It got to be something of a regular event. In the early months of 1982, Xebec chairman and chief executive officer Jim Toreson would stretch out in his Sunnyvale, Calif., office and listen as the visitors from Southeast Asia made their pitches. Hailing from such low-wage havens as Taiwan and Korea, they would lay out elaborate offshore manufacturing schemes that, they promised, would turn Toreson's eight-year-old disk-drive controller company into the next high-growth superstar of Silicon Valley.

On the surface, the argument seemed so simple. Given Xebec's well-developed engineering and marketing capabilities, Toreson could use offshore manufacturing to ramp up his production, taking full advantage of both low-wage labor and a grab bag of tax holidays, low-interest loans, and other government largess. Most of his competitors seemed to be doing it -- the faster he followed suit, and the better off Xebec would be.

"The temptations were tremendous," recalls the 42-year-old entrepreneur, who has steered Xebec from $14.7 million in sales in 1981 to $158 million in the fiscal year ending in September 1984. "But as an engineer and an American, the whole thing made me sick. It implies we can't make anything in this country anymore. It just created a sense of challenge in me to prove everything they were saying was bullshit."

Wounded national pride aside, Toreson was thinking strategically as well. For too long, he believes, American entrepreneurial companies have relied on continued technological breakthroughs to fuel their long-term growth. Yet "the distinction between being a commodity producer and high tech doesn't exist in our world anymore," Toreson says. "You have to be both or you can be neither. The question is, How do you maintain your edge? It's not enough to be brilliant marketers or designers. You have to combine good product technology with the best manufacturing technology. You need the one-two punch or you'll end up, in the long run, as consultants to the Japanese."

Determined to avoid that fate, Toreson decided to bet Xebec's future on a radical new idea: staying home. Although profits have declined due to the current shakeout in the microcomputer marketplace, Xebec is now spending some $30 million to implement an onshore manufacturing strategy. Toreson directs a far-flung operation that includes two highly automated factories in Nevada, another in Pennsylvania's Lehigh Valley, a major computer-aided design center in Reno, Nev., and $20 million worth of the latest generation of IBM robots. His long-term goal: to turn Xebec into a billion-dollar company by the end of the decade.

"At some point," he says emphatically, "Americans have to take up the Japanese challenge in mass production and with high quality. We have to stop this tendency to say, 'Screw it, let the dumb Japanese or dumb somebody elses do the manufacturing.' The problem is, they usually turn out not to be so dumb."


As a disk-drive controller company, Xebec competes in one of the most volatile sectors of the turbulent high-technology marketplace, and as a result, the company's production problems might seem something less than typical. But if you examine Toreson's reasoning carefully -- and in particular, his ideas on how to manufacture successfully at home -- the broad impications of his decision become clear. The future of all American manufacturing, and, perhaps, of the American economy itself, may hang on the kind of choices Jim Toreson had to make.

This is not to suggest that Toreson is part of any back-to-the-U.S.A. mass movement. While a handful of other leading entrepreneurial companies (such as Micron Technology, Tecstor, and Apple Computer) are pursuing strategies similar to Xebec's, there is little evidence that the decades-long trend toward offshore manufacturing is slowing down. In fact, most small American high-tech businesses seem increasingly resigned to a future in which they will serve, in the words of one Hong Kong engineering manager, as "prototype shops" for more efficient offshore producers.

"You do what you do best and let us do what we do best," suggests George Koo, the Chinese-born president of Microelectronic Business International Inc., an equity investment firm in Mountain View, Calif., that actively solicits American companies for Taiwanese high-tech manufacturing ventures. "Americans are great at inventing and marketing products, but we are the ones who know how to manufacture well and cheaply. The people in the Far East are better. They have more of a work ethic and work longer and harder than you do."

Koo's arguments echo the widely accepted belief that America's decline in manufacturing is really no cause for concern -- that it is simply a proper and natural outgrowth of our evolution toward an "information economy" dominated by service industries. Megatrends author John Naisbitt, a leading proponent of the offshore-is-beautiful scenario, reassures us that while Japan may be number one in manufacturing, that is like being a "new champion in a declining sport." The events of the past three decades, however, can be seen in a much less reassuring light.

Granted, some export-oriented American companies, such as IBM Corp. and Hewlett-Packard Co., have used their foreign manufacturing operations as an adjunct to continued domestic production, and as a means of effectively penetrating foreign markets. But other companies' moves offshore reflect a very different, and dangerous, strategic focus. Faced with mounting competition from Asian producers, these companies have simply closed up shop at home and moved their factories overseas, sometimes even contracting to have their products built for them by natural competitors in such countries as Japan.

