Introduction
Climate-induced disasters are disasters that are caused by naturally occurring environmental hazards (e.g., blizzards, cyclones, droughts, floods, severe storms, tornadoes, tsunamis, wildfires, etc.).Reference Shaluf1 Over the past several decades, the frequency and severity of climate-induced disasters have increased in the United States and resulted in substantial amounts of preventable human morbidity and mortality.Reference Prelog and Miller2 The health impacts of climate-induced disasters are linked to interactions between factors related to the disaster event, disaster exposures, and the population impacted by a disaster. Within the United States, rural and urban environments are exposed to the same types of climate-induced disaster events, but rural populations are considered to be particularly vulnerable to adverse health effects.Reference Kappel, Chavez and Melnick3, Reference Betka, Bergren and Rowen4 This is largely due to historical marginalization and neglect, which have left rural environments with relatively limited resources, infrastructure, and capacity to support disaster response efforts.Reference Kappel, Chavez and Melnick3–Reference Douthit, Kiv and Dwolatzky8 Beyond these considerations, geographic disparities persist, and rural populations within the United States are typically older, sicker, less educated, less insured, and of lower socioeconomic status when compared to urban populations.Reference Callaghan, Kassabian and Johnson6, Reference Parker, Horowitz and Brown9, Reference Randolph, Thomas and Holmes10 Because of this, rural populations in the United States face distinct public health challenges as they relate to preparing for and responding to climate-induced disaster events.Reference Kappel, Chavez and Melnick3, Reference Betka, Bergren and Rowen4
To date, much of the research on public health preparedness for climate-induced disasters in rural contexts has focused on estimating the risks posed by climate-induced disasters and the resilience of various systems against these hazards.Reference Kapucu, Hawkins and Rivera11, Reference Cutter, Ash and Emrich12 While important lines of inquiry, these considerations neglect individual-level preparedness, which is considered a fundamental method for reducing the risk posed by climate-induced disasters.Reference Rivera13 Within the United States, numerous organizations—including the Federal Emergency Management Agency, the Centers for Disease Control and Prevention, and the American Red Cross, among others—have promoted individual-level preparedness by recommending that individuals develop evacuation plans, compile emergency preparedness kits, stockpile supplies, and familiarize themselves with warning systems and signals, among other actions.Reference Murphy, Cody and Frank14, 15 Individual-level preparedness, however, can be modified by numerous factors that modify the ability and desire of individuals to prepare for climate-induced disasters, including socioeconomic and demographic considerations, as well as risk perception.Reference Levac, Toal-Sullivan and O’Sullivan16
Recognizing that rural and urban populations differ across many of these aspects, it is reasonable to hypothesize that rurality may alter individual-level preparedness attitudes and behaviors for climate-induced disasters. The purpose of this study, then, is to compare individual-level preparedness for climate-induced disasters in rural versus urban environments, and to investigate how rurality is related to demographic and socioeconomic variables that influence climate-induced disaster preparedness attitudes and behaviors.
Methods
Survey Development and Distribution
A survey was developed to characterize attitudes and behaviors regarding preparedness for climate-induced disasters. The survey queried on self-reported demographic, socioeconomic, and experience predictor variables, as well as attitudes toward climate-induced disasters and specific preparedness behaviors. Demographic, socioeconomic, and experience variables included rurality, gender, age, race, education, annual household income, political affiliation, region of residence, residence in an area that experiences natural disaster events, and previous experience with natural disaster events. Questions regarding attitudes and preparedness behaviors included: (1) Is it important to be prepared for natural disasters? (2) Are you concerned that natural disasters will become more severe in the future? (3) Are you concerned that natural disasters will become more frequent in the future? (4) Do you have a personal or family evacuation plan in case of a natural disaster? (5) Do you have an emergency kit in your home? (6) Are you confident that you know where to get emergency information in case of a natural disaster? (7) Are you confident that you have the supplies you need if you were to be impacted by a natural disaster? (8) Are you confident that you know how to protect yourself during a natural disaster? Once finalized, the study materials were translated into Spanish, and materials were submitted to the Texas A&M University Institutional Review Board for ethical review, which declared the study exempt from full review.
