## 465. J. M. Keynes to Harrod

, 14 August 1935[a]

[Replies to 464 , continues at 466 ]

I mean here I think the same that you mean. But it did not seem to me necessary to put "prime" in front of "factor". I was considering prime cost as made up of user cost and factor cost, each of the constituents and also the sum of the two being, of course, further divided into marginal and average. I do not see that it is necessary to sub-divide supplementary costs into supplementary factor costs and something else. Thus, I mean by factor costs those current disbursements to the factors of production which could be avoided by not using the machinery. However, it would be very easy to make any further sub-division you like if there were any occasion for it. I will deal with your other point later.

Do I ever affirm or deny that "the propensity to save has to be equated to the propensity to invest" or words to that effect? [1] If so, I should not in a revised draft. The propensity to save is a schedule or function, not a quantity at all, namely the function relating different levels of net income to the amounts saved at each level. The inducement to invest (I no longer speak of a "propensity" to invest [2] ) is the relation between the schedule of marginal efficiency of capital and a given rate of interest. A change in the rate of interest will change both the propensity to save and the inducement to invest; but there is no sense in which it can be said to equate them. If we take everything else as given, the rate of interest determines the equilibrium level of income at which consumption plus investment provides the amount of employment, the supply price of the output of which is equal to that level of income. But if entrepreneurs make mistakes and offer a different level of employment, saving and investment will still be equal.

One can (as assumed) only invent a meaning for the classical ideas, if one assumes that income and employment cannot change. On this assumption the rate of interest causes the propensity to save and the inducement to invest to change in such a way relatively to one another that income and employment are always the same whatever the rate of interest. In this case one can abolish the function "propensity to save" and introduce a new one, namely the curve which intersects the family "of propensities to save" corresponding to different rates of interest at the point of each corresponding to the <+> unchangeable volume of income. This is the classical supply curve of savings. But if income and employment are capable of changing, there is no longer any such thing.

However I will try to make matters clearer when I get to redrafting that chapter.

R. F. Harrod Esq., Sennen Cove Hotel, Sennen Cove, nr Lands End, Cornwall.

- 1. See letter
464 ,
[jump to page] .
2. In the Table of Contents of The General Theory sent to Harrod on 26 June 1935 (letter 451 ), Keynes had corrected in his own hand the title of Book IV from "The Propensity to Invest" into "The Inducement to Invest."