Previous research on climate change opinions has largely focused on individual or national levels, neglecting significant local variations within countries. This is problematic because effective climate change mitigation and adaptation strategies require local-level support and action. This study addresses this gap by examining CCOs at a fine-grained scale in Germany, a country with a complex history and diverse regional characteristics. The research questions are: 1. Do significant geographic variations exist in public CCOs within Germany? 2. What contextual factors explain the observed geographic differences in CCOs? The study's importance lies in its potential to improve our understanding of the factors driving climate change opinion divergence and inform the development of more effective, locally tailored climate policies. Germany offers a compelling case study due to its historical division, its role as a leader in climate change politics, and the presence of both strong environmental movements and right-wing populist parties that deny climate change. The study aims to determine whether similar patterns observed in the U.S. regarding geographic CCO polarization are also present in Germany, a country with distinct geographic, political, economic, and cultural characteristics.

Existing studies have shown substantial local-level variation in CCOs, particularly in the United States. However, these studies often focus on individual-level predictors projected to local scales, neglecting contextual factors. While sociodemographic characteristics like gender, age, education, income, and political beliefs predict individual CCOs, this alone doesn't explain differences between places with similar sociodemographic structures. Research on urban-rural divides in environmental attitudes offers mixed results, with some studies showing urban dwellers exhibiting greater climate change awareness. Similarly, the relationship between economic development and environmental concern is ambiguous. This study will consider both factors and explore additional contextual elements, such as local cultures, which can influence CCOs through the transmission of values and norms within communities. The study is novel in investigating contextual factors in a country other than the U.S., avoiding assumptions of generalizability from the US's unique political and social landscape.

The study utilizes data from the Green Socio-Ecological Panel (Green SOEP), a representative household survey conducted in Germany between 2012 and 2015. The survey includes detailed information on households' sociodemographic characteristics, geographical locations, and climate change opinions. Three dimensions of CCOs are assessed: belief in climate change, concern about its consequences, and perceived importance of collective responses. To account for the methodological challenges of analyzing public opinions at a fine-grained spatial scale, the study employs actor-based clustering with spatial discontinuities, which smooths data from 4,667 municipalities while considering spatial weights based on geographical proximity. This approach is validated through cross-validation and comparison with multilevel regression with post-stratification (MRP) at the regional level. Four contextual factors are tested using both spatial smoothing and MRP methods: urban vs. rural municipalities, prospering vs. declining municipalities, local green cultures (approximated by 1994 Green Party votes), and East vs. West Germany. Multilevel regression models are used to analyze the relationship between household-level and regional-level factors in predicting CCOs. The models incorporate individual-level predictors (gender, age, education, income, political affiliation, environmental concern) and regional-level predictors (urbanity, prosperity, green culture, East/West division). The dependent variable is a binary outcome indicating a positive response (belief, concern, and perceived importance of collective response).

The study reveals substantial geographic variations in CCOs across German municipalities. Maps illustrate significant clustering of opinions, with regional variations in the percentage of people who believe in climate change (71-89%), are concerned (44-70%), and perceive the importance of collective action (71-94%). Significant differences emerge between urban and rural areas, with urban areas demonstrating higher CCOs. A more pronounced difference is found between prospering and declining areas, with a 6.4% difference in overall CCOs between the most prosperous and most declining regions. This exceeds the urban-rural divide by a factor of three. The analysis confirms the importance of historical and cultural contexts. Municipalities with above-average Green Party votes in 1994 (indicating a strong green political culture) show significantly higher CCOs today. A striking East-West divide persists, with West Germany consistently exhibiting higher CCOs than East Germany across all three dimensions. Multilevel regression models demonstrate that individual-level factors (gender, education, political affiliation, environmental concern) are significant predictors of CCOs. Regionally, a positive association is found between prosperity and CCOs, while the East-West divide remains highly significant. Local green cultures in 1994 significantly predict CCOs in 2012-2015, with the impact being more pronounced for those without a personally held green attitude, suggesting a strong influence of local context on the general population's views. The East-West divide accounts for 53% of the variance in CCOs. Robustness checks controlling for selective migration after reunification do not change the results.

The findings extend previous research by demonstrating that geographic polarization in CCOs is not unique to the U.S., but also exists in Germany with its unique socio-political context. The study shows the importance of considering both socio-demographic and contextual factors when understanding local variations in climate change opinions. The findings highlight the interplay between individual attitudes and the broader regional context, which includes both geographical factors (urban vs. rural, prosperous vs. declining) and long-term cultural and political influences (Green political culture, East-West divide). The persistence of the East-West divide demonstrates the long-lasting influence of historical events and distinct political cultures. The strong influence of local green cultures suggests that the social context can shape climate change opinions even among individuals who do not personally identify with green ideologies. This has important implications for targeting climate change communication and policies.

This study reveals significant geographic variations in public CCOs within Germany, driven by a complex interplay of individual and contextual factors. The findings emphasize the importance of tailoring climate change policies to local contexts. Future research should investigate how extreme weather events influence CCOs at the local level and examine temporal changes in local opinion over time. Further analysis is also needed to elucidate the causal mechanisms behind the identified correlations, especially regarding the influence of historical and cultural factors.

The study's cross-sectional nature limits the ability to infer causality. While the data covers a four-year period, it does not capture changes in opinion that may have occurred due to subsequent extreme weather events. The use of 1994 Green Party vote share as a proxy for local green culture may not perfectly capture the full complexity of regional cultural values. The reliance on survey data may also introduce biases related to sampling and response rates.