Exploring the association between public transit use and overweight and obesity: Evidence from Taiwan's county-level panel data

Climate change is intensifying extreme weather events and power outages, posing life-threatening risks to older adults who depend on electricity for medical devices, temperature regulation, and daily care. No nationwide study has systematically examined where power-outage risks overlap with the vulnerabilities of older adults. This study quantifies spatial patterns of power outage exposure among older adults across the United States (U.S.) from 2014 to 2021. We analyzed county-level power outage data covering 3108 counties, integrated with demographic data on older adults living alone and nursing home resident characteristics. We measured three outage dimensions: duration, frequency, and intensity. Using spatial clustering methods, we identified geographic hotspots where high outage exposure coincides with vulnerable populations, including those requiring intensive care and minority nursing home residents. The southeastern U.S., particularly Texas, Louisiana, and Florida, experiences high-frequency outages alongside large concentrations of older adults and minority populations. During the study period, Texas generated 386 county-year observations showing clustering in which Hispanic nursing home residents faced frequent outage interruptions. In Florida, 155 county-year observations were characterized by high-intensity outages affecting disproportionately large numbers of customers coinciding with high proportions of older adults requiring intensive care. Counties with higher proportions of minority nursing home residents were consistently associated with greater outage intensity compared with predominantly White counties. Tropical cyclones and extreme heat amplified these disparities. These findings reveal systematic inequities in power-outage exposure that threaten vulnerable older adults. The spatial clusters represent priority areas for infrastructure investment, emergency preparedness, and backup power resources. As climate change intensifies extreme weather and strains electrical grids, targeted interventions in high-risk regions are urgently needed to protect aging populations from compounding health risks.

Municipal consolidation, a widespread form of local government restructuring, has attracted growing scholarly attention worldwide. The majority of research on municipal consolidation investigates impacts instead of motives. Using prefecture- and county-level data from China, this study comprehensively examines the local drivers of county-to-district conversion (CTD) events during the 2010s, a period marked by a significant wave of CTDs. The results show that cities with a developable land shortage, a single district, or a higher economic ranking within a province are more likely to implement CTD. All else equal, counties in closer proximity to the central city, more lagged behind the city in development, or having a higher fiscal revenue per capita are more likely to be consolidated. Together, these factors explain about 40% of the odds of CTD at both the city and county levels. These findings highlight the importance of local incentives and characteristics in shaping jurisdictional changes and provide guidance for mitigating selection bias in future impact evaluations of municipal consolidation.

As a crucial national ecological barrier, China’s Southern Collective Forest Region (SCFR) plays an essential role in maintaining regional ecological security and promoting sustainable development. Understanding the mechanisms driving the evolution of its ecosystem service value (ESV) is of great significance. Based on county-level data from 2000 to 2023, this study integrated the equivalent factor method, spatial autocorrelation analysis, the XGBoost-SHAP model, geographically and temporally weighted regression (GTWR), and partial least squares structural equation modeling (PLS-SEM) to examine the spatio-temporal evolution patterns and driving mechanisms of ESV in the SCFR. The results showed that ESV in the SCFR exhibited an overall downward trend, with a cumulative loss of 1973.77 × 108 CNY. This was primarily due to marked reductions in hydrological and climate regulation services. The spatial distribution of ESV exhibited a significant heterogeneity—higher in the southwestern and southeastern mountainous regions, and lower in the northern plains and coastal zones, with the center of gravity shifting first to the northeast and then to the southwest. Local spatial autocorrelation revealed relatively stable “High–High” and “Low–Low” clustering characteristics, where high-value clusters were consistently distributed in core forest zones, while low-value clusters overlapped highly with urban agglomerations. Socio-economic factors exerted a significantly stronger influence on ESV than natural factors. Population density (POP), land use intensity (LUI), and gross domestic product (GDP) were identified as the dominant drivers, exhibiting distinct non-linear threshold effects and significant spatio-temporal heterogeneity. PLS-SEM analysis further quantified LUI as the dominant direct inhibitory pathway on ESV, highlighting urbanization’s indirect negative effect mediated through intensified LUI. Meanwhile, terrain effects were confirmed to positively influence ESV indirectly by constraining LUI and modulating local climate. The analytical framework of “threshold identification–spatio-temporal heterogeneity–causal pathway analysis” proposed in this study elucidated the complex driving mechanisms of ESV evolution, providing valuable guidance for ecological restoration evaluation and differentiated environmental governance.

