This study examined how and why extension workers shared farming videos with farmers, revealing divergent appropriation patterns and their implications for digitization in agriculture. 294 extension workers in Bihar (India) were asked to circulate three wheat agronomy videos with farmers. Extension workers’ circulation of these videos was observed using link tracking, phone surveys, and follow-up interviews. Results were analyzed using a novel analytic framework based in affordance theory. Exten…
Read moreThis study examined how and why extension workers shared farming videos with farmers, revealing divergent appropriation patterns and their implications for digitization in agriculture. 294 extension workers in Bihar (India) were asked to circulate three wheat agronomy videos with farmers. Extension workers’ circulation of these videos was observed using link tracking, phone surveys, and follow-up interviews. Results were analyzed using a novel analytic framework based in affordance theory. Extension workers varied widely in how much, how, and why they shared the farming videos. This variation was underpinned by extension workers’ differing incentives and goals. In other words, extension workers heterogeneously appropriated—rather than homogeneously deployed—the practice of sharing farming videos. Some but not all of these appropriations were desirable from the perspective of service managers. For theory, extension workers’ appropriations of farming videos demonstrate that prevalent conceptualizations of digital agricultural technologies do not account for the adaptation of these technologies by farmers and other actors in agricultural innovation systems. For digital agriculture evaluators, the findings caution against the prevalent focus on averaging effects of interventions and highlight the need to examine the variability of these effects within and across interventions. For extension service managers, the findings emphasize the importance of engaging extension actors with farmer-aligned incentives and goals. This study was limited in focusing on the video-sharing behaviors of human extension actors and not on algorithmic extension actors, like YouTube or farming advisory chatbots powered by large language models. However, the findings have implications for both: just as human actors variably appropriate digital tools, algorithmic extension actors also embed implicit goals that shape how agricultural information circulates. Future research should examine the goals and behaviors of these algorithmic actors that have increasing influence in agricultural innovation systems.
