Spatial Autoregressive Model Research Articles

Policy Imitation Based on Similarities: The Diffusion of Public–Private Partnerships

ABSTRACTPolicy imitation likely occurs among similar jurisdictions. The extant literature primarily focuses on geographic proximity or ideological congruence, while overlooking multifaceted similarities in the diffusion of public policies. This study explores the adoption and spread of public–private partnerships (PPP) across China's cities, specifically examining the extent to which three distinctive similarity measures influence PPP policy imitation. In addition to geographic proximity and economic resemblance, equivalent positioning in China's administrative hierarchy captures a managerial aspect of similarity to understand policy imitation among resembling cities. Three corresponding spatial autoregressive panel models are used, respectively. The findings suggest that cities in China imitate their geographic, economic, and administrative peers while directing PPP investments. Relative closeness in the administrative hierarchy of cities exerts a greater influence on policy imitation than economic resemblance and spatial proximity.

Tests for spatial dependence and temporal heterogeneity in time-varying coefficient spatial autoregressive panel data model

Time-varying coefficient spatial autoregressive panel data model is a powerful tool to simultaneously deal with spatial dependence and temporal heterogeneity in spatial panel data analysis. However, little work has been devoted to related statistical inference issues, which largely limits the application scope of the model. In this paper, we develop two generalized-likelihood-ratio-statistic-based bootstrap tests to detect spatial dependence of the response variable and temporal heterogeneity of the regression relationship, respectively. The simulation studies show that both tests are of accurate size and satisfactory power, and are quite robust to non-normality of error distribution. A house price data set is finally analyzed to demonstrate the application of the proposed tests in detecting spatial dependence and temporal heterogeneity.

SPATIAL PANEL MODELING OF PROVINCIAL INFLATION IN INDONESIA TO MITIGATE ECONOMIC IMPACTS OF HEALTH CRISES

Probabilistic statistical modeling simplifies complex issues, including economic and health challenges, by applying inductive statistics. Spatial panel modeling, using Queen Contiguity weighting, has proven to be essential for analyzing inflation expenditure patterns during health crises, such as COVID-19 in Indonesia. This study highlights the impact of inflation on national economic stability and explores the inter-provincial relationships that influence inflation dynamics across expenditure groups. The purpose of this study is to develop a spatial panel model to address this gap, offering insights for policy and recovery strategies. The results reveal significant spatial interdependence in provincial inflation data, underscoring the role of spatial factors in economic analysis. Two models are identified: Spatial Autoregressive Model with Random Effects (SAR-RE) before the crisis and Spatial Error Model with Random Effects (SEM-RE) during the crisis. Transportation facilities consistently affect inflation, demonstrating the effectiveness of spatial panel modeling in guiding policies for economic stability and recovery.

Оценка значимых факторов экономического развития муниципальных образований региона: пространственный анализ (на примере Ростовской области и Краснодарского края)

The study aims to identify the significant factors and their direct and external effects (externalities) that influence the economic productivity of municipal entities in the Rostov Region and Krasnodar Krai. The principal tasks focus on characterizing the economic space and delineating territorial clusters of accelerated development alongside zones of economic decline in the specified regions. A spatial econometric approach has been applied. For the purposes of this analysis, authors designed SAR (spatial autoregressive model) and SEM (spatial error model) models based on municipal statistics data for 2023. Model calculations revealed that the common significant factors influencing economic productivity in both Rostov Region and Krasnodar Krai include the volume of investments in fixed capital, the average monthly wage of employees in organizations, and the density of paved local public roads. Moreover, for Rostov region, an additional growth factor at the municipal level of municipalities is the average number of employees of organizations. Using the tools of spatial econometrics, the presence of a negative spatial autocorrelation of economic dynamics between neighboring municipalities of the region has been proved. Furthermore, for both subjects, based on calculations of local spatial aggregation indicators (LISA), territorial clusters of advanced development and zones of economic recession were identified. The results obtained can serve as a basis for developing a differentiated approach to determining the directions and tools for managing socio-economic development at the subregional level.

