Expected climate change consequences and their role in explaining individual risk judgments

Measurements The survey included an open-ended question so that respondents could answer in an unprompted and unrestricted manner based on their immediate and subjective associations, understandings, and feelings about the effects of climate change. The exact formulation was: ‘Climate change may affect different countries in different ways. What do you think will be the most important effect of climate change on [France/Germany/Norway/the UK]?’. Personal worry about climate change was assessed by asking respondents ‘How worried, if at all, are you about climate change?’, which could be answered on the following scale: 1 (Not at all worried), 2 (Not very worried), 3 (Fairly worried), 4 (Very worried), 5 (Extremely wor- ried). There were n = 28 missing responses to this question, including those choosing the answer options ’Don’t know’ or ‘Refuse to respond’; the latter option was included only in France. The question ‘Overall, how positive or negative do you think the effects of climate change will be on [France/Germany/Norway/the UK]?’ was used as a risk evaluation measure using the following scale: 1 (Entirely positive), 2 (More positive than negative), 3 (Neither posi- tive nor negative), 4 (More negative than positive) and 5 (Entirely negative). ‘Don’t know’ and ‘Refuse to respond’ were categorized as missing (n = 198). Respondents who answered ‘there is no such thing as climate change’ on an earlier question in the survey were not asked to answer the open-ended question or to indicate how positive or negative the impacts of climate change will be on their country.

Socio-demographic characteristics were included in the analyses based on literature identi- fying gender, age, and political orientation as relevant for explaining individual differences in climate change beliefs [see e.g., 29]. Gender (0 = Men, 1 = Women) and age (0 = 15–24, 1 = 25–34, 2 = 35–44, 3 = 45–54, 4 = 55–64, 5 = 65+) were treated as categorical variables. Political orientation was measured with the question: ‘In politics people sometimes talk of “left” and “right”. Using a scale from 0 to 10, where 0 means the left and 10 means the right, where would you place yourself on this scale?’ (n = 399 missing values, including the answers ’Don’t know’ and ‘Refuse to respond’). A summary of descriptive statistics and correlations among closed-ended questions can be found in the supporting information, S1 Table. Statistical analysis A coding scheme was developed in German and later translated into Norwegian, English, and French. Using content analysis [30], responses were sorted into six categories, five of which correspond to the causal levels described in Bo¨hm and Pfister’s [11] multi-level framework. The sixth category included answers reflecting expectations of hardly any impacts of climate change on one’s own country; for more details, see Table 1. Subcategories were used in some cases to characterize the responses more specifically or to differentiate between relevant char- acteristics. For example, a response categorized as Level 4 could either reflect a rather abstract answer such as ‘We hardly have seasons anymore’, or a more concrete answer specifying that

the consequence would influence plants ‘Crops like corn will benefit, others will yield less har- vest. Pests attacking plants will multiply’ (subcategory 1). Each response could be sorted into one, several, or none of the categories, depending on how many categories were expressed by the respondent. An overview of the coding scheme, including all subcategories, is provided in the supporting information, S2 Table. The coding was conducted independently by two native speakers of the country’s language. Intercoder reliability was assessed as proportion agreement, calculated as the number of cod- ings where the two coders gave the same code divided by the total number of codings. The interrater reliability was 0.924 for the UK, 0.972 for Norway, 0.982 for Germany, and 0.976 for PLOS ONE Expected climate change consequences and individual risk judgments PLOS ONE | https://doi.org/10.1371/journal.pone.0281258 February 15, 2023 4 / 15