394. Harrod to G. Haberler , 5 November 1934 [a]

[Replies to 393 , answered by 396 ]

Christ Church, Oxford

5 November 1934

My dear Haberler

I am very interested in your idea of arbitration. I have myself often advocated that something of this sort should be done in cases of disputes among economists. I think perhaps "conciliation"--to use the language of industrial disputes--would be more appropriate in the case of academic disagreement, since "a man convinced against his will is of the same opinion still." I will certainly raise this question with Marschak [b] & Opie, [1] when I get an opportunity, but, though open to conviction on any point, I still feel very far indeed from the threshold of conviction on this. I tried to bring my state of mind before you when I said in an earlier letter that I feel on this as I think you would feel if asked to deny the truth of a properly stated quantity theory. [2]

To get down to detail. I assume that your admirable office keeps a copy of your letter. I refer to page 2. You offer alternatives (a) and (b), saying that (b) is more in harmony with current usage. [3] I agree. I have always thought along the lines of (b). So we can leave (a) on one side.

On the theme of (b) you do not write very much. You write: "If you decide for this alternative you have an increase in investment, without a corresponding increase of savings. (During the following periods, after the newly created money has become wholly or in part income in somebody's hand, the situation might be different; but that is another question)." That is all you say about (b) and that therefore is all I have at present to deal with.

Consider it. In my view if a bank makes a loan of £100 to an entrepreneur and he does not use it, viz. he leaves it lying in his account at that bank, that is not an investment of £100. Do you wish to challenge that? Next, if this entrepreneur, whom we call A, buys some raw materials from entrepreneur B out of [c] B's stock of same, and B depletes his stock of goods by this amount and adds to his money balance by this amount, that is not investment. For if you imagine A and B an integrated concern, this latter case is the same as the former. A + B have not used the loan to increase the output of producers' goods but have merely let it lie as an extra balance. Consequently investment only begins when A or B or both begin to pay out some or all of the £100 as income to factors of production. Consequently I challenge your sentence in brackets. I do not think the period in which the newly created money is paid out as someone's income is a following period but a simultaneous period, and I do not think what happens then is another question but the question.

Simultaneously, then, with the investment of £100, there is an increase of consumers' income of £100. What do the consumers do with it? You might suppose that simultaneously they do nothing with it (spending it somewhat later). But in that case there is an increase of saving of £100 and the new investment of £100 is exactly balanced by a new saving of £100. 1

But suppose, seeing their income rise, they simultaneously spend £50. This involves a rise in entrepreneurs' incomes of £50. Thus simultaneously with the original investment of £100, there is a total rise of incomes of £150 of which only £50 is spent and \ £100 saved. Further if the entrepreneurs simultaneously spend £50 more, there is an [d] additional rise of other entrepreneurs' incomes of £50, viz. a total rise of £200 and total expenditure of £100 and \ saving of £100. Or if the entrepreneurs invest the £50, viz. pay it out to other factors, there is a total investment of £150, rise of income of £200, rise of expenditure of £50 and new saving of £150.

You may trace out the path of the original new loan of £100 as you will, making what hypotheses you please, but the same result always ensues--that you have additional saving of £100.

This brings me back to my point that unless you define savings (or investment) in a funny way, the age-long assumption that savings = investment, which is nothing more than the assumption that the amount of a commodity sold must be equal to the amount bought, is true. [4]

I quite agree that it is not possible to find out something about the real world by re-defining terms. But it may be possible to explain a discovery already made by using terms defined in a somewhat different way. This, I think, is what Keynes did. But such a method of exposition is attended with danger, because the unusual definitions may bewilder the mind of the reader. This has happened with the Treatise. Very few (even of distinguished economists) have followed what Keynes has said there. [5] In particular Keynes has given popularity to the notion that Investment and Saving need not be equal, many of those taking up this idea forgetting that it presupposed an unusual definition of saving and would be quite invalid on an ordinary definition of saving. But Keynes' doctrines may be explained in other terms without using his curious definitions.

Hawtrey's article contained many interesting things. I am not sure how far it was meant to be a criticism of Keynes' central position, but, if it was, it was based, I believe, on an ignoratio elenchi. [6]

I do not want at this moment to play the part of defender of Keynes. All I said in my earlier letter [7] was that his views ought to be recorded as an interesting contribution to trade cycle theory not yet refuted. The essence of his view is this, I think. If a market rate of interest gets established, which is not consistent with regular advance, or which, I think we may therefore say, is not equal to the equilibrium rate, no forces are at once brought into being to bring it back to the equilibrium rate. This is contrary to what usually happens in a market. If he is right in this, he has got onto the track of the kind of phenomenon which is required to explain the cycle. I quite agree with the point well expressed by Hayek in his Monetary Theory, that general equilibrium analysis leads us to expect that if there is a departure from equilibrium at some point in the system forces come into play leading to a restoration. Yet the cycle phenomena suggest a wider and wider departure. [8] Now Hayek's own explanation I regard as totally fallacious for reasons already given. Keynes gives the kind of explanation that Hayek desiderates and I am not yet convinced--I know of no published refutation except on some points of detail--that Keynes' explanation is wrong.

Yours sincerely

Roy Harrod

  1. 1. See note 5 to letter 393 .

    2. Letter 388 , [jump to page] .

    3. Letter 393 , [jump to page] .

    4. Letter 388 , [jump to page] .

    5. A similar point was raised by Kahn in letter 391 , [jump to page] .

    6. Harrod probably refers to R. G. Hawtrey, "Public Expenditure and Trade Depression", Journal of the Royal Statistical Society, Part III, 1933, in particular pp. 447, 464-65 and 474. Harrod's doubts as to the objective of Hawtrey's critical remark indicates that he probably was not aware of the more explicit criticism advanced in Hawtrey's The Art of Central Banking, London: Longmans, 1932, pp. 343-44, to which Haberler referred in letters 393 , [jump to page] , and 396 , [jump to page] .

    7. Letter 378 , [jump to page] .

    8. F. A. Hayek, Monetary Theory and the Trade Cycle (1933), in particular pp. 42-44 and 70-71.

    1. a. ALS, four pages on four leaves, in GH Box 66. TL transcription Cc, with Ms corrections, four pages on four leaves, in HP IV-395-422; further Ccs also in NKP NK/3/5/7-10 and LoN 10B/12653/12653.

      b. Ms: «Marchak», Ts: «Marshak» (transcribed as "Marshall" by the typist, the final double l was corrected in k, probably by Haberler).

      c. Ms and Ts: «of his B's».

      d. Ms: «a an».


1. Haberler marked this paragraph with two vertical lines in the margin.


Welcome page

top of page

Return to index of this section

Go to previous page

Go to next page