IntroductionClimate change is no longer a distant threat; its impacts are reaching all corners of the globe, including affluent countries1,2. In such countries, communities are increasingly confronting severe climate-related challenges-frequent floods, rising sea levels, and extreme weather events are particularly impacting economically disadvantaged areas, given the major levels of inequality within some wealthy nations. This reality highlights that climate vulnerability is not confined to low-income or developing regions; it extends to vulnerable communities within wealthy countries as well3,4.In this context, decreasing vulnerability is of vital importance. It is not merely about bouncing back from adverse events but about developing the capacity to adapt to and mitigate these impacts, thus reducing overall vulnerability. Understanding which climate policies effectively enhance resilience is necessary. This study aims to identify which types of policies are most successful in decreasing vulnerability, ensuring that resilience strategies are both impactful and equitable across diverse communities.In response to the escalating effects of climate change, many communities are adopting various climate adaptation laws aimed at mitigating risks and enhancing resilience5. However, the effectiveness of these policies can vary significantly6. This study highlights that while some policies are widely implemented, their impact on increasing climate resilience is not uniform. Understanding which types of policies are most effective is essential for guiding future legislation and maximizing the benefits of climate adaptation efforts.To better understand the dynamics between climate change, policy interventions, and societal impacts, this research is informed by the Social-Ecological Systems (SES) theory7,8. SES theory emphasizes the interconnectedness of social and ecological components within a system, highlighting how human actions and policy decisions can influence the resilience of both natural and human systems to environmental stressors9. By applying the SES perspective, this study examines how different climate policy features and implementation levels affect the resilience of New England’s communities, considering both their social and ecological dimensions.From an SES perspective, policies that enhance the adaptive capacity of both social and ecological systems, such as those supporting infrastructure improvements and ecosystem restoration, are expected to reduce vulnerability. Evidence suggests that enhancing social dimensions, including community engagement and local governance, contributes significantly to resilience in social-ecological systems10,11. Conversely, policies that disrupt these interconnections or fail to account for them may inadvertently increase vulnerability12. Understanding which climate policies effectively enhance resilience is vital. In this context, promoting public participation in decision-making processes and fostering cooperation among stakeholders are essential strategies13,14.Despite the growing body of literature on climate adaptation policies15,16,17,18,19,20, several gaps remain. Quantitative analyses that comprehensively assess the real-world impacts of a broad spectrum of climate policies across a specific region, like New England, are still limited. Moreover, the differential effectiveness of specific policy features and implementation levels in reducing social vulnerability, as measured by indices like the SVI, requires further empirical investigation. This study contributes to the literature by providing a large-scale quantitative evaluation of over 1200 policies in New England, examining the nuanced impacts of specific policy characteristics on social vulnerability using robust econometric methods, and offering insights into the effectiveness of policies targeting different climate hazards in a specific regional context.This study aims to identify which types of policies are most successful in decreasing vulnerability, ensuring that resilience strategies are both impactful and equitable across diverse communities. The implications of these findings underscore the need for integrated approaches that consider both ecological and social factors while highlighting the importance of robust institutional frameworks that support these interactions. Overall, the synthesis of SES theory with climate policy analysis presents a comprehensive framework for addressing the multifaceted challenges posed by climate change.To understand how, who, what, and when climate policies can effectively enhance resilience within the interconnected social and ecological systems of New England, this study seeks to address the following questions, guided by the principles of Social-Ecological Systems (SES) theory:Policy Features and SES Resilience: How do different features and types of climate policies, such as those impacting infrastructure or ecosystem health, influence the resilience of New England’s social-ecological systems?Jurisdictional Impact on SES Vulnerability: Who—in terms of implementation jurisdiction (state, local, tribal, etc.)—plays the most effective role in reducing social vulnerability within the context of New England’s interconnected social-ecological systems?Climate Goals and SES System Stability: What specific climate goals or focuses, considering the dynamics of social-ecological interactions, appear to be most effective in maintaining stability and reducing vulnerability in New England’s systems?To illustrate the current landscape of climate resilience, Fig. 1 displays vulnerability levels across New England. This map reveals significant variations in resilience across the region, underscoring the necessity of effective and targeted climate policies. By analyzing these variations, this work aims to uncover which policy characteristics contribute most to reducing vulnerability and enhancing resilience.Fig. 1: Vulnerability scores in New England states by county.The mapped data displays vulnerability scores across New England, sourced from the Social Vulnerability (SVI) index42. While the map shown here indicates vulnerability scores, Supplementary Note 1 shows how environmental resilience explicitly maps onto the county level.Full size imageThe primary objective of this paper is to analyze how various climate policies, including their features, types, levels of implementation, and focuses, influence the reduction of vulnerability across New England. Based on these insights, actionable recommendations are provided to guide policymakers in developing more effective climate adaptation strategies and enhancing community resilience against climate change.As the impacts of climate change intensify globally, the development and implementation of effective adaptation policies have become paramount for reducing community vulnerability. While climate change adaptation has garnered significant attention in recent decades, leading to a burgeoning body of literature, gaps remain in our empirical understanding of policy effectiveness21,22. Existing research on climate adaptation policies has explored diverse facets of this challenge. For instance, numerous studies have focused on the theoretical underpinnings of adaptation planning, emphasizing frameworks such as adaptive governance and resilience-building strategies23,24. A substantial portion of the literature, often employing qualitative case studies and policy analyses, has documented various local, regional, and national adaptation initiatives, highlighting best practices and common challenges in policy implementation. For example, integrated coastal zone management (ICZM) has been widely regarded as an effective tool for addressing coastal vulnerabilities, providing examples from various geographic contexts, such as the Mediterranean Sea, where the integration of science and policy has shown promise in enhancing coastal resilience25,26. Furthermore, researchers have emphasized the need for innovative methodologies to assess the ecological and economic impacts of coastal