BEER Lab Meeting - Mirko Reul

The world's 500 million smallholder farmers produce two-thirds of the food but struggle with low productivity and vulnerability to climate change.

Development interventions that alleviate these stressors with in-person encouragement towards adopting sustainable practices are challenging to scale, despite their proven benefits.

We present an experimental design that tests if (1) descriptive norms and (2) risk framing can increase farmer engagement with information on sustainable practices in a chatbot, which would allow encouragement at scale.

Our study leverages an existing WhatsApp chatbot that supplies over 100'000 farmers in Madhya Pradesh (India) with information on market prices.

After recruiting 2'000 principal farmers from the active user population and surveying their farming profile and risk preferences, we treat participants with a message on natural farming practices that have known benefits but low prevalence in the region.

Our factorial treatments frame the message to induce (1) perceptions of beneficial practice adoption by peers, and a preference for adoption over either (2.1) certain losses or (2.2) uncertain gains from non-adoption.

We expect that compared to a neutrally framed message, all five treatment conditions increase farmer engagement with detailed practice information and step-by-step implementation guidelines, which are sent sequentially by the chatbot.

We further expect that the peer adoption treatment increases the diffusion of practice information to non-participants over the diffusion of market information.