The survey was then uploaded to Qualtrics XM, and Lucid—an online survey platform—was used to recruit respondents. Lucid uses multiple sourcing methods and techniques to recruit survey respondents through an opt-in online marketplace that offers incentives to survey respondents. The survey ran from April 24 to June 3, 2024, and surveys were administered in both English and Spanish. Eligibility criteria for this analysis included: (1) being at least 18 years of age, (2) residing in the United States, and (3) self-reported residence in an urban or rural environment.
Statistical Analysis
Descriptive statistics were used to characterize the reported sociodemographic characteristics, and preparedness attitudes and behaviors for rural and urban populations. Chi-square tests of independence were used to inspect for differences among the study populations.
Multiple logistic regression was used to investigate the relationship between sociodemographic characteristics and preparedness attitudes and behaviors for those that were found to differ significantly between rural and urban populations. Education was dropped from regression models, as it was highly correlated with annual household income. A screening analysis was performed to inform the regression models, in which a logistic regression model for each sociodemographic characteristic was fit individually while also allowing an interaction with rurality. Variables with interaction P-values less than 0.1 (P <.10) were included in the final model for a given preparedness attitude or behavior. Following the screening analysis, multiple logistic regression models with interaction terms and robust standard errors were used to examine the associations between rurality, sociodemographic characteristics, and preparedness attitudes and behaviors. The threshold for statistical significance was set at an alpha of 0.05 for all data analysis (P <.05), other than the screening analysis. Data analysis was conducted in October 2024 using StataSE v.18.
Results
A total of 1,680 individuals were included in the analysis, of which 816 (48.57%) reported living in rural environments and 864 (51.43%) in urban environments (Table 1). Rural respondents most often identified as female, 18-34 years of age, white, with an annual income of less than $25,000, politically Independent, from the South, living in an area that experiences climate-induced disasters, and having previous experience with disasters; urban respondents most often identified as male, 18-34 years of age, white, with an annual income of less than $25,000, politically Democrat, from the South, living in an area that experiences climate-induced disasters, and having previous experience with disasters. The distribution of all individual characteristics was significantly different between rural and urban populations, except for living in an area that experiences climate-induced disasters and having previous experience with climate-induced disasters.
Reported respondent characteristics for rural and urban subpopulations (n = 1680)
Respondents from both rural and urban environments most often reported believing that preparedness for climate-induced disasters is important, being concerned about the severity of future disasters, being concerned about the frequency of future disasters, having a personal or family evacuation plan, having an emergency kit at home, having necessary supplies in case of a disaster, knowing how to protect oneself during a disaster, and knowing where to access emergency information in case of a disaster (Table 2). However, chi-square tests revealed statistically significant associations between rurality and having a personal or family evacuation plan (X 2 (1, n = 1680) = 6.23, P = .012), knowing where to access information in case of a disaster (X 2 (1, n = 1680) = 7.25, P = .007), being concerned about the severity of future disasters (X 2 (1, n = 1680) = 11.17, P = .001), and being concerned about the frequency of future disasters (X 2 (1, n = 1680) = 11.22, P = .001).
Reported attitudes and behaviors regarding preparedness for climate-induced disasters for rural and urban subpopulations (n = 1680)
* P <.05
Screening analyses suggested potentially meaningful interactions between rurality and age, income, political affiliation, and region of residence for select outcomes (Table 3). Results from the multiple logistic regression model investigating concern about the severity of future climate-induced disasters showed, compared to referent groups, the odds of reporting concern were significantly higher for those between 35 and 54 years of age, for those living in an area that experiences disasters, and for those previously impacted by a disaster, while the odds were significantly lower for those with Republican or Independent political identities (Table 4). Predicted probabilities also suggest that Democrats were significantly more likely to report concern about the severity and frequency of future climate-induced disasters than Republicans or Independents, in both urban and rural environments (Supplemental File). Significant interactions between rurality and age and income were also observed, indicating that differences in rural versus urban responses differ along these characteristics. Compared to their urban counterparts, the prevalence of reporting concern about the frequency of future climate-induced disasters was significantly lower for rural individuals aged 35-44 years and with an annual income of $75,000-$99,999 (Figure 1; Supplemental File).
