Spatial Regression Research Articles

Economic and institutional determinants of environmental health and sustainability: Spatial and nonlinear effects for a panel of worldwide countries

Research background: This study identifies the key factors influencing environmental health across a global panel of countries, focusing on protection from environmental hazards, as informed by the existing literature, while also shedding light on novel aspects of these causal relationships. Purpose of the article: This study aims to reveal, through a comprehensive review of the relevant literature, the underexplored phenomena of spatial diffusion and contagion of national environmental behaviors and the nonlinear dynamics between environmental performance and its determinants, acknowledging the significant diversity in the characteristics and behaviors of the countries studied. Methods: Spatial analysis and econometric methods, including spatial panel regression alongside dynamic panel models using threshold techniques, were employed to meet the study’s objectives. Findings & value added: This study’s major finding is that environmental performance across nations shows significant clustering influenced by economic and institutional factors. This clustering effect arises from spatial contagion and diffusion processes, as evidenced by spatial panel regression analysis. Furthermore, this study demonstrates that variations in environmental behavior can be attributed to differing levels of development and specific internal conditions within countries. Notably, a country’s gross domestic product and the proportion of industries in its economy have a substantial effect on its environmental health practices, establishing distinct impact thresholds. This research enriches academic dialogue by illustrating, through these thresholds, that in less developed countries, an increased industrial share leads to environmental degradation. Moreover, the influence of the other examined factors varied depending on the category of the country under review, highlighting the nuanced effects of economic and institutional variables on environmental outcomes.

Inequalities in urban green space distribution across priority population groups: Evidence from Tāmaki Makaurau Auckland, Aotearoa New Zealand

Growing evidence has shown that urban green space (UGS) in neighbourhoods can substantially enhance residents' well-being. This study focuses on the possible inequitable UGS distribution among priority population groups of city-dwellers in the region of Tāmaki Makaurau, Aotearoa New Zealand. UGS availability was calculated using coverage ratio of service area and Overlap/cumulative opportunity. UGS accessibility was measured using Gaussian-based two-step floating catchment area. Bivariate analysis, spatial regression, ordinary least squares estimation, and Moran's I were employed to explore the potential inequality between UGS distribution and priority population groups. The results revealed that most deprived neighbourhoods tended to have more UGS availability but less accessibility. Children, adolescents and elderly were likely to have less potential opportunity to choose different UGS to visit. Areas with a higher proportion of Māori had less UGS availability and Pacific populations had less accessibility. Neighbourhoods with a higher proportion of Asian people had more UGS availability but less accessibility. This study adds new evidence to UGS inequality in priority population groups by contributing a regional investigation in a developed country. The results can inform policy makers and planners on where and how to implement UGS and on which groups to focus on to prevent further disparities.

Spatial analysis of American cutaneous leishmaniasis in the state of Amazonas.

To evaluate, using spatial analysis, the occurrence of American Cutaneous Leishmaniasis (ACL) and analyze its association with the municipal human development index (MHDI) and deforestation in the state of Amazonas, Brazil, from 2016 to 2020. This ecological study, carried out from January 2016 to December 2020, included the 62 municipalities of the state of Amazonas. The incidence rate of ACL was determined in space and time. Using Multiple Linear Regression by Ordinary Least Squares (OLS) and Spatial Autoregressive Regression (SAR) models, the relationship between incidence rates and Human Development Index (HDI) and deforestation was analyzed., The high- and low-risk clusters were identified by employing the Getis-Ord Gi* statistic. A total of 7,499 cases of ACL were registered in all 62 municipalities in the state. Most cases were in male (n=5,924; 79.24%), with the greatest frequency in the population aged from 20 to 39 years (n=3,356; 44.7%). The incidence rate in the state of Amazonas was 7.34 cases per 100,000 inhabitants-year, with the municipalities of Rio Preto da Eva and Presidente Figueiredo showing the highest rates (1,377.5 and 817.5 cases per 100,000 population-year, respectively). The ACL cases were clustered into specific areas related to those municipalities with the highest incidence rates. The SAR model revealed a positive relationship between ACL and deforestation. The occurrence of ACL was evident in a variety of patterns in the state of Amazonas; the high incidence rates and persistence of this disease in this state were linked to deforestation. The temporal distribution showed variations in the incidence rates during each year. Our results can help optimize the measures needed to prevent and control this disease in the state.

