468. Harrod to J. M. Keynes , 21 August 1935 [a]

[Follows on from 467 , continues at 469 ]

21 August 1935

Having re-read Ch. 15 [1] I am more depressed about it than I was before, when in the first flush of enthusiasm I was ready to swallow anything.

Let me make this preliminary point. I dont for a moment believe that the authors to whom you refer conceived the possibility of saving being anything else than equal to investment. The inequality between saving & investment was a new-fangled idea propounded by you in the Treatise and since taken up by others & mis-applied. [2] So that when you argue that once we see that saving must be equal to investment, this argument falls, they would reply that they had always supposed that saving must be equal to investment and that that was the basis of their argument. You rightly say that S = I corresponds to the truism that sales = purchases. But what is it that compels sales to be equal to purchases, when every one is perfectly free to sell or not to sell and to buy or not to buy? The traditional answer is price, which in the case of S = I is the rate of interest. (The parallel between S (= I) and commodities which cannot be put to stock is complete. And the existence of stocks is not of the essence of supply & demand analysis but a complicating factor.)

I agree with your para on Galley 70, [3] beginning "Certainly the ordinary man ...." He thinks an increased propensity to save will lower the rate of interest. The next point you ought to go on to (vide your nonsense paragraph beginning "Now the analysis ..." [b] ,   [4] ) is that the rate of interest may not fall because there is another relevant variable in the situation, viz. the level of income. The increased propensity to save may, instead of lowering the rate of interest, merely lower the level of income by reducing the multiplier, and establish a new equilibrium with the old rate of interest and less employment.

Then you should go on to say that not only may this be so, but in certain circumstances it must be so, viz. if there is no change in the liquidity preference schedule or the quantity of money. Q.E.D.

The classical school went wrong, one, [c] because they did not perceive that there was another dependent variable in the situation, viz. the level of income, and, two, because they did not appreciate that there are other forces which if unchanged would compel the rate of interest to continue to be what it is, and thus, so to say, compel the change in the situation to act not on the rate of interest but on the other dependent variable, viz. the level of income.

Then you can come back to your delusion of the "ordinary man, banker, civil servant" etc. (vide para 1 on Galley 70 [5] ) and point out that when an individual performs an act of saving there may be no net addition to aggregate saving (and in certain circ[umstance]s. there can be none) because his particular saving may merely act on the multiplier and not on the rate of interest (and in certain circ[umstance]s. must do so).

In the bottom para of 70 [6] when you say that saving and investment are twin determinates, you are in perfect agreement with the classical system. Only the classical system made them determined by something different. For the classical school they were determined by the marginal efficiency of capital and the propensity to save, for you by the same marginal efficiency of capital, but on the other side by the rate of interest, which for you is determined by something quite different. But the propensity to save comes into your system too, but in a different way, viz. in determining the value of the multiplier.

I dont follow the paragraphs beginning "The second approach of the classical theory ..." [7] because I only know of one approach in any reasonably modern theory--tho with troublesome terminological wrangling.

The last 2 paras of the ch. [8] seem to me to be O.K.

N.B. I cant help thinking that the doctrine of the multiplier ought to be explicitly set forth somewhere here rather than later. 1

Where the classical school undoubtedly begins to get into difficulties is when in their monetary department they begin to discuss the effect of the rate of interest on the level of activity. Because once they begin to do this, they recognize the level of income to be a variable (whereas it is essential to their pure theory of interest to assume it constant) and they recognize some kind of nexus between interest and the level of income. This however is a development which takes place in the department of money and may be held to undermine their general theory; but this is no excuse for attacking the general theory as inconsistent or confused on its own premises.

I send off these notes as I write them.

R. F. H.

  1. 1. Chapter 14 of the General Theory (1936) (reference is to the table of contents sent by Keynes on 26 June: letter 451 ).

    2. A few months earlier, the definition of saving and investment given by Keynes in the Treatise on Money (1930) was the subject of a prolonged exchange with Kahn (see letters 375 , 382 , 383 , 387 , 389 , 391 , 392 , 402 ), was discussed with Haberler (for instance letters 388 , [jump to page] ; 393 , [jump to page] ; and 394 , [jump to page] ), and was on the background of several exchanges with Robertson (see for instance letter 306 ).

    3. Keynes, Collected Writings, vol. XIV, p. 475; General Theory (1936), p. 177.

    4. Ibid., following paragraph.

    5. Ibid., paragraph in the middle of p. 177 of the General Theory (1936).

    6. Collected Writings, vol. XIV, p. 476; General Theory (1936), p. 183 last paragraph.

    7. Collected Writings, vol. XIV, pp. 476-77.

    8. Collected Writings, vol. XIV, pp. 478-79; General Theory (1936), pp. 184-85.

    1. a. ANI, three pages on three leaves, numbered from the second, in JMK GTE/1/314-16. Printed in Keynes, Collected Writings, vol. XIII, pp. 544-46.

      b. Ms: «...)».

      c. Pair of commas added to facilitate the reading.

      d. Keynes's mark in the margin.

1. Keynes marked this passage with a vertical line in the margin.

Welcome page

top of page

Return to index of this section

Go to previous page

Go to next page