The Free Trade Fallacy
[Reprinted from The New Republic, vol. 188,
no. 12 (28 March 1983)]
In the firmament of American ideological convictions, no star burns
brighter than the bipartisan devotion to free trade. The President's
1983 Economic Report, to no one's surprise, sternly admonished
would-be protectionists. An editorial in The New York Times,
midway through an otherwise sensibly Keynesian argument, paused to add
ritually, "Protectionism might mean a few jobs for American auto
workers, but it would depress the living standards of hundreds of
millions of consumers and workers, here and abroad."
The Rising Tide of Protectionism has become an irresistible topic for
a light news day. Before me is a thick sheaf of nearly interchangeable
clips warning of impending trade war. With rare unanimity, the press
has excoriated the United Auto Workers for its local content
legislation. The Wall Street Journal's editorial ("Loco
Content") and the Times's ("The Made-in-America Trap")
were, if anything, a shade more charitable than Cockburn and Ridgeway
in The Village Voice ("Jobs and Racism"). And when
former Vice President Mondale began telling labor audiences that
America should hold Japan to a single standard in trade, it signaled a
chorus of shame-on-Fritz stories.
The standard trade war story goes like this: recession has prompted a
spate of jingoistic and self-defeating demands to fence out superior
foreign goods. These demands typically emanate from overpaid workers,
loser industries, and their political toadies. Protectionism will
breed stagnation, retaliation, and worldwide depression. Remember
Perhaps it is just the unnerving experience of seeing The Wall
Street Journal and The Village Voice on the same side, but
one is moved to further inquiry. Recall for a moment the classic
theory of comparative advantage. As the English economist David
Ricardo explained it in 1817, if you are more efficient at making wine
and I am better at weaving cloth, then it would be silly for each of
us to produce both goods. Far better to do what each does best, and to
trade the excess. Obviously then, barriers to trade defeat potential
efficiency gains. Add some algebra, and that is how trade theory
continues to be taught today.
To bring Ricardo's homely illustration up to date, the economically
sound way to deal with the Japanese menace is simply to buy their
entire cornucopia - the cheaper the better. If they are superior at
making autos, TVs, tape recorders, cameras, steel, machine tools,
baseballs, semiconductors, computers, and other peculiarly Oriental
products, it is irrational to shelter our own benighted industries.
Far more sensible to buy their goods, let the bracing tonic of
competition shake America from its torpor, and wait for the market to
reveal our niche in the international division of labor.
But this formulation fails to describe the global economy as it
actually works. The classical theory of free trade was based on what
economists call "factor endowments" - a nation's natural
advantages in climate, minerals, arable land, or plentiful labor. The
theory doesn't fit a world of learning curves, economies of scale, and
floating exchange rates. And it certainly doesn't deal with the fact
that much "comparative advantage" today is created not by
markets but by government action. If Boeing got a head start on the
707 from multibillion-dollar military contracts, is that a sin against
free trade? Well, sort of. If the European Airbus responds with
subsidized loans, is that worse? If only Western Electric (a U.S.
supplier) can produce for Bell, is that protection? If Japan uses
public capital, research subsidies, and market-sharing cartels to
launch a highly competitive semiconductor industry, is that
protection? Maybe so, maybe not.
Just fifty years ago, Keynes, having dissented from the
nineteenth-century theory of free markets, began wondering about free
trade as well. In a 1933 essay in the Yale Review called "National
Self-Sufficiency," he noted that "most modern processes of
mass production can be performed in most countries and climates with
almost equal efficiency." He wondered whether the putative
efficiencies of trade necessarily justified the loss of national
autonomy. Today nearly half of world trade is conducted between units
of multinational corporations. As Keynes predicted, most basic
products (such as steel, plastics, microprocessors, textiles, and
machine tools) can be manufactured almost anywhere, but by labor
forces with vastly differing prevailing wages.
With dozens of countries trying to emulate Japan, the trend is toward
worldwide excess capacity, shortened useful life of capital equipment,
and downward pressure on wages. For in a world where technology is
highly mobile and interchangeable, there is a real risk that
comparative advantage comes to be defined as whose work force will
work for the lowest wage.
In such a world, it is possible for industries to grow nominally more
productive while the national economy grows poorer. How can that be?
The factor left out of the simple Ricardo equation is idle capacity.