A novel framework for assessment of human impact of floods: Demonstrated for the Indian subcontinent

Flood damages are a result of mutual interaction between flood hazard and socio-economic factors. The increase in flood damages over time can be attributed to various enhancements in the magnitude and frequency of floods as well as socio-economic development in the floodplains. The quantification and analysis of the human dimensions of flood damages is challenged due to irreversibility and lack of understanding of complex-human flood interactions, especially in less developed regions. Since flood damage assessments play a critical role in designing effective flood preparedness plans, we suggest including the human dimension in evaluating damages. We present a framework for the assessment of the human impact of floods by incorporating statistics on the number of people exposed, displaced, and fatalities recorded for each flood event. The framework involves categorizing human impact, identifying spatial clusters of different categories, and exploring the role of various hazard and socio-economic characteristics in determining impact. While categorizing and clustering of impact helps to distinguish between regions prone to different impact, understanding the role of socio-economic and demographic factors is expected to help in developing tailor-made strategies to mitigate flood disasters. The methodology broadly consists of categorization using unsupervised learning, spatial clustering using Local Indicators of Spatial Autocorrelation (LISA), and spatial regression. The results indicate that the impact varies from one region to another and so does the role of hazard and socio-demographic variables in determining impact. Moreover, the framework developed may be used to study the relevance of various socio-hydrological theories and their role in determining the human impact of floods.

Regional comparison of socio-demographic variation in urban E-scooter usage

In recent years we have witnessed explosive growth in the shared, free-floating, electric scooter industry. While still controversial in many North American cities, a number of large e-scooter operators have managed to carve out a piece of the urban transportation landscape. As these vehicles shift from novelty services to increasingly reliable modes of short personal travel, the discussion has turned to investigating who exactly benefits from these micromobility services and who are being left behind. Though population surveys have been administered to identify the socio-demographic characteristics of e-scooter riders in the past, little work has linked these characteristics through trips, or investigated the regional variation in these demographic factors. In this work we explore the variability and similarities in e-scooter rider characteristics across three major U.S. cities. To accomplish this, we apply a Moran’s Eigenvector Spatial Filtering linear regression model and compare our results to more commonly used spatial regression approaches. Our results indicate that the spatial filtering approach outperforms other methods in identifying socio-demographic characteristics of e-scooter users, across multiple regions. We find that many socio-demographics associated with e-scooter usage are regionally variant, despite younger users making up the core user base in all cities. There are variations in usage based on gender, income, and race across cities with Black and Hispanic populations remaining underserved. The implications of these findings are discussed.

Identifying local foci of tuberculosis transmission in Moldova using a spatial multinomial logistic regression model

BackgroundMultidrug resistant tuberculosis (MDR-TB) represents a major public health concern in the Republic of Moldova, with an estimated 31% of new and 56% of previously treated TB cases having MDR disease in 2022. A recent genomic epidemiology study of incident TB occurring in 2018 and 2019 found that 92% of MDR-TB was the result of transmission. The MDR phenotype was concentrated among two M. tuberculosis (Mtb) lineages: L2.2.1 (Beijing) and L4.2.1 (Ural). MethodsWe developed and applied a hierarchical Bayesian multinominal logistic regression model to Mtb genomic, spatial, and epidemiological data collected from all individuals with diagnosed TB in Moldova in 2018 and 2019 to identify locations in which specific Mtb strains are being transmitted. We then used a logistic regression model to estimate locality-level factors associated with local transmission. FindingsWe found differences in the spatial distribution and degree of local concentration of disease due to specific strains of Beijing and Ural lineage Mtb. Foci of transmission for four strains of Beijing lineage Mtb, predominantly of the MDR-TB phenotype, were located in several regions, but largely concentrated in Transnistria. In contrast, transmission of Ural lineage Mtb had less marked patterns of spatial aggregation, with a single strain (also of the MDR phenotype) spatially clustered in southern Transnistria. We found a 30% (95% credible interval 2%-80%) increase in odds of a locality being a transmission cluster for each increase of 100 persons per square kilometer, while higher local tuberculosis incidence and poverty were not associated with a locality being a transmission focus. InterpretationOur results identified localities where specific Mtb transmission networks were concentrated and quantified the association between locality-level factors and focal transmission. This analysis revealed Transnistria as the primary area where specific Mtb strains (predominantly of the MDR-TB phenotype) were locally transmitted and suggests that targeted intensified case finding in this region may be an attractive policy option.