Study on the Impact of Green Finance and Environmental Regulations on Haze Pollution in China

With the rapid economic growth in China, the years from 2012 to 2017 witnessed a severe outbreak of haze pollution. In response, numerous enterprises were shut down or forced to reduce production, leading to significant negative impacts on the economy and society. Consequently, the scientific management of haze pollution has become a focal point of China’s environmental research. This paper explores the influence of green finance and environmental regulations on haze pollution, employing spatial autoregressive and spatial Durbin models to conduct an empirical analysis based on panel data from 30 Chinese provinces spanning 2000 to 2017. The following conclusions were thus drawn: from 2010 to 2017, the green finance index of the Beijing–Tianjin–Hebei region and Liaoning Province, which suffered severe haze pollution, showed an overall downward trend; efforts to strengthen environmental regulations played an important role in reducing air pollution, especially the PM2.5 level; the improvement in the green finance level can help reduce haze pollution; and green finance has an obvious spatial spillover effect. A number of policy recommendations were proposed on this basis, including strengthening inter-regional cooperation and exchanges in green finance, promoting the scientific application of environmental regulations in areas with serious haze pollution, strengthening the implementation of green finance strategies in areas with serious haze pollution, and establishing regional green finance demonstration zones. The purpose of this study is to provide meaningful support for China’s scientific management of haze pollution through an analysis of the impact of green finance and policy optimization. The innovative aspects of this research lie in its pioneering use of spatial econometric models to investigate the role of green finance and environmental regulations in controlling haze pollution. The findings reveal that advancements in green finance can significantly alleviate haze pollution.

A Marginal Maximum Likelihood Approach for Hierarchical Simultaneous Autoregressive Models with Missing Data

Efficient estimation methods for simultaneous autoregressive (SAR) models with missing data in the response variable have been well explored in the literature. A common practice is introducing measurement error into SAR models to separate the noise component from the spatial process. However, prior studies have not considered incorporating measurement error into SAR models with missing data. Maximum likelihood estimation for such models, especially with large datasets, poses significant computational challenges. This paper proposes an efficient likelihood-based estimation method, the marginal maximum likelihood (ML), for estimating SAR models on large datasets with measurement errors and a high percentage of missing data in the response variable. The spatial autoregressive model (SAM) and the spatial error model (SEM), two popular SAR model types, are considered. The missing data mechanism is assumed to follow a missing-at-random (MAR) pattern. We propose a fast method for marginal ML estimation with a computational complexity of O(n3/2), where n is the total number of observations. This complexity applies when the spatial weight matrix is constructed based on a local neighbourhood structure. The effectiveness of the proposed methods is demonstrated through simulations and real-world data applications.

Fixed effects spatial panel interval-valued autoregressive models and applications

Interval-valued data has garnered attention across various applications, leading to increased research into spatial interval-valued data models. The integration of uncertainty variables into spatial panel data models has become crucial. This paper presents a spatial panel interval-valued autoregressive model with fixed effects, utilizing the parametric method. The quasi-maximum likelihood method is employed for parameter estimation, and its consistency and asymptotic properties are discussed. Additionally, three special cases and two degenerated models derived from our framework are presented, elucidating their significance in spatial statistics. Monte Carlo simulations are used to validate the fitting and forecasting performance of our proposed models across diverse scenarios. Furthermore, the models are implemented in real-world air quality and house price datasets for forecasting purposes. Through rigorous experimentation, the superior performance of the models is demonstrated. These results highlight the practical utility of the spatial panel interval-valued autoregressive models in addressing spatial data challenges.

Nigeria's malaria prevalence in 2015: a geospatial, exploratory district-level approach.

This study used data from the second Nigeria Malaria Indicator Survey (NMIS) conducted in 2015 to investigate the spatial distribution of malaria prevalence in the country and identify its associated factors. Nigeria is divided into 36 states with 109 senatorial districts, most of which are affected by malaria, a major cause of morbidity and mortality in children under five years of age. We carried out an ecological study with analysis at the senatorial district level. A malaria prevalence map was produced combining geographic information systems data from the Nigeria Malaria Indicator Survey (NMIS) of 2015 with shape files from an open data-sharing platform. Spatial autoregressive models were fitted using a set of key covariates. Malaria prevalence in children under-five was highest in Kebbi South senatorial district (70.6%). It was found that poorest wealth index (β = 0.10 (95% CI: 0.01, 0.20), p = 0.04), mothers having only secondary level of education (β = 0.78 (95% CI: 0.05, 1.51), p = 0.04) and households without mosquito bed nets (β = 0.21 (95% CI: 0.02, 0.39), p = 0.03) were all significantly associated with higher malaria prevalence. Moran's I (54.81, p<0.001) showed spatial dependence of malaria prevalence across contiguous districts and spatial autoregressive modelling demonstrated significant spill-over effect of malaria prevalence. Maps produced in this study provide a useful graphical representation of the spatial distribution of malaria prevalence based on NMIS-2015 data. Clustering of malaria prevalence in certain areas further highlights the need for sustained malaria elimination interventions across affected regions in order to break the chain of transmission.

