Extinguishing the Blaze: Reducing Crop Residue Burning in India

Policy Context

Crop residue burning in India contributes up to 40% of the air pollution during winter months- often pushing the air quality index 100 times over the safe limit (The Guardian, 2023). Existing evidence highlights various factors contributing to crop residue burning- credit constraints (Jack et al. 2022), narrow time frame to prepare soil for the next crop (Kumar et al. 2015) and informational constraints about sustainable crop residue management (IGC 2022). Since there is limited evidence on the impact of directly reducing timing and informational constraints on crop residue burning, the research team focuses focus on evaluating the impact of removing these barriers on burning outcomes.

Early maturity seed varieties developed by Punjab Agricultural University could mitigate time constraints between rice harvest and wheat sowing. This reduces the pressure on farmers to quickly burn crops and promotes sustainable crop residue management, while still leaving farmers no worse-off than before (PAU Ludhiana 2018). Using a randomized controlled trial, this project will evaluate the impact of subsidies for early maturity variety seeds and technical training about sustainable residue management on burning outcomes.

Study Design

The study is designed as an RCT, with the farmers randomized into one of the three groups- ‘subsidy only’ group, ‘subsidy and technical training’ group and a control group:

1. Seeds subsidy: Farmers in the ‘subsidy only’ and ‘subsidy and technical training’ groups will be offered 50% subsidy on purchase of early maturity variety of seeds, covering their land holding size or 10 acres, whichever is lower. The offers will be extended to farmers a month before they make their purchasing decisions (usually mid-to late April).
2. Technical training: Farmers in the `subsidy and technical training’ group will receive technical training in sustainable methods of residue management before and after the harvest season (September-November).

Randomization is done at the farmer level, and stratification at the village level. The team will divide a 930 farmer sample equally between the three groups to maximize power. For the primary outcome- percentage of area burnt, this sample size is powered to detect an effect of 10 percent of the control mean.

Results and Policy Lessons

Forthcoming.