399. Harrod to G. Haberler , 10 November 1934 [a]

[Follows on from 398 , replies to 396 , answered by 401 ]


10 November 1934

My dear Haberler

While you have to neglect the affairs of the world in order to write me letters, I only have to neglect my pupils. So that by ordinary principles you may expect me to get the last word. He who has the last is not of course necessarily right.

I should like to say that I do always think of income, savings and investment as flows in time. When one says amount of new saving one ought to express it by . But I dont think I have confused successive periods. [1]

In your last letter you seek to rebut me by reference to a static state. "If, now, a person gets more money on the first day of the week and he spends the additional money with his ordinary income, we can surely not say that he saves, just because he keeps the money on the average (say) half a week in his pocket." I do not see the force of this argument. Why should I not say that he saves? You are not suggesting that we need not bother to call this saving because the amount involved is trifling! Is it that you wish to argue that he cannot be saving because in the static state there is no saving by definition? But there is no net saving by definition; this does not mean that there is no saving by anyone. If our friend gets a rise of income in a static state, someone [b] else's income must be reduced, and, since M is constant, his holding of money will be reduced in the early days of the week. He therefore will be dis-saving and his dis-saving will be equal to the saving of our friend. I cannot therefore see any force in this argument from a static state.

Your second (and only other) argument is that if factors of production spend money coming in to them in respect of new investment simultaneously with the receipt of that money, the income flowing in to entrepreneurs as a result of that expenditure belongs to a second period. [2] But why? If one income accrues simultaneously with another, how can the former be said to belong to a period subsequent to the latter?

Of course I grant that the factors who receive money income in connexion with the new investment will probably not increase their expenditure by the whole amount simultaneously (especially if the rise of income considered involves a discontinuity). In so far as they do not, they undertake saving which pro tanto balances the investment. But any increase of expenditure that occurs simultaneously must be counted as an additional simultaneous increase of income to other entrepreneurs. Then the same considerations apply in the analysis of what they do with that. I do not admit that we have yet got into the second period.

On the position of Keynes, which is, of course, quite distinct from this issue, I am not prepared to argue that anything is gained by his strange definition. I am prepared to agree with you. What I do want considered is his proposition that if the market rate of interest departs from the equilibrum rate, our economic system does not provide forces (as it does in the case of other prices) tending to restore the equilibrium. [3] You might then say this only means that [c] there is no equilibrium rate, since by definition an equilibrium position is that to which there will be [d] a return after a random force has disturbed it. I agree that it does mean that there is no equilibrium rate in this sense. Linked with this proposition, is the proposition to which I am provisionally prepared to subscribe (tho' not with a feeling a certain assurance, with which I subscribe to the proposition that saving investment is impossible except on odd and peculiar definitions) that there is no equilibrium volume of output, and in particular that, even on the supposition that there were no rigidities, there are not any forces in our system tending to a position in which productive resources are fully utilized. Generally, the amount of productive resources that are utilized is determinate. If it is true that there is no equilibrium rate of interest in the proper sense, one might define the equilibrium rate in some special way, such as that consistent with regular advance. Keynes' proposition would then be that if the market rate diverges from this there are no forces in the system tending to bring about a restoration. All this is rather loose and I particularly emphasize that I dont put it at all on the same plane of reasoning as that which I advance concerning the equality of saving and investment.

I cannot help asking--is it fair to bother you with all this. <+> way in which I justify my action is that if you are writing this survey, [4] it is as well that you should know the kind of thoughts which may find expression in future publications from my pen or that of orthers. This knowledge might do something to forearm you, even if you cant accept the positions, and enable you to couch your conclusions in a form that would less easily be upset by such publications than they would be if you were working in ignorance of them.

Yours very sincerely

Roy Harrod

  1. 1. Letter 396 , [jump to page] .

    2. Letter 396 , [jump to page] .

    3. See also letters 378 , [jump to page] and 394 , [jump to page] to Haberler, and letter 392 , [jump to page] to Kahn.

    4. G. Haberler, Systematic Analysis of the Theories of the Business Cycle (1934). See letter 364 , in particular note 1 for context.

    1. a. ALS, two pages on one leaf, in GH Box 66.

      b. Ms: «some one».

      c. Ms: «than».

      d. Ms: «will a return».

Welcome page

top of page

Return to index of this section

Go to previous page

Go to next page