Viner on International Trade
Since the comparative absence of methodological discussion in the literature of the theory of international trade is a condition whose persistence need not, in my opinion, be deplored, this methodological note is presented in the spirit of Henry Sidgwick's famous lecture on the futility of lectures. One of the methodological criticisms which has occasionally been made against the theory of international trade is that its exponents have not formulated an adequate definition of its scope and objectives, so that it fails to deal with matters properly within its range and perhaps concerns itself with questions which do not fall within its legitimate boundaries. I find it difficult to conceive what useful purposes the formal definition of the scope of a discipline can serve, except the purposes of editors of encyclopedias and administrators of educational institutions, whose responsibility it may be to prevent overlapping, to obtain full coverage, and to arbitrate jurisdictional disputes. No damage is likely to be incurred by economics if serious consideration of these jurisdictional questions is confined to those for whom it is an unavoidable occupational responsibility.
It is indeed arguable that energy spent in trying to define the proper limits of disciplines is often worse than energy wasted, since preoccupation with such definition often arises from an inadequately suppressed desire to confine analysis to one's own private set of assumptions and concepts. In the absence of precise delimitation of the scope of a field there will, it is true, tend to be much overlapping and much raggedness of boundaries. Overlapping, however, is, outside of encyclopedias with crowded pages and the curricula of universities with strained budgets, an evil of a minor order. The waste of effort which may result from it is more than counterbalanced by the mutual stimulation of the overlapping disciplines which it tends to provide, and by the safeguards which it sets up against degeneration of the individual disciplines into formal and lifeless academic systems whose orginal organs of contact with the problems of real life and with the development of thought in other fields have become atrophied through more or less deliberate disuse. The opposite evil, too restricted a scope, with consequent neglect of promising areas of investigation, is a more genuine one, and definition may conceivably serve to expose its existence and to indicate its specific nature, but a sample demonstration of how the discipline would be improved by an extension of its scope would seem to be a much more effective means of securing such extension. It is surely reasonable to expect the economist who urges a novel program of investigation upon his fellows to demonstrate his own faith in its possibilities and to give some concrete evidence that this faith is not misplaced by himself executing some portion at least of his program.
The discussion, however, often turns not on the propriety of the existing limits of the theory of international trade, but on the appropriateness of the doctrine's label to its contents. It has been repeatedly objected that the term “international trade” or “foreign trade” in the label is misleading, on the ground that the theory deals with trade between regions, irrespective of whether or not these regions are “nations” or “countries.” Edgeworth remarked that: “International trade meaning in plain English trade between nations, it is not surprising that the term should mean something else in political economy.” 1 That the theory was not concerned solely, or was not applicable solely, to trade between sovereign nations was recognized from the start. The writers, from Hume on, when expounding some doctrine in this field in terms of trade between countries would stop to point out that it was applicable also to trade between regions or provinces within a country. John Stuart Mill, when asked whether Ricardo was correct in stating that the same rule which regulates the relative value of native commodities does not regulate the value of the products of different countries, replied in the affirmative, but said that he would substitute “places” for “countries” in the proposition.2 Bastable toyed with the idea of substituting “interregional” for “international,” but concluded that: “‘interregional’ would prove a troublesome word; it is better therefore to adhere to the old term.3 Ohlin adopted this “troublesome word,” but in giving to his important book the title “Interregional and International Trade” he seems to imply that even for him the former term does not fully embrace the latter.
Finding flaws in labels is much easier than finding patently superior substitutes. If what has gone under the label of the theory of international trade was simply an investigation of the spatial aspects of trade, “interregional trade” would be a highly appropriate label. It would have the merit that it stressed the main methodological difference between the theory of domestic (or “closed economy”) trade, as ordinarily formulated, and the “theory of international trade,” namely, the assumption by the former of a single market without spatial dimension and the assumption by the latter of at least two spatially distinct markets, each without internal spatial dimensions, but with substantial obstacles to the movements of factors of production and, in some cases, of commodities across the frontiers. But if the theory of international trade were distinguishable from the theory of domestic trade only by the recognition by the former, and the exclusion by the latter, of the existence of space in some abstract sense, it would have been surprising if someone had not long ago offered a demonstration that the theory of international trade could be absorbed into the theory of trade in general with gain to the latter and without loss form the abandonment of the former by introducing into the equations of the theory of trade in general additional s (for “space”) terms in the manner in which the complex economic problems arising out of the temporal flux of phenomena have recently been solved for us by the introduction into the equations of general equilibrium theory of magical t (for “time”) terms. Examination of the actual assumption, explicit and implicit, of the theory of international trade reveals, however, that the role of “space” in the theory of international trade is too varied and elastic to be adequately disposed of by any such simple stratagem.
