Measuring and enhancing resilience of United States rural communities in the context of climate variability

Abstract

This article serves as an introduction to the Special Issue section “Measuring and Enhancing Resilience of United States Rural Communities in the Context of Climate Variability.” To set the stage for this section, we review how climate hazards impact rural areas and synthesize insights that emerge across the issue’s four papers, noting their policy relevance and highlighting opportunities for continued research. We argue that emerging data tools can help program designers and policy makers better support the resilience of rural areas, but that doing so remains complicated by heterogeneity in resources and vulnerabilities across rural areas.

Introduction

Rural areas play a key role in the agricultural, industrial, and recreational life of the United States. Rural areas, however, remain vulnerable to a variety of shocks, including climate-related hazards. The Federal Emergency Management Agency’s (FEMA) National Risk Index dataset, which summarizes natural hazard risk across the United States, shows that rural areas exhibit, on average, higher social vulnerability and lower community resilience than urban areas (Zuzak et al. 2023). Building the resilience of rural areas requires an acknowledgment that their geographic, social, and economic characteristics shape their vulnerability and adaptive capacity in important ways.

This special issue section addresses the need to better understand the resilience of rural communities in the United States, especially to increasing climate variability. The papers it contains provide insight into the challenges of defining rural resilience, measuring outcomes associated with rural resilience, and creating policies that can help build rural resilience. In this introductory article, we set the stage by reviewing how climate hazards impact rural areas. We also note how characteristics of rural areas might make them attractive destinations for internal migration when disasters – climate-related or otherwise – hit densely populated urban areas. We then synthesize insights that emerge across the section’s four papers, note their policy relevance, and highlight opportunities for continued research.

Background: climate hazards and rural areas

The Federal Emergency Management Agency (FEMA) defines natural hazards as “environmental phenomena that have the potential to impact societies and the human environment” (FEMA 2025).¹ The environmental phenomena themselves are routine occurrences as part of Earth’s natural climate. Climate disasters, on the other hand, occur when these natural hazards intersect with human systems. The consequences of these clashes are often significant. Between 1980 and 2024, 403 climate hazard disaster events have been estimated to have impacts each exceeding one billion dollars. Of those, severe storms and tropical cyclones, flooding, and drought are the most common (NOAA 2025a).

Both rural and urban areas are meaningfully affected by climate disasters. The way that rural and urban areas prepare for, experience, and recover from climate disasters, however, is influenced by their differing economic organization, geographic distribution, and physical characteristics. Rural populations differ from urban populations in employment, education, natural amenities, access to healthcare, and broadband connectivity (Davis et al. 2022). Smaller populations in rural areas make it harder to share the cost of disaster preparation and recovery. As working age populations in rural areas shrink, a larger share of the community (both the very young and very old) depends on a smaller workforce (Farrigan et al. 2024). Additionally, rural areas are more likely than urban areas to have local economies built around place-based industries such as agriculture, recreation, or resource extraction. Across the United States, 17 percent of non-metropolitan counties are farming dependent, 15 percent are recreation dependent, and four percent are mining dependent (USDA-ERS 2025b). In total, 60 percent of non-metropolitan counties are categorized as dependent on a single industry. In contrast, only 33 percent of metropolitan counties are industry dependent.² High levels of dependence, particularly in place-based industries, makes rural counties especially vulnerable to climate disasters.

In the remainder of this section, we highlight some of the ways in which two climate disasters – drought and flooding – impact rural areas. While not the only climate disasters to affect rural areas, drought and flooding are among the most frequent and capture important elements of the challenge of building resilience in rural areas.

Like all of the hazards identified above, drought is a natural phenomenon, only becoming a disaster when it interacts with the demands placed on water and other natural resources by human use systems (Wilhite, Svoboda and Hayes 2007). All regions of the United States are affected by drought and, since 1980, drought ranks as the third most common cause of billion-dollar disaster events (NOAA 2025a). The scope of drought events, often affecting entire regions of a country and persisting for multiple years, poses a particular challenge: it is extremely hard for individuals and communities to support one another when everyone faces a shock at the same time. For example, the drought conditions that began in the Western region of the United States in 2012 persisted in some states, most notably in California, until 2016 (NOAA 2025a). As their geographic impact peaked in September of 2012, 65.5 percent of land in the Lower 48 states faced drought conditions according the the U.S. Drought Monitor.³

For rural communities, both the direct and indirect consequences of drought can be significant. The direct effects of drought include reduced crop, rangeland, and forest productivity, increased wildfire risk, reduced water availability, increased livestock and wildlife mortality, and deteriorated fish and wildlife habitats (Fu et al. 2013; Wilhite, Svoboda and Hayes 2007). Indirectly, decreased farm incomes due to drought mean less spending in associated industries and local communities, higher food prices, higher unemployment, and reduced tax revenues (Wilhite, Svoboda and Hayes 2007). As a concrete example, the negative economic impact of the 2014–2016 drought on California agriculture was estimated at $3.8 billion (Lund et al. 2018). In 2015 alone, the drought was estimated to have caused 21,000 lost jobs in California agricultural production and associated industries. Severe drought events routinely have consequences of or exceeding this magnitude. Across the 32 droughts resulting in billion-dollar drought disasters recorded by National Oceanic and Atmospheric Administration since 1980, the average estimated cost was $11.5 billion per event. What’s more, while drought is most commonly associated with effects on agriculture, its impacts on tourism, industry, and social welfare are believed to be under-estimated (Fu et al. 2013).

