334. Harrod to E. Cannan , 12 December 1933 [a]

[Replies to 330 and 333 , answered by 335 ]

Christ Church, Oxford

12 December 1933

Dear Cannan

I was just about to write to you concerning letter no. 1., left on one side till the rush of term should be over, when I received letter no 2.

With regard to the first, I agree with you that there is no need to apply the discreditable word deficit to the balance of expenditure covered by the issue of money for inflationary purposes. [1] But I am not so clear that the issue should take the form of notes. The proportion between people's deposit accounts at banks and notes in wallet or cash box is a matter of convenience and no purpose could be served by attempts to distort this proportion. [2] If the govt [b] began paying judges etc. with currency notes, they would just pay them into the banks and draw cheques on the account so acquired. The notes would--except in so far as the greater purchasing power led to a rise of prices and some consequent demand for more 1 currency, but not for as much more currency as the increase of bank deposits--go back to the Bank of England and raise the deposits of the commercial banks with the B of E. Precisely the same result would follow if the govt. arranged for a Ways & Means advance of sufficient amount and paid the judges with cheques on the B of E. Purchasing power would be increased by the same amount and all subsequent results would be the same. The deposits of the Joint Stock Banks [c] at the B of E would be increased and the B of E assets would be increased by so many govt i.o.u's: on your method the i.o.u's would repose in the Issue Dept and an equivalent number of notes in the reserve of the Banking Dept: on my method the i.o.u's would repose in the Banking Dept. On my method the absurdity of paying the judges in notes would be avoided--that is the only difference. 2

But as a matter of fact I dont think the Ways & Means advance is the right method of inflating. For the result would be greatly swollen cash reserves of the Joint Stock Banks and they at their wits end to know to increase their other assets in proportion. And if much inflation of this sort was attempted, e.g. £200 or £300 million, their position would become intolerable. They would have to add to their investments by £2000 or £3000 m. to keep their conventional proportion 3 , which would be an absurdly top-heavy position.

Therefore I suggest that the right method of inflating by say £200 m. is to get a ways and means advance of say £20 m. or get the B of E to buy £20 m. extra Treasury Bill and sell about £180 m. Treasury Bill to the Joint Stock Banks. The govt. would then have injected £200 m new money without putting the banks into an anomalous position. There would possibly be some consequential demand for extra cash to meet higher pay-rolls or even to allow people with banking accounts to carry more money about with them and this would be dealt with simply. The B of E would transfer the Treasury Bills to the issue dept. and pay out such extra notes as were needed. [3]

One further difference between my way and yours. On my way the govt is involved in an interest charge on the Treasury Bills--anyhow on those sold to the Joint Stock Banks--for those sold to the B of E could be held in the Issue Dept by arrangement. I dont think this is a serious practical objection, because interest rates are so low. By the time they rose, it would be time for the govt to think of paying off debt. But it is a difficulty in principle and I am not sure that the proper solution isnt to make the Joint Stock Banks hold interest free govt obligations in a certain proportion to their capital in a time of inflation. Since these assets would be additional to the assets they would otherwise acquire and could only possibly be held because the govt was in fact inflating I dont think it would be unfair to the Joint Stock Banks to say--you will make no profit on these extra govt. obligations. This smacks rather of Douglas-ism or some such heresy, but shouldnt, I think, be turned down solely on that account.

If you insist on paying judges in notes you get the same result, provided you say to the Joint Stock Banks--you must expect your cash basis to be doubled or trebled, but do not get alarmed and do not seek to acquire other assets in proportion.

The aggregated bank balances would then stand at, say,

Cash

400

Call money

200

(A)

Discounts

200

Advances

800

Investments

400

instead of the usual

Cash

200

etc.

(rest same).

It would not I submit be desirable to get to this position:

Cash

400

Call Money

200

Discounts

200

Advances

800

Investments

2,000

which would be reg[arde]d to restore the proportion of cash to other assets to the usual level.

On my plan this balance sheet would be as follows

Cash

200

Call money

200

(B)

Discounts (non interest bearing)

200

Discounts (interest bearing)

200

Advances

800

Investments

400

A and B are really the same except for formalities.

I wonder if you agree with this.

You also say "no one seems to doubt that R[oosevelt]. can raise prices by a sufficient issue of greenbacks." [4] I think there is a great deal of muddled thought about greenbacks. Surely the real truth is that there is no doubt that R can raise prices by meeting a sufficient amount of govt. expenditure by paper rather than taxation.

If the govt. issued greenbacks and used the proceeds to pay off certificates of indebtedness held by the Federal Reserve Banks, this would have no effect on prices at all.

With regard to Cole's book you have me at a disadvantage as you have probably read more than I have. I cant help feeling that your quarrel with Radice is to some extent verbal. [5]

What I do submit with earnestness is that there is not much difference of principle between what happens when a bank "issues notes" to me and when it credits my account with so much. Apart from hand to hand transfers, my account can only be credited with so much (or I can only receive notes from the bank) either (1) if I deposit 4 gold or (2) receive a loan. (I call bank investments a species of bank loan). If the banks gave money away the position would be different. If the govt owed the banking system it could create deposits by paying cheques to its servants uncovered by taxation. It so happens that a govt can issue notes without owing a bank. The note issuing dept. can indeed be thought of as a sort of bank on which other banks are not allowed to encash their drafts (the notes).

It is because govt.s havent happened to open a bank of their own and pay out drafts on it which the public and other banks can only convert into credit accounts at the said bank, that the fallacy has arisen that there is something fundamentally different between the issue of notes and the crediting of so much to the banking account of some one.

To your final query about notes outstanding and cheques outstanding, [6] I would say that the cheques outstanding correspond to the number of notes (unknown) at any time going through the post. With notes outstanding must be compared outstanding deposits drawable on by cheque.

I have written much--too much. I have probably written unguardedly. I hope you endow this with the sense it was intended!

Yours very sincerely

R. F. Harrod.

  1. 1. Letter 330 , [jump to page] .

    2. The banks' "endeavour to reserve a definite ratio between their cash holding and the volume of their deposits" is discussed in Harrod, "Currency and Central Banking" ( 1933:13 ), pp. 138-40.

    3. A proposal along similar lines was later suggested in a letter to The Times: "Meeting a Trade Recession" (Harrod 1938:11 ).

    4. Letter 330 , [jump to page] .

    5. Letter 333 , [jump to page] . Reference is to E. A. Radice, "Commercial Banks and Credit", in G. D. H. Cole (ed.), What Everybody Wants to Know about Money, London: Gollancz, 1933, pp. 162-225.

    6. Letter 333 , [jump to page] .

    1. a. ALS, six pages on three leaves, in CP 1033/208-10.

      b. Punctuation after the abbreviations is not consistent, and is reproduced here as in the original.

      c. Ms: «joint stock banks».


1. Underlined by Cannan, who pencilled in the margin: "kitten chasing its tail if the <+> is excluded".

2. Cannan drew a horizontal line in the margin.

3. Cannan marked this sentence in the margin with two exclamation marks.

4. Cannan underlined the work, and pencilled a question mark in the margin.


Welcome page

top of page

Return to index of this section

Go to previous page

Go to next page