362. D. H. Robertson to Harrod , 5 August 1934 [a]

[The exchange continues at 363 ]

Trinity [College, Cambridge]

5 August 1934

Dear Roy,

Do not attend to this unless or until you feel inclined.

I have been returning to your rehabilitation of comparative cost, [1] and wondering how far it does dispose of the difficulties which Ohlin [2] etc. find in the classical theory. I think my difficulty centres on uncertainty as to your conception of rent. It appears to comprise only what I call Shovian rent, viz. the excess earnings derived by some bits of factors of production from being more useful then in their existing use than in any other [3] (your p. 25). But so far as I can see what Ohlin is concerned about is "Pigouvian rent", viz. the enhanced earnings in all uses of a (quite possibly homogeneous) factor of production which arise from the expansion of an industry, or set of industries, which make particularly large use of it (Cf Ec. of Welfare, 3 rd ed t , p. 805 [4] ). Objection can be brought against the validity of this concept in connection with the cost curve of a single small industry (cf Ec. of Imperfect Competition, [5] p. 118): but if we are examining big changes in industrial structure due to the opening up of foreign trade, it is surely to the point. E.g. the opening up of trade between Australia and England tends to increase the scarcity of land in general in Australia and increase rent per acre in all uses, and to have the opposite effect in England: (and to diminish the scarcity of (say) skilled labour in Australia and increase it in England.) This I take to be Ohlin's main contention: and if it is true, and an important part of the whole truth, such enhancements of land-savings as wheat production in Australia is expanded cannot (once crude Ricardianism is discarded) be ruled out of account as "rents",--the whole of them (an old unity of land as well as new) are transfer costs, i.e. prices which must be paid to prevent the land going to other uses, and cannot be put into a separate category from any other costs. {The argument that they are "offset" by enhanced incomes of landlords (p. 26) seems to me to prove nothing,--all costs are incomes!}.

What I am suggesting is that while your argument is valid as regards the Shovian rents with which it avowedly deals, it is an ignoratio elenchi of the real difficulty,--viz. the alteration in the relative values per unit of the factors of production (assumed uniform) due to the opening up of trade.

Please convince me,--I want to be convinced, for I hate the Ohlin method of approach, which throws overboard the great classical achievement of showing how the monetary epiphenomena--relative levels of incomes or rates of exchange--are the result of the underlying "real" situation.

I expect you will say that the whole answer lies embedded in that footnote on p. 26, but I should like it extricated, please, and expounded in terms of my question about what you mean by rent!

But again,--not unless or until you feel inclined.



  1. 1. Harrod, International Economics ( 1933:10 ), chapter 2 § 3. On the "rehabilitation" see Robertson's letter 231 to Harrod of 6 April 1932, [jump to page] .

    2. B. Ohlin, Interregional and International Trade, Oxford: at the University Press, and Harvard: at the University Press, 1933, in particular appendix III, pp. 571-90.

    3. G. F. Shove, "The Representative Firm and Increasing Returns", Economic Journal XL, March 1930, in particular pp. 99-100.

    4. Pigou, Economics of Welfare (1928), Appendix III, § 20.

    5. J. V. Robinson, The Economics of Imperfect Competition (1933), p. 118n.

    1. a. ALS, two pages on two leaves, in HP IV-990-1069d/18.

Welcome page

top of page

Return to index of this section

Go to previous page

Go to next page