Does confidence influence demand and responsiveness for information? It’s the question Jared Gars, a PhD candidate from the University of Wisconsin, set out to address in his job market paper titled: “Confidence and Information Usage: Evidence from Soil Testing in India,” which was presented on February 2nd as part of NCID’s Weekly Seminars. This question is of great relevance as information interventions are used by policymakers to incentivize behavioral change on various circumstances, especially in developing countries.
In 2015, the Government of India launched a centrally-sponsored “Soil Health Card Scheme” which cost approximately 85 million dollars. The objective was to provide soil health information to 140 million farmers on a triennial basis in order to secure long-term soil health, as in recent years there has been an imbalance in the use of fertilizer in India. Nevertheless, acquisition of this information does not imply compliance. Farmers may feel overconfident and ignore recommendations due to their extended experience, which is as high as 30 years, on average.
It is in this context that the paper uses data on randomly treated farmers across villages in rural India. The authors obtain measurements of farmers’ confidence on optimal input usage, and are able to observe their decisions throughout the season, which allows them to estimate responsiveness to the program. Farmers’ trust towards the system were recorded as well.
Results provide strong evidence that as confidence increases, farmers will be less willing to follow recommendations provided by the soil health program. Note that confidence is measured as the dispersion of a farmer’s beliefs on optimal input usage, such that, those who are less disperse are considered to be relatively more confident. Furthermore, it seems that lower responsiveness, defined as a change in input usage due to the acquired information, is associated to higher confidence and ability.
Finally, other important factors for the successful implementation of this program are the credibility of the provided information and the farmers’ literacy rate. Farmers who are literate or have greater trust in the system, are willing to pay more for this information. However, there is no evidence that this factors have an impact on responsiveness.