Beyond Stiglitz: A Dynamic Theory of Share Tenancy

James Roumasset
3:00PM - 4:15PM,
Friday, November 2, 2001
Miller Seminar Room SSB 515


Abstract

Stiglitz's canonical model of share tenancy not only launched principal-agency theory but serves as a prototypical justification of government intervention in a second-best world. The model has never survived an empirical test; indeed, the plausibility of the model has not even been demonstrated. Closer inspection of Stiglitz's theory reveals that his central proposition is incorrect on a priori grounds alone. Moreover, when the model is simulated using Philippine data, high levels of risk aversion are not associated with more reliance on sharing and less on fixed payments as Stiglitz claims. For risk aversion beyond moderate levels, it is just the reverse.

This compelling case against the conventional wisdom demands an alternative theory. We suggest that the primary advantage of sharing is its resiliency in the face of uncertainty. Rent contracts are more prone to breach and therefore less conducive to learning-by-doing and other forms of investment. The resiliency effect is demonstrated in a double-shirking model with Bayesian learning.

The talk will be loosely based on two papers with Mark Deweaver, available by clicking below:

Mark Deweaver and James Roumasset, "A Correction and Prima Facie Test of the Canonical Theory of Share Tenancy"

Mark Deweaver and James Roumasset, "Opportunism without Guile: Transaction costs, Bayesian Decision-Making, and Tenure Choice"

(Adobe Acrobat is required to read)