Such strategies are often justified in terms of immediate cost savings, but the long-term costs may be staggering. U.S. companies that eschew domestic manufacturing for offshore production may end up losing both their market share and their technological edge. "The thing that makes offshore production so dangerous is that it puts you in the mentality of letting the other guy do the hard stuff," says Harvard Business School professor Robert Hayes, a leading expert on manufacturing strategy. "When the other guy enters the market, he's worked with the process on a daily basis, has a sense of the wider potential of the technology, of possible applications that you wouldn't have been thinking about. He's got all the advantages. Eventually, he who can do nothing but sell is at a great disadvantage."

Perhaps nothing more starkly reveals the risks of this fast-track offshore strategy than the American debacle in consumer electronics. Virtually the uncontested world leaders in this field during the 1960s, American companies started shifting production to low-wage countries like Taiwan in the late 1960s and '70s as a means of countering the assault of efficient Japanese manufacturers. The American companies were committed to using low wages as their "process innovation," according to New York City attorney James Millstein, who follows the consumer electronics industry closely. And this tactic made them reluctant to implement the kind of new, automated manufacturing techniques being used in Japan, or to bother applying such U.S.-invented technologies as solid-state circuitry and video recording.

In part due to such shortsightedness (and despite protectionist measures covering such items as television sets), by 1982 the United States was importing nearly half of all consumer electronics products sold here -- some $4.5 billion worth. Nor is the situation getting better. As William Relyea, an electronics industry analyst with F. Eberstadt & Co., states flatly: "Leadership in consumer electronics has passed to Japan, and it isn't coming back."

This loss of manufacturing leadership -- repeated in other fields, such as automobiles and steel -- has become so serious that American merchandise trade deficits reached $61.1 billion in 1983 and topped $100 billion in 1984. Prowess at the "declining sport" of manufacturing in such nations as Japan and Korea has generated economic growth rates that are more than twice as great as this country's over the past two decades. And these same countries are now launching a powerful assault on the information industries that even Naisbitt sees as crucial to America's long-term economic future.

Nevertheless, enterpreneurs in these emerging industries seem hell-bent on following the consumer electronics companies' lead. Instead of solving their manuacturing problems at home, many of high tech's most prominent young companies (including such now troubled businesses as TeleVideo Systems Inc. and Atari Corp.) have chosen to shift their production overseas -- all in search of an illusory panacea to the fundamental manufacturing challenge posed by the Japanese.

Nowhere has this trend been more pronounced than in the disk-drive industry. Even as Toreson gears up for onshore production of his new Owl disk drive, nearly all the established industry leaders -- among them prime Xebec competitors Tandon Corp. and Seagate Technology -- have committed themselves fully to offshore production.

Many analysts, looking at the short-run costs, encouraged these manufacturers to go offshore, and their early success made this advice look good. Yet today, these same companies suffer from some unanticipated problems that followed the offshore decision (see "Winning the Battle, Losing the War?" page 60). Tandon has developed a reputation for uncertain quality, and its inventories have risen to unhealthy levels. Seagate also shows signs of trouble, attributable in part to poor coordination among the marketing, engineering, and manufacturing divisions that resulted from the company's move to Singapore.

"It's becoming clear that for Tandon and Seagate, offshore wasn't the salvation they thought it would be," says Blake Downing, a partner and computer-industry analyst at the San Francisco-based investment bank of Robertson, Colman & Stephens. "Maybe we're at a stage of economic development where we're simply not very good at manufacturing. It sure seems that way."


But what if it is true that we are "not very good at manufacturing?" Does it really matter? When you can head off to a faraway production heaven with deferred taxes; cheap credit; low-cost, highly motivated labor; and even state-of-the-art equipment, who needs homegrown assembly-line skills?

"Every week it seems I get two or three propositions telling me why I have to move my operations to Hong Kong or Jamaica," claims Leonard Bleininger, president of Tecstor Corp., a four-year-old manufacturer of high-density disk drives in Huntington Beach, Calif. "No one is going around trying to convince you to build your plant in this country. Nobody seems to think that there may be advantages to manufacturing at home."

Yet Bleininger, like Jim Toreson and a number of other American entrepreneurs, believes that there are compelling advantages to staying onshore. In the short run, they argue, these include such benefits as better quality control, ability to adapt quickly to changing markets, better communication with customers, and in some cases -- surprisingly -- lower costs. Even more important, they say, are the long-term gains. For companies to survive in an increasingly automated future, they will need to control manufacturing automation, developing the human resources to take full advantage of the tremendous investment it represents. This is no less true for an aluminum foundry in Missouri, where manufacturing offshore was never in the cards (see "Lessons for the Rust Belt," page 59), than it is for a go-go high-tech start-up like Xebec, which could put a plant in Taiwan in a matter of months if it wanted to.

