Can informed buyers improve goods quality? Experimental evidence from crop seeds

Working Paper
Published on 20 January 2022


We study whether adding informed buyers to a market can improve the quality of goods supplied by sellers, in an environment where goods quality is difficult to observe. To do so, we implement a market-level intervention, randomizing rural markets in Kenya into a community-wide information campaign. Small-scale maize (corn) farmers in treated market areas were trained to identify hybrid maize seed that is quality verified under national seed regulations. We do so in a region where there are widespread concerns about deceptive counterfeits and other uncertified seeds of lower quality. In a first main result, we find that observable markers predict seed quality, and treatment increased knowledge of these markers and affected seed purchase decisions. We show that substantial gains in yields were experienced among subgroups of informed buyers that had larger gaps between baseline seed quality and national standards. We do not see effects on seed quality offered to uninformed buyers in the same communities, as revealed by data from secret shoppers. These patterns can be explained by a model in which sellers' ability to discriminate allows lower-quality products to continue to be offered to uninformed buyers, even while informed buyers experience benefits and their increasing numbers hurt firm profits. Consistent with the model predictions, we find that treatment caused some seed sellers to exit the market, and we do not detect changes in seed prices. Taken together, the findings document new stylized facts and provide evidence relevant for boosting yields of a staple crop crucial for food security. More generally, they provide lessons concerning the role of improved consumer information in disciplining firms in low information environments.


Eric Hsu

Yale University

Anne W. Wambugu

Strathmore Energy Research Centre