Impacts of cooperative membership on banana yield and risk exposure: Insights from China

Abstract The yield‐increasing effects of agricultural cooperative membership have been widely examined in the literature. However, so far, little is known about whether cooperative membership has the potential to reduce farmers’ exposure to production risk. We address this gap by estimating the impacts of cooperative membership on expected yield, yield variance (variability), and yield skewness (exposure to downside risk), using data collected from a survey of 626 banana farmers in China. We employ an endogenous switching regression model to address the selectivity bias issue. The empirical results show that cooperative membership increases banana yield by 3% and reduces the variance and downside risk exposure by 60% and 114%, respectively. The results are supported by robustness check estimates, using propensity score matching and inverse probability weighted regression adjustment models. Additional analysis reveals that the treatment effects of cooperative membership vary among members with different household and farm characteristics. Our findings suggest that agricultural cooperatives can be an effective institutional arrangement for reducing production risk and crop failure and point to the need for policies and programmes in developing cooperatives and increasing membership involvements of smallholder farmers.