Gottfried Haberler, The Theory of Comparative Cost Once More (1929)

Haberler’s theoretical article defends comparative cost against two contemporary objections: that monetary trade can sustain an equilibrium contrary to comparative advantage, and that specialization need not increase aggregate output. Its two-part structure separates the conditions governing trade from the productive gains obtainable through specialization. Haberler preserves the principle’s relevance to the economic choice between free trade and protection while clarifying what production costs alone can establish.

The first section addresses James W. Angell’s argument that international exchange follows money prices, whose relationship to labor costs may reverse comparative advantages. Haberler provisionally accepts the labor theory of value underlying Angell’s example. A country can have lower money prices in both of two commodities, but its exports of both do not establish a stable counterexample: the other country must finance its imports. Introducing additional commodities that it can export restores the relevant comparison. Comparative advantage must be assessed between exported and imported goods, not simply between two goods that might both belong to the export group.

Haberler generalizes this reasoning algebraically. National price–labor relations and the exchange rate establish a common threshold separating exports from imports. The resulting inequalities imply that each exported commodity has a comparative cost advantage relative to each imported commodity. Money prices therefore mediate the principle rather than overturn it. Yet the proof also reveals its limits:

To be sure, the line that separates export from import articles (designated by the quotient l₂/l₁r) cannot be drawn, if we know only the costs of production.

Costs constrain the ordering of exports and imports without independently locating the dividing line. Haberler thus extends Mill’s recognition that comparative costs require supplementation by demand: indeterminacy concerns not only exchange ratios but also which commodities enter each side of trade. This qualification strengthens the theory by narrowing its claim appropriately.

It is by no means a serious criticism, only an elaboration of the classical theory.

The second section considers A. F. Burns’s contention, following Pareto, that complete specialization need not produce an aggregate surplus. Haberler accepts the proposition but rejects the inference that it defeats comparative cost. His example contrasts the United States with Luxembourg: even if America has a comparative advantage in steel, abandoning all wheat production in exchange for Luxembourg’s small output would be absurd. The appropriate adjustment is asymmetric—Luxembourg specializes in wheat while America shifts only a corresponding portion of its resources into steel.

Therefore partial specialization leads always to a more attractive combination, while under complete specialization this is not necessarily true, as Mr. Burns has demonstrated.

The conceptual distinction is between advantageous reallocation and indiscriminate complete specialization. Under the example’s assumptions, partial specialization permits an aggregate output gain where complete specialization may fail. Haberler adds that, with many traded commodities and constant costs, joint production is confined to the commodity at the import–export margin.

The article closes by identifying a deeper issue than either objection it has answered:

A much more serious objection is the one advanced by Pareto, by Professor Viner recently in his brilliant criticism of Angell's book (p. 622), and by Professor Mason in his suggestive article, "The Doctrine of Comparative Cost" (in this Journal, xli [1926], 63); namely, the objection that the comparative-cost theory builds on an old-fashioned and abandoned labor theory of value.

Haberler postpones that reconstruction to another occasion. This article’s contribution is consequently precise: it reconciles monetary price guidance with comparative cost, distinguishes cost rankings from a fully determined trade pattern, and separates gains from specialization from the demand for complete specialization. Its defense of the classical principle depends on making these distinctions explicit, while leaving its labor-value foundation open to further revision.