Urban–rural disparities in hospital admissions for depression in Austria: A spatial panel data analysis

Medical practice variation in mental healthcare is a useful indicator for policymakers aiming to improve the efficiency of healthcare delivery. Previous studies have shown strong regional variation in healthcare utilisation in Austria, which seems to be a by-product of regionalised institutional rules and healthcare service mix rather than epidemiology. We use a set of routine municipality-level healthcare data on hospital admissions for depressive episodes of adult Austrian patients from 2009 to 2014 to examine spatial patterns in healthcare utilisation in mental health. Our data contains 93,302 hospital episodes by 65,908 adult patients across 2114 municipalities. We estimate a random-effects spatial autoregressive combined model to regress log hospital admission rates on hospital supply and urbanicity as proxies for municipality healthcare service mix alongside demographic and socioeconomic controls. We find that admissions for depression are substantially higher in suburban municipalities compared to rural areas and in municipalities with hospitals compared to those without. The spatial structure suggests positive spatial spillovers between neighbouring municipalities. Our main results are stable across virtually all model specifications used for robustness and show that healthcare service mix and supply of hospital services strongly correlate with spatial patterns of hospital admission rates in the population. Promoting timely access to high-quality primary care and early-stage treatments may reduce the burden of avoidable depression-related hospitalisations for patients and public budgets, and close a gap of unmet need for care of vulnerable populations.

The impact of improving element quality on the comprehensive production capacity of grain: evidence from high-standard farmland construction

PurposeChina’s overall grain production efficiency is currently relatively low, and how to improve it is key to high-quality grain development. High-standard farmland construction fills the gaps in grain production, providing a path for improving comprehensive grain production capacity.Design/methodology/approachThis paper analyzes grain production data from 281 Chinese cities and uses the meta-Malmquist index to calculate total factor productivity (TFP) of grain and the spatial autoregressive model with spatial error (SARAR) model to analyze the impact of high-standard farmland construction on TFP of grain. Finally, it tests the model for robustness and heterogeneity.FindingsTFP of grain has room for improvement, and technological efficiency has a greater impact on it than technological progress. TFP of grain in the eastern region is significantly higher than that in central and western regions. High-standard farmland construction can significantly improve TFP and technology efficiency of grain, and its economic benefits are directly affected by the differences in regional grain economic development. High-standard farmland construction improves TFP of grain through agricultural mechanization, and its impact is heterogeneous; TFP of grain in the (poor, plain) eastern region is more likely to share its benefits.Originality/valueThis paper evaluates whether high-standard farmland construction has achieved its purpose of enhancing comprehensive grain production capacity. It offers novel insights into elevating element quality and comprehensive grain production capacity, provides a theoretical reference from the perspective of element quality and draws corresponding policy implications for high-standard farmland.

Geopolitical Risks and Stock Market Volatility in the SAARC Region

Abstract This study examined the stock market volatility of the member states of the South Asian Association for Regional Cooperation (SAARC) and geopolitical risk (GPR). The analysis period covered January 2014 to March 2024, and the time–frequency wavelet method was used to process the data. The time-varying parameter vector autoregression and spatial autoregressive models helped determine the dynamic connectedness of volatility in the analyzed states. The findings revealed similar stock market connections in Bangladesh and India. In addition, a comparative analysis of stocks in India and Pakistan led to the identification of common elements. The connection between geopolitical concerns and Sri Lankan stocks was the strongest and increased in intensity after 2019. GPR and Nepal’s stock market maintained a continuous but low-intensity relationship. The dynamic connectedness between member states’ stock markets was limited during the review period. The study results could encourage SAARC governments to bridge their political differences to ensure that South Asia becomes a strong partner in the global economy. Equally, our results can benefit investors, financial institutions, regulatory authorities, and governments.

Potentials for Reducing Spatial Inequalities in Innovation: A Spatial Econometric Perspective

ABSTRACTThis study explains disparities in regional innovation through specific features of regional knowledge bases, including recently discussed concepts such as, technological relatedness, knowledge complexity, and technological complementarity among neighboring regions and regions connected in R&amp;D networks. We employ a spatial autoregressive panel model for 405 European regions to estimate the effects of these characteristics. While being connected to complementary regions and having a high region‐internal technological relatedness are conducive to regional innovation, knowledge complexity has no positive effect interestingly. In illustrative convergence scenarios, we demonstrate the potential of increasing relatedness and complementarity to reduce inequalities in Europe, pointing to important policy implications.

Spatial Patterns of Emergency Dental Care Utilization: Assessing the Influence of Social Vulnerability and Dental Provider Availability.

