See accompanying letter


pp. 4-5. [5] This is all ignoratio elenchi. The L in M = L(r) changes from day to day, according to the conditions, like any other demand function, and in particular according to the price level. Your criticism is like some one attempting to negate the proposition that a fall in price stimulates demand, by reference to a lower price conjoined with lower demand at a time when the price-levels, the level of incomes, people's tastes etc. had all changed.

p. 7. [6] If the propensity to save is nil, you have the economics of full employment, viz. the classical system.

When you say "I do not need to take so extreme an illustration", I should reply that you do, since the remaining argument on p. 7 is quite inconclusive. Your conclusion about 3/4 people being out of work is no more paradoxical than your premises which involve their saving 4 times as much whatever the unemployment. The sentence which concludes this para (on p. 8) is entirely unsupported.

p. 8-9 [7] Keynes offers a perfectly adequate explanation of why rates fall in slump and rise in boom. The amount of liquidity required depends on the volume of trade and the level of prices (so says J.M.K. and, I should have thought, everyone!). In a slump the volume of trade and the level of prices fall and unless the quantity of money is reduced pro rata, which it usually is not, the rate of interest vary in accordance with M = L(r) be expected to fall. And to rise in the boom for the same reason.

(N.B. I dont feel altogether happy about the liquidity preference theory of interest, as a complete theory, for reasons which I have stated in my book. [8] But if you get rid of that you are one equation short. It means that the true theory of interest is still to seek, not that there is any theory in existence, that makes sense!)

p. 11. [9] Of course you are perfectly entitled to define Saving & Investment as you like, and work out any theory you can on the basis of any definition you like to give. The worst of allowing inequality between Saving and Investment is that everyone has his own definition. I have carried on a considerable correspondence in the last few years about this topic with economists in various parts of the world, and I find that every economist has a different one. I havent come across yours before.

It is quite futile to scold Maynard for not having adopted yours. He might equally well be scolded for not having adopted any of the other economists', whom I have corresponded with (e.g. DHR's [10] ).

The objection to yours is the difficulty of drawing the line. When the Canadian Wheat pool accumulates its holdings in the definite hope of getting a better price, is this voluntary or involuntary? 1

p. 13. [11] You assume rather too readily that stocks accumulate in the slump. This is true of course of basic materials. But so far as processed goods, goods in shops etc. there is a decumulation, and tho the volume of change may not be so great, one must remember that these goods have higher value. 2

p. 13. [12] The charge made in the last paragraph that Keynes uses investment in a different sense in the remainder of the book is unsubstantiated and untrue.

The sentence beginning "This is patently untrue if ..." is wrong. Income in the period of accumulation will be higher than it otherwise would be whether the accumulation is voluntary or involuntary. 3 The argument you attempt here indicates that you have not sufficiently considered the doctrine of the multiplier.

p. 14. [13] I dont think there is much in the point that large public works might be financed by a reduction of the stocks of materials and finished goods. No doubt that might be a short period effect. 4 It is mentioned by Maynard somewhere. [14]   5 It is a question of the quantitative 6 importance of this.

If Maynard's point is otherwise correct about the euthanasia of the rentier, [15] it is not likely to be invalidated because there are a few odd bricks lying about and a few unsold <lots> at the time of starting.

p. 15-16. [16]   7 Here we run into fallacy again. 1. If the banks allow an overdraft, this increases the quantity of money, and, since some one must hold it, this extra money will precisely absorb the extra saving, which will not be available for the capital market. If you object, Ah, but these people will want to invest this money, the answer is--quite so; this will bring the rate of interest down; liquidity preference is the same and the banks have increased the quantity of money. 2. If the business lets its credit balance decline, again someone [c] else must hold the money released, and that much saving is held off the capital market. But if that someone else wants to invest his money, that will tend to reduce the rate of interest--the root fact here being a decline of liquidity preference on the part of the business which works with a smaller cash balance. You say this is malgré lui. [17] Not entirely. He might sell securities. But what Maynard is trying to do is to find some general concepts covering these particular cases.

pp. 17-18. [18] This is pure Maynard doctrine. 8 (Though you say that in his book you can find no reference to it!) Where he differs from you is in his opinion that there is no reason to suppose that the consequent fall in the rate of interest will be sufficient to restore full employment in any sense of that expression.

pp. 21-4. I have already recognized that this is a matter of fact, about which if you were right, the practical importance (tho not the theoretical interest) of J.M.K.'s theory would be much diminished.


