532. H. D. Henderson to Harrod , 26 February 1936 [a]

[Replies to 528 , answered by 533 ]

14, Upper Park Road, London #

26 February 1936

My dear Harrod,

Many thanks for your very sweet-tempered letter. I quite realise how wounding my attitude must be in that I appear, far more than I really mean, by the way--to be challenging the utility of what you have been working at for some years past. And that being so, I think you show a most forbearing spirit in receiving my criticisms as you have done.

I can never restrain myself in controversy from going rather farther than I really mean, and in appearing to dismiss with contempt the whole of the sort of analytical work you have been engaged on, I certainly went farther than I meant. Let me try to restate my position as follows. A highly abstract analysis, if logically coherent, as yours certainly is, can claim a value which is not to be despised in the sphere of intellectual and aesthetic satisfaction. It may further have a practical utility in relation to the study of concrete problems by suggesting the right questions to ask. This is all the utility I would claim for the simpler parts of orthodox economic theory, (apart from the market laws of supply & demand) that it does really help in suggesting the questions to ask, i.e. the points to which to direct attention in any particular problem. I am sceptical as to the utility of the more elaborate analytical work of modern times (which certainly dates back to Pigou--I was always on Clapham's side over the "empty economic boxes" [1] ) because I have never found it helpful in suggesting the right questions to ask. But that is a personal scepticism of mine, possibly over-hasty, which you of course must disregard, and I am being entirely unreasonable if I appear at any point to suggest that you should reconstruct your line of argument to meet my personal crotchets.

At the same time, that section of your book on profits and prices [2] seems to me to raise another methodological issue which I do beg you to consider carefully. You seem to me to be using your analysis not merely for the purpose of arranging the questions you ask in an ordered form but as premisses of fact by which you seek to establish conclusions relevant to the concrete short-term world. That seems to me exposed to an objection to which your previous work is not exposed, but here again if you do not agree you must of course disregard me, and with the above I will drop finally all criticism of a methodological character.

There is, however, one point which it may be of positive help to you that I should develop more fully. I should like you to reflect further on the real significance of the contrast between industry and agriculture in regard to the way in which depression reacts upon output and prices. The main cause of the difference is not, I think well expressed by saying that the marketing conditions of agriculture are more competitive. [3] That factor comes in; but it is the fact that it works in conjunction with an essentially distinct one, namely, that the scale of output in agriculture is decided in advance of the contracts for sale, that really accounts for the phenomenon. The typical cotton spinner, as Lee reminded us, quotes for orders in a perfectly competitive market, but he does not (for the most part) produce except to meet the orders so obtained. [4] Accordingly he is unlikely to accept orders appreciably below his prime costs of production. Thus however competitive the conditions may be in which the contract for sale is made, prices are not likely to fall appreciably below costs, if the contract for sale precedes the decision to produce. Conversely output will be restricted in exact proportion to the decline of demand.

In the case of most agricultural products, on the other hand, the higgling of the market takes place after the goods are already produced, or at least after the decision to produce them has been made. It may often happen, accordingly, that the supply greatly exceeds the demand and then, under competitive conditions, the only limit to the extent to which prices can fall is set by whatever mechanism exists for carrying stocks forward to another year. Again the fact that prices have been unremunerative in one year will exert no necessary effect on the output of the next year, except in so far as it drives agriculturalists out of existence through bankruptcy; for in deciding how much to produce the next year the competitive nature of the market again comes in and deprives the farmer of any ground for hoping that he can secure a better price by producing less himself.

Now the distinction between producing for orders and producing in advance of the contract for sale is one that cuts right across the distinction between competitive and semi-monopolistic market conditions. Scott-Stokes [b] is a semi-monopolist but he determines his scale of output in advance of his sales: [5] so presumably does Lee.

