Causal Inference on Networks Under Continuous Treatment Interference: An Application to Trade Distortions in Agricultural Markets

Extensive work has been done to assess the role played by externalities in policy interventions. Studies dealing with this issue in experimental settings abound in the literature. Much less attention, however, has been devoted to the case when data comes from observational studies. In this paper, we address this gap by presenting a methodology to draw causal inference in a non-experimental setting subject to interference. Specifically, we develop a generalized propensity score-based estimator that allows estimating both direct and spillover effects of a continuous treatment. In order to showcase this methodology, we investigate whether and how spillover effects shape the optimal level of producers' support in agricultural markets. Our results show that, in this context, neglecting interference may lead to a downward bias when assessing policy effectiveness.