A new paper (gated) by Michael Carter, Rachid Laagja and Dean Yang shows, using a randomized fertilizer subsidy, that reducing costs increases adoption, but also, somewhat in opposition to previous research and importantly, that adoption is persistent into the following season.
First, we provide one of the first randomized controlled trials of the impact of an input subsidy program, and the first to measure impacts on a range of important household outcomes beyond fertilizer use itself. The only previous study using randomized methods is Duflo et al. (2011), who estimate impacts of fertilizer subsidies on fertilizer use alone (in rural Kenya). We show positive impacts of input subsidies (in Mozambique) on a range of outcomes beyond input use, including farm output, household consumption, assets, and housing quality.
Second, we find positive effects of input subsidies that persist up to two annual agricultural seasons beyond the season in which the subsidies were offered. This result contrasts with Duflo et al. (2011), who find no persistent impact of either heavy (50%) subsidies for fertilizer or the well-timed nudge of offering free delivery at the time of the previous harvest. Both treatments raise fertilizer use in the season they are provided, but impacts are very close to zero and not statistically significantly different from zero in the next season.
Having spent a lot of time lately with a friend writing a book on fertilizer and the apparent failure to launch of Africa’s Green Revolution, my thoughts immediately go to whether the fertilizer available on the market is real and how perceptions of fake fertilizer are affecting the decisions of farmers to continue (or not) using fertilizer in their fields.
Luckily, a few people are looking into this and maybe we’ll have some answers soon.