Mises Daily Articles
Free Trade and Factor Mobility
In a recent New York Times op-ed piece, "Second Thoughts on Free Trade," Senator Charles Schumer and economist Paul Craig Roberts argue that the typical arguments for free trade, while perhaps valid in the days of Ricardo, are no longer relevant in today's economy of multinational corporations and high-speed telecommunications. According to our authors:
Two recent examples illustrate this concern. Over the next three years, a major New York securities firm plans to replace its team of 800 American software engineers, who each earns about $150,000 per year, with an equally competent team in India earning an average of only $20,000. Second, within five years the number of radiologists in this country is expected to decline significantly because M.R.I. data can be sent over the Internet to Asian radiologists capable of diagnosing the problem at a small fraction of the cost.
Our authors acknowledge that most economists would interpret the above developments "through the classic prism of 'free trade,'" and that most economists would "label any challenge as protectionism." Nonetheless, our authors claim that "these new developments call into question some of the key assumptions supporting the doctrine of free trade."
So what is wrong with David Ricardo's law of comparative advantage, which showed that per capita productivity is highest when each country specializes in those industries in which its workers are relatively superior? According to Schumer and PCR:
[W]hen Ricardo said that free trade would produce shared gains for all nations, he assumed that the resources used to produce goods—what he called the "factors of production"—would not be easily moved over international borders. Comparative advantage is undermined if the factors of production can relocate to wherever they are most productive: in today's case, to a relatively few countries with abundant cheap labor. In this situation, there are no longer shared gains—some countries win and others lose.
Schumer and Roberts are correct in their discussion of Ricardo's demonstration, which did indeed assume that factors of production were fixed and that only final goods could be shipped across international borders. However, Schumer and Roberts are wrong when they conclude that free trade is good only under these assumptions. (For an excellent discussion see Mises's treatment in Human Action.)
The case for free trade is more general than Ricardo's narrow demonstration. The per capita consumption of a group of individuals will be lower if an external penalty (e.g. a tariff) is placed on their ability to trade with people outside of the group. This is true whether or not the tools with which these people work, or the people themselves, are allowed to leave an arbitrary geographical region.
Tariffs are Taxes
Before getting into the heart of the matter, let us remind ourselves of a basic but important fact: Tariffs are taxes. No matter how sophisticated the argument, when someone opposes free trade, what that really means is that he favors the placement of taxes (or similar restrictions) on consumers. In the present context, Schumer and Roberts are implicitly proposing to make Americans richer by raising taxes on some of the goods available for purchase.1 They implicitly argue that trade barriers reduce consumption when capital is immobile (as Ricardo demonstrated), but all that changed with the invention of the Internet. Now, according to our authors, barriers to trade will raise living standards. Such an argument should puzzle any free market economist.
"Two Recent Examples"
Another preliminary observation is that the two "recent examples" (quoted in the first block paragraph above) really have nothing to do with factor mobility, and are no different from standard objections to classical free trade. Since the economist Paul Craig Roberts says that he agrees with free trade when Ricardo's assumptions hold, he should therefore understand that his two examples demonstrate the beneficial operation of the market.
Recall that in the first case, our authors lament a New York firm that lays off 800 American software engineers and hires an "equally competent team in India." The firm does this because the wages of the American engineers are $150,000 while the Indians must be paid only $20,000. Apparently this is an example of the insidious influence of multinational corporations.
But how does this example differ from the standard (naïve) objection to free trade? Suppose that an Indian securities firm hired Indian software engineers for $20,000 each, and drove the American firm out of business (when its American customers gave their business to the Indian firm), thus putting its 800 software engineers out of work? We can imagine the American firm running to Congress asking for a tariff. Its lobbyists would argue that because American engineers must be paid $150,000, they can't possibly compete with the Indian securities firm.
There is really nothing different between this case and the one cited by our authors, except for the nationality of the individuals who own the company in question. Do our authors really want to argue that it is good (for Americans in general) when an Indian firm hires Indian engineers and puts American engineers out of work, but that it is bad when an American-owned multinational does the same thing?
We can apply similar reasoning to our authors' second example, in which the speed of data transmission threatens the future of American radiologists. Remember, Schumer and Roberts argue that this case does not fall under the assumptions used by classical economists in support of free trade, because here the dreaded "Asian radiologists capable of diagnosing the [medical] problem at a small fraction of the cost" aren't simply using resources in Asia and then shipping over the final good to compete with the products of American radiologists. No, in this Information Age context, what is happening is that one of the "factors of production" involved in the service—i.e. the MRI data of the patients—can be transmitted to the Asian radiologists, who perform their analysis in Asia and then transmit their opinion back to the US.
But does this difference really flip the conclusion reached by Ricardo, Bastiat, and other free traders? After all, with only a slight modification we can change the example to one clearly falling under the classical assumptions of immobile factors. Let us suppose that an Asian firm invents a disposable black box, programmed by Asian radiologists, which it sells to doctors all over the world. Into this imaginary device, a doctor feeds a patient's MRI data and then, after a few moments of whirring and buzzing, out pops an expert diagnosis.
