American Economic - Masthead Fighting For American Companies - Fighting for American Jobs United States Business and Industry Council
Current Trade Deficit: - Presented by The Robert A. Stranahan Lectures
Fatal Flaws in the Theory of Comparative Advantage
Thursday, November 06, 2008
Commentary by USBICEF Contributor Ian Fletcher

Free traders tell us that David Ricardo’s famous theory of comparative advantage proves that free trade always benefits both trading partners and only ever harms special interests.  But as its inventor knew perfectly well, his theory is not a blank check for free trade, but a conditional theory that depends upon certain assumptions that may or may not hold.

Because these assumptions often don’t hold in the real world, the theory’s direct applicability to contemporary American policymaking is sharply limited, as 30-plus years of trade deficits – not to mention significant industrial hollowing out – should indicate to even the theory’s most ardent advocates.

Assumption #1:  There are no externalities.  

An externality emerges when the price of a good does not reflect its full economic cost or value.  The classic negative externality is environmental damage, which does economic harm but doesn’t raise the price of the product that caused it.  Goods from a country with lax pollution standards will be relatively cheap, encouraging other countries to import too much.  As a result, practicing free trade in these circumstances will benefit the pollution-haven country economically, but harm both it and the importing countries environmentally.  And since pollution respects no national borders, no country on the planet will win from an environmental standpoint.

The classic positive externality is technological spillover, which occurs when the technology an industry develops benefits parts of the economy that are not its customers and thus do not contribute to its profits.   A domestic industry that generates technological spillovers may be wiped out by subsidized foreign competition because the price system did not accurately capture all the value it was producing.  This result, moreover, will not only hurt the importing industry and country, but it also can hurt the exporting country (and the entire world) by retarding the pace of technological progress.

Assumption #2:  Nations trade only goods and services, not debt and assets.  

At its heart, comparative advantage flows from the fact that nations face different costs for producing goods.  Therefore, its mechanism will malfunction if prices cease to be determined by these costs.  But this is precisely what will happen if nations cease trading goods produced today and start trading goods produced yesterday (assets) or promised for tomorrow (debts).  The production cost of such goods today, by definition, is zero, so the economy will mistakenly think that it has a limitless supply of “free exports” and will import too much.  And the United States’ current practice of financing its trade deficits by selling off assets and assuming ever more debt shows that this mistake is indeed being made – on a gigantic scale.

Countries and populations facing debts that are a rising percentage of their assets will never be economic winners over significant lengths of time.  Unless they change their policies and practices, they will eventually run out of assets to sell.  Such countries will, moreover, incur yet another economic cost:   Demand for their currencies to pay for their assets will push the value of those currencies above the level that their trade deficits would otherwise maintain.  As a result, the debtor countries’ exports will be curbed and their imports inflated, undermining their national finances and wealth-creation opportunities for their domestic producers.

And long before these free trade arrangements turn debtor countries and populations into economic losers, they will turn them into political losers – as growing foreign control of their asset base inevitably translates into growing foreign influence over all aspects of national life.

Assumption #3: Factors of production are domestically mobile.  

The theory of comparative advantage assumes that free trade will reallocate factors of production from sectors with comparative disadvantage to sectors with comparative advantage.   But if factors can’t reallocate, this mechanism breaks down.  If labor can’t move between industries – say, because of skill mismatches – then comparative advantage won’t shunt workers out of sunset industries into sunrise ones, but into unemployment.

Indeed, this problem brings up a further weakness in this assumption of easily mobile domestic factors of production – the related assumption that full employment always prevails.  In other words, comparative advantage theory assumes not only that workers and their characteristics are perfectly fungible, but also that the sunrise industries will always be willing and able to absorb them.
Yet if differing productivity levels are one big factor distinguishing sunset from sunrise industries, it’s easy to identify one reason this assumption fails the reality test: All else being equal, the more productive industries will be creating fewer jobs than the less productive.  Recalling the infrequency with which national economies have created and sustained full employment reveals another gap between comparative advantage assumptions and the real world.

Assumption #4:  Factors of production are not internationally mobile.

The theory of comparative advantage says that free trade optimizes a nation’s economy by driving factors of production to their highest-value uses within it.  But this maxim relies upon the assumption that it is impossible to drive these factors right out of the nation in question.  If this happens, then the factors will be driven to their highest-value use in the entire world instead.  This transnational movement will optimize the world economy, but is not necessarily advantageous for a particular nation.  In fact, comparative advantage will no longer even apply, but will be replaced by absolute advantage, which governs competition between nations.

In 1950, it seemed obvious that Michigan had comparative advantage in automobiles and Alabama in cotton.  But by 2000, automobile plants were closing in Michigan and opening in Alabama.  This may have benefited Alabama, but it did not necessarily benefit Michigan.  (It might have if Michigan had been transitioning to industries with higher value-added than automobiles, but this didn’t happen.)

Moreover, here again the fallacious assumption of full employment is at work.  To continue the example, the U.S. automobile industry (along with other high-value industries) has been unable to create all the jobs needed to absorb all the displaced Michiganders and their counterparts nationwide.  At least as important, given the thorough globalization of trade flows, these high-value industries have been unable to create all the jobs needed to absorb all the American workers displaced by Chinese, Indian, and other foreign labor.
Assumption #5: Long-term growth is caused by short-term efficiency.  

