Spatial Autoregressive Model Research Articles

Ridge regularization for spatial autoregressive models with multicollinearity issues

Ridge regularization for spatial autoregressive models with multicollinearity issues

Inequality characteristics and influencing factors of CO2 emissions per capita in Jiangsu Province, China.

Analyzing the inequality characteristics and influencing factors of CO2 emissions per capita (CEPC) is conducive to balancing regional development and CO2 emissions reduction. This study applied the Gini coefficient and Theil index to investigate the CEPC inequalities during 2005-2017 at the county level in Jiangsu Province, China. Considering the spatial spillover and interaction effects, the factors influencing CEPC were analyzed by a hierarchical spatial autoregressive model. The results showed that the inequalities in CEPC first increased and then decreased at the inter-regional, and inter-county levels. The spatial pattern of CEPC was stable, and there was a significantly positive spatial autocorrelation of CEPC at the county level. The High-High type counties were mainly located in Sunan (southern Jiangsu). The spatial interaction effects of the CEPC between the prefecture and county levels indicated that governments at the prefecture level should integrate their county governments to reduce the CEPC. Moreover, carbon intensity, GDP per capita, land urbanization, and industrial structure play an important role in reducing CEPC. Our findings provide a scientific basis for formulating reasonable and effective carbon emission reduction policies.

Spillover effects of microcredit on the GDP of surrounding cantons in Ecuador

PurposeThe aim of this paper is to analyse the spillover effects of microcredit on the economy of Ecuador, with a particular focus on its potential as a poverty alleviation mechanism.Design/methodology/approachTo address our research questions, we take into account the distance between cantons (Ecuador’s own administrative distribution) by adopting a spatial autoregressive (SAR) model. To this end, a database will be constructed with macroeconomic information about the country, broken down by canton (administrative division of Ecuador), and in a 2019 cross section, with a total of 1,331 microcredit operations in all 221 of Ecuador’s cantons.FindingsWe find a positive effect of microcredit on these neighbouring regions in terms of wealth generation.Research limitations/implicationsWe acknowledge that this study is limited to Ecuadorian cantons. Nonetheless, it is crucial to emphasize that focussing on an under-represented developing country like Ecuador adds significant value to the research.Practical implicationsFacilitating access to microcredit is one of the main solutions to address the goals proposed in the sustainable development goals (SDGs).Social implicationsMicrocredit activity contributes to the creation of value and wealth in Ecuador, exerting a spillover effect in neighbouring areas that can generate positive multiplier effects and alleviate poverty. For all of the above reasons, our proposal for the country is to support and promote microcredit as one of the main solutions to address the goals proposed in the SDGs.Originality/valueThe novelty of this study lies in the use of spatial econometrics to observe the indirect effects of microcredit on the regions bordering the canton in which it was issued, thus examining the spatial effects of microcredit on wealth distribution.

Best linear and quadratic moments for spatial econometric models with an application to spatial interdependence patterns of employment growth in US counties

SummaryWe provide a novel analytic procedure to construct best linear and quadratic moments of the generalized method of moments estimation for a large class of cross‐sectional network and spatial econometric models. These moments generate an estimator that is asymptotically more efficient than the quasi‐maximum likelihood estimator when the disturbances follow a non‐normal and unknown distribution. We apply this procedure to a high‐order spatial autoregressive model with spatial errors, where the disturbances are heteroskedastic. Two normality tests of disturbances are developed. We apply the model to employment data in US counties, which demonstrates spatial interdependence patterns of regional employment growth.

Modeling Spatial Regimes With Smooth Transitions

In this paper, we introduce a new way of modeling spatial regimes using smooth transitions. We propose an autoregressive spatial lag model where a logistic function captures structural variation in the spatial lag parameter. In the regime-switching spatial lag model with smooth transitions, the effect of the spatial neighbors depends on the transition variable that governs the regime switch. An LM test for detecting nonlinearity is derived, and a simulation study, where the properties of the test are investigated, is conducted. The test shows good power in relatively small samples with moderate deviation from linearity. An empirical application is included, where data on median house prices from the Boston area is used to explore the spatial dependence between census tracts. A smooth shift between regimes governed by the variable property tax rate is found. The smooth spatial lag model performs slightly better than the regular spatial lag model in terms of model fit and capturing spatial dependence.

