How tailored climate information affects attitudes towards climate policy and psychological distance of climate change

National-level climate policies require broad public support grounded in perceptions of risks, needs, and fairness of policy paths. Psychological distance (PD)—spatial, temporal, social, and hypothetical—has been posited to influence climate engagement and policy support, yet evidence is mixed. People in lower-exposure countries often perceive climate change as psychologically distant, potentially reducing engagement. Communication strategies that localize and personalize climate information may reduce PD and increase support, but some studies suggest reduced PD does not necessarily increase policy support. In Finland, a climate policy pioneer aiming for carbon neutrality by 2035, understanding links between tailored climate impact information, PD, and policy attitudes is needed. This study introduces an interactive tool presenting localized climate scenario impacts and personalized vulnerability, and assesses effects on knowledge, PD, and policy attitudes in a representative Finnish sample. Research questions: RQ1: How do Finns regard and actually perform on knowledge about local climate impacts under different emission scenarios, and is knowledge linked to attitudes? RQ2: What is the level of PD to climate change in Finland and differences by demographics? RQ3: Does tailored climate scenario information affect (a) knowledge, (b) PD, and (c) attitudes toward climate policies?

Prior work links climate risk perception to policy support but finds inconsistent evidence on PD’s role. Reviews report mixed findings on whether climate change is perceived as distant and whether decreasing PD increases engagement or support. Proposed communication strategies include localizing impacts, emphasizing global identity, concretizing local consequences and solutions, tailoring messages to risk type, and using immersive media. Some studies (e.g., Mildenberger et al., Schuldt et al.) show personalized or localized risk messages can sometimes reduce concern or fail to boost support. Keller et al. recommend focusing PD research on specific contexts (impacts, policy support, behaviors). In Finland, surveys show high concern and support for mitigation but varying support for instruments and limited urban–rural differences; higher risk perception predicts support for some policies. A gap remains on how tailored climate information affects PD and policy attitudes in Finland.

Design: Mixed-methods study combining climate scenarios (downscaled), an interactive visualization tool producing personalized outputs, and a national survey administered before and after tool use to measure knowledge, PD, and attitudes. Climate scenarios and models: Used Climate Impact Lab GDPCIR downscaled CMIP6 outputs (0.25°) for precipitation and daily Tmin/Tmax under SSP1-2.6 and SSP2-4.5, with trend-preserving bias correction. Ensemble included FGOALS-g3, INM-CM5-0, ACCESS-CM2, ACCESS-ESM1-5, BCC-CSM2-MR, CMCC-CM2-SR5, CMCC-ESM2, EC-Earth3, GFDL-ESM4, HadGEM3-GC31-LL, MIROC-ES2L, MIROC6, MPI-ESM1-2-HR, NESM3, NorESM2-MM, UKESM1-0-LL, CanESM5; ensemble means computed. Air quality (PM2.5) projections from ECHAM-HAMMOZ (1.9°). Data accessed via Microsoft Planetary Computer. Metrics displayed: Changes from present (2020 average of 2018–2022) to 2040 (2035–2045) in (1) heat wave days (streaks above 22 °C; Finland averages: +18% SSP1-2.6, +24% SSP2-4.5), (2) heavy rain days (>15 mm; +14% SSP1-2.6, +16% SSP2-4.5), (3) duration of winter (days <0 °C; −5% SSP1-2.6, −7% SSP2-4.5), (4) PM2.5-attributable disease burden (DALYs) using RR1=1.062 per 10 µg/m³, BoD=132,000 per million; average change by 2040: −37% SSP1-2.6, 0% SSP2-4.5, (5) household electricity consumption for heating (proxy for heating demand; −3% SSP1-2.6, −10% SSP2-4.5). Vulnerability metric: Aggregated (geometric mean, equal weights) across seven factors scored 1–6: fraction of artificial surfaces within 2 km, flood risk area status (location-based), age, income, education, pre-existing medical conditions, and social isolation. Data sources: CORINE land-use and Finnish Environment Institute flood risk datasets (CC BY 4.0). Vulnerability displayed on a qualitative scale from very low to very high. Interactive tool: Web tool (https://www.climateguide.fi/articles/how-could-climate-change-affect-you/) where users input location (Finland), birth year, education, income, medical conditions, and living alone status. Output: radar chart comparing five metrics for SSP1-2.6 vs SSP2-4.5 by 2040 relative to present, plus a color bar indicating personal vulnerability; numerical descriptions accompany graphics. Survey: National sample representative of Finnish population (age, gender, geography; ages >15) recruited by Online Research Finland Ltd; fielded 29 June–18 July 2023 (Finnish; translated to English). N=1017; birth years 1942–2007 (Mo=1971); 142 distinct domiciles (37.6% urban); income in eight categories (<€10k to >€100k; mode €30–40k), 40.7% university educated, 32.6% with relevant medical condition, 38.9% living alone, 49.7% men, 49.8% women, 0.5% other. Procedure: Baseline survey (demographics; claims on climate/policy/PD; two-question knowledge test on expected direction/magnitude for five metrics under increasing vs decreasing emissions), then use interactive tool, then immediate post-tool survey repeating claims. Measures: Claims mapped to aspects (knowledge, PD spatial/social/temporal, attitudes). Likert responses: 1=fully disagree, 2=partly disagree, 3=partly agree, 4=fully agree, 0=no opinion. Knowledge test scoring: correct direction=1 point; correct direction+magnitude=2 points (max 20). PD indicators computed from claim responses onto a 4-step scale from proximal to very distant for spatial, social, temporal dimensions; total PD is average across dimensions. Analysis: Percent shares and changes, means and changes, linear regression, paired samples t-tests (e.g., PDsocial and total PD before vs after). Thematic analysis of open-ended responses to classify recurring themes (changes induced by tailored info; additional information needs).

