The climatic impacts on rice yield in the Indian state of Odisha: an application of Just-Pope production function and quantile regression.

Abstract

The impacts of climate change on Indian agriculture are well documented. However, there is a dearth of research addressing the inter-regional diversities in the impacts. Furthermore, existing studies are mostly restricted to the impacts on mean agricultural yield, overlooking the impacts on yield variability. This study investigates climatic impacts on rice yield in Odisha, a coastal Indian state, employing district-level panel data from 1995 to 2017. This study uses the Feasible Generalized Least Squares (FGLS) and quantile regression approaches to estimate how climatic impacts are realized on mean yield, yield variability, and conditional quantiles of yield distribution. The use of the agro-climatic zone level dummies empowers the study to account for local weather and soil conditions. The results reveal substantial heterogeneity in the climatic impacts across agro-climatic zones. A marginal increase in the maximum temperature reduces rice yield by 279-325kg/ha across different agro-climatic zones, largely constituting the state's coastal region, but increases yield by 340-766kg/ha across the zones lying in the western region. Conversely, an increase in minimum temperature increases rice yield by around 211kg/ha in the coastal region, but reduces by 333-744kg/ha in the western region. In addition, an increase in precipitation reduces yield by 0.318 kg/ha in some parts of the coastal region, but increases yield by 0.268 0.881 kg/ha across other regions. Similar heterogeneity is also identified across conditional quantiles of yield distribution. The heterogeneous impacts call for strategic policy intervention specific to zones and quantiles.