The Role of the Social Network in Adopting New Turfgrass Varieties: An Analysis of Twitter Data

This study examines the effect of social learning on new turfgrass variety adoption decisions using data from 231 turfgrass professionals’ Twitter accounts between 1 Jun 2018 and 31 Dec 2019. To determine the social learning effect, we decompose networking effects into social learning, individual-level and group-level similarities, herd behavior, and clustering effects. Our study estimates a spatial autoregressive probit model that directly incorporates the social network structure to account for unobservable networking effects and potential reflection problem. A Bayesian estimation procedure is used to alleviate the convergence problem caused by the complexity of model specification. Empirical results show that the social learning effect positively influences the new technology adoption and was greater than herd behavior effect. The results also suggest that turf professionals rely more on suggestions and information from online social networking among themselves than recommendations from advisors.

Connected Community Classification (C3): Development, Validation, and Geospatial Application for Population Health Promotion and Equity.

Social determinants of health (SDOH) impact population health. Leveraging community-level strengths related to SDOH through a social infrastructure perspective can optimize health behaviors and health outcomes to promote health equity. Our aims were to develop, validate, and apply the Connected Community Classification (C3) as comprehensive community-level measure of protective SDOH and structural factors in the Four Corners states region of the United States. C3 was developed using an iterative principal component analysis of publicly available data mapped to 5 SDOH domains. Regional clustering of C3 by zip code tabulation area (ZCTA) was identified using spatial autocorrelation methods. In adjusted spatial autoregressive models, we analyzed the association of C3 with high-risk health behaviors and chronic disease prevalence using publicly available data for population-level estimates of fruit and vegetable intake, physical activity, obesity, smoking, alcohol use, coronary heart disease (CHD), diabetes, and cancer. C3 was found to be reliable and valid; a C3 value of 10 indicates communities with greater connection (high), while a value of 1 indicates communities with greater separation (low) to social infrastructure. Lower connection, as measured by C3, was significantly inversely associated with lower fruit and vegetable intake, lower physical activity, and higher rates of obesity, smoking, CHD, diabetes, and cancer. C3 was significantly positively associated with heavy alcohol use. These findings demonstrate that communities connected to social infrastructure have better population health outcomes. C3 captures protective community attributes and can be used in future applications to assist health researchers, practitioners, nonprofits, and policymakers to advance social connection and health equity in geographically diverse underserved regions.

Drivers of Tree Canopy Loss in a Mid-Sized Growing City: Case Study in Portland, OR (USA)

The benefits of the urban tree and tree canopy (UTC) are increasingly crucial in addressing urban sustainability. Yet, increasingly evident from earlier research is the distributional inequities of UTC and active efforts to expand tree plantings. Less is known about the dynamics of UTC loss over time and location. This study aims to understand the dynamics of UTC change, especially canopy loss, and to investigate the drivers of the loss. This study draws on a high–resolution dataset of an urban canopy in Portland, Oregon, USA, assessing changes in UTC from 2014 to 2020. By integrating demographic, biophysical, and policy data with UTC information, we use a spatial autoregressive model to identify the drivers of UTC loss. The results reveal an unexpected spatial distribution of UTC change: less gain in the neighborhoods with the least UTC, and greater loss in the neighborhoods with moderate UTC. This study identifies four primary drivers of UTC loss: socioeconomic characteristics, urban form, activities on trees, and residential status. Factors such as population density, race, and income have an impact on canopy loss, as well as the building footprint and the number of multifamily housing units; residential statuses, such as the proportion of owner-occupied housing and residential stability, impact canopy loss.

Robust variable selection with exponential squared loss for the partially linear varying coefficient spatial autoregressive model

Robust variable selection with exponential squared loss for the partially linear varying coefficient spatial autoregressive model

Local Walsh-average-based Estimation and Variable Selection for Spatial Single-index Autoregressive Models

Local Walsh-average-based Estimation and Variable Selection for Spatial Single-index Autoregressive Models

Model Checking in Partially Linear Spatial Autoregressive Models

This article proposes two new classes of nonparametric tests for the correct specification of linear spatial autoregressive models based on the “integrated conditional moment” approach. Our test statistics are constructed as continuous functionals of a residual marked empirical process as well as its projected version. We derive asymptotic properties of the test statistics under the null hypothesis, the alternative hypothesis, and a sequence of local alternatives. The proposed tests do not involve the selection of tuning parameters such as bandwidths and are able to detect a broad class of local alternatives converging to the null at the parametric rate n − 1 / 2 , with n being the sample size. We also propose a multiplier bootstrap procedure that is computationally simple to approximate the critical values. Monte Carlo simulations illustrate that our tests have a reasonable size and satisfactory power for different types of data-generating processes. Finally, an empirical analysis of growth convergence is carried out to demonstrate the usefulness of the tests.

