A Question for the Pigou Club
5 May 2008 at 9:07 am Peter G. Klein 9 comments
| Peter Klein |
A few years ago Greg Mankiw coined the term Pigou Club, a label for those (like himself) who advocate higher Pigouvian taxes on gasoline consumption and other high-carbon-footprint activities. Personally, I don’t find the Pigouvian analysis very convincing, in this or the more general case. First, the idea of efficient Pigouvian taxes and subsidies ignores subjective value and the Hayekian knowledge problem. How can government officials possibly choose tax or subsidy amounts that compensate for the actual harm suffered by, or benefit enjoyed by, all possible third parties for all activities generating externalities? The problem is several orders of magnitude more complex than what is typically described in the the textbooks. As a mechanism design problem, it is as difficult as the general socialist calculation problem itself (and you know how I feel about that). Second, the Pigouvian approach ignores comparative institutional analysis altogether. What are the efficiency consequences of establishing, empowering, and funding a government agency to compute and implement Pigouvian taxes and subsidies? Where will the tax revenues go? How will the subsidies be financed? What are the effects of these distortions?
My preference is to treat “negative externalities” as torts, with property titles assigned by the homesteading principle rather than Coasean wealth maximization criteria. (Essentially the Rothbardian view.)
But my main beef with today’s Pigouvians is that they cherry-pick a case here and there — taxes on gasoline, primarily — without fully pursuing the implications of the analysis. If increasing gasoline taxes is efficient, why stop there? What other market failures should the state be empowered to remedy? Here’s my question, specifically:
Please name the activities you believe deserve Pigouvian subsidies. For each activity provide the efficient subsidy amount, explain how this was calculated, and say how the revenues should be raised.
I don’t recall Mankiw discussing Pigouvian subsidies on his blog. Greg, help us out!
Entry filed under: - Klein -, Classical Liberalism, Law and Economics.
Organizations and Markets 
1.
Marcin Tustin | 5 May 2008 at 6:20 pm
How would the right to not be polluted against be homesteaded?
2.
spostrel | 5 May 2008 at 7:32 pm
For any Pigou skeptic, I recommend reading the late Aaron Wildavsky’s (with David Fogerty and Claude Jeanrenaud) “At Once Ubiquitous and Elusive, the Concept of Externalities is Either Vacuous or Misapplied,” Chapter 3 of Cultural and Social Theory (1998). He ties a skeptical treatment of the externality concept to the (Mary) Douglas-Wildavsky “Cultural Theory,” explaining how the four cultural types in that grid-group theory (Fatalists, Individualists, Hierarchs, and Egalitarians) respond to putative externalities (withdrawal, denial, expert control, attack markets). Definitely thought-provoking.
3.
Peter Klein | 6 May 2008 at 9:12 am
Steve, thanks for the cite. Marcin, this approach treats pollution as trespass, so the right not to be polluted against is implied by ownership of land or some other physical space. Homesteading refers to the process by which the property right is obtained (in essence, Lockean first use or legitimate transfer).
4.
Araglin | 6 May 2008 at 3:39 pm
Professor Klein:
To play devil’s advocate for a moment, I wonder if the very knowledge and/or calculation argument(s) you posed against the efficient administration of Pigouvian taxes would be equally applicable to your preferred torts-based approach:
That is, how does one go about determining the right level of compensation for trespass or nuisance after an invasion of justly-acquired property rights has occurred? Is the amount entirely arbitrary (or, is it possible to achieve a rough proportionality between the gravity of the harm sustained and the compensation to be awarded)?
One can’t simply ask the victim how much he would have had to be paid ex ante to have consented to the invasion (because, once the tort has already been committed, the victim has an incentive to exaggerate his appraisal). Especially when it comes to unique and/or unreplaceable goods that are not traded on thick markets in both damaged and undamaged forms, how can there even be a right amount of compensation, given the standard subjectivist arguments against the medieval conception of the “just price”?
One might say, let common law courts work it out, but how would you advise said courts to go about making such determinations?
Thanks,
Araglin
5. A Pigouvian Challenge « The Everyday Economist | 6 May 2008 at 3:54 pm
[…] is dispersed and cannot be communicated to a single mind. Thus, I was delighted to see this challenge from Peter Klein: Please name the activities you believe deserve Pigouvian subsidies. For each […]
6.