P-values from screening tests for interactions with rural residence and various demographic and socioeconomic characteristics for select preparedness risks and attitudes (n = 1680)
* P <.10
Adjusted odds ratios for select attitudes and behaviors regarding preparedness for climate-induced disasters (n = 1680)
* P <.05
Prevalence of reporting concern about the severity of climate-induced disasters in the future for rural and urban populations.
Regarding reported concern about the frequency of future climate-induced disasters, compared to referent groups, the odds of reporting concern were significantly higher for those living in an area that experiences disasters and for those previously impacted by a disaster, while they were significantly lower for those with Republican or Independent political identities.
For evacuation plans, the odds were significantly lower for those identifying as female, and significantly higher for those living in an area that experiences disasters and those previously impacted by a disaster. A significant interaction was also observed between rurality and income, which indicates that rurality affects the relationship between income and having an evacuation plan. More specifically, the prevalence of reporting that one had an individual or family evacuation plan was significantly higher for rural individuals with an income between $25,000 and $49,000, compared to their urban counterparts (Figure 2; Supplemental File).
Prevalence of reporting having an evacuation plan for rural and urban populations.
Regarding knowledge about where to access information, compared to referent groups, the odds of reporting that one knew where to access information were significantly higher for those 35 years of age or older, those with an annual household income of at least $25,000, those who lived in an area that experiences disasters, and those who had been previously impacted by a disaster.
Discussion
Rural and urban populations in the United States are exposed to the same types of climate-induced disasters but differ across many demographic and socioeconomic aspects that may hold implications for preparedness attitudes and behaviors. This analysis compared individual-level preparedness in rural versus urban environments and investigated the impacts of rurality on several demographic and socioeconomic variables that may influence attitudes and behaviors. The results suggest that preparedness attitudes and behaviors are generally shared across environments, but that rurality does interact with several population characteristics to hold important implications for individual-level preparedness for climate-induced disasters.
Results from this study revealed that nearly 88% of rural residents considered preparedness for climate-induced disasters to be important. The Rural Healthy People 2030 survey—which included medical professionals, government officials, academic researchers, and other individuals whose primary responsibility is to improve rural health—ranked emergency preparedness a relatively low health behavior priority (9/14) and overall priority for rural health (36/62).Reference Callaghan, Kassabian and Johnson6, Reference Ferdinand, Bolin and Callaghan7 While this research involved a different study population and did ask respondents to rank the importance of emergency preparedness against other health priorities, as was done in the Rural Healthy People 2030 survey, the results from this study suggest that individuals residing in rural environments do view emergency preparedness as an important health priority.
Only four of the attitudes and behaviors investigated in this study differed significantly between populations residing in rural and urban environments: having a personal or family evacuation plan, knowing where to access information in case of a climate-induced disaster, being concerned about the severity of climate-induced disasters in the future, and being concerned about the frequency of climate-induced disasters in the future. Of these, rurality was only found to play a significant role in influencing the odds of having an evacuation plan and the odds of being concerned about the severity of future climate-induced disasters.
Notably, the role that rurality plays appears to be mixed. On one hand, significant interactions existed between rural status, age, and income, suggesting that rural individuals between 35 and 44 years of age and those with an income between $75,000 and $99,999 were less likely than their urban counterparts to be concerned about the severity of future climate-induced disasters. These results may be explained, in part, by politics and education. Many climate-induced disasters are inherently linked to climatic processes and cycles, and political ideology represents a frequently replicated and discussed predictor of climate skepticism and concern in the United States. The existing literature suggests that those self-identifying as politically liberal generally express greater levels of concern about climatic and environmental change compared to those who identify as politically conservative.Reference McCright and Dunlap17–Reference Hornsey, Harris and Bain22 Indeed, meta-analysis has demonstrated that political ideology is a better predictor of climate skepticism than age, gender, race, and education.Reference Hornsey, Harris and Bain22 The results from this study align with these previous findings and suggest that the odds of reporting concern about the severity of future climate-induced disasters were significantly lower for Republicans and Independents. Further, when examining the prevalence of self-reported Republican or Independent political affiliations in rural and urban study populations, the greatest difference between study subpopulations occurred in the 35-44 years age group, with rural individuals in this age group reporting a much higher prevalence of Republican or Independent ideologies (Supplemental File). This may explain why significant differences were observed between rural and urban populations in this age group but not others.