All-cause midlife mortality rates have been increasing since 2010 in the United States. Using data from 1970 to 2010, this study investigates the association between county-level changes in economic, social, and employment sectors and changes in midlife mortality rates the occurred between 2010 and 2018. The study employs a novel approach to analyze temporal trends. County-level mortality data for 2009-2019 were obtained from the Centers for Disease Control and Prevention (CDC), while decennial data for 19 indicators-covering socioeconomic conditions, social factors, and employment sectors-were obtained from IPUMS NHGIS time series tables and the US Bureau of Economic Analysis, Economic Profile by County. Data were examined for 3,069 (97.6%) of the 3,143 U.S. counties and county equivalents. Absolute changes in county characteristics were measured over ten possible comparison periods: single decades, two decades, three decades, and four decades. LASSO regression was used to identify significant predictors and assess their impact over multiple time periods. While changes in some county characteristics (e.g., households headed by single mothers, employment in certain sectors, college education, and labor force participation), tend to be associated with higher or lower mortality risk; in many cases the strength and direction of observed associations differed depending on time period, place, and race. These results reveal the importance of historical and contextual factors in understanding mortality trends and highlight the complex interplay between social determinants and health outcomes. This study provides insights into the drivers of midlife mortality and a nuanced look at the temporal dynamics and geographic variations in mortality trends. By identifying critical time periods and specific predictors associated with mortality changes, the study informs policy and public health efforts aimed at reducing mortality disparities and improving population health outcomes.

When 287(g) immigration enforcement agreements surged from 135 to 1,035 across U.S. agencies in 9 months during 2025, adoption patterns revealed surprising dynamics of local policy diffusion under federal-state coercion. Using comprehensive county-level data and Cox proportional hazards models with time-varying covariates, I find that geographic proximity dominated adoption decisions despite unprecedented top-down pressure. Each neighboring adopter increased a focal county’s adoption likelihood by 25%, the single strongest predictor. State-level peer effects operated independently, creating compounding regional pressures. Fiscal findings challenge conventional wisdom: counties more dependent on intergovernmental transfers were 64% less likely to adopt, while debt burdens and deficits showed no effects, revealing that institutional capacity, not fiscal desperation, determines responsiveness to federal incentives. State mandates accelerated adoption but did not eliminate geographic variation in timing. These findings demonstrate that spatial networks structure local policy decisions even when higher governments actively direct outcomes, with important implications for understanding federalism, policy diffusion mechanisms, and the limits of coercive intergovernmental relations.

This study aims to determine the effect of the COVID-19 pandemic on changes in late HIV diagnoses by race/ethnicity across Ending the HIV Epidemic (EHE) priority jurisdictions in the US. We analyzed annual county- and state-level (when county-level data were unavailable) late HIV diagnosis data in EHE priority jurisdictions from local epidemiological profiles and the AIDSVu between 2017-2022. Descriptive analyses were conducted to examine the percentages of late diagnoses across racial/ethnic groups before and after the onset of the pandemic. We then used interrupted time-series analysis to assess changes in both the level and trend of late diagnosis percentages across non-Hispanic White, non-Hispanic Black, and Hispanic populations. We included a total of 31 jurisdictions. The highest percentages of late HIV diagnoses were more often reported among Black or Hispanic populations. There were statistically significant (P<0.05) downward trends in late diagnosis percentages before the pandemic among the Black population in Kings County, NY(-3.8%), Georgia(-0.8%), and Los Angeles County, CA(-4.5%), and among the Hispanic population in Georgia(-4.9%). Subsequently, there were significant immediate increases in 2020 in the Black population in Kings County, NY(11.5%), Georgia(2.5%), and Los Angeles County, CA(10.8%), and among the Hispanic population in Georgia(11.5%). Additionally, there was a significant annual increase in the trend after the onset of the pandemic in both the Black(1.0%) and Hispanic populations(4.8%) in Georgia. In contrast, no statistically significant changes were found in the White population. The COVID-19 pandemic appears to have exacerbated existing race/ethnic disparities in late HIV diagnoses among several EHE jurisdictions.

