Population health is one of the most important indicators of human development, and the mortality rate, as a key measure, is influenced by health-related, medical, and social factors. The aim of this study is to examine the spatial effects of government health expenditures on health outcomes (mortality rate) across the provinces of Iran during the period 2011–2021. To this end, provincial-level data and a spatial weight matrix are employed to identify spatial dependence among provinces. First, the existence of spatial autocorrelation in mortality rates is confirmed using appropriate diagnostic tests. Then, the Spatial Durbin Model is applied to analyze the impact of government health expenditures on health outcomes. The estimation results indicate that mortality rates exhibit significant spatial dependence, such that changes in mortality rates in one province can affect neighboring provinces. Moreover, government health expenditures, the number of physicians, and the number of hospital beds have a negative and statistically significant effect on mortality rates, highlighting the important role of expanding health infrastructure and services in reducing mortality. In addition, the level of education shows a significant negative effect on mortality rates, confirming the importance of socio-economic factors in improving health outcomes. Based on these findings, the design and implementation of health policies require a regional and spatial approach, so that through efficient allocation of resources and consideration of spatial spillover effects, a sustainable reduction in mortality rates can be achieved at the national level.

Introduction

Human capital has long been interpreted narrowly in economic thought, but it has become increasingly clear that labor resources cannot be understood primarily in quantitative terms. This has led to a completely new approach to the field of economic activity. According to this broader approach, the qualitative dimensions of human capital have gained increasing importance. Healthcare is one of the most vital sectors in any national economy. Improving health status significantly increases human capital through greater productivity and longer working lives through reduced illness, leading to improved production and consumption, as well as returns on investment. Health studies have shown that health spending and investment in health systems are determinants of population health status and mortality rates. Health spending usually includes financial resources spent on medical services, preventable diseases, primary and specialized care, vaccinations, and health insurance. The aim of this study is to investigate the spatial effects of government health spending on health outcomes (mortality rates) in the country's provinces during the period 2011–2021. For this purpose, provincial data and a spatial weight matrix were used to identify spatial dependencies between provinces.

Methods and Material

This study considers the relationship between health expenditures and health development based on Grossman's (1972) health production function and the "social determinants of health" by the World Health Organization, a macro health production function model similar to Zhao et al. (2024) as equation (1):
 
(1) 
The spatial units in this study comprise 31 provinces of the country during the period 2011–2021. To investigate and measure potential spatial effects, the Spatial Lag Model (SLM), the Spatial Error Model (SEM), and the Spatial Durbin Model (SDM) are employed. The SDM is derived from the integration of the SLM and SEM. The SLM captures spatial dependence by incorporating a spatially lagged dependent variable into the regression framework. The SEM, on the other hand, assumes that spatial correlation among variables is reflected in the regression error term rather than through the inclusion of a lagged dependent variable. In contrast, the SDM not only accounts for spatial spillover effects of the dependent variable across neighboring regions, but also incorporates the influence of independent variables in adjacent regions on the dependent variable of a given region. Consequently, the SDM is widely used for empirical testing of spatial spillover effects (Elhorst, 2017). To determine which spatial econometric model is more appropriate, Elhorst (2017) proposed a testing procedure. Specifically, two hypotheses of the panel SDM are examined and summarized using the Wald test and the likelihood ratio (LR) test based on the restrictions θ = 0 and θ + ρβ = 0. If both null hypotheses are rejected, the panel SDM should be employed. If θ = 0 and both the LM test and the robust LM test indicate spatial dependence in the dependent variable, the SDM reduces to an SLM. If θ + ρβ = 0 and the LM and robust LM tests suggest that the residuals exhibit spatial autocorrelation, the SDM simplifies to an SEM.

Results and Discussion

Each variable consists of 341 observations. The average population mortality rate, used as a health indicator, is 4.5 percent, with minimum and maximum values of 3.4 percent and 6.7 percent, respectively. Health expenditures, measured as a percentage of provincial gross domestic product, have an average value of 5.76 percent, with minimum and maximum values of 4.65 percent and 7.98 percent, respectively. The Moran’s I index for each variable is significantly positive, indicating the presence of spatial autocorrelation in health status, health expenditures, physicians per capita, hospital beds per capita, and education per capita across the 31 provinces of the country. The coefficient of the spatial model is significantly positive under all three spatial weight matrices. This finding indicates the existence of strong spatial spillover effects in provincial health outcomes in Iran. Therefore, the relationship between health status and government health expenditures should be examined from a spatial economics perspective. Investment in health services improves access to medical care, enhances disease prevention, and raises the quality of treatment. Increased health spending is generally associated with earlier disease detection, better management of chronic conditions, and a reduction in preventable mortality. In contrast, insufficient health expenditures may weaken the health system and lead to higher mortality rates, particularly among vulnerable population groups.

Conclusion

This study employed spatial econometric models to examine the effects of health expenditures, physicians per capita, hospital beds per capita, and education level on provincial mortality rates. The results reveal a significant spatial dependence in mortality rates across provinces, indicating that the health and socioeconomic conditions of each province affect not only its own outcomes but also those of neighboring provinces. This finding underscores the necessity of adopting spatial approaches in public health analyses. The estimations further show that increased health expenditures are significantly associated with lower mortality rates, highlighting the importance of targeted investment in health infrastructure and services. Moreover, increases in the number of physicians and hospital beds exert a significant negative effect on mortality, emphasizing the critical role of physical and human access to healthcare services. Finally, education level, as a key socioeconomic determinant, exhibits a significant negative impact on mortality rates, suggesting that improvements in education contribute to reduced mortality through healthier behaviors, greater awareness, and more effective utilization of healthcare services.