Lawrence Solum has an excellent post today on Nobel laureate economist Ronald Coase and the famous Coase theorem. The Coase theorem is a principal foundation of modern neoclassical law and economics, of course; indeed, arguably the principal foundation. Solum's post is highly recommended.
Solum's post reminded me that I have been meaning to review Coase's book The Firm, The Market, and the Law. This is principally a collection of Coase's seminal works, although it does contain some useful new material. In particular, the opening chapter is entirely new and shows how a consistent theory of firms and markets, as well as a unique conception of economics and economically-oriented scholarship, runs through Coase's work from the 1930s to the late 1980s (when the book was published).
Coase is best known for two seminal articles. The earlier article "The Theory of the Firm" is the seminal work on the so-called nexus of contracts theory of the firm, as well as an early source for the transaction cost branch of the New Institutional Economics. The nexus of contracts model treats the firm not as an entity, but as an aggregate of various inputs acting together to produce goods or services. Employees provide labor. Creditors provide debt capital. Shareholders initially provide equity capital and subsequently bear the risk of losses and monitor the performance of management. Management monitors the performance of employees and coordinates the activities of all the firm's inputs. The firm is simply a legal fiction representing the complex set of contractual relationships between these inputs. Besides emphasizing the importance of examining the various contracts making up the firm, however, Coase's fundamental insight was that the contractual nature of the firm does not preclude an element of command and control absent from market transactions. If a corporate employee moves from department Y to department X he does so not because of change in relative prices, but because he is ordered to do so. In other words, markets allocate resources via the price mechanism but firms allocate resources via authoritative direction. The set of contracts making up the firm consists in very large measure of implicit agreements, which by definition are both incomplete and unenforceable. Under conditions of uncertainty and complexity, the firm's many constituencies cannot execute a complete contract, so that many decisions must be left for later contractual rewrites imposed by fiat. It is precisely the unenforceability of implicit corporate contracts that makes it possible for the central decisionmaker to rewrite them more-or-less freely. The parties to the corporate contract presumably accept this consequence of relying on implicit contracts because the resulting reduction in transaction costs benefits them all.
Even better known, and even more central to transaction cost economics, however, is Coase's later article "The Problem of Social Cost," which also is reprinted in full here. In that article, Coase laid a critical foundation of modern law and economics - the so-called Coase theorem. The Coase theorem has been formulated in various ways, but one useful statement might be that: "When the parties can bargain successfully, the initial allocation of legal rights does not matter." Suppose a steam locomotive drives by a field of wheat. Sparks from the engine set crops on fire. Should the railroad company be liable? In a world of zero transaction costs, the initial assignment of rights is irrelevant. If the legal rule we choose is inefficient, the parties can bargain around it. Put another way, according to the Coase theorem, rights will be acquired by those who value them most highly, which creates an incentive to discover and implement transaction cost minimizing governance forms.
The Coase theorem has been widely criticized. The second major set of new material in this book is a chapter entitled "Notes on the Problem of Social Cost," in which Coase answers the more serious criticisms. That essay provides a useful intellectual history of the Coase theorem, as well as a trenchant defense of its main claims. One of the less-well informed criticisms of Coase is that he assumes transaction costs are zero. He does not, as this new essay makes clear. Indeed, as Coase points out, the interesting cases are those in which transactions costs are non-zero. In a world of positive transaction costs, however, the parties may not be able to bargain. This is likely to be true in our example. The railroad travels past the property of many landowners, who put their property to differing uses and put differing values on those uses. Negotiating an optimal solution will all of those owners would be, at best, time consuming and onerous. Hence, the allocation of legal rights becomes quite important.
In any event, this is a book that clearly belongs on the shelf of any economically-minded lawyer or legal-minded economist.
UPDATE: Brian Leiter writes:
I've heard it said by various people, including, as I recall, Cooter, that Ellickson's book Order Without Law (1991) shows that the Coase theorem is false (or, more precisely, that the predictions that would be generated from the theorem are false). Comment? I'm genuinely curious about this.