management initiatives22, reinforcing the importance of understanding local adaptation measures.Furthermore, researchers have investigated the socioeconomic and environmental factors influencing vulnerability to climate impacts, often utilizing composite indices to map and understand susceptible communities. For instance, studies have employed the Social Vulnerability Index (SVI) to quantify social vulnerability to natural hazards, demonstrating its utility in various assessments27. Efforts have also been made to categorize and assess different types of adaptation measures, from infrastructure development to ecosystem-based approaches, and their potential efficacy28. The integrated approach to planning and management of coastal areas highlights the effectiveness of localized strategies in addressing the unique challenges posed by climate change and the necessity for a diverse range of tools and methods to enhance adaptive capacity29,30.Despite these valuable contributions, a significant gap persists in comprehensive, quantitative analyses that rigorously assess the real-world impacts of a broad spectrum of climate policies across specific regions. While qualitative studies offer rich contextual insights, they often lack the generalizability and statistical power needed to identify causal relationships between policy features and vulnerability reduction across diverse settings23,31. This analytical gap is particularly pronounced in regions like New England, where detailed, large-scale quantitative evaluations of adaptation policy outcomes remain limited. Without such robust empirical evidence, policymakers face challenges in discerning which specific policy characteristics are most effective, where to strategically allocate resources, and how to design interventions that genuinely reduce vulnerability24. This study directly addresses this gap by providing a large-scale quantitative evaluation of over 1200 adaptation policies implemented across New England.This research is scientifically relevant for several key reasons. First, by employing robust econometric methods, including Fixed Effects regression models and Staggered Treatment Difference-in-Differences (DID), the literature moves beyond descriptive accounts to statistically isolate the causal effects of various policy features and implementation levels on social vulnerability27. This rigorous approach yields generalizable insights that are vital for evidence-based policymaking24. Second, the application of SES theory provides a sophisticated theoretical lens, recognizing the complexities between human actions, policy interventions, and environmental systems32. This framework enables us to understand not just whether policies reduce vulnerability, but how they do so within complex, interconnected systems, offering a more holistic understanding than approaches that might compartmentalize social and ecological factors26.Third, by focusing on New England, a region particularly susceptible to diverse climate hazards such as sea-level rise, extreme weather events, and inland flooding, these findings offer guidance for regional stakeholders. Identifying which policies are most successful in decreasing vulnerability in this specific context is paramount for developing impactful, equitable, and locally relevant resilience strategies that can serve as models for other vulnerable regions21,28.To interpret the quantitative findings in a way that reflects the complexity of real-world adaptation efforts, it is essential to anchor them within the robust theoretical framework of SES, as it offers such a foundation by highlighting the interconnections between human institutions and ecological processes. Especially in the context of climate law and resilience, SES helps unpack how social and environmental systems interact, providing a powerful lens through which to evaluate whether policies are truly reducing vulnerability and enhancing adaptive capacity.At the heart of the SES framework is the idea that changes in one part of the system (e.g., environmental degradation) can trigger ripple effects in another, like increased socioeconomic instability33. For example, government decisions around resource use and environmental protection directly impact ecological conditions, which then affect how communities cope with climate-related stressors34. This back-and-forth relationship shows why it is necessary to evaluate climate laws not just by their environmental outcomes, but also by their effects on people (especially those most vulnerable to climate shocks)35.The SES framework also gives us tools to explore what makes a system resilient. Such tools may include adaptability, access to resources, and the capacity to recover from disruptions36. By breaking down elements of climate policy (e.g., regulations, community engagement efforts, and economic incentives), we can better understand which strategies actually reduce vulnerability. For instance, strong governance that includes diverse voices can enhance a community’s ability to respond and adapt to climate impacts37. The framework also highlights the importance of ecosystem services (e.g., clean water, food, protection from natural hazards), which are often lifelines for communities38.To apply the SES framework in practice, researchers can analyze climate laws by identifying their key components and tracing how these influence social and ecological outcomes. This often involves building statistical models that can tease apart the relationships between policy features and indicators of resilience39. Such analysis can reveal not only which laws are working, but why they are effective (or not).In short, the SES framework helps us see the full picture: how climate laws operate within complex systems of people and nature, and what that means for building resilience. It offers a grounded, flexible approach for evaluating policy impacts and ultimately supports the design of more effective, inclusive, and sustainable climate responses33,36,40.ResultsTo assess how different climate policies influence vulnerability reduction, a series of Ordinary Least Squares (OLS) Fixed Effects regression models are employed. It is worth noting, however, that this indicates short-term vulnerability with the lag rather than long-term effects, which may tell a different story. Additionally, this study does not claim to measure the effectiveness of specific policies but rather examines how these policies influence vulnerability on a broader scale. These models analyze the impact of policy features, types, levels of implementation, and goals on decreasing vulnerability (detailed model specifications are available in the “Methods”).This analysis draws on the Resilience and Adaptation in New England (RAINE) database, which includes data on 1232 policies and plans from Massachusetts, New Hampshire, Maine, Connecticut, Rhode Island, and Vermont, spanning from 2000 to 202341. Additionally, the Social Vulnerability Index (SVI) from the CDC and ATSDR is integrated, providing county-level data on factors affecting vulnerability for the years 2000, 2010, 2014, 2016, 2018, 2020, and 202242. This combined dataset creates a panel of 476 county-year observations, allowing for the evaluation of the effectiveness of climate policies in reducing vulnerability across New England.How effective are different policies? Dissecting the features and types that shape vulnerabilityTo understand how various policy features affect vulnerability levels, the top five most frequent policy features are analyzed: Economic Resilience, Ecosystem and Natural Resources, Government Bylaws and Ordinances, Infrastructure Built, and Social and Environmental Justice (for detailed definitions of each policy characteristic throughout the analysis, see the “Data” section) The findings indicate that policies incorporating Government Bylaws and Ordinances and Infrastructure Built are associated with a decline in vulnerability levels (see Fig. 2).Fig. 2: The impact of policy features on climate vulnerability.The coefficient plot illustrates the