Understanding the Coase Theorem

Last year on my old blog, I stirred some controversy with a flippant post on the Coase Theorem. The Coase Theorem states that even in the presence of externalities, the market will reach an efficient solution as long as it’s possible to trade those externalities without transaction costs. Supposedly, this happens regardless of the initial assignment of property rights: it doesn’t matter whether the law gives me the right to peace and quiet, or my neighbor the right to play obnoxious music all day long. As long as we can write a contract about it, we’ll reach the efficient outcome.

This raises some natural questions, as I mentioned in the original post:

Suppose that we inhabited a world where farmers had no property rights against encroachment by cattle. In this world, the first thing I’d do is start a company (preferably with an nefarious-sounding name like Multinational United) that raised cattle specially bred to ravage cropland. I would locate these cattle near a large number of farmers and announce that each they had to pay me $100,000 to avoid being overrun by the herd. For most farmers, this would be worthwhile, and I’d make a bundle of money.
Of course, this would be inefficient: if any farmers didn’t find it worthwhile to pay the $100,000, their farms would be destroyed. And even if we assumed that evil Matt had perfect information and could charge a special rate to each farmer to make sure that it was worthwhile for everyone to pay, this would be inefficient in a different way. Payments that varied according to ability to pay would effectively be a tax on productive assets, which in the long run would discourage the accumulation of those assets.

These sound like very distorted incentives! How does the Coase Theorem get away with claiming that we’ll reach the “efficient” outcome? (Incidentally, cattle herds were the example in Coase’s original article, The Problem of Social Cost.)

That’s because the Coase Theorem would say that all the farmers, before spending time or money building up the capital in their farms, would have signed contracts with me where I pledged (in exchange for side payments) not to bother them with mutant, crop-destroying cattle.

Of course, there’s still an inefficiency: since I get side payments in exchange for the pledge not to bother people with mutant cattle, I have an incentive to undertake costly investments in socially worthless technologies that threaten cropland (which are necessary to make my threat credible in the first place). How does the Coase Theorem resolve this? Well, I have to sign a contract with every farmer in the world before I spend any time breeding mutant cattle, where I pledge to refrain from developing the technology in the first place (in exchange for further side payments, of course).

But to be in a position to pull off this deal, I have to be a smartass who is able to credibly threaten that he will a develop new technology with the potential to destroy farms across the world. My parents (who will benefit from any wealth I receive) are now inefficiently incentivized to raise a smartass who offers no value to society. The Coase Theorem requires that they sign a contact with farmers across the world while I’m still a baby, pledging to raise a normal human being in exchange for side payments. Of course, this creates an (inefficient) incentive to make a baby who shows even the potential to become an evil rent-seeking mastermind. And so on…

The Coase Theorem is only really “true” in one setting: at the beginning of the universe, everyone who will ever exist gets together and writes a contract that covers every contingency that will ever conceivably occur.

In any other situation, where we negotiate Coasian contracts after economic decisions have already been made, anticipation of the fact that your returns will be bargained away in future contracts makes you less willing to undertake investments in the first place. That’s inefficient. And the Coase Theorem offers no way around it.

Edit: As a link to this post points out, Coase himself might have agreed with the sentiment here: he viewed his result as a consciousness-raising device to emphasize the centrality of transactions costs and noncontractibility, not as some kind of proof that Pigouvian remedies are never necessary. He said as much in his 1990 Nobel lecture: “However, I tend to regard the Coase Theorem as a stepping stone on the way to an analysis of an economy with positive transaction costs.” Unfortunately, I think that the spirit of Coase’s interpretation is often forgotten, and the theorem is taken to imply much more than it can legitimately offer.



Filed under micro

21 responses to “Understanding the Coase Theorem

  1. “And the Coase Theorem offers no way around it.”

    perhaps yes, perhaps no.

    Arrow-Debreu equilibria are implemented by continous trading of long lived assets, one asset for each source of uncertainty. So if there were only N sources of uncertainty you only need N assets (maybe it was N+1).

    In that case the possibility of your parents being a doctor evil and raising you to be a mini-me must be spanned by that uncertainty and thus the payoffs by those assets.

    You don’t need side bets you’d need long-lived transferrable “bonds” that pay their coupon if you behave. One can argue that such things exist, the “bond” in question is the social bond of living in this free society where the rule of law holds. The coupon is the benefit thus derived and is passed on through generations.

    New family lines are new nodes of uncertainty and also coincide, automatically, with the issue of a new bond. Dynamic completeness…

  2. Just to clarify, what I’m saying is that well designed government is the market’s efficient solution.

    The coase thereom implies we should have government.

    • Wayne

      That’s a jump. I would say The Coase Theroem implies (or rather, reduces to the point) that the assignment (or rather, distribution (or rather, homesteading outcome)) of property does matter.

  3. Darf Ferrara

    I don’t completely agree with Adam, but I think that he captures the essence of the solution. It gives legislatures, and especially judges a way to think about how to assign property rights that allows for economically efficient bargains. is an excellent book that gets into how Coase is used in a legal setting.

