Market Unraveling and Quality Control in Agriculture

We ask a simple research question: do price incentives, cheap DNA verification technology, or both, reduce information asymmetries at the very earliest stage of the agricultural value chain? Determining which of these tools matters most could stimulate growth in seed markets for new varieties, which will help farmers adopt high-value crops and diversify their income streams. These outcomes are crucial for farmers to achieve, as the extreme weather conditions they face due to climate change are more regularly threatening their crop yields and productivity. Moreover, by ensuring that high-quality seeds are produced, more farmers may be willing to invest in them and successfully achieve higher yields through proper management.

2025-06-15

2025-11-30

Contract Acceptance; Genetic match; Contract Success; Delivered Quantity; Production Processes and Effort; Production Amount/Sales; Satisfaction with contract

See PAP attached

We will explore the use of two possible levers to improve seed quantity and quality: higher prices (incentives to growers) and DNA testing. These two interventions have the potential to deal with problems of moral hazard and adverse selection they may be causing this market to unravel. Higher incentives may be necessary to offset the costs growers incur when multiplying seed and to attract higher-quality producers back into the market, helping to mitigate problems of adverse selection in these markets. However, even high-quality growers may decide not to bear the costs of producing pure seeds if there is little risk that inspectors will reject their output. In this case, implementing DNA testing may be necessary to ensure compliance with best practices and overcome problems of moral hazard.
To formally test which strategies are most effective at improving the quality of seeds that growers supply, we will run an RCT across 408 villages with 4 farmers in each village. Each village will be assigned to one of three treatment arms and a control group:
1. Control - Will receive a small incentive if they follow a simple set of best practices for producing groudnut.
2. Incentive - Will receive a large incentive if they follow a simple set of best practices for producing groudnut.
3. DNA testing - Will receive a small incentive if the DNA verification confirms that the seeds are genetically pure
4. DNA testing + Incentive - Will receive a large incentive if the DNA verification confirms that the seeds are genetically pure

Not available

Computer

Village

Yes

408 villages

1632 farmers

4 farmers per village

It is important to note that we are likely to be better powered than most agricultural experiments because most of our main outcomes are precisely measured binary variables (did the farmer agree to multiply the seed, did they ultimately multiply the seed (and how much). We compute that with 100 clusters in each treatment arm, four farmers in each cluster, an inter-cluster correlation of 0.1 and 80% power, the standardized minimum detectable effect is 0.2 standard deviations. This effect size is of a reasonable magnitude - this is roughly a 6-percentage point increase in acceptance of the peanut growing contract.