The Myth of Comparative Advantage

Today's "free trade" is not only the last remnant of laissez-faire -- it is its least deserving remnant, full of wholesale foul play, deception, currency manipulation, predatory techniques, and other violations of its rules.
This post was published on the now-closed HuffPost Contributor platform. Contributors control their own work and posted freely to our site. If you need to flag this entry as abusive, send us an email.

One of the main pillars of international (as well as general) economics is the "Law of Comparative Advantage" (from now on, for brevity, "LCA"). When Paul Samuelson, a preeminent economist and the father of the theory of international trade, was asked to name just one economic principle that was both non-trivial and true, he -- according to legend, with lightning speed -- retorted: "How about the Law of Comparative Advantage?"

Well, every legend has to be debunked. He had to think for three years to make that repartee. Clearly, true and not-trivial tenets of economics are not abundant.

Anyway, the issue is more serious: LCA is not a true "law" -- at best, it is a rule of thumb, not often applicable. (For instance, by the 1980s Paul Krugman has shown that LCA may at best apply only to some commodity-like products, where the "learning curve" doesn't matter.) LCA violates common sense and reality in many other directions, too.

(A.) Paul Samuelson, in the earlier editions of his Economics textbook cautiously states LCA as follows (slightly truncated here): "If any region specializes in the products in which it has a comparative advantage (greatest relative efficiency), trade will be mutually profitable for both countries." He warns, however, that this doctrine "is unassailable" only "when properly stated." So what does "properly stated" mean? Nearly a full large page of the textbook is then devoted to "Qualifications." Almost each one is currently impossible to meet. For instance, prices of the traded products, wages, and currency rates have to be fully and instantly changeable, so that trade, in fact, becomes barter. Movement of money and technology across borders is prohibited, and so on.

Clearly, all requirements combined would be met substantially later than hell would freeze. (In later editions, co-authored with William Nordhaus, he is less cautious. For instance, as dollars flew out of the US borders, the constraint of zero trade balance was deleted. Pity.)

(B) LCA in Samuelson's formulation says only that "trade will be mutually profitable"; it does not say, "Both countries will be better off." And indeed, how can it state that?

Economics forbids any interpersonal comparisons of utilities even between individuals; therefore it cannot construct a utility function for a society. How can it then construct a worldwide "intercountry" utility function, compare well-being of entire nations under much more complex conditions, and declare that "the world economy as a whole, and each country in it, benefit from trade"? Suppose that a country has indeed been more profitable, but its society is now more polarized. How can one state anything about its well-being?

Samuelson was well aware of that problem. For instance, in his much-talked-of 2004 article in the Journal of Economic Perspectives about further limitations of LCA (it breaks down under the weight of technological progress in a developing country, as Ralph Gomory and William Baumol have noted in their 2001 book Global Trade and Conflicting National Interests), he admits the fundamental difficulty even for the purely economic part of the problem, but proposes a solution: to compare geometric means, calculated for each country over its trading sectors of industry. OK, but geometric mean is much older than economics. If it indeed can help comparability, why cannot we make interpersonal comparisons of utilities between individuals; if it cannot help even in this modest task, how can we use it for entire countries? (The well-known logical conundrum: "If this is the meat, where is the cat? If this is the cat, where is the meat?")

The whole discipline of international economics becomes largely an exercise in futility.

(C) Each trading country has a large number of losers from trade. They further complicate comparing the countries. What can we do about them?

In 1825, at the ripe age of 19 (clever much beyond his age but perhaps a little lacking in worldly experience), John Stuart Mill proposed that the country's winners willingly compensate its losers. Has anyone heard of any volunteers? Or of any actual case of such compensation, substantial and long-term, during the past 185 years?

But the difficulty goes even deeper than actual implementation. Before compensating all losers, we have to define who wins and who loses. Two prominent economists, Nicolas Kaldor and John Hicks, proposed two diametrically opposite criteria for that definition. Then, in 1941, Tibor de Scitovsky had shown that a free trade policy could be beneficial only if it simultaneously satisfied both conflicting criteria. A little difficult, isn't it? As a free trader Douglas Irwin puts it in his 1996 book Against the Tide: An Intellectual History of Free Trade, "This undermined the hope that a single and unique metric could be employed to evaluate the welfare effects of various policies." In essence, this is the same problem as in the (A) above. Even if we forget about social implications, no serious economic problem has an unambiguous answer. There exist no good one-handed economists, requested by Truman.