Surface water, river, and coastal flooding are also common and costly. Approximately 75 percent of Presidential disaster declarations are associated with flooding (NOAA 2025b). Flooding is the most common source of billion-dollar weather and climate disasters (NOAA 2025a) and is estimated to have contributed 7.4 percent of total economic losses associated with natural disasters in the United States between 1980 and 2023 (Abegaz, Wang and Xu 2024).

Heavy rain and flooding can impose large costs on rural residents, even in counties with well-diversified local economies, by damaging dispersed infrastructure and reducing water quality (Abegaz, Wang and Xu 2024). While the 2020 United States Census reports that only 20 percent of the nation’s population live in rural areas, those areas hold 71 percent of the country’s road miles and 72 percent of bridges (USDT 2024). In addition to more road miles, roads in rural census tracts are more likely to be categorized as slightly, moderately, or highly rugged than roads in urban census tracts (USDA-ERS 2025a). Both the extent of road infrastructure and its topography in rural areas have implications for moving goods, accessing services, and even evacuating populations during flood events and other disasters.

While characteristics like dispersed infrastructure and a concentrated economic base contribute to the vulnerability of rural areas to climate disasters, they also help define opportunities for growth and improved resilience. For example, focused economic drivers present clear targets for investments in infrastructure and workforce development. The location and lower population density of rural areas also create inherent amenities that may increase rural areas’ attractiveness as migration destinations when urban areas experience downturns or disasters. This was visible as nonmetropolitan areas in the United States reversed nearly a decade of population decline following the COVID-19 pandemic, not only growing but growing at a faster rate than metropolitan areas (Davis et al. 2022; Johnson and Lichter 2019). The largest increases in population were seen in areas that also offered desirable amenities, such as robust infrastructure to support remote work, recreation opportunities, and retirement destinations (Davis et al. 2022), again creating clear targets for strategic investment.

Climate disasters in urban areas may result in similar migration patterns to those observed during the COVID-19 pandemic. Across the country, rural counties are estimated to have lower overall risk of coastal and river flooding, heat waves, hurricanes, ice storms, and strong winds than urban counties (Zuzak et al. 2023). While in-migration to rural areas in response to these disasters may tax existing infrastructure in the short run, the additional workforce and spending power may help provide much-needed resources to in the long run.

As every rural community is unique, it is important that resilience planning take into account a community’s specific strengths and vulnerabilities. Hazard mitigation plans built on community-specific characteristics have proven effective at generating more robust mitigation programming and disaster response (Burby 2005). At the same time, such planning efforts have largely been shown to be incomplete, despite policymakers’ awareness of areas for improvement (Horney et al. 2017).

Research insights: defining resilience and identifying evolving data needs

The papers in this special issue section highlight a number of challenges associated with defining resilience in the context of rural communities and evaluating policies meant to boost rural resilience. In the simplest of scenarios, designing effective policies and evaluating the programs created under those policies is challenging. It is necessary to precisely enumerate goals and define clear metrics for evaluation. Even then, well-matched data is required to put evaluation plans into practice. As noted throughout this special issue, however, supporting rural communities in becoming and remaining resilient is far from the simplest of scenarios.

Insight 1: defining resilience-related program or policy goals is important and non-trivial

While resilience has become a common term and there is broad acknowledgment that “more resilient” is better than “less resilient,” an actionable definition of resilience has been slow to emerge. Crawley, Daigneault and Bowen (2025) document the wide range of definitions and outcomes that have been associated with resilience in the community and regional development literature. As the authors note, a precise definition is a prerequisite for measurement. We take it a step farther and note that measurement is a prerequisite for evaluation.

Rural communities are diverse and complex. Even adopting what Crawley, Daigneault and Bowen (2025) identify as the emerging consensus that resilience involves elements of withstanding, recovering from, and reorganizing in response to shocks, what is needed to build resilience differs from one rural community to the next. Moreover, policies are likely to target elements of resilience or resources that support an element of resilience rather than resilience itself.

The challenge of defining appropriate goals and the limited extent to which solutions can be ported from one community to another likely contribute to the prevalence of incomplete mitigation planning noted by Horney et al. (2017). Identifying community-specific priorities and resources, the importance of which is highlighted by Cain (2022), can be both time consuming and expensive. Often, reaching consensus on priorities is complicated by the existence of diverse interests and differing evaluations of risk and benefit. Still, the importance of well-defined goals and sound evaluation plans for resilience-related policies should not be understated.

Insight 2: existing data can be combined in new ways to generate insight into the opportunities for and challenges facing rural areas

Fortunately, given the challenge identified in the previous item, we live in a data-rich time. Much of the data necessary to better understand the diverse characteristics of rural areas, to measure resilience, and to evaluate policies seeking to improve components of resilience exist, though they reside in a dispersed set of survey, administrative, and earth observation data repositories. Drawing on emerging data sources and combining existing data sources in new ways can contribute to well-defined and well-measured resilience goals.