results of the “Feature" models. It combines data from the Resilience and Adaptation in New England (RAINE) database, which details the stock of climate policy features for each county-year, and the CDC/ATSDR Social Vulnerability Index (SVI), which measures vulnerability levels in the same communities. In this plot, negative coefficients indicate a reduction in vulnerability to climate hazards (see Supplementary Note 2 for the corresponding table).Full size imageConversely, the inclusion of Economic Resilience in policies did not show a statistically significant impact on vulnerability reduction. Unexpectedly, policies featuring Social and Environmental Justice were associated with an increase in vulnerability43. This suggests that while policies aimed at promoting justice are well-intentioned, current implementation may not effectively address or mitigate vulnerabilities as intended.The observed results highlight several intriguing dynamics in how different policy features impact vulnerability. Policies that focus on Government Bylaws and Ordinances, and Infrastructure Built tend to effectively reduce vulnerability. This outcome is likely because these features address fundamental aspects of climate adaptation-such as enhancing natural defenses, creating robust legal frameworks, and improving infrastructure resilience-that directly contribute to a community’s ability to withstand climate impacts4,44,45,46,47.The increase in vulnerability linked to Social and Environmental Justice policies could result from inadequate implementation or resources, indicating that these policies, despite their equity focus, may not yet be effectively translating into tangible resilience improvements48,49. Alternatively, the challenges faced by vulnerable communities may be so complex that policies focusing solely on justice without comprehensive support may inadvertently fail to reduce vulnerability.To exemplify these patterns substantively, policy characteristics that contribute to increased vulnerability primarily include social and environmental justice features, certain types of policy plans, and specific policy goals. Social and environmental justice-oriented policies, while designed with long-term equity in mind, can sometimes exacerbate vulnerability due to poor implementation, lack of funding, or weak enforcement. For instance, the Guidance Policy for Considering Environmental Justice by Rhode Island’s Department of Environmental Management integrates environmental justice into contaminated site remediation50. It emphasizes stakeholder engagement and policy development to ensure fair community participation51. However, these policies often take longer to show measurable outcomes, as they prioritize procedural justice over immediate risk reduction, leaving contaminated sites hazardous for longer periods52. Additionally, while such policies aim to empower communities, their focus on equity and long-term outcomes can inadvertently lead to unintended consequences, such as rising housing costs, displacement, or increased regulatory burdens53. These outcomes can worsen conditions for vulnerable populations in the short term54. Political resistance, bureaucratic challenges, and misalignment with local community needs further hinder the effectiveness of SEJ policies, making it difficult for them to reduce vulnerability at scale55. In contrast, infrastructure policies, shown to decrease vulnerability in this analysis, tend to provide more immediate benefits through tangible physical improvements, which may make them appear more effective in the short run56.To decrease vulnerability through climate-resilient infrastructure, the Connecticut Adaptation Resource Toolkit (CART) equips local governments with essential adaptation resources, prioritizing infrastructure and built environment solutions57. By providing planning frameworks, funding opportunities, legal guidance, and technical support, CART enables municipalities to implement resilience-focused projects, such as strengthening clean drinking water infrastructure against coastal flooding. These tangible investments enhance adaptive capacity and reduce climate-related risks.The analysis of different plan types-Adaptation Plans, Case Study Implementations, Climate Mitigation Documents, Disaster Recovery Plans, and Resilience Plans-reveals varying effects on vulnerability, as illustrated in Fig. 3. For Adaptation Plans and Resilience Plans, the results indicate no statistically meaningful relationship with vulnerability.Fig. 3: The impact of plan types on climate vulnerability.The coefficient plot illustrates the results of the “Plan Type” models. It combines data from the Resilience and Adaptation in New England (RAINE) database, which details the stock of climate policy and plan types for each county-year, and the CDC/ATSDR Social Vulnerability Index (SVI), which measures vulnerability levels in the same communities. In this plot, negative coefficients indicate a reduction in vulnerability to climate hazards (see Supplementary Note 3 for the corresponding table).Full size imageClimate Mitigation Documents show an increase in vulnerability. This may occur because these documents often prioritize reducing greenhouse gas emissions on a broad scale, which can overlook the immediate needs of the most vulnerable communities. Similarly, Disaster Recovery Plans are found to increase vulnerability, as alluded to in previous literature58. These plans focus primarily on managing the aftermath of disasters rather than preventing them or addressing underlying vulnerabilities. This approach might effectively ‘pocket’ funds for future recovery needs, potentially neglecting immediate risks and inadvertently leaving communities more exposed to subsequent events.Case Study Implementation plans (which provide an in-depth examination of a situation, report, project, or plan41) are the only plan type associated with a decrease in vulnerability. This effectiveness stems from their context-specific approach. By leveraging community needs and subsequently tailored solutions, these plans address the unique needs of local populations. The practical, evidence-based interventions provided by case studies are better suited to mitigate specific vulnerabilities and offer targeted strategies that directly improve resilience59. This localized focus ensures that the solutions are relevant and actionable, making them particularly effective in reducing vulnerability.To decrease vulnerability through policy, government bylaws and zoning ordinances play a major role in embedding climate resilience into local regulations. For example, Cambridge, Massachusetts, implemented Climate Resilience Zoning in February 2023, requiring new developments and major renovations to account for future flood risks60. By mandating flood protection measures based on projected climate data, these policies ensure that urban development is adapted to long-term environmental changes. Such regulatory frameworks strengthen community resilience by proactively mitigating climate-related risks.Specifically, mitigation and disaster recovery plans, while necessary for addressing immediate threats, may inadvertently increase vulnerability by diverting resources away from long-term adaptation goals. For example, the Massachusetts State Hazard Mitigation Plan focuses on reducing emissions but allocates fewer resources toward long-term infrastructure adaptation projects61. Similarly, the Connecticut Comprehensive Emergency Management Plan prioritizes rapid recovery after disasters but lacks provisions for climate-resilient infrastructure, potentially leaving vulnerable communities exposed to future risks62. Furthermore, the Maine Climate Council’s Climate Action Plan dedicates significant resources to emission reduction, which may divert attention from adaptation strategies such as flood resilience infrastructure and ecosystem protection in vulnerable areas63. By emphasizing short-term disaster recovery and mitigation measures, these plans may leave communities ill-prepared for the ongoing