This website uses cookies.

By clicking the "Accept" button or continuing to browse our site, you agree to first-party and session-only cookies being stored on your device to enhance site navigation and analyze site performance and traffic. For more information on our use of cookies, please see our Privacy Policy.

  1. Home
  2. ASSA Annual Meeting
  3. 2027
  4. Preliminary Program
  5. Information, Learning, and Agricultural Technology Adoption in Developing Countries
« Back to Results

Information, Learning, and Agricultural Technology Adoption in Developing Countries

Paper Session

Sunday, Jan. 3, 2027 8:00 AM - 10:00 AM (EST)

Marriott Marquis Washington DC
Hosted By: American Economic Association
  • Chair: Nicholas Swanson, Cornell University

The Economic Consequences of Knowledge Hoarding

Luisa Cefala
,
Cornell University
Nicholas Swanson
,
Cornell University
Franck Irakoze
,
University of Burundi
Pedro Naso
,
Saint Louis University

Abstract

Social learning is an important source of knowledge diffusion in low-income countries. However, because developing country markets are often highly localized, individuals with social ties may compete more directly for the same economic rents, creating incentives for individuals to “hoard” their knowledge. This paper studies the impact of knowledge hoarding on the diffusion of profitable skills and technologies in rural Burundi, and measures its aggregate and distributional consequences for the village economy. In a field experiment covering 223 villages (labor markets), workers skilled in high-return agricultural technologies are encouraged to share their knowledge with unskilled individuals. We randomize at the local labor market level whether the unskilled worker is a competitor (i.e., someone from the same labor market) and whether the training is about a technology with rivalrous rents (row planting, which commands a wage premium in the labor market). We first establish that knowledge hoarding indeed reduces social learning. When incumbents are matched with an individual from the same labor market, knowledge transmission occurs only 3% of the time but reaches 43% if the unskilled worker is not a competitor. In contrast, transmission of technologies with nonrivalrous rents (e.g., composting) is high regardless of the unskilled worker’s identity. Next, we show that knowledge hoarding creates winners and losers: by hoarding knowledge, incumbents earn 6% more, and the skilled equilibrium wage is 3% higher. Instead, unskilled workers’ earnings and farm output are 7% and 20% lower, respectively. Altogether, knowledge hoarding reduces technology adoption by over 20%, suggesting substantial yield losses. Finally, our results suggest that fear of social sanction is a mechanism that sustains knowledge hoarding among the incumbents, highlighting how social ties can foster social learning but also inhibit it when knowledge diffusion threatens incumbents’ rents.

Learning from Self and Learning from Others: Experimental Evidence from Bangladesh

Florence Kondylis
,
World Bank
John Loeser
,
World Bank
Mushfiq Mobarak
,
Yale University
Maria Jones
,
World Bank
Dan Stein
,
Giving Green

Abstract

Can decentralizing demonstration accelerate learning about new technologies? This paper randomizes access to a fixed
demonstration kit for new flood-saline-resilient seeds across villages in Bangladesh, with demonstration either by a
single farmer or spread across many farmers. In the short run, higher learning from self and others under decentral-
ization increases technology adoption. In the long run, the impacts of any demonstration persist, but the additional
impacts of decentralization vanish. A Bayesian model of learning the returns to a new technology suggests belief
dispersion caused noisy adoption along the learning path, and farmers’ expected gains from demonstration are four
times higher under decentralization.

How Mechanistic Explanations Reshape Learning and Behavior: Evidence from a Fertilizer Choice Experiment in Eastern Uganda

Anirudh Sankar
,
Stanford University
Jessica Rudder
,
Oregon State University

Abstract

Mechanistic explanations—descriptions of a system through the causal interactions of its parts—play a key role in human cognition and scientific progress. Despite their importance, we lack systematic evidence on whether and how mechanistic explanations help lay decision-makers interpret information in complex economic environments. We evaluate the causal impact of including mechanistic explanations in an information intervention: public demonstrations of fertilizer use for smallholder tomato farmers in Eastern Uganda. In all demonstrations, extension officers showcased the impact of a recommended fertilizer recipe. In the treatment group, officers also explained the mechanisms underlying the recipe’s effects—introducing the language of macronutrients and the causal processes linking nutrients, soil features, and plant growth. We collected detailed data on beliefs and behaviors from 797 farmers in a lab-in-the-field experiment conducted at the demonstration site and followed up with them over two growing seasons. In the lab-in-the-field, treated farmers generalized more effectively—making better substitution and arbitrage decisions among fertilizers and achieving 9% higher simulated profits in an incentivized fertilizer-application task. At endline, treated farmers’ real fertilizer choices reflected improved nutrient timing and balance, and their yields were 14% higher.

Generative AI for Agricultural Advisory in Kenya

Joshua Deutschmann
,
University of Chicago

Abstract

We evaluate FarmerChat, an AI-powered agricultural advisory chatbot developed by Digital Green, using a cluster-randomized trial in 600 villages and 3,000 smartphone-owning farm households in Nakuru County, Kenya. We estimate effects on farmers' knowledge, adoption of recommended practices, yields, and income. AI chatbots can deliver tailored agronomic advice at low marginal cost, but evidence on their effectiveness is scarce. A key design concern is that lower-income farmers may be furthest from the productivity frontier and least exposed to peers who have transitioned to higher-value practices, raising the cognitive cost of formulating ambitious queries — an "idea trap" in which the farmers with the most to gain engage least with the tool. We cross-randomize a video intervention featuring local farmers who describe attainable production frontiers for common crops and explain how they closed yield gaps. We estimate whether expanding farmers' opportunity sets complements access to AI-enabled advice.
JEL Classifications
  • O1 - Economic Development
  • O3 - Innovation; Research and Development; Technological Change; Intellectual Property Rights