Some Problems in the Pure Theory of International Trade

Gottfried Haberler’s “Some Problems in the Pure Theory of International Trade” is a single theoretical article in static international-trade theory and welfare economics. Its scope is deliberately restricted: two countries, two commodities, given external terms of trade, and no dynamic or strategic terms-of-trade analysis except in the later discussion of infant industry.

The present article discusses certain elaborations and applications of the now familiar and widely used presentation of the theory of comparative cost in terms of opportunity cost.

Haberler begins from the standard production-opportunity curve and insists that the tangent at a production point has two distinct meanings: a market exchange ratio and a marginal transformation ratio. The free-trade case depends on their coincidence under competitive conditions, flexible factor prices, and no external economies or diseconomies.

It is very important to keep these two meanings apart, for only under special conditions will the two ratios be identical.

The welfare argument is not made by treating “the community” as a single chooser. Haberler is wary of community indifference curves when trade changes income distribution. Instead, he adopts the compensation criterion: trade raises welfare if redistribution could make everyone better off, even if actual losers remain.

It is not necessary that income will actually be redistributed so that everybody will in fact be better off; there will practically always be some individuals who are worse off than before.

Much of the article is then an exercise in controlled deviation from the benchmark. Haberler does not deny market imperfections; he argues that naming imperfections is not enough to justify protection. The analyst must show that the relevant distortion is persistent, important, and works against freer trade rather than in its favor.

A mere enumeration of possible imperfections and deviations from the ideal case does not prove more than the possibility that certain controls might be beneficial (provided of course that they are efficiently administered—which amounts to assuming quite a lot).

His most important conceptual move is to separate factor immobility from factor-price rigidity. Immobility alone may reduce gains from trade but need not destroy them; the serious welfare problem arises when wages or other factor prices cannot adjust, so resources become unemployed before their opportunity cost has fallen to zero.

It can be easily shown, however, that what really causes trouble and may make trade detrimental and justify protection is rigidity of factor prices, which may or may not be associated with immobility of factors.

In that case the price line no longer expresses the social marginal transformation ratio. Trade can move the economy to a lower welfare position, and protection may be a second-best remedy. Haberler is careful, however, to state this as a possibility, not a general refutation of comparative advantage.

What we have proved is the possibility of an unfavorable outcome, not its necessity.

The later sections extend the same logic to external economies and diseconomies. Here the distortion is not unemployment but the divergence between private and social cost. If producers do not internalize economies available from expanding an industry, the country may specialize in the “wrong” direction. Yet Haberler again places the burden of proof on protectionists.

It is one thing, however, and an easy one at that, to point out the possible consequences of a divergence of private and social cost and to cite instances in which such a divergence is likely to arise.

The infant-industry argument is treated as related but distinct. Its claim is intertemporal: protection may move production along the present curve in a way that shifts the future curve outward through learning, skills, and improved methods.

The essence of the infant-industry argument is that a movement on the transformation curve will bring about an irreversible shift of the curve itself.

Even here Haberler’s conclusion is symmetrical and skeptical. Learning may occur in export industries as well as import-competing ones; external effects may be negative as well as positive. The article’s lasting relevance lies in this second-best discipline: trade intervention can be justified only by diagnosing the specific domestic distortion, not by invoking imperfection in general. Its core thesis is that free trade remains the benchmark when prices reflect social opportunity costs, while departures from it require exact proof that private prices mislead production in a welfare-reducing direction.