299. Harrod to Joan Robinson , 23 March 1933 [a]

[Replies to 298 ]

51 Campden Hill Square, W.8.

23 March 1933

Dear Joan

This is awful. We must reach agreement. I will explain what I take us to be quarrelling about. I think that my cost curve for the equilibrium firm is definitely related to the supply price of the industry in your third type of case, which I would agree with you is the most frequent--and, may I add, most probable. [1] It is this third case clearly that is relevant to Trade Cycle analysis, and I believe that this cost analysis has got to be brought in to make the Trade Cycle analysis complete. [2]

There are 2 conditions for long period equilibrium. 1. Profit must not be so high as to attract new firms. 2. Profit on marginal fixed equipment must be equal to the firm's own supply price of fixed equipment. In your article you certainly seemed to say that normal profits are those at which new firms are just not tempted to enter in and to say that in long period equilibrium firms must be earning a normal profit. [3] I hold that in long period equilibrium a firm may be earning less than normal profit in this sense; but it must be earning normal profit at the margin in my sense (2 sup.).

Since it must be earning normal profit at the margin, it is proper to draw my kind of total cost per unit curve. Then cost need not be decreasing in long period equilibrium. Let a rise in demand occur; new firms will not be drawn in; the price will tend to rise and give a supra-normal profit (my sense) on fixed equipment at the margin. Existing firms will increase their fixed equipment, but no reduction of cost will ensue. In my sense costs are constant at the old & new equilibrium. You, by charging a higher rate of profit (viz. the normal in your sense) make them decreasing in both. The fact remains that a rise of demand of your third type will not lead to a lower supply price. I want to draw cost curves for the equilibrium firm so as to bring out what happens to supply price in response to your third type of change in demand.

You say that my curve does not bear any particular relation to price. Here I cannot follow. You say "as soon as demand falls output is determined solely by marginal revenue and marginal cost." [4] I agree. In the short period by short period marginal cost; in the long period by long period marginal cost (taking cost to include my normal profit).

e.g. [b] [fig. 1]

N.B. the long period marg. should never rise but is badly drawn here.

Alternative positions of long period equilibrium:--[figures 2 and 3] [c]

Fig. 2:-- 1. Short period marg. = long period marg. = marg. revenue.

2. Price = short period av. = long period av.

3. Marg. profit = av. profit.

Fig. 3:-- 1. [d] Short period av. = long period av. = short period marg. = long period marg. = marg. revenue.

2. Price > short period av. etc.

3. Av. profit > marg. profit (i.e. > normal profit (my sense)) It must not of course exceed normal profit (your sense).

Very badly drawn, I fear.



  1. 1. Joan Robinson's inquiry into the effects of an increase in demand on prices showed that prices can either increase, decrease or remain constant according to the manner in which the demand curves move. She discussed in more detail the conditions and implications of each case, the third and more frequent one being that of a price fall, which "would occur if the increase in demand were spread evenly over the whole market [...] and if the new firms were set up [...] in between the old firms". This would imply an increase in the elasticity of demand. See Robinson, "Imperfect Competition and Falling Supply Price" (1932), p. 553.

    2. Harrod developed this theme in "Doctrines of Imperfect Competition", where he argued that decreasing costs in the long and short period provide the endogenous de-stabilizing element enabling the economic system to fluctuate without being tied to a state of permanent equilibrium (Harrod 1934:3 , pp. 465-70; see D. Besomi, The Making of Harrod's Dynamics, 1999, pp. 20-24 and passim).

    3. J. Robinson, "Imperfect Competition and Falling Supply Price" (1932), pp. 546-47.

    4. Letter 298 , [jump to page] .

    1. a. ALS, two pages on one leaf, in JVR vii/191/7-8.

      b. At this point, Harrod drew a diagram, with the caption "Long period marg joins long period av. when long period av. become[s] constant". However, the picture was not clear; he therefore crossed it out, and added: "somewhat more helpful figure on the back". Under the figure, he commented: "This fig. brings out the 4 curves better".

      c. Ms: figures 1 and 2. Numbering is silently changed throughout.

      d. Ms: punctuation is often missing in the abbreviations in the description of fig. 3, and is silently reinstated throughout.

Welcome page

top of page

Return to index of this section

Go to previous page

Go to next page