and future challenges posed by climate change. However, it is worth noting that increasing adaptive resilience is not the goal of these plans.A “case study implementation” style policy can significantly reduce vulnerability by drawing on localized experiences and tailoring solutions to specific regional contexts. This approach allows policymakers to learn from past events, particularly how certain communities or sectors have responded to climate change impacts, and apply these lessons to broader resilience-building efforts. For example, the New England Climate Adaptation Project (NECAP) implemented a case study-driven framework that encouraged municipalities to adopt climate adaptation strategies proven successful in nearby regions64. By showcasing real-world examples of successful adaptation, such as flood protection measures in coastal Massachusetts communities, this approach not only strengthens local resilience but also ensures that policies are based on tested, context-specific solutions. Moreover, the case study style empowers local governments to engage with their communities, adapting general policy frameworks to the needs of their populations. This practical, evidence-based approach is more likely to result in effective and sustainable climate adaptation measures, ultimately reducing vulnerability.In summary, the “how” of decreasing vulnerability effectively includes a focus on infrastructure improvements, using a case study-level analysis for plan implementation, and incorporating strategies into government bylaws and ordinances. These targeted approaches appear to build resilience by addressing immediate needs and embedding adaptive measures into local regulations.Who’s making a difference? The role of implementation levels in reducing vulnerabilityUnderstanding which actors are most effective in reducing climate vulnerability is vital for both theory and policy. Different levels of governance-State, Organization, Town, and Tribe-possess varying capacities, resources, and community engagement strategies, which may lead to different outcomes. Analyzing these levels allows us to evaluate how decentralized versus centralized responses shape climate resilience, and to what extent local governance structures can outperform larger-scale interventions.As illustrated in Fig. 4, the analysis reveals that most levels of implementation show limited effectiveness in reducing vulnerability-except for the tribal level, which emerges as significantly more impactful. This finding underscores the potential of place-based, culturally grounded adaptation strategies in addressing climate risks.Fig. 4: The impact of implementation level on climate vulnerability.The coefficient plot illustrates the results of the “Implementation Level” models. It combines data from the Resilience and Adaptation in New England (RAINE) database, which details the stock of climate policy implementation in each implementation group for each county-year, and the CDC/ATSDR Social Vulnerability Index (SVI), which measures vulnerability levels in the same communities. In this plot, negative coefficients indicate a reduction in vulnerability to climate hazards (see Supplementary Note 4 for the corresponding table).Full size imagePlans implemented by Indigenous Tribes are particularly effective in reducing vulnerability. This success may be due to Tribes’ deep, place-based knowledge and their integration of traditional practices into adaptation strategies65. Additionally, external support for tribal initiatives often recognizes the high vulnerability of these communities and provides targeted, context-specific assistance, enhancing the effectiveness of their resilience measures66. State-level policy implementation is also shown to reduce vulnerability. This effect may be attributed to the ability of states to allocate funds to otherwise under-resourced countries that are particularly vulnerable, enabling governments to address specific issues more effectively67,68.Substantively, regarding implementation levels, evidence shows that the state-level and tribal-level policies decrease vulnerability. State governments, with their authority and resources, can implement large-scale initiatives like Massachusetts’ State Hazard Mitigation and Climate Adaptation Plan, which focuses on infrastructure resilience and sustainable land use to reduce vulnerability to climate risks69. These plans are informed by regional climate projections, ensuring relevance to local conditions, and, at the state level, are highly resourced.Tribal-level policies show a decrease in vulnerability for Indigenous communities facing disproportionate climate impacts. Tribal governments leverage traditional knowledge with modern science, crafting strategies that protect both cultural practices and natural resources. For example, the Narragansett Tribe’s Native American Climate Change Adaptation Plan focuses on coastal erosion and sustainable fisheries, reflecting the tribe’s unique vulnerabilities70. By addressing local environmental, social, and cultural needs, both state and tribal policies are key in reducing vulnerability and fostering resilience.What climate goals deliver? Assessing effectiveness in vulnerability reductionThe analysis of policy focuses-Extreme Heat, Flooding, Saltwater Intrusion, Sea Level Rise, and Storm Surge-reveals important trends in vulnerability reduction, as depicted in Fig. 5.Fig. 5: The effect of the climate goal on vulnerability.The coefficient plot illustrates the results of the “Goals” models. It combines data from the Resilience and Adaptation in New England (RAINE) database, which details the stock of climate policy goals and focuses for each county-year, and the CDC/ATSDR Social Vulnerability Index (SVI), which measures vulnerability levels in the same communities. In this plot, negative coefficients indicate a reduction in vulnerability to climate hazards (see Supplementary Note 5 for the corresponding table).Full size imagePolicies aimed at Flooding, Extreme Heat, Sea Level Rise, and Storm Surge are associated with an increase in vulnerability. This may be because these policies often do not prioritize enhancing community-level resilience. Instead, they might focus on broader or less targeted interventions, potentially neglecting the specific needs of vulnerable local populations71.More specifically, policies aimed at flooding, extreme heat, sea-level rise, and storm surge often prioritize coastal areas, which can lead to uneven resilience efforts across the state. For instance, as shown in Fig. 1, counties along the Appalachians exhibit systematically higher vulnerability, yet these regions are frequently overlooked in climate discussions. The “Blue Ridge and Northern Highlands” area is one of the most drought-prone72 regions of the country, and, when it does rain, suffers from flooding73,74 exacerbated by coal mining practices and warming temperatures. These environmental changes are also damaging local forests, water quality75, and increasing wildfire risks.This coastal-centric focus not only neglects the serious vulnerabilities faced by rural and less developed areas but may also widen the inequality divide as resources are disproportionately allocated to wealthier coastal regions. This highlights a gap in climate policies: to be more effective, they should incorporate comprehensive strategies that address community-specific vulnerabilities, ensuring that resilience measures are integrated into the broader climate adaptation framework. Climate policies must adopt a more holistic approach that considers the unique vulnerabilities of all regions, thereby enhancing community-level resilience and ensuring equitable, targeted interventions for all affected populations.An important point to consider when discussing policies targeting climate impacts such as flooding, extreme heat, sea-level rise, and storm surges is the potential for a circular relationship between climate vulnerability and policy interventions. While the “Methods” section provides some causal evidence in terms of the