Speaker: Simone Piras (The James Hutton Institute)

Abstract
Innovation adoption by small producers is key to promoting sustainable food systems in developing countries. However, some innovations require collective management and thus cooperation. Public Good Games (PGGs) are well-established settings to experimentally assess individuals’ willingness to cooperate. Within the H2020 project FoodLAND, we ran PGGs among smallholders in Tanzania, Kenya, and Tunisia. In all countries, the decision how much to contribute to the public good was repeated twice, and in two, the design was changed between rounds. In Tanzania, we alternated different multiplication factors, corresponding to Marginal Per Capita Returns to the public good of 0.20 and 0.15, respectively. In Kenya, we played one round with unequal (two levels) and one with equal endowments, keeping the same total endowment at group level. We find that individual contributions do not vary significantly depending on the rate of return unless smallholders face a lower rate in the second round, in which case the efficiency of cooperation declines. Opposite to our hypothesis, group-level contributions were higher with unequal endowment; however, less endowed players contributed less in both relative and absolute terms. We further control for group size, which does not seem to affect cooperation, and for round-order effects, finding that smallholders reduce their contribution if they have cooperated relatively more than their group members in the previous round. Our results can help derive recommendations for the successful dissemination of collaborative innovations among smallholders farmers.