  1. 1. Refers to Henderson, "Mr. Keynes's Theories. Marshall Society, Cambridge. May 2nd, 1936". Of this paper two versions survive: one is a TD, 23 pages with autograph correction (in HP II/182 and HHP 10/2), the other a CcTDI, 24 pages (in HP II/182; a further incomplete copy HHP 10/2). Harrod's comments refer to the latter, which seems to be the first version (in spite of manuscript notes on the front pages in the HHP copies specifying the opposite). A comparison between the two versions offers little clue as to their order, apart from an autograph corrections on a passage of the second version which was identical to the first version, suggesting that if the order were reversed, then the correction (if contemporary) would have been typed in. However, together with the second version Henderson sent an additional paper including some remarks he left out in revising (see letter 554 ). These appear in the 24-pages version, but not in the other one (see note 2 to letter 554 ). The second version, including the autograph additions, was later published as "Mr. Keynes's Theories", in H. D. Henderson, The Inter-war Years and Other Papers, edited by H. Clay, Oxford: Clarendon Press, 1955.

    2. Harrod had been affected by laryngitis (letter to Henderson, 21 March 1936, in HHP 22A/6).

    3. Henderson's paper was to be read at the Marshall Society on 2 May 1936. However, the Minutes of the Society do not record the event (Marshall Society, minutes, Marshall Library, Cambridge, Marsoc 1 and Marsoc 2).

    4. Keynes, The General Theory (1936).

    5. This passage corresponds to pp. 168-69 of the version published in The Inter-war Years.

    6. Having interpreted the General Theory as denying that "a change in the propensity to save could possibly affect the rate of interest", Henderson subjected this proposition to a "simple test": "Suppose you have a closed community in which the propensity to save is nil. Is it not evident that in such a community the rate of interest must rise to a level which is high enough to choke off new investment altogether, and that all that you could do by increasing the quantity of money would be to maintain a vast inflation of prices and money incomes without necessarily securing full employment, and without succeeding in keeping the rate of interest low? But I do not need to take so extreme an illustration." The paragraph concludes as follows: "From a long-period or equilibrium standpoint, the dependence of the rate of interest on the two factors [the propensity to save and the marginal efficiency of capital] which Mr. Keynes deliberately excludes from this theory is undeniable and fundamental." These passages were omitted from the second version.

    7. This passage approximately corresponds to pp. 170-71 of the version published in The Inter-war Years. The paragraph, however, concludes as follows: "To explain that, you must bring back the factors which Mr. Keynes has excluded from his theory; the supply of savings, the demand for them, and the relations between the two."

    8. Harrod, The Trade Cycle, ( 1936:8 ), pp. 120 and 107n.

    9. This passage of the first version criticises Keynes's notion of saving and investment as always exactly equal to each other. Henderson proposed as an alternative "to use the word investment in its ordinary sense to mean investment that is deliberately made" (p. 12). The whole discussion of the "perversity" of Keynes's terminology and its consequences (pp. 11-18 of the first version) does not appear in the second version, but was elaborated into a brief note on "Savings and Investment": see letter 554 .

    10. Dennis Robertson (the correspondence mentioned by Harrod is reproduced here as the exchange beginning at letter 477 ). Harrod also discussed the topic with Kahn (exchange beginning at letter 382 ), Haberler (exchange beginning at letter 378 ), and Durbin (letter 404 ).