direction of policies to vulnerability reduction, it is important to note that areas with high baseline vulnerability to these climate impacts may be more likely to implement protective policies. This, in turn, could create a feedback loop: areas already experiencing significant climate impacts may adopt more robust policies, but these very impacts continue to increase, thereby raising overall vulnerability despite mitigation efforts. Thus, while policies aiming to address these climate hazards are vital, the underlying vulnerability might persist or even increase due to the severity of ongoing climate change. This is also why state-level resources are so important. While models in the “Methods” section have incorporated these causal inference techniques, this circular relationship requires further attention when interpreting policy effectiveness and impact.In conclusion, this study leverages a comprehensive methodological approach to assess the impact of climate adaptation policies on community vulnerability across New England. A fixed effects model with clustered standard errors is utilized to account for unobserved heterogeneity at the community level. Principal Component Analysis (PCA) is employed to address multicollinearity among climate policy variables. Control variables from the United States Census and American Social Survey and the Climate Resilience Screening Index (CRSI) are incorporated to account for factors influencing community characteristics and climate exposure. Furthermore, the linearity of the models is assessed, and Granger causality tests are employed to strengthen the causal inferences drawn from the analysis. By combining these techniques, this study provides a robust framework for evaluating the effectiveness of climate adaptation policies in reducing vulnerability across communities.The findings of this study reveal that climate policies that focus on enhancing infrastructure, incorporating case-specific context, and prioritizing governance are associated with decreased levels of vulnerability. While individual policy characteristics were analyzed in isolation, it is likely that a combination of these vulnerability-reducing features working together—such as in the “Saving the Great Marsh” project—results in more robust and effective policies. This approach not only highlights which aspects of policies are functioning well and which are not, but also enables a more targeted and informed strategy for policy planning and adjustment. By understanding the interplay of policy features, types of plans, implementation levels, and goals, we can better evaluate their overall effectiveness in reducing climate vulnerability. Future initiatives should emphasize case-specific, evidence-based approaches that directly respond to the needs of vulnerable communities to strengthen resilience against climate change.The success story of Essex, MA: how comprehensive policy transformed climate resilienceThe “Saving the Great Marsh” project exemplifies an effective climate adaptation policy by integrating key elements such as bylaws and ordinances, infrastructure building, local-level implementation, and a case study approach. Focused on restoring the resilience of the Great Marsh in Essex County, Massachusetts, the project incorporates innovative ditch remediation techniques to address the ecosystem’s vulnerabilities76. A significant factor contributing to the project’s success is its robust funding at the state level, including grants from the National Coastal Resilience Fund and the MassBays grant. This financial support not only facilitates restoration efforts but also generates substantial employment opportunities, helping to alleviate socioeconomic vulnerabilities and adding an important layer to the project’s comprehensive approach. Figure 6 displays the resilience map of Massachusetts, highlighting Essex County as a region with high resilience levels.Fig. 6: Vulnerability scores in Massachusetts by county.The mapped data displays vulnerability scores across Massachusetts, sourced from the Social Vulnerability (SVI) index42. Essex County, which houses the Great Marsh restoration project, is depicted in the northeast most the county. Ideally, by addressing both environmental and socioeconomic factors, this project and others like it will enhance community resilience and reduce vulnerability. While the map shown here indicates vulnerability scores, Supplementary Note 6 shows how environmental resilience explicitly maps onto the county level. For Figs. 1 and 6, vulnerability scores, which serve as the main dependent variable of this paper, are shown to indicate baseline vulnerability for substantive discussion. However, resilience scores, which more explicitly capture certain environmental factors, are mapped in Supplementary Figs. 1 and 2.Full size imageDiscussionThis analysis offers insights into the nexus between climate policies and community vulnerability, particularly through the lens of SES theory. SES theory highlights the interdependencies between social practices and ecological conditions, underscoring that effective adaptation strategies must incorporate both environmental and societal dimensions. Policies centered around Government Bylaws and Ordinances, along with Infrastructure, demonstrate a marked decrease in climate vulnerability. Such outcomes suggest that these policies effectively enhance community resilience by reinforcing natural defenses and establishing strong legal frameworks while upgrading and investing in essential infrastructure. Existing literature purports that local institutions are pivotal in adapting to climate change, as they help align policy interventions with community-specific needs, thereby fostering resilience77.Furthermore, context-specific methodologies in Case Study Implementation Plans have been shown to effectively tailor approaches directly to local vulnerabilities, enhancing adaptive capacity78. The successful engagement of Indigenous Tribes to leverage traditional knowledge within policy frameworks underscores the importance of localized adaptation strategies. This finding resonates with Leichenko and Silva’s79 assertion that climate change vulnerabilities and responses are deeply entangled within political, social, and economic processes, emphasizing the vital role of localized governance in adaptation efforts. Policymakers must recognize that targeting various outcomes can complicate the repercussions on community vulnerability.While this study highlights significant patterns regarding the effectiveness of specific climate policies, it does not establish definitive causal relationships between policy features and outcomes. The absence of systematic benchmarks to measure the explicit impacts of policies on vulnerability limits our ability to understand these phenomena. Accordingly, this paper focuses on identifying correlations rather than isolating direct causation, which still allows for valuable insights aligned with the SES framework, considering both social trends and ecological impacts80. However, future research must delve deeper into the causal mechanisms by which these policy characteristics influence community vulnerability, particularly in the context of broader climate adaptation frameworks.This research contributes to the existing body of literature by demonstrating the necessity for policies that not only pursue stated objectives but also effectively enhance resilience in practice81. Policymakers should prioritize context-sensitive approaches that mitigate the unintended consequences of climate policies, as noted by Morrow and Bowen, who advocate for cross-sectoral policy integration to address vulnerabilities in multiple domains such as health, transportation, and food security82. By addressing the socio-ecological dynamics highlighted in SES theory, these findings stress the importance of cultivating holistic climate adaptation strategies that leverage local capacities and institutional knowledge to comprehensively mitigate vulnerabilities.In light of these discussions, future