The Global Economy: Who Gets HurtAaron Bernstein
You'll have to excuse the McAmis family of Greeneville, Tenn., if they cringe at talk of free trade. After 12 years of assembling Magnavox TVs at a North American Philips Corp. plant, Allen McAmis was laid off in February and his $13-an-hour job moved to Juarez, Mexico, where it pays $2 an hour. He was recalled in June to fill in for workers on sick leave. But Allen and his wife, Sherry, who also works at the plant, fear that Philips will move more jobs south. The couple have cut out birthday gifts and allowances for their children, Christina and Brian, and survey the future with dread. If Allen loses his job again and "I get laid off, too, we'll be a welfare family," Sherry says.
Ever since the British economist David Ricardo advanced the theory of comparative advantage in 1817, conventional economic wisdom has held that the benefits of dropping trade barriers--lower prices and higher growth--outweigh the loss of jobs and pay that some workers suffer in the process. Most analysts have held to this view as global competition soared in the 1980s and the U.S. negotiated a free-trade pact with Canada and contemplated one with Mexico. But now, a flurry of studies by disparate economists--everyone from middle-of-the-road academics to ardent defenders of Reaganomics--are finding that the tradeoff isn't nearly that simple.
SERVICE SURPLUS. These analyses still call growing trade a net plus. But they also pinpoint more precisely than ever who gets hurt, and the scope of their findings is sobering. In the 1980s, the first and worst hit by global trade were factory workers whose companies competed with foreign rivals. As these people were laid off or suffered wage cuts, they created a glut of job candidates that helped hold down pay among the 64 million workers, across a wide spectrum of industries, who never went beyond high school. Only the college-educated did well. These 54 million Americans, blessed with high skills and fortified by the fact that new technologies boosted demand for the work they do, were insulated both from foreign competition and the struggles of the less educated.
These trends sound familiar, but their broadest implication is news: The increase in trade bears much of the blame for an unprecedented surge in income inequality between the most- and least-educated halves of the U.S. work force. To cite the extremes, the real wages of high school dropouts have fallen by up to 20% since 1979, while real incomes of employees with more than four years of college have grown by 8%, according to the Economic Policy Institute, a liberal think tank in Washington, D.C. Marvin H. Kosters, an ardent free-trade advocate and director of economic policy studies at the conservative American Enterprise Institute, has just edited a book called Workers and Their Wages, which tries to explain the new inequality. It results, he says, from the potent combination of "trade and technology."
What makes the conclusion especially unsettling is that trade is bound to pick up. U.S. and Mexican negotiators are close to reaching a North American Free Trade Agreement that will phase out many trade barriers between the two. This will no doubt create U.S. jobs as exports to Mexico increase--but will also send more lower skilled ones south. Meanwhile, the seven major industrial countries are trying to expand the 108-nation General Agreement on Tariffs & Trade (GATT), which will leave labor-intensive industries such as textiles more vulnerable to foreign rivals. And countries such as Malaysia and China are rapidly boosting exports. All this "will have a dramatic impact on lower skilled workers in the U.S.," predicts Edward E. Leamer, a trade economist at the University of California at Los Angeles.
CLASS WARFARE? The impact will reach much further than that. Declining pay for the bottom half may not slow U.S. growth, since average incomes should rise as the top half does better. But widening inequality poses other problems. The poverty rate could stay up. The tab for welfare and unemployment could mount, inflating taxes. Ultimately, resentment of the wealthy could reach a boiling point, leading to ferocious attacks on executive pay and even to more riots like last May's in Los Angeles. "One possibility is for us to become a class society like those in Latin America," which have unequal distributions of wealth and chronically unstable governments, says Richard B. Freeman, an economist at Harvard University. "That's the direction we're headed."
This will be a potent issue in the Presidential campaign. Democratic candidate Bill Clinton, who in accepting his nomination vowed that "American companies must act like American companies again--exporting products, not jobs," will attack the Reagan-Bush legacy of wage stagnation for all but the most well off. To boost competitiveness, Clinton proposes to spend $60 billion by 1996 on education and training. At the Republican convention on Aug. 17, President Bush is likely to argue that his proposals for cuts in spending and taxes would have spurred growth already if only the Democratic Congress had enacted them.
