Near the end of his 1817 treatise, "On the Principles of Political Economy and Taxation," David Ricardo advanced the "law of comparative advantage," the idea that each country—not to mention the world that countries add up to—would be better off if each specialized in the thing it did most efficiently. Portugal may be more productive than Britain in both clothmaking and winemaking; but if Portugal is comparatively more productive in winemaking than clothmaking, and Britain the other way around, Portugal should make the wine, Britain the cloth, and they should trade freely with one another. The math will work, even if Portuguese weavers will not, at least for a while—and even if each country's countryside will come to seem less pleasingly variegated. The worker, in the long run, would be compensated, owing to "a fall in the value of the necessaries on which his wages are expended." Accordingly, Ricardo argued in Parliament for the abolition of Britain's "corn laws," tariffs on imported grain, which protected the remnants of the landed aristocracy, along with their rural retainers. Those tariffs were eventually lifted in 1846, a generation after his death; bread got cheaper, and lords got quainter.
The case for free trade, embodied in deals like the North American Free Trade Agreement, the Permanent Normal Trade Relations with China, and the proposed Trans-Pacific Partnership, remains at bottom Ricardo's. And the long run is still working out pretty much as he assumed. The McKinsey Global Institute, or M.G.I., reported in 2014 that some twenty-six trillion dollars in goods, services, and financial investments crossed borders in 2012, representing about thirty-six percent of global G.D.P. The report, looking country by country, reckoned that burgeoning trade added fifteen to twenty-five per cent to global G.D.P. growth—as much as four hundred and fifty billion dollars. "Countries with a larger number of connections in the global network of flows increase their GDP growth by up to 40 percent more than less connected countries do," the M.G.I. said.
But the case against free trade seems also to be based on Ricardo's premises, albeit with heightened compassion for the Portuguese weavers and British wheat farmers. American critics—Bernie Sanders, earnestly; Donald Trump, cannily—argue that trade decimated U.S. manufacturing by forcing American products into competition with countries where wages, labor, and environmental standards are not nearly as strong as those in America, or by ignoring how some countries, especially China, manipulate their currency to encourage exports. Sanders launched his post-primary movement, "Our Revolution," in late August, with an e-mail to potential donors. The most conspicuous demand to rally the troops was opposition to the T.P.P. "Since 2001, nearly 60,000 manufacturing plants in this country have been shut down," the e-mail said, "and we have lost almost 5 million decent-paying manufacturing jobs. NAFTA alone led to the loss of almost three-quarters of a million jobs—the Permanent Normalized Trade Agreement with China cost America four times that number: almost 3 million jobs." These agreements "are not the only reason" why manufacturing in the United States has declined, the e-mail goes on, but "they are important factors."
Such evidence for why the T.P.P. should be thrown out is hard to dispute, since the e-mail doesn't say what jobs were gained because of past deals, or explore what other "factors" may be important. President Obama, the champion of the T.P.P., may grant that certain provisions of the deal might be strengthened in favor of American standards without agreeing with "Our Revolution" on what's bathwater and what's baby. What's clearer is that the anti-trade message is hitting home, especially among the hundred and fifty million Americans, about sixty-one per cent of the adult population, with no post-high-school degree of any kind.
The investor Jeremy Grantham in July wrote an op-ed in Barron's noting that some ten million net new jobs were created in the U.S. since the lows of 2009 (the actual number being fifteen million), while "a remarkable 99 percent" excluded people without a university degree. That's a crisis, not of unemployment but of unemployability, which backshadows skepticism about the T.P.P. and trade as a whole. Trump's lead over Hillary Clinton among less-well-educated white voters remains solid, in spite of his alleged sexual predations; a large number of voters remain drawn to his grousing about the balance-of-trade deficit—which he presents as if it were a losing football score. Clinton has apparently decided to pass up the teachable moment, pretty much adopting Sanders's anti-trade line, though her private views almost certainly remain more nuanced. In an e-mail exposed by the WikiLeaks hack, purporting to detail a conversation between Clinton aides, she allegedly told Banco Itau, a Brazilian bank, in 2013, that she favored, at "some time in the future," a "hemispheric common market, with open trade and open borders"—a curious case of a leak embarrassing a candidate by showing her to be more visionary and expert than she wants to appear.
The anxiety is understandable, but the focus on trade deals seriously underestimates the changes that have reshaped global corporations over the past generation. Trade, increasingly, is mostly not in finished goods like Portuguese wine. It is, rather, in components moving within corporate networks—that is, from federated sources toward final assembly, then on to sales channels, in complex supply chains. An estimated sixty per cent of international trade happens within, rather than between, global corporations: that is, across national boundaries but within the same corporate group. It is hard to shake the image of global corporations as versions of post-Second World War U.S. multinationals: huge command-and-control pyramids, replicating their operations in places where, say, customers are particularly eager or labor is particularly cheap. This is wrong. Corporations are hierarchies of product teams, which live in a global cloud. "Made in America" is an idealization.
