Community Supported Agriculture (CSA) contracts allow consumers to buy claims on a farm’s future production. In turn, the consumer provides working capital to the farm during the growing season. CSA contracts also provide risk management for farmers with limited access to Federal crop insurance by transferring part of the farm’s risk to the consumer. We derive a theory of CSA contract pricing for the two most prevalent types of CSA contracts: yield contracts, in which consumers receive a percentage of the farm’s production, and weight contracts, in which consumers receive fixed quantities. We develop a two-period model in which expected utility maximizing producers and consumers engage in CSA contracting in the first period based on anticipation of yields and spot prices in the second period. Using the model, we generate several testable hypotheses to be explored in future research. Additionally, we present an overview of the data necessary to test the propositions and potential challenges that might arise in related empirical work.
That is the abstract of a new article by my long-time friend and former grad-school colleague Jaclyn Kropp and her coauthor Tom Sproul in the American Journal of Agricultural Economics.
This is fascinating for a few different reasons. First off, from the point of view of novelty, this is exactly the kind of topic that I think agricultural and applied economists should be spending some time working on, given the rise in CSA popularity. Second, from a contract-theoretic perspective, I found it interesting that there were two types of contracts, as highlighted in the abstract: a contract wherein production risk is shared by the producer and the consumer (yield contract), and a contract where all risk is borne by the producer (weight contract). Finally, I like that the paper pays attention to both the theory and, in a more limited fashion, the empirics of CSA.
Having written a dissertation on agrarian contracts, this does not surprise me, but I was unaware that the institution of CSA came in both those contractual shapes. Comparing this to land tenancy agreements, yield contracts for CSAs are reminiscent of sharecropping arrangements, whereas weight contracts are reminiscent of fixed rent contracts.
Where I am less confident is in the latter half of Jackie and Tom’s paper, where they discuss the kind of data required to test those claims. Again, this is informed by the work I have done in applied contract theory where, as one reviewer once put it, “everything is endogenous to everything.” Credible inference on the behavior of economic agents in contractual relationships is marred by three major sources of endogeneity:
- Unobserved heterogeneity among agents (here, CSAs),
- Unobserved heterogeneity among principals (here, CSA subscribers), and
- Unobserved heterogeneity among principal-agent matches (here, specific CSA–CSA subscriber matches).
And that’s without even mentioning reverse causality and measurement error. Perhaps the best way to credibly test the claims put forth by Jackie and Tom would be to work with a handful of producers who offer CSA subscriptions who will let researchers randomize the type of contract offered to each potential subscriber along with an encouragement design to “move” subscribers to accept one contract over the other–but this seems like a lot of work for what represents a small sector of the US agricultural sector.