Spatial diffusion of public-private partnership (PPP) in China: A county-level analysis

This paper explores the spatial diffusion of Public-Private Partnership (PPP) in China, assessing the effects of regional boundaries, including political and functional boundaries, and transaction costs on this process. We conduct an empirical analysis using county-level panel data from 2014 to 2020 and employ a combination of geographic information system (GIS) and spatial autoregressive models to analyze the spatial factors influencing PPP diffusion in China. Our findings reveal three key points: First, local governments' PPP investment positively correlates with neighboring regions, indicating a contagious effect. Second, the contagious effect is more significant within the same urban agglomeration, demonstrating a boundary effect and illustrating the impact of spatial rescaling on spatial diffusion. Third, the contagious effect is more pronounced for PPP sectors with lower output measurement difficulty, suggesting transaction costs influence spatial diffusion. These findings offer valuable insights into barriers to spatial diffusion and strategies to promote the spatial diffusion of innovation.

The distribution and conservation of areas with microendemic species in a biodiversity hotspot: a multi-taxa approach.

Microendemic species are species with very small geographic distributions (ranges). Their presence delimitates areas with microendemic species (AMs), denoting a spatial unit comprising at least one population of at least one microendemic species. AMs are assumed to be distributed distinctively and associated with specific ecological, historical, and anthropogenic attributes. However, the level of influence of these factors remains unclear. Thus, we studied the distribution patterns of microendemic species within the Brazilian Atlantic Forest to (a) identify the region's AMs; (b) evaluate whether ecological (latitude, altitude, distance from the coastline), historical (climate stability), and anthropogenic (ecological integrity) attributes distinguish AMs from non-AMs; and (c) assess the conservation status of the Atlantic Forest's AMs. We mapped the ranges of 1,362 microendemic species of angiosperms, freshwater fishes, and terrestrial vertebrates (snakes, passerine birds, and small mammals) to identify the region's AMs. Further, spatial autoregressive logit regression models were used to evaluate whether latitude, altitude, distance from the coastline, Climate Stability Index, and ecological integrity can be used to discern AMs from non-AMs. Moreover, the AMs' conservation status was assessed by evaluating the region's ecological integrity and conservation efforts (measured as the proportion of AMs in protected areas). We identified 261 AMs for angiosperm, 205 AMs for freshwater fishes, and 102 AMs for terrestrial vertebrates in the Brazilian Atlantic Forest, totaling 474 AMs covering 23.8% of the region. The Brazilian Atlantic Forest is a large and complex biogeographic mosaic where AMs represent islands or archipelagoes surrounded by transition areas with no microendemic species. All local attributes help to distinguish AMs from non-AMs, but their impacts vary across taxonomic groups. Around 69% of AMs have low ecological integrity and poor conservation efforts, indicating that most microendemic species are under threat. This study provides insights into the biogeography of one of the most important global biodiversity hotspots, creating a foundation for comparative studies using other tropical forest regions.

Variable selection via penalized quasi-maximum likelihood method for spatial autoregressive model with missing response

Spatial autoregressive model is widely concerned in the economic field, whereas when the data is missing, variable selection and parameter estimation of the model is quite challenging. Based on this, we discuss the variable selection in spatial autoregressive model with missing data. Under the condition that errors are independent and identically distributed, we have developed a penalized quasi-maximum likelihood method to achieve variable selection and parameter estimation simultaneously in the presence of missing responses. The method’s theoretical properties, including consistency and asymptotical normality, are established under certain assumptions. Meanwhile, an improved expectation–maximization algorithm is provided for optimizing the penalized quasi-maximum likelihood function. Simulations are conducted to examine the proposed method and assess the finite-sample performance. Additionally, we present a practical example to illustrate the method’s application.