The product manager of the Chevrolet Volt, which Obama singled out at the time of the auto bailout, told me in 2009 that the car was a kind of Lego build: the design was developed by an international team in Michigan, the chassis came from the U.S., the battery cells from Korea, the small battery-charging engine from Germany, the electrical harnesses from Mexico, suspension parts from Canada, and smartphone-integration software from Silicon Valley. More and more, the design of products and services happens in distributed hubs. The serendipitous sourcing of technologies and customer characteristics, the lowering of transaction costs, the trade flows enabled by accelerated financing and logistics—all of these—presume growing network integration and social media, the latter increasingly important to glean marketing data.
The point is that each component, and each step in production, adds value differently. Where value is added will depend on what corporate accountants call "cost structure": how much of the component or step requires local materials, or unskilled labor, or skilled labor wedded to expensive capital equipment, or high transportation costs, and so forth. Some components, like the fuel injectors assembled into the Volt's German engine, required high-technology production systems. Labor was a trivial part of the cost structure, and the engine could be built in the highest-wage region on the planet. Harnesses, in contrast, which required a much higher proportion of manual labor—and were relatively easy to ship—were inevitably sourced from places where workers make as little as a tenth as much as Americans, thousands of miles away. No tariffs can reverse this trend, and no currency manipulation can drive it. Between 2005 and 2012, the M.G.I. found, thirty-eight per cent of trade derived from "emerging economies," up from fourteen per cent in 1990.
In principle, the production of virtually any component or assembly operation could be robotized and moved to high-wage countries—but only so long as demand is great enough, and design specifications stable enough, to justify huge scale and hundreds of millions, if not billions, in upfront investments. But that's not good news for American workers with little education. Robotic lines employ few, comparatively highly trained workers. The Atlantic's Andrew McGill reported recently that American manufacturing output has actually continued to grow steadily, but employment in manufacturing peaked at nineteen and a half million, in 1979; today, only twelve million work in manufacturing. Deloitte figured that, in 2011, some six hundred thousand manufacturing jobs went unfilled in the U.S., largely owing to a dearth of skilled workers. So global trade only puts jobs that are already at risk more seriously and immediately at risk. Besides, the danger to the unskilled in higher-wage regions is not really new; the Rust Belt began forming because of North Carolina, not because of South Korea.
Nor does America's trade deficit indicate what anti-trade campaigners suppose. Douglas Irwin (my colleague at Dartmouth) writes in Foreign Affairs that the American economy is usually growing and creating jobs when the deficit grows, and it decreases when the economy is contracting. The deficit reflects, among other things, large capital flows from abroad, which sustain the dollar as a reserve currency; the deficit has remained roughly unchanged at about two to three per cent of G.D.P. over the past seven years. The jobs at Carrier in Indiana, which management is moving to Mexico—and Trump made much of, threatening tariffs on air-conditioners—have been more than made up for this year by jobs at Subaru and Honda plants in the state.
"We understand that many of our citizens are frustrated by the pace of globalization and feel they're not experiencing the benefits of international trade," Obama said after the latest G-20 meeting, in Hangzhou, China; the challenge, he said, is to "spur economic growth, to boost free trade, and build a fairer economy that truly works for all." How fairer? The answer cannot be working to unravel the global networks and trade flows that converge in a knowledge economy; the advisory firm Ocean Tomo estimated last year that eighty-seven per cent of the assets of S. & P. 500 corporations are now "intangible": the management of science and technology, the chemistry of teams, network effects, cosmopolitan values.
Fairness, rather, requires more government retraining programs, portable health care, cheaper community colleges, state investments in clean energy—in short, educating motivated workers and engendering lower-skilled jobs that cannot be exported. The most important of the latter will come from major infrastructure projects, like Boston's "Big Dig," which has been rashly vilified for its cost overruns and construction mistakes but eventually created thousands of well-paying, low-skill construction jobs, and ultimately joined the city's Financial District to the Seaport area, catalyzing scores of billions in new building, tourism, and business development. George Gendron, the former editor of Inc. Magazine, told me that the new "innovation district" in the Seaport, once derelict, is, thanks to the Big Dig, the most expensive real estate in Boston. "General Electric is putting up its new global headquarters in the Seaport, which will be networking with dozens of startups from M.I.T. and Harvard," he said.
Sanders's movement is calling for a trillion-dollar investment in this kind of infrastructure, supported by higher taxes on the wealthy; the Clinton campaign wants a little more than a quarter of a trillion, with no new taxes and an infrastructure bank. Trump, for his part, wants to spend "billions" along with a tax cut; his first major supporter, New Jersey Governor Chris Christie, famously killed a new commuter-train tunnel from the Garden State to New York City. These differences might well have been the focus of campaign coverage, and might yet be raised in the final debate. It is a sad tribute to Ricardo that they are overshadowed by a simplified resentment for trade. "Blaming trade for the loss of manufacturing jobs is an easy mark, and misguided, something like blaming immigration," Gendron said. "If somebody else is working, and you are not, that has to be stopped. There's a better way, but it will cost us."