ABSTRACT
In the light of the what we have seen of Coase’s discussion of Pigou in ‘The problem of social cost’, the way in which he begins that discussion is unusual. He registers an agreement. Coase is careful to distinguish his criticisms of Pigou’s arguments for intervention, of which we have seen he is profoundly critical, from an argument of Pigou’s with which he agrees. The argument for intervention in Part 2 of Economics of Welfare is prefaced with a four page introduction in which Pigou quotes the distinguished economist Edwin Cannan to the following effect:
This is a most significant argument with which to agree, for it is, of course, a rejection of the invisible hand.73 It is as well to quote the most substantial description Adam Smith gives of the invisible hand:
I also quote Smith’s most famous illustration of the operation of the invisible hand:
However, it is not to either of these passages that Pigou directs his criticism of the invisible hand, but to the following:
Pigou evidently thinks that Cannan’s criticism of the invisible hand disposes of this basic argument against intervention and proceeds, in the remaining 284 pages of Part 2 of The Economics of Welfare, to set out the apparatus for distinguishing between private and social net products that has since grounded intervention. But Coase’s criticism of Pigou allows us to see that Pigou has missed his mark in Smith, and that Pigou’s choice of quotation from Smith is ample evidence of this. In this quotation, Smith is setting out an argument against intervention in functioning markets which, of course, so tells against the ‘industrial policy’ which has been adopted in the UK since 194577 that it could (with stylistic alterations) easily be passed off as a general conclusion about that policy. Pigou obviously is disinclined to accept this argument and, were he to set out a case for intervention which showed awareness not merely of market failures but of the difficulties of improving on the position produced by the market, one would be compelled to consider it at length. But, to take only his most important general case, Pigou is prepared to consider ‘permanent bounties’ to shift the entire industrial system away from equilibrium based on a particular distribution of goods to an equilibrium based on another distribution (in which less of a good disliked for moral and political reasons or more of a good favoured for such reasons would be produced), even though: ‘[t]he conditions in which bounties are likely to have this effect’, he fairly says: ‘are somewhat special.’ He concludes, however, that: ‘it can be proved that, in certain states of demand and supply, some rates of bounty must have this effect,’ and, the clinching point, those states ‘can readily be depicted in a diagram.’78
