769. Harrod to N. Kaldor , 24 May 1938 [a]

[Replies to 767 , answered by 771 ]

Ch[rist]. Ch[urch, Oxford].

24 May 1938

My dear Kaldor

I think you must excuse me for taking a long time to understand your letter; you drew some diagrams, but, as you did not put in the new demand curves after the rise of real wages, I could not see--and still cant--how they illustrate your point.

What I piece together is this. An individual may employ increments of capital by producing more or using more per unit of product. As he has a rising supply schedule of capital, he uses available supplies in extending output and does not "deepen" the productive process (to use Hawtrey's word) [1] as much as he would if he could get an unlimited supply of capital at as low a rate as he can get some.

When a rise of real wages occurs you assume that this restricts his market. [2] He then proceeds to use his available supplies of capital to deepen [b] his process. But this is not inconsistent with my view. Suppose in the old equilibrium the marginal supply price of capital to him is 8%. Then in the old equilibrium the marginal yield from deepening was 8%. When his market is restricted capital becomes available for deepening at less than 8% and he deepens accordingly. In the new equilibrium the marginal yield of deepening is say 7%. Owing to the restriction of the market he is not using so much capital altogether and so can move back to a lower point on his supply schedule. But what is this but saying that for him the rate of interest has fallen to 7%? However much his market was restricted the fact remains that unless he could get capital for deepening at less than 8% (i.e. at less than the old marginal rate) he would not deepen. You may say, Oh, but the market rate may still be what it was before, say 5%. But if the argument is to have any general significance, we must suppose that this fall in the marginal supply price of capital occurs for every or anyhow for the representative individual. And if that happens I see no meaning whatever in saying that the market rate has not fallen. If each individual can get the marginal capital he requires at a lower rate I say that the market rate has fallen.

You may say that the market supply schedule of capital need not have altered. I did not say that it need. What has happened is that the market price of capital has fallen (owing to a fall in the demand for it).

Your argument amounts to this. If a rise in real wages restricts the amount of labour in employment, then there being more capital available per unit of labour employed the rate of interest will tend to fall. Given the initial effect of a rise in real wages I do not deny this. What I do deny is that there can be any substitution of capital per labour unless the rate of interest does fall.

You may stress the distinction between the yield of consols and the marginal supply price of capital to individual employers. There may be something in this. But I should say that in a condition of general plethora such as that envisaged all rates would be likely to be affected.

To recapitulate. We compare 2 equilibria A and B, in which technology is the same but in B real wages are higher. This is possible because we take it that more of the other factors are used per unit of labour. If the total supply of other factors are not increased the employment of labour in B will be lower. If the amount of capital used in B is less than that in A by the same amount that labour is less, the amount used per unit of labour and the equilibrium rate of interest is the same. If more capital is used per unit of labour the equilibrium rate is lower.

My main point, which I think is often overlooked, may be put simply thus. If real wages are raised for manual workers and the real wages for brain workers are the same, there will be a tendency to use less [3] brain workers per unit of manual workers in the productive process. If real wages are raised for manual workers and the rate of interest remains the same, there will not be a tendency to use more capital per unit of manual workers. (You might argue that if the rate of interest remained the same and the real rate of wages went up, the real reward for capital had gone up too: it would if you measure a unit of capital as waiting for time t in respect of a unit of labour: but not if you measured a unit of capital as waiting for time t in respect of a unit of consumption goods).

I hope I have made my point clear.


Roy Harrod

  1. 1. Hawtrey, Capital and Employment (1937), p. 36.

    2. At this point of the manuscript, Harrod crossed out the following passage: "(This assumption may be legitimate in the case of an individual but it would be fallacious to argue to production as a whole, i.e. to the representative individual.)".

    3. Presumably a slip of the pen for "more".

    1. a. ALS, six pages on three leaves, in NKP NK/3/30/86/6-11. Photocopy in HP NC.

      b. Ms: «deepening».

Welcome page

top of page

Return to index of this section

Go to previous page

Go to next page