Monday, September 2, 2013

How not to honor Coase's legacy... acting like anyone that still talks like a Pigovian is too dense or statist to really appreciate Coase. I've already seen a little of it. I'm sure we'll see a lot more. Coase starts the second section of The Problem of Social Cost thus:
"The traditional approach has tended to obscure the nature of the choice that has to be made. The question is commonly thought of as one in which A inflicts harm on B and what has to be decided is: how should we restrain A? But this is wrong. We are dealing with a problem of a reciprocal nature. To avoid the harm to B would inflict harm on A. The real question that has to be decided is: should A be allowed to harm B or should B be allowed to harm A?"
It's true. This is an important question. But in practice most people talk about Pigou because they have thought about that question and answered it or because the parties have resolved it themselves but we think that the resolution came about under duress.

And I promise you, that doesn't mean you need to explain Coase to us.

I told my students on the first day that economics is progressive and although sometimes adversarial (say, Friedman vs. Coase in a seminar room) it is not factional (there is no great Pigou vs. Coase war - ignore people who act like there is).


  1. Now, I'm not one to argue that Coase implies a corner solution for externalities. But one word that you use - duress - gives me a lot of pause. Even if the solution was made under duress, it will still be Pareto optimal, no? The duress card is too easy to play because a) you can't show it is *not* true and b) it has to do with the distribution of rents, not efficiency. Or maybe I'm just being dense and you can cite some random bargaining model which makes it about efficiency again.

    The whole duress thing just drives me insane because it's that type of concept that economists have been and should be trained to throw out as self-serving cheap talk unless there is really really good evidence otherwise. I don't think there is.

    1. It's not an easy one to apply and it's going to be the sort of thing where it will always be contested when applied. But this is all I mean by it: let's say we think that B just should never have to pay off A to get him to refrain from imposing a cost. It's just a matter of justice. If we saw B paying off A we might be tempted to say this is a private solution so there's nothing wrong with it. However, if B pays off A because B knows that real justice isn't in the cards.

      In that sense, yes, it's Pareto optimal. But I know of no rule that says that economists are obligated to like Pareto optimal solutions.

      When we invoke any kind of welfare economics as a foundation for normative claims we are ALWAYS bringing assumptions about standing to the question. There are usually fairly specific and quite reasonable assumptions about standing that would lead someone to invoke Pigou even if they like Coase. That's what I'm saying here.

    2. Daniel, one of the primary points of the whole analysis is the symmetry implicit in the entire concept of an externality. If what you care about is wealth max, then you shouldn't care about the moral question of whether I am morally right to force someone not to pollute, or whether I am right to force someone to accept my pollution. In the world of positive transaction costs, you assign the right in whatever way that cuts down on transaction costs because you want them to bargain to the efficient solution.

      Once you are AT the efficient solution, everything else is just a transfer. If you want to fix inequality or whatever caused by moving to the efficient solution, there are better ways YOU know of for fixing that problem.

      In other words, if assigning property rights in a way that minimizes transaction costs has allowed the parties to bargain to a solution, there is a *presumption* that it is a more efficient arrangement than would have taken place via Pigovian mechanisms.

      And again, if you find something unsavory or unjust about the wealth maximizing outcome, there are more efficient ways of addressing that than to arbitrarily invalidate the contract.

    3. Right, so if you don't agree with the symmetry - perhaps for moral reasons - there is still an important role for Pigou along side Coase.

      This is fine: "In the world of positive transaction costs, you assign the right in whatever way that cuts down on transaction costs because you want them to bargain to the efficient solution."

      IF that's the question you're interested in. And who says it should be?

      It's easier to see the point with a more extreme example, like rape (as I mention below). There is absolutely no doubt that rapists derive real utility from rape. There's nothing debatable at all about this. There's nothing illegitimate about saying "I don't want to talk about assigning property rights in a way that reduces transaction costs and compensating low-income rape victims after the fact with a transfer - I simply don't care about the rapist's utility. I don't care about symmetry in this case".

