Most of the six acres of land we had bought with the house was planted with vines, and these had
been looked after for years under the traditional system of metayage : the owner of the land pays the
capital costs of new vine stock and fertilizer, while the farmer does the work of spraying, cropping and
pruning. At the end of the season, the farmer takes two-thirds of the profits and the owner one-third.
—Peter Mayle, A Year in Provence
The excerpt from Mayle's famous travelogue through the south of France describes the sim-
ple structure of typical cropshare contracts and generates a number of interesting questions.
One wonders how long the contracts are, and whether they are written? How often are they
renewed, and why does the owner pay for fertilizer but not for spraying and pruning? Are
the terms similar for other crops? As powerful as neoclassical economics is, it can only
address issues related to prices and volumes of trade—it is silent on how trade is organized.
The transaction cost paradigm, however, is intended to answer and capable of answering
A telling example of the distinction between transaction cost and neoclassical economics
is illustrated on a recent back cover of the Journal of Political Economy ( JPE ) addressing
Mayle's observation. The JPE regularly features excerpts from literary sources on its back
cover, accompanied by titles that cleverly link economic theory to the subject at hand. In its
December 2000 issue the JPE reprinted Mayle's quote with the neoclassical title “Cobb and
Douglas Visit the South of France.” The reference to a constant-returns-to-scale production
function with coefficients on labor of two-thirds and on capital of one-third was too much to
resist. This seemingly trivial example shows how a purely neoclassical approach is incapable
of explaining contracts and organization. As Coase pointed out over sixty years ago, the
nature of the firm hinges on transaction costs. A pure neoclassical model has nothing to say
about organizational issues. The allocation of resources is independent of all distributions
of property rights in the neoclassical model.
The Transaction Cost Approach
The transaction cost approach is related to other economic models of organization. 1 For
example, models found in modern contract theory or agency theory routinely contain
transaction cost elements, either explicitly or implicitly. These models often start with
unobservable effort, actions, or types. In fact, the recent work by Baker, Gibbs, and Holstr om
(1994), Holmstr om and Milgrom (1994), Lazear (1995), and Prendergast (2002)—with its
focus on empirical phenomenon, abstraction from risk sharing, and analysis of multiple
incentives—is hard to distinguish from transaction cost economics. 2 Likewise, the work
of Hart (1995) and the recent work in the economics of property law and contract law
(for example, Ellickson 1991; Hansmann 1996) use models of organization related to the
transaction cost model described here. What then makes the transaction cost approach any
different, if at all?