770. E. H. Phelps Brown to Harrod , 25 May 1938 [a]

33 Museum Road, Oxford

25 May 1938

Dear Roy,

I have read this address [1] with great enjoyment, and I think that you can be quite sure that it is well suited to its purpose: I do not see how the exposition could be improved. There are two passages which I might mention as particularly pleasing to me: the first the account of the right place of methodology with which you open; [2] the second, though this a relatively minor matter, your treatment of "the comparability of different men's utilities". [3] (I remember a discussion with Jay in which he convinced me that the proposition "Smith would sooner have one more orange than two more apples" cannot claim greater certainty than "The Aga Khan will feel the loss of a shilling less keenly than Hodge the labourer".)

The remarks which occur to me should perhaps be taken rather as contributions to the discussion than as suggestions for alterations, but perhaps I might set them down. There are two main points.

The first concerns the distinction which you draw between "The Economic Criterion" and "Static Theory". [4] (a) The distinction between them is clear in this sense, that the former begins with a "should"--that persons should get what they prefer--whereas the latter is concerned only with matters of fact. Should I only be bringing back the failure to distinguish the two which you regret, if I said that the application of the criterion is bound up with the static theory?--for example, Pigou's principle of the equimarginal returns to a productive factor [5] seems to me only a particular case of the Law of Demand as you formulate it in its most general form--homogeneity against heterogeneity--as the principle of Static Theory. Thus it seems to me as if the "analytic map" of the first department is really the static framework of the second. (b) This difficulty about the dividing line is enforced by the treatment you give to Static Theory itself. You regard it as providing causal laws: [6] but causal laws require a reference to more than one point in time, and this reference seems to me to be excluded by definition from static theory. [7] Thus the elementary supply curve and demand curve diagram gives us the price at which quantity supplied equals quantity demanded; but we cannot say that the market will tend to settle at this price until we know the sequence of adjustments--until we know, for example, that there is no lag there such as gives us a "cobweb" diagram: but this means dynamics. In a later passage you say (p. 40), "for every proposition purporting to relate to the succession of events it must be possible to point to the empirical evidence. ... If empirical evidence is lacking, the proposition can be no more than a definition of the terms which it employs". [8] Can this be reconciled with the view that Static Theory yields causal laws? [9]

Thus my unconsidered impulse would be to make the distinction between (1) static theory yielding the simultaneous analytic map and worked out by the "scarcity principle", and (2) the bringing of [b] this to bear on problems of prescription by introducing the premise that men should get what they prefer.

The second remark expresses an uneasiness rather than an argument. It concerns the consideration of a distinct economic end, a criterion of economic welfare. (a) You consider an example (p. 15) [10] in which there are ends other than that of economic welfare to be sought, but the particular case you consider seems an extreme one, in that the non-economic end is given complete priority and has no more or less about it. Such a case is surely easier to deal with than the usual problem, in which we have the possibility of a greater or less degree of self-sufficiency (or other such "political" end) and correspondingly a less or greater degree of economic welfare: and we have to decide what is the best pattern on the whole. This decision can be given only in the light of a single end over all, that of "the good society". Indeed I feel that prescription can never be made except in the light of a single conception of welfare, and that "economic welfare" cannot be separated out. 1 (b) This difficulty is associated in my mind with the likelihood that in fact the movement of a people towards one partial end such as economic welfare will bring with it consequences affecting the attainment of other partial ends. Economic welfare, for instance, points to a higher standard of living: but suppose the sociologist finds that as societies attain a higher standard they also show a "weakening of morale", an atrophy of the creative energies, a flight from parenthood? Particular outcomes apart, it does seem likely that the attainment of greater economic welfare will have its necessary social consequences, both welcome and unwelcome, and therefore I feel that we cannot prescribe with reference to an economic criterion alone.