Let's look first at the immediate benefits that, according to onshore proponents, a home-court strategy has to offer.

1. THE CUSTOMERS GET TO KICK THE TIRES. Leonard Bleininger's company sells a line of sophisticated disk drives. In many cases, these drives, which cost from $6,000 to $50,000 apiece, have to be altered to fit the speciications of individual customers. In an average month, for instance, Tecstor may have to turn out as many as 15 versions of its basic drive.

To meet the exacting needs of these customers, Bleininger believes, there must be close personal contact between Tecstor's engineering and production staff and representatives of a purchasing company, something all but impossible to achieve over 10,000 miles and with severe language and cultural barriers.

"Maybe you can save a few dollars by building parts in Singapore," says Bleininger, whose company will have sales of around $10 million for the fiscal year ending this month, double last year's. "But when you're dealing with $100,000 machines, you have to meet the specifications of the systems house -- and to do that from offshore would be ruinously expensive. We have to be very close to what the customer wants or we're sunk."

Priam Corp., a manufacturer of a broad range of disk drives, rejected offshore production in favor of a highly automated, $10-million factory in San Jose, in part to maintain closer links with present and potential customers. "One advantagewe have found is that our customers can come here and kick the tires," says Joe Smith, vice-president of operations. "They feel good about it. They can see the process we use, the product being made. They're a lot less likely to do that in Singapore."

With computer product life cycles narrowing, explains Priam president William J. Schroeder, working shoulder-to-shoulder with customers on the specifications of the next generation is increasingly imperative. And "75% of our potential customers are in the United States," he says. Last November, Priam projected a $3-million to $4-million loss in the fourth quarter of calendar-year 1984 after making $3.2 million on sales of $44 million in the first half of the year. But Schroeder doesn't blame his current problems on his onshore location. After all, he points out, one of his keenest competitors -- Fujitsu Ltd., a $5.4-billion Japanese computer and electronics company -- announced plans last September to set up its new disk-drive operation near Portland, Ore.

2. QUALITY CAN COME FIRST. Of course, a huge company like Fujitsu can fall back on manufacturing expertise developed back in Japan. But young American companies don't have such an option. If they want to match the high standards of their Japanese competitors, Tecstor's Bleininger believes, they first must perfect their manufacturing techniques under their own roofs.

In building long-run relationships with customers, Bleininger says, quality and serviceability are simply more important than any slight price advantage. "There are costs to being offshore -- such as poor quality and being out of touch -- that people just don't figure on," he explains. "When Tandon went offshore, he became dependent on offshore for his quality."

The same kind of offshore quality problems have surfaced in the semiconductor industry. For much of his career, Fritz Beyerlein built offshore semiconductor-assembly plants for such companies as Signetics, Rockwell International, and RCA. Now vice-president in charge of manufacturing for Cypress Semiconductor in San Jose, Beyerlein finds that onshore assembly produces quality improvement simply through management synergies. With his highly automated assembly operation next door to Cypress's engineering section, for instance, Beyerlein can provide assembly-related input in the early stages of Cypress's product designs -- something offshore assembly managers can rarely do.

3. ITS EASIER TO FILL ORDERS FAST -- AND CHANGE WHEN THE MARKET CHANGES. Beyerlein sees speed of delivery as another strong argument for onshore assembly operations. Products assembled offshore can take up to six weeks from shipment to return to the U.S. market. At Cypress, on the other hand, turnaround time can be as short as three days. This time differential can make a critical difference to a start-up that needs to show it can fill customers' shifting needs quickly and efficiently.

"In a start-up in this business, you can't fool around with long lead times," says Cypress president T. J. Rodgers. "If a customer wants our design, he wants it because we have it first. He doesn't want to wait six weeks for it to come back from Penang."

Nor is the speed advantage confined to high-technology industries. Quick turnaround, for example, can be crucial in the garment industry, which, like electronics, has been hard hit by foreign competition. (Between 1976 and 1981, apparel imports more than doubled, to better than $6.7 billion; they now account for more than half of the total market.)

The garment industry moves at a pace that would make even a semiconductor engineer's head spin; in women's apparel, for example, a manufacturer needs a new product line every six months. For domestic manufacturers, such as R.J.M.J. Inc., of New York City, responding quickly to the constant changes in fashion can help compensate for the cost advantages usually associated with offshore production.