Tree abundance, species richness, or species mix? Exploring the relationship between features of urban street trees and pedestrian volume in Jinan, China

Accumulating evidence has confirmed that urban greenery, especially street trees, is beneficial to walking behaviors. Existing street greenery exposure-walking behavior studies have focused on the quantity of greenery, which is often measured by normalized difference vegetation index (NDVI) or green view index (GVI). However, some important qualities of street trees, the dominant component of urban greenery, were often overlooked, due to the labor-intensive and expensive nature of on-site surveys. To address this issue, we proposed a cutting-edge deep learning technique to identify street tree species at the individual tree level. We established a citywide dataset of all street trees with species information via 185,831 Baidu Street View images (BSV) in Jinan, China. Population-level walking intensity, measured by pedestrian volume, was retrieved from BSV images using Baidu AI. We further adopted spatial regression models to investigate the association between pedestrian volume and street tree characteristics, including street tree abundance (number of street trees), species richness (number of unique tree species) and species mix (the degree of diversity of tree species). The built environment and urban greenery covariates were adjusted in the models. The results indicate that the street tree abundance and species mix are positively associated with pedestrian volume. Species richness is not associated with it. Besides, spatial mismatch is identified between abundance and species mix of street trees in the study area. Hence, to facilitate walking behavior and deliver related health benefits, it is necessary to develop fine-grained measures of street greenery features.

Spatio-Temporal Evolution Characteristics and Influencing Factors of INGO Activities in Myanmar

Myanmar is among the regions with the most frequent activities of International Non-Government Organizations (INGOs). Analyzing the spatio-temporal patterns of these activities holds crucial importance for optimizing organizational coordination and enhancing governmental oversight. This study focuses on the spatio-temporal evolution characteristics and influencing factors of INGO activities in Myanmar from 2010 to 2021, utilizing spatial autocorrelation and regression analysis. The results show that the number of INGOs in Myanmar has shown a gradual slowdown in growth trends, with the number of activities exhibiting a wave-like pattern, primarily driven by spontaneous activities of INGOs. The spatial distribution of INGO activities in Myanmar is concentrated in the southern plains, with the core located in Yangon, Naypyitaw, and Loilen. Furthermore, there is significant spatial polarization in the hotspot area of INGO acticities. The hotspots followed an evolutionary path from “South Myanmar” to “North Myanmar” and then back to “South Myanmar”. INGO activities in Myanmar are more focused on the local economic level, urbanization level, medical level, education level, and total population size, providing the necessary support and services for the local society and making up for the “government malfunction” and “market malfunction”.

Testing the Zipf’s law under heterogeneous urban-rural hierarchies with spatial quantile regressions

ABSTRACT Despite the widespread recognition of Zipf’s law in urban economics, empirical evidence reveals deviations from the strict adherence to its postulated parameters. In our contribution to this ongoing debate, we applied Zipf’s law to population density of a metropolitan hierarchy in Greece (1033 municipalities) over a substantial timeframe (1961–2021). To overcome the specific limitations of global Ordinary Least Squares (OLS) models, cross-section spatial quantile regressions had been performed. The outcomes unveiled heterogeneous rank-size relationships for urban municipalities corresponding with the first and the second quartile of density distribution. These relationships became more homogeneous in rural municipalities (third and fourth quartiles). Consistent with the predictions of Zipf’s law, slope coefficients exhibited a notable increase over time, approaching 1 for mixed urban-rural municipalities. Conversely, they remained slightly below 1 for strictly urban municipalities and slightly above 1 for strictly rural ones. Space was found to play a significant role during late urbanization (1980s-1990s), suggesting an impact of agglomeration economies on metropolitan hierarchy. Suburbanization exerted a reverse effect resulting in more heterogeneous hierarchies since the late-1990s. Our findings underscore the intricate nexus between deviations from Zipfian patterns in urban-rural population distribution and the spatially explicit dynamics of city life cycles.

