Fritz Machlup, “Relative Prices and Aggregate Spending in the Analysis of Devaluation” (1955)

This single-author journal article is a methodological critique of devaluation theory. Machlup’s target is not devaluation as policy, but the analytical claim that Alexander’s “income-absorption” approach supersedes the older elasticities or relative-prices approach.

The purpose of this article is to examine the supposedly superior approach and to give a comparative evaluation of both.

Machlup first grants the weaknesses of the relative-prices method. Supply and demand curves for foreign exchange cannot simply be treated as stable, because devaluation changes costs, incomes, income distribution, and the allocation of resources. Yet he argues that Alexander’s aggregate-spending framework has parallel weaknesses. Its key identity, (Y \equiv A+B), is useful because it makes visible that the trade balance is the difference between national income and domestic absorption.

The trade balance is negative when the nation absorbs more than its income.

But Machlup’s central point is that this accounting truth is not yet a causal theory. Alexander’s model asks how devaluation changes income, how income changes absorption, and how devaluation directly changes absorption. Machlup reconstructs this carefully—idle-resources effects, terms-of-trade effects, cash-balance effects, redistribution, money illusion, and other absorption effects—before arguing that the model smuggles relative-price reasoning back in while claiming to replace it.

The most important omission is resource reallocation. Alexander recognizes fuller employment and terms-of-trade changes as sources of real-income change, but not the possibility that devaluation raises real income by moving existing resources into more valuable uses.

There are three ways in which an increase in real national income can be achieved: through fuller employment of the available productive resources; through their better utilization and more economic allocation; and through more favorable terms of trade.

This move restores the price system to the center of the analysis. Devaluation alters the relative prices of imports, exports, import substitutes, and domestic goods; these changes induce substitution in consumption and production. For Machlup, such substitution effects cannot be reduced to a marginal propensity to absorb. They may change real absorption even when money absorption is unchanged, and they are indispensable for understanding how resources actually shift toward exports or away from imports.

A second conceptual move is Machlup’s attack on reasoning from identities. National-income equations classify ex post magnitudes, but they do not by themselves explain what causes what.

Such equations usually serve a useful purpose in aiding the organization of the analysis. But they may easily tempt an analyst into "implicit theorizing," illegitimately deducing causal relationships, and overlooking the shifting meanings of terms in different contexts.

This criticism structures the middle of the article. Machlup distinguishes real income from output in an open economy, domestic-money trade balances from foreign-money trade balances, and causal relations from accounting equivalences. A devaluation may make a deficit look worse in domestic currency while improving it in foreign currency; an equation alone cannot answer the policy question. Hence his blunt warning:

As a matter of fact, the equation gives very little insight into causal relationships.

Machlup also brings monetary policy back into the picture. A continuing import surplus without autonomous capital inflow presupposes reserve sales and credit creation. Thus the effect of devaluation cannot be assessed without specifying the behavior of the money supply, banking system, and fiscal authorities. Spending propensities are not more “given” than elasticities: they can shift with expectations, policy, inventories, government budgets, and the devaluation itself.

The article’s final comparison is deliberately balanced. Machlup does not defend a naïve elasticities approach; he accepts that aggregate spending matters. But he rejects the claim that absorption analysis can replace supply-and-demand analysis. Foreign supply and demand conditions, domestic substitution possibilities, and terms-of-trade effects all require elasticities, even when the argument is framed in aggregate terms.

Neither of the two “alternative” sets of tools can be spared; both are needed.

The article’s relevance lies in this synthesis. It is a mid-century intervention against both mechanical Marshallian elasticity formulas and overconfident Keynesian aggregation. Machlup’s lasting contribution is to show that devaluation analysis must combine relative prices, aggregate spending, monetary policy, and institutional constraints rather than elevate any one of them into a complete theory.