Robert Ellickson’s book “Order Without Law” to which Brian Leiter refers is a deep and very important study of how ranchers and their neighbors in Shasta County, California, resolve disputes -- mainly territorial (such as cattle wandering onto a neighbor’s property, where they do damage. I do not think Ellickson regards his study as disproving the Coase Theorem. I don’t read Ellickson as denying that the Coase Thoerem holds under conditions of zero transaction costs. I read Ellickson as arguing that one important source of transaction costs is learning about one’s initial legal entitlements. For most residents of Shasta County, learning about the true law of property was too costly, let alone the cost of resorting to formal dispute resolution systems. (There is a dispute in the literature as to whether the relevant transaction costs are real or not. In other words, is the ranchers’ ignorance rational or a form of cognitive error?) So they resorted to informal common-sense norms to resolve disputes among themselves. Ellickson then summarizes his dispute with Coase as follows: "Coase overstates the influence of law. His error lies in his implicit assumption that people can effortlessly learn and enforce their initial legal entitlements, and that they confront transaction costs only when they attempt to bargain from their legal starting positions. In a world of costly information, however, one cannot assume that people will both know and honor law." (Order Without Law at 281.)
Actually, in an odd way, I think Ellickson reaffirms the Coase theorem. One implication of the Coase theorem is that, if transactions costs are zero, the actual content of legal rules does not matter. The usual move by law and economics scholars (and it is one I have made myself) is to recognize that transaction costs are often non-trivial. In such contexts, we typically argue that the government should facilitate private ordering by selecting the majoritarian default. As I read Order Without Law, however, Ellickson is saying that if parties face a high transaction cost barrier to learning what the formal rules of law are, the actual content of the rules again does not matter. People will just ignore the law and solve disputes by accepted community norms. The odd way in which this reaffirms the Coase Theorem is as follows: When transaction costs are high, the initial assignment of legal rights does matter, just as Coase predicts. Unlike Coase, however, Ellickson says that the relevant assignment of rights is effected by social norms rather than formal law. After which, bargaining occurs when the majoritarian default norm does not work for the particular parties in question. In sum, its just a question of whether law or social norms provides the relevant assignment of rights.
Finally, I should point you or interested readers to the Cooter-Ulen web site, where they assert: “In a result that comes down somewhere between the Coase Theorem and its critics, Ellickson’s study found that social custom, rather than the law, governed the relationships in Shasta County between farmers and cattlemen.” http://www.cooter-ulen.com/property.htm
UPDATE 2: Put another way, to the extent that the Coase theorem is understood to say that when transaction costs are high the content of formal law matters, Ellickson disproves it. To the extent the Coase theorem is understood to say that when transaction costs are high the content of the relevant rule -- whether law or social norm -- matters, he does not.
I should also note that part of the confusion is engendered by the so-called "Coasean parable" a.k.a. the "Parable of the Cattle." In The Problem of Social Cost, Coase told the following parable: A rancher's cattle wanders onto a farmer's land and harms some crops. Coase posited that, provided the parties could bargain with one other at low cost, the legal rule had only distributional consequences -- i.e., it affected who paid for the damages -- but not allocational consequences -- i.e., it did not affect whether the land was used for grazing or crops. Coase's parable, of course, tracks the situation Ellickson studied in Shasta County quite closely. Ellickson finds that the legal rule had neither allocational nor distributional consequences. Instead of being determined by the content of formal rules, the distributional issues were decided by social norms. Hence, the parable is "disproved" to that extent.
UPDATE 3: Reader JI writes:
For those concerned about Ellickson's "disproof" of the Coase Theorem, I would strongly recommend Daniel Farber, "Parody Lost/Pragmatism Regained: The Ironic History of the Coase Theorem", 83 Va. L. Rev. 397 (1998).
In short, Farber's point is thus: The so-called Coase Theorem was intended as a parody of neoclassical economics, and their assumptions of zero transaction costs. In case this pint was not clear from "The Problem of Social Costs", Coase explained this point at greater length in "Notes on The Problem of Social Costs" in his collection "The Firm, the Market and the Law". Guido Calabresi, among others, also makes this point in "The Pointlessness of Pareto: Carrying Coase Further", 100 Yale L.J. 1211 (1991). See also Jeanne L. Schroeder, "The End of the Market: A Psychoanalysis of Law and Economics", 112 Harvard L. Rev. 483 (1998-9). The list goes on, but back to Farber and Ellickson.
