popularity based on the seven PPC scores. Leave-one-out cross-validation was employed to test out-of-sample predictive power.
   • Causal Inference: To address potential reverse causality (motivated reasoning where support shapes perception), the authors constructed "popularity-neutral" PPCs by predicting perceptions at the midpoint of the support scale. Additionally, the experiment tested whether shifting PPCs via expert statements causally shifted policy support.
   • Moderation: Society-level institutional quality (Corruption Perceptions Index and Press Freedom Index) was analyzed as a moderator for the effectiveness of positive expert statements.

Key Results

• Predictive Power of PPCs: Three to four perceived consequences explain the vast majority of variance in policy popularity. Specifically, fairness, effectiveness, and low cost for households were statistically significant positive predictors. Together, these three dimensions accounted for 84% of the cross-policy variance in global popularity. The full model (including all seven PPCs) achieved an $R^2$ of 0.863, with a leave-one-out cross-validation $R^2$ of 0.784.
• Global Consistency: While overall support levels vary by society (higher in societies with greater climate worry and lower in wealthier societies), the relative ranking of policies is highly consistent across societies ($r = .95$ between high and low-income societies; $r = .99$ between high and low climate-worry societies). Variance decomposition revealed that only 0.9% of the total variance in ratings was attributable to differences between societies, compared to 13.1% between policies.
• Causal Mechanism: The popularity-neutral analysis confirmed that the relationship between PPCs and popularity holds even when removing the channel of motivated reasoning. The experiment further demonstrated a causal link: negative expert statements significantly reduced policy support (by 0.64 points on a -2 to +2 scale) by shifting negative perceptions.
• The Communication Dilemma: Contrary to the hypothesis, positive expert statements did not increase support on average. In fact, they often backfired. An exploratory analysis revealed that the effect of positive statements was heavily moderated by institutional credibility. Positive statements increased support only in societies with high institutional quality (low corruption, high press freedom). In societies with lower institutional quality, positive statements tended to reduce support or have no effect, suggesting that in these contexts, positive messaging is perceived as untrustworthy propaganda.

Significance and Claims
The paper claims to reframe the question of climate policy support from "who supports climate action?" to "which policies are popular?" The authors argue that policy popularity is driven by a small set of widely shared perceptions regarding fairness, effectiveness, and cost.

• Practical Utility: The study suggests that "PPC profiling" can provide policymakers with a reliable, empirical basis for anticipating public response before a policy is proposed. By identifying policies that are perceived as effective, fair, and low-cost, governments can avoid the political costs of enacting unpopular measures that are later reversed.
• Communication Strategy: The findings highlight a critical communication dilemma. While negative information about policy consequences is universally credible and reduces support, positive information is only effective in societies with high institutional trust. This implies that in contexts with low institutional credibility, positive framing may be counterproductive, and policymakers cannot rely solely on communication to overcome resistance to policies perceived as costly or unfair.
• Policy Design: The authors conclude that the space for politically viable climate action is not empty but is populated by policies with specific consequence profiles (e.g., tree planting, which was the most popular policy) rather than the punitive measures (e.g., meat taxes) that have historically faced backlash.

The paper modestly notes that its experimental treatments were brief and survey-embedded, and that the formation of consequence perceptions is an open question for future research. However, it asserts that the identified recipe—fairness, effectiveness, and affordability—offers a concrete, evidence-based guide for designing climate policies that can sustain public support globally.