  4. Of course, I would like to say that your example of an incentivised (and nefarious) mutant cattle breeder is ridiculous… but unfortunately I’ve had personal experience arguing with a libertarian friend on something quite similar. (In particular, that it would be “efficient” for people to farm animals and threaten to kill them in inhuman ways lest some wealthy animal lover(s) paid them not to).

    Let me say, though, that you are giving Coase short shrift. His theory may have been hijacked by free market environmentalist ideologues, but he was much more nuanced on these matters himself. In fact, his argument was largely about considering which type of arrangement (e.g. Pigovian vs market negotiation) would allow society to sort out externalities at least cost in different circumstances. Read the final paragraph of the very paper you quote if you don’t agree.

    Final point on your particular example. You would need to assume that the breeding mutant cattle is the most profitable activity available to you. Given that you own all the grazing land (I’m not sure how you could separate the property right against invading cattle otherwise?), this seems unlikely.

  5. Donald A. Coffin

    Two comments.
    First, as I recall, Coase assumed that income effects of the transfers were negligible.
    Second. Many years ago, I read an article published in a law review (Temple’s?) in which the author noted that one alternative to Coasian baergaining was that the owner of the cattle and the owner of the corn farm would be circling each other through the corn fields with their shotguns, waiting for a clear shot. Or, Coase only even potentially “makes sense” when alternatives other than bargaining and side payments are somehow excluded.

  6. JWalker

    Remember that the Coase Therom works in the absence of transaction costs. However, externalities almost always have transaction costs associated with them. Such as the time and effot the farmer has to spend to protect his farm outside of farming (the market activity). Then the key becomes haw to minimize transaction costs. This is why in many cases regulation is better than litigation. The costs are lower. Coase (like general equilibrium models out of Chicago for example) assumes a perfect world. We need to think about the imperfect world. Mark Thoma in his blog talks about the Peter Temin interview on the subject of general equilibrium macroeconomics.

  7. Joe Beaulieu

    A hidden assumption that you don’t mention is wealth effects. Coase assumes that the allocation of property rights, which makes the receiver richer, doesn’t matter for the efficient outcome. This can be very important in real world situations. Consider Iceland, who distributes fishing rights based on catches in the mid 1980s. Iceland allows these fishing rights to be traded, but it still ends up with a fishing industry that relies on small-scale fishermen and is probably inefficient from a productivity standpoint. I think it reaches that equilibrium because it allows guys to be their own boss and fish as they like, even though they could sell the permit and do something else and probably be richer (money wise). Make them buy or rent the permit, and I’ll bet you have a more concentrated industrial fishing industry.

    • The theme of this post is that even if wealth effects don’t matter for reaching the efficient outcome in period 2, in period 1 the anticipation of period 2 wealth transfers will lead to a distortion in economic decisions, as everyone angles for a greater share of the period 2 surplus. The only way you can circumvent this is by writing a giant contract in period 1 that covers every conceivable contingency for the rest of history. (This is in some sense just a restatement of classic papers on the “hold-up” problem: see, for instance, Grossman and Hart who discussed these issues in another context.)

      The Iceland example is very interesting—I think the Coase Theorem would say that the distribution of fishing rights today should be efficient, but that there might have been overinvestment in fishing in the 1980s due to anticipation of the distribution mechanism. Of course, one reason the distribution of fishing rights today might not actually be efficient is that people anticipate a revision to the allocation mechanism; maybe at some point Iceland will decide to give more permits to small-scale fishermen operating in the mid-2000s. I suspect that the possibility of such a contingency is built into the price.

  8. e

    My understanding was that Coase suggested that we should make an effort to define property rights as clearly as possible and produce institutions that ease negotiations within the range of reasonable allocations of property rights. As opposed to idiosyncratic or worse discretionary ways of resolving externalities. So although you seem right to me, I don’t think you really dent Coase’s value as a heuristic, and frankly although I hear a lot of complaints about libertarians justifying unjust allocations of property rights with Coase, I’ve never actually seen a libertarian do so.

  9. Scott Sumner

    Matt, Your description of the Coase Theorem is inaccurate. The Coase theorem is about transactions costs, it does not say environmental regulation is not needed. Coase never said that, and it was never taught that way at Chicago. I’m afraid you’ve attacked a straw man here.

    I think the Coase theorem is most useful in cases where people believe there are externality problems, but because there are no transactions cost problems, no GOVERNMENT regulation is required. Cases like indoor smoking in the workplace. There may need to be regulation, but the employer already has the right incentives to produce the optimal amount of regulation. That’s the sort of case where Coase’s theorem can lead to interesting policy conclusions.

    • Well, as most people interpret it, the Coase Theorem says that we’ll reach an efficient solution (albeit maybe one with bad distributional consequences) in the absence of transactions costs. Since there are no obvious transaction costs in my example, I don’t see why it’s not a valid critique.

      Implicit in my example, of course, is an extreme kind of transaction cost: non-contractibility (which is equivalent to a transaction cost of infinity). My point is that the implications of non-contractibility need to be taken very, very seriously: the question isn’t only whether a current externality problem can be settled by a contract (even a zero transaction-cost one), but also whether anticipation of this contract and the resulting transfers of surplus will cause agents to change their decisions in previous periods—and that it is wildly implausible that a monster contract containing both these decisions and the future externality will be drawn up at date 0.