    11. The passage runs as follows: "a tendency towards an accumulation of unsold stocks of goods is one of the leading features of a slump".

    12. Henderson claims that the issue is not a purely terminological one:

    • Having used his definition for the purpose of establishing his contention that savings and investment must always equal one another, Mr. Keynes forgets about it for the remainder of his book. Whenever he uses the word subsequently, he uses it in my sense of the term, to mean deliberate investment only. [See note 9 to this letter.] Take his doctrine of the multiplier, which tells us that an increase in investment will entail an increase in income, larger than you might expect at first sight. This is patently untrue if the increase in investment is represented by an undesired accumulation of unsold goods. In that case, income will not increase at all. Clearly Mr. Keynes is using the word investment in its ordinary sense in this connection." (First version, pp. 13-14)

    13. The passage runs as follows: "The investment in the public works may be balanced in whole or in part by negative investment of the involuntary type, that is to say by a reduction of stocks of materials and finished goods below their normal level."

    14. Keynes, General Theory, in Collected Writings, vol. VII, pp. 123-24.

    15. Keynes, General Theory, in Collected Writings, vol. VII, pp. 221 and 375-76.

    16. Henderson explores the consequences of the "supposition that people spend next month only half as much as they spent last month" in order to illustrate the difficulty in explaining the causal link between a divergence between saving and investment and movements in interest rates:

    • Stocks of goods accumulate; and these stocks of unsold goods must be financed somehow. The cash position of any business which finds that its sales are declining must become weaker. Either its balance at the bank will decline, or its overdraft will increase, to the extent of its accumulation of unsold goods. Suppose, for the sake of simplicity, that it normally works with an overdraft, so that what happens is that its overdraft increases. Its need for an increased overdraft is tantamount, it can be argued, to an increase in the demand for capital in the capital market precisely equal to the increase in supply represented by the additional saving that caused the accumulation of unsold goods. Thus, unless there is a simultaneous change in other relevant factors, an increase in saving will leave the relations of demand and supply in the capital market just as they were before, increasing the demand side as much as the supply side, so that there is no reason why the rate of interest should fall.

    17. The passage runs as follows: "A business whose balance at the bank has fallen, say, from £5000 to £2,000, because its sales have fallen sharply has not preferred to become less liquid. It has become less liquid malgré lui, as a consequence of the tendency for savings to exceed investment" (p. 17 of the first version).

    18. Henderson examines the case of a slump caused by saving exceeding investment. This entails "a decline in employment, diminished incomes and a diminished ability to save." Entrepreneurs therefore need less cash to pay wages; moreover, if "retail prices fall, the cash required to make retail purchases will also diminish. Thus money, actual cash, will flow back to the commercial banks; their cash position will be strengthened; and, without any increase in the quantity of money in the ordinary sense or in Mr. Keynes's sense, there will be an abundance of money in the money market which will help to reduce interest-rates. In the case of a boom the opposite phenomenon appears [...]. In the whole of Mr. Keynes's book I can find no reference to or recognition of these well-known and long-discussed phenomena."

    1. a. ALS, four pages on two leaves, and ANI, four pages on two leaves, in HHP Box 22A.

      b. Ms: one or two illegible words.

      c. Ms: «some one» (also in the next occurrence).

1. Not a <demerit> [Henderson's note in the margin].

2. Not in one early stage [Henderson's note in the margin].

3. Why? Higher than if by <+> [b] [Henderson's note in the margin].

4. The <+ don't vary; &> interest rates would rise [Henderson's note in the margin].

5. pp. 123-4 [Harrod's note in the margin].

6. No. Question of a tendency in this direction [Henderson's note in the margin].

7. Interest rates would rise and <+> would fall [Henderson's note in the margin].

8. I cannot find c[hapter]. & v[erse]. So I suppose what I say is rather unfair. But I am quite clear that it is. [Harrod's footnote]

Welcome page

top of page

Return to index of this section

Go to previous page

Go to next page