International climate adaptation assistance: Assessing public support in Switzerland

PLOS ONE | https://doi.org/10.1371/journal.pone.0317344 February 12, 2025 18 / 26 PLOS ONE Support for adaptation assistance in Switzerland preference for the Philippines thus deserves additional exploration. Filipino residents in Switzerland do not constitute a large community (6,423 individuals, [92]). Still, similarly to Algerians, they are more present than Bangladeshi and Kenyan citizens (around 1,704 and 1,853 respectively, [92]). Whether Swiss residents hold particularly positive social represen- tations of the Filipino community, and whether this explains our results, however, should be researched further. Because of the multiple factors encompassed by geographical location and the absence of a direct test of the proposed mechanisms that would allow for disentangling them, additional studies are needed to understand what specific features of the recipient country drive citizens’ favourability of policy packages. Interestingly, however, our results suggest that the identity of the recipient country influences policy favorability only in specific subsets of the sample (Supporting information S8 Fig, S9 Table, S10 Fig, S11 Table, S12 Fig, S13 Table). Indeed, the significant positive effect for the Philippines and the negative effect of Algeria disappears amongst young respondents (18–34 years old), amongst highly educated ones, and respondents with medium to high incomes. In other words, the results are insignificant for each of the four recipient countries in these sample sub-groups. This may imply that racial and cultural preju-

dices are lower among the younger generation and demographics with higher education and income (Supporting information S4 Fig, S5 Table, S6 Fig, S7 Table, S8 Fig, S9 Table). At the same time, however, the position of respondents on the political spectrum (Support- ing information S23 Fig, S24 Table), their beliefs about integration (Supporting Information S12 Fig, S13 Table), and their pride towards Switzerland do not significantly alter our main results in this respect (Supporting Information S14 Fig, S15 Table). Benchmarking and support for adaptation assistance Recall that the conjoint experiment is nested within a framing experiment. Might the respon- dents’ favorability of aid or immigration be conditioned by benchmarking with other OECD countries (Treatment 1), or with Switzerland’s existing policies (Treatment 2)? After all, the general public may not have deep knowledge about actual aid and immigration levels within Switzerland or in its peer countries. Thus, providing context might reduce the knowledge gap and shape the policy preferences of respondents. We find no significant differences across treatment groups. In further analysis, the interaction of aid and immigration with treatment dummies does not produce statistically significant results (Supporting Information S16 Fig). This is broadly consistent with Uji et al. [26] which reports that benchmarking with respect to past Japanese policy and peer G7 countries did not produce statistically significant effects. Conditional effects of political ideology and cultural beliefs

Finally, to further explore the robustness of the main findings of our experiment, we also explored the conditional effect of political ideology and cultural beliefs as factors likely to influence the favorability of policy packages, given the highly contentious immigration and climate policy debates. Studies of climate politics in Switzerland note the consistent divide between right-wing and left-wing coalitions on climate policies [72–74,95]. Research on pub- lic opinion and foreign aid has similarly emphasized the variation across the left-right political ideology spectrum [23,24,46]. While conservatives tend to oppose migration, we do not find political ideology to condition the favorability of policy packages either for the aid or immi- gration dimensions of the policy package (Supporting Information S23 Fig, S24 Table). Finally, we examine how individual beliefs might condition preferences for policy pack- ages, even after controlling for ideology. Consider the interaction effects of cultural views based on consideration of language affinities (Supporting Information S10 Fig, S11 Table),