Our aim was to examine the association between emergency dental care use and dental providers' availability and ZIP code social vulnerability index. This cross-sectional observational study mapped variations in emergency dental care and analyzed their association with social vulnerability using generalized spatial two-stage least-squares to address spatial correlation. To perform spatial autoregressive modeling to examine how dental provider capacity and social vulnerability influence emergency care use across neighboring counties, accounting for ZIP Code Tabulation Area (ZCTA) spillover effects, we used secondary data from the Maryland Health Services Cost Review Commission, InfoUSA, and the US Census American Community Survey. We focused on emergency dental care usage in ZIP Code Tabulation Areas, using emergency department visits and inpatient stays per 1000 residents for dental conditions, as the dependent variable. Highly vulnerable ZCTAs saw 5.9× more non-traumatic dental condition (NTDC) ED visits, 5.3× more chronic dental ED visits, 3.3× more NTDC inpatient stays, and 1.3× more chronic dental inpatient stays than less vulnerable areas. For all four measures of emergency dental care use, increased dental provider availability was associated with reductions in use of emergency dental care (ED), while higher social vulnerability was associated with increased use. For example, an increase of one dental provider per 1000 residents was associated with a reduction of 28.2 non-traumatic dental ED visits, while a high social vulnerability index was associated with an increase of 88.6 non-traumatic dental ED visits. There was an association among dental provider availability, social vulnerability, and the use of emergency dental care.

Interaction of environmental quality with agricultural productivity and other determinants in Java-Bali, Indonesia

Abstract This study aims to analyze the direct and spillover effects of agricultural productivity and other determinant variables on environmental quality in Java-Bali Island from 2015 to 2022. The data is in the form of a panel, sourced from the Central Bureau of Statistics and the Indonesian Ministry of Environment. The method used to answer the study objectives is through a spatial autoregressive model (SAR) with a weighting matrix of three nearest neighbors. The results show that the agriculture, forestry and fisheries sectors and the industrial sector are the biggest contributors to environmental damage in Java-Bali. Meanwhile, increasing real GRDP per capita and the burden of dependent workers can actually improve environmental quality. Meanwhile, democracy has not been able to play an important role in environmental issues. The implication is that a strong commitment is needed to change production habits, especially the agricultural and industrial sectors that are more environmentally friendly with innovation and utilization of technology. The concentration of democratic activities in responding to environmental issues also needs to be strengthened.

Changes in the European Union households' consumption structure and the sustainable development

The subject of this study is the relationship between the households' consumption structure changes and sustainable development. This research was conducted for the European Union (EU) countries in the years 2008–2020. Additionally, in the investigation, the considered area is divided into two separate regimes: countries that joined the EU before 2000 (referred to as Old EU Member States) and countries that joined the EU later (called New EU Member States). It is widely known that the territorial units in the first regime are relatively wealthier than those in the second group. As a result of differences in wealth between the defined regimes, consumer behaviours can vary. Sustainable consumption (SC) is one of the major aspects of sustainable development. SC indicates consumer choices that are both intentional and effective, and these choices demonstrate a positive impact on environmental and/or social conditions. In this research, the influence of households' consumption changes on the environmental pillar of sustainable development is evaluated. Caring for the natural environment is one of the main challenges today. In this research, another approach to measure environmental development is proposed. So far, different types of footprints have been used. This analysis considers most of the environmental indicators defined in the Sustainable Development Goals (SDG's). The environmental development of EU countries is evaluated based on the synthetic measure of development proposed by Hellwig. Moreover, the spatio-temporal sensitivity models are estimated and verified. The sensitivity parameter of these models shows the impact of consumption structure changes on environmental development, wherein the consumption structure is categorized into four groups of goods: services, non-durable, semi-durable, and durable. The spatial autoregressive (SAR) models are also estimated and verified. The investigation results point out the differences in the influence of consumption structure changes on sustainable development depending on the considered group of countries.

Remotely Sensed Variables Predict Grassland Diversity Better at Scales Below 1,000 km as Opposed to Abiotic Variables That Predict It Better at Larger Scales

