When Does Trade Benefit Everyone?

Lessons from Krugman (1981)

Liberalizing trade should increase the total product of the economy, but it is not guaranteed to be a Pareto improvement. Some people may be made worse off by the change, and with very specific parameters and an assumption of declining marginal utility of wealth, it may even be possible for welfare to decline. Topalova (2010) found, for example, that trade liberalization in India increased extreme poverty because the very poor were bound to particular areas, and could not easily relocate into different areas. How can we predict when trade liberalization will be a Pareto improvement, and when will it not be?

Paul Krugman’s 1981 paper, “Intraindustry Specialization and the Gains from Trade” provides a clear and tractable model for determining when it will be, and when it will not. Krugman, before he became more famous as a political columnist, revolutionized trade theory. His 1979 paper, “Increasing Returns, Monopolistic Competition, and International Trade”, reformed our view of trade away from factor-based Ricardian and Heckscher-Ohlin models of trade.

The Ricardian view is simple. Suppose two countries could produce two goods — cheese and olives, say. If one is better at producing cheese and the other olives, naturally they will specialize and trade. This trade is somewhat miraculous insofar as even if one country were better at producing both, it would still benefit both if they still produce only one good, (if we make a small assumption of declining marginal utility from any given good. This is, in essence, the assumption that consumers have a taste for variety). Portugal might produce cheese and olives better than England, but trade is still advantageous. Presume Portugal could produce six olives and three cheeses, and England one olive and two cheeses. Left to their own devices, without trade, Portugal could consume six olives and England two cheeses; or if they prefer an equal amount of both, Portugal would allocate 2/3rds of its resources to produce cheese and consume two of each, and England would allocate 2/3rds of its resources to olives and consume 2/3rds of each. Suppose Portugal produced more olives, and England only cheese. Then Portugal could exchange one olive for one cheese, and everybody has more. Trade occurs anytime when companies or people or countries have different tastes, preferences, or productive abilities, with the implication that identical traders have no use for trade. The precise goods which are produced is determined by the factors which each country has in abundance relative to others — land, labor, or capital.

Krugman asks what happens if, instead of perfect competition, we find ourselves in a state of monopolistic competition? There are economies of scale which are internal to the firm. A simple example would be if there were a fixed cost to start a business, and then it cost the same to produce every additional unit. The larger the firm, then, the lower the average cost per unit. Trade will occur even when everyone has identical tastes and endowments.

This solved many of the weird inconsistencies of the standard trade models of the time. No longer must we be confused why both Japan and the United States produce automobiles, despite having different factor endowments. Honda specializes in producing Hondas, and Ford specializes in producing Fords. Both of these have declining marginal costs, and we are richer for exchanging the two than if we didn’t.

Now, in the later paper, Krugman wants to answer when trade liberalization will be painless. For convenience sake, let’s say there are two factors of production. It goes without saying that if both factors should increase, then the transition is painless. If the two countries have identical factor endowments, then in a state of total autarky they will produce the same mix of good. Liberalizing trade will cause the industries to differentiate their products, but will not change the utilization of their endowments. By contrast, if a country has different factors of production that another, then liberalizing trade will cause some factors to lose out. What’s more, the more substitable the goods are, the less similar the factors have to be for it to be a Pareto improvement.

This gives us clear predictions for what will happen when we liberalize trade. If we liberalize trade with a country like China which has far more labor and much less capital available than us, industries which are labor intensive will suffer, while those dependent on capital will gain. This is exactly what we saw, as shown by Autor, Dorn, and Hanson (2016) in their paper on the China Shock. They even quote Krugman (and Obstfeld, it’s from a textbook) on the very first page! “Owner’s of a country’s abundant factors gain from trade, but owner’s of a country’s scarce factors lose. … Compared with the rest of the world the United States is abundantly endowed with highly skilled labor and … low-skilled labor is correspondingly scarce. This means that international trade tends to make low-skilled workers worse off — not just temporarily, but on a sustained basis.”

This is hardly a call for restricting trade. In the long run, allowing more trade always increases incomes. But if we are concerned about the political effects, the first places to start are trading with countries which have similar endowments. In addition, we should allow the migration of workers, just as we allow the migration of capital. The incomes of the owners of the scarce factors suffer in the long run only if they are unable to move where their skills are more valued. We can see too how restrictions on the building of housing cause not only losses in the quantity of quality of housing, but in preventing people from exploiting their comparative advantage. Being in a world of the second-best may change what is optimal, but we should strive for the best of all possible worlds.