      “Cases like indoor smoking in the workplace. There may need to be regulation, but the employer already has the right incentives to produce the optimal amount of regulation.”

      I agree that government regulation is not necessary in this case, but this seems like a rather weak implication of the Coase Theorem, if the theorem is even needed at all. In my mind, the Coase Theorem deals with resolving a genuine externality (i.e. one party’s decision affects another party in a way that has not been previously agreed upon in a contract), whereas a decision about indoor smoking in the workplace is a decision about the conditions of a relationship that is always covered by a contract to begin with.

  10. wintercow20

    Isn’t your illustration enlightened by Coase? It takes two to tango. Common law recognizes that “coming to the nuisance” is not grounds for compensation, and so it would seem that “bringing the nuisance” would be treated similarly. That of course is not the Coase Theorem in operation, but is a result of courts recognizing an insight of Coases – that “solutions” to externality problems require the lowest cost avoider to handle the problem. That is all I read the Coase Theorem as saying. And the two implications of this view are these:

    (1) That a pigovian tax is a special case of the Coase Theorem when certain conditions hold

    (2) Allowing/encouraging negotiation in the presence of certain regulatory institutions can actually make outcomes less efficient.

  11. Isn’t your illustration enlightened by Coase? It takes two to tango. Common law recognizes that “coming to the nuisance” is not grounds for compensation, and so it would seem that “bringing the nuisance” would be treated similarly. That of course is not the Coase Theorem in operation, but is a result of courts recognizing an insight of Coases – that “solutions” to externality problems require the lowest cost avoider to handle the problem.

    The aspect of the Coase Theorem I’m addressing here is the claim that it doesn’t matter who has the initial right: regardless of whether the polluter has the unlimited right to pollute (unless he has signed it away in a contract) or the victims of pollution have the right not to suffer from dirty air (unless they have ceded it in a contract), they will bargain and reach the right outcome. In particular, if the polluter has the default right, then the potential victims of pollution will sign a contract where the polluter cuts back in exchange for side payments. My point is that this creates an inefficient incentive to be a polluter in the first place—and the way that the Coase Theorem would hypothetically deal with this problem (by requiring everyone to sign a monster contract before any economic decisions anywhere in the world are made) is unrealistic.

    It sounds like you’re saying that courts will treat parties “bringing the nuisance” (e.g. building a polluting factory close to a major city) as the ones responsible for compensating others, not the other way around. I absolutely agree, and when it is restricted to this setting the Coase Theorem is much more practically useful. But the claim it makes (one that is cited surprisingly often) is more ambitious, and I think it’s valuable to recognize its failings in a world with even minor noncontractibility.

  12. Nick Szabo often promoted a critique of the Coase Theorem (more specifically, David Friedman’s use of it) based on similar logic of bringing a nuisance for extortion. I never got David Friedman to respond, but I did get one from Johnnie Lin.

  13. Scott Sumner

    Maybe I overreacted. Coase won a Nobel prize for his lifelong study of transactions costs. Indeed Coasian economics is the study of transactions costs. So I get frustrated when people think Coasian economics is the denial of transactions costs. I think all environmental economists and law and economics-types understand this, but lots of people who just have a vague recollection of the Coase Theorem from an undergrad econ class probably due suffer from the misconception you mentioned.

    Certainly the extortion possibility is well understood; I recall that being discussed in the 1970s when I learned the Coase Theorem. Much of law and economics is about designing optimal property rights systems to minimize that possibility.

    I’m glad you agree on smoking; if only my commenters could have seen the point when I made that observation in a post. They mostly disagreed.

  14. Jim Glass

    Coase said he didn’t intend the Coase Theorem as a theorem — Stigler presented it as the Coase Theorem — and that he did mean it as a counterfactual to highlight the importance of transaction costs.

    Quoting him at Reason:

    Q: Could you state the Coase Theorem? How do you explain it to people?

    Coase: … Take the case of a newly discovered cave. I say, whether the law says it’s owned by the person where the mouth of the cave is or whether it belongs to the man who discovers it or whether it belongs to the man under whose land it is, it’ll be used for growing mushrooms, storing bank records, or as a gas reservoir according to which of these uses produces the most value. The law of property determines who owns something, but the market determines how it will be used.

    It’s so obvious to me that I couldn’t understand the fuss. All it says is that the people will use resources in the way that produces the most value, that’s all. I still think it’s an obvious point. You wouldn’t think there was a need for a Coase Theorem, really.

    But the people at the University of Chicago thought it was an error … I don’t know whether you’ve had a conversation with Milton Friedman, but an argument with Milton Friedman is a pretty strenuous affair. He’s very good. He’s very fair, but he doesn’t let you slip up on anything … when at the end of whatever the time was — say, an hour — I found I was still standing, I knew I’d won. Because if Milton can’t knock you out in a few rounds, you’re home…

  15. Scott Sumner

    This new article by Demsetz is quite interesting, and seems at least loosely related to your argument:


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s