AbstractGlobal spatial patterns of vascular plant diversity have been mapped at coarse grain based on climate‐dominated environment–diversity relationships and, where possible, at finer grain using remote sensing. However, for grasslands with their small plant sizes, the limited availability of vegetation plot data has caused large uncertainties in fine‐grained mapping of species diversity. Here we used vegetation survey data from 1,609 field sites (&gt;4,000 plots of 1 m2), remotely sensed data (ecosystem productivity and phenology, habitat heterogeneity, functional traits and spectral diversity), and abiotic data (water‐ and energy‐related, characterizing climate‐dominated environment) together with machine learning and spatial autoregressive models to predict and map grassland species richness per 100 m2 across the Mongolian Plateau at 500 m resolution. Combining all variables yielded a predictive accuracy of 69% compared with 64% using remotely sensed variables or 65% using abiotic variables alone. Among remotely sensed variables, functional traits showed the highest predictive power (55%) in species richness estimation, followed by productivity and phenology (48%), spectral diversity (48%) and habitat heterogeneity (48%). When considering spatial autocorrelation, remotely sensed variables explained 52% and abiotic variables explained 41%. Moreover, Remotely sensed variables provided better prediction at smaller grain size (&lt;∼1,000 km), while water‐ and energy‐dominated macro‐environment variables were the most important drivers and dominated the effects of remotely sensed variables on diversity patterns at macro‐scale (&gt;∼1,000 km). These findings indicate that while remotely sensed vegetation characteristics and climate‐dominated macro‐environment provide similar predictions for mapping grassland plant species richness, they offer complementary explanations across broad spatial scales.

Variable selection of partially functional linear spatial autoregressive model with a diverging number of parameters

In this article, we consider variable selection of partially functional linear spatial autoregressive model with a diverging number of parameters, in which the explanatory variables include infinite dimensional predictor procedure, treated as functional data, and multiple real valued scalar variate. By combining series approximation method, two-stage least squares method and a class of non convex penalty function, we propose a variable selection method to simultaneously select significant explanatory variables in the parametric component and estimate the corresponding parameter related to spatial lag of the response variable. Under appropriate conditions, we derive the rate of convergence of the series estimator of the functional and parametric component, and show that the proposed variable selection method processes the oracle property. That is, it can estimate the zero components as exact zero with high probability, and estimate the non zero components as efficiently as if the true model was known beforehand. Simulation result show that our proposed variable selection method has better finite sample property.Notably, in the case where the correlation among the explanatory variables in the parametric component is low, the proposed variable selection method performs well. An application of the proposed variable selection method serves as a practical illustration.

How Does Tax Avoidance Affect a Firm's Investment Inefficiency in Taiwan's Semiconductor Industry?

ABSTRACTThis study examines an important issue regarding how tax avoidance affects a firm's investment inefficiency in Taiwan's semiconductor industry. In addition, the dynamic effect of a firm's investment inefficiency and spillover effect of investment inefficiency from competitors are also explored. This study adopts 44 listed firms from 2013 to 2021 and constructs the “competitor‐dependence matrix” to replace the “spatial‐dependence matrix” in the dynamic spatial autoregressive model. According to the difference generalized method of moments estimation, the primary finding of this study is that the tax avoidance has a statistically and significantly positive effect on a firm's investment inefficiency while it is overinvestment, but a negative effect while it is underinvestment. Moreover, the dynamic effect is statistically insignificant, but the spillover effect is statistically and significantly negative.

Variable selection of nonparametric spatial autoregressive models via deep learning

With the development of deep learning techniques, the application of neural networks to statistical inference has dramatically increased in popularity. In this paper, we extend the deep neural network-based variable selection method to nonparametric spatial autoregressive models. Our approach incorporates feature selection and parameter learning by introducing Lasso penalties in a residual network structure with spatial effects. We transform the problem into a constrained optimization task, where optimizing an objective function with constraints. Without specifying sparsity, we are also able to obtain a specific set of selected variables. The performance of the method with finite samples is demonstrated through an extensive Monte Carlo simulation study. Finally, we apply the method to California housing price data, further validating its superiority in terms of variable selection and predictive performance.

Beyond Parametric Bounds: Exploring Regional Unemployment Patterns Using Semiparametric Spatial Autoregression

Abstract Background It is a well-known phenomenon that nonlinearities that are inherent in the relationship among economic variables negatively affect the commonly used estimators in the econometric models. The nonlinearities cause an instability of the estimated parameters that, in particular, are unable to capture a local relationship between the response and the covariate. Objectives The main aim of the paper is the simultaneous consideration of spatial effects as well as nonlinearities through an advanced semiparametric spatial autoregressive econometric model. The paper seeks to contribute to empirical studies of regional science focused on the application of semiparametric spatial autoregressive econometric models. Methods/Approach We outline an approach that can be used to correct nonlinearities by incorporating a semiparametric idea within the framework of econometric models. We use an expansion by penalised basis splines that are highly flexible and are able to capture local nonlinearities between variables. Results In the empirical study, we fit different econometric models that attempt to explain the dynamics of the European Union's regional unemployment. Conclusions The results show that regional unemployment exhibits significant spatial dependence, indicating interconnectedness among neighbouring regions and suggesting the adoption of a semiparametric spatial autoregressive model for improved modelling flexibility, surpassing traditional parametric approaches.