There are two very minor textual points I might mention. p. 5, just below halfway, sentence beginning, "Or, he may show that conclusions ..." This is deep philosophical water, but I didn't see the force of the antithesis: e.g. mayn't a proposition in Euclid be both "the laboriously won fruit of a complicated chain of reasoning" and "all the time implicit in the premises"? [11] --p. 29, top line. There is the Giffen case to be allowed for here. I take it that this case arises because bread and meat are inferior and superior forms of the same "commodity", the minimum energy ration; and the only question of principle raised is the impossibility of delimiting "a commodity", which is not in point in the context. Perhaps though a saving parenthesis might be put in. [12]

I had no idea this letter would be so long; but let me just add, that I do think your paper is full of interesting contentions which make up a real contribution to the subject, and the test of this is that it deals with the central problems without ever blanketing them under balanced generalities or stirring up the flames of tribal warfare.



We look forward to seeing you here on Monday.

  1. 1. Phelps Brown refers to "Scope and Method of Economics", Harrod's presidential address to be read before Section F of the British Association for the advancement of science, which was later published as Harrod ( 1938:15 ).

    2. Harrod, "Scope and Method of Economics" ( 1938:15 ), pp. 384-85.

    3. Harrod, "Scope and Method of Economics" ( 1938:15 ), p. 396.

    4. The "economic criterion [...] deals with the nature and authority of the prescriptions given on the basis of the analytical map" (Harrod, "Scope and Method of Economics", 1938:15 , p. 389), which "arises [...] from a simultaneous chart or survey of the economic field, and the main work of the cartographer is analysis and classification" (p. 387). Static theory "is the theory of value and distribution which considers the scope and validity of the causal knowledge derived from the Law of Demand" (p. 389).

    5. Pigou, The Economics of Welfare (1920), part II, chapter 2, section 5, pp. 117-18, cited in "Scope and Method of Economics" ( 1938:15 ), pp. 386 and 394.

    6. Harrod, "Scope and Method of Economics" ( 1938:15 ), p. 398 top.

    7. Phelps Brown refers to the distinction between statics and dynamics suggested by Ragnar Frisch, according to which in a static system "all the variables belong to the same point of time", while in dynamics "we consider not only a set of magnitudes in a given point of time and study the interrelations between them, but we consider the magnitudes of certain variables in different points of time, and we introduce certain equations which embrace at the same time several of these magnitudes belonging to different instants" ("Propagation Problems and Impulse Problems in Dynamic Economics", in Economic Essays in Honour of Gustav Cassel, London: Allen & Unwin, pp. 171-72; see also "On the Notion of Equilibrium and Disequilibrium", Review of Economic Studies 3, February 1936, p. 100).

    8. Harrod, "Scope and Method of Economics" ( 1938:15 ), p. 406 top.

    9. In his address, Harrod introduced two distinctions at once: one between the analytical (purely deductive) map, which from the logical point of view yields nothing more than tautologies, and the causal laws which must have an empirical basis. The first distinction thus regarded the scopes of induction and deduction ("Scope and Method of Economics", 1938:15 , pp. 386-87). The second distinction was between statics and dynamics, the former concerned with a state of rest and the latter with the increase or decline of economic magnitudes (pp. 402-3). The two distinctions overlap, as in both statics and dynamics one must distinguish between the purely deductive part of the theory, and the axioms--based on generalizations derived from "a few basic empirical laws" (p. 404)--on which such deduction is based. See D. Besomi, The Making of Harrod's Dynamics (1999), chapter 5.

    As the draft of the Presidential address does not survive, it is not clear whether Phelps Brown's failure to distinguish the two problems originated from an obscure statement of the earlier formulation.

    10. Harrod, "Scope and Method of Economics" ( 1938:15 ), p. 391, last paragraph.

    11. Harrod, "Scope and Method of Economics" ( 1938:15 ), p. 385, second full paragraph.

    12. Harrod inserted a footnote to p. 399 referring to Marshall's discussion of the Giffen case.

    1. a. TLS with autograph corrections in pen, three pages on three leaves, in HP II-78E.

      b. Ms: «bringing this».

1. Harrod ticked this sentence in the margin in pencil.

Welcome page

top of page

Return to index of this section

Go to previous page

Go to next page