inquiries should explore the implementation challenges and specific practices associated with Social and Environmental Justice policies. Additionally, the influence of community engagement and traditional ecological knowledge must be considered to optimize climate policy efficacy. As the dynamics of climate change evolve, longitudinal studies tracking the long-term impacts of various climate policies in diverse contexts will be vital in refining resilience-building methodologies and ensuring sustainable adaptation efforts.MethodsThis section provides a comprehensive overview of the data and methodology used in the analysis, as well as additional robustness checks. It outlines the data, control variables, principal components, and statistical techniques employed to evaluate the impact of various facets of climate policies on community vulnerability. Specifically, the dataset consists of 1232 policies from the Resilience and Adaptation in New England (RAINE) database and integrates Social Vulnerability Index (SVI) data, census data, and climate exposure data. First, the Data and Model Specification are discussed. Then, insights on robustness are provided using Principal Component Analysis (PCA) to address multicollinearity, Fixed and Random Effects regression models, Hausman tests, Granger causality tests, and difference-in-differences (DID) analyses.DataThe dataset in this analysis consists of 1232 policies from the Resilience and Adaptation in New England (RAINE) database. To assess how climate policy design affects regional resilience, this study investigates three core questions: (1) How do different features and types of climate policies-such as those targeting infrastructure or ecosystem health-influence the resilience of New England’s social-ecological systems (SES)? (2) Which levels of implementation jurisdiction (state, local, tribal, etc.) are most effective in reducing social vulnerability across interconnected SES landscapes? and (3) What specific climate goals or focuses appear most effective at maintaining system stability and reducing vulnerability, given the dynamics of social-ecological interactions? These questions are empirically addressed by integrating Social Vulnerability Index (SVI) data, U.S. Census demographic data, and climate exposure metrics. Ordinary Least Squares (OLS) Fixed Effects regression models are employed to estimate the effects of policy features, types, and jurisdictional levels on SES resilience outcomes41,42,83,84. To address multicollinearity among correlated policy attributes, the analysis includes sets of Principal Components (PCs) representing various dimensions of climate policy design.The main dependent variable of this work leverages the Social Vulnerability Index (SVI), developed by the Centers for Disease Control and Prevention (CDC) and the Agency for Toxic Substances and Disease Registry (ATSDR)42. The SVI is a composite measure designed to identify communities with heightened susceptibility to harm during and after hazardous events. It aggregates 16 social factors across four key themes: Socioeconomic Status (e.g., poverty, unemployment), Household Characteristics (e.g., age, disability), Racial and Ethnic Minority Status (percentage of minority population), and Housing Type and Transportation (e.g., housing density, vehicle access). These factors are statistically combined to generate percentile rankings at the county level within New England, where a higher percentile indicates greater social vulnerability and, by extension, a potentially lower baseline capacity for resilience.This study interprets the SVI as an inverse measure of resilience. A community with a high SVI score is considered to have lower inherent resilience due to the presence of factors that can impede its ability to prepare for, respond to, and recover from climate-related shocks. Conversely, a lower SVI score suggests a greater capacity for resilience. Therefore, the analysis examines how climate adaptation policies influence these underlying social vulnerabilities, with the expectation that effective policies will contribute to a reduction in SVI scores over time, signifying an increase in community resilience across New England.Below details the exact definitions of the policy characteristics drawn from the RAINE Database41. First, for policy Features, the following are included:Economic Resilience: The potential to withstand disturbances to economic systems, including critical infrastructure (communications, drinking water, wastewater, and energy), resources, businesses, and services.Ecosystem Services: The economic value of maintaining or creating natural processes or systems that provide and regulate clean air, water, and soil.Bylaws/Ordinances/Codes: Local legislative actions to promote town or regional resilience.Built Infrastructure: Non-residential private and public buildings, roads, and highways are captured under transportation infrastructure.Environmental Justice: The fair treatment and meaningful involvement of all people regardless of race, color, national origin, or income with respect to the development, implementation, and enforcement of environmental laws, regulations, and policies.Additionally, for Plan Types, these include:Adaptation Plans: Plans that identify actions to avoid, benefit from, or deal with current and future climate change. Adaptation can take place in advance (by planning before an impact occurs) or in response to changes that are already occurring.Case Studies: Provides an in-depth examination of a situation, report, project, or plan.Climate Mitigation: Actions to reduce greenhouse gas emissions or to enhance the capacity of natural systems to absorb greenhouse gases from the atmosphere.Disaster Recovery Plans: Plans that guide communities in strengthening resilience to disasters and climate change, focusing on recovery and adaptation strategies.Resilience Plans: Plans that outline strategies to reduce communities’ vulnerability to flooding and support long-term recovery after a flood.In terms of the Implementation Levels, the following categories are included:State: Plans or actions implemented at the state level to address climate change impacts and resilience.Organization: Plans or actions implemented by organizations, including non-profits, businesses, or other entities, to address climate change impacts and resilience.Town: Plans or actions implemented at the municipal level to address climate change impacts and resilience.Tribe: Plans or actions implemented by tribal governments to address climate change impacts and resilience.For policy Focuses, or goals, these include:Extreme Heat: Periods of unusually hot weather, often defined as days with temperatures above 95 °F, projected to increase with climate change.Flooding: Overflow of water onto land that is usually dry, often exacerbated by climate change impacts.Saltwater Intrusion: The movement of saline water into freshwater aquifers, which can lead to contamination of drinking water sources.Sea Level Rise: The increase in the level of the world’s oceans due to the effects of global warming.Storm Surge: An abnormal rise of water generated by a storm, over and above the predicted astronomical tides.Sociodemographic variables are included as control variables to account for underlying differences in community characteristics that may independently influence social vulnerability and resilience outcomes, ensuring that the estimated effects of climate policies are not confounded by these baseline factors. The socioeconomic control variables are derived from the tidycensus data and encompass various social and demographic factors85. These factors align with the Morelli et al.86 climate adaptation screening focus areas, in being known to influence community vulnerability. They are also inherently related to the potential effectiveness of climate adaptation policies. These control variables include:Below 150% Poverty Rate (% of population)Unemployment Rate (% of population)Housing Cost Burden (% of population)Minority Household Rate (% of population)Single Parent Household Rate (% of population)Mobile Home Housing Rate (% of population)English Second Language (ESL) Rate (% of population)The variable that baseline climate exposure is derived from the Climate Risk Index84. This inclusion enables the analysis to account for the inherent vulnerability of communities to climate-related risks, thereby enhancing the robustness of the findings.To address potential multicollinearity among climate policy variables, Principal Component Analysis (PCA) is employed. PCA reduces the dimensionality of the data by creating a new set of uncorrelated variables, known as principal components, that capture the maximum variance in the original data87,88.Specifically, sets of principal components derived from the “Feature,” “Plan Type,” and “Goal” categories are controlled for (see Figs. 7, 8, and 9, respectively, for the Principal Component breakdown for model inclusion).Fig. 7: “Feature” principal component analysis.The first four principal components of “Feature” are observed to explain the necessary variation in the main models, so these components are included in the analysis.Full size imageFig. 