Free Trade Doesn't Work

by Ian Fletcher

  There are a number of reasons for this. For a start, A taking 100 percent of the world’s retainable industries would mean it taking industries for which it is fundamentally ill-suited, despite its luck in winning them. (As noted, a nation can sometimes be a “bad” winner.) When an industry would have been significantly more efficient in B, A would be better off allocating its factors of production to other industries and importing those goods from B instead. Why? Because A has to balance the monopoly profits from winning a retainable industry against the Ricardian cost of producing for itself goods that B could have produced more cheaply.

  We already know about Ricardian costs: these are the costs a nation incurs when it defies free trade and makes for itself things it should import. Nothing in Gomory and Baumol’s theory makes these costs go away. But their theory introduces monopoly profits that sometimes outweigh them. Would America be better off if we produced for ourselves the fine watches we currently import from Switzerland? Maybe, because this would mean hosting a sophisticated high-paying industry. But maybe not, if we turned out to be a grossly inferior producer and the extra wages and profits were outweighed by the cost of ending up with inferior and more expensive watches. Ricardo understood the Ricardian cost side of this perfectly, but not the monopoly profits side. In his world, there are no monopoly profits because all industries operate under perfect competition all the time—and without scale economies and lockout, it is impossible to have an international monopoly or quasi-monopoly.

  There is another problem with A capturing every retainable industry: this would mean dividing up A’s finite labor force and capital among them. This would spread A’s labor and capital too thin to realize the maximum possible scale economies in each. America has such a large economy that we tend not to take this issue seriously, but for most nations this is a serious issue. Finland, for example, has a world-class position in cellular phones (Nokia). But Finland is probably too small to support both that and world-class positions in avionics, nanotechnology, fiber optics, and genetic engineering. So nation A has to balance its monopoly profits against a scale-economies loss, too.

  The final problem is that capturing every retainable industry would mean depriving B of retainable industries of its own. As a result, B wouldn’t have enough high-value exports to afford very many imports from A, which would have enabled A to increase its exports from its own best industries even more. So A has to balance its monopoly profits against a trading loss as well.

  A good analogy to A’s bundle of tradeoffs would be to an exceptionally lucky jack-of-all-trades who manages to monopolize all the most lucrative jobs in a small town. He is the town’s lawyer, its doctor, its banker, et cetera. Obviously, up to a point, this might be very lucrative. But if he takes it too far, he will make less profit than if he relinquished a few jobs to other people. His best move? Pick some desirable occupation, whether or not someone else would be better at it, but a) he shouldn’t grab any job he’s terrible at, b) he shouldn’t spread himself too thin, and c) he shouldn’t lock everyone else out of good careers entirely. Point a) means there’s no point imposing meritocracy on yourself if this just means somebody else taking your dream job. But equally, you shouldn’t try to be your own doctor if you’re no good at medicine. Point b) means that working several jobs means several paychecks, but you probably can’t make a success out of a profession you practice for only a few hours a day. Point c) means that it’s nice to have the best job in town, but unless other people have good jobs too, you won’t have many customers.

  Clearly, the old Ricardian logic has its place in a Gomory-Baumol world. It just isn’t the whole story anymore. The successful pursuit of economic self-interest, for both nations and people, is a tradeoff between simply grabbing what one wants and submitting to various demands of efficiency. Nations, like people, benefit from importing and exporting in order to allocate their finite productive abilities to their most productive activities. But nations, like people, also benefit by capturing monopoly positions (nations capture good industries, people hold good jobs) whether or not this is efficient per se. (As noted in Chapter Five, nations trade for the same reasons people do.)

  Gomory and Baumol’s analysis also warns us that when someone else holds a lucrative job, they will use it to extract monopoly profits from us. And they may not be the best person for the job, anyway, merely someone who managed to entrench themselves in it. This reasoning extends to nations: not only is it advantageous to win good industries, but it is disadvantageous to end up at the mercy of other nations that have—because this will entail paying them tribute in the form of their monopoly profit margins. (And they may not be the best possible suppliers, anyhow.) Americans are only really conscious of this problem in the case of OPEC, which is a natural oligopoly. But if we end up 30 years from now with a solar-powered economy dependent on Japanese-made solar cells, we will be in the exact same position, especially because American-made solar cells might have been cheaper and better, if we had gotten our act together and developed this industry.

  WIN-WIN VS. WIN-LOSE TRADE

  There is a twin to nation A in the above analysis: its trading partner nation B. When A has 100 percent of the world’s retainable industries, B has 0 percent; when A has 90 percent, B has 10 percent, and so on. What does B’s experience of the above scenarios look like? Let’s plot both nations on the same graph and see (on the next page).751

  The interesting thing about this graph is that in the two side regions (labeled Zone of Mutual Gain), any change in A’s industry capture that benefits A, benefits B also. Whenever A captures more industries and its GDP goes up, B’s GDP goes up, too. And whenever A captures fewer industries and its GDP goes down, B’s GDP goes down, too. So anything either nation does to benefit itself also benefits its trading partner. This means a world in which economic rivalry does not exist because all outcomes are win-win. (This is how Ricardians believe the world works all the time.) However, as one can also see, in the region at the center, things work very differently. There, gain for one nation coincides with loss for the other. In this region, economic rivalry is a fact of life.

