Bioeconomic feedbacks from large-scale adoption of transgenic pesticidal corn in the Philippines

Farmer control of agricultural pests raises the possibility of bioeconomic feedbacks and spillovers, whereby greater aggregate effort exerted on pest control lowers overall pest densities. This in turn decreases individual growers’ marginal incentives for pest control. While economists have written theoretically about such feedbacks or modeled it in simulations of bio-invasions, they rarely measure it econometrically. Here we adapt an instrumental variables methodology developed for discrete choice endogenous sorting models in the environmental and urban economics literatures to study bioeconomic feedbacks in pest control. As a methodological innovation, we introduce use of censored regression methods to handle 0% or 100% market shares in hedonic second-stage analysis of fixed effects in discrete choice models. We apply these methods to study area-level adoption and potential feedbacks from individuals’ decisions to adopt transgenic Bt corn, using a panel dataset from the Philippines. In a conceptual model, we generate the hypothesis that greater areawide deployment of Bt crops should reduce individual farmers’ incentives to use this technology, ceteris paribus. Our econometric estimation supports the hypothesis that greater areawide use of Bt attenuates individual incentives to use these varieties. In terms of economic significance, this feedback effect implies a mean long-run price elasticity for the Bt trait 67% lower than that implied by an econometric model ignoring it. Examining whether this estimated feedback relates to areawide pest suppression, we find farmers’ expectations about infestation from the main pest targeted by Bt crops are significantly reduced by higher areawide Bt deployment. We conclude by discussing the welfare and yield implications for these areawide bioeconomic feedbacks.