Neoclassical economics, following Arthur Pigou, felt that the ideal world that should exist, in which transactions will be most efficient, is a world without transaction costs. Their (influential) policy recommendations were therefore based on the premise that the government should strive to eliminate transaction, or "social", costs, thereby making the market more efficient by having it emulate, as much as possible, a market without transaction costs.
Coase attacked this position on two grounds. First, he argued that transaction costs are often reciprocal. Take the case of the sparks from the train burning wheat in the surrounding field. The typical neoclassical solution to this social cost would be to tax the train the amount of damage caused to the wheat farmer. But the more economic efficient outcome might be to have the farmer not plant wheat in such close proximity to the train tracks. In fact, forcing the railroad to reimburse the farmer for damaged wheat may encourage the farmer to plant surplus wheat, that would not otherwise be profitable for him to plant, in close proximity to the railroad in order to collect damages from the railroad.
Coase's second, and most important argument, is one of empiricism. He argued that neoclassical economics' "solution" to the problem of social costs, indeed economic analysis as a whole, was/is too theoretical. Pigou's solution to the problem of social costs was based completely on theory, and a one-size-fits-all theory to boot. Coase argued that economics and economic policy recommendations should be based on empirical, pragmatic observation fo the situiation at hand. To fruther the cause of empirical, pragmatic economics, he founded the Journal of Law and Economics, which is mostly devoted to such study (see, for example, Morrison, Winston and Watson, "Fundamental Flaws of Social Regulation: The Case of Airplane Noise", 42 J. Law & Econ. 723 (1999)).
Back to Farber. Since Coase was an advocate of empirical, pragmatic economics, he used the theorem later on dubbed the Coase Theorem as a parody of the neoclassical economic world of supposing no transaction costs. The outcome of such a model is ludicrous - legal entitlements don't matter, and the most efficient outcome is always attained. This is similar to physics without friction. However, since the hard-core neoclassicists were so engrosed in their own theories, they failed to notice the parody in the Coase Theorem. But since they did realize that accepting the Coase Theorem proved how ridiculous their models were, they set out to prove the Coase Theorem wrong. The way to prove the Coase Theorem wrong was to go out and show that in real life situations initial entitlements DO matter. So by failing to grasp the parody in the Coase Theorem (Parody Lost), these econmists inadvertently wound up engaging in pragmatic, empirical economic study (Pragmatism Regained) - which is the type of study Coase advocated - without adopting, or even understanding, Coase's arguments. What these neoclassical economists failed to realize is that it is impossible to empricially disprove the Coase Theorem, since it exists in a decidedly non-empirical world. In the real world, there will ALWAYS be transaction costs (see, for instance, Calabresi, "Transaction Costs, Resource Allocation and Liability Rules - A Comment", 11 J. Law & Econ. 67 (1968)).
One further point. Ironically, a correct understanding of the Coase Theorem can lead one towards libertarianism (Coase) or towards more advocacy of further governmental intrusion into the market (Calabresi). But on second thought, this isn't so ironic. Coase advocated empiricism and pragmatism - tools which are far less wont to lead to forgone conclusions than theory (witness the monotony of the academic eltie).
I think JI's comment is insightful, but I take issue with it in at least one respect. I tend to think that Farber (and JI) overstate the case in treating the Coase theorem as a "parody," at least to the extent they ascribe intentionality to Coase in that regard. Coase himself said in "The Firm, the Market, and the Law" that his intent "was not to describe what life would be like in such a world [i..e, one without transaction costs] but to provide a simple setting in which to develop the analysis and, what was even more important, to make clear the fundamental role which transaction costs do, and should, play in the fashioning of the institutions that make up the economic system." (13) Having said that, however, it is certainly true that Coase objected strenuously to the tendency among later economists to describe "the world of zero transaction costs ... as a Coasian world." (174)