8: “Plan Type” principal component analysis.The first three principal components of “Plan Type” are observed to explain the necessary variation in the main models, so these components are included in the analysis.Full size imageFig. 9: “Goals” principal component analysis.The first five principal components of “Goals” are observed to explain the necessary variation in the main models, so these components are included in the analysis.Full size imageFor the purposes of this analysis, this approach allows for the inclusion of policy-relevant information without introducing unnecessary collinearity in the model. Mathematically, PCA involves the following steps outlined in Equations (1)–(4):Equation (1), for Centering the Data, is:$${X}_{c}=X-\bar{X}$$(1)where X is the original data matrix and \(\bar{X}\) is the mean of each variable.Equation (2), for the Covariance Matrix, is:$$C=\frac{1}{n-1}{X}_{c}^{T}{X}_{c}$$(2)where C is the covariance matrix and n is the number of observations.To find the principal components, the Eigenvalue Problem is solved in Equation (3):$$C{\bf{v}}=\lambda {\bf{v}}$$(3)where λ are the eigenvalues and v are the corresponding eigenvectors.The Principal Components Z can be computed in Equation (4) as:$$Z={X}_{c}{\bf{V}}$$(4)where V is the matrix of eigenvectors corresponding to the largest eigenvalues.This technique effectively mitigates the biasing effects of multicollinearity, which arises when predictor variables exhibit high correlation, ultimately resulting in unreliable coefficient estimates. Having delineated the data, the next step is model specification.Model specificationThis section outlines the model specification and validation process. It discusses the assumptions of linearity and heteroskedasticity, and compares fixed effects and random effects models. The Hausman test is conducted to determine the most appropriate model for the analysis.Several methods are employed to ensure the robustness of the results. First, the linearity of the relationships between the independent variables and the dependent variable is assessed using scatter plots and residual plots (see Fig. 10). Linearity is a key assumption of regression analysis, and ensuring its validity strengthens the interpretability of the model’s results. Residual plots are examined to identify any patterns or outliers that might indicate violations of model assumptions.Fig. 10: Total policy on vulnerability.This scatter plot displays the relationship between total climate policies implemented in each county-year and social vulnerability index scores across New England. Each point represents a county-year observation, with the trend line showing the overall association between policy implementation and vulnerability reduction.Full size imageWhile the linearity appears to be satisfactory, clustered standard errors based on Huber-White adjustments are employed across all models to account for potential heteroskedasticity89,90.Furthermore, skewness and kurtosis for the independent variables were checked to assess their distributions and determine which variables to log transform. The variables Single Parent, Unemployment, Housing Burden, Poverty, Minority, Mobile Homes, ESL, and Climate Exposure were identified and logged to address concerns related to skewness and non-linearity. The logged versions of these variables were included in all models.Given that the data is linear and in panel format, a determination is made regarding whether a random effects or fixed effects model is more suitable. The fixed effects model is shown in Equation (5):$${\rm{ClimateVulnerability}}it={\beta }_{0}+{\beta }_{1}{\rm{PolicyComponent}}_{it}+{\beta }_{{\rm{Controls}}}+{v}_{i}+{\epsilon }_{it}$$(5)where ClimateVulnerabilityit denotes the dependent variable for individual i at time t, while PolicyComponentit encapsulates the feature, plan type, implementation level, or goal of the policy (alternatively referenced as FEATUREit, PLANit, IMPLEMENTATIONit, or GOALit). Specifically, FEATUREit signifies the policy feature variable (Economic Resilience, Ecosystem and Natural Resources, Government Bylaws and Ordinances, Infrastructure Built, or Social and Environmental Justice), while PLANit indicates the plan type variable (Adaptation Plans, Case Study Implementations, Climate Mitigation Documents, Disaster Recovery Plans, and Resilience Plans). Moreover, IMPLEMENTATIONit refers to the variable representing the level of implementation (Organization, State, Town, or Tribe), and GOALit is the policy goal variable (Extreme Heat, Flooding, Saltwater Intrusion, Sea Level Rise, or Storm Surge). In this context, β0 represents the intercept, β1 is the coefficient associated with the feature variable, βControls signifies the composite index of control variables, ui denotes the individual-specific effect (fixed effect), and ϵit is the idiosyncratic error term.The random effects model is specified in Equation (6):$${\text{ClimateVulnerability}}={\beta }_{0}+{\beta }_{1}{\text{PolicyComponent}}_{i}+{\beta }_{{\rm{Controls}}}+{u}_{i}+{\epsilon }_{it}$$(6)In contrast to the fixed effects model, which accounts for individual-specific effects (ui) that are assumed to be correlated with the independent variables, the random effects model incorporates a term (vi) that captures individual-specific variations while assuming these variations are uncorrelated with the explanatory variables. This assumption allows for the inclusion of time-invariant variables in the analysis. A Hausman test is conducted to evaluate the relative appropriateness of these models. Equation (7) represents the Hausman test:$$H=({\hat{\beta}}_{FE}-{\hat{\beta}}_{RE})^{\prime} {[{\text{Var}}({\hat{\beta }}_{FE})-{\text{Var}}({\hat{\beta}}_{RE})]}^{-1}({\hat{\beta }}_{FE}-{\hat{\beta }}_{RE})$$(7)where \({\hat{\beta }}_{FE}\) indicates the coefficients estimated by the Fixed Effects Model, and \({\hat{\beta }}_{RE}\) signifies the coefficients estimated by the Random Effects Model. The variance of the coefficients from the Fixed Effects Model is denoted by \({\text{Var}}\,({\hat{\beta}}_{FE})\), while \({\text{Var}}\,({\hat{\beta }}_{\rm{RE}})\) represents the variance of the coefficients from the Random Effects Model. The results of the Hausman test are displayed in Table 1:Table 1 Hausman test resultsFull size tableThe results of the Hausman test indicated a preference for the fixed effects model over the random effects model. Consequently, the fixed effects model is specified in the primary analysis. While the fixed effects model is estimated in the main text, it is noteworthy that the random effects models demonstrate robustness to the findings, as shown in Supplementary Note 7. Furthermore, the high-dimensional effects captured in the random effects models are well-suited to account for the hierarchically structured Baseline Climate Exposure.A fixed effects model is employed to account for unobserved