Background: Long-term PM2.5 exposure is a major health risk, yet the heterogeneity of its chemical components and sources is often overlooked in risk assessments. Methods: Using county-level census data for 1990, 2000, and 2010 in China, we employed a difference-in-differences (DID) study design to associate all-cause mortality with long-term exposure to PM2.5 from five sources and by five components. We applied log-linear models and several mixed exposure models (weighted quantile sum (WQS), ridge regression, random forest, XGBoost, and an ensemble model) to capture effect heterogeneity with cross-validation for evaluating the prediction performance and Shapley additive explanations (SHAP) values for interpretability. Results: Each 10 μg/m3 increment in PM2.5 was associated with a 3.16% increase in all-cause mortality (95% confidence interval [95% CI]: 2.13-4.20%). Industry (3.46%, for each interquartile range increment), power (2.32%), and residential usage (1.46%) were the most hazardous sources; ammonium (6.46%), nitrate (4.23%), and black carbon (3.93%) were the most potent components. Random forest and the ensemble model showed superior predictive performance, identifying industrial-sourced black carbon as the leading contributor to PM2.5-associated mortality. Conclusion: The health effects of PM2.5 differ considerably by the source sector and chemical component. Industrial black carbon exhibits higher toxicity and should be given priority when designing the air-pollution-control strategies.

ABSTRACT Governments often prioritize arts and culture in strategies to revitalize economically depressed communities. However, little is understood about the broader impact of governance structures in cultural districts on community-related outcomes. To start this process, we ask: What governance structures and local development aims shape cultural districts? We develop a unique relational database on cultural districts’ characteristics and governance structures. Data comprises county-level data from state cultural resources and arts agencies, the National Register of Historic Places, and US census data on community and socio-spatial characteristics of the cultural districts. This exploratory study focuses on the Southern US, where many of the poorest counties and cities in the United States are located. Understanding the governance structures and development aims of these cultural districts has implications for capacity building within these organizations, especially among nonprofits, to develop collaborations and alliances for mutual and broad social, economic, and environmental impact.

ABSTRACT Enhancing rural tourism green total factor productivity (RTGTFP) is vital for sustainability. Using Chinese county-level data (2018–2021), this study develops an RTGTFP indicator system incorporating multi-stakeholder inputs, rural welfare, and environmental outputs. Evolution is evaluated via the Super-SBM model and ML index. Technology-Organization-Environment (TOE) conditions are examined through dynamic fuzzy-set Qualitative Comparative Analysis (fsQCA) across steady (2018–2019), shock (2020), and recovery (2021) periods. Results show significant temporal variation; no single TOE dimension is necessary for high RTGTFP. Instead, multiple pathways emerge, including environment-anchored, technology-organization synergy, and comprehensive synergy models. Phase comparisons reveal shock contingency: no robust configuration exists for 2020, while 2021 shows a narrower environment-led solution.

The implementation of COVID-19 vaccination in the United States has revealed substantial disparities driven by geography, socioeconomic conditions, and political ideology. This study examines the association between these factors and COVID-19 vaccination uptake across 360 counties in four U.S.-Mexico border states, characterized by distinct socio-political traits. Using county-level data, this study employed multivariable regression analysis and GIS mapping to assess the effects of income, education, employment, age, race, ethnicity, occupation, metropolitan status, border status, and political affiliation on Dose 1, Dose 2, and booster vaccination rates. The analysis showed that Dose 1 vaccination rates were significantly higher in border counties and metropolitan areas. Democratic population share and per capita income were positively associated with vaccination uptake. Dose 2 vaccination rates exhibited patterns similar to those observed for Dose 1. Booster vaccination rates were positively associated with Democratic affiliation, the proportion of the population with at least a high school education, and the share of individuals aged 65 years and older. In contrast, unemployment rates were negatively associated with booster uptake. Racial and ethnic composition was also associated with vaccination outcomes: higher Black population shares were associated with lower Dose 1 vaccination rates, whereas higher Native American population shares were associated with higher vaccination rates. Booster uptake was higher with larger shares of the Asian population but slightly lower with larger shares of the White population. COVID-19 vaccination uptake in U.S.-Mexico border counties was associated with a complex interaction of geographic, socioeconomic, demographic, and political factors. These findings underscore the importance of targeted, context-specific public health strategies to reduce vaccination disparities and improve booster coverage in border regions.