Early last year, for instance, R.J.M.J., which manufactures women's slacks and shorts, ordered a small amount of a fabric called French canvas.So did most of its offshore competitors. But by the height of the season, the fabric was selling so fast that the retailers were screaming for more. Companies that rely on offshore manufacturers, which have to order their fabric and garments six to eight months ahead, could not respond to this unexpected demand. But R.J.M.J. president Robert Shipman could simply order the fabric overnight and have it processed at his wholly owned factory in Westminster, S.C., as well as through two trusted subcontractors in Florida and Georgia.

"You need a crystal ball in this business," says Shipman. "But once I knew it was hot I could move right away. I was able to call up for 100,000 yards of French canvas here in New York and start turning it around in three weeks. The importer can't do that."

So what started off as 50,000 yards of French canvas ended up as 300,000 yards, and R.J.M.J.'s sales of slacks made out of the material soon reached $3 million, up from around $60,000 the year before. Particularly gratifying for Shipman have been the 35 new accounts -- some of which used to rely almost totally on offshore lines -- generated by his French canvas coup.

"We killed the offshore guys with our flexibility in manufacturing," says Shipman, who expects to earn a 15% before-tax profit on sales of $20 million for the fiscal year ending this month. "Now we get calls from guys who never talked to us before."

4. WHAT'S MORE, YOU ACTUALLY CAN SAVE MONEY. There are even ways in which onshore manufacturing operations can be flat-out cheaper than going offshore, at least according to Fritz Beyerlein. This new development relates to the decreasing percentage of assembly cost taken up by labor.

As recently as the mid-1970s, Beyerlein notes, integrated circuit assembly was highly labor intensive; with U.S. operators making at least $3 an hour while their Filipino counterparts made as little as 15?, the differential more than made up the expense of flying fabricated circuits to the Philippines for assembly and re-export to the United States. But today, wage rates in the Philippines have jumped to about one-tenth, instead of one-twentieth, of U.S. rates, while new automated wire-bonding equipment -- which increases each worker's productivity from 125 units an hour to as many as 5,000 units an hour -- has reduced drastically the labor content in the assembly process. Given these changes, Beyerlein believes that the direct cost of onshore assembly at Cypress is less than it would be overseas. "Labor just isn't the key factor anymore," he concludes.

Equally important, onshore manufacturing allows for a more efficient use of capital. Because Cypress makes relatively expensive circuits (they average around $8 each), time-consuming offshore production would have tied up the company's capital in a very expensive way. Even though offshore assembly would save about 3.2? per chip in labor costs, that would be more than offset by combined shipping and customs costs of 2.5?, and an additional 16? in opportunity costs, which are calculated as the capital cost involved in holding inventory.

"I've been looking at this situation for years, and now -- with automation -- the numbers really add up," says Cypress chairman L. J. Sevin, who is also the company's lead venture capitalist. "Some people just look at the labor rates, but it's inventory cost that matters. It's simply cheaper to sell a part in one week than in five or six. You have to figure what you could have done with the inventory, or the money you could have made simply by pulling the interest on the dollars you have tied up in the part."


Yet to hear Jim Toreson tell it, all these benefits are less important than the long-term strategic advantage he expects to gain by controlling the means of his own production. This is where Toreson really parts company with the post-industrial theorists -- and Xebec's fascinating, risky gamble in Nevada can be seen as an alternative model of the American economic future. It is a model with two basic premises: first, that automation is the key to modern manufacturing; and second, that for a company to automate successfully, its people -- particularly those on the factory floor -- must play a central role.

As Toreson and his team planned to expand Xebec from a small-scale producer of controllers into a large-volume disk-drive maker, they rejected the offshore examples of Tandon and Seagate, seeking inspiration instead from Japanese competitors whose successes have long depended on their manufacturing muscle. Most important, however, was Xebec's analysis of the new strategic focus at IBM, the company that accounts for nearly half of Xebec's current sales. IBM has committed more than $11 billion over the past five years to land, buildings, and equipment, in large part to boost its manufacturing capacity and productivity. A company spokesperson confirms that while it continues to buy components overseas, IBM has committed itself to manufacturing onshore virtually all products sold in the United States.

"We took a hard look at what IBM was doing," explains Marcia Glow, Xebec's executive vice-president of operations. "They seemed to be taking a long-range strategy when, for example, they put $350 million into automating their typewriter plant. They realized they had to get control of their destiny by improving their manufacturing technology."

