295. Harrod to Joan Robinson , [mid-March 1933] [a]

[Answered by 296 ]

51 Campden Hill Square, London W.8.

[mid-March 1933] [1]

Dear Joan

Perusal of the current number of the E.J., especially Shove's sec. 3 [2] suggests to my mind the following point; will you tell me if it contains an obvious blunder?

If entry into a trade is not free, your definition of normal profit [3] necessarily involves not a particular rate, but a range of rates, e.g. not 6%, but anything between 8% and 4%. But in order to draw an average cost curve, it is necessary to specify a particular rate of profit.

How shall we choose which? You might say--let us take the rate which is actually being earned; unfortunately this will be different for every level of output, the rate for the equilibrium level cannot be known, until the equilibrium has been determined and the equation of the curve is necessary to determine it. An approach might be made algebraically but it would be very complicated, and a much simpler geometric method is open to us.

There is no doubt which is the appropriate rate of normal profit to choose; namely that rate which the firm itself regards as a sufficient inducement to make a long period outlay. In the long period a firm will push its long period outlay to the point at which the marginal revenue due to it yields what it regards as a proper rate. This is the normal rate appropriate to the construction of an average total cost curve. It may be greater or less than the average yield of its fixed equipment. It may be greater or less than profit in other trades, but it will probably bear a closer relation to profit in other trades than the average yield.

The difference between us is that in drawing a total cost curve, you assumed that fixed plant is reckoned as earning that rate of profit which it [b] does earn on the average, I that rate of profit which in the long period it does earn at the margin. If these rates co-incide the total cost curve will be tangential to the demand curve, but not otherwise. [4] Yet within the sphere of the "otherwise," equilibrium in every sense is possible. i.e. marginal revenue due to an increase of fixed plant will yield a return on it such as just to justify the hopes of the firm, and both the marginal return and the average return to fixed plant will lie within the limits, inside of which neither new firms are attracted nor old firms fall away.

With my definition of normal profit, constant costs are possible. In equilibrium, in that case, marginal cost = average cost = marginal revenue. Price exceeds marginal revenue and average profit exceeds marginal profit. In computing costs, fixed plant is taken as required to yield a certain rate of revenue--the firm's own supply price of fixed plant--and the marginal outlay on plant does yield this rate of revenue.

In perfect competition, but only in perfect competition, marginal rate of profit on fixed plant tends to be equal to average rate of profit.

I hope this point is clear. I ought to have seen it as soon as your article appeared. [5]

I am afraid I cant make anything of the Prof.'s contribution [6] --can you? Perhaps light will dawn on further study



  1. 1. Undated. However, the reference to the March issue of the Economic Journal and the date of Joan Robinson's reply indicate that the letter was written towards the second week on March 1933.

    2. A. C. Pigou, "A Note on Imperfect Competition", Economic Journal XLIII, March 1933, pp. 108-12; G. F. Shove, "The Imperfection of the Market. A Further Note", ibid., pp. 113-24; J. V. Robinson, "[A Comment on Shove's Note]", ibid., pp. 124-25. Reference is to Shove's discussion of Joan Robinson's views on "normal profits" and equilibrium (pp. 119-21), as expressed in "Imperfect Competition and Falling Supply Price" (1932).

    3. Robinson defined normal profits as the level of profits which neither attracts new firms to a particular industry, nor induces any firms to leave the industry: "Imperfect Competition and Falling Supply Price", p. 546.

    4. Robinson had suggested that the two curves are tangential to each other in a letter to Harrod of 13 December 1931 (letter 224 , [jump to page] ), where she pointed out a mistake in Harrod's diagram on p. 571 of "The Law of Decreasing Costs" ( 1931:2 ). Harrod accordingly corrected the error in "Decreasing Costs: An Addendum" ( 1932:6 ), p. 492.

    5. Harrod later expounded the arguments debated in this exchange with Joan Robinson in "A Further Note on Decreasing Costs" ( 1933:7 ). In particular, the remark that Robinson's definition of normal profits involves a range of values in on pp. 337-38; Harrod's alternative proposal is on p. 338; he disputes the necessity that demand and total cost curves are tangential to each other on pp. 339-40; and he rejects the conclusion that in imperfect competition a firm must be subject to decreasing costs on p. 340. These conclusions were also restated in "Doctrines of Imperfect Competition" ( 1934:3 ).

    6. Pigou's article cited in note 2 to this letter.

    1. a. ALS, two pages on two leaves, in JVR vii/191/3-4.

      b. Ms: «is».

Welcome page

top of page

Return to index of this section

Go to previous page

Go to next page