DYNAMICS, TRADE AND MONEY IN THE CORRESPONDENCE BETWEEN ROY HARROD AND DENNIS ROBERTSON

The scientific correspondence between Harrod and Robertson was initiated by Harrod’s criticism of Robertson’s Banking Policy and the Price Level (1926).⁷ Harrod first wrote on 18 May 1926 (letter 2) raising at once the following “salient point”: Much of your argument depends on the view that justifiable expansions and contractions as defined by you are desirable. Why are they desirable? You give reasons on p. 22 why you think some instability in output desirable. But the reasons mentioned there (and I can’t find any others) don’t seem particularly directed to show that the special form of instability constituted by the so-called “justifiable” expansions and contractions is desirable. They seem to me to show that perhaps some instability, that, presumably, of less degree than we have been accustomed to in the past, is good, but by no means precisely how much is good. Thus, suppose the “hypothetical group member” or “the actual workman” of p. 19 were able to govern output according to their own self interest, there would still, according to the arguments of ch. 2, be some instability. Would not that be enough? Or if you want more, why stop at the “justifiable”? Why not have some of that due to “secondary” causes? It seems to me that you have been led away by purely aesthetic interests to identify that more moderate amount of instability which we really need (as shown on p. 22.) with that which we would get: (i) if secondary causes were removed; and (ii) if control of output stayed where it is now – in the hands of the entrepreneur. I don’t see how you can say to the banks more than “damp down fluctuation a bit, but leave some fluctuation, as that is healthful for the body economic”.He added two notes to his letter, in the first of which he commented upon the four proposed courses of policy outlined by Robertson on pp. 25–26 of his book. In the second note Harrod suggested that Robertson’s calculations in Appendix I to Ch. 5 of Banking Policy assumed the following behaviour of the public: (i) People do not allow for the effect of their withholding on the price level (this is reasonable). (ii) They are ignorant as the future course of inflation (or do nothing to meet it). (iii) On this basis they decide what withholding is necessary to restore H, decide that it would be too much effort to restore it at once, and…spread the restoration over K – 1 days. It so happens that by choosing K – 1 their 2 errors (or failures to take everything into account) cancel each other out, and they do effect the restoration in that time. If K or K – 2 had been chosen, this would not have been so.Harrod further argued that Robertson’s “so-called reasonable assumption of a restoration in K – 1 days is purely arbitrary,” and that “all this reasoning is rendered of doubtful value by the fact that we must suppose an alteration in view as to ‘the appropriate proportion between Real Hoarding and Real Income’ during the process of inflation. Not only will people not replenish H at once, but they may well voluntarily reduce it.”