To control its destiny, Xebec had to solve two enormous problems at once. When Toreson took Xebec public two years ago, the company was making a disk-drive controller that dominated a narrow market niche. But the staff knew the time was coming when controllers, then a separate element from the disk drive, would simply be folded into a chip. Thus the first challenge was to develop a drive with an embedded, rather than a freestanding, controller before anyone else did. The second was to acquire the kind of mass-manufacturing capacity needed to survive in a viciously competitive field.

Rejecting offshore solutions to this second problem, Toreson still had to find an environment more suitable than Silicon Valley for low-cost mass production. So in 1983, he moved the bulk of Xebec's controller manufacturing operation to a plant in Gardnerville, Nev., a small town in the high desert country about an hour's drive south of Reno. To guard against possible communications problems with the new plant, he also moved Xebec's corporate headquarters to a suite of offices just outside the San Jose airport, and purchased a twin-engine Cheyenne.

Gardnerville is no Singapore, but it does have substantial cost advantages over Xebec's old production facilities in Sunnyvale. For one thing, Nevada, unlike California, has no corporate income tax. And wages run 10% to 15% below those in the Valley.

But perhaps more important, the new plant gives Xebec an almost perfect laboratory in which to apply the latest thinking about automated production. "This plant is sort of an experiment," explains Gardnerville manager Michael Rainey. "We are developing a new kind of manufacturing intelligence here."

Using some of the $20 million worth of robots Xebec is buying from IBM, Rainey and his team oversee a facility that is already heavily automated. An IBM Series 1 computer monitors the production lines and the position of each assembled board, and collects all relevant data. Robots and other automated equipment perform numerous tedious tasks, such as component insertion, wave soldering, and even some test functions.

The automation process, which continues at a gradual pace, has already improved Xebec's product quality while cutting "substantially" the per-board cost, according to Volker von Detten, Xebec's manager of advanced manufacturing, engineering, and automation. By late spring, von Detten predicts, the plant will increase its capacity by 33%, from 150,000 to 200,000 controllers a month -- without adding a single employee.

A far greater payoff, however, may come from the knowledge Xebec is developing about the automation process itself. Lessons learned at Gardnerville can be applied to the far more complex task of automating the company's recently opened disk-drive plant in nearby Carson City. And with successfully automated production, Xebec's new Owl drive can be offered with a quality level and at a price to match those of potential competitors from Japan or anywhere else. "You want to learn as much as you can about your manufacturing process before you go into producing your next technology," explains von Detten. "Here's where we're trying to learn about how to make a new product that might not be reaching its potential for two years. That's how you develop the confidence to compete."

The key to that confidence may lie with a new buzzword -- "upskilling." It implies a radically different view of how workers and new machines should interact. "Automation isn't like managing a sweatshop," Toreson says. "To succeed, we need to move assembly workers up into being knowledge workers." And he has made this belief a central part of Xebec's corporate strategy.

To best develop the automation process, explains Marcia Glow, Xebec must maintain high employee productivity and loyalty. A no-layoffs policy at Gardnerville, for instance, allays fears that robots will take people's jobs, and training programs teach workers new functions -- such as computer operation and robot maintenance -- that they will need in the automated environment.

During the past three years, Glow notes, Xebec has continually improved its sales-per-employee ratio, reaching a high of $240,000 (compared with an American industry average of $90,000 and a Japanese average of $130,000). With automation and upskilling, she believes, the company can achieve its ultimate goal: "a billion dollars in sales with only a thousand employees."

In return for its pledge to turn assembly line personnel into knowledge workers, Xebec expects employees to commit themselves to the company. That is one reason it put its production facilities in Nevada and Pennsylvania. As Rainey puts it: "We wanted an atmosphere where our people could learn the very basics of roboticization. We wanted people who would stay with us as we learned, who wouldn't go next door for another 10%."

It is far too early to proclaim Xebec's experiment a success. The company's Nevada operations -- and its automation strategy -- remain largely untested, and even if they work out, they won't immunize Xebec from the maladies now plaguing most microcomputer-related businesses. In September 1983, for instance, the company had to write off $2.7 million owed by the bankrupt Victor Technologies Inc. And according to company insiders, Xebec exacerbated its problems by purchasing expensive computer chips from brokers -- chips that have now dipped drastically in value. As a result of these and other problems, Xebec, despite a nearly threefold jump in sales last year, has been saddled with just-above break-even earnings for the year.

But the same policies seem already to be proven at Micron Technology Inc., a semiconductor manufacturer in Boise, Idaho. Like Xebec in Nevada, Micron is located in an area short of experienced semiconductor manufacturing workers. And with some manufacturing talent, such as experienced operators of assembly equipment, in short supply even in Silicon Valley, training workers for the automated factory has become a major priority for Micron's management.

