Hawtrey and the Keynesian Multiplier: A Response to Ahiakpor

I am not persuaded by James Ahiakpor’s denial (this issue) that Ralph Hawtrey was a pioneer in using a geometric series to derive a finite-valued spending multiplier. Hawtrey’s 1928 Treasury memorandum (quoted by Davis 1983, appendix) worked out a numerical example of the change in equilibrium income resulting from the imposition of taxation to pay reparations, with a leakage into imports from successive rounds of spending. His December 1930 Macmillan Committee working paper presented a numerical example of the multiplier, with leakage into saving (in Keynes 1971–89, vol. 13; cf. Davis 1980), while in The Art of Central Banking (1932) Hawtrey replaced his numerical example with an algebraic analysis. This derivation of a finite-valued multiplier stands as an analytical contribution, regardless of Hawtrey’s acceptance of Kahn’s priority in publication (compare with the opening sentence of Ahiakpor’s conclusion) or of Hawtrey’s support for stabilization by means of monetary policy rather than fiscal policy. As I pointed out in my 1997 article, understanding the comparative-static multiplier (understanding the IS goods-market equilibrium condition, to use terminology from later in the 1930s) does not imply belief in the effectiveness of fiscal policy in changing aggregate demand, since that also depends on the LM money-market equilibrium condition. As I noted, Hawtrey later identified liquidity preference as Keynes’s key innovation, that is, as something not previously part of Hawtrey’s own analysis. Similarly, Hawtrey’s analytical contributions in 1928, 1930, and 1932¹ are in no way nullified by his dissent from Keynesian economics in the 1950s, a dissent recounted by Ahiakpor. Indeed, one need not move ahead twenty years in Hawtrey’s career: I drew attention to a 1933 article by Hawtrey expressing views on fiscal policy contrary to those of Keynes and Kahn.

Ahiakpor does not accept Hawtrey as a multiplier pioneer because he attributes to Hawtrey a general equilibrium analysis inspired by Say’s law in which “saving is not a leakage from the expenditure stream as the Keynesian argument claims; and the classics too argued that saving is spending” (this issue, 895). Ahiakpor’s note 4 provides quotations from David Ricardo and John Stuart Mill expounding the law of markets. What Ahiakpor fails to notice is that for Hawtrey to have accepted the law of markets, in the form that Ahiakpor defends, would have made nonsense of his whole approach to economic fluctuations and stabilization, because in that case changes in aggregate demand would have no effect on output whether brought about through public works or through monetary policy. While Hawtrey rejected fiscal policy as a means of influencing output and employment, he consistently argued for a monetary theory of economic fluctuations and for the effectiveness of monetary policy.

Ahiakpor seems to hold that, since saving and investment are equal in equilibrium (neglecting foreign trade and government), an increase in the rate of saving does not act as a leakage. The Keynesian analysis, on the contrary, distinguishes decisions to save from decisions to invest, and looks to the level of income as the equilibrating force bringing saving into equality with investment (more generally, as in IS-LM, the level of income and the interest rate would both play a role). In critiquing the Keynesian multiplier analysis, Ahiakpor would be better advised to turn to Irving Fisher’s two-period consumption model (Fisher 1907, 409), which shows that, given perfect credit markets, the present discounted value of expected lifetime income is the relevant budget constraint for consumption, which implies a small marginal propensity to consume out of transitory changes in income. Even Keynes, in a Treasury memorandum during the Second World War, found his absolute-income consumption function too simplified for practical policy analysis, arguing that “a remission of taxation on which people could rely only for an indefinitely short period might have very limited effects in stimulating their...