Ronald Coase has died at the age of 102. I am therefore reposting, with minor changes, the appreciation I wrote a few years ago for his 99th birthday.
In the theory of externalities—that is, costs imposed involuntarily on others—there have been exactly two great ideas. The first, forever associated with the name of Arthur Cecil Pigou (writing about 1920) is that things tend to go badly when people can escape the costs of their own behavior. Factories pollute too much because someone other than the factory owner has to breathe the polluted air. Nineteenth century trains threw off sparks that tended to ignite the crops on neighboring farms, and the railroads ran too many of those trains because the crops belonged to someone else. Farmers keep too many unfenced rabbits when they don’t care about the lettuce farmer next door.
Pigou’s solution—and it’s often a good one—is to make sure that people do feel the costs of their actions, via taxes, fines, or liability rules that allow the victims to sue for damages. Do a dollar’s worth of damage, and you’re charged a dollar.
Pigou endorsed this policy not because it seems fair, though it does seem fair to many, but because it yields, under what he believed to be very general conditions, the optimal amounts of damage. We don’t want too much pollution, but we don’t want too little, either, given that pollution is a necessary by-product of a lot of stuff we enjoy. Pigou offered a proof—now standard fare in all the textbooks—that his policies lead to the perfect compromises, in a sense that can be made precise.
The second great idea about externalities sprang full-blown from the mind of a law professor and subsequent Nobel prize winner named Ronald Coase, who stunned the profession in 1960 by pointing out that Pigou’s argument runs both ways. If you breathe the pollution from my factory, I’m imposing a cost on you—but at the same time, you’re imposing a cost on me. After all, if you lived somewhere else, you wouldn’t be complaining about the smoke and I wouldn’t be getting punished for it.
This insight—so simple once stated, but thoroughly astonishing to the economists of 1960 (I’ve heard tales of this astonishment from several of the participants in Coase’s historic seminar)—means that in a case of externalities, pure theory can never tell you who should bear the costs; you’ve got to look at the specifics of the case. Take those spark-throwing railroad trains. Pigou says: There are too many fires because the railroads don’t care; let’s make them reimburse the farmers for all the crop destruction, and then they’ll care. Coase says: Wait a minute. Often, farmers can prevent fires at very low cost by not planting quite so close to the tracks. True, the railroads don’t currently care about the crop damage. But if you reimburse the farmers, then the farmers won’t care, and you’ll get too many crops planted too close to the tracks. The best way to prevent fires might (or might not) be to grant the railroads complete legal immunity.
And as for that rabbit farmer—the one who lives next to the lettuce farmer and lets his rabbits run wild—Pigou would have insisted that the rabbit farmer cover the damages. Coase is more evenhanded. There are a lot of ways for the rabbit farmer to solve this problem: Put the rabbits in cages, or file their teeth down, or raise a different breed of rabbit, or move away, or switch to keeping geckos. There are also a lot of ways for the lettuce farmer to solve this problem: Fence the lettuce, or spray it with rabbit repellent, or move away, or switch to growing barley. If the rabbit farmer is immune from lawsuits, he’ll have no incentive to implement his solutions. But if the lettuce farmer is routinely reimbursed for lost lettuce, then he’ll have no incentive to implement his solutions. Which outcome is worse? That depends on whose solutions are better. Pure theory can’t answer that question.
How did Pigou—and every other economist in the world—manage to miss this point until Coase came along? According to Coase, it’s because they were obsessed with the faulty notion of “fault”—the idea that if there’s a problem, it must be someone’s fault, and we should begin by identifying that someone. But in the rabbit/lettuce example, who’s really at fault? It’s true that if there were no rabbits, the lettuce wouldn’t get eaten. But it’s equally true that if there were no lettuce the lettuce wouldn’t get eaten. The problem is that rabbit farmers should not be next to lettuce farmers, and when you put it that way, you’re forced to recognize the fundamental symmetry of the situtation.
What, then, should courts and legislators do? Coase had a lot to say about this also, beginning with the observation that it’s sometimes a really really good idea to encourage antagonists to talk to each other. From this beginning sprang the entire intellectual framework usually called Law and Economics.
Coase’s Nobel Prize winning paper is surely one of the landmark papers of 20th century economics. It’s also entirely non-technical (which is fine), and (in my opinion) ridiculously verbose (which is annoying). It’s littered with numerical examples intended to illustrate several different but related points, but the points and the examples are so jumbled together that it’s often difficult to tell what point is being illustrated. I frequently assign my students the task of distilling all of the main ideas into two or three pages, and they frequently succeed. But pioneering work is rarely presented cleanly, and Coase was a true pioneer.
RIP Coase! Here’s hoping the next life (if there is one) is as productive as the first.
I teach public finance and this weeks lesson is on externalities, so the death of Coase will make my classes more interesting and meaningful. I’d like to think that Coase would smile if he knew that even his death provided external benefits to others. No matter how highly regarded Coase is, he is underrated as an economist.