  Gomory and Baumol are definitely onto something here because they have managed to bridge the gap between the Pollyannaish “international trade is always win-win” Ricardian view and the overly pessimistic “international trade is war” view. The former view is naïve and, due to dubious assumptions #5 (capital is not internationally mobile) and #7 (trade does not induce adverse productivity growth abroad), untenable even without their insights. The latter view ignores the fact that economics precisely isn’t war because it is a positive-sum game in which goods are produced, not just divided, making mutual gains possible.

  So here, at long last, we have a theoretical framework that can accommodate economic reality as we actually experience it, not just lecture us on what “must” happen as Ricardianism does. It’s both a dog-eat-dog and a scratch-my-back-and-I’ll-scratch-yours world. Economics has finally given common sense permission to be true.

  Gomory and Baumol’s analysis is obviously just the beginning of a whole new way of looking at international trade, which will require decades of elaboration before it becomes a complete new theory of the world economy. But the importance of their work should be obvious already. It will have enormous repercussions for economics (and the real-world policy decisions that depend on it) in the years to come.

  COMPARATIVE ADVANTAGE IN THE RIGHT INDUSTRIES

  What are the policy implications of Gomory and Baumol’s work? Basically that a wise nation will willingly let other nations have their share of the world’s industrial base, but will try to grab the best industries for itself. Then it will sit back (here’s where laissez faire plays its legitimate role) and let the rest of the world compete—head to head, driving the price down through the perfect competition in free markets it seeks to avoid for itself—to produce for it the things it doesn’t want to produce at home.

  Here Ricardo’s ghost rears its head ye
t again: comparative advantage remains a valid principle, but a nation’s best move is not simply to trade according to the comparative advantage it already has. It is to seek comparative advantage in the best industries. Ricardianism is about finding the best use for the comparative advantage one already has (mistaking this for the entire question); Gomory and Baumol are about what kind of comparative advantage it is best to have.

  The top retainable industries will be the best ones to have comparative advantage in, followed by the lesser retainable industries. What makes one retainable industry better than another? It ideally should be large, strongly retainable, and with a long future potential for innovation, and many spinoff industries, ahead of it. As Michael Porter explains, referring to such industries as “structurally attractive” ones:

  Structurally attractive industries, with sustainable entry barriers in such areas as technology, specialized skills, channel access, and brand reputation, often involve high labor productivity and will earn more attractive returns to capital. Standard of living will depend importantly on the capacity of a nation’s firms to successfully penetrate structurally attractive industries. The attractiveness of an industry is not reliably indicated by size, rapid growth, or newness of technology, attributes often stressed by executives and by government planners...By targeting entry into structurally unattractive industries, developing nations have often made poor use of scarce national resources.752

  The industries it is worst to have are the nonretainable, no-scale-economies industries. These behave according to the old Ricardian model: head-to-head free-market competition that drives down prices, profits, and wages. Cynically speaking, these are what a shrewd nation wants its trading partners to specialize in.

  But this doesn’t mean a shrewd nation wants its trading partners to be destitute. Then they would have few tradable industries of any kind and low productivity in those they did have. This would make it impossible to realize significant gains by trading with them. Nobody gets rich trading with Kalahari Bushmen, no matter how shrewd, efficient, or even downright exploitive they are, because Bushmen just don’t have that much stuff in the first place.753 Instead, the ideal trading partner is one that perfectly complements a nation’s own more sophisticated economy. The ideal trading partner is less like a slave (the colonial exploitation model) than like the perfect employee. He skillfully performs all the tasks his employer doesn’t want to perform, freeing that employer to perform more-valuable tasks. But he isn’t so skillful that he threatens his employer’s entrenched position doing the tasks he wishes to reserve to himself. Every lawyer wants an efficient paralegal; no lawyer wants one so skilled that she sets up a competing legal practice!

  The ideal trading partner thus has the highest possible productivity in the industries that a nation doesn’t want to compete in, but low productivity in those it does want to compete in.754 For example, because Japan is a net importer of oil, Japan should want all oil exporting nations to be the most efficient possible oil producers, as this will provide Japan the cheapest possible oil. But Japan should not want Kuwait to become an efficient producer of cars!

  There is a fundamental asymmetry here: a productivity increase in Japan’s car industry will always benefit Kuwait, by enabling Kuwait to buy cheaper cars. But a productivity increase in Kuwait’s car industry (from zero, as it does not currently have one) will not necessarily benefit Japan. (The only time it might is if Kuwait started building car parts that enabled Japan to build better cars, or if Kuwait started building cheap cars, enabling Japan to stop producing these for itself and produce more-expensive ones instead.) Productivity gains by the leader (Japan) are always win-win, but gains by followers (Kuwait) can sometimes be win-lose.755 This is why the most visible pressures for protectionism appear in threatened leading nations, while the most successful protectionism and industrial policy are visible in catching-up nations. Once one views these phenomena under Gomory-Baumol assumptions, they make perfect sense, while Ricardian thinking sees only misguided complaining driven by special-interest politics and unnecessary protectionism of industries that would have succeeded anyway.