The theory of comparative advantage is a case of static analysis: It looks at the facts of a single instant in time and determines the most advantageous response at that instant.  But it says nothing about how these facts may change tomorrow, or more importantly, how one might cause them to change in one’s favor.  So even if comparative advantage correctly tells us the most efficient use for our stock of factors of production today, given their productivities in various industries, it does not tell us the best way to raise those productivities tomorrow.  

Productivity, however, is the essence of economic growth, and in the long run a lot more important than squeezing every drop of advantage from the productivities we have today.  Sacrificing a bit of short-term efficiency to get new industries started is the classic “infant industries” exception to free trade; but its scope is now understood to be considerably broader, as parts of every dynamic industry are always “infant.”

The economics of long-term growth are very complex, and rife with path dependencies and positive externalities that are not even touched on by the theory of comparative advantage.  This is a major reason why nations that have openly rejected the theory as a guide to policy, like Japan and China, have been able to succeed so brilliantly without it.

Assumption # 6:  There are no economies of scale.  

In the presence of scale economies, nations that reach large-scale production first can entrench themselves in industries simply because they were first.   As revealed by Ralph Gomory and William Baumol in their path-breaking 2001 book Global Trade and Conflicting National Interests, this fact has vast significance.  It means that the global distribution of industries depends upon accidents of industrial history – or on farsighted government policies.  And if these policies and their effects are not somehow offset, the continuation of free-trade policies by certain countries will not even produce the most efficient possible worldwide outcome.  

Moreover, economists and business leaders know that certain industries constitute global oligopolies that can reap exceptional profits and pay exceptional wages even in the face of cheap foreign labor.  Free trade alone will not necessarily assign these industries to the United States.  Again, however, it can aid foreign countries with active industrial policies aimed at prying them from American hands.

Assumption # 7:  There is no cross-border investment

The theory of comparative advantage requires that capital not be significantly mobile between nations.  David Ricardo himself knew this perfectly well.  As he put it, using his standard example of the trade in English cloth for Portuguese wine:

The difference in this respect, between a single country and many, is easily accounted for, by considering the difficulty with which capital moves from one country to another, to seek a more profitable employment, and the activity with which it invariably passes from one province to another of the same country.

It would undoubtedly be advantageous to the capitalists of England, and to the consumers in both countries, that under such circumstances the wine and the cloth should both be made in Portugal, and therefore that the capital and labor of England employed in making cloth should be removed to Portugal for that purpose. (The Principles of Political Economy and Taxation, p. 83)

But it would not, he continues, necessarily be advantageous to the workers of England!  This is, of course, exactly the problem Americans experience today when cheaper foreign goods replace goods produced here: Capitalists like the higher profits and consumers like the lower prices, but workers don’t like the lost jobs.  Most consumers, though, are workers, and there is no guarantee that under free trade they gain more in the former capacity than they lose in the latter.

Having observed that capital mobility would undo his theory, Ricardo then argues why capital will not, in fact, be mobile:

Experience, however, shows that the fancied or real insecurity of capital, when not under the immediate control of its owner, together with the natural disinclination which every man has to quit the country of his birth and connections, and entrust himself, with all his habits fixed, to a strange government and new laws, check the emigration of capital. These feelings, which I should be sorry to see weakened, [emphasis added] induce most men of property to be satisfied with a low rate of profits in their own country, rather than seek a more advantageous employment for their wealth in foreign nations.

Yet this assumption has been wholly invalidated by the recent tendency for global capital flows to exceed global trade flows by orders of magnitude.  Moreover, although much of this capital flows among countries with similar languages and customs (e.g., within the European Union, between the United States and Canada and the United States and Britain, among East Asian countries), much of it also flows among countries with little in common culturally or socially (e.g., between the high-income countries of Europe and North America to the low-income countries of East Asia and Latin America).

The importance of capital flows to trade flows, and to the modern utility of  Ricardo’s assumptions, is magnified by the fact that so much transnational capital investment is devoted to the creation of trade-oriented – and, more specifically, export-oriented – activity.


The dissection of the foregoing list of assumptions should make clear that while the theory of comparative advantage is a valid and useful tool of economic analysis, it is not the only point that economics has to make about who wins and who loses in international trade.  It is simply not valid, even according to the theory itself, to use it as a rubber stamp to “prove” that 100% free trade 100% of the time with 100% of the world’s nations is good for America.   That would only be true if all of the above assumptions were satisfied in reality, and they don’t even come close.

In fact, the current form of globalization means that they get further away from being satisfied every day.  In the 1950s, when these assumptions were much closer to reality, free trade may have been a winning move for America, but those days are long gone.  It is free traders, not protectionists, who are living in the past and sticking their heads in the sand.

Ian Fletcher is an economist in private practice in San Francisco, consulting mainly to hedge funds and the occasional politician and news organization.

Sign up below to have AEA news headlines, articles, and reports delivered directly to your mailbox. 100% free and private.
PDF Download View our plan to save American Manufacturing
PDF Download Multimedia educational shorts on key trade issues