659. J. M. Keynes to Harrod , 20 April 1937 [a]

[Replies to 657 ]

46, Gordon Square, Bloomsbury #

20 April 1937

Dear Roy,

I don't really like the formula , since M is here the average multiplier, which requires re-interpretation when net investment sinks to zero (zero investment does not mean zero income). Moreover, it suggests that y and R are independent variables; and that is your Achilles' heel. On the contrary, I should hold that y is determined by the rate of interest and the state of expectation, that M is determined by individual psychology and institutions, and that, in the short period, R is dragged at the chariot wheels of y and M. I am sure that a great deal of what you say in your book is influenced by a tacit presumption in favour of R being an independent variable.

On the other hand, the strength of your approach lies in the idea that R has, so to speak, a normal long period value which is a function of the rate of interest. When changes in y and M force R below its normal value at the existing rate of interest, there is a strong inducement to investment and expansion; when R has reached its normal value [b] this inducement is exhausted; and the exhaustion of this inducement will (apart from further incentives due to a change in the rate of interest) probably bring R above its normal value which will be an equally strong deterrent to investment.

After the crisis R will be super-normal which will tend to make the recession cumulative. With the passage of time capital will wear [c] out until R is no longer much above normal, and some lucky stimulus will get the ball rolling up again. In the upward phase R will be sub-normal. But with the further passage of time it will be normal again, and then we must expect the slump.

The recovery of R to normal will operate as a non-recurrent stimulus, exactly on all fours with the recovery of working capital to normal (which is a part of R) on which I have laid so much stress.

Thus you will see that I fully accept the application of your ideas to the Trade Cycle along the above lines. What I meant was that the factors determining the normal value of the Relation have little or no bearing on the Trade Cycle; and it was those I thought you were discussing in much of your book. Moreover I question the contention that the Relation involves the theoretical inevitability of recession--I do not question its probability in present circumstances.

I should prefer to say that when R has recovered to its normal value at the existing rate of interest, we are deprived of the adventitious aid on which we have been depending and are thrown back on the fundamental conditions of the growth, which must be the counterpart of net saving in excess of zero. We can tackle the normal value of R with a view to reducing it by means of the rate of interest--and that we must do. But we have no direct means of influencing the actual value of R. That is, as I have said, dragged at the chariot wheels of y and M; and for direct measures we are thrown back on ways of stimulating y (one way of which is to increase the normal value of R) and moderating M.

It has dropped out in your treatment, I feel, that, given the rate of interest, y is a function of the widely fluctuating state of expectation (for which you substitute "the expectation of a steady growth of consumption", which does not hold in the short period). If y increases because of a change for the better in expectation, price changes will necessarily ensue sufficient to modify R to suit the new value of y in conjunction with the multiplier. Neglecting changes in M, R is in the short period a function of y such that, when y rises, R falls. This will always be true in spite of the fact that the normal long period value of R is a function of the rate of interest; the normal value operating through the repercussions on y whenever the actual value departs from the normal.

As I mentioned before, I shall be lecturing on your book next term (much better instructed than if I hadn't written to you). [1] But I am not proposing to write anything--at any rate not until after you have carried your ideas a stage further, which I hope you'll soon do.

Yours ever,

J M Keynes

R. F. Harrod Esq., Christ Church, Oxford.

  1. 1. Harrod, The Trade Cycle ( 1936:8 ).
    1. a. TLS with autograph corrections, five pages on five leaves, numbered from the second, Cc in JMK CO/3/100-103, original in HP II-73. The latter folder encloses an offprint of D. H. Robertson, "[Review of] The Trade Cycle, by R. F. Harrod", The Canadian Journal of Economics and Political Science 3, February 1937, pp. 124-27. The letter is printed in Keynes, Collected Writings, vol. XIV, pp. 177-79. Reproduced by kind permission of the Provost and Scholars, King's College, Cambridge.

      b. Ts: «... its normal this inducement ...».

      c. Ts: «willwear».


Welcome page

top of page

Return to index of this section

Go to previous page

Go to next page