THE NEW BUSINESS CYCLE
It used to be housing and autos. But now, high tech rules. And a stall there could stagger the economy
Six years into the economic expansion of the 1990s, and the living is easy. Almost 3 million jobs have been added over the past year alone, and consumer confidence is soaring. Company coffers are brimming with profits, the stock market is in the stratosphere, and inflation is actually falling.
Perhaps best of all, to many it seems the good times can go on for years. That old bugaboo of capitalism--the business cycle--has been tamed, according to today's conventional view. Rather than experiencing the booms and busts of old, the economy is on a steady growth path. Companies are avoiding past excesses, using computers and improved communications to manage inventories better and to boost workers' productivity. "Information technology has doubtless enhanced the stability of business operations," said Federal Reserve Chairman Alan Greenspan in his Feb. 26 testimony before Congress. His biggest worry? An overheated stock market.
But it is not the stock market or inventories or even inflation that will determine how long this expansion will last. Nor will auto sales, housing starts, or any of the traditional cyclical indicators give the first warnings of an impending recession, as they did in 1979 and 1989.
Instead, there is a new business cycle, tied to the health of the high-tech sector. Riding a wave of technological optimism, the computer, software, and communications industries have grown at a pace far exceeding the rest of the economy over the past three years, helping to extend the expansion.
But this exuberance has a price: With high tech having grown so big, the economy is now vulnerable to a high-tech slowdown in a way that was never true before. And there are already troubling signs of weakness in those industries that, if prolonged, could foreshadow a wider slump--as well as a steep decline in the stock market.
The distinctive character of the new business cycle raises the odds of policy mistakes. Like generals fighting the last war, the Fed seems to be focusing on traditional cyclical indicators, such as retail sales, new-home starts, and industrial production, that have all been upbeat recently. These are "still what drives the business cycle," says Fed Governor Laurence H. Meyer. Meanwhile, semiconductor shipments and other measures of high-tech growth hardly are mentioned by Greenspan and other policymakers, despite their economic importance. "The Fed needs to look not just at inflationary measures but also at what part of the technology cycle we are in," says G. Dan Hutcheson, president of VLSI Research Inc., a Silicon Valley research firm.
Certainly there is little doubt that high technology has replaced the traditional cyclical industries as the main driving force for growth. In the past three years, the high-tech sector has contributed 27% of the growth in gross domestic product, compared with 14% for residential housing and only 4% for the auto sector. Over the past year, a stunning 33% of GDP growth has come from information-technology industries, propelled by everything from the Internet boom to the rise of direct-broadcast satellite television (chart, page 60).
The unique nature of an expansion led by high technology explains why the U.S. has been able to sustain a lower unemployment rate with faster growth and less inflation than economists ever believed possible. Despite strong demand and rising wages for programmers, network technicians, and other high-tech workers, inflationary pressures are counteracted by constantly falling prices for such products as computers and communications equipment. Meanwhile, with the rest of the economy growing at a meager 1.8% annual rate, the demand for workers outside of high tech has not been strong enough to drive up wages. A BUSINESS WEEK analysis shows that real wages for nonsupervisory workers outside high tech have risen by just 0.3% in the past year, hardly enough to trigger inflation.
But the business cycle has not disappeared. To the contrary: High technology is more volatile than the automobile industry, with the biggest swings driven by new technologies (chart, page 64). When the new-product pipeline temporarily slows, as it did in 1985 and 1989, demand can fall sharply. But a hot new technology, such as the Internet, can send sales skyrocketing just as suddenly. "Every time we thought something about our business was less cyclical," says Andrew S. Grove, chief executive at Intel Corp., "the next cycle was bigger than the earlier one."
There's another factor that could compound the impact of the tech cycle: High tech has grown so large and important that there's a feedback loop to the rest of the economy. When times are good, fast-growing high-tech companies throw off money that fuels general prosperity, which in turn sustains the demand for high-tech products. An expanding high-tech sector spends big dollars on everything from advertising to new buildings to cleaning services. High-tech profits boost stock prices, making investors more willing to spend. And rising wages and bonuses for high-tech workers finance the purchases of new cars and homes. Indeed, high-tech jobs and industries have accounted for roughly 20% to 25% of the real wage and salary growth over the past year, according to new BUSINESS WEEK calculations, making them the key support for consumer spending.
But if and when things turn sour, look out. Information technology--now the single largest line in many corporate capital budgets--will make a tempting target for cost-cutting if the product cycle slows, or if the rest of the economy should slip because of, say, a Fed rate hike. Inventories are lean, and with 3.5 million jobs eliminated since 1989, Corporate America has squeezed most of the fat out of its workforce. Next time the economy slows, the only way tight-pressed companies can save money will be to delay nonessential info-tech projects. "Our industry used to be immune from the business cycle," says Eric A. Benhamou, CEO and chairman of 3Com Corp. "We're no longer flying under the radar. There's just too much money being spent."
