McCloskey on the Coase Theorem

The Coase Theorem is an important part of economics research, agricultural or otherwise. The standard definition of the theorem is that in a world with zero or very low transactions costs, bargaining over a disputed right between two parties will lead to an efficient allocation of resources regardless of which party is legally awarded the property right. (If you are familiar with the Coase Theorem, you can skip the next paragraph.)

A classic example (which is found in Coase's paper "The Problem of Social Cost" is that of the railroad and the farmer. As trains traveled down the track in the 1900s, they emitted sparks that potentially could set fire to farmers' crops. If this were to happen, who should pay for the lost crop? In a world where transactions costs (that is, costs associated with acquiring information, monitoring the other party, or executing the trade) are zero, it doesn't matter who a court decides is responsible. If it's less costly for the railroad to put on a spark-catching device, the farmer will pay the railroad company to do so. If it's less costly for the farmer to plant his crops farther from the railroad, the railroad company will pay the farmer to do so.

Of course, we don't live in a world of zero transactions costs, so it does matter to whom the court grants the property right. Transactions costs also make it difficult for governments to figure out how to solve the problem. If the government decided it wanted to make farmers reduce their planting to solve the problem, there would be significant costs associated with determining just how far to restrict planting away from the tracks. In the opposite case, the government (that is, the taxpayer) would have to incur a cost to determine precisely which device the railroad should use to contain the sparks. These costs don't disappear in the cases of tradeable permits in fisheries or pollution. The government still has to determine the extent and details of these artificial markets.

It seems strange, but Coase himself maintains that the theorem (as described in the first two paragraphs of this post) is not about his work. He says that it's really George Stigler's theorem, but that Stigler was kind enough to name it after him. This is where McCloskey comes in. Her Summer 1998 column in the Eastern Economic Journal makes some bold claims about the Coase Theorem and, I think, provides some interesting insights, which I reproduce below.

Her interpretation is at odds with the way most interpret the Theorem. Her interpretation, which she says is Coase's (and I have reason to believe it is... Coase spent a lot of time arguing against what he called "blackboard economics"), is a caution to people who take the typical form of the theorem as an argument for "markets are always best" or "government solutions are always best."


She puts her point most succinctly in an e-mail conversation she prints in the column:
(McCloskey's own comments on the conversation are in brackets.)

Weisskopf: I have been reading Bob Cooter's piece on the "Coase theorem" in The New Palgrave. It strikes me as ... quite consistent with the view that whether externalities
can best be addressed by state action or private market transactions is
always an empirical issue. I take it we agree on this. [Yes, we do, and with Ronald
Coase.] I was just wondering ... whether you would include Cooter among the
dozen or so in your corner. 

McCloskey: No, Bob gets it wrong, with the other N - 12 folks. He gets it wrong also in
his book with Tom Ulen, Law and Economics (p. 5; in a Stiglerian vein they call
the "Coase theorem" = Adam Smith's theorem a "remarkable conclusion," echoing
Stigler). Dick Posner gets it wrong in his book, The Economic Approach to Law
(4th ed., pp. 8, 22, 285, most egregiously on p. 636, and then throughout; he gets it
right once, on p. 50f). These excellent scholars get it wrong for the same reason.
They do not appreciate the doctrinal history from which Coase was writing [and
which Medema and Richard Zerbe illuminate in a recent article about Coasean
economics (1997)], and in particular they do not appreciate Coase's distaste for
blackboard economics, Pigou/Samuelson, and easy solutions to difficult problems.
The so-called "Coase theorem" is framed in a Benthamite spirit wholly obnoxious
to Coase. 

Weisskopf: [I here put words into Weisskopf's mouth]: Can that be right? They all got
it wrong? 

McCloskey: I don't expect you to believe it, since no one does except the Blessed Dozen,
and Ronald Coase himself. In Coase's introduction to his little collection of substantive articles, The Firm, the Market and the Law (1988), you can find statements
that sound like the "Coase" theorem, but more that sound like the Coase
theorem, no scare quotes, the actual theorem. If you read all of Coase's works you
will see that the actual theorem, besides being the prose meaning of the last few
pages of "The Problem of Social Cost" (which most economists appear not to have
read), is consistent with the rest of Coase's views. Coase's second published paper,
the famous one in 1937 on the firm, says that transaction costs matter. So have all
his papers, over and over again. The version of the "Coase" theorem that Stigler
popularized says the opposite. 

Weisskopf: If I understand your point, the true Coase theorem implies that one cannot
in general efficiently internalize an externality by taxing/subsidizing whoever is
generating the negative/positive externalities, because (in light of transactions
costs) this would generally not result in the right to the resource affected going to
the person who values it the most. 

McCloskey: Yes, that's right. Another way of putting it, as Coase does, is to point out
that "whoever is generating the externality" is not a meaningful notion. Coase has
the famous, and confusing, railway/farmers example: who "causes" the burnt fields
of corn, the railway which makes the sparks, or the farmers who plant imprudently close to the line? [The question arises naturally in the common law; on a
blackboard by contrast you can't see it: equalize marginal private and social cost,
that's all]. A better example is noise pollution around airports. We usually think
of the airplanes as thè. cause. But wait. Suppose that there were no ears close to
the airport. (Or that the ears were easily protected from the noise.) In that case the
noise would be harmless, and it would be silly to curb it. So the presence of ears is
just as much a "cause" as the vibrations in the planes' motors. Where then should
the Pigou/Samuelson tax be placed? 

Weisskopf: Rather than implying that the state ought to get out of the business of
dealing with externalities, doesn't this imply that the state ought to concentrate on
defining and transferring property rights instead of taxing/subsidizing [in the
Pigou/Samuelson way] ... getting the entitlements right rather than getting the
prices right? 

McCloskey: Yes, it certainly does, which is why the notion that Coase justifies doing
nothing is wrong. Coase, who emphasizes transaction costs, says in effect: "Because
the transaction costs are high it matters where the resources start out.
Pigovian taxes are not going to get the right solution, except by happy accident.
Face up to the hard facts of life: transaction costs are high." [Which is another
way of saying that Second Best sometimes applies.] Amazingly, economists have
understood him to be saying, "Transaction costs are low. Relax." [Which is another
way of saying that First Best always applies.] But there's a further point that Coase makes in every paper he's written, which may explain why people such as George Stigler think he is saying that we do already live in the best of all possible worlds. The point is that "getting the entitlements right" is devilishly difficult with the governments we actually have. Coase is forever saying that this or that proposal for a public policy entails knowing things that no economist can in fact know. He claims, with considerable empirical evidence, that in many cases laissez faire will be in practice better than what we will get from actual governments - though neither is perfect (we live in a second-best world, that is, a world of transaction costs). The methodological point is that Coase does not claim to have proven laissez faire on a blackboard. He says in effect, "If you look at the FCC or the lighthouses or the law of liability you see that governmental attempts to guide things minute-by-minute - as you say, Tom, 'getting the prices right'- don't work very well. Maybe it's better to just deal the cards and play. But in this veil of tears there are no guarantees. It may not work like some curves you have drawn. Life is hard. Knowledge is scarce. Grow up and admit that you can't extract policy from a couple of lines on a blackboard."