Analyzing the relationship between the experience of intimate partner violence and female internet use in Nigeria

Female internet use is abysmally low (17.48%) in Nigeria, and it has become an issue of serious concern due to its negative impact on their health and economic well-being. While most of the scholarly debates centers on the influence of individual socio-economic characteristics and geographical factors to understand and improve female internet use, this study contributes to the literature by interrogating the role of women's experience of intimate partner violence (IPV) and other socio-economic and geographical factors based on the social theory of internet use. We conduct spatial and logistic regression analyses using the 2018 cross-sectional Demographic and Health Survey of 41,821 women aged 15–49. The spatial analysis shows a significant concentration of female internet non-use in the Northern region relative to the South, while the binary logistic regression analysis indicates that the experience of IPV, age, wealth status, education, being married, urban location, and residing in Northern Nigeria are significant predictors of female internet use. The study recommends tailoring support systems and interventions that acknowledge the unique challenges faced by survivors of IPV and the improvement of the socio-economic conditions of women to achieve greater internet use.

Spatial Regression Modeling of Child Survival on the Distribution of Births and Deaths in Kenya Based on the Kenya Demographic and Health Survey (KDHS) 2022

This study used spatial mapping techniques to examine the distribution of births and deaths in Kenya and their relationship with various factors related to child survival, such as maternal age, education, wealth, and access to health services. Data were obtained from the 2022 Kenya Demographic and Health Survey (KDHS). Spatial autocorrelation analyses were conducted to identify clusters of high or low child mortality rates. The results showed significant spatial autocorrelation in child mortality rates, indicating that neighboring areas had similar mortality rates. Factors such as maternal education, wealth, and access to health services were found to be significantly associated with child mortality rates. These findings can inform targeted interventions and policies to reduce child mortality rates in Kenya, particularly in areas with the highest risk of mortality.

Analysis of the funding of social services from a spatial approach in Andalusia (Spain)

In Andalusia, Community Social Service Centres (CSSCs) are funded by regulations following demographic, geographic, economic, and social disadvantage criteria. This study aims to analyse the geographical distribution of funding per inhabitant of CSSC by catchment area in Andalusia in 2019, and to study the statistical associations between funding and a range of demographic and socioeconomic indicators. The study spatial units (n = 184) included the catchment areas of CSSCs and, in the case of intramunicipal areas in large municipalities, they were grouped at the municipal level. Spatial autocorrelation measures were used to identify spatial clusters of high/low funding rates per inhabitant. Later, nonspatial and spatial regressions were applied to search for associations with different indicators (sex ratio, ageing index, dependency index, emigration rate, immigration rate, unemployment rate, population density, and employment rate in the primary sector). The geographical distribution of the funding of social services in Andalusia was not random since the analyses identified several spatial clusters with significantly high (hot spots) and low (cold spots) funding per inhabitant (p < 0.05). The funding rates were significantly (p < 0.05) and directly associated with the ageing index and the percentage of primary sector employees, and indirectly with the proportion of foreigners in the population and the population density. The hot spots were mainly located in rural and deprived areas, while the cold spots were in urban areas. The variables related to the regulated funding distribution criteria were not fully associated with higher financing. Instead, other additional variables showed significant associations (p < 0.05), such as primary-sector workers and foreign populations. The results showed that spatial analyses may support service assessment and decision-making in social policy.

Geographic isolation and learning: Evidence from rural schools in Colombia

Rural schools are usually behind in terms of learning, and this could be partly related to geographical isolation. We explore this hypothesis, assessing the causal effect of the distance between schools, towns, and State capitals on elementary school test scores. We use granular administrative records from Colombia and estimate spatial regression discontinuity models. Results indicate that the distance to both towns and State capitals negatively affect students test scores. The differences in educational inputs, such as teachers education attainment and contract stability, partly explain these gaps. A program providing monetary incentives to teachers in difficult access areas fails to attenuate the isolation effect.

Spatio-Temporal analysis of exports of cultural products and their affecting factors for spatial distribution.