Kim and Fannin (2025) demonstrate to utility of combining data from various types of decision-makers, using household-level, business-level, and public-level data to create a community vulnerability measure. Because resilience is multifaceted, incorporating data from both place-based (business and local government) and people-based sources can provide a fuller picture than using a single data source alone. As the authors note, this approach can help communities identify features that serve as the basis for strengths and vulnerabilities as they plan investments for the future.

Many features of rural areas can be seen as both resources in building resilience and challenges in maintaining resilience. Dobis et al. (2025) highlight one such feature by using census tract topography to create measures of ruggedness. The Area Ruggedness Scale (ARS) and the Road Ruggedness Scale (RRS) provide a useful complement to the already established Natural Amenities Scale by reflecting both the amenity value (such as access to recreational activities and open spaces) and the disamenity value (like higher cost of obtaining services and vulnerability of infrastructure) that rugged terrain brings to rural areas.

The topography data that underlies the ARS and RRS is appealing as it has national coverage, meaning it can be used to create very local measures of ruggedness. Additionally, it does not involve individual surveys that are expensive to conduct and must overcome participation and sampling difficulties. This is a characteristic shared by earth observation data, which can provide program designers and evaluators with high frequency and high resolution data to monitor conditions in near-real time. Benami, Ramanujan and Cecil (2025) provides a primer on earth observation data and its integration into program management and evaluation.

Insight 3: decisions about how data are used to administer and monitor programs have significant consequences

As more expansive and precise data become available, using data for program administration and evaluation requires more decisions around data source and processing. Benami, Ramanujan and Cecil (2025) demonstrate that choice of data source and scale can have significant implications for programs that support the management of natural hazard risk. In the context of a rainfall index insurance product for perennial pasture, rangeland, and forage used to feed livestock in Texas, the authors show that both payout frequency and level are sensitive to data source and spatial resolution, even when index zones are held fixed. Their results hold meaningful implications from the point of view of participant welfare and programmatic cost. Furthermore, they show that “finer” is not necessarily “better” when it comes to spatial resolution, which leaves the onus on program designers to make informed decisions.

The question of data source and scale is further complicated by the fact that not all data decisions are elective. In working with data from different sources, especially on a multi-dimensional concept like resilience, data conformability issues are likely to arise. Dobis et al. (2025) provide guidance on potential issues surrounding data aggregation and suggest a number of practical approaches, which they demonstrate in the context of aggregating census tract measures of the ARS and RRS to the county level.

Policy relevance and future research priorities

Research into rural resilience, particularly in the context of climate-related threats, remains extremely relevant. In CPI-adjusted dollars, the 1980s saw a total of 33 “billion-dollar” climate disasters. That number rose to 57 in the 1990s, 67 in the 2000s, and 131 in the 2010s (NOAA 2025a). Many of these events directly affected rural communities, often in devastating ways. Even when disaster events take place outside rural areas, inter-connected supply chains and changing migration patterns transmit their influence across the country and around the globe.

The four papers in this special issue section highlight some of the challenges and opportunities facing policy makers working to improve the resilience of rural communities in the United States. Each paper reiterates the multidimensional nature of resilience and the unique characteristics of rural communities, providing important reminders for effective policy design. To conclude this introductory article, we note three promising areas for future research related to rural resilience stemming from the papers in this special issue section.

1. 1. Resilience is multidimensional. Programs aimed at increasing resilience often target single outcomes that support a community’s overall resilience, such as infrastructure, economic diversification, or food access. Future research can explore how different dimensions of resilience interact. For example, both economic institutions and social institutions play key roles in preparing for and recovering from climate disasters. Research into how these institutions reinforce or substitute for each other in generating resilience is warranted.

2. 2. The unique characteristics of rural areas shape both their vulnerability and adaptive capacity. The impacts of negative shock events can be direct (on individuals and communities where they occur) or indirect (such as impacts on communities connected or adjacent to those experiencing shocks). For example, farming-dependent counties might experience shocks directly from drought, while other counties experience indirect shocks from supply chain disruptions. The economic profile of these counties shapes the consequences of the drought. Future research might study how unique characteristics of rural areas, such as economic connectivity and geographic isolation, influence their vulnerability and adaptive capacity.

3. 3. Evidence-based policy design and evaluation requires well-matched data. Reliable data and well-executed evaluations are needed to demonstrate program efficacy and identify opportunities to improve program design or targeting. Future research might prioritize the collection, integration, and use of data sources to inform the development of targeted and evidence-based policies. Such data can additionally be used to learn from communities that demonstrate a high level of resilience, despite the vulnerabilities they face.

The points highlighted in this introductory article are by no means a complete summary of the insights contained in the issue’s papers. Each paper makes its own contribution by providing data-driven insights of use to researchers and policymakers alike. In both the results they provide and discussion they facilitate, the papers that follow further the goal of creating effective, evidence-based polices that enhance the resilience of rural communities to climate related threats.