Subsidies and the African Green Revolution: Direct Effects and Social Network Spillovers of Randomized Input Subsidies in Mozambique - A Reproducibility and Replicability Assessment of Carter et al. (2021)

In this paper, we reproduce and test the robustness of the main results in Carter, Laajaj, and Yang (2021), who estimate the causal effect of a government-implemented input subsidy program for Mozambican maize farmers on maize yields and technology adoption. They find positive and persistent effects, and significant spillovers that benefit the social network of subsidized farmers. First, we reproduce the paper's main findings. Second, we conduct additional checks and show that the results are not sensitive to the inclusion of covariates highly correlated with the outcome of interest or to a sample restriction that focuses on clusters with enough variation in the treatment variable. Only the result associated with the long-run effect on fertilizer use is sensitive to the definition of the dependent variable or the choice of econometric specification. Overall, we can robustly reproduce most of Carter, Laajaj, and Yang's (2021) main results.