  DEFENDING RETAINABLE INDUSTRIES

  We have just scratched at a serious issue. Retainability is real, but it is not absolute and it does have to be defended over time as old industries decay into obsolescence and new ones emerge, rendering yesterday’s entrenched positions irrelevant. As noted earlier, retainability exists because scale economies entrench productivity differences, which entrench industry assignments among nations. But what if productivity differences are not due to entrenched scale economies, but are just there, for whatever reason? Then, as long as they last, the same implications will flow from them that flow from entrenched productivity differences—only these implications will be unstable. So the same Gomory-Baumol analyses will apply. Unfortunately, this also means that the win-lose implications of the Gomory-Baumol analysis will apply, too. So there will be a Zone of Conflict and a Zone of Mutual Gain. There will be harmony and rivalry. There will be winners and losers. Productivity gains by one nation will either help or hurt the other, depending on where on the graph they both are.

  This is extremely important. Because it finally gives us a sound theoretical basis for saying what ordinary Americans (and extremely sophisticated international businesspeople) tend to regard as obvious, even if most American economists do not:

  Foreign productivity growth can take entire industries away from us without conferring any compensating benefits.756

  This is neither an illusion promoted by special interests nor a myth invented by demagogic politicians. Here, Ricardianism is completely out of touch with reality and catastrophically wrong.

  The illusion of permanent advantage that transient productivity differences create explains a fact noted in Chapter Six:757 nations at the peak of their economic power, like Britain in 1860 or the U.S. in 1960, can garner the mistaken impression that free trade is universally beneficial to them. But this illusion depends upon productivity differences between them and their rivals that will not endure forever. This situation also resembles a problem we encountered before, in milder form, with dubious assumption #7 (free trade does not induce adverse productivity growth abroad). By previously analyzing the problem without using multiple-equilibrium ideas, we merely noted that foreign productivity growth can roll back existing gains from trade. This is much worse.

  ARE INDUSTRIES NATIONAL?

  There is one final, and very important, caveat to the whole Gomory-Baumol analysis. When we noted that half the world’s large passenger aircraft are built in Seattle and two-thirds of its fine watches in Switzerland, the reason this is true is that the American aircraft and Swiss watch industries are relatively localized. That is, the companies that make up these industries, though they do have international operations, are still significantly tied to their national home bases. (As noted in Chapter One, this is still largely true even for most multinational corporations.)758 As a result, it is meaningful to talk about the national location of these corporations and thus about national industries.

  This is critical, because scale economies mostly reside in companies, not industries as such. As a result, if companies are truly internationalized, the Gomory-Baumol analysis will get no traction on them, and the more internationalized they are, the less traction it will get. As a result, the less any solution based on this analysis will succeed. Luckily, it is, of course, free trade itself that tends to denationalize companies, so protectionism can push back at denationalization. Successful protectionism based on Gomory and Baumol’s ideas will therefore be a two-pronged policy: first, to maintain in corporations a sufficiently national character for policies imposed at the national level to work, and second, to impose the right national policies.

  Strictly speaking, matters are even more complicated than this. What really matters actually isn’t how national a company’s production is, but how national the scale-economy stages of its production are. The supply chain of a
product from raw materials to the consumer can have widely differing scale economies at different stages. The classic example is consumer electronics, where the manufacture of silicon chips, liquid crystal displays, and other key components enjoys huge scale economies, but the final assembly of these components into a finished device enjoys comparatively few. The former activities depend upon massive R&D, require expensive and sophisticated machinery for physical fabrication, and employ highly trained engineers and scientists. The latter activities often require almost no R&D and are done by hand (perhaps on a 1920s-style assembly line) by semiskilled labor, often in developing countries. One possible strategy this implies is to be fairly agnostic about where companies locate overall, but seek to capture the segments of their production that enjoy scale economies. This is, in fact, roughly the strategy pursued by Singapore, which (unlike Japan) has no internationally recognizable brand-name companies but has managed, through aggressive industrial policy, to systematically capture the high-value segments of foreign corporations’ supply chains in electronics and other industries.

  Chapter 11

  The Natural Strategic Tariff

  As we have seen, free trade depends upon a long list of assumptions that are always dubious and often false. So there is no justification for assuming that it will be the best policy. But while this necessarily opens up the possibility that some kind of protectionism could be better, it doesn’t give an obvious formula for what form that protectionism should take. Which imports, that is, should we tax, how much, and when? We have only scraps of insight here and there, based on the various flaws in free trade we have examined, with no synthesis uniting them into a coherent policy implication. Even Gomory and Baumol’s breakthrough insights, while they do tell us what a better outcome than free trade would look like (more good, i.e. scale economy or retainable industries), do not tell us how to attain that outcome. Merely stapling together the tariff wish lists of every complaining industry in America will do no good, as this embodies no particular rational economic strategy.