The downside of the new business cycle could have dramatic consequences for employment, investment, and growth. In Silicon Valley, in Boston, and in other high-tech hotbeds across the country, a multitude of software companies are staffing up in expectation of 20% annual growth, hiring hordes of programmers, testers, and technical writers who would not be needed if high-tech sales slowed. "It's a speculative bubble," says Larry Kimbell, director of the UCLA/Anderson Business Forecasting Project, "and there will be a lot of very disappointed people" when the boom slows.
If profits drop sharply enough, most high-tech companies would have to curtail new-product development and the construction of factories. Venture-capital funds, now going begging, would dry up. As the breakneck pace of technological change slowed, buyers would have less reason to upgrade immediately to the next generation of computers or software, dampening demand even more. "If our rate of innovation slowed, and the buying momentum consequently lessened, presumably that could have repercussions on the whole economy," says Grove.
A slowdown in high tech could hold dire consequences for the stock market. From 1993 to the end of 1996, high-tech stocks have been the market leaders, pulling other sectors along with them. Over that three-year stretch, high-tech stocks have produced blistering annual returns of 35%, compared with 20% for the S&P 500. The surge in tech stocks has given many of them sky-high price-earnings multiples on the expectation of soaring future growth. An unexpected slowdown in high-tech demand could send those stocks tumbling, knocking the rest of the market from its record levels.
FALL-OFF. There are some signs that such a high-tech slowdown may already have started. For one thing, tech stocks, as measured by the Morgan Stanley High Tech Index, are down 15% since mid-January. For another, demand for high-tech equipment is weakening. Chip sales dropped by 3% from December to January and are down 14% from a year earlier, according to a Mar. 13 report from the Semiconductor Industry Assn. Government figures show consumer spending on computers rising at its slowest rate since 1992, while unfilled orders for information-technology equipment are shrinking for the first time since 1994. "We knew retail was slow, but now the corporate side--where most of the money is made--is showing signs of weakness as well," says Robinson-Humphrey Co. analyst Robert Anastasi, who follows computer sales. Adds Richard C. O'Brien, economist for Hewlett-Packard Co.: "We think we're in the late stages of a business cycle, and we're exhibiting the lower growth rates you'd expect," because of lower capital spending.
Of course, this high-tech slowdown may turn out to be a blip in a remarkable upward trajectory. Cheaper computers from Compaq Computer Corp., Packard Bell NEC Inc., and others could stimulate demand at the low end of the market. The Internet may continue to open up a range of new applications. And even an extended high-tech pause may be cushioned if global demand rebounds.
But even if high tech is the future of the economy over the long run, that does not preclude a bumpy ride along the way. In many ways, this period resembles the second half of the 19th century, also a time of massive investment in a new technology: the railroads. From 1869 to 1893, the miles of rail track quadrupled, and rail shipping costs dropped dramatically, opening up large parts of the country for manufacturing and commercial agriculture. The railroads themselves consumed much of the U.S. steel and coal production and accounted for almost 20% of all investment. Overall, the railroads' expansion fueled an economy that grew an average of 5% annually.
GROWING FORCE. But the long-term growth in this period conceals two sharp downturns, both linked to the railroads. The panic of 1873 was caused largely by railroad overexpansion, leading to a slew of railroad bankruptcies and an abrupt decline in new investment. And the mini-depression that started in 1893 was greatly aggravated when overbuilt railroads sharply curtailed the construction of new tracks.
Like the railroads in the late 19th century, high tech is the leading sector of the economy. The Information Revolution started in the mid-1970s, when the price of computing power began to plummet. But until recently, high tech simply was not big enough to influence the whole economy. From 1983 to 1993, high-tech spending in dollar terms grew no faster than the rest of the economy.
But over the past three years, the high-tech sector has skyrocketed while the rest of the economy has slowed down. High tech, as measured by BUSINESS WEEK, totaled some $420 billion in 1996 (page 68). Consumers and businesses now spend $282 billion in the U.S. on information technology hardware alone, making it larger than any of the traditionally cyclical sectors such as autos and construction. That's 17% more than U.S. purchases of new motor vehicles and parts, 49% more than spending on new homes, and 168% more than commercial and industrial construction.