Sample and tool usability: N=1017; 81% agreed the tool was clear/illustrative. Some (n=131) reported technical issues (e.g., tool not opening/firewall), noted qualitatively. Knowledge and perceived knowledge: Before tool, 72.5% partly/fully agreed they know how climate change will affect them locally by 2040 if emissions increase; after tool, 80.6%. “Fully agree” increased by 50.5% relative (+10.5 percentage points). Actual knowledge test average score was 6.4/20, with no significant demographic differences. Lowest average score (3.5/20) among those strongly agreeing they know local impacts and strongly disagreeing that science justifies mitigation; highest (8.0/20) among those strongly disagreeing they know local impacts and strongly agreeing that science justifies mitigation. Most relevant metrics: Heat wave days was most relevant for 34.0% of respondents; PM2.5-attributable diseases for 25.8% (also the metric with largest improvement under SSP1-2.6); electricity consumption for 19.5%; duration of winter for 14.2%; heavy rain days for 6.6%. Attitudes toward policy: Overall, the tool reduced “No opinion” responses and modestly shifted some attitudes. - Scientific basis for mitigation: Stronger agreement correlated with higher education; no tool effect highlighted. - Justified to advance climate action in Finland: No significant changes before vs after; no urban–rural or demographic differences. - Finland’s 2035 carbon neutrality goal justified: “Fully agree” increased by 3.9 percentage points overall after the tool; among the youngest, “Partly disagree” rose by 11.4 percentage points. Larger positive shift among men than women. - Policies harmful to my life: Among age>65, “Fully disagree” most popular (27%) and unchanged post-tool. Among younger respondents, “Partly agree” was most popular (31%); both “Partly agree” and “Fully agree” were reinforced post-tool. Agreeing views increased by up to 11 percentage points among high-income (>€70,000) respondents. Urban vs rural: urban respondents shifted more to “Fully agree,” whereas rural showed a shift toward “Partly disagree,” indicating tailored information made wealthy/urban respondents view policy impacts more negatively while rural shifted oppositely. - Policies can alleviate/mitigate local changes: Majority partly/fully agreed across ages; “Fully agree” increased by 6.9 percentage points overall, with larger increases among age<30 (+8.5 pp) and urban dwellers (+7.4 pp); no change among rural dwellers. Psychological distance (PD): - Spatial PD: Majority perceived impacts as proximal or hardly distant; minimal change after tool (slight shift from “proximal” to “hardly distant”). Elderly (>65) showed majority in “proximal.” - Temporal PD: 34.4% viewed impacts as fairly distant; age<30 had 43.8% fairly distant; age>65 had 29.3% temporally proximal. After tool, slight increases of +1.5 percentage points in both “fairly distant” and “very distant.” - Social PD: Significant increase after tool (paired t-test, p<0.001); pre-tool 34.5% socially proximal. Post-tool, responses shifted from “proximal/hardly distant” to “fairly/very distant.” PDsocial increased by 10% on average and up to 15% among those born 1990–1999. Urban respondents remained more socially proximal than rural; rural PDsocial increased by 15% post-tool. Respondents with basic diseases perceived higher social distance than those without; PDsocial increased by 12% among those without medical conditions, and remained effectively unchanged among those with conditions. - Total PD: Increased significantly overall after tool use (paired t-test, p<0.001), with increases in “fairly” and “very distant” classes. Total PD increased in all age groups except 1950–1959 cohort, in all income groups, in all education levels except the lowest, and among those without basic diseases. Urban total PD increased by ~3% on average; rural by ~11%. Open responses: Effects of tailored info (n=328): More info/reinforcement (43.7%); surprised by results (42.4%); motivation/empowerment (7.0%); emotions (6.1%); increased skepticism (0.9%). Information needs (n=625): Global perspective/Finland vs others (26.0%); objective/unpolitical truth (14.4%); socio-economic impacts (13.8%); details on actions & impacts (12.2%); impacts on nature (10.5%); nothing—expressed skepticism (10.5%); fairness (3.9%); nothing—formed opinion (3.3%); different time span (2.2%); emotional rhetoric (1.1%); health impacts (1.1%); positive impacts/motivation (1.1%).