Cross-Border Migration and Its Influence on Mental Health Dynamics in Thailand: An Empirical Analysis of Provincial Data

While previous research has examined the impact of cross-border migrants on economic development and other social issues, limited studies have focused on the potential effects of migrants on the mental health of the native population. This paper investigates the impact of cross-border migrants on the mental health of natives in Thailand, considering various mental health disorders. This study analyzes 77 provincial data sets from 2016 to 2021 using a spatial panel autoregressive model. The findings suggest that the presence of migrants did not have an adverse influence on the mental health of the native population at a 5% significance level. Notably, there is a reduction in the prevalence of substance abuse disorders associated with the presence of migrants at the 5% significance level. Additionally, unemployment, education, public health services, technological infrastructure, health risk behavior, and age are significant contributors to the mental health of natives at a 5% significance level. The study underscores the need for a comprehensive approach to developing effective mental health policies and initiatives that address these multiple contributors.

Análise espacial de acidente vascular cerebral isquêmico na Espanha: os papéis da acessibilidade à saúde e do desenvolvimento econômico

Abstract: Ischemic stroke is a major cause of mortality worldwide; however, few studies have been conducted to measure the impact of the distribution of healthcare services on ischemic stroke fatality. This study aimed to explore the relationship between three ischemic stroke outcomes (incidence, mortality, and fatality) and accessibility to hospitals in Spain, considering its economic development. A cross-sectional ecological study was performed using data on hospital admissions and mortality due to ischemic stroke during 2016-2018. Gross geographic product (GGP) per capita was estimated and a healthcare accessibility index was created. A Besag-York-Mollié autoregressive spatial model was used to estimate the magnitude of association between ischemic stroke outcomes and economic development and healthcare accessibility. GGP per capita showed a geographical gradient from southwest to northeast in Spain. Mortality and case-fatality rates due to ischemic stroke were higher in the south of the country in both women and men aged 60+ years. In women and men aged 20-59 years a EUR 1,000 increase in GGP per capita was associated with decreases in mortality of 5% and 4%, respectively. Fatality decreased 3-4% with each EUR 1,000 increase of GGP per capita in both sexes and in the 20-59 and 60+ age groups. Decreased healthcare accessibility was associated with higher fatality in the population aged 60+. Economic development in southwest Spain would not only improve employment opportunities but also reduce ischemic stroke mortality. New health related strategies to improve hospital accessibility should be considered in more sparsely populated regions or those with worse transport and/or healthcare infrastructure.

Spatial Autoregressive Models with Generalized Spatial Disturbances

Spatial Autoregressive Models with Generalized Spatial Disturbances

Analysis of the impact of e-commerce on rural quality development and its spatial spillover effects

Abstract This study addresses the challenge of variable selection in partially linear additive spatial autoregressive models through a novel approach that leverages orthogonal projection based on the QR decomposition technique. This method ensures the validity and compatibility of examining the influence of e-commerce on rural high-quality development while minimizing bias in spatial spillover benefits analysis. By constructing a comprehensive evaluation index system, the research applies a partially linear weighted spatial autoregressive model with orthogonal projection to assess e-commerce’s impact on rural development. Analysis of baseline regression for selected villages, supplemented by heterogeneity tests, threshold effects, and spatial analyses, reveals a significant relationship: a one-unit increase in e-commerce activity corresponds to a 0.0038 unit increase in high-quality rural development. This growth comprises both a direct effect of 0.0011 units and a spillover effect of 0.0027 units, highlighting e-commerce’s pivotal role in promoting rural prosperity and its beneficial spillover to adjacent areas.

Freight Transport and Land Use Interaction: an Analysis Approach Based on Floating Car Data

Freight transport in urban areas affects city liveability and, in turn, it is affected by the territorial characteristics and by the location of firms, factories and production activities. Therefore, the relationship between freight flow and land use must be investigated, combining knowledge derived from the analysis of data on freight transport and territorial attributes. This work considers a dataset of floating car data (i.e., a sample of freight vehicles moving in the study area) and a dataset of territorial data (e.g., number of activities and employees) in order to calibrate a model able to foresee the production of tours from each zone of the study area. In particular, a linear and a spatial autoregressive model are calibrated considering aggregated data related to the commercial activities in the study area.

Composite quantile regression based robust empirical likelihood for partially linear spatial autoregressive models

Composite quantile regression based robust empirical likelihood for partially linear spatial autoregressive models