The clout of the high-tech sector shows up clearly in the labor market. Help-wanted ads plead for project managers, systems analysts, and help-desk technicians. Meanwhile, companies' continuing need for assistance with their computer systems has created a bull market for the management consulting industry, which is adding jobs at a rate of more than 40,000 per year. "We're experiencing 40% growth in the information-based consulting area," says Roger Siboni, deputy chairman and chief operating officer of KPMG Peat Marwick.
All told, there are more than 9 million workers now in the high-tech sector. Each new job creates additional jobs across the economy, generating a "multiplier" effect. Historically, manufacturing was thought to have the biggest multiplier effect in the economy, which is why it played such a big role in business cycles. The expansion of an auto plant, say, would ripple through the economy, boosting hiring at steel mills and other suppliers. Since factory workers were well paid, jobs would be created at nearby stores and even construction companies to supply and house the new workers.
There's growing evidence, however, that high tech may now have a larger multiplier effect in the U.S. than traditional manufacturing industries such as autos. As U.S. manufacturers have become more efficient, an expansion in output no longer requires massive hiring at factories. And with suppliers increasingly spread out in Mexico, the Far East, and Europe, the impact of an increase in auto sales, say, is diffused globally.
By contrast, creating a new chip design or a new software program is a labor-intensive effort that relies almost exclusively on well-paid domestic workers. A study of Microsoft Corp.'s impact on the Washington state economy showed that each Microsoft job created 6.7 new jobs in the state, compared with a 3.8 multiplier for Boeing Co. The difference? Almost $800 million in stock option income for Microsoft workers in 1995, the year of the study. These dollars spill over in the form of increased purchases of such things as cars and homes. In addition, Boeing outsources a lot of the work on each plane. As a result, "Microsoft spends three to four times more per person in the local economy than Boeing does," says Richard S. Conway Jr., the Seattle-based economist who conducted the study.
High-tech jobs tend to pay well--and salaries are rising--even the jobs that aren't so glamorous. Bureau of Labor Statistics data show that weekly wages for production workers in the communications-equipment industry rose by almost 7% in the past year, while earnings for nonsupervisory workers in the software and computer services industry rose by almost 5%. Intel distributed $620 million in profit-sharing and bonus money to its more than 40,000 employees for 1996.
Despite strong growth, an expansion driven by high tech is less likely to spur price inflation. In a traditional business cycle, prices would rise and productivity growth would slow as factories hit their capacity limits. But high-tech industries, such as semiconductors and software, are different. Creating a program or a microprocessor requires a big investment. But then the cost of actually producing the chips or software for sale is relatively low. The result is a virtuous circle: Rising demand drives average costs down, making it possible to charge lower prices and boosting demand even further.
OPTIMISM. How long can this high-tech expansion continue? Technology companies are banking on a continuing stream of new and improved products to fuel growth well into the future. "With lower costs and more advanced technology, you can do applications that weren't economically possible before," says Benjamin Anixter, vice-president for external affairs at Advanced Micro Devices Inc. At the bottom, falling prices may pull in new demand. The latest low-priced computer from Compaq, for example, "is selling extremely well to customers we didn't sell to before," says Robert A. Gunst, CEO of retailers Good Guys Inc., based in Brisbane, Calif.
Most growth forecasts are upbeat. Venture-capital firm Hambrecht & Quist predicts that the number of PCs sold in North America will rise by 17% in 1997. Looking further out, Gartner Group Inc. projects that spending on hardware and software will rise by about 8% annually for the rest of the decade.
Yet the history of the information-technology industries is anything but smooth. The biggest high-tech downturn in the past two decades came in 1985, when demand petered out just before the introduction of Intel's 80386 microprocessor. And in the first half of 1990, another product transition helped drive down business spending on computers even before the start of the recession of 1990-91.
Now, the industry may be facing another pause. With orders falling off, there is no obvious "must-have" product on the horizon to stimulate demand, as the Internet did. Intel's much-ballyhooed Pentium MMX chip is turning out to offer only a minimal improvement in speed for existing software. "It hasn't set the corporate world on fire," says Anastasi. "Usually, new products create a lot of excitement, but that hasn't happened yet." High-definition TV, which will eventually create enormous sales, will not hit the airwaves until 1999, while the low-orbit satellite systems that promise a wealth of new applications are nowhere near ready.
DANGER. The dominant role of high-tech spending gives a bigger meaning to any drop-off in demand. An economy is vulnerable to a downturn when it develops "imbalances," to use Greenspan's term--that is, when one sector or industry expands too fast. For example, consumer spending and office construction far outpaced the rest of the economy during the 1980s. So when the recession came in 1990, these sectors crashed, making the downturn much worse.