Similarly, educational attainment is recognized as a demographic factor that can impact disaster risk perception.Reference Wachinger, Renn and Begg23 Typically, in the United States, individuals with lower levels of education perceive less risk from climate-induced disasters.24 The regression analyses in this study did not include educational attainment because it was highly correlated with income, but when examining the prevalence of attaining an associate’s degree or higher in rural and urban subpopulations, the greatest difference between study subpopulations occurred in the $75,000-$99,999 income bracket, with rural populations reporting lower levels of education compared to urban populations (Supplemental File). Therefore, the significant differences in concern about severity may be explained, in part, by differences in educational attainment in the rural and urban populations included in the study.
On the other hand, a significant interaction also existed between rurality and income that suggested that rural individuals with an annual household income between $25,000 and $49,999 were more likely to report having an evacuation plan, when compared to their urban counterparts. More specifically, results suggest that in rural populations, as income increases, so too does the prevalence of having an evacuation plan. Whereas, in urban populations, the prevalence of having an evacuation plan decreases for those with an annual income between $25,000 and $49,999, before increasing for higher income groups.
Several factors may also explain this result. Climate-induced disasters can result in significant societal disruptions due to damage to systems and infrastructure. In the aftermath of disaster events, a diverse suite of stakeholders can act to fill these gaps and remedy the provision of basic services as a part of disaster response and recovery efforts.Reference Sledge and Thomas25 However, because of the relative isolation and lack of capacity in rural environments, service provision can be more difficult in rural environments. These challenges could render the need to evacuate—and, accordingly, to develop evacuation plans—greater in rural environments when compared to urban environments. It is also noteworthy that the only significant interactions between rurality and income were observed for those reporting an annual income between $25,000 and $49,999. The financial burdens associated with certain individual-level preparedness behaviors, such as stockpiling supplies or procuring backup systems, are not negligible. Those of higher socioeconomic status in rural areas may have greater capacity to obtain these items, rendering their need to develop evacuation plans relatively less important compared to those of lower socioeconomic status.
Another plausible explanation for this result is rural values. In the United States, existing research has suggested that, when compared to urban populations, rural populations place a greater emphasis on values such as self-reliance, self-sufficiency, individualism, and distrust of outsiders.Reference Killian, Moon and McNeill26, Reference Keller and Owens27 These values may result in rural populations being more likely to develop evacuation plans that enable them to take matters into their own hands and render them less dependent on others or external assistance during disaster responses.
To note, these results are only of value if they are translated into practice by public health departments, emergency management agencies, and government officials involved in the response to climate-induced disasters. This research suggests that leverage points exist for changing preparedness attitudes in rural communities as a means of ensuring that climate-induced disasters do not exacerbate existing health equities between rural and urban environments. Public education and awareness campaigns represent, perhaps, the most fundamental opportunity for influencing attitudes surrounding preparedness for climate-induced disasters. Results from this work suggest that a majority of individuals residing in rural environments already consider climate-induced disaster preparedness to be important, so rural populations may also be open to learning about how climate-induced disaster risk could change in the future. This is especially true if campaigns seek to tailor messaging and materials.Reference Hawkins, Kreuter and Resnicow28 This could be accomplished, for example, by linking campaigns with rural values and deemphasizing the links to climate change, which may cause some to disengage due to cultural and identity politics.
