Climate change significantly increases vulnerability to rising sea levels, higher temperatures, and altered weather patterns. Even optimistic scenarios limiting global warming to 1.5°C cannot fully avoid these challenges. High-risk, densely populated areas must adapt by investing in mitigation strategies. A crucial concern is the adaptation gap – the discrepancy between current adaptation levels and those needed to effectively reduce climate risk impacts. Understanding this gap is essential for informing effective policy. Data scarcity has historically hindered systematic research on this gap; details on existing infrastructure and future adaptation plans are often unavailable at the city level. Existing literature relies on case studies, surveys, or socioeconomic proxies for adaptation. This study addresses this limitation by analyzing a large dataset of financial disclosures to quantify adaptation gaps and identify associated factors.

Prior research on urban climate adaptation has been limited by data availability, relying heavily on case studies, surveys, or socioeconomic indicators to approximate adaptation levels. Studies have highlighted the importance of various factors in adaptation, including behavioral changes (influenced by political beliefs and risk perception), financial resources (as adaptation measures can be expensive), and good governance (manifested in long-term planning horizons). This study builds upon this existing research by utilizing a novel data source (financial disclosures) and a large sample of US cities to empirically test the relationship between these factors and the adaptation gap.

The study uses a unique dataset of 8726 hand-collected financial disclosures from 431 US cities between 2013 and 2020. A city-specific adaptation dictionary was created and validated to capture hard and soft adaptation strategies for flood risk through textual analysis of budgets, annual comprehensive financial reports (ACFRs), and bond prospectuses. The main adaptation measure is the number of adaptation-related sentences per city-year, chosen due to its strong correlation with adaptation activities. The study combines this textual data with independent flood risk assessments, hand-collected partisanship data, and financial and socioeconomic city characteristics. Several validation tests were conducted to ensure the accuracy of the adaptation measure, including correlations with capital improvement expenses, insurance discounts, and municipal bond spreads. The adaptation gap is defined as an indicator variable, equaling one when a city's actual adaptation level is lower than predicted based on its flood risk. Regression analyses were used to examine the relationships between the adaptation gap and factors like political affiliation (Republican vs. non-Republican cities), financial constraints (measured by unrestricted-fund-to-expense ratio and total debt per capita), and planning horizons (measured by capital budget outlook). Further analyses explore the heterogeneity in adaptation gap distribution across cities with different flood risks, constituent beliefs, state grant availability, and household incomes.

The study reveals significant variation in adaptation levels across US cities. While a positive correlation exists between flood risk and adaptation, a substantial adaptation gap is observed: over half of high-risk cities show less adaptation than predicted based on their flood risk exposure. Contrary to expectations, political affiliation (Republican vs. non-Republican) shows little association with the overall adaptation gap, except for a slightly higher likelihood of a soft adaptation gap in Republican-led cities. However, financial constraints and shorter planning horizons are strongly associated with a higher probability of an adaptation gap. Cities with one standard deviation lower unrestricted-fund-to-expense ratio are 6.6% more likely to have an adaptation gap. Similarly, cities with a one-year shorter budget planning horizon are 4% more likely to have an adaptation gap. These effects are more pronounced in high-flood-risk areas and counties where constituents strongly believe local officials should address climate change. State grants and high local household income do not fully offset these constraints. The explanatory power of financial constraints on the adaptation gap diminishes over time, whereas the effect of short planning horizons persists.

The findings highlight the importance of addressing financial constraints and short-term planning horizons to close the adaptation gap. The lack of a significant relationship between political affiliation and the adaptation gap suggests that even in cities where climate change skepticism is prevalent, practical actions are taken to mitigate flood risks to protect the local economy and residents. This contrasts with the more prominent partisan divide often observed in climate change mitigation discussions. The study emphasizes the need for strategies to alleviate financial constraints, including exploring private financing and public-private partnerships, and for implementing longer-term planning mechanisms to overcome myopic decision-making. The finding that adaptation mitigates the flood risk premium in municipal bonds suggests that investments in adaptation can improve a city’s financial standing, potentially creating a positive feedback loop.

This study demonstrates the prevalence of an adaptation gap in US cities, emphasizing the need to address financial constraints and short planning horizons. While political affiliation does not appear to be a major driver of this gap, financial limitations and short-term planning significantly hinder adaptation efforts. Future research should focus on establishing causal links and exploring additional factors contributing to the gap. The methodology presented here, using textual analysis of financial disclosures, offers a scalable approach for broader studies on urban climate adaptation across diverse geographical contexts.

The study focuses primarily on flood risk adaptation, neglecting other climate-related risks (e.g., extreme temperatures, droughts, wildfires). Certain adaptation measures, such as zoning and building codes, are difficult to quantify through textual analysis. The analysis does not establish definitive causal relationships between the identified factors and the adaptation gap. Future research should explore these limitations and conduct causal inference analyses.