Distributional impacts of climate change on agricultural total factor productivity in India

Abstract

This paper assesses the distribution of climate change impacts on agricultural Total Factor Productivity (TFP) across districts in India. Combining the district-level TFP, estimated using multiple rounds of nationally representative agricultural surveys, with climate and other controls, the relationship between production efficiency and climate is estimated for two time points: 2002–2003 and 2012–2013. The estimated climate response function suggests that for every 1 °C rise in temperature, agricultural productivity reduces by ∼4.5%. Using estimated climate sensitivity and regionally downscaled climate projections, the study further assesses the impacts on agricultural TFP across districts over the mid-century. By 2050, TFP in agriculture is projected to decline for all the states considered in the study. The latter-period (2012–2013) climate response function projects more adverse impacts compared to the early-period (2002–2003) response function. The results also show increase in the magnitude of impacts over time, indicating that Indian agriculture has become more climate sensitive.

Keywords:

• Climate change
• agriculture
• total factor productivity
• India
• impact assessment
• distributional effects

JEL CODES:

• D24
• Q15
• Q51
• Q54

Acknowledgement

The authors would like to thank the anonymous reviewers for their valuable comments on the paper. The authors thank Dr. A. Balasubramanian and Mr. Rishabh Mahendra for their research assistance.

Disclosure statement

The authors declare no potential conflict of interest.

Notes

1 This “output per unit of input” or “index” interpretation of Total Factor Productivity is attributed to Stigler (1947). This interpretation got a theoretical support through its linkages to the production function in a neoclassical framework suggested by Solow (1957) and is commonly known as the “Solow residual”. The term TFP however has contradictory interpretations in the literature (see Lipsey and Carlaw 2004). In its most vague interpretation, it has been termed as “measure of our ignorance”.

2 The interpretation of TFP containing measurement errors needs qualification in empirical studies on TFP measurement. While the issue of measurement errors could potentially be severe, several tools such as stochastic frontier production function estimations have the ability to address this issue.

3 The study by Letta and Tol (2019) is an exception. The study is based on the growth modelling framework suggested in Dietz and Stern (2015) to examine the effects of changes in weather and climate on growth in TFP across countries.

4 Globally as well as in the Indian context most literature on TFP study the “growth empirics” relating to TFP focusing largely on the time trends in TFP growth, thereby ignoring the spatial patterns. However, climate change impacts are likely to have heterogeneous effects across other welfare dimensions (e.g. income class) and population sub-groups (e.g. small farmers or older population). See Pattanayak et al. (2021) for a discussion of welfare consequences associated with climate change impacts on agricultural TFP.

5 Besides poverty reduction, other consequences of productivity improvements in agriculture include, but are not limited to, changes in employment generation, improvements in livelihood options, increased standard of living, greater investments and improvements in rural infrastructure.

6 See Dell et al. (2014) for a broad overview of both strands of the literature with their respective advantages and disadvantages. The first strand of the literature employs panel data regression models, pioneered by Deschênes and Greenstone (2007). The second strand of literature relies on cross-sectional regressions (also known as the Ricardian approach) propounded by Mendelsohn et al. (1994).

7 For the general formulation of TFP growth in terms of input-cost shares, see Jorgenson and Grilliches (1967) and Hall (1989). See Aiyar and Dalgaard (2005) and Hall and Jones (1996) for the formulation of TFP growth in the context of a cross-section of countries.

8 Value of land reported in the 59th round were self-reported or ‘notional’ values, whereas normative estimates of value of land was considered in the 70th round, potentially correcting for the errors in the notional values (see Report on “Key Indicators of Debt and Investment in India”, Ministry of Statistics and Programme Implementation, 2014). While some discrepancy in estimates of land values could be expected, the approach to systematically identify such discrepancy is not documented.

9 The periods 2002–2003 and 2012–2013 correspond to the survey years for NSS Situation Assessment Survey of farmers 59th and 70th rounds respectively, which provide the land value information but not the wage information. Data for the wage regression was taken from the NSS rounds of Employment and Unemployment closest to the NSS Situation Assessment Surveys in both periods. The survey years and survey rounds corresponding to the NSS situation assessment survey have been interchangeably used in this study. Alternatively, the survey years and rounds are also referred as period 1 and period 2.

10 Data for climate change projections have been obtained from Climate Change Information Portal (http://www.climatevulnerability.in/).

Additional information

Notes on contributors

Anubhab Pattanayak

Anubhab Pattanayak is Assistant Professor at Madras School of Economics, Chennai. He has obtained his Master’s degree and Ph.D. in Economics from Anna University, Chennai and University of Madras, Chennai respectively. His Doctoral research has assessed the impacts of climate change on Indian agriculture focusing on rice. He has also several years of professional experience in corporate/consultancy organizations. He has been awarded the India-IIASA YSSP fellowship during his doctoral studies. He has six years of teaching experience and has published in peer-reviewed journal articles, chapters in edited books, and contributed to research project in the area of climate change and environment.

K. S. Kavi Kumar

K. S. Kavi Kumar is professor at Madras School of Economics, Chennai. Kavi Kumar has over 20 years of experience in research, teaching and industry. After receiving Bachelor’s and Master’s degrees in Engineering, he completed his Ph.D. in Development Economics in 1998 from Indira Gandhi Institute of Development Research, Mumbai with focus on “Climate Change Impacts and Indian Agriculture”. Dr. Kavi Kumar has been associated with various international and national institutions including The World Bank, International Institute for Applied Systems Analysis, Potsdam Institute for Climate Impact Research, GIZ, Tata Energy Research Institute, Institute of Economic Growth etc. for various research assignments. Since 1999, he has been working with Madras School of Economics. At Madras School of Economics, Dr. Kavi Kumar had coordinated the activities of the Centre of Excellence in Environmental Economics supported by the Ministry of Environment and Forest, GoI during 2008 to 2015.

Lavanya R. Anneboina

Lavanya R. Anneboina is an independent researcher based in Chennai, India. She has obtained her Masters in Environmental Economics from University of York, UK and Ph.D. in Economics from University of Bath, UK. She has worked as a Consultant at Madras School of Economics, Chennai during 2014–2017 and as Assistant Professor at Tata Institute of Social Sciences (TISS), Mumbai during 2012–2013. She has contributed to a number of research projects, working Papers and has published in peer-reviewed journals including Ecosystem Services and Economic and Political Weekly.