Overall, our results reveal the limitations of market-based producer organizations to provide the same long-term support as state-backed commodity boards for smallholder producers. We suggest that these organizations need some sort of subsidies (think reinsurance) to survive in the long term.

When we match participant performance in the game to administrative data on member deliveries, we find that treatment effect of the additional income decreases in years of membership. New members are particularly vulnerable, and we suggest a focus on them (education and maybe subsidies).

Of the four elements, the additional income treatment reduces side-selling the most: it increases loyalty of the participants by 10%. Nudge reminders of services, small market shocks, and production shocks have no effect. Punchline: cooperative members deviate when they are liquidity constrained.

We find a strong preference for price certainty: participants allocate on average 82% of their harvest to the certain-price buyer. Moreover, 22% of the participants always allocate their entire harvest to the certain-price buyer. We randomly include nudge reminders of additional services.

Over 60 growing seasons, participants roll a die to find out their harvest and rough outside buyer price. They allocate their harvest between a certain-price buyer and an outside buyer. Then they find out the realized price of the outside buyer. Half receive additional income from another source.

Without a stick, can we use a carrot? We design a framed field experiment to figure out which of four factors -- harvest size, outside buyer price, framing of the cooperative, or additional income -- is driving the side-selling behavior.

Our partner cooperative sets a stable price (green) during the growing season that usually beats the market price. During the pandemic year, the international price of coffee (orange) quickly surpassed this price. Member deliveries (bars) plummeted. The cooperative's viability was threatened.

These cooperatives use forward contracts up the value chain to fund member services. These contracts are based on an expected volume of member delivery. The problem comes when the producer-members don't market their production through the cooperatives. The cooperatives can't force them to.

Since the market-oriented reforms in 1990s, agricultural cooperatives have emerged as an alternative to state-backed commodity boards as a way to provide price insurance, microcredit, and technical assistance to smallholder producers across the developing world.

But my question — is it possible to get published in an economics journal these days with a matching method? I have seen them used in conjunction with diff-in-diff as an alternate to specifying control variables. It is not always feasible to do an RCT.

Or also this recent paper about membership in a honey cooperative.

onlinelibrary.wiley.com/doi/abs/10.1...

I wrote my master’s thesis on the effects of membership in a coffee cooperative and found that it is used a lot in that space. See Mojo (2017). www.sciencedirect.com/science/arti...