Agriculture

The impact of climate change on Indian crops

  • Blog Post Date 13 June, 2023
  • Articles
  • Print Page
Author Image

Madhu Khanna

University of Illinois at Urbana-Champaign

khanna1@illinois.edu

Author Image

Surender Kumar

Delhi School of Economics

skumar@econdse.org

This article looks at deviations in the impact of short-term weather events and long-term climate change on yields of rice, maize and wheat, and finds that the negative impact of temperature is higher in the short run than in the long run, implying that farmers were able to adapt to long-term changes in climate. It also discusses the impact of precipitation on crop yields, and concludes by highlighting the need for customised farm management policies to adapt to climate change. 

Climate change and the growing frequency of extreme weather events are wreaking havoc on crop yields, and forcing farmers to take adaptive measures to limit damage to their crops. Most studies that purport to measure the effects of climate change are looking at year-to-year changes, which are representative of variations in weather and not climate (see, for example, Deschênes and Greenstone 2007, Schlenker and Roberts 2009, Miao et al. 2016, Zhang et al. 2017, Malikov et al. 2020, among others). 

The effects of long-term climate change are not uniform across regions, productivity zones, and crops. In a recent study (Kumar and Khanna 2023), we look at three major crops – rice, maize and wheat – over 50 years (1966-2015) to understand the long- and short-term effects of climate change on crop yields. 

Measuring the impact of climate, not weather 

Weather change is a short-term phenomenon, and may be different from long-term changes. If farmers observe changes in long-term patterns of climatic conditions, it is expected that they would adapt by changing their farming and irrigation practices. 

We intend to find out whether the effect of short-term deviations in extreme temperature and precipitation are significant and have a negative effect on yields relative to their long-term averages. If these effects are absent in the long-term, it implies that farmers are adapting to climate change.   

We use quantile regression models1 to determine if farmers were adapting to the long-term changes in climate. This approach allows climate change impacts to systematically vary by the productivity distribution of crop yield – it explicitly allows distributional heterogeneity in the effects of climatic variables on crop yields in low-, average- and high-productivity areas. This gives us a complete account of the relationship between the distribution of the dependent variable and its determinants. With this methodology, we use 50 years’ data on temperature, precipitation, the length of the growing season, and crop yields to formulate different models for short- and long-term responses of crops. 

Our findings 

We find that increasing temperatures and declining precipitation reduce crop yields, but also increase dispersion of the observed yields. We also find that the farmers were able to adapt to changes in temperature for rice and maize, but not in case of wheat. For rice, a 1o C increase in temperature led to a 6% decline in crop yields in the short run and a 4% decline in the long run. For maize, the effect was a 9% and 1% decline in the short and long run respectively. We observe that farmers are customising their strategies across different regions and crops – for example, heat-prone districts would naturally fare better to higher temperatures compared to districts in colder regions. 

Moreover, we find that the impacts are higher at the lower tail of the productivity distribution but are lower at the upper tail. A 1o C increase in temperature lowers rice and wheat productivity by 23% and 9% respectively in the first quantile, but the damage is only 6% and 5% at the ninth quantile. Interestingly, farmers who worked in areas that were less productive differed in their response to those who worked in areas where the yields were higher. This could be explained by the fact that in less productive areas farmers take more adaptation measures due to larger impacts of temperature increase. For example, higher productivity regions already have more irrigation facilities and are less dependent on the monsoon – as a result, the difference between long-term and short-term impacts is negligible.   

A 100-millimetre increase in precipitation enhances rice yields by 2.4% on average, but rice productivity increases by 3.5% at the lowest tail of the productivity curve, as compared to only about 1% for the top 10% of producers. Less productive regions benefit more from the increase in precipitations relative to high productive regions, consistent with the fact that high productive regions have better irrigation facilities. 

We find the impact of 1o C rise in temperature leads to a 3% decrease in wheat productivity in both the short and long run. Since we observe that there is no significant difference, it suggests the absence of any adaptation. The short and long run response of median wheat yield to precipitation follow a similar pattern, and there is no significant difference in the magnitude of yield response functions. Unlike rice, a 100 mm increase in precipitation negatively affects wheat productivity by 2.5%. 

Conclusion 

In sum, we find that the effects of weather and climate change are asymmetric over the distribution of yield curve – farmers at the lower tail of the productivity curve experience larger effects in comparison to the farmers working at the upper tail of the distribution. It is interesting to note that significant adaptation measures are taken at the lower tail of yield distribution curve. 

However, the adaptation to changing weather and climate conditions is not uniform across crops, implying that the customised farm management policies will be useful for adapting to changing weather and climate conditions. 

Note:

  1. Quantile regression estimates the conditional median (or other quantiles) of the response variable across values of predictor variables, whereas the method of least square estimates the conditional mean of the response variable.

Further Reading 

  • Deschênes, Olivier and Michael Greenstone (2007),‘The Economic Impacts of Climate Change: Evidence from Agricultural Output and Random Fluctuations in Weather’, American Economic Review97(1): 354-385.
  • Kumar, Surender and Madhu Khanna (2023), “Distributional heterogeneity in climate change impacts and adaptation: Evidence from Indian agriculture”, Agricultural Economics, 54(2): 147-160.
  • Malikov, Emir, Ruiqing Miao and Jingfang Zhang (2020), “Distributional and temporal heterogeneity in the climate change effects on US agriculture”, Journal of Environmental Economics and Management, 104: 102386.
  • Miao, Ruiqing, Madhu Khanna and Haixiao Huang (2016), “Responsiveness of Crop Yield and Acreage to Prices and Climate”, American Journal of Agricultural Economics, 98(1): 191-211.
  • Schlenker, Wolfram and Michael J Roberts (2009), “Nonlinear temperature effects indicate severe damages to US crop yields under climate change”, Proceedings of the National Academy of Sciences, 106(37): 15594-15598.
  • Zhang, Peng, Junjie Zhang and Minpeng Chen (2017), “Economic impacts of climate change on agriculture: The importance of additional climatic variables other than temperature and precipitation”, Journal of Environmental Economics and Management, 83: 8-31. 

No comments yet
Join the conversation
Captcha Captcha Reload

Comments will be held for moderation. Your contact information will not be made public.

Related content

Sign up to our newsletter