Cultural products constitute a significant portion of global trade, and understanding their export patterns can shed light on economic trends, trade dynamics, and market opportunities. This study conducted the spatio-temporal analysis of exports of cultural products, exploring the relationship between various influencing factors and their impact on the spatial distribution of these exports. Leveraging a diverse dataset encompassing 55 BRI countries for the period of 2005-2022, this research employs advanced spatial analysis techniques, including spatial autocorrelation and spatial regression models, to examine the spatial patterns and determinants of exports if cultural product exports. Moreover, this study delves into the multifaceted determinants affecting the spatial distribution of these exports. The findings of this study reveal significant spatio-temporal variations in the exports of cultural products. Spatial autocorrelation analysis indicates the presence of spatial clustering, suggesting that regions with high cultural product exports tend to be geographically close to each other. The spatial regression models further identify several key factors like economic development, productive capacities, cultural tourism, information development and human capital influence the spatial distribution of these exports. The findings of the study reveal that there is strong spatial relationship for exports of cultural products in BRI countries. The findings of this research contribute valuable insights for policymakers, businesses, and stakeholders regarding a deeper comprehension of the driving forces behind the spatial distribution of these cultural products, facilitating informed decision-making processes to optimize strategies for promoting and sustaining the trade of cultural products in an increasingly interconnected world.

Effects of cultural landscape service features in national forest parks on visitors’ sentiments: A nationwide social media-based analysis in China

National Forest Parks (NFPs) represent the highest level of the forest park system in China and have long been considered to cultivate positive experiences among visitors through their rich cultural services. However, there is limited knowledge of the relationship between the cultural dimension of landscape services (CLSs) in NFPs and visitors’ positive experiences. This study represents the first effort to investigate the effects of NFP CLSs on visitors’ expressed sentiments and aims to further identify the relative importance of various CLS features. A total of 267 NFPs across 31 provinces in China were selected as case studies, and approximately 300,000 visitors’ online comments from a widely used travel platform (CTrip) were obtained for sentiment analysis using natural language processing. A novel systematic framework for assessing CLSs was proposed, encompassing five principal features and 19 indicator subcategories. Ordinary least-squares and spatial regression models were used to reveal associations between NFP CLSs and expressed sentiments, whereas the XGBoost model and Shapley Additive exPlanations (SHAP) values were employed to identify the relative importance of NFP CLS features. Three major findings were observed: First, the feature scoring of CLSs in NFPs was unevenly distributed across the country, with scores along the eastern seaboard relatively higher than those inland in the northwest. Second, beneficial associations between the five principal CLS features and visitors’ expressed sentiments were suggested. Third, the influence of landscape aesthetics was particularly prominent in promoting visitors’ positive sentiments, followed by education, physical activities and health, recreation and tourism, and cultural heritage and spiritual practices. These findings provide valuable insights for health-oriented NFP management, policies, and planning.

Allocative Efficiency, Plant Dynamics and Regional Productivity: Evidence from Germany

The productivity gap between East and West Germany is a long ongoing discussion among the public and policy makers. Regional disparities still appear to be substantial. In this paper, we shed light on the role of allocative efficiency as a region’s driver of productivity disparities. We show that over 50 percent of the East-West productivity gap is associated with a less efficient labor allocation in former East Germany. Controlling for the heterogeneity among German federal states, we perform spatial regression on official firm-level data (AFiD), revealing that the regional differences in allocative efficiency are significantly associated with trade openness, competitive intensity, economies of scale and labor mobility.

Integrating traffic pollution dispersion into spatiotemporal NO2 prediction

Accurately predicting ambient NO2 concentrations has great public health importance, as traffic-related air pollution is of major concern in urban areas. In this study, we present a novel approach incorporating traffic contribution to NO2 prediction in a fine-scale spatiotemporal model. We used nationally available traffic estimate dataset in a scalable dispersion model, Research LINE source dispersion model (RLINE). RLINE estimates then served as an additional input for a validated spatiotemporal pollution modeling approach. Our analysis uses measurement data collected by the Multi-Ethnic Study of Atherosclerosis and Air Pollution in the greater Los Angeles area between 2006 and 2009. We predicted road-type-specific annual average daily traffic (AADT) on road segments via national-level spatial regression models with nearest-neighbor Gaussian processes (spNNGP); the spNNGP models were trained based on over half a million point-level traffic volume measurements nationwide. AADT estimates on all highways were combined with meteorological data in RLINE models. We evaluated two strategies to integrate RLINE estimates into spatiotemporal NO2 models: 1) incorporating RLINE estimates as a space-only covariate and, 2) as a spatiotemporal covariate. The results showed that integrating the RLINE estimates as a space-only covariate improved overall cross-validation R2 from 0.83 to 0.84, and root mean squared error (RMSE) from 3.58 to 3.48 ppb. Incorporating the estimates as a spatiotemporal covariate resulted in similar model improvement. The improvement of our spatiotemporal model was more profound in roadside monitors alongside highways, with R2 increasing from 0.56 to 0.66 and RMSE decreasing from 3.52 to 3.11 ppb. The observed improvement indicates that the RLINE estimates enhanced the model's predictive capabilities for roadside NO2 concentration gradients even after considering a comprehensive list of geographic covariates including the distance to roads. Our proposed modeling framework can be generalized to improve high-resolution prediction of NO2 exposure – especially near major roads in the U.S.

