While the health impact of COVID-19 in most African countries appears modest, the impact of social distancing measures, closing of markets and reduced mobility is felt across the board. Domestic, labor-intensive and traditional food value chains and the smallholders they serve appear to be particularly affected. During a systemic shock where idiosyncratic risk coping strategies fail, collective or organizational resilience becomes of the essence to protect the livelihoods of smallholders. In this study, we have used pre- and during-shock data on agricultural cooperatives from Southeast Africa to understand how resilient these smallholder-owned organizations are. We find that many organizations could not countervail market-disruptions and fell into a state of dormancy during the pandemic. One reason for this is that collective decision-making was heavily affected by the banning of gatherings. Only a few organizations devised innovative solutions to maintain the market linkages of rural smallholders. The lack of resilience demonstrated by most cooperatives appears to be associated with organizational immaturity, large membership size, elite capture and limited business-orientation, which underscore a general lack of managerial capital.

The dependence of rainfed agriculture on weather shocks means that managing risk is an important preoccupation of smallholders in Niger. In the absence of objective probabilities that can be attached to rainfall conditions, farmers formulate their own beliefs about uncertain outcomes. The role that these beliefs play in household decision making, and thus the potential gains associated with more skilled forecasts, is not yet well understood. In the current paper, we aim to fill this gap in the literature. In what follows we lay out an inter-temporal model of farm household decision making and apply this to recent survey data of rural households in Niger to shed light on whether and to what extent household behavior regarding investments in risky inputs is affected by expected rainfall conditions. We find that a farmer’s belief that rainfall conditions are favorable is associated with more spending on risky inputs and that returns to input use are higher when we control for the bias induced by omitting household decision-making on input use from the production process. Our results suggest that there is potential for weather information to induce profit-maximizing behavior of risk-averse farmers.