(D) The trade balance-of-payments of a country is a common good, and, under present conditions, it is hit whenever anybody buys an imported good. Clearly, that balance is an externality. To take into account externalities, Arthur Pigou proposed the so-called "pigovian taxes and subsidies"; Baumol and others continued his work. How can economics then avoid putting on the trade balance a pigovian tax or a constraint? (In Capitalism and Freedom, Milton Friedman is not afraid though of trade deficits, because "What would the Japanese exporters do with the dollars? They cannot eat them, wear them, or sleep in them." I do not understand why China does not listen to this Nobel, the most influential economist of the end of the 20th century. It stupidly accumulates 2.5 trillion of smackers and does not sleep in them. Instead, it buys Africa wholesale, buys everything that is not nailed to the floor and has some natural resource, economic, or geopolitical value, becomes a monopolist in rare earth metals, and does some other dopey things -- e.g., every year increases its military budget by up to 20 percent, maybe more, if their accounting is not quite up to par.)

(E) In the real (not abstract economic) world not all trading partners are created equal. Various countries follow different trading policies, use or do not use different dirty tricks. Accordingly, they have to be dealt with differently. Plus, international trade is an ages-old, venerable, and most effective tool of diplomacy: it would be extremely stupid not to make distinction between friendly and adversarial trading partners. Trading sanctions is one example; containing an adversary (a la Kennan) -- another; giving trading privileges to our ally Japan -- the third. Trading balance with each country, or each trading block, must be subject of individual consideration. Pigovian taxes for different countries or trading blocs may not be identical.

(F) Furthermore, if we equalize conditions for all countries (say, by selling import certificates at auctions), a country with large currency reserves may overbid other contestants, buy all certificates, and -- after using whatever is needed for its own exports -- sell the rest to other countries at cheaper prices, but with a potential of political or economic blackmail.

(G) Many economists transfer the beneficial impact of free trade to far-away and nebulous future. But they always forget to discount these future benefits because of delayed gratification, as done in every cost-benefit analysis.

(H) Even would the theory of international trade be correct, its projected results would be completely unacceptable to majority of Americans. These conclusions are therefore hushed up. For instance, it has been rigorously proven (by three groups of eminent economists) that, even under full employment, and even if all theoretical and practical difficulties were "assumed away," wages in all free-trading countries would be equalized at a sufficiently low "weighted" level. Basically, importing goods is equivalent to letting in the workers that produced these goods.

Do not forget that wages in some countries are 30 to 40 times less that in the US. Until recently, the average daily wage of an American worker was 135 dollars; that also was the average monthly wage of a Chinese worker. Not for nothing one of Samuelson's "Qualifications" of LCA was a request of a moderate difference in wages between countries. (In 1999-2000 the US labor organizations demanded that US would not admit China into WTO without "addressing its systematic and egregious violations of workers' rights," in particular -- until its wages satisfy standards of international labor bodies. Clinton prevailed; W happily followed in his steps, patriotically exporting the US manufacturing base. Another example of "triumph of hope over experience." One of the worst and most costly triumphs in the history of mankind.)

Under persistent unemployment, and under the conditions of the new industrial revolution, it will be much worse than under full employment: in every category of labor, excessive supply will overwhelm any possible growth of demand. And, as pointed out by Alan Blinder (Princeton, a former Vice Chairman of Fed), education and skills may not help Americans because of an enormous pool of unemployed and less-paid professionals and skilled workers in the rest of the world. China alone needs 17 to 20 million new jobs a year, and -- for political stability - needs them badly. According to Blinder, up to 30% of American jobs can be exported offshore, including 100% of manufacturing jobs.

Let us finish discussion of international trade theory with revealing just one more of fundamental faults of LCA. Last, but very far from the least.

(I) In distinction from the rest of economics, which is based on cost-benefit analysis, LCA takes into consideration only benefits. It promises us no-cost free lunch. In theory, a country is able to specialize without any adjustment costs; in the real world, these costs might easily bury "the advantage."