Now, high tech faces the same danger. Over the past three years, business spending on information-technology gear has risen by almost 45%. Meanwhile, spending on labor has risen by only 19%. Even if information technology has become a strategic asset for many businesses, its very size makes it a logical place to cut or postpone when times get tough. "After all," says Paul Saffo of the Institute for the Future, "you can only defy gravity if you're small."
Much of the planned increases in information spending can be deferred if the economy softens. Consider KPMG Peat Marwick. It is planning to spend $250 million over the next three years on information technology, more than double its historic spending pattern. Out of that, there's little flexibility in $150 million, but the remaining $100 million is more variable, says Siboni. "Building Web-centric information systems might become less critical."
The changing nature of corporate info-tech spending at companies will make it easier to make quick cuts. Rather than hiring a large internal staff of programmers, companies are relying on temporary workers and outside consulting firms, which can be easily sloughed off during tough times. At temporary-worker giant Manpower Inc., the division that supplies technology workers is growing at 40% per year.
Small businesses will not be reluctant to hold off spending if the economy turns down. "I have a lot of control over what I buy," says Josh Moritz, CEO of DMTG Inc., a small New York-based advertising firm. "I can delay purchases or buy used machines."
Consumers have even more reason to cut back during recessions. Over the past three years, the amount of money spent on home computers has risen by 55%, far in excess of income growth. Most consumers do not need a new computer. Instead, like autos, they buy the newest model if they can afford it. "A dominant force in the PC world is becoming consumer markets and hype," says William C. Rosser, a Gartner Group analyst. "That means you're more likely to see cycles like other consumer industries."
If and when the cutbacks come, they will not be confined to computers and software. Instead, the downturn is likely to spread to the telephone and networking industries as well. When the phone system was used mainly for voice calls, demand for lines was driven by population growth. But more and more, investment by phone companies is driven by the need for additional phone lines for computers and more capacity for data connections rather than basic voice service. The sudden prominence of the Internet, for example, forced Ameritech Corp., the Chicago-based telephone company, to boost its capital spending by $100 million to $200 million in 1996, to keep up with the need for more lines.
"PAYBACK." To be sure, there are forces that could prevent a high-tech downturn from happening anytime soon. Intel and Microsoft, the giants of personal computing, are investing hundreds of millions to create new products and to help other companies do so. "If the cycle is pointing down or is softer, then the only thing we can do to cope with that is to pump up our products," says Grove.
High-tech executives are also counting on competitive pressure forcing corporations to continue spending on information technology even in a recession. "The payback customers get from our products is very clear, very justifiable," says Mark Eppley, chairman of Traveling Software Inc., a Bothell (Wash.) maker of software for remote access. "Our business is growing very fast--independently of the economy."
A decline in U.S. sales could be cushioned by growth in overseas sales. Telecom deregulation in Europe may open the floodgates for cutting-edge computers and communications gear, where U.S. companies excel. Developing countries represent a huge untapped market that will become increasingly important. "There's a growing middle class in India, and it's as large as all of France," says Donald Macleod, chief financial officer at National Semiconductor Corp. "That's where the long-term opportunity lies."
But relying on global demand is no panacea. North America still makes up almost half the global market for computers, software, and peripherals. The latest data show chip sales falling faster in Asia than in the U.S. And growth in Japan and South Korea is expected to slow in 1997, according to DRI/McGraw-Hill forecasts, with Europe's economies still sluggish. In areas such as Japan, Germany, and France, "our business is still in most people's terms reasonably good," says John T. Chambers, CEO of Cisco Systems Inc. "But it's not good in Cisco terms."
Closer to home, the growing importance of high-tech poses new problems for economic policymakers such as the Fed. The tools of monetary policy--interest rates--do not work very well on high tech. In particular, it matters more whether compelling new products and technologies are available. When high-tech industries are riding a technology wave, they can power their way through slowdowns in the macroeconomy, no matter what the Fed does. "Whether the interest rate is 7% or 9% doesn't matter when you're growing at 30% per year," says AMD's Anixter.
So far in this expansion, the country has been lucky. The surge in high tech in 1995 and 1996 came just when autos and housing turned sluggish. But now the rest of the economy is turning up, judging from recent data. As a result, forecasters are raising their projections of growth in the first quarter to 3% or more, leading to speculation that the Fed will soon boost interest rates.
But ignoring the health of the information technology industries can produce bad policy. With order growth for computers and semiconductors having slowed noticeably, the economy may be far weaker than the traditional indicators make it seem. This raises the possibility that a rate hike by the Fed could be a serious mistake at this point.
Is a high-tech-led recession inevitable? Hardly. But for the first time, it's possible, and the odds are rising. The new business cycle is alive and well in Silicon Valley.By MICHAEL J. MANDEL With Andy Reinhardt in San Francisco and bureau reportsReturn to top