Lead Papers National Symposium on Climate Change and Rain-fed Agriculture 18-20, February, 2010

Abstract

Climate change impact studies on agriculture are broadly divided into agronomic-economic approach and Ricardian approach. The Ricardian approach, similar in principle to the Hedonic pricing approach of environmental valuation, has received significant attention due to its elegance and also some strong assumptions it makes. Broadly, this approach attempts to estimate climate response function using crosssectional evidence on land values (or, farm level net-revenue as used in the present analysis) and climate, while controlling for all other cross-sectional variations. This paper attempts to extend the existing knowledge in this field by specifically addressing two important issues: (a) extent of change in climate sensitivity of Indian agriculture over time; (b) importance of accounting for spatial features in the assessment of climate sensitivity. The analysis based on four decades of data suggests that the climate sensitivity of Indian agriculture is increasing over time, particularly in the period from mid-eighties to late nineties. This finding corroborates the growing evidence of weakening agricultural productivity over the similar period in India. The results also show presence of significant positive spatial autocorrelation, necessitating estimation of climate sensitivity while controlling for the same. While many explanations may exist for the presence of spatial autocorrelation, this paper argued that inter-farmer communication could be one of the primary reasons for the spatial dependence. Field studies carried out in Andhra Pradesh and Tamil Nadu through focus group discussions provided limited evidence in this direction.