Every winter, a thick blanket of smoke descends over northern India. The culprit? Millions of farmers burning rice residue after harvest to quickly clear their fields for the next planting season. This practice, known as crop residue burning, contributes massively to air pollution—causing tens of thousands of premature deaths annually across India. The smoke travels hundreds of kilometres, affecting major cities like New Delhi. Despite bans and penalties, burning persists. Why? Because farmers face a genuine dilemma: mechanized harvesting leaves behind 8-12 inches of rice stalk, and they have only 2-3 weeks between harvesting rice and sowing wheat. Burning is cheap, fast, and familiar. Alternative methods exist—machines that can incorporate residue into soil, balers that bundle it for sale—but these cost money, require coordination, and many farmers worry they’ll hurt crop yields. The question isn’t whether burning is bad (the evidence is overwhelming), but rather: what constraints prevent farmers from adopting cleaner alternatives, and how can policy effectively address them?
Tackling Time, Volume, and Knowledge Constraints
In my job market paper, I investigate this question through a large-scale field experiment with 1,000 farmers in Punjab. I test whether interventions aimed at addressing the underlying time and information constraints can reduce burning more effectively. The intervention centered on subsidizing “early-maturity variety” (EMV) rice seeds that mature 20-25 days faster than traditional varieties. This seemingly simple change addresses multiple constraints simultaneously: it gives farmers an extra 3-4 weeks between harvests, reducing time pressure; the shorter plants produce less residue per acre, making alternatives to burning more manageable; and adoption provides a natural entry point for training farmers in sustainable residue management techniques like mulching, composting, and baling.
I randomly assigned farmers to three groups: a control group receiving no intervention, a group receiving a 50% seed subsidy (about $2.60 per acre), and a group receiving both the subsidy and hands-on training sessions about alternate methods of managing residue before and after harvest. The results reveal how these constraints interact. Both treatment groups reduced burning significantly, with particularly large effects from the combined subsidy-and-training approach. Control farmers fully burned about 3 acres on average; the seed subsidy alone reduced this by roughly 1 acre, while adding training reduced it by approximately 1.5 acres. The reduction came primarily from eliminating full burning (where both upper residue and basal stalks are burned) rather than partial burning (where only upper residue is burnt). Relative to a control mean of 3.1 acres under EMV cultivation, farmers in treatment groups adopted EMV seeds on 2.6-3.6 additional acres and were 8-10 percentage points less likely to report facing time constraints. They also significantly increased their use of sustainable alternatives like balers and mulchers. Interestingly, even farmers who received only subsidies—without formal training—reported improvements in their knowledge of sustainable practices, suggesting that EMV adoption itself triggers learning. However, the training made a measurable difference in one crucial way: it reduced equipment access barriers. Farmers who received training were 12 percentage points less likely to report delays in renting residue management machinery, while subsidy-only farmers showed no improvement.
The effects are significantly cost-effective. Every dollar spent generated $28-32 in social benefits from avoided greenhouse gas damages. The health benefits alone—preventing one acre of burning—exceed the intervention cost by roughly 400-460 times. Compared to a Payments for Ecosystem Services (PES) program in the same region (which I discuss below), this approach achieves each additional unburned acre at 53-60% lower cost.
Addressing Credit Constraints and Distrust Through Payment Design
A complementary approach to reducing burning involves direct financial incentives—paying farmers not to burn through PES contracts. Jack et al. (2025) conducted a randomized trial testing different PES contract designs in Punjab. Their key insight concerns when farmers receive money. Standard PES contracts pay farmers only after verifying compliance (no burning). But this creates potential problems: farmers must trust they’ll actually receive payment later, and they may lack cash upfront to pay for alternatives like equipment rental. The researchers compared three groups: farmers receiving no contract, those offered standard PES (full payment after verification), and those offered partial upfront payment (some money paid unconditionally upon enrollment, with the remainder paid after verification). They find that partial up-front payments reduce burning by 7.7–11.5 percentage points relative to the control group, while purely ex-post payments have no measurable effect. This research demonstrates that contract design details matter enormously for environmental programs in settings with weak institutions.
The Political Economy of Enforcement
Even well-designed incentives face implementation challenges in contexts where enforcement depends on local officials whose incentives may not align with environmental goals. Dipoppa et al. (2024) examine why burning bans fail despite official penalties. Using satellite fire data across South Asia, they show that burning increases when smoke drifts into neighboring administrative areas and reduces when smoke pollutes their own jurisdiction. This interjurisdictional externality fundamentally undermines top-down enforcement approaches. However, leadership characteristics can matter. Jagnani et al. (2025) find that women political leaders in India significantly reduce crop residue burning in their constituencies. Women-led constituencies show substantially lower fire counts during harvest seasons. The authors suggest women leaders may prioritize environmental and health outcomes differently and are more likely to favour policies curbing residue burning. This finding highlights that who governs matters for environmental outcomes, particularly when policies require sustained local enforcement rather than one-time interventions.