      That's an extreme case, but in most Pigovian cases that is how people are approaching the question. There are limits. Polluters should be able to produce something. But there's a point where I simply don't recognize their standing. In that world - in the normatively non-symmetrical world - Pigou is fine, particularly if we find something inherently offensive in a rape victim paying off a rapist.

      Of course sometimes we live in normatively symmetrical worlds and sometimes we live in normatively non-symmetrical worlds. That's why Pigou and Coase are both important as far as I'm concerned.

    4. Assume zero transaction costs. In that world, there would still be severe punishments against rape because social costs of rape are much higher than social benefits - the utility of the rapist. I don't think a Coasian bargain is appropriate, but the solution isn't Pigovian, either. The analogy fails on its own grounds.

      This isn't a matter of who should have "standing;" it's a matter of what will maximize net benefits. Too much bullshit is allowed to get by because people treat these issues as lexicographic, and your framing is exactly an example of that. It's what puts us in situations where we ignore the social benefits of allowing trees to be cut down and re-grown because some endangered beetle lives there. We do stupid things because we don't grant the consumers and lumber producers standing in the issue.

  2. Who is the person who prompted you to make this post, Daniel Kuehn? Could you please link it to us?

  3. I am paying Daniel $83 on the side to not link to my comment.

  4. Daniel, one of the main reasons I read your blog is for posts like this. Before encountering you, I never really thought much about how annoying a patronizing post (on Coase, subjectivism in economics, micro foundations in macro, etc.) must be.

    Having said that, I think you often minimize the element of truth in these complaints. Take a different example: cardinal utility functions. When an Austrian complains that mainstream economists believe in cardinal utility, it's standard to pull out Hal Varian or Mas-Collel and show that no they don't.

    But, how many actual PhD economists who teach undergrads understand that stuff? I am quite sure that half of my cohort at NYU thinks utility functions map to cardinal units of psychic utility, even if (in the first year) they had to do a problem on one homework set about utility representation theorems.

    1. I may be overly optimistic, and you may be overly pessimistic, and that probably depends on the question at hand, it's true. One of the justifications I like to whip out to convince myself is that I haven't had the illustrious educational career that you have had, and I got it and my cohort seems to have gotten it. So I have a hard time believing people at NYU don't get it. Actually the higher caliber the PhD program, the more time you spend mapping sets and preference relations onto continuous functions and working with the math that maps one onto the other. So I'd think they digest that more.

      Then again, to play my own devil's advocate, one of my pet peeves is when people treat welfare economics as if its positive when it's actually normative. That's this whole point above of asking why exactly I'm obligated to give a damn about Pareto optimality. Sure it's a nice benchmark and a nice way of talking about the sense in which economies are efficient, but must I import it as my own value system? Surely not.

      So somehow I need to square my sense that most economists understand ordinal vs. cardinal with my sense that they take a far too objective attitude towards welfare economics. It seems like that's not entirely consistent.

      One solution might be that they compartmentalize. They understand what's going on with cardinal utility and ordinal preferences just fine but they like speaking authoritatively about fundamental questions of social welfare - they like having an answer to the "ought" question - so they bracket off what they know when they're finding optimal positions.

      And speaking of bracketing things off, the other point I raise with this that I do stand by is that in some cases the simplifying step of assuming cardinality is fine if, for example, you are teaching a sophomore the last econ class he'll ever take and it's an intermediate micro class. In that case I would personally rather he get marginal rate of substitution and the logic of optimizing behavior with an incorrect cardinal utility assumption than spend time on the ordinal/cardinal mapping and not have him grasp the optimization as strongly. So we tell these useful fictions to students. That's of course a different case than the population of PhDs that you're referring to.

