Beyond Belief

Some economists are beginning to doubt the benefits of free trade. What’s wrong with them?

Paul Samuelson, an undisputed titan of 20th-century economics, was once challenged by the mathematician Stanislaw Ulam to name a single proposition in all social science that was both true and nontrivial. It took a while, but Samuelson finally thought of a good answer: the principle of comparative advantage. This classical theory was true, Samuelson explained, as a matter of mathematical deduction, and its nontriviality was “attested by the thousands of important and intelligent men who have never been able to grasp the doctrine for themselves or to believe it after it was explained to them.”

ONETIME DEFENDER of free trade Paul Samuelson, in 1950.

The doctrine in question, devised by David Ricardo in 1817, makes a strong claim about the gains that accrue from trade. It would be difficult to exaggerate the centrality of the idea in modern economics. For nearly 200 years, the principle of comparative advantage, and the ideas about economic policy that flowed from it, divided the world into two camps: those with basic economic literacy, and the rest. Understanding this idea, and advocating it to the world, was part of what it meant to be an economist—especially an American economist.

Lately things have changed. Some of America’s most eminent economists, including Samuelson himself, have edged away from that earlier consensus. Their support for liberal trade is far more tepid and tentative than before. The shift is both momentous and disturbing. Just why it happened is a mystery.

To understand this issue, one must first understand what the principle of comparative advantage does not say. Intuitively, trade between two countries will make both better off so long as each is especially good at making something different from the other. They should specialize and trade to maximize this absolute advantage—as Adam Smith had earlier advocated in The Wealth of Nations. Ricardo’s idea was subtler—and remarkable. He showed that there are mutual gains from trade even when one country is better at producing everything. All that matters is that its margin of superior efficiency is greater for some products than for others. The two countries should still specialize, even on the basis of this comparative advantage—if England is slightly worse than France at making wool, and much worse at making wine, it should specialize in wool and trade. However unproductive a country may be, even if it is uncompetitive across the board, it gains from trade.

The proof is a few lines of math, as anyone with an hour or two’s training in economics will know. Nearly everybody else, as Samuelson said, finds the idea difficult to believe. People sometimes think they believe it when they don’t. The term comparative advantage is widely used, to be sure, but absolute advantage is what the politician or pundit usually has in mind. If I may speak for those with an hour or two’s training, we used to find this confusion quite gratifying. “What happens if a country has no comparative advantage in anything?” people would ask, gravely. How we laughed.

Ricardo’s theory did not cover every circumstance. Exceptions to its general rule (potential benefits from protecting “infant industries,” for instance) were recognized long ago. Nor did the theory claim that everyone gains; only that gains exceed losses. Nonetheless, the idea and the literature that grew up around it created a strong presumption that free trade was best. Exceptions were narrow and often exotic. The case for liberal trade commanded close to universal agreement among America’s leading economists.

Astonishingly, Samuelson himself struck the heaviest new blow against the discipline’s confidence on the issue. In 2004, aged 89, he roused himself to a burst of indignation, in the pages of the American Economic Association’s Journal of Economic Perspectives, at conventional defenses of globalization. They were simplistic, he later told The New York Times. Under certain circumstances, he pointed out, the losses from trade could exceed the benefits, not just for particular industries but for the economy as a whole. BusinessWeek said Samuelson and others were “beginning to question the basic tenets of free-trade theory … Is it possible that David Ricardo’s economic analysis doesn’t work for the 21st century?” Trade skeptics were exultant. Vindication from on high!

Since then, mainstream economists’ disenchantment with the old near- certainties has continued to build, and at a gathering rate. Top-tier orthodox economists such as William Baumol, Alan Blinder, Paul Krugman, and Brad DeLong—invariably prefacing the point with the words “Although I’m no protectionist—have expressed new fears about what imports and offshoring are doing to living standards. Lawrence Summers, Bill Clinton’s treasury secretary from 1999 to 2001 (and before that a revered mainstream scholar), is the most surprising new doubter. In a July Financial Times forum, he wrote, “It is not even altogether clear that [globalization] benefits America in aggregate.”

Why has this happened? What has changed? I wish I could tell you. No new theoretical insight has emerged to challenge the old pro-trade presumption. Samuelson’s 2004 article was mistaken for that in some quarters, but was actually a muddled combination of erroneous new theory on offshoring and uncontested old theory about monopoly power in global markets. The novel part was conclusively rubbished by Columbia’s Arvind Panagariya, who showed that Samuelson had got his offshoring model plain wrong. So far, anyway, nobody has explained why offshoring needs a new theory; it is just another kind of trade. The old part—the idea that the United States might see its income fall if trade drives down the prices of its exports—was already encompassed by the earlier consensus.

Jagdish Bhagwati, a preeminent trade scholar, also of Columbia, helped to theorize this danger in the 1950s. As he explained, “immiserizing” trade can arise only under quite unusual circumstances. The principal benefit to a country from openness to trade is cheaper imports. In the ordinary case, additional imports may put some domestic producers out of business, but the displaced capital and labor get applied to more efficient new uses, so the economy as a whole still gains. Suppose, however, that the country had previously been collecting some monopoly profits on its exports. If new trade drives up its production of those goods, their price might fall—and in theory, they could fall by enough to outweigh the gains from cheaper imports. In practical terms, this is an unimportant exception. Bhagwati himself remains a trenchant defender of liberal trade.

"When the facts change, I change my mind,” said John Maynard Keynes. “What do you do, sir?” Theory aside, the past decade has supplied a lot of new facts—not least, rising inequality and a protracted stagnation of middle-class earnings. It seems natural to blame the quickening pace of globalization for this. Again, however, no careful examination of the new facts on earnings shows trade or offshoring to be more than minor culprits. When you look closely, the shifts in earnings and the shifts in trade fail to marry up: The periods when imports have risen most rapidly are not the periods when wage pressures have been most intense. Studies suggest that labor-saving technology is a much more powerful force. No empirical work even comes close to supporting the claim that globalization is failing to benefit America in the aggregate. Countless studies have shown, and continue to show, the benefits of trade. Yes, some industries have shrunk or disappeared because of trade, and their workers have suffered the consequences; the pace of this change has probably quickened lately; better policies to insure and compensate the victims are surely called for. But to say this is very different from supposing that open markets hurt the United States as a whole.

The new trade doubters, as their disclaimers insist, are not yet trade-policy revisionists. For the moment they would mostly agree that a presumption in favor of open markets is best—but a weaker presumption than before, to be expressed more tentatively. Militant support for free trade, they seem to feel, will offend opinion at a time when so many are experiencing economic stress. The way to maintain open markets is a softer sell—one that acknowledges, and even seems to validate, popular fears that trade is to blame. But this posture is surely unwise. The doubters are influential, especially in Democratic politics, and their seeming conversion is itself altering the political climate. Would the campaign positions of the leading Democratic contenders for the presidency—so different from Bill Clinton’s position on the issue—be so hostile to trade otherwise? Ideas do matter, and for the first time, America’s leading orthodox economists are failing to speak with one voice in defense of liberal trade.

Samuelson once regarded the principle of comparative advantage—the modern theory of the gains from trade—as nontrivial. I would go a little further and say it was the greatest gift that economic wisdom ever bestowed on humankind. In a way that Samuelson did not envisage, the doubts that he and others have expressed threaten to make that idea trivial after all—dismissed as nothing more than an arresting curiosity, apt to be oversimplified by blinkered pro-trade types, with no real policy content and no claim on politicians’ attention. What a tragedy that would be. And not just because economics would no longer have an answer to Stanislaw Ulam’s question.