Principles for effective risk communication also recommend that education campaigns seek to surpass knowledge provision—which is insufficient for prompting behavior change—and also provide clear, consistent, and actionable guidance that allows individuals to respond as needed and appropriate.Reference Campbell, Roper-Fetter and Yoder29 These campaigns would also benefit from messages being repeated in diverse fora. Indeed, “simple, clear messages, repeated often, by a variety of trusted messengers” represents one heuristic for effective risk communication.Reference Maibach, Uppalapati and Orr30 Communication campaigns in rural environments may require special considerations due to a relative paucity of communication outlets, which make it difficult to repeat messages often.Reference Campbell, Roper-Fetter and Yoder29 This, ultimately, underscores the importance of community engagement and partnerships. Two kinds of partners may be especially effective for targeting messaging in rural environments: those who are influential in one specific community (i.e., opinion leaders) and those whose influence reaches across communities (i.e., gatekeepers/bridge builders).Reference Katz and Lazarsfeld31 Ensuring that educational and awareness campaigns engage both types of partners in rural environments will be critical for maximizing the reach, acceptance, and uptake of climate-induced disaster preparedness messaging in rural environments.
Limitations
There are several limitations associated with this study. First, these data are cross-sectional and present a snapshot of attitudes and behaviors at a single moment in time. Results from this research and other efforts suggest that experience with disasters can result in changes to attitudes, behaviors, and perceptions.Reference Sattler, Kaiser and Hittner32–Reference Bourque34 As a result, should climate-induced disasters become more frequent and severe in the future, as has been predicted, attitudes and behaviors regarding climate-induced disasters may also change. Future work would benefit from longitudinal study designs that allow for research to be conducted over longer periods of time.
This study also relies on self-reported attitudes and behaviors, and the scope of this study renders it prone to social desirability bias, or the idea that respondents may try to align answers with what they perceive to be socially acceptable.Reference Bergen and Labonté35 For example, social desirability bias could have resulted in individuals reporting that they believed preparedness for climate-induced disasters to be important or that they had an emergency kit at home, when, in fact, they do not. It is difficult to determine how this bias may have impacted the results of this research that sought to compare preparedness attitudes and behaviors across environments that differ in relative rurality and sociocultural identities. Relatedly, respondents also self-reported their rural or urban status in this study. The concept of “rural” is multifaceted and means different things to different individuals. For example, population size and density, principal economic activity, service availability, and proximity to other population centers can all factor into rural-urban classification taxonomies.Reference Hart, Larson and Lishner5 Thus, what is rural to one individual may be suburban or urban to another. Future work should seek to validate these results using more defined measures of rurality and also investigate the accuracy of self-reported rurality measures.
Conclusions
This study compared individual-level preparedness for climate-induced disasters in rural versus urban environments and examined the impact of rurality on other demographic and socioeconomic characteristics that may influence preparedness attitudes and behaviors. Results suggest that a majority of climate-induced disaster preparedness attitudes and behaviors do not differ significantly across rural and urban populations. However, select attitudes and behaviors—namely, expressing concern about the severity of future climate-induced disasters and developing evacuation plans—do differ significantly between rural and urban environments, and rurality interacts with other demographic and socioeconomic characteristics to modify these relationships. In the context of a future that will likely witness climate-induced disasters increase in both frequency and severity, future qualitative and mixed-methods research should continue to investigate how cultural identities influence attitudes and behaviors for climate-induced disaster events as a means of promoting equitable public health preparedness and response.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1017/dmp.2025.10225.
Acknowledgments
We extend our sincere thanks to Tiffany Radcliff (Texas A&M School of Public Health) and Kirk Niekamp (USA Center for Rural Public Health Preparedness) for their roles in supporting this project.
Author contribution
MRB: Formal analysis, methodology, visualization, writing—original draft; writing—review and editing. CCB: Conceptualization, data curation, methodology, project administration, supervision, writing—review and editing; MR—Conceptualization, data curation, methodology, project administration, writing—review and editing; JBM: Conceptualization, funding acquisition, methodology, project administration, supervision, writing—review and editing.
Funding statement
Support for this work was provided by a Seedling Research Grant from the Texas A&M University Health Science Center.
Competing interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this manuscript.
Open access