    2. On this post I don't think I'm being too optimistic. Questions of justice and right and wrong pervade externality discussions. This question that Coase treats as an open question in the article is not an open question for people. It's not that they miss the logic of the Coase theorem, it's that they reject the very idea that the swimmers downstream should be expected to pay the power plant as a moral question. Of course they COULD pay the power plant. A rape victim could also presumably pay her rapist not to rape her. After all, the rapist is harmed by not being allowed to rape, right? But nobody's going to claim that you don't understand Coase if you were to rule that out from the get-go.

    3. I think it depends on whether the power plant or the swimmers were there first. Standard law & economics says to look for the least-cost avoider.

  5. Great post- Coase is definitely one of the economists who created, as Peter Boettke calls it, "an invitation to inquiry". There are few dogmatically settled beliefs when one reads Coase. His writing does not show that property allocation does not matter- as a naive Coase Theorem suggests- nor that initial allocation is everything and actors cannot solve externality problems- as a naive Pigovian theory suggests. Instead, the situations when improvements are "organic" or require a visible hand of the state depend upon the exact circumstance. This is what I see as the stress in Coase and one of the points that is most important.

    Between "The Nature of the Firm" and "The Problem of Social Cost", I see the foundation for much of the work of Ostrom, Williamson, Demsetz, Alchian, and countless others. People are not omniscient, yet see solutions that the pre-Coase theory denies. Here I am thinking of Ostrom's work on common pools. Yet, people are self-interested, meaning they care more about themselves and those closest than strangers, and make decisions in their own interest. Therefore, the definition, often by the state, of property rights is extremely important in the real world. The enforcement and possible alteration of property rights can have a real improvement in the world- ignoring extreme subjectivism. Here I am thinking of Demsetz and Alchian's work as following Coase.

    Any economist who can inspire one of those authors is noteworthy. He was truly one-of-a-kind and I'm glad to see a post explaining the complexity of Coase. (Although I do believe, anecdotally, that economists tend to error on the Pigovian side)

    1. Excellent thoughts!

      I took a look at the Wikipedia page on the Coase Theorem and - without defending every single thing on there because I can't remember every single thing on there - one part I liked was that they listed four different institutional and transaction cost arrangements - variations on a single situation - that Coase would have considered "rational". I think this illustrates your point well. Coase wasn't talking about one state of the world and one right answer - he was introducing transaction costs and bargaining and thereby making the whole suite of potential outcomes to discuss much richer.

      I like the connection to Ostrom, of course.

      Part of the us vs. them attitude, I think, comes from the caricature of what a "state solution" and a "market solution" is. If you view states as some kind of exogenous imposed force or if you view markets as some kind of atomistic non-cooperative free for all (both wrong), then you tend to develop a very factionalized sense of these discussions.

      If you view it as a spectrum of voluntary institutions - from transactional to something more collective - then Pigou and Coase and Ostrom all fit together much more effortlessly.

      That last point is more stylistic and interpretive, of course.

    2. I think the "state solution" and "market solution" point you made fits it Coase's thinking. For him, abstract theory was not what economics was about. Instead, economics was about understanding the world as it exists now; I can appreciate this focus without necessarily supporting his entire method.

      In the world today, the state and the market are constantly interacting and driving each other. I would even argue that Buchanan/Tullock is growth on this understanding of both as pivotal to the process. Yes- a theorist can explain how an all state or all market solution would be better, maybe even optimal. Yet, that is not the world we live in, nor is it likely one of our immediate opportunity sets. At this point, Coase found it more beneficial to analyze a world with both markets and a stat. Both driven by real people who are neither all sinner not all saint- which is good for us students looking at economics as an inquiry, not a solution to learn.

    3. I am not, btw, making a "mixed economy" point, although that of course is always important to keep in mind too. I'm making the point the if we think about non-market collective solutions to problems, the state is not in some categorically different category from non-state institutions. Humans get together and build states in the same way and for the same reasons that they build non-state institutions (for example, of the sort that Ostrom spent a lot of time thinking about).

  6. "Humans get together and build states in the same way and for the same reasons that they build non-state institutions"

    e.g for profit


All anonymous comments will be deleted. Consistent pseudonyms are fine.