Now according to our authors, if American radiologists were put out of business because doctors could obtain MRI analysis more cheaply by using the imported Asian black boxes, then no tariff protection would be warranted. This would be a classic example of comparative advantage, in which the Asian product could provide the service more efficiently than its American counterparts. Although the American radiologists might be worse off because of the Asian innovation, Americans in general would be richer, because of the lower price of MRI analysis made possible by the cheap imports. Some poorer Americans might actually owe their very lives to the low-cost analysis made possible by the imported boxes.
So again I ask: Do our authors really want to argue that this second, imaginary case would not warrant tariff protection for the American radiologists, while the first case would? Do they really think there is a significant difference between Asian radiologists programming their knowledge into a black box and shipping it overseas, versus Asian radiologists (after receiving MRI data) typing their diagnosis into a computer and emailing it overseas? If the Asian innovation represents a boon to the average American in the hypothetical example, why shouldn't we likewise applaud the real-world ability of Asian radiologists to diagnose "the problem at a small fraction of the cost"?
A Sympathetic Treatment
Although we have seen that the "two recent examples" cited by our authors do not importantly differ from standard mercantilist (and long-since refuted) arguments against free trade, we should still take a moment to understand what has led our authors—in particular, a generally free market economist like Paul Craig Roberts—astray. I think Roberts may have had something like the following in mind:
First let's imagine a scenario with immobile factors of production (and which thus falls under Ricardo's classical assumptions). Suppose that a US capitalist owns a tractor, and that there is a US farmer who is an expert at growing coffee. For a while the capitalist hires the farmer to use his tractor and grow coffee, which is sold to US consumers.
But then this happy arrangement is interrupted when a Brazilian farmer, without using any tractors, grows coffee and ships it to the United States. Because the Brazilian farmer is willing to work for only the equivalent of 25 cents per day, his coffee is much cheaper than the US brand. In the face of this merciless competition, the US capitalist fires the farmer and hires a different farmer to grow tomatoes with his tractor.
Now we can imagine the displaced US coffee farmer running to Congress and demanding a tariff on Brazilian coffee in order to "protect jobs." But economists like Paul Craig Roberts could inform the coffee farmer that, as Ricardo showed way back in the 19th century, Brazil has the comparative advantage in coffee production. It is more efficient to use US labor and capital goods in the production of crops like tomatoes. The reason US workers don't accept wages of 25 cents per day is that their labor is far more productive in other lines.
So much for the first scenario, with immobile factors of production. But now let's relax this classical assumption. Assume that the US capitalist, facing competition from the Brazilian coffee growers, still fires the US coffee grower. But now, instead of using his tractor in a different industry in the United States, our capitalist ships his tractor to Brazil where he hires a Brazilian coffee grower. (We assumed that this option was not profitable in our discussion above, perhaps because shipping costs were too high.)
Again, we can imagine the displaced US coffee grower running to Congress, demanding tariffs on Brazilian coffee. He could correctly point out that a high enough tariff would render US coffee production profitable once again, and thus prevent the US capitalist from shipping his tractor to Brazil. Surely not only the wages of US coffee growers, but average US wages, would be higher if Congress enacted policies to keep capital goods within the United States!
At first, this type of reasoning contains a superficial plausibility. After all, how can the free trader possibly argue that US citizens will have a higher standard of living if US workers have fewer capital goods with which to augment their labor? After the tractor is shipped to Brazil, won't fewer physical things be produced in a given time period within US borders?
Yes, other things equal, fewer physical goods will be produced within US borders when capital goods are shipped to other countries. But so what? What matters is not how many physical items are assembled in the US, but rather how much enjoyment US citizens get from the goods that they end up consuming. And even in the case of the tractor being shipped to Brazil, US consumers benefit from the Brazilian imports.
What happens is this: The Brazilian coffee growers are (by stipulation) willing to work for 25 cents per day, and now on top of this they are aided by a US tractor. Thus their productivity will increase, meaning the price charged to US consumers for Brazilian coffee will fall even lower. Although it is true that fewer physical things will be produced in the US, nonetheless these items will still fetch more consumption goods through international trade.
Whatever it is that Brazilians are importing from the US in exchange for their coffee, will now see an increase in demand because of the higher productivity (and hence real purchasing power) of the Brazilian coffee growers, due to their use of the new tractor. (This is simply the operation of Say's Law.) In short, just as lower coffee prices made US consumers better off in our first scenario when the tractor was merely shifted to another US industry, so too do much lower coffee prices make US consumers better off when the tractor is shifted to Brazil.
The case for free trade remains solid. The citizens of the US are not made richer by raising taxes or other barriers to foreign consumption goods, and this is true whether factors of production are immobile (as Ricardo assumed) or mobile. We should not fear the cost-cutting advancements in data transmission, or the improved skills and education of foreign workers. On the contrary, we should welcome these developments because they mean lower prices for imported goods and services, and hence a higher standard of living for Americans.
- 1. In fairness, I should note that Schumer and Roberts say, "Old-fashioned protectionist measures are not the answer, but the new era will demand new thinking and new solutions." However, nowhere do they specify how their anti-free trade position differs from "old-fashioned protectionist measures," and as a consumer I for one am not eager to see what Senator Chuck Schumer has in mind.