"Automation is usually associated with a misleading concept that people lose their jobs. All it has meant here is that people increase their skills," explains Juan A. Benitez, Micron's vice-president of engineering. At Micron, automation is carried out piecemeal. Benitez has increased the proportion of automation by 30% over the past two years, but he always consults with workers on the assembly line and says, "We don't automate until the operators are ready and see the need."

Due in part to this policy, Benitez claims, workers at Micron -- most of whom also own company stock -- have approached the automation process with both enthusiasm and a sense of responsibility. To upgrade skills, the company makes generous use of videotapes, quality circles, and on-the-job training. Each employee is encouraged to reach for more and more responsibility. And with the company's sales growing rapidly (in the quarter ending November 30, 1984, sales were at $37.2 million, up from $8.3 million a year earlier), there is likely to be plenty of room for upward mobility even with the use of labor-saving equipment.

To ease the way for upgrading, Micron has a strict policy of not hiring managers from outside the company. Rather, management has promoted assemblers to numerous key positions, including manager of assembly and manager of fabrication production. "No matter how much equipment you have," Benitez insists, "it's the people who do it for you. The whole process is people intensive. It keeps morale up to know they can go up as we grow and automate. We're looking for commitment -- and in exchange we give a career, not just a job. After all, we expect those same people to be running Micron in 20 years."


What Benitez, Toreson, and the other made-in-U.S.A. strategists are saying is this: If American companies hope to play a significant part in the industrial future, they must make a serious commitment to a long-range manufacturing strategy. The great industrial companies of the American past -- from the steel empire of Andrew Carnegie right up to IBM -- were based on something more profound than any single technological opportunity. They were built on a vision of the future, on a commitment to be a leader not just for the next year, but for the decades to come.

Even if Toreson hasn't yet built a great American industrial company, at least he has the audacity to dream that dream -- and that may be his greatest contribution.His approach marks a break with the trend toward the sort of "disposable companies" that have become increasingly commonplace in recent years. Often started by brilliant and innovative entrepreneurs, these companies rush through a market window, use cheap offshore labor to boost production, and then, shortly after enriching their owners, fall victim to their lack of planning and foresight.

"It's hard to fight it this way instead of being a fast-growth darling of Wall Street," says Tecstor's Bleininger, "but in the long run, it's better to be on a little slower track and control what you're doing. That's why, when a lot of those offshore companies are gone, we'll still be here. You can't worry about the herd."

Perhaps as important, such entrepreneurs as Bleininger don't waste their most valuable resource -- people -- by laying them off in search of short-run profits. That approach almost always shatters the bond between the CEO and employees that seems to separate great companies from those that shine briefly and die.

"It really affected morale when they started laying off the production people," recalls one Seagate engineer. "We weren't affected directly in the engineering group, but we know a lot of loyal employees got it, people who were working 12 hours a day for the company. . . . It makes everyone wonder, 'What's going to happen next?' and produces a lot of distrust. There just doesn't seem to be any sense of long-term planning. I guess nobody really expects they are going to retire at Seagate."

The marketplace itself is not likely to get easier for those who rush offshore to solve their manufacturing problems. Donald F. Taylor, a former Tandon vice-president in charge of marketing and product development, predicts that the tightening of product cycles -- sometimes to as short as six to nine months -- will wreak havoc on companies that plan to profit solely by being first through a market window.

"When you shorten the product cycle so quickly, from two years to a few months, there won't be time to make those quick, early profits anymore," says Taylor, who is now a private consultant. "The time will come when the cycle is so short that nobody will be able to make money before the Japanese come in. Then the whole house of cards will fall apart."

But nothing will prevent this collapse unless there are dramatic changes in the way the financial community approaches the problem. For the past few years, most analysts, venture capitalists, and investment bankers seem to have accepted the idea that entrepreneurial American companies are somehow intrinsically incapable of competing here at home against Japanese and other foreign companies. While a few venture capitalists have committed themselves to domestic production, many of the companies with a strong long-term commitment to domestic manufacturing, such as Xebec, have been built without conventional venture capital and against the prevailing wisdom of Wall Street.

Indeed, this "wisdom" may not change until someone -- perhaps Toreson -- proves that it can be done. "You know, when we went out to raise our money, everyone expected we'd be driven out, that Tandon would roll over us. They told us we couldn't deliver the product at the right price," Toreson recalls, breaking into a slight grin. "Well, they didn't understand, and they still don't. You can't fight the Japanese by running from them. You have to match them by being better the way they are better. You don't build a sand castle -- you build a house of stone."