If America's autos (or steel tubes, or machine tools) are manufactured
more productively than a decade ago but less productively than in
Japan (or Korea, or Brazil), and if we practice what we preach about
open trade, then an immense share of U.S. purchasing power will go to
provide jobs overseas. A growing segment of our productive resources
will lie idle. American manufacturers, detecting soft markets and
falling profits, will decline to invest. Steelmakers will buy oil
companies. Consumer access to superior foreign products will not
necessarily compensate for the decline in real income and the idle
resources. Nor is there any guarantee that the new industrial
countries will use their burgeoning income from American sales to buy
American capital equipment (or computers, or even coal), for they are
all striving to develop their own advanced, diversified economies.
Against this background of tidal change in the global economy, the
conventional reverence for "free trade" is just not helpful.
As an economic paradigm, it denies us a realistic appraisal of second
bests. As a political principle, it leads liberals into a disastrous
logic in which the main obstacle to a strong American economy is
decent living: standards for the American work force. Worst of all, a
simple-minded devotion to textbook free trade in a world of
mercantilism assures that the form of protection we inevitably get
will be purely defensive, and will not lead to constructive change in
the protected industry.
The seductive fallacy that pervades the hand-wringing about
protectionism is the premise that free trade is the norm and that
successful foreign exporters must be playing by the rules. Even so
canny a critic of political economy as Michael Kinsley wrote in these
pages that "Very few American workers have lost their jobs
because of unfair foreign trade practices, and it is demagogic for
Mondale and company to suggest otherwise." But what is an unfair
trade practice? The Common Market just filed a complaint alleging that
the entire Japanese industrial system is one great unfair trade
To the extent that the rules of liberal trade are codified, they
repose in the General Agreement on Tariffs and Trade (stay awake, this
will be brief). The GATT is one of those multilateral institutions
created in the American image just after World War II, a splendid
historical moment when we could commend free trade to our allies the
way the biggest kid on the block calls for a fair fight.
The basic GATT treaty, ratified in 1947, requires that all member
nations get the same tariff treatment (the "most favored nation"
doctrine), and that tariffs, in theory at least, are the only
permissible form of barrier. Governments are supposed to treat foreign
goods exactly the same as domestic ones: no subsidies, tax
preferences, cheap loans to home industries, no quotas, preferential
procurement, or inspection gimmicks to exclude foreign ones. Nor can
producers sell below cost (dumping) in foreign markets....
In classical free trade theory, the only permissible candidate for
temporary protection is the "infant industry." But Japan and
its imitators, not unreasonably, treat every emerging technology as an
infant industry. Japan uses a highly sheltered domestic market as a
laboratory, and as a shield behind which to launch one export winner
after another. Seemingly, Japan should be paying a heavy price for its
protectionism as its industry stagnates. Poor Japan! This is not the
place for a detailed recapitulation of Japan, Inc., but keep in mind
The Japanese government, in close collaboration with industry,
targets sectors for development. It doesn't try to pick winners
blindfolded; it creates them. It offers special equity loans, which
need be repaid only if the venture turns a profit. It lends public
capital through the Japan Development Bank, which signals private
bankers to let funds flow. Where our government offers tax deductions
to all businesses as an entitlement, Japan taxes ordinary business
profits at stiff rates and saves its tax subsidies for targeted
ventures. The government sometimes buys back outdated capital
equipment to create markets for newer capital.
The famed Ministry of International Trade and Industry has pursued
this essential strategy for better than twenty years, keeping foreign
borrowers out of cheap Japanese capital markets, letting in foreign
investors only on very restricted terms, moving Japan up the product
ladder from cheap labor intensive goods in the 1950s to autos and
steel in the 1960s, consumer electronics in the early 1970s, and
computers, semiconductors, optical fibers, and just about everything
else by 1980. The Japanese government also waives antimonopoly laws
for development cartels, and organizes recession cartels when
overcapacity is a problem. And far from defying the discipline of the
market, Mm encourages fierce domestic competition before winnowing the
field down to a few export champions....
The Japanese not only sin against the rules of market economics. They
convert sin into productive virtue. By our own highest standards, they
must be doing something right. The evident success of the Japanese
model and the worldwide rush to emulate it create both a diplomatic
crisis for American trade negotiators and a deeper ideological crisis
for the free trade regime. As Berkeley professors John Zysman and
Steven Cohen observed in a careful study for the Congressional Joint
Economic Committee last December, America, as the main defender of the
GATT philosophy, now faces an acute policy dilemma: "how to
sustain the open trade system and promote the competitive position of
American industry" at the same time.
Unfortunately, the dilemma is compounded by our ideological blinders.
Americans believe so fervently in free markets, especially in trade,
that we shun interventionist measures until an industry is in deep
trouble. Then we build it half a bridge.