Lung cancer is the leading cause of cancer-related mortality in the United States. While smoking cigarettes remains the dominant risk factor, declining smoking rates have drawn attention to Radon. Despite the known risks from individual exposure to radon, how long-term lung cancer trends vary by county-level Radon risk classification remains unknown. Using SEER-8 cancer registry data from Connecticut, Iowa, New Mexico, and Utah, we compared lung cancer incidence between counties classified by the EPA as low-risk and high-risk (>4.0 pCi/L) for radon exposure. Constructing a random effects Generalized Least Squares regression, which adjusted for county-level smoking prevalence, we quantified the excess lung cancer rate attributable to radon risk across decade, sex, and histologic subtype. Compared to low-risk counties, counties at high risk of radon exposure had a significantly higher overall lung cancer incidence: 13.5+cases per 100,000 person-years (95% CI: 10.0, 17.1). This gap between low and high-risk counties was largest for Adenocarcinoma and small cell carcinoma, and larger in males than females. However, only in females did we observe the gap in lung cancer incidence between low- and high-risk counties grow decade after decade. This study illustrates how county-level radon data can be leveraged to enhance cancer surveillance and guide prevention and control strategies. Further research is needed to evaluate the impact and cost-effectiveness of radon prevention or mitigation policies, especially in females. We argue that public health systems, over the next fifty years, must prioritize eliminating the two main drivers of America's deadliest cancer.

Abstract Authoritarian regimes have long faced governance challenges arising from decentralization, as central governments struggle to control local government behavior due to information barriers. This paper argues that promoting the disclosure of local government information to the public is an effective strategy to alleviate issues associated with decentralization. Using a policy mandating the disclosure of local government information on social governance as a quasi-natural experiment, we examine its impact on a critical governance challenge emerging from decentralization – the decline of county seats. Using China’s county-level panel data (2015–2022) and a difference-in-differences approach, we find that local information disclosure significantly promotes county seat development. Specifically, it increases land allocation for both economic and public service purposes, thereby breaking the vicious cycle between deficient public services and economic stagnation. Heterogeneity analysis indicates that these effects are more pronounced in counties with higher citizen responsiveness and more constrained fiscal capacity.

This paper investigates the impact of casino openings on residential housing prices in Arkansas, focusing on Hot Springs, Pine Bluff, and West Memphis, along with their respective counties. Following the policy change in 2018, we utilize the synthetic control method with city- and county-level Zillow housing price data (2013–2025) to evaluate market responses. We find that the city of Hot Springs and its encompassing Garland County experienced a statistically significant increase in housing prices following the transition to full casino gaming in April 2019. In contrast, Pine Bluff, West Memphis, and their counties (Jefferson and Crittenden) showed no discernible change in housing prices attributable to their respective casino openings. Robustness checks using Realtor.com listing price data (2016–2025) yield consistent results. This mixed evidence suggests that the effects of casino openings on housing markets are not uniform and are likely influenced by preexisting local characteristics and market conditions.

Counties have been continuously serving as basic administrative divisions throughout ancient and modern China, and population has always been the focus of national governance. Thus county-level population data is of great significance for understanding spatial pattern of governance in ancient China, and is also basic data for studying urban system of ancient China. However, due to twists and turns in history, most of the county population statistics before Qing Dynasty (1644~1911AD) have been lost, and need to be obtained through data reconstruction process. Based on publicly available historical statistical data on prefecture level, this study combines raster terrain data and vectorized historical geographic data, simulates population distribution of 4 representative dynasties through geo-calculation method, and estimates the population data of 6,089 counties in the years 2, 742, 1102 and 1820AD. By preliminarily reconstructing population distribution and county-level population sequence of representative time slices, the evolution of key administrative districts and habitat regions in different eras has been revealed.