heterogeneity at the community level (see Equation (5)). This approach is particularly suited to the study design, as it recognizes that communities may possess inherent characteristics that influence their vulnerability scores, beyond the variables included in the model91. By incorporating random effects, these variations are accounted for, leading to more accurate estimates of the true effects of climate policies on community vulnerability. where ClimateVulnerability is the dependent variable, and PolicyComponenti represents the feature, plan type, implementation level, or goal of the policy (otherwise referred to as FEATUREi, PLANi, IMPLEMENTATIONi, or GOALi). Specifically, FEATUREi denotes the policy feature variable (Economic Resilience, Ecosystem and Natural Resources, Government Bylaws and Ordinances, Infrastructure Built, or Social and Environmental Justice), while PLANi indicates the plan type variable (Adaptation Plans, Case Study Implementations, Climate Mitigation Documents, Disaster Recovery Plans, and Resilience Plans). Furthermore, IMPLEMENTATIONi refers to the level of implementation variable (Organization, State, Town, or Tribe), and GOALi is the policy goal variable (Extreme Heat, Flooding, Saltwater Intrusion, Sea Level Rise, or Storm Surge). In this equation, β0 represents the intercept, β1 is the coefficient for the feature variable, βControls denotes the index of control variables, ui is the individual-specific effect (random effect), and ϵit is the idiosyncratic error term.Another reason the random effects model is optimal is that the climate exposure control variable is only available at the spatial level, not the temporal level. This is because it represents aggregate climate stress across a region, not changing over time, and comes from the CRSI index92. The random effects model can effectively account for this spatial variation in climate exposure.RobustnessThis section provides a detailed overview of the robustness checks conducted to ensure the validity of the results. It discusses the use of Principal Component Analysis (PCA) to address multicollinearity, Granger causality tests to establish temporal relationships, and difference-in-differences (DID) analysis to assess the causal impact of climate policies.The concept and measurement of climate vulnerability inherently encompass various sociodemographic factors and climate variables. To ensure that the models are robust against any unnecessary multicollinearity, Principal Component Analysis (PCA) is employed (see Equations (1)–(4)) on all climate vulnerability (CV) socio-demographic factors and climate variables. The results of this analysis are presented in Fig. 11, illustrating the correlations among the principal components.Fig. 11: Principal component analysis of socioeconomic & climate control variables.The first seven principal components explain the necessary variation in the models.Full size imageFurthermore, Fig. 12 depicts how these principal components correlate with the outcome variable. Notably, Principal Component 1 (PC1) exhibits a high correlation, prompting its exclusion from the model. Instead, the analysis utilizes Principal Components 2 through 7 as predictors. This approach ensures that the outcome variation is not unnecessarily influenced by certain predictors.Fig. 12: Principal component analysis correlation with dependent variable.This plot shows how each of the first seven principal components from the socio-economic and climate control variables correlates with the dependent variable (Social Vulnerability Index). PC1 shows the highest correlation and was excluded from models to avoid multicollinearity.Full size imageThe results remain robust and comparable to those obtained from the main models, as shown in Fig. 13 and Table 9 in Supplementary Note 8. The only difference is that in these models, the State level of implementation shifts to statistically significant at the .1 level.Fig. 13: Main models (Figs. 2–5) with PC 2–7 controls.Robustness check using principal components 2–7 as controls. These coefficient plots replicate the main analyses from Figs. 2–5 while controlling for principal components 2–7 of socio-economic and climate variables instead of individual control variables. Results demonstrate consistency with the main findings.Full size imageFurthermore, Fig. 14 shows the simple bivariate fixed effects model outcomes only including the policy-related principal controls to ensure the isolated effect of the policies (see corresponding output in Supplementary Note 9).Fig. 14: Bivariate models corresponding to the main models in Figs. 2–5.These plots show the relationship between policy variables and vulnerability using only policy-related principal components as controls, removing socio-economic controls to isolate the direct effects of climate policies on vulnerability.Full size imageThese models show robustness, with some small changes. The Town-level of implementation now leads to lower climate vulnerability, leaving only the Organizational level as ineffective. Additionally, differences are observed in the policy features, with Government Bylaws and Ordinances, and Infrastructure Built slightly shifting to significance only at the 0.1 level.To account for the influence of political factors on climate policy outcomes, Fig. 15 presents the results of the fixed effects model, which incorporates the percentage of the county’s population that votes Republican, as sourced from the National Neighborhood Data Archive (NaNDA)93. Including this variable helps capture the potential impact of political ideology on the formulation and implementation of climate policies. Given that political party affiliation often shapes policy preferences and priorities, understanding the political disposition in a county allows for a better understanding of how political context may influence vulnerability to climate change. This approach highlights the importance of considering local political dynamics when assessing the effectiveness of climate policies and their potential to mitigate vulnerability.Fig. 15: Political control models corresponding to the main models in Figs. 2–5.These analyses replicate the main findings while including the percentage of county population voting Republican as an additional control variable to account for potential political influences on climate policy effectiveness.Full size imageThese models show robustness with the main models, but with ecosystem natural resource features moving to be statistically significant in reducing vulnerability, while the policy goal of saltwater intrusion has lost its statistical significance. The ratio of the county that has voted republican is not statistically significant in any of the models. See the corresponding table output in Supplementary Note 10.Additionally, looped Granger causality tests are conducted to explore the temporal relationships between climate policies (indicated here by Total Climate Policies in a given county-year) and SVI scores. Granger causality is a statistical technique used to assess whether one time series variable can predict another (see Equation (8)). In this context, the test assists in determining whether the implementation of climate policies precedes any measurable changes in community vulnerability scores94.$${Y}_{t}=\alpha +\mathop{\sum }\limits_{i=1}^{p}{\beta }_{i}{Y}_{t-i}+\mathop{\sum }\limits_{j=1}^{q}{\theta }_{j}{X}_{t-j}+{\epsilon }_{t}$$(8)where Yt signifies the climate readiness score, which is the dependent variable being forecasted, and Xt stands for the adaptation laws, the independent variable under examination for Granger causality. The symbols p and q represent the number of lags for Y and X, respectively. The constant term is denoted by α, with βi and θj representing the coefficients for the lagged variables. Lastly, ϵt indicates the error term in the model.The results of the Granger causality test (see Table 2) show statistical significance, indicating a causal relationship between climate adaptation policies and reduced vulnerability (p