Tailored, localized climate scenario information increased respondents’ confidence and understanding of climate impacts but produced nuanced and demographic-dependent effects on policy attitudes and PD. While the tool generally reduced “no opinion” responses and increased perceived potential for policies to alleviate local changes, some groups—particularly younger and high-income respondents—interpreted tailored information as indicating more harmful or negative policy impacts. PD findings indicate that spatial proximity is already widely perceived, whereas temporal and especially social dimensions are more variable and sensitive to tailored information; PDsocial increased significantly post-tool, notably among younger and healthier respondents and rural residents. These patterns suggest vulnerability factors (older age, lower income, medical conditions, urban domicile) can make tailored information feel personally relevant and closer, while low-vulnerability individuals may experience increased distance. The results align with prior research indicating that personalizing/localizing climate risks is not a guaranteed path to higher engagement or policy support. Effective communication may need to balance personalized local impacts with global context and comparative perspectives to avoid inadvertently increasing psychological distance among low-vulnerability groups.

This study introduces a reproducible approach that combines downscaled climate scenarios, a personalized vulnerability metric, and an interactive tool to examine effects on knowledge, psychological distance, and climate policy attitudes in Finland. Tailored information improved perceived and actual knowledge (though test scores remained modest) and increased perceived potential of policies to alleviate local changes. However, effects on attitudes and PD varied by demographics and vulnerability: low-vulnerability groups tended to experience increased social/temporal distance and, in some cases, more negative views of policy impacts. Tailored climate information is thus not a panacea for reducing PD or boosting policy support. Future research should incorporate control groups (e.g., global-only information and combined tailored+global information), enable comparative results across locations and countries, and include broader metrics (socio-economic and biodiversity impacts) to better understand how information content shapes PD and policy attitudes. Replication in regions with larger projected impacts could test generalizability.

Key limitations include the absence of a control group (e.g., receiving only global information or both tailored and global information), limiting causal attribution of tailored information effects; the specific selection of metrics in the visualization tool, which might have influenced responses (e.g., some participants requested more information on nature/biodiversity impacts); and the context of Finland where projected impacts are relatively modest, potentially contributing to increased PD among low-vulnerability participants. Equal weighting in the vulnerability aggregation may not reflect locally optimal weights. Some respondents experienced technical issues with the tool display. Findings are based on a Finnish sample and may not generalize without replication in other contexts.