I just adore Coase’s idea of “reciprocal harms” (also, with respect to railway sparks, Coase noted that one source of the problem was the law itself: it was the British Parliament that had given railways immunity for damages over 100 pounds), but the notion of “transaction costs is a bit fuzzy” … anyways, thanks for the link to Coase’s verbose social cost paper … I’m surprised the copyright police at Jstor haven’t messed with it yet
In the case of the rabbit/lettuce farmers, is the conflict really symmetrical though? After all the rabbit farmer damages the property of the lettuce farmer and not vice versa. Would you say that this is exactly counterweighted be the implicit damage dealt to the rabbit farmer for not being able to have rabbits run freely on his property? How do property rights figure into Coase’s argument?
Excellent question!
The classic example in Law and Economics is the fence out/fence in controversy, illustrated in song and story: If your cow damages my crop, what remedy? Cue the song “The Farmer and the Cowman should be Friends.”
In the wide open spaces of the West, a farmer who failed to put up strong fences around his crops had no claim against the owner of the cow. But as more farmers arrived, gradually the laws would change to impose the liability on the owner of the cow to control his cattle – that is, a duty to fence the cattle in. Cue the song “Don’t Fence Me In.” I’m told you can trace the settlement of the American West by the dates that the laws change from Fence Out to Fence In.
But gosh, doesn’t it just seem intuitive that the farmer has the better property rights claim? Intuition is a funny thing. How long ago was it that my right to smoke was deemed intuitive, and your preference to breathe smoke-free air was deemed an unreasonable intrusion on my right?
The “Coase Theorem” states that where transaction costs are low, socially optimal results derive from having clear property rights, even if the rights are assigned arbitrarily. But in a world where transaction costs are high – especially a world when the number of parties to a potential transaction is almost infinite – the allocation of property rights influences the ability to achieve socially optimal outcomes. Consider fee simple ownership of land: If I value consent over use of force, how many parties would I need to contract with to secure “property rights” to a piece of land? Answer: All parties – both now and into an indefinite future. That’s a big transaction cost. Thus the philosophically arbitrary choice to recognize property rights in land – to privilege the claims of one person to exclude others – has huge social benefits. The alternative choice – to recognize everyone’s equal claim to use a piece of land – has the advantage of egalitarianism; it also leads to the Tragedy of the Commons.
The moral is 1) property rights are socially defined, not received from heaven, and 2) people who build their moral philosophies on the idea the property rights are sacrosanct are building on a foundation of sand. (Which is not to say that other moral philosophies have a firmer foundation….)
The most impressive thing is that Mr. and Mrs. Coase were married 75 years.
What about the relative rights of breeders and the childfree? Not only are the spaces of the childfree invaded by the brood, but the childfree are charged high taxes to support and encourage the breeding.
It’s as if the farmers were taxed to support the railroads or the lettuce growers taxed to support the breeding of rabbits.
And the farmer increased in prosperity because the railways allowed him to get his produce more quickly and cheaply and to a wider market.
The lettuce grower prospered because he had a customer right next door, and the rabbit producer had a cheaper source of rabbit feed, since it required no shipping costs.
The simple solution would be for the lettuce farmer to grow and leave part of his crop accessible to the rabbits, by agreement, and charge the rabbit farmer. Or for the farmer to realise if he did not keep his rabbits out of the lettuce, he would drive away a supplier of cheap rabbit-feed.
Or the lettuce farmer could set snares, get a gun and go into the pie making business.
So getting the antagonists to talk works out better than pokey nose Government.
#4 – What if the farmer decides to plant lettuce right next to the rabbit guy? Often we think about who’s harmed in terms of who was there first.
In which case recognizing pre-existing property rights can both preclude conflict / simplify resolution and square with a general sense of fairness.
#5 – I like your summary, but your moral seems to infer too much. I see Coase as being eminently practical and focusing on overall welfare while leaving the equity of the particular transfers to the judges. Surely most of us can think of examples where fairness trumps efficiency. Perhaps slavery is an “arbitrary assignment of property rights”? I think most people feel there is more to laws against murder or even theft than pure social convention. I know you like to suggest there can be no fundamental rights simply because people have to agree to respect them. But supposing there were, all we could do is try to establish and protect them right? Of course there will be conflict but that does not automatically invalidate the principle, rather the purpose of courts is to sift through the gray areas of interpretation. Or is might makes right all there is?
Moving beyond Coase:
Is there an anthropologist in the house? To the extent that certain ideas are universal and “self-evident” or whatever, presumably we’d find them expressed in all societies. So, what do we observe?
I suspect pretty much all societies have some prohibition on homicide. That said, societies seem to give very different content to these prohibitions. Thus, the US is still tampering with the boundaries of when killing in self-defense is justified; see “Stand Your Ground Laws.” Many jurisdictions authorize the death penalty; many do not. Killing during war is deemed different than killing during peacetime. Some societies sanction honor killings; some grant quasi-sanctions to it (e.g., killing someone who is having sex with your spouse is treated as manslaughter rather than murder, because society grants a degree of license to people who react violently to such circumstances).