Can Jim Toreson's automation-and-people-centered manufacturing strategy be applied outside the high-technology field? Glenn Stahl, an aluminum foundryman, thinks it can. He has been using the same approach for years.

With his slight Southern accent and down-home manner, Stahl seems the antithesis of a Silicon Valley entrepreneur. And as the president of an aluminum foundry in the nation's heartland, the 72-year-old Stahl isn't exactly the sort to be courted by offshore manufacturing promoters from the Far East. Yet he, like Toreson, has learned that the best way to compete with offshore manufacturers -- which now control an estimated one-quarter of the U.S. casting market -- is to perfect the art of production at home.

"Too many foundries have stayed still. People get killed by imports that way," Stahl says. "The only way to do it is to invest in the latest machinery and have a higher level of technology than they do. That gives you a flexibility and ability to diversify they can't compete with."

At a time when capital spending at most U.S. foundries has been falling, Stahl Specialty Co. has continued to buy from $2 million to $4 million worth of new equipment every year.Today the company's facilities in Kingsville, Mo., a hamlet of 365 people an hour's drive from Kansas City, boast more than a dozen of the most modern numerically controlled machine tools, an IBM System 38 computer for payroll, and a $400,000 McDonnell Douglas CAD/CAM system.

This huge capital expenditure reflects Stahl's strong personal commitment to his company. Over nearly four decades, Stahl, who owns 57% of the company, has refused to take anything more than his salary out of the business. His net worth on paper is probably in the millions, but he still lives in a modest, $40,000 one-story house and continues to work -- "taking it easy," he says -- some 40 hours a week. "I don't have the slightest inclination to drive a Cadillac or stay in Palm Springs," he explains. "I just enjoy seeing the company grow and challenging the employees."

If growth and challenge are the ticket for Glenn Stahl, he must be a happy man indeed. Shipments of U.S. aluminum castings declined 35% from 1979 to 1983, but Stahl Specialty has enjoyed consistent profits recently, with sales jumping from $24 million in 1980 to more than $41 million in 1984, and has boosted its work force from 310 to 520 since 1981.

This remarkable performance would not have been possible, Stahl believes, without his massive investment in new equipment. The new machines -- along with dozens of sophisticated foundry processing tools that the company makes itself -- have allowed Stahl to expand his production capability dramatically. A decade ago, for instance, the vast majority of the company's sales consisted of such relatively unsophisticated items as lawn mowers and barbecue grills. But today Stahl Specialty's products range from the castings for IBM 3380 storage devices to engine parts and brake cylinders for Deere, Caterpillar Tractor, Bendix, Ford, and Delco Products.

"Quite honestly, if we had stuck with the lawn mower and barbecue business, we'd be out of business," says company controller Dan Davis. "You can't afford not to modernize and diversify. Now we can do the sophisticated stuff they can't do in Taiwan. IBM doesn't want to fool around with a casting that has to be accurate to one one-thousandth of an inch. In that, the machines make all the difference."

Beyond the simple investment in machinery, however, Stahl Specialty -- like Xebec and Micron Technology Inc. (see main text) -- has given the "upskilling" of employees a central role in the company's rapid modernization. Training gets a high priority: From the moment workers are hired, for instance, they are shown extensive videotapes on a whole range of foundry skills. Stahl also draws heavily on resources from nearby Central Missouri State University. Professors are brought in to lecture on the latest technical developments. Students receive extensive internships at Stahl Specialty, which has recruited many key managers and technical personnel from the school.

Yet despite the university tie, Stahl, like Micron vice-president of engineering Juan Benitez, insists that new production employees work their way up through the ranks. "We want to work with them as raw material and teach them the whole trade," Stahl says. "Often kids come out of school with no practical knowledge and tend to have too much concern about how things were done in the past. They sometimes don't respect the fact that sometimes people on the floor know how to do it better. We can't afford to have people around here who look down their noses at people who produce things."


Nowhere has the offshore-on-shore debate been more clearly joined than in the disk-drive industry. At first, the offshore side -- represented by such fast-growing superstars as Tandon Corp. and Seagate Technology -- appeared to be winning big. Venture capitalists, analysts, and investment bankers backed their Far Eastern manufacturing strategies, and most of these backers have yet to change their minds. E. F. Hutton vice-president Bob Grandhi, for example, hails Seagate founder Alan Shugart and Tandon founder Sirjang Lal Tandon as "fathers of this industry," who correctly realized that they couldn't compete effectively at home. "They believed that they had to knock the price down and pass on that price advantage to the customers," says Grandhi. "The only way to accomplish that was offshore."

But while offshore manufacturing helped these companies win the early battles, it may be setting them up to lose the war.