Abstract This study treats the selection of land conservation and intensive use model counties as a quasi-natural experiment. Using Chinese county-level panel data, we evaluate the multidimensional impacts of the land conservation and intensive use policy (LCIUP). We find that LCIUP reduced PM2.5 concentrations in counties while simultaneously lowering per capita GDP, exerting a positive effect on environmental quality but a negative inhibitory effect on economic growth, showing a distinct environment–economy asymmetry. LCIUP restricts industrial land supply and curbs the entry of polluting enterprises, but fails to facilitate industrial transformation and upgrading. Counties reliant on secondary industries face significant industrial transformation lock-in challenges, while those with substantial market potential can achieve dual economic and environmental goals. Green finance policies effectively complement LCIUP to promote industrial transformation and upgrading, whereas technological innovation and talent attraction policies currently lack such synergy. Cost–benefit analysis confirms that LCIUP’s marginal environmental benefits outweigh economic losses.

Mapping cancer incidence is crucial for analyzing and visualizing patterns across geographic areas. While many studies map cancer incidence at subnational levels (e.g., state, county), publicly available county-level data, especially for less common cancers, is limited. Using data from the NAACCR CiNA research database (2005 to 2019), we developed spatio-temporal hierarchical models to smooth/predict annual age-group-specific case counts for all U.S. counties. We compared Poisson and zero-truncated Poisson likelihoods and various priors. Model performance was assessed using Deviance Information Criterion (DIC), Weighted Akaike Information Criterion (WAIC), and average absolute relative deviation (AARD). Modeled age-adjusted rates were mapped to visualize spatial and temporal patterns. Modeled age-adjusted rates were produced for 16 selected sex-specific cancer sites across 3,109 counties from 2005 to 2019. AARD values varied by site and context, being lowest for common cancers and populous counties and highest for rare cancers and sparsely populated areas. Compared with maps of observed rates, modeled maps were smoother and more coherent, filling gaps and reducing extreme values driven by small case counts, and preserving large-scale geographic gradients and temporal trends. The standard Poisson hierarchical mixed-effects model showed superior accuracy and computational efficiency and was selected for final estimation. As expected, the most accurate predictions are for more common cancer sites in more populous areas, and the least accurate predictions are for rarer cancers in areas with lower population. The resulting estimates and maps could support surveillance, trend analysis, disparity identification, targeted interventions, and broader research efforts.

Addressing social determinants is critical to achieving the End TB Strategy targets. However, the impact of tuberculosis (TB) programmatic indicators amenable to short-term interventions on TB incidence remains underexplored. We aimed to identify modifiable social determinants and evaluate their impact on TB incidence in different risk-specific areas. We compiled individual-level surveillance data on reported pulmonary TB (PTB) cases and county-level data on 12 social determinants between 2011 and 2018 across 108 counties in Shaanxi Province, China. A Bayesian spatiotemporal model quantified associations between PTB incidence rate and social determinants. Model-estimated relative risks (RRs; x-axis) and their annual percentage change (APC; y-axis) were integrated to construct an RR–APC plane diagram for classifying different risk-specific counties. Counterfactual analysis estimated potential reduction in PTB incidence rate if the referral and tracing percentage for PTB patients and suspected cases or the percentage of PTB cases confirmed bacteriologically in these counties were increased to 2018 provincial benchmarks. Five social determinants were significantly associated with PTB incidence rate. A 1% increase in the referral and tracing percentage for PTB patients and suspected cases, per capita housing area, and the percentage of PTB cases confirmed bacteriologically corresponded to reductions in PTB incidence rate of 0.39%, 0.39%, and 0.04%, respectively. In contrast, a 1% increase in the percentage of population aged ≥ 50 years and in the percentage of illiterate population was associated with increases of 0.35% and 0.07%, respectively. The RR–APC plane diagram identified three types of risk areas: four counties in Ankang Municipality with rapidly escalating risks, three counties in Yulin Municipalities with persistently high risks, and three newly emerging high-risk counties in Ankang and Shangluo Municipalities. Counterfactual analysis suggested that increasing the referral and tracing percentage to 94% could reduce PTB incidence rate by 9.24/100,000 to 38.54/100,000 (relative reductions of 8.49% to 28.06%) across different risk-specific counties, and simultaneous improvement in both indicators could achieve reductions of 10.48/100,000 to 38.99/100,000 (9.62% to 28.38%). The referral and tracing percentage for PTB patients and suspected cases and the percentage of PTB cases confirmed bacteriologically are modifiable social determinants on TB incidence rate. Strengthening the effectiveness of referral and tracing mechanisms and expanding bacteriological diagnostic capacity across risk areas may substantially reduce TB burden and accelerate progress towards End TB Strategy targets.

