In his final comment responding to critiques of his book Burning Down the House, Professor Andrew Koppelman responds to this VC post by Ilya Somin noting:
Somin does not dispute my claim that sometimes, large regulatory programs are justified. But, he says, the characteristic failures of democratic governance "amount to a systematic relative advantage of the private sector that should create a presumption against state control. The problem isn't limited to one or a few specific areas of government policy." This is not, however, the sort of question that is appropriately addressed with presumptions. As I say in the book, "whether this is so in any particular case cannot be resolved without attention to the local evidence." (69) Presumptions are not a substitute for such evidence. Sometimes libertarians can supply it: we are well rid of the Civil Aeronautics Board and the pre-1980 Interstate Commerce Commission restrictions on trucking. But sometimes the evidence points the other way. . . .
The modern regulatory state is a mighty complex enterprise, and it's hard to make reliable generalizations across the whole. The most powerful case for intervention is presented by problems of externalities, positive or negative, in which if government doesn't do something it just won't get done. Libertarian presumptions, as lately deployed in the Supreme Court, have crippled the capacity of the federal government to address climate change and Covid. This is not a gain for liberty.
This comment made me think of Ronald Coase, whose work on "externalities" and transaction costs is routinely quoted, but often mischaracterized or misunderstood. Indeed, while Coase's work is often cited for the proposition that the presence of externalities (and transaction costs) justifies governmental intervention, that was not how Coase understood his own work.