So far as I understand it, your system of analysis takes no account of this very important distinction between working for orders and producing in advance of orders. It would be possible, I imagine, to fit it in by using some of the terms of the analysis in an artificial sense. One might, for example, regard the whole cost of production, in the case of commodities produced in advance of sale, as a sort of overhead from the short period standpoint and say that the marginal costs involved in selling rather than not-selling what has already been produced are negligible. But if one were to use the term "marginal costs" in this way most of the premisses assumed as to the relationship between marginal costs and wages and the like would obviously cease to hold good. Alternatively, if you use marginal costs in the more ordinary sense (though the ordinary meaning is rather vague to my mind) then it seems to me that no propositions as to the relationship of marginal revenue and marginal costs are true in the short period whenever commodities are produced in advance of the contract for sale. This means that the relationship does not hold true in the short period in the case of a very large range of industry as well as of agriculture.

This brings me to my moral relating to your prices and profits section. Your law of the diminishing elasticity of demand [6] seems to me reasonable a priori, and in conformity with the facts, so far as we know them, both of the long period and of the short period. Subject to my personal crotchet against the word elasticity, [7] I have no quarrel with it whatever, and I am quite open-minded as to the degree of importance that may attach to it. I demur only when you say that this alone can account for the agreed facts as to the relative movements of profits, prices and wages during a slump. [8] Those facts, as it seems to me, are easily accounted for in other ways. The above argument gives [c] an indication of one of the ways in which they can be accounted for, and since you can only advance on the other side an argument that depends entirely on the assumption that marginal revenue cannot diverge in a short period from marginal costs, an assumption which in turn ignores the fact that many goods are produced in advance of the contract of sale, your contention seems to me wholly unconvincing.

I do not, in other words, seek to disparage your law, but I maintain with considerable confidence that the claim which you make for it in this section cannot be substantiated.

We both appreciate, I think, that the scope of my criticisms is limited to a very small part of your book, however vital it may appear to you. Though disliking as a matter of a personal crotchet the identity between savings and investment approach, [9] I think that your management both of the exposition of Keynes's views and of the way in which large fluctuations of savings may occur is really a masterly bit of work, as I think I told you.

Yours ever,

Hubert D. Henderson

R.F. Harrod, Esq., Christ Church, Oxford.

  1. 1. J. H. Clapham, "Of Empty Economic Boxes", Economic Journal XXXII, September 1922, pp. 305-14.

    2. Harrod, The Trade Cycle ( 1936:8 ), pp. 75-88.

    3. The conditions of agricultural production are discussed in The Trade Cycle (Harrod 1936:8 ), pp. 33-35. For Henderson's comments in this connection see letter 524 , [jump to page] .

    4. Lee was the second entrepreneur interviewed by the Oxford Economists' Research Group on 21 February 1936 (on the formation and activities of the OERG see note 10 to letter 433 ). Harrod reported that "The typical cotton spinner sells in an organised market to the manufacturer. The merchant, having got his orders, get competing tenders from the manufacturers. It is a very free competitive market. The dyers and bleachers charge ring prices from the merchants in whose hands the goods are by that stage" ("Oxford Economic Research Group. Visit of Sir Kenneth Lee on 21.ii.36", marked "confidential", TDI, in HCN 5/11 and ABP 46, p. 3).

    5. H. F. Scott-Stokes was interviewed by the OERG on 31 January 1936 (the minutes, a draft and the final version of the report on the visit, drafted by Harrod, "Visit of Mr. H. F. Scott-Stokes on 31.i.36", are in HCN 5/1 ): see note 1 to letter 529 R.

    6. Harrod, The Trade Cycle ( 1936:8 ), pp. 21-22 and 82-85. See letters 524 , [jump to page] , and 528 , [jump to page] .

    7. See letters 524 , [jump to page] and 527 , [jump to page] .

    8. Harrod, The Trade Cycle ( 1936:8 ), p. 75.

    9. Saving and investment were discussed a few weeks later, in an exchange on Henderson's criticism to the General Theory: see letters 540 , [jump to page] , 541 , [jump to page] and 542 , [jump to page] .

    1. a. TLS with autograph corrections and additions, five pages on five numbered leaves, in HP IV-480-484. Cc with autograph additions in HHP Box 22/A6.

      b. Ms: «Scott Stokes».

      c. Ts: «gives,» (the extra comma was left by mistake after crossing out part of the sentence).


Welcome page

top of page

Return to index of this section

Go to previous page

Go to next page