Neither candidate embraces protectionism, and other strategies would work. For one, Corporate America could compete internationally by adopting a high-skills strategy. Rather than push pay to the lowest common denominator, companies such as Deere, Ford, and Motorola are training employees to improve their skills, boost productivity--and keep jobs at home. If most companies did likewise, the ranks of the less skilled would shrink, easing the glut of them. To hasten this trend, Washington could spend more on education and training, for example, or curtail immigration of low skilled workers.
For the moment, however, Washington seems unwilling to act. And for every Motorola or Ford, a trend-setter such as AT&T is turning high-paying jobs into low-wage ones. Even those that upgrade worker skills, such as Ford Motor Co. and General Electric Co., still shift work to cheap-labor countries--thus pursuing both approaches.
Meanwhile, the bottom half faces another ominous trend. Service industries--banks, insurers, and retailers, among others, which employ 75% of the work force--are automating jobs and displacing lower skilled workers. Since 1990, Sears has dumped 21,000 positions this way. "If service productivity rises in the '90s, the question is how neutral this will be between educated and less educated workers," says University of Chicago economist Kevin M. Murphy. "If the gains come mostly in low-skilled jobs, the story will be the same as for manufacturing in the '80s."
The specter of inequality hasn't changed most economists' conviction that globalization is good. Economic theory has held for decades that as foreign countries sell more to the U.S., they use the revenue to buy U.S. exports. They also set up shop and create jobs in the U.S., as the Japanese and Germans are doing. The latest example: A $400 million BMW plant that will bring 2,000 jobs to Spartanburg, S.C. There are other benefits, too, such as a freer exchange of technology and managerial ideas. Historically, the net benefit of all this in terms of jobs and growth has been difficult to quantify. Still, economists project that a new GATT deal would add 0.5% a year to world GDP growth, and that free trade with Mexico would lift output in both countries and create tens of thousands of new U.S. jobs.
TECHNO TRADE. There's at least equal proof of the link between trade and inequality. Wolfgang F. Stolper and Paul A. Samuelson, the father of neoclassical economics, first expressed the concept in 1941, with elegant equations that built on Ricardo's comparative advantage theory. If goods trade freely, the idea goes, prices will equalize, and so will production costs. To compete, countries must specialize where they have a relative edge. Low-wage countries will make labor-intensive goods, while those with capital will do better in technology-intensive products. Thus, U.S. low-skilled work should flow overseas, or wages of low-skilled workers must fall.
This didn't happen for much of this century, largely because trade--the sum of exports and imports--was less than 10% of U.S. gross domestic product. In the past 20 years, as that figure grew to 25%, imports began to displace low-skilled workers or depress their wages in industries such as textiles, auto parts, and electronics. But the U.S. wage gap didn't widen: The baby boomers, plus an influx of working women, produced a surplus of college-educated employees, which kept their incomes from rising faster than those of the less educated.
These forces reversed in the 1980s. The rush of college graduates eased, reflecting the baby bust and the slowdown of women entering the work force. New technology investment, spurred in part by foreign competition, swelled demand for college grads such as engineers and technicians, so their pay rose. Meanwhile, 350,000 less skilled immigrants entered the U.S. each year, and soaring imports eliminated factory jobs. Soon, "workers displaced by trade were competing in service industries," says Harvard economist Lawrence F. Katz.
American Telephone & Telegraph Co. illustrates how all this happens. Since 1984, it has eliminated 21,000 blue-collar manufacturing jobs in the U.S., leaving 32,000. It also has created 12,000 mostly lower paying factory jobs elsewhere. Meanwhile, its U.S. white-collar work force has grown by 8,000, to 118,000. "We're losing low-skilled jobs overseas or to technology," says William J. Warwick, president of microelectronics manufacturing at AT&T. "I need more engineers, designers, and high-skilled technicians, but fewer people who make circuit boards, the lowest-skilled job."
Jacqueline A. Gregory proves his point. A 29-year-old electrical engineer at AT&T's Allentown (Pa.) microelectronics facility, she was hired last September to help develop software that AT&T customers use to design semiconductor chips. AT&T created Gregory's group of six engineers shortly before she signed on, then added four positions. "We're now trying to hire three more people," she says. And Gregory is back in college part-time, aiming for a masters degree.
