Mr. Keynes' Theory of the "Multiplier": A Methodological Criticism

Haberler’s article is a focused methodological critique of Keynes’s multiplier, not a denial that public works may have secondary effects. He reconstructs Keynes’s formal scheme—income, consumption, investment, the marginal propensity to consume, and the multiplier in wage-units—to argue that the celebrated result is secured by definition rather than by empirical discovery. Once the multiplier is defined as the reciprocal of one minus the marginal propensity to consume, its magnitude has not yet been independently explained.

In reality nothing more than that a new name is given to the multiplier.

The central objection is that Keynes presents a tautological relation as if it were a causal law. For Haberler, to infer a large multiplier from a large marginal propensity to consume is valid only if that propensity has been established apart from the multiplier formula. Otherwise the explanation merely restates the same unknown magnitude in another vocabulary.

By assuming something about the marginal propensity to consume he assumes something about the multiplier, but this is no more an explanation of the multiplier that pauvreté is an explanation of poverty.

Haberler connects this criticism to what he regards as a similar defect in Keynes’s earlier monetary theory, where definitions of saving, investment, profits, and losses were made to carry explanatory weight. The real question, he insists, is concrete: what further employment and income follow from a given new investment, such as public road-building? Keynes’s method, in Haberler’s view, evades that empirical problem by renaming the aggregate relation that needs explanation.

Keynes approaches the problem by means of a terminological roundabout way, that is to say, by giving the magnitude in which we are interested another name.

A major part of the essay turns on the ambiguity of “marginal propensity to consume.” In its formal aggregate sense, it is simply the algebraic counterpart of the multiplier. In its ordinary psychological sense, it refers to people’s tendency to spend additional income. Keynes’s discussions of objective and subjective motives for consumption may illuminate the second meaning, but they cannot determine the first without further assumptions about time lags, leakages, monetary circulation, displaced investment, and expectations.

His terminology exemplifies the paradox of poverty in the midst of plenty.

Haberler’s discussion of saving sharpens the same point. Since Keynes defines aggregate saving as equal to aggregate investment, it is misleading, Haberler argues, to say that income must rise until the public is induced to provide the saving required by new investment. On Keynes’s definitions, the saving is already present in the accounting identity; what remains unexplained is the causal process by which expenditure, production, credit, and employment adjust.

The conclusion is therefore methodological. Haberler grants that Keynes offers important observations about consumption behavior, public works, monetary conditions, and confidence, but denies that the multiplier theory successfully organizes them. The article’s enduring claim is that macroeconomic identities can clarify accounting relations while misleading theory when they are treated as mechanisms of causation.