Are High-Immigrant Neighborhoods Disadvantaged in Seeking Local Government Services? Evidence from Baltimore City, Maryland

Abstract To modernize public service delivery, U.S. communities increasingly rely on 311 systems for residents to request government services. Research on 311 systems is relatively new, and there is mixed evidence on whether 311 can help bridge the gap between disadvantaged communities and governments. This study draws from research on immigration, race/ethnicity, and differential engagement to explore the link between immigrant concentration and 311 usage. We use longitudinal data on 311 requests in Baltimore City, Maryland (2014–2019) and spatial panel regression analysis to show that neighborhood racial/ethnic structure and the national policy environment can significantly influence whether immigrant concentration is a barrier for 311 service-seeking. Specifically, we find that immigrant concentration reduces 311 requests in high-immigrant neighborhoods with Latino or Black concentration, but not in high-immigrant neighborhoods with White/Asian concentration. We also find that in Latino high-immigrant neighborhoods, the relationship between immigrant concentration and 311 requests appears mainly after 2017, when the federal government adopted hostile immigration policies. By establishing and contextualizing the relationship between immigrant concentration and 311 usage, the study contributes to a deeper understanding of civic participation and the connection between immigrant communities and government.

Spatial analysis of food and water-borne diseases in Ahmedabad, India: Implications for urban public health planning

Spatial analysis of infectious diseases can play an important role in mapping the spread of diseases and can support policy making at local level. Moreover, identification of disease clusters based on local geography and landscape forms the basis for disease control and prevention. Therefore, this study aimed to examine the spatial-temporal variations, hotspot areas, and potential risk factors of infectious diseases (including Viral Hepatitis, Typhoid and Diarrhea) in Ahmedabad city of India. We used Moran's I and Local Indicators of Spatial Association (LISA) mapping to detect spatial clustering of diseases. Spatial and temporal regression analysis was used to identify the association between disease incidence and spatial risk factors. The Moran's I statistics identified presence of positive spatial autocorrelation within the considered diseases, with Moran's I from 0.09 for typhoid to 0.21 for diarrhea (p < 0.001). This indicates a clustering of affected wards for each disease, suggesting that cases were not randomly distributed across the city. LISA mapping demonstrated the clustering of hotspots in central regions of the city, especially towards the east of the river Sabarmati, highlighting key geographical areas with elevated disease risk. The spatial clusters of infectious diseases were consistently associated with slum population density and illiteracy. Furthermore, temporal analysis suggested illiteracy rates could increase risk of viral hepatitis by 13 % (95 % Confidence Interval (CI): 1.01–1.26) and of diarrhea by 18 % (95 % CI: 1.07–1.31). Significant inverse association was also seen between viral hepatitis incidence and the distance of wards from rivers. Conclusively, the study highlight the impact of socio-economic gradients, such as slum population density (indicative of poverty) and illiteracy, on the localized transmission of water and foodborne infections. The evident social stratification between impoverished and affluent households emerges as a notable contributing factor and a potential source of differences in the dynamics of infectious diseases in Ahmedabad.

Regional economic resilience, productivity growth and sectoral interconnectedness

This paper develops a theoretical model for the conceptualization and empirical analysis of regional economic resilience. Our production function framework is applied to empirically assess the effect of pre-crisis conditions on regional resilience accounting for exogenous changes in technology and changes in production factors. Spatial regression analysis indicates that intersectoral linkages are important factors for explaining the resilience of regions to negative exogenous shocks. A key finding of the study is the positive effect of the interconnectedness of agriculture, construction, and electrical, optical and transport equipment sectors on the ability of European Union regions to absorb the adverse effects of the 2008 economic crisis.