BORDER CROSSING. Dreama B. Fields, 33, (cover photo) has had the opposite luck. She worked for nine years as an assembler and packer at an AT&T plant that made transformers and other devices in Radford, Va. Then, in 1990, after moving work to Matamoros, Mexico, for five years, AT&T shut the plant. By then, 800 of Radford's 2,100 blue-collar jobs had migrated south. Labor comprised about 50% of production costs at Radford, says Warwick. He won't give the figure in Mexico. But workers there earn an average $2.35 an hour in wages and benefits, vs. the $13 Fields earned.
Fields has suffered less than many of her peers because her husband, James, has a job as a state park ranger. Still, the couple adopted a more modest lifestyle so she can attend community college. They buy second-hand clothes for Audrey, 4, and often don't spend the $10 a month it costs to let Candi, 13, enter bowling tournaments. Fields hopes accounting classes will help her get work as a payroll clerk--a $5-an-hour job in Radford. "It will be years before I get back to where I was," she says.
Even workers who still have jobs feel the fallout. From 1979 to 1989, as the sum of exported and imported goods rose from 55% of U.S. manufacturing output to 82%, the average pay and benefits of U.S. factory workers fell 6% after inflation. Meanwhile, plant productivity rose 42%. In fact, factory pay in Europe and Japan, where companies and governments train workers, gained on the U.S. average after inflation and exchange-rate adjustments. German factory compensation has passed U.S. levels for the first time in modern history.
This trend probably won't turn around soon. UCLA's Leamer recently tried to project how much free trade would cost less skilled workers in the future. To do so, he looked at how trade changed prices for goods from 1972 to 1985 in 38 labor- and capital-intensive manufacturing industries, then figured the pay effect on professional and blue-collar employees across the economy.
He found that trade raised demand for higher-skilled work, as Stolper and Samuelson predicted. So the average annual pay of 17 million professionals was 9%, or $1,900, higher than it would have been without trade growth. But annual pay of more than 90 million other workers was $465, or 3%, lower than it would have been otherwise. Leamer predicts that lower-skilled workers will suffer similar declines in the 1990s--and that professionals will reap similar gains--if manufacturing trade grows the 60% it did in the period he studied. "Gains from the expansion of trade will continue to be very unequally distributed," he adds.
This may be true even if trade growth slows, as the mere threat of exporting jobs keeps pay down. In June, Philips wrung concessions from the McAmises and 1,600 colleagues who have so far survived the transfer of 900 jobs to Juarez. The International Union of Electronic Workers accepted $1.50 an hour less for new hires than current workers get--and no benefits. These conditions will apply to 600 or more laid-off workers who may be recalled. The union consented, officials say, because Philips threatened to shift all TV production to Mexico and leave 300 distribution jobs in the U.S. Says F. Joseph Brang, the Greeneville plant manager: "There is great pressure to move work to Mexico because of the low labor costs there."
Free-trade advocates such as Massachusetts Institute of Technology economist Rudiger Dornbusch argue that the loss of lower-skilled jobs is not only inevitable in a world awash with cheap labor--but may be a blessing in disguise. If trade growth indeed spurs exports of value-added products, where the U.S. has an edge, then higher-paying jobs should spring up to replace lower-paying ones that are lost, he says.
POSITIVE STATIC. But for this to occur, most companies must try the approach taken by those such as Motorola. Today, only 44% of its 100,000 employees work in the U.S., vs. nearly 100% in 1960. Part of the shift overseas reflected a desire to manufacture in its best markets. But nearly half of the change stemmed from a search for low wages, says A. William Wiggenhorn, Motorola's vice-president for training and education. Then, in the early 1980s, Motorola reexamined itself in light of moves it was making to just-in-time inventory control, total quality management, and teamwork systems. More than cheap labor, "today, we want high quality and a short time to market," says Wiggenhorn. "So we don't move employment to low-cost countries anymore."
Take Motorola's cellular phone business. When the company started making the phones in 1983, it eliminated larger parts by putting their functions on computer chips. This required less assembly, cutting labor costs to 10% to 15% of production costs, vs. 25%-plus in previous products. The company also switched to a labor relations strategy that encourages its 3,000 cellular phone employees in Arlington Heights, Ill., to boost quality and productivity.