Until very recently, the belief in offshore production seemed to be borne out by the results. Aided by the low cost of labor at its facilities in India and Singapore, Tandon was able to ramp up large-scale production long before most of its competitors, and enjoyed spectacular growth, with sales skyrocketing from $6.6 million in 1979 to more than $400 million last year. Seagate enjoyed a similar takeoff, with sales rising from $9.8 million in 1981 to $343.9 million in fiscal 1984. Last fall, yet another disk-drive producer, the Long-mont, Colo.-based MiniScribe Corp., decided to shift most of its high-volume production overseas.

In announcing its move, MiniScribe cited competitive pressures as a major cause. And it is certainly true that increasing competition at home and abroad, vicious price-cutting, and an unexpected slowing in the growth of the microcomputer market have disrupted an overcrowded marketplace. Yet, had MiniScribe looked more closely at Tandon and Seagate's experience in the Far East, it might not have been so eager to follow their lead. Instead of proving a defense against Japanese competition, as its promoters planned, the offshore strategy may have caused a whole new set of problems for the disk-drive industry.

Both industry analysts and former Tandon executives, for instance, blame offshore operations for exacerbating Tandon's reputation for inferior quality. "It's tough to control quality across the nation, but it's more of a problem across an ocean," says Bill Frank, senior vice-president at InfoCorp, a market research firm specializing in the information processing industry. Tandon, he adds, "forgets it's very expensive to make a poor quality product. I have heard that, at times, 30% of the Tandon drives intended for IBM have had to be reworked before shipping. . . . If [Tandon] had overcome that quality problem, he'd have been king of the hill -- but he couldn't."

Paul Merten, who for two and a half years served as a manufacturing engineer at Tandon, elaborates on this theme. "Communications was the biggest problem," says Merten, now a manufacturing executive at Microcomputer Memories Inc., in Van Nuys, Calif. "Engineering changes took a long time in the pipeline.It's all facsimiles, telephones, and telexes. Sometimes it could take up to two months to see the changes stateside. Meanwhile 10,000 -- or 30,000 -- units have been shipped without the proper changes."

This reputation for uneven quality, along with problems relating to the introduction of new product lines, may help explain the spectacular rise in Tandon's inventories, which jumped from $128 million in June 1983, to more than $209 million a year later. David Moy, an analyst at Morgan Stanley & Co., notes that the rate at which inventory is turned over by other companies in the same industry is often twice as fast. Commenting on the inventory problem, John Levinson, who follows the company for Goldman, Sachs & Co. -- and who strongly supports its offshore strategy -- admits that "Tandon has had notorious quality problems for years," and calls offshore production "one of the big factors."

The unhealthy inventory situation may also have been partially responsible for a drop in Tandon's earnings for the fiscal year ending in September 1984. Despite an attractive increase in annual sales and a respectable increase in annual net income, the company suffered a disastrous quarter, with sales down slightly from the comparable period in 1983 and a net loss of $724,000 -- Tandon's first quarterly loss since its founding.

At the end of the quarter ending September 1984, Seagate -- the other big offshore disk-drive maker -- saw its quarterly sales volumes cut nearly in half from the previous quarter's $100-million level. Analysts and company insiders alike blame the decline on delays caused by poor coordination among the marketing, engineering, and manufacturing divisions, created in part by the company's move last year to Singapore. Explains one top Seagate engineer: "Engineering is so removed from manufacturing because the product is being made overseas. We don't even see the problems that crop up for weeks. It's not like you can go and check it out yourself on the line."

Nor did offshore production do much, in the long run, for the disk-drive company founded by Seagate's Shugart in 1973. Shugart Associates was purchased in 1977 by Xerox Corp.; in January, Xerox reported that it took an $85-million write-off and operating loss for 1984 while announcing plans to close its operations and eliminate more than 1,650 jobs. At the same time, Xerox also announced its intention to sell the Shugart line of 5 1/4-inch disk drives to Shugart's prime offshore contractor, Matsushita Communication Industrial Corp.

Ironically, the dreaded Japanese -- the oft-cited excuse for the offshore shift -- seem undeterred by their cost-cutting American rivals. Despite wage rates three or more times higher than those in such countries as Korea and Taiwan, highly automated Japanese producers seem poised to dominate such key, mass commodity areas as 5 1/4-inch floppy disk drives (sales of which are expected to double by 1985 from last year's levels). Japanese companies have increased their world market share of these drives from 28% in 1983 to an estimated 50% for 1984, with 1985 projections showing them likely to capture a mammoth 60% of total sales, more than double U.S. production.