Similarly, I suspect pretty much all societies have some concept of private property, resulting in some prohibition on theft, yet various societies give different content to these prohibitions. I understand that Germanic leaders regarded kingship as a private family business, and did not expect to raise revenues from the people they governed. Roman taxation was a new – and very unwelcome – innovation to them. And there are various views about the extent to which a person owns land, or owns his or her child.
Slavery has been practiced broadly. And while slavery is now outlawed everywhere, it seems to be widely practiced still – especially in the sex trade. Also, many nations still sanction a military draft and/or prohibit women from obtaining abortions. Moreover, the extent to which I can effectively coerce people into laboring for me is influenced by dynamics that vary by society. I might want to threaten you or your family as a means of securing your labor, but a society that provides greater police protection might impede this strategy. Similarly, I might threaten to withhold food and shelter as a means of securing your labor, but a society that provides a greater social safety net might impede this strategy, too.
Governments establish various prohibitions that have the effect of taking property from people, as illustrated in 5th Amendment jurisprudence (“No person shall be … deprived of life, liberty, or property, without due process of law; nor shall private property be taken for public use, without just compensation.”) If government bans the trade in eagle feathers, government may have effectively destroyed 99% of the value you have in your inventory; are you entitled to compensation or not? The line between just regulation and unjustified “takings” is pretty vague.
I don’t mean to preclude the idea that some universal norms exist. But I suspect they are much narrower than we like to think.
Coase pointed out (I think) that the overall cost where transaction costs are absent will be the same wherever you place the property rights or liability. The cost to the individuals concerned will be very much affected.
Examples. Lettuce farming makes 100 profit. Rabbit farming makes 100. Rabbit damage is 10. Fencing rabbits out of lettuce costs 5 and fencing rabbits in costs 10. If the lettuce farmer is protected, then rabbit farmer either pays 10 to fence in the rabbits, or pays 5 to the lettuce farmer to fence the rabbits out of the lettuce. Without transaction costs, he will pay the lettuce farmer 5. If rabbit farmer is protected then lettuce farmer pays 5 to fence rabbits out. Total output is 195 in either case.
If the fence cost 15, then it is not socially beneficial to build the fence. If the rabbit farmer is liable, he compensates the lettuce farmer 10 for the damage rather than building the fence. If the rabbit farmer is protected, the lettuce farmer just bears the 10 loss from the rabbits.
However, say there were another location away from the lettuce that the rabbit farmer could have used, where he would only make 98: an effective cost of 2.
If the lettuce farmer is protected, in the original location the rabbit farmer would be forced to bear the 5 cost for the fence, so he will choose the other site at cost of 2. Total output = 198. If the rabbit farmer is protected, he will choose the original site and the lettuce farmer must erect the fence. Total output = 195. The social benefit is different.
Of course this is just a case of non-zero transaction costs. When the rabbit farmer is setting up, the lettuce farmer could pay him 2 to use the second location if he knew there woud be 5 damage by the rabbits. This requires knowledge of the damage that will be done – a much more difficult thing that evaluating actual damage.
Coase only talks about what is economically efficient. We may believe that other things are more important. A theocracy for example may value adherence to religious principles more highly than efficiency. As a society, we may also value other things more highly than efficiency. This is fine, as long as we acknowledge it and realise this is what we are doing.
I think we are also a bit mixed up whether certain principles are fundamental or are the best practical way we can achieve other goals such as efficiency. Since transactions are not cost-free, we have no choice but to assign rights and protections. We should be clear about whether we do this to approximate as best we can the efficient outcome, or whether we are doing it in deference to a higher principle.
#11 — I think we’re saying the same thing, Coase does not speak to whether property rights are fundamental or merely conventional in nature. But to me if anything he seems to affirm the case for assigning them and from there shifting control back to people (e.g. even where policymakers can facilitate a solution they are instructed to find the agreement people would reach privately).
On the broader tangent I would just add:
#10 — “To the extent that certain ideas are universal and “self-evident” or whatever, presumably we’d find them expressed in all societies”
Mmm to me this understates the political achievement we call the American experiment (people in power voluntarily subjugating themselves to the citizenry). One could describe this as an exceptional moment of elevating certain fundamental truths. I think you raise some interesting examples where rights may (or may not) come into conflict (eminent domain is a different can of worms), but that does not a foundation of sand make.
Thanks Prof Coase.
I really think that because of some condition of social optimum, or at least social function, establishing fault in a situation is a ubiquitous phenomenon. It seems to be part of a system of ethical debits and credits agents keep with one another, always keeping track of who owes whom favors. Reciprocity and its ability to cohere groups into productive teams with divisions of (social and economic) labor has been under-studyied by economists.
Though Coase’s observation that there are a number of solutions to any conflict of interest, and tradeoffs in adopting those solutions, and that we ought to consider them remains strong.