Tweet [1]
In the September 2003 Freeman I argued that those who respond to the case for free trade by asking “But what about…?” do not, in fact, call the case for free trade into serious question [2]. My column is below the fold.
My Virginia license plate, adorning both bumpers of my Japanese car, reads FRE TRDE. I always mention this fact to audiences so they know exactly where I stand on the question of how free consumers should be to spend their incomes on foreigners’ goods and services.
I am proudly, completely, confidently, and unconditionally a free trader. I find none of the many “butwhatabouts” to be effective arguments against free trade.
What, you ask, is a butwhatabout? A butwhatabout is the most common response of a free-trade skeptic, issued in the form of a supposedly killer question—as in “But what about the trade deficit?” or “But what about the fact that many foreign workers are paid a mere fraction of what Americans earn?”
Butwhatabouts are bountiful and varied. But the most frequent butwhatabout is this: “But what about the worker who loses his job to foreign competition? How do you justify free trade to him?” The suggestion is that free trade is inappropriate for real-world policy if its advocates cannot satisfactorily justify their case to workers rendered unemployed by foreign trade.
Because I encounter this particular butwhatabout so often, I know that it’s one whose answer matters and hence one that should be rebutted with care.
I begin my answer by expressing sincere recognition that losing a job is an unfortunate and often devastating experience. I truly pity workers who, through no fault of their own, find themselves unemployed. Then I ask, “But what about the worker who loses his job to protectionism? How do you justify high tariffs and import restrictions to him?” You see, it’s a common mistake of free-trade skeptics to overlook the fact that protectionism inevitably causes some workers to lose jobs.
When government imposes a tariff or import restraint on, say, Americans who seek to purchase foreign steel, one object of this policy is to protect jobs in the U.S. steel industry. And, indeed, by making it more costly for American consumers to buy foreign steel, government artificially props up demand for American steel and thus keeps some workers employed in steel plants who would otherwise lose their jobs.
But by artificially reducing American purchases of foreign steel, protectionism also reduces the number of dollars that foreign steel producers receive. With fewer dollars to spend, foreigners must reduce the amount of goods and services that they purchase from American producers—say, from U.S. lumber suppliers and software firms. Also, because Americans are forced by protectionism to pay more for steel, they must spend less on other goods and services, many of which are produced domestically.
Employment in these domestic industries falls. Some American workers who would have kept their jobs in the absence of the steel tariffs are, through no fault of their own, thrown into unemployment. So it is no argument against free trade to point out that it eliminates the jobs of some domestic workers. Preventing free trade—protectionism—has the same effect.
While no essential differences distinguish workers who lose jobs as a consequence of free trade from workers who lose jobs as a consequence of protectionism (hence making invalid the argument that free trade is cruel or otherwise suspect because it causes some particular job losses), there are at least two differences that distinguish the jobs eliminated by free trade from those eliminated by protectionism. These differences are worker productivity and wages.
Workers in protected industries are less productive than workers in industries that survive without protection. The reasons are two. First, shielding a firm from competition weakens its incentives to increase efficiency. Second and more important, firms that enjoy a comparative advantage over foreign competitors do not need protection from imports. So the industries and firms that seek protection typically are those that do not enjoy a comparative advantage in producing the things they will produce behind the tariff wall. These firms are among the country’s least-efficient producers.
Being among the country’s least-efficient producers means that the per-worker output of protected firms is generally lower than the per-worker output of firms that are efficient enough to survive market competition without protection. Therefore, competitive wages in protected industries generally are lower than competitive wages earned by equally skilled domestic workers whose jobs depend on free trade. In short, jobs made possible by free trade generally are more highly paid than jobs maintained by protectionism.
Protectionism protects jobs that are comparatively inefficient; it eliminates jobs that are comparatively efficient. This is hardly a means of fostering a strong domestic economy.
Consumers Eliminate Jobs
But an even more fundamental response to the butwhatabout-the-unemployed-worker question is pertinent—namely, that free trade is not unique in causing some jobs to disappear while causing others to arise. Any substantial change in consumer spending or savings will eliminate some jobs and create others.
For example, if changing tastes prompt consumers to spend more time working out in gyms and less time bowling, some bowling-alley employees will lose their jobs. Some might remain painfully unemployed for a long time. Butwhatabout them? Do those of us who support consumer freedom have the burden of persuading unemployed bowling-alley workers that such freedom is desirable? Does the fact that nearly all changes in consumer spending will destroy some particular jobs mean that government should consider freezing consumer spending in place now and forever?
Of course not.
Whenever government intervention is proposed, it is imperative to avoid being misled by superficial facts. No fact about voluntary exchange is more superficial than the particular piece of geography in which a seller happens to be situated—no ifs, ands, or butwhatabouts.