LIGHTHOUSES. Atlantic Blog points to a D-Squared Digest reprise of the public good property of lighthouses. D-Squared puts Ronald Coase on his fact checking list. I'd suggest he check some of his own facts as well.
"Second on methodology, while the source of my sympathy for Coase is his methodology; his determination to always look for the real world example rather than the 'blackboard economics' proof, this is one area in which he and Samuelson both screwed up, and it was by adopting the arrogance which is the hallmark of the economics profession. Samuelson had a good argument about the optimal provision of lighthouses, and if he'd studied the history, he would have come up with the Isle of Man and Sir John Clayton examples to show that non-excludability of lighthouse services led to real problems, which caused real ships to avoidably crash. But he had to, as I say, put a f*****g cherry on top by overextending his sensible blackboard argument into a generalisation about the world. Coase then, correctly and admirably, called him on it by falsifying the generalisation, but then fell into the trap of forcing the pieces to fit into his own grand blackboard generalisation -- that free market negotiation between willing participants could always solve problems of resource allocation so long as property rights were well defined."
But wait, there's more ... one of the Deirdre McCloskey columns in Atlantic Economic Journal (Or is it the Eastern???) takes a closer look at "Problem of Social Cost" and notes that the real fun begins when transaction costs are present. That's the real point of departure, whether you're a Coasian or a Samuelsonian. To simply assert that costly information 'justifies' government action is to deprive yourself of a research opportunity. Where information is costly, there might be many ways of conserving on the costs of information.