There is no better example of the lethal combination of protectionism
plus market-capitalism-as-usual than the steel industry. Steel has
enjoyed some import limitation since the late 1950s, initially through
informal quotas. The industry is oligopolistic; it was very slow to
modernize. By the mid-1970s, world demand for steel was leveling off
just as aggressive new producers such as Japan, Korea, and Brazil were
flooding world markets with cheap, state-of-the-art steel.
As the Carter Administration took office, the American steel industry
was pursuing antidumping suits against foreign producers - an avenue
that creates problems for American diplomacy. The new Administration
had a better idea, more consistent with open markets and neighborly
economic relations. It devised a "trigger price mechanism,"
a kind of floor price for foreign steel entering American markets.
This was supposed to limit import penetration. The steelmakers
withdrew their suits. Imports continued to increase.
So the Carter Administration moved with characteristic caution toward
a minimalist industrial policy. Officials invented a kind of near-beer
called the Steel Tripartite. Together, industry, labor, and government
would devise a strategy for a competitive American steel industry. The
eventual steel policy accepted the industry's own agenda: more
protection, a softening of pollution control requirements, wage
restraint, new tax incentives, and a gentlemen's agreement to phase
out excess capacity. What the policy did not include was either an
enforceable commitment or adequate capital to modernize the industry.
By market standards, massive retooling was not a rational course,
because the return on steel investment was well below prevailing
yields on other investments. Moreover, government officials had
neither the ideological mandate nor adequate information to tell the
steel industry how to invest. "We would sit around and talk about
rods versus plate versus specialty steel, and none of us in government
had any knowledge of how the steel industry actually operates,"
confesses C. Fred Bergsten, who served as Treasury's top trade
official under Carter. "There has never been a government study
of what size and shape steel industry the country needs. If we're
going to go down this road, we should do it right, rather than simply
preserving the status quo."...
The argument that we should let "the market" ease us out of
old-fashioned heavy industry in which newly industrialized countries
have a comparative advantage quickly melts away once you realize that
precisely the same nonmarket pressures are squeezing us out of the
highest-tech industries as well. And the argument that blames the
problem on overpaid American labor collapses when one understands that
semiskilled labor overseas in several Asian nations is producing
advanced products for the U.S. market at less than a dollar an hour.
Who really thinks that we should lower American wages to that level in
order to compete?
In theory, other nations' willingness to exploit their work forces in
order to provide Americans with good, cheap products offers a deal we
shouldn't refuse. But the fallacy in that logic is to measure the
costs and benefits of a trade transaction only in terms of that
transaction itself. Classical free-trade theory assumes full
employment. When foreign, state-led competition drives us out of
industry after industry, the costs to the economy as a whole can
easily outweigh the benefits. As Wolfgang Hager, a consultant to the
Common Market, has written, "The cheap [imported] shirt is paid
for several times: once at the counter, then again in unemployment
benefits. Secondary losses involve input industries... machinery,
fibers, chemicals for dyeing and finishing products."
As it happens, Hager's metaphor, the textile industry, is a fairly
successful example of managed trade, which combines a dose of
protection with a dose of modernization. Essentially, textiles have
been removed from the free-trade regime by an international
market-sharing agreement. In the late 1950s, the American textile
industry began suffering insurmountable competition from cheap
imports. The United States first imposed quotas on imports of cotton
fibers, then on synthetics, and eventually on most textiles and
apparel as well. A so-called Multi-Fiber Arrangement eventually was
negotiated with other nations, which shelters the textile industries
of Europe and the United States from wholesale import penetration.
Under M.F.A., import growth in textiles was limited to an average of 6
percent per year.
The consequences of this, in theory, should have been stagnation. But
the result has been exactly the opposite. The degree of protection,
and a climate of cooperation with the two major labor unions,
encouraged the American textile industry to invest heavily in
modernization. During the 1960s and 1970s, the average annual
productivity growth in textiles has been about twice the U.S.
industrial average, second only to electronics. According to a study
done for the Common Market, productivity in the most efficient
American weaving operations is 130,000 stitches per worker per hour -
twice as high as France and three times as high as Britain. Textiles,
surprisingly enough, have remained an export winner for the United
States, with net exports regularly exceeding imports. (In 1982, a
depressed year that saw renewed competition from China, Hong Kong,
Korea, and Taiwan, exports just about equaled imports.)
But surely the American consumer pays the bill when the domestic
market is sheltered from open foreign competition. Wrong again.
Textile prices have risen at only about half the average rate of the
producer price index, both before and after the introduction of the
Now, it is possible to perform some algebraic manipulations and show
how much lower textile prices would have been without any protection.
