Improving public support for climate action through multilateralism

Public concern about climate change is widespread, yet costly climate policies often face substantial political opposition, driven by distributional conflicts and the global public goods nature of mitigation. Policymakers have invested in international frameworks (e.g., UNFCCC) to coordinate action and potentially increase domestic support for carbon pricing. The authors argue that voters value climate benefits and are more willing to support domestic action when others act as well, because: (1) multilateral action increases expected policy effectiveness (greater likelihood of achieving sustainability goals), and (2) multilateralism resonates with reciprocal fairness norms, increasing willingness to bear costs when others are contributing. The study investigates whether, and through which mechanisms, multilateralism increases support for costly climate policies along a continuum from decentralized coordination to flexible institutional arrangements without enforcement (e.g., nationally determined contributions under the Paris Agreement).

Early studies suggest that publics value multilateral cooperation on climate change. More recent contributions argue that public support for costly climate policies is largely independent of other countries’ actions. The authors revisit this debate by testing whether public support is interdependent with foreign climate policy choices. They reference work on public goods provision and cooperation (Ostrom; Axelrod; Taylor), institutional design of climate agreements (Bechtel & Scheve 2013), conditional cooperation (Tingley & Tomz 2014), and studies on support for international climate action and funding. The paper positions itself against claims of irrelevance of other countries’ behavior by providing experimental evidence on effectiveness beliefs and reciprocity norms as mechanisms linking multilateralism to support for climate action.

Two experimental studies were fielded across France, Germany, the United Kingdom (April 2019; N=2000 per country), and an additional United States sample (Dec 18, 2018–Jan 3, 2019; N=4075). Surveys were administered online by YouGov using matched sampling to approximate representative adult populations. Weights were applied; results are similar unweighted. Study 1: Vignette experiment on carbon tax support randomized multilateralism. Respondents evaluated a domestic carbon tax framed either as unilateral or as implemented by the respondent’s country and other major economies. Approval was measured on a 1–10 scale and analyzed at/above the midpoint as support. A cross-randomized effectiveness prime provided either no information (control) or statements that most experts expect the policy to avoid a few (low effectiveness) or most (high effectiveness) damaging consequences of climate change. This enabled an eliminated effects framework: estimating average treatment effects (ATE) of multilateralism, average controlled direct effects (ACDE) under fixed effectiveness beliefs, and eliminated effects (ATE−ACDE) to assess mediation by effectiveness. Study 2: Multilateral conjoint experiment on policy design. Respondents in France, Germany, the UK (same 2019 surveys) and the US (2018 survey) evaluated pairs of multilateral climate policy profiles specifying attributes for both their own country and other major economies: monthly household cost levels (approx. low/medium/high/very high, country-specific amounts), cost paths over time (constant, increasing, decreasing), and investment allocation between adaptation and mitigation (shares). Attributes were fully randomized, allowing estimation of causal effects of foreign policy choices on domestic support and assessment of reciprocity (qualitative and exact) by interacting own and other countries’ cost levels. Methods notes: Linear regressions with controls for gender, age, income, education, employment, and country fixed effects; survey weights applied; 95% CIs reported. Expectation measures about effectiveness and sustainability outcomes were aligned with treatment assignments. Pre-registration: AEA RCT Registry #AEARCTR-0004090. Data and code are available on Harvard Dataverse (Excel, Stata 16, R used).

- Study 1 (vignette experiment): In the control (no effectiveness info), about 53% supported a unilateral carbon tax (not statistically different from 50%). Mentioning that other major economies also implement a carbon tax increased support to about 59% (+6 percentage points; ~+11% over unilateral baseline), yielding a statistically significant majority. Under high-effectiveness information, support rose from ~57% (unilateral) to ~62% (multilateral) (+5 pp). When effectiveness was framed as low, switching from unilateral to multilateral had almost no effect. Eliminated effects analysis indicates that a substantial share of multilateralism’s impact operates through increased perceived effectiveness; ACDEs under fixed effectiveness are smaller, implying mediation by effectiveness beliefs. - Study 2 (multilateral conjoint): Domestic costs strongly reduce support: moving from low to medium costs lowers support by ~6 pp (France: €28→€56; Germany: €39→€77; UK: £15→£30; US: $3→$107), and moving to very high costs lowers support by ~17 pp (France: €118; Germany: €154; UK: £60; US: $213). Increases in other countries’ costs also reduce support, though less than domestic costs, suggesting sensitivity to the burden abroad (potentially via aversion to induced reciprocity costs or other-regarding preferences). Constant or decreasing multilateral cost paths are preferred over increasing paths. Higher mitigation investment by other countries increases support: when others invest at least 60% in mitigation, support increases by ~5 pp—sufficient to offset the support loss from foreign costs rising from low to medium. Reciprocity: Aversion to own costs diminishes when other countries incur higher costs (qualitative reciprocity). For example, when others’ costs are very low, increasing own costs to medium reduces support by ~7 pp; when others’ costs are very high, the same increase reduces support by ~5.5 pp. Exact reciprocity (matching own and others’ cost levels) yields additional positive interaction effects on support (average marginal interaction effects > 0). Overall, multilateralism boosts approval primarily via higher perceived effectiveness and reciprocity-consistent fairness norms.

Findings show that multilateral framing of climate policy increases public support for costly action by enhancing beliefs about policy effectiveness and aligning with reciprocal fairness norms. This addresses the research question by demonstrating that citizens’ willingness to bear costs depends on whether other countries also contribute. Policy implications: pursuing climate action through international cooperation can build broader domestic coalitions for carbon pricing; existing international interdependencies and institutions can be leveraged to activate conditional cooperation; and specific multilateral design features—such as constant/decreasing cost paths and high mitigation shares abroad—further increase support. The conjoint reveals public preferences over both domestic and foreign policy attributes and highlights the importance of reciprocity in mitigating cost aversion. These results challenge claims that mass support is independent of other countries’ policies and underscore the role of multilateralism in overcoming domestic political obstacles to decarbonization.

The paper demonstrates that multilateralism meaningfully increases public approval of costly climate policies. Two cross-national experiments reveal that: (1) multilateral framing increases support by around 5–6 percentage points on average, with effects concentrated when policies are perceived as effective; (2) public support is sensitive to foreign policy choices regarding costs, cost paths, and mitigation investments; and (3) reciprocity moderates cost aversion, with higher foreign effort reducing opposition to domestic costs and matched contributions yielding additional support. These contributions suggest that international cooperation can cultivate the public goodwill necessary for effective carbon pricing and other climate measures. Future research should extend the multilateral policy cognition framework to other sustainability domains (e.g., biodiversity, forest conservation, air pollution, energy transitions, waste, ocean acidification) and examine how institutional features, enforcement, and distributional compensations interact with reciprocity and effectiveness beliefs across diverse polities.