Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
This study investigates how exposure to short and long-term measures of rainfall shocks and past crop diversification decisions influence subsequent diversification in Malawi and Tanzania. We use balanced household panel data combined with corresponding historical monthly weather data to achieve our aim. Crop diversification is studied as a state-contingent risky investment decision that households make before the state of nature is revealed. We model crop diversification decisions using correlated and dynamic random effects panel Poisson and Tobit models that control for unobserved heterogeneity in household crop diversification decisions plus initial conditions that may influence crop diversification across space and time. We establish that smallholder farmers in Malawi and Tanzania respond to rainfall shocks by diversifying their crop portfolios and that crop diversification decisions are state-dependent. Thus, farmers with knowledge and experience gained from past crop diversification have elevated chances to intensify subsequent crop diversification and adapt to recurrent rainfall shocks. Heterogeneity analysis reveals that the relatively better-off farmers with sufficient land and non-land household assets to a greater extent diversify the crop portfolio post-drought shock exposure. Pro-poor policies addressing seed access constraints are needed to support diversification and adaptation to shocks.