One such computation places the cost of each protected textile job at
several hundred thousand dollars. But these static calculations are
essentially useless as practical policy guides, for they leave out the
value over time of maintaining a textile industry in the United
States. The benefits include not only jobs, but contributions to
G.N.P., to the balance of payments, and the fact that investing in
this generation's technology is the ticket of admission to the next.
Why didn't the textile industry stagnate? Why didn't protectionism
lead to higher prices? Largely because the textile industry is quite
competitive domestically. The top five manufacturers have less than 20
percent of the market. The industry still operates under a 1968
Federal Trade Commission consent order prohibiting any company with
sales of more than $100 million from acquiring one with sales
exceeding $10 million. If an industry competes vigorously
domestically, it can innovate and keep prices low, despite being
sheltered from ultra- low-wage foreign competition - or rather, thanks
to the shelter. In fact, students of the nature of modern managed
capitalism should hardly be surprised that market stability and new
investment go hand in hand.
The textile case also suggests that the sunrise industry/sunset
industry distinction is so much nonsense. Most of America's major
industries can be winners or losers, depending on whether they get
sufficient capital investment. And it turns out that many U.S.
industries such as textiles and shoes, which conventionally seem
destined for lower-wage countries, can survive and modernize given a
reasonable degree of, well, protection.
What, then, is to be done? First, we should acknowledge the realities
of international trade. Our competitors, increasingly, are not free
marketeers in our own mold. It is absurd to let foreign mercantilist
enterprise overrun U.S. industry in the name of free trade. The
alternative is not jingoist protectionism. It is managed trade, on the
model of the Multi-Fiber Arrangement. If domestic industries are
assured some limits to import growth, then it becomes rational for
them to keep retooling and modernizing.
It is not necessary to protect every industry, nor do we want an
American MITT. But surely it is reasonable to fashion plans for
particular key sectors like steel, autos, machine tools, and
semiconductors. The idea is not to close U.S. markets, but to limit
the rate of import growth in key industries. In exchange, the domestic
industry must invest heavily in modernization. And as part of the
bargain, workers deserve a degree of job security and job retraining
Far from being just another euphemism for beggar-thy-neighbor, a more
stable trade system generally can be in the interest of producing
countries. Universal excess capacity does no country much of a favor.
When rapid penetration of the U.S. color TV market by Korean suppliers
became intolerable, we slammed shut an open door. Overnight, Korean
color TV production shrank to 20 percent of capacity. Predictable, if
more gradual, growth in sales would have been preferable for us and
for the Koreans.
Second, we should understand the interrelationship of managed trade,
industrial policies, and economic recovery. Without a degree of
industrial planning, limiting imports leads indeed to stagnation.
Without restored world economic growth, managed trade becomes a nasty
battle over shares of a shrinking pie, instead of allocation of a
growing one. And without some limitation on imports, the Keynesian
pump leaks. One reason big deficits fail to ignite recoveries is that
so much of the growth in demand goes to purchase imported goods.
Third, we should train more economists to study industries in the
particular. Most economists dwell in the best of all possible worlds,
where markets equilibrate, firms optimize, the idle resources
re-employ themselves. "Microeconomics" is seldom the study
of actual industries; it is most often a branch of arcane mathematics.
The issue of whether governments can sometimes improve on markets is
not a fit subject for empirical inquiry, for the paradigm begins with
the assumption that they cannot. The highly practical question of when
a little protection is justified is ruled out ex ante, since
neoclassical economics assumes that less protection is always better
Because applied industrial economics is not a mainstream concern of
the economics profession, the people who study it tend to come from
the fields of management, industrial and labor relations, planning,
and law. They are not invited to professional gatherings of
economists, who thus continue to avoid the most pressing practical
questions. One economist whom I otherwise admire told me he found it "seedy"
that high-wage autoworkers would ask consumers to subsidize their pay.
Surely it is seedier for an $800-a-week tenured economist to lecture a
$400-a-week autoworker on job security; if the Japanese have a genuine
comparative advantage in anything, it is in applied economics.
Fourth, we should stop viewing high wages as a liability. After World
War II, Western Europe and North America evolved a social contract
unique in the history of industrial capitalism. Unionism was
encouraged, workers got a fair share in the fruits of production, and
a measure of job security. The transformation of a crude industrial
production machine into something approximating social citizenship is
an immense achievement, not to be sacrificed lightly on the altar of "free
trade." It took one depression to show that wage cuts are no
route to recovery. Will it take another to show they are a poor
formula for competitiveness? Well-paid workers, after all, are