Last year, for instance, Lemel Lewis, who makes circuit boards for phones, joined a plantwide team formed to reduce static in the air, which causes defects as boards are produced. The team tested various antistatic packages used to move the boards along the assembly line and "discovered that a clamshell-shaped package is the best," says Lewis. The change has helped boost quality and keep Motorola ahead of its rivals abroad.
A few companies are even trying the higher-skills approach in apparel, one of the most labor-intensive industries. In traditional piecework, each worker at a machine sews the right pocket, say, onto 30 garments that arrive in a bundle. Then the bundle moves to the next machine for the left pocket. Recently, companies such as Russell, Levi Strauss, and Sarah Lee Corp.'s Hanes have begun to create teams of 30 to 50 workers that make an entire garment. Operators within each team move between machines to eliminate bottlenecks. And they, not supervisors or engineers, decide how to set up the line.
Levi Strauss & Co. just began converting its 27 U.S. plants last fall, so it doesn't yet know if teams boost productivity. But other benefits are clear. Each plant now can make a bundle of 30 jeans in seven hours from start to finish instead of the six days it often took as bundles got stuck at different stages of the old system. So a factory can rapidly switch styles and sizes to respond to consumer demand.
MORE GRADS? At Levi's 432-employee Blue Ridge (Ga.) plant, which last spring became the first to convert entirely to teams, defects fell to 1.9% of production in April, from 2.6% during the same month in 1991. "Our factories couldn't compete with overseas ones on a pure cost basis," says Peter Jacobi, Levi's president of global sourcing. "If the team system succeeds, there are a lot of products, like shirts and Dockers jeans, that we could produce here" at current wage rates.
Foreign factories may not be far behind, however. One of General Motors Corp.'s highest-quality plants is in Ramos Arizpe, Mexico. And AT&T's Warwick uses teamwork in Juarez. "The average education level in Mexico is about the ninth grade, but it doesn't take more than that to use new manufacturing techniques," he says.
If U.S. companies' efforts to boost skills don't reverse wage inequality, it's unclear what will. Until recently, Kosters, among others, thought labor market forces might narrow the gap. He notes that college enrollment rates jumped from 33% of 14- to 24-year-olds in 1984 to 39% in 1990. If the lure of higher pay hastens this trend, then in theory the supply of college grads should eventually match the demand for them. And a resulting shortage of lesser skilled workers could ease the downward pressure on pay for these people.
The problem is, this scenario requires a lot more college grads. John H. Bishop, an economist at Cornell University, puts the figure at 1.2 million a year, vs. 1 million now. Both President Bush and Democratic candidate Clinton support tuition-loan programs that would help more students finish college. But boosting the graduation total will be hard to do, given that there are only 25 million 18- to 24-year-olds today, vs. 30 million in 1979. In fact, says Bishop, based on these demographics, "I project a slowdown in the growth of college-educated workers and a continuing escalation of wage premiums for college graduates."
Companies and government could help even things out by training more existing workers. Washington now spends about $4 billion a year on training. Employers spend $30 billion more, mostly to upgrade the skills of professionals. To match Japan and Germany, U.S. industry should be investing $15 billion more a year, says the American Society for Training & Development. "Even if you need $60 billion a year in training and education, we should spend it," argues MIT's Dornbusch. "What else can you do, given that we're now seeing the results of not doing so, like the L. A. riots?"
Economists once thought that the inequality of the past decade occurred mainly because sluggish productivity gains slowed growth--and because lower-wage workers suffer most when growth stagnates. They assumed that as productivity rose, wage inequality would recede. But global competition may now prevent this. "You can't expect an automatic link anymore between the overall economy and less educated workers," says Harvard economist Katz.
This isn't an argument for protectionism, with all its negative effects. Instead, it should prompt debate on ways to ensure that the benefits of free trade don't go just to America's top half. "If there really is a net gain from trade, we ought to compensate the losers" with subsidized training or lighter tax burdens, argues Andrew M. Sum, an economist at Northeastern University in Boston. Otherwise, there will be no lifeboats for Allen McAmis, Dreama Fields, and millions more whose ship is sinking.