Hal Finney wrote:
>Coase's theorem basically says that you can solve externalities by
>introducing property rights. ...
>(These are two sides of the same idea: there is a "right to emit sparks
>that may cause fires", and in one case the railroad starts off owning it,
>and in the other case the farmers start off owning it.)
>Either way, the result is the same, according to Coase. Neglecting
>transaction costs (reasonable if there aren't too many farmers to
>negotiate with), you get a socially optimal balance between spark
>suppression and fire fighting. The distribution of wealth is different
>... (I should mention that my understanding of Coase's theorem comes
>from secondhand sources; I have not read his original work.)
Coase's original paper gave no theorem, and no one else has bothered to write one down because it's almost a tautology: if you assume negotiations are efficient (never choosing something when some other option would have made both parties better off), that the parties between them control enough relevant parameters, and that deals can be enforced, then the results of those negotiations are efficient. It's a useful tautology though, as it makes people think about what could make negotiations inefficient.
It is far from clear, however, that neglecting transaction costs is reasonable when there aren't too many farmers to negotiate with. There are lots of other factors that can lead to inefficient negotiations, such as asymmetric information.
>Coase's theorem is controversial among libertarians. Some will seize on
>the idea of property rights to solve problems usually left to government.
>... On the other hand, introducing new property rights smacks of coercion
>in itself, to many people (as we have seen here). Property rights in
>clean air and water may be acceptable; property rights to unobstructed
>views (common in some areas) begin to cross the line; and intellectual
>property rights are outside the pale.
I think the key problem is that most "libertarians" don't like simple consequential analyses of legal & political questions, preferring axiomatic analyses instead. If they can get someone to accept a "no coercion" axiom, then given enough time they expect to be able to convince them of the "right" libertarian views on things. If they engage in a consequentialist debate, however, talking about what policies lead to death, poverty, illness, etc., libertarians fear they may have to admit that in some situations the "wrong" policies may lead to better consequences. Or they may be forced to invoke an usual consequence like "reduces liberty".
Ideologues of other political views also fear simple consequential analyses. They want to invoke axioms like "everyone has a right to health care" or unusual consequences like "commodification" to avoid possible unpleasant outcomes of simple consequential analysis.
I embrace simple consequential analysis as a basis for policy discussions. I accept that it may sometimes favor anti-libertarian views. I also accept that simple consequential analysis may sometimes mislead us either because it neglects unusual consequences or because there really are some axioms we do not want to violate.
I accept these features because simple consequential analysis seems our best chance for creating broad intellectual consensus on policy questions. People do form self-serving beliefs about the details of which particular situations lead to which simple consequences. But evidence slowly erodes these views, and I see much less prospect for erosion of disagreements about what axioms to accept or what unusual consequences to include in analyses.
Robin Hanson
hanson@econ.berkeley.edu http://hanson.berkeley.edu/ RWJF Health Policy Scholar FAX: 510-643-8614 140 Warren Hall, UC Berkeley, CA 94720-7360 510-643-1884