Climate change impacts on agriculture: do spatial spillovers matter?

The relative lack of research concerning the potential impacts of climate change on different sectors in developing countries, especially the Middle Eastern countries, as the essential prerequisite of climate policy actions has made these countries the frontline against climate impacts. To fill this gap, we use a non-market valuation model to assess the future impacts of climate change on agriculture in Iran. In this study, the relationship between farmland net revenue, as a proxy for land values, and climate change is investigated using a long-spanning Ricardian framework. For farm variables, we take into account novel hydro-climatic variables and climate extreme indices by taking advantage of a high-resolution meteorological dataset to tackle the sparse distribution of weather stations in Iran. For non-farm variables, we consider the pressure of rapid urbanization and migrations from rural areas in addition to socio-economic variables. This study also contributes to the body of literature through methodological improvement by taking advantage of spatial panel econometrics to develop a more robust and consistent model against spatial dependency, spatial heterogeneity, and omitted factors extraneous to the agriculture sector. The estimated coefficients are then employed in projecting long-run welfare impacts on the agricultural sector under several climate change scenarios based on the sixth phase of the Coupled Model Intercomparison Project (CMIP6) in 2050 and 2080. The results show that although climate change probably could have deleterious impacts on agriculture when we see the whole picture, its impacts would highly depend on climate zones and geographical locations. Generally, counties in snow and warm-temperate climate classes would be less susceptible to climate impacts than arid and semi-arid counties. Besides, climate change could even be beneficial for agriculture in a few counties owing to a decrease in cold extreme events frequency and intensity and an increase in growing season lengths and effective growing season degree-days. Thus, we suggest that these positive factors of climate change should be included in empirical studies to avoid overestimating the disruptive impacts of climate change. Finally, we argue that overlooking spatial dependency and spatial heterogeneity in Ricardian models could substantially affect impact assessments.