It has been alleged that what differentiates the theory of international trade from domestic-trade theory is solely the assumption in the former that there are transportation costs for commodities or factors and abstraction from transportation costs in the latter. Objection is then made to the differentiation of the two theories on the grounds that: there is no such difference in the facts; that if any differences exist in fact between internal and international transportation costs the differences are relative rather than absolute; and, finally, that transportation costs are often in fact greater between regions within a country than between countries. If these considerations are well-founded and relevant, there would nevertheless still be room for two theories, one abstracting from transportation costs while the other makes them its special concern. There would be no point, however, in labeling the latter the theory of “international” trade, and “interregional” would seem a highly appropriate substitute. But transportation costs are commonly abstracted from in both theories, and while spatial obstacles to movement are a special concern of the theory of international trade, these spatial obstacles do not, or need not, consist of transportation costs.
In their rare methodological dicta, the classical exponents of the theory of international trade explained that they were assuming international immobility and complete internal mobility of the productive factors. Mobility assumptions were important for part of their theorizing, but the mobility which they assumed to be absent internationally was a different kind of mobility from that which they assumed to be present internally. What underlay their analysis was the assumption of international place immobility of the factors of production, irrespective of occupation, and the assumption of internal occupational mobility of the factors of production, irrespective of location, and for a large part of their analysis only the former assumptions was significant. Much of the criticism of the mobility assumptions of the classical theory of international trade as unrealistic is irrelevant because it fails to note this distinction between types of mobility.
It will be conceded at once that contrast between an international immobility and an internal mobility, if valid at all, is valid only as a relative and not as an absolute contrast. But a relative difference in mobility, provided it is a substantial difference, suffices as a foundation for a separate theory of international trade. The differences in degrees of mobility of the factors of production, moreover, seem obviously to be great when countries are being considered and to be minor, or non-existent, or in the reverse direction, when neighboring regions within a country are being considered, if the mobilities being compared are place mobility between areas, on the one hand, and occupational mobility within areas, on the other.
There is from the entrepreneurs' point of view perfect occupational mobility of a factor of production within a country if any desired quantities of its services can be hired or purchased by any industry at the same terms as by any other industry. In the long run, occupational mobility of “disposable capital” and of natural resources must approach closely to perfection. Because of occupational preferences on the part of labor and because of non-competing occupational labor groups, there appears to be, however, a substantial departure from perfect occupational mobility of labor, whether from the entrepreneurs' or the laborers'4 point of view, even in the long run.
An appropriate criterion of perfect long-run international mobility of the factors is the existence of sufficient place mobility to prevent persistent international differences in their money rates of return in similar occupations. There is obviously, in this sense of the term, zero mobility of natural resources: existing immigration restrictions suffice, today at least, to guarantee almost zero international mobility of labor; but in normal times at least there is a high mobility of capital and of entrepreneurial skill. These international immobilities of labor and natural resources are all that is needed as a basis for a separate theory of international trade even if there were perfect international mobility of capital and imperfect internal occupational mobility of all the factors, although any variation in the mobility assumptions as a matter of course makes necessary a variation of some portion of the analysis and conclusions of the theory.
There are additional reasons why “international” is a more appropriate term than “interregional” for the theory of international trade, given its traditional range of interests. In the development of the monetary aspects of the theory of the mechanism of international trade, the classical economists had generally in mind a particular area, England, partly because it had a single monetary and credit system and partly because it was for them an area of special interest. “Countries” fit these two considerations much more generally than do regions within a country. In the analysis of gain from trade, attention was definitely centered upon particular boundaries, enclosing areas of community of interest, and these areas were also generally countries or nations. As Sidgwick remarked: “it is only in the case of foreign trade that the investigation of the conditions of favorable interchange excite practical interest; because it is only in this case that there has ever been a serious question of governmental interference with a view of making the interchange more favorable.” 5 In inductive investigations within the field of the theory of international trade, the unit of investigation has almost invariably been a “country,” partly because this was an area of special interest to the investigator, and partly because the concentration of public interest on country units has resulted in a relatively much greater supply of statistical information for such units than for “regions.” The subsidiary field of study of international economic policies confines its attention to the obstacles to economic intercourse, natural, institutional, statutory, administrative, which are associated with or correspond to national frontiers: import duties, immigration restrictions, differences in commercial law and commercial practices, differences in language, tastes, customs, etc. The theory of international trade is therefore to a large extent a genuine theory of trade between nations. Both by design and as an incidental by-product, it is also in large part an economic theory of regionalism. It is often something in between these two. Except for zealots in definition, this flexibility in its boundaries has not been a source of difficulty or confusion.
