473. Harrod to J. M. Keynes , 20 September 1935 [a]

[Replies to 472 , answered by 474 ]

51 Campden Hill Square, London W.8.

20 September 1935 [b]

Dear Maynard

By all means use my construction of a family of classical supply curves if you feel so disposed.

To return to your point that the Classical Theory doesnt make sense. You write: "the assumption of a fixed income is too all embracing and does not allow for those things to vary which the classical theory assumes to vary. For, if either the propensity to save or the marginal efficiency curve changes, the level of income changes." [1] Yes, according to you; not necessarily according to the classical theory. And if it does change, it does not do so by the right amount for your theory. Take an increased propensity to save. Why will income vary? The level of income, remember, is uniquely determined by the marginal disutility of labour and its marginal productivity. Which of these varies? If the rate of interest falls, as classical theory supposes it to do, then the productivity curve [c] moves to the right--future product being discounted at a lower rate--and income goes up! An increase in the marg. eff. of capital might also cause income to go up (e.g. if due to an invention making both labour & capital more effective, or to the opening of a new foreign market). But there is no guarantee in classical theory that it will go up by the right amount to make a rise in the rate of interest un-necessary. So that tho' income cant be taken as constant in the face of these changes, the variations in it induced by variations in the propensity to save or m. e. of cap. are not according to classical theory of the order of magnitude you require for your theory.

Return to the increased propensity to save. The classical theorists supposed a fall in the rate of interest. A pure miracle, say you, with your liquidity preference in mind. You claim that income will fall sufficiently to reduce saving to the level at which it stood before the change in propensity and so allow equilibrium with the old rate of interest. But doesnt this fall in income look a little like a miracle? With the level of income uniquely determined by the labour disutility & productivity functions it is not only a miracle but an impossibility. By your employment theory, (viz. disutility of labour determining only maximum level of income), you make appropriate variations of income a possibility. By your theory that interest is otherwise determined, viz. by liquidity preference, you make it a necessity.

What I think is that you are apt to forget the position before your own mind began to work along these lines. I also think that, while your definition of the marginal efficiency of capital is one of the most brilliant and important parts of the book, you are wrong in supposing that the idea (tho' in a vaguer & cruder form) was not present in the classical writers. You get it in Marshall's review of Jevons (1871)! [2]

I dont find the long passage from Marshall (ch. 16 [3] ) nonsense. I find it harder to understand some of your criticisms. I cannot find any confusion between interest and quasi-rent. According to both him & you, the current rate of interest = the marginal efficiency of capital. But whereas according to you the rate of interest is independently determined, according to him the two unknowns, amount of investment and rate of interest are jointly determined by the marginal efficiency of capital schedule and the propensity to save (from a level of income taken as given). The value of capital other than that involved in the current marginal investment is determined by its quasi-rent and the rate of interest. 1 Subject to the ordinary difficulties of expressing any of these things in language, the M[arshall]. passage seems to me crystal clear and your criticism in the text comparatively difficult to follow.

What I dont like about the chapter is the implied allegation that Marshall's system of thought was fundamentally confused, whereas I believe the confusion to be due to your failure to think yourself back into the system of thought you have abandoned. But for those who are still in it, I believe your chapter will seem captious and lacking in comprehension.

I should be delighted to come for a night one day next week.



  1. 1. Letter 472 , [jump to page] .

    2. Harrod's statement seems somewhat over-optimistic: see A. Marshall, "Mr. Jevons Theory of Political Economy", Academy, 1 April 1871, now in Pigou, Memorials of Alfred Marshall, London: Macmillan, 1925, pp. 93-99.

    3. Keynes, Collected Writings (1971-89),vol. XIV, pp. 479-81, General Theory (1936), pp. 186-87.

    4. See letter 474 , [jump to page] .

    1. a. ALS, three pages on three leaves numbered from the second, in JMK GTE/1/414-16. Printed in Keynes, Collected Writings,vol. XIII, pp. 560-61.

      b. Ms: misdated «20.xi.35».

      c. Ms: «curves».

1. Keynes noted at the top of the first page: "M[arshall] believed that an increase in the quantity of money reduced the rate of interest." [4]

Welcome page

top of page

Return to index of this section

Go to previous page

Go to next page