The objective of our paper is to provide an explanation for the lack of joint adoption by farmers of cleaner technologies in banana production, specifically fallow period (FP) and disease-free seedlings (DFS). Our hypothesis is that while these technologies are synergistic from an agronomic and environmental perspective, and thus efficient from a social interest perspective, they are substitutable rather than complementary from a farmer’s private interest perspective. In other words, farmers receive lower returns from adopting both technologies together than from adopting them in isolation. To test this hypothesis, we present a unified empirical framework for assessing complementarity. We estimate a structural model of complementarity that overcomes the unobservable heterogeneity bias found in previous models using a database of 607 banana farmers in the French West Indies. Our results support our hypothesis, showing a substitution effect between FP and DFS rather than a complementarity effect. Moreover, we observe a contrasting profile of adopting farmers: smallholders who are reluctant to change adopt FP, while more specialized farmers who anticipate a pesticide ban adopt DFS. A public policy that promotes joint adoption should compensate smallholders for the cost of the DFS technology, while compensating more productive farmers for leaving their land fallow.

Pollution from fossil fuel use is a global problem. Studies have shown that a worsening of environmental quality has adverse effects on worker productivity and health. In this study, there is an inexhaustible natural resource that deteriorates environmental quality and affects productivity. There also exists a perfect substitute clean backstop, which is initially too costly to operate and whose costs can be reduced through investments in knowledge. Depending on the endowment of environmental quality, the optimal solution shows that the planner should only use the resource or only the backstop until a constant steady state is reached in which the polluting resource and backstop are used in fixed proportions. We show that investments in alternative technologies from the very beginning can help an economy make the eventual switch to clean energy sources, thereby attaining better environmental quality.