Introduction Amid global economic shifts and rural decline, promoting rural industrial integration has become a core strategy for China’s Rural Revitalization and sustainable farmers’ income growth. However, existing studies mostly focus on macro-policy interpretation or case descriptions, lacking mechanism testing and heterogeneity analysis based on large-sample county-level data. Jiangxi Province, a typical agricultural province with distinctive terrain and urgent rural industrial revitalization needs, provides an ideal empirical context. This study aims to systematically explore the impact, transmission mechanisms, and heterogeneous characteristics of rural industrial integration on farmers’ income. Methods Using panel data from 95 counties in Jiangxi Province during 2014–2023, we constructed a Rural Industrial Integration Index (RIII) covering four dimensions: chain extension, multi-functionality expansion, multiformat integration, and technology penetration. A combination of two-way fixed effects models, mediating effect models, and full quantile regression methods was employed to conduct empirical analysis. Results The RIII exerts a significantly positive effect on farmers’ income: every 1% increase in the RIII leads to a 3.1–3.3% growth in rural residents’ Per Capita Disposable Income (PCDI). Non-agricultural employment serves as a key transmission mechanism, with its mediating effect contributing 23.74% of the total effect. The income-increasing effect exhibits notable heterogeneity, being more pronounced in middle-low income counties, hilly terrain areas, and central Jiangxi regions. Discussion This study enriches agricultural economics theories by verifying the causal relationship between rural industrial integration and farmers’ income. The findings provide empirical evidence for formulating regionally differentiated rural industrial integration policies, offering critical decision-making insights for boosting farmers’ income growth and advancing rural revitalization strategies in Jiangxi and similar agricultural regions.

Wheat plays a critical role in global food security; however, its vulnerability to rising temperatures introduces significant uncertainty about future yields in a changing climate. Although earlier studies have linked higher temperatures to yield reductions, the moderating influence of terrain elevation on crop-climate interactions remains insufficiently explored. We combine 42 years of county-level yield data (1982-2023) with phenology-specific climate exposures to evaluate how terrain elevation shapes U.S. winter wheat responses to extreme heat and precipitation. Using an econometric framework with county-specific temperature thresholds for extreme degree days, we identify a critical elevation cutoff at 350 meters that delineates two distinct yield-climate regimes. Low-elevation counties exhibit faster long-term yield growth but greater vulnerability to late season heat stress compared to high-elevation counties. Indeed, the late season coincides with the grain-filling stage, which is critical for winter wheat yield outcomes. Rolling-window estimates further reveal that heat-related yield losses have intensified since the 1980s, with late season penalties nearly doubling. Trends also indicate stronger vulnerability among low-elevation counties, especially in the recent period (2004-2023). These findings demonstrate that topography fundamentally mediates climate risks to wheat production. Adaptation may therefore require not only a latitudinal but also an elevational redistribution of wheat cultivation, reshaping the geography of U.S. production under climate change. More broadly, the results underscore the importance of integrating terrain elevation into climate-crop assessments to improve yield projections and inform adaptation strategies across diverse agricultural systems.