¹ Of course, the now not-so-new welfare economics sought to circumvent the “short-run” difficulty by asserting that more trade is preferable if those who benefit can overcompensate those who are injured. By implication, there was nothing wrong with economic theory, and it was left to the politicians to legislate compensation if they were so concerned about losses. However, it will be shown that actual compensation would alter the market outcome of opening trade, and so various policy dilemmas would still remain. (See Kaldor, N., “Welfare Propositions of Economics and Interpersonal Comparisons of Utility,” Economic Journal, 09, 1939 Google Scholar, for the original statement of the compensation argument.)

² There are at least three major sources of gain from more external trade. First, and traditionally the most emphasized, is the reallocation of resources to increase national specialization in production. Second, and only explicitly noted in the last decade or so by Haberler and others (see below), is the gain from internationally redistributing a given national output. Third, and perhaps more important but seldom specifically mentioned, is the addition to the national output that depends on imports of producer goods. The size of a nation's production block is itself dependent on trade. This last idea has been developed by the author in an article as yet unpublished.

³ Samuelson, P. A., “The Gains from International Trade,” this Journal, V, no. 2, 05, 1939, 195.Google Scholar His thesis is in general terms and so logically includes the extremely short-run case of immobile resources considered here. Factor prices and quantities are ignored in this comment; Samuelson handles them analogously to commodity prices and quantities.

⁴ Professor Haberler, G. (“Some Problems in the Pure Theory of International Trade,” Economic Journal, 06, 1950)Google Scholar also rejects Samuelson's argument, for apparently similar reasons, and terms it anthropomorphic.

⁵ Ibid.

⁶ Lipsey, R. G., “The Theory of Customs Unions: Trade Diversion and Welfare,” Economica, 02, 1957.Google Scholar

⁷ At Q the consumer indifference curves of Y and X are tangent, and they are both on the contract curve, CC′; the only equilibrium point on this contract curve is Q, because only there is a line from A tangent at the same point to an indifference curve of both X and Y.

⁸ Graphically, if one wishes to suppose that each household produces some x and some y, the price would have to revolve around a point inside the box, indicating the quantity of x and y produced by both households. In all cases it is assumed that households ignore their possible influence on price. There is full competition in buying and selling output.

⁹ For example, Q is on the collective consumer indifference curve X″. All the individual consumer indifference curves, which when summed at the given price, AP″, gave Q, are “known.” For each unit change in x it is conceptually possible, for each household, to ascertain the indifferent change in y. These two related and indifferent changes, when aggregated for all X households, provide a point on the collective curve X″, away from Q, which can be designated O. Then an Q stock of x and y, uniquely distributed, can make the X households as well and no better off than they are at Q. All other distributions will not leave each and every household indifferent.

¹⁰ The expenditure of duty collections by government is ignored. Perhaps such duties as it does levy against certain countries are prohibitive. Or protection may come from import quotas.

¹¹ The first requirement is that of Kaldor (“Welfare Propositions”) and Hicks, (“Foundations of Welfare Economics,” Economic Journal, 12, 1939).Google Scholar The second and additional requirement is that of Scitovsky, T. (“A Note on Welfare Propositions in Economics,” Review of Economic Studies, 11, 1941).Google Scholar The extra condition prevents the possibility that, if compensation is not actually paid, each situation may seem preferable to the other.

¹² There is no assurance that, if trade were opened, the QR units of x and y that the X households could afford to pay would, when added to the position of the Y households at S, raise them again to indifference curve Y″. The fact that indifference curve Y′ intersects QR at T is immaterial here, for the Y households are not at T but at S in the open situation. Hence the need to employ obverse reasoning.

¹³ An important critique of this entire approach has been made by Little, I.M.D. (A Critique of Welfare Economics, Oxford, 1957 Google Scholar, esp. chap, vi) in which he stresses as an additional test of a “good” change that the redistribution of welfare be “desirable”; also Mishan, E. J. (“A Survey of Welfare Economics,” Economic Journal, 06, 1960)Google Scholar has recently summarized the logical problems inherent in any use of welfare economics.

¹⁴ Samuelson, , “The Gains from International Trade,” 195.Google Scholar

¹⁵ Perhaps because X′ and Y′ have a coincident segment.

¹⁶ And if open price weights are used with the quantities on X″ and Y″ that make these indifference curves tangent to a line parallel with AP′, the closed situation will always appear exaggeratedly worse than the open one, even though R and S almost coincide and there be little gain from trade.

¹⁷ The paired consumer indifference curves for X and Y households, which together comprise the contract curve, must have slopes other than AP″ at all points other than Q if there is to be a unique equilibrium without trade. From Q, along CC′ towards the Y origin, these tangencies will probably have slopes representative of slightly higher y prices for x. And the X household consumer indifference curve passing through H (see X″ in Figure 2) will have a slope representative of a lower y price for x than that at R on X′.

¹⁸ Always recognizing that transformation possibilities must also be considered in the real world, which will mean more trade volume, smaller changes in household incomes, and reduced price changes in the long run.

¹⁹ Initially, although such a supposition is convenient rather than essential, one can assume that a group trades freely or not at all with other groups.

²⁰ In this case it is necessary to think of the X households of A and D together having an offer-consumption curve, and similarly the Y households of both groups, with a resultant price of pad; the reasoning of the previous sections indicates that outside trade and accompanying price changes will benefit one class of households more than the other class is injured.

²¹ A group and a subset are here analytically equivalent. Both contain X and Y households. A subset includes two or more groups.

²² Among other things, the group must then worry about further enlargements of trade that will injure it but which it cannot subsequently prevent.

²³ Many writers, comparing only the absolute costs of imports, have assumed that each country should favour universal free trade because it can then obtain every good at the cheapest source after allowing for transportation.

²⁴ Additional references are Samuelson, P. A., “Evaluation of Real National Income,” Oxford Economic Papers, 01, 1950 Google Scholar; Baldwin, R. E., “The New Welfare Economics and Gains in International Trade,” Quarterly Journal of Economics, 02, 1952 Google Scholar; and Little, I. M. D., “Welfare and Tariffs,” Review of Economic Studies, XVI (2), no. 40.Google Scholar