Even forgetting about the social costs of possible unemployment, LCA therefore excludes -- among many other items -- the costs of re-training, relocation, and readjustment of workers of an industry that is liquidated. Since the theory of international trade deals with multi-industry models, this means that all those workers must be "homogeneous" jacks-of-all-trades, able at a day notice go full speed into any of the growth industries. Textile industry women of South Carolina, when their factories are closed, must start next morning at Boeing or Microsoft. Doing Windows, rather than windows.

Adam Smith strived for efficiency by specialization of individual workers. Ricardo generalized the approach of Smith by specializing whole countries. But he achieved that goal only by sacrificing the specialization of individuals; Ricardo contradicts Smith. (He contradicted Smith in another area, too, but I won't go into that technical issue.)

To avoid an unpleasant discussion, free trade economists needed to find the superwomen and supermen that met the necessary "do-all" skill requirements in manifold industrial fields of international trade. As far as I know, they were not able to find even a single industrial example of such proficiency.

Moreover, they also were not able to find even a single non-industrial example of the "do-all" proficiency among the lower-level workers, those to be displaced, those who need such multi-skill proficiency.

It is arrogant, irresponsible, and cruel to recommend closing down whole industries, with so much at stake and enormous human cost for millions of people, on a basis of a theory that cannot find even a single appropriate and relevant example that would support its main assumption -- the assumption about the homogeneity of the industrial workers. Moreover, cannot even invent such an example!

Perhaps no science in the world was ever-able get away with providing a theory without a single realistic example supporting it -- except economics. (One of the reasons why the "mainstream" economics is a pseudo-science, and LCA - a pseudo-law.)

How do economists find any jacks-of-all-trades at all? They shy away from industrial skills like a devil from holy water. Samuelson has in earlier editions of his Economics textbook a lawyer, who is also an excellent typist; she hires a secretary to do typing. In later editions, this example has been deleted. He knew the score.

In International Economics, Paul Krugman and Maurice Obstfeld give us Babe Ruth, who was both a superb pitcher and a super-superb catcher. Of course, he moved to catching.

By all means, the cigar goes to Gregory Mankiw. In his 1998 Principles of Economics (advance on royalties 1,400,000 bucks, the largest ever for an economics textbook) his multi-skill man is Michael Jordan, who is able both to play basketball and - surprise, surprise -- to mow his lawn. He chooses the ball and hires for mowing a girl-next-door.

But Mankiw properly feels his example to be inadequate and changes it. Not by finding an industrial worker though. Not by changing mowing grass to some meaningful craft. Not by giving an additional skill to the girl-next-door, rather than to Jordan.

Instead, in a later edition, he replaces Jordan by Tiger Woods and the girl by Forrest Gump.

The present role model turns out as slightly tarnished now (because of interaction with other models), so - if Tiger were not rehabilitated -- Mankiw would have to replace Woods. A new edition would be necessary, which would require additional work. Poor guy (only metaphorically -- a new advance on royalties should console him). A silver lining, indeed.

All that would indeed be hilarious, were it not so scary.

To sum up, LCA is invalid, inapplicable, and irrelevant in the real world of trade imbalances; global movement of capital, technology, research, and management skills; worker specialization; persistent large-scale unemployment; huge wage-level gaps between countries; "sticky" prices, wages, and currency rates; technological progress; "learning curves"; production overcapacity; geopolitical and economic instability; and unprecedented uncertainty. This list of forbidden by the theory but unavoidable conditions of the 21st century can be prolonged further. (And I do not even include here the "unfair-trade" conditions that maybe -- just maybe -- can be removed, such as predatory trading and currency manipulation.)

Among the best sayings of George Orwell, "There are some things only intellectuals are crazy enough to believe" takes a place of honor. LCA is one of these things.

Well, let us give the "mainstream" economists the benefit of the doubt. Maybe they do not really believe LCA -- just pretend to believe it. For some noble purposes, like getting a tenure. Intellect restored, but scientific integrity a little under par.

Today's "free trade" is not only the last remnant of laissez-faire -- it is its least deserving remnant, full of wholesale foul play, deception, currency manipulation, predatory techniques, and other violations of its rules, with the perps not even trying to conceal those violations. To call the existing international trade "free," and to use LCA to justify it, is the top of unscrupulous audacity. I do not understand how people of integrity can do it and then call themselves "scientists."

Go To Homepage

Popular in the Community