Peter Klein | 6 May 2008 at 4:22 pm
Araglin, that’s a very good question. A few responses:
1. In some cases the appropriate remedy may be injunction, rather than damages, rendering the question moot.
2. Even in those cases, of course, there is a “knowledge problem,” as the court must make a determination of liability. I.e., the property rights have to be established through some real-world process, one potentially subject to error, bias, etc. for the reasons you mention. But, the general presumption in this approach is that common-law type remedies, while imperfect, will tend to be better than administrative remedies. Chicago types will justify this claim with arguments about the efficiency of the common law. Even absent those arguments, though, a key point is that the decisions of common-law courts in such cases are local and made on a case-by-case basis, as opposed to the one-size-fits-all administrative remedies of the Pigouvian model.
3. There is also a normative issue. The tresspass approach holds that if A’s smoke damages B’s property, A should compensate B for the damage. Imposing, say, a federal tax on A’s activities, none of which benefits B directly, is hardly an equivalent remedy, even abstracting from the knowledge and calculation problems we’re discussing. If B holds legitimate title to the property in question, then his rights have been violated. Moreover, suppose B values his undamaged property so much (e.g., for sentimental reasons) that no amount of damages will compensate him. In the torts-based approach B can enjoin A to eliminate the trespass altogether, an outcome that isn’t possible in the Pigouvian model.
7.
Sam | 7 May 2008 at 5:13 pm
Re 2: “local and made on a case-by-case basis” could also be described as “arbitrary and capricious”. Whether that is better or worse than an arbitrary Pigou charge is, I think, a rather open question.
Re 3: Surely the argument for Pigou taxes in the case of CO2 emission and the like must be that individual polluters and victims are hard to identify in the case of an atmospheric pollutant.
If I own a river bank, and you discharge industrial waste into the river which converts my fragrant riverside home to a toxic dump, there is an identifiable culprit, an identifiable victim, and the tort system will work just fine. Making you pay a Pigou tax for the right to pollute my riverbank would be rather less just.
If, on the other hand, my property is polluted by climate change caused by atmospheric CO2 or my journeys around town are inhibited by congested traffic, I have suffered (a small amount of) harm, but the culprits are “everyone that burns stuff” or “everyone that drives in my town”. I don’t see any hope for a tort claim sorting that out equitably, even when my claim becomes a class action on behalf of everyone that lives in my town. A court in this case would first have to do the Pigou tax calculation to value the wrong done to me, then assign it amongst all my neighbors without any reliable information as to how much pollution each of them is responsible for.
The answer can’t be “nobody’s harmed too badly, so let’s just pollute”. The tort system isn’t going to help you. The answer could be “we use our democratic structures to decide how much pollution is acceptable, and then regulate polluters” but that again leaves the problem of deciding how much pollution is acceptable, which comes down to almost exactly the calculation that you need to do for a Pigou tax.
Once you’ve set that number, a tax is a more efficient way of imposing the constraint than hard regulatory limits.
8.
Araglin | 7 May 2008 at 5:41 pm
Sam and others,
On the question of how a Rothbardian on property rights might be able to go about obtaining redress when no one culprit’s contribution was a necessary condition of the harm, I think the following paper written by Professor Roderick Long is quite suggestive (although, so far I have seen zero discussion of the paper within libertarian circles):
http://praxeology.net/SmallContributions-REVISED.doc
Cheers,
Araglin
9.
Peter Klein | 7 May 2008 at 9:12 pm
Sam, of course, no feasible system is perfect. The argument for common-law, rather than administrative, remedies builds on an extensive literature suggesting that the decisions of common-law courts are, on average, much better than those made by legislatures. (See Hayek and Leoni for the more general case, and Rubin, Priest, Posner, etc. for the strong-form Chicago efficiency version.) The point is not that individual judges always get it right, but that a decentralized system based on precedent has built-in mechanisms for correcting errors, systematically, mechanisms not present under administrative procedure. Anyway, it is a comparative-institutional argument.
The technical problems of defining vicitims and tortfeasors in cases like air pollution have also been discussed extensively in the literature. No one can specify, ex ante, how the courts would work this out. The historical record suggests, however, that common-law courts were actually pretty good at making determinants of liability even in difficult cases with many agents. (BTW the approach I’m describing applies only to cases of actual property-rights violations, not pecuniary externalities, traffic congestion, etc.)
Finally, I’m enjoying the excellent discussion but so far no one has offered an answer to the question posed in the original post. Can some Pigouvian out there help?