Williams has complained, however, that the assumption in the theory of international trade of international immobility of the factors prevents it from taking into account the important economic consequences of the substantial international migration of the factors which have actually occurred throughout the past.6 It must be admitted that a theory which always assumes complete international immobility of the factors would be as inappropriate by itself for the economic analysis of the effects of migration of these factors as a theory of trade in general resting on strictly static assumptions has proved to be as an instrument for the analysis of business fluctuations. But this would constitute a valid criticism of the theory of international trade only if the latter professed to answer questions relating to the effects of international migration of the factors of production. The theory of international trade has departed sufficiently from its usual adherence to the assumption of international immobility of the factors of production to provide us with the only body of analysis of any pretensions relating to the mechanism of transfer of capital. But with the myriad long-run economic effects of the international migration of capital, or of labor, the theory of international trade has not dealt nor pretended to deal. While there is no doubt a valuable contribution still to be made by the theory of international trade in this connection, it seems to me that it is to the economic theorist, the economic historian, and other specialists, that we must mainly look for significant results in this field. Particularly in the field of immigration of labor, to whose vast specialized literature, as far as I know, no international trade theorist except Ohlin has made any contribution of consequence, it would probably sound like passing strange doctrine to the specialists in the field that they really were encroaching all the time on the legitimate boundaries of the theory of international trade. But it may be taken for granted that the specialists in industrial history or in immigration would welcome with open arms any genuine contribution to the analysis and solution of their problems which any specialist in international-trade theory has it within his power to make.
Williams makes another, and to me a completely novel, criticism of the theory of international trade, on the ground that, as a theory of benefits from territorial division of labor, its conclusions contradict its premises of internal mobility of the factors. Trade means national specialization for the world market. “Specialization is thus the characteristic feature and the root idea of international trade. But specialization is the antithesis of mobility, in this case of domestic movement of productive factors.” 7 What I understand him to mean is that national specialization, by leading to greater population and greater accumulation of capital within a country than could be productively employed within that country if access to foreign markets were cut off, brings into being (presumably by domestic growth as well as by immigration, since England is used as an illustration) such great increases in the amounts of the factors of production that the factors in large part have no satisfactory alternative to production for export, would with the cessation of foreign trade have either to starve or emigrate, and therefore have no internal occupational mobility. The effects of foreign trade on the amounts and rewards of the factors postulated by Williams are in kind quite in keeping with the expectations of the classical writers, though I cannot recall any instance of forecasts on their part so optimistic in degree. In default of careful investigation, I have no reason to doubt that the classical economists in general, and not only John Stuart Mill, whom Williams cites, overlooked the adverse effect on average occupational mobility of a great expansion in capital and population dependent on foreign trade for their employment. The alleged contradiction between the mobility assumptions and the conclusions of the theory of international trade, nevertheless, seems to me to be spurious. The relevant mobility assumption of the theory is not that occupational mobility is a consequence of national specialization, but that it is a prerequisite thereof, which instead of being a questionable proposition approaches closely to being a truism. Notwithstanding the passage cited by Williams from Mill, which seems to deny the possibility that foreign trade may result in a loss of average occupational mobility by the factors of production, I feel certain that Mill—or Ricardo, or Cairnes, or Marshall, or Taussig—would gladly have given assent to the proposition that a country can profitably employ more capital and can support a larger population at a given standard of living under trade than with foreign trade cut off, which seems to be the gist of William's argument.
Papers relating to political economy, II, 5.
Political Economy Club, Minutes of proceedings, VI (1921), 291.
Theory of international trade, 4th ed., 1903, p. 12, note.
The ability of labor freely to choose its occupation and the ability of entrepreneurs, whatever their occupation, to hire labor at uniform rates are, of course, different, though related, concepts. Both are significant for the doctrine of comparative real costs, but only the latter is relevant for the theory of mechanism.
The principles of political economy, 2d ed., 1887, p. 216.
J. H. Williams, “The theory of international trade reconsidered,” Economic journal, XXXIX (1929), especially pp. 205–09. It has conspicuously failed to do so in his own case. Cf. his Argentine international trade under inconvertible paper money 1880–1900, 1920.
“The theory of international trade reconsidered,” loc. cit. p. 203.