95. H. B. W. Joseph to Harrod , 26 November 1925 [a]

[Continues at 96 ]

New College, Oxford #

26 November 1925

My dear Harrod,

You know much more than I do about economic theory, & why in particular the notion of marginal value/utility [b] is essential. [1] I suppose it is wanted for explaining how the prices of goods in a market come to be adjusted, & what is meant by the elasticity or inelasticity of demand. I don't think anything I said inconsistent with what is true in those branches of economic theory, so far as the theory makes use in them of the notion of marginal utility. Clearly I am prepared to give more for 10 lb. of sugar when I have none, than when I already have 20 lb., if I have the same period of consumption & occasions for use in view. But I should like to make a great distinction between the measurement of a jam-manufacturer, who asks whether if he spends so much in buying more sugar, he will be able to make a larger profit that will recoup him for his expenditure, & the so-called measurement of the house-holder, who asks whether the extra satisfaction which he will get through his additional purchase of sugar will recoup him for the loss of other satisfactions that the money could have bought him. I don't think there's any measurement in the second case. But that doesn't prevent my preferring to spend my money one way rather than another. That there is "a homogeneous ground of all preferences;" this homogeneous something is capable of "being measured", I utterly dispute. [2] I don't think that is the issue in Eth. I. vi., because Aristotle wasn't imputing to those he criticized a quantitative doctrine of good. [3] He was denying that particular goods are qualitatively homogeneous, but it didn't occur to him that it was necessary to deny that you could measure them if they were. Triangles are qualitatively homogeneous; but there isn't more triangularity in one set of triangles than in another [c] . Of course you could measure their <duration>. I don't even maintain that I can't prefer a picture by <Francis Haley> to a photograph of the King more than I prefer the latter to an oleograph of the King. And if you are going to speak (metaphorically) of distance between them, I will allow that the first distance is greater than the second, but not that it is so many times as great. All this business of measurement requires that you could properly use numbers. It isn't necessary that your numerical estimates of the amount of this homogeneous something in your alternatives should be precise. Lack of precision will of course make your results inaccurate, for as long as there is as much precision in the measurement of one alternative as of the other that doesn't matter, since you want the result merely for determining which is greater. But I don't think much assignment of values occurs. I only know of the recorded instance, to be found in H. Spencer's autobiography: [4] & he preferred the alternative for which he brought out the lesser quantity of advantage. Certainly I am not aware of ever having proceeded that way, & I suppose I make preferences. What I want to do first is to become as explicitly conscious as I can of the nature of the alternatives. I may help myself by setting off against each other certain details in each; but those acts are <minor> acts of preference, & are not preceded by ascertaining in the details set off quantities of a homogeneous something. It is possible that the details are themselves homogeneous--e.g. if I have to choose between two routes for a journey, with sea-passages in each, & if I like the sea, it will be a ground of preference that the passage is longer by one route than by the other. But if one route includes a sea-voyage, & the other a chance of seeing the Dresden gallery, I don't & can't reduce these details each to so much of the same, before discovering which I prefer. That is what your theory seems to me to require, if you really think it out.

A paper on the commensurability of all advantages (why not "of all goods"?) might be a very good subject for the Philosophical: [5] only I should like so to convert you that you wouldn't wish to write it. As to mischief done to students of economics in Modern Greats by raising these issues, I hardly know what to say. Up to a point I think the sciences do well to neglect metaphysical issues. It is of no consequence to an arithmetician to determine whether an integer is a class of classes or an instance of an Ã-sòmblgto q Ãrihmá q. But when he begins to philosophize & tell me that it's the first, I'm prepared to think I have as good a right to an opinion as he. As [d] long as the economist is only considering how the fact that men's preferences are affected by the quantities they already possess, of the things proposed for exchange or purchase, will in turn affect the prices of things & the quantities of them bought & sold, I've nothing to say. But when he begins to tell me about a balance of sacrifice & satisfaction, & really to philosophize about his economic facts, I think I ought to investigate his doctrine as a question of philosophy (or psychology--but it's not merely that--it concerns things as well as the soul). And since economics is after all a study of human behaviour, which arithmetic isn't, the philosophical issues are much more germane to it, than to arithmetic.

Would you like to come & dine one night, & talk it over? Tomorrow or Sat. would do, or next Tues. If tomorrow, I suppose I must say dress: each day, 7.0 p.m.

Yours ever

H. W. B. Joseph

  1. 1. This discussion originates from a paper Joseph had read at the Political Economy Club. Harrod reported of F. Y. Edgeworth's reaction, and of his appeal to the younger generation in defence of the notions of total and marginal utility, in a letter to J. M. Keynes of 25 March 1926 (here reproduced as letter 106 , [jump to page] ).

    2. The subject was taken up again in an exchange between Joseph and Harrod regarding Harrod's paper on "Morals and Arithmetic" read at the Philosophical Society in November 1926: see in particular letter 129 R.

    3. Aristotle, Nicomachean Ethics, I:6 (in The Complete Works of Aristotle. The Revised Oxford Translation, ed. by J. Barnes, vol. 2, Princeton: Princeton University Press, 1984)

    4. H. Spencer, An Autobiography, 2 vols, London: Williams and Norgate, 1904.

    5. This remark probably played a part in stimulating Harrod to write "Morals and Arithmetic" for the Philosophical Society, here reproduced as essay 3 : see note 2 to this letter, and letters 122 , 123 and 129 R.

    1. a. ALS, ten pages on three folded sheets (the first two written on four sides), in HP IV-580.

      b. Ms: «utility» was written, probably as an afterthought, between the lines above the word «value».

      c. This sentence was inserted as an afterthought.

      d. Ms: «As as».

Welcome page

top of page

Return to index of this section

Go to previous page

Go to next page