عوامل مؤثر بر انتشار دی‌اکسیدکربن استان‌های کشور: رویکرد اقتصادسنجی فضایی

Abstract
Introduction
Climate change driven by the accumulation of greenhouse gases has become one of the most serious global environmental challenges. Among these gases, carbon dioxide (CO₂) is the most significant pollutant generated by the production, transformation, and final consumption of energy. In Iran, the economic structure is highly dependent on fossil fuels, and energy consumption has remained relatively high because of energy-intensive production technologies, low energy prices, and the expansion of industrial and urban activities. As a result, CO₂ emissions have increased continuously over time.
At the provincial level, emissions are not distributed uniformly. Provinces differ substantially in terms of population density, industrial structure, income level, climatic conditions, and patterns of energy consumption. Moreover, environmental phenomena are not necessarily confined within administrative boundaries. Geographical proximity and interregional interactions may generate spatial spillovers, meaning that the emissions or energy-related characteristics of one province can affect emissions in neighboring provinces. Ignoring such spatial dependence may lead to biased or inconsistent estimates and, consequently, misleading policy implications.
Previous domestic studies on the determinants of CO₂ emissions in Iran have mainly focused on economic and social variables without explicitly accounting for spatial interactions among provinces. This study extends the existing literature by incorporating spatial dependence in provincial CO₂ emissions and estimating both the direct and indirect effects of the explanatory variables.
Objective
The main objective of this study is to identify and estimate the determinants of CO₂ emissions across Iranian provinces while explicitly accounting for spatial spillover effects. More specifically, the study aims to examine the effects of provincial per capita income, energy prices, industrialization, population density, and hot and cold temperature indices on the density of CO₂ emissions.
A further objective is to determine whether CO₂ emissions in one province are affected by emissions and explanatory variables in neighboring provinces. By distinguishing among direct, indirect, and total effects, the study seeks to provide a more accurate understanding of how economic, demographic, industrial, and energy-price variables influence provincial emissions. This distinction is particularly important for environmental policymaking, as policies implemented in one province may generate external effects in adjacent provinces.
Research Method
The study uses provincial panel data for 28 Iranian provinces over the period 2004-2011. Since official data on provincial CO₂ emissions are not directly available, CO₂ emissions are estimated using provincial energy consumption data and standard emission coefficients for different energy carriers. These emissions are then divided by the area of each province to obtain CO₂ emission density, which serves as the dependent variable.
The explanatory variables include provincial per capita gross regional product, energy price, the industrialization rate, population density, and hot and cold temperature indices. The energy price variable is calculated as a weighted average of the prices of different energy carriers consumed in each province. The industrialization variable is measured as the share of industrial value added in provincial gross domestic product. Population density is calculated as the ratio of provincial population to provincial area. The hot and cold temperature indices are constructed based on deviations from desirable seasonal temperature levels.
Methodologically, the study applies spatial panel econometric techniques. First, Moran’s I statistic is used to test for spatial dependence in CO₂ emission density across provinces. The positive and statistically significant Moran’s I statistic confirms the presence of spatial dependence. The Spatial Durbin Model is then estimated, as this model allows for spatial dependence in both the dependent variable and the explanatory variables. Wald tests are used to examine whether the Spatial Durbin Model can be simplified into a Spatial Autoregressive Model or a Spatial Error Model. The test results reject both restrictions, indicating that the Spatial Durbin Model is the appropriate specification. Finally, the spatial Hausman test supports the use of fixed effects. Therefore, the final model is estimated as a Spatial Durbin Model with spatial fixed effects.
Results and Discussion
The empirical results confirm the existence of statistically significant spatial spillovers in provincial CO₂ emissions. The estimated spatial autoregressive coefficient is positive and significant, indicating that CO₂ emission density in each province is affected by emission density in neighboring provinces. This finding suggests that environmental pollution in Iran has an important spatial dimension and that provincial emissions cannot be analyzed independently of neighboring regions.
The estimated direct effects show that provincial per capita income has a positive and significant effect on CO₂ emissions. A one percent increase in per capita income increases CO₂ emission density by approximately 0.29 percent. This result is consistent with the view that, at the observed income levels, higher income is associated with higher production, greater energy demand, and consequently higher emissions.
Energy price has a negative and statistically significant direct effect on CO₂ emissions. A one percent increase in energy prices reduces CO₂ emission density by approximately 0.70 percent. This result indicates that energy pricing policies can play an important role in reducing emissions. In the Iranian context, where energy prices have historically been low and subsidized, this finding suggests that energy price reform may contribute not only to improving energy efficiency but also to reducing environmental pollution.
The industrialization rate also has a positive and significant effect on provincial CO₂ emissions. A higher share of industry in provincial production increases energy demand and, consequently, CO₂ emissions. This result reflects the energy-intensive nature of industrial activity in Iran. However, it does not imply that reducing industrial activity is a desirable policy. Rather, it highlights the need for cleaner technologies, improved energy efficiency, modernization of industrial equipment, and greater use of low-carbon energy sources.
Population density has the largest positive direct effect among the explanatory variables. The results show that higher population density significantly increases CO₂ emission density. This finding is consistent with the concentration of energy consumption, transportation demand, residential fuel use, and economic activity in densely populated provinces and metropolitan areas.
The hot and cold temperature indices are not statistically significant in the estimated model. This suggests that, within the data and specification used in this study, climatic temperature variations do not have a robust measurable effect on provincial CO₂ emission density.
The decomposition of effects into direct, indirect, and total effects provides further insight. The indirect effect of energy price is statistically significant, implying that changes in energy prices in neighboring provinces can affect CO₂ emissions in a given province. This confirms the importance of spatial spillovers and shows that energy and environmental policies may have cross-provincial implications. By contrast, the indirect effects of industrialization and population density are not statistically significant, suggesting that their effects are mainly concentrated within the province itself.
Conclusion
This study investigated the determinants of CO₂ emissions across Iranian provinces using a spatial panel econometric approach. The findings show that provincial CO₂ emissions are spatially dependent and that ignoring spatial spillovers may lead to incomplete or misleading conclusions. The results indicate that per capita income, industrialization, and population density increase CO₂ emission density, whereas higher energy prices reduce emissions. In addition, the significant spatial spillover effect confirms that environmental pollution in one province is related to conditions in neighboring provinces.
The policy implication is that environmental and energy policies in Iran should not be designed solely at the level of isolated provinces. Since emissions and some policy variables generate spatial effects, coordinated regional and national policies are required. Energy price reform can be an effective tool for reducing CO₂ emissions, but it should be accompanied by complementary measures such as improving industrial energy efficiency, modernizing production technologies, promoting cleaner energy sources, and taking into account socioeconomic differences among provinces. Overall, the results emphasize the need for spatially informed environmental policymaking in Iran.