The Dismal Science

The IT Split

Why Japan’s tech industry bombed while America’s boomed.

One of the most remarkable developments of the last 10 years was the dual reversal of fortune by the world’s two largest economies—the simultaneous economic ascendance of the United States and decline of Japan. Fiscal and monetary policies certainly contributed in both instances. However, new research shows that microeconomic factors, especially the different organization and strategies of the two countries’ information technology sectors, may have been more important.

Information technology may matter less than “New Economy” promoters once claimed, but the strength and length of the 1990s U.S. boom clearly owed much to IT. According to first-rate analysis by the Commerce Department, IT companies, while constituting less than 8 percent of the GDP, accounted for nearly 30 percent of U.S. economic growth from 1995 to 2000. Harvard’s Dale Jorgenson and the Fed’s Kevin Stiroh trace more than three-fourths of increased U.S. productivity growth to IT advances and strong business investment to deploy those advances.

Throughout the ‘80s, many Americans envied Japan its technological prowess, but there is a strong case that Japan’s IT weakness magnified its slow growth and weak productivity gains in the ‘90s. One of Japan’s top economists, Rizaburo Nezu of the Fujitsu Research Institute, has recently documented the various and complex ways in which, less than a generation after Japanese firms pioneered the miniaturization and marketing of electronics, Japan has “disappeared from world IT competition.”

Today, the world’s 10 largest IT companies include six American firms but only one Japanese company, the mobile phone and ISP giant NTT DoCoMo. Japan’s domestic macroeconomic troubles cannot explain this. IT companies draw on global markets and technologies. For the last decade, companies could borrow money in Tokyo to develop or buy IT technologies more cheaply than in New York. And since 1990, Japanese firms have invested in R & D and in IT hardware at higher rates than their U.S. counterparts.

Yet the ‘90s saw Japanese IT fail spectacularly, market segment by market segment. The main culprits were the flawed strategies and organization of Japanese IT companies. The semiconductor industry is, in many ways, the very heart of IT development. Twelve years ago, U.S. and Japanese firms each controlled more than 40 percent of the world market. Over the next decade, their corporate strategies diverged; by 2000, American chip makers accounted for 55 percent of global semiconductor sales, and Japanese companies were left with just 25 percent. Leading chip makers in Japan (and Europe) tried to produce all kinds of chips. This strategy required them to diffuse their R & D resources. As they fell behind technologically, their market share eroded. Their counterparts in America (and Korea) concentrated on a few, select classes of chips and vastly expanded their market share.

American firms today control two-thirds of the world PC market and 80 percent of global server sales. In each case, Japanese producers account for less than 5 percent. Again, contrasting production strategies may explain the Japanese’s also-ran status. The American companies that have come to dominate PC and server sales—Dell, IBM, Sun Microsystems, Compaq, and HP—concentrate on procurement, logistics, and after-sales service and leave most of the computer design and production to firms like Taiwan’s Quanta and Acer. Japan’s once-great computer makers—Fujitsu, Toshiba, and Sony—have persisted in doing almost everything at home and for themselves, an increasingly improvident strategy in an era of globalization.

Flawed strategy also helps explain the failures of Japanese software firms, not one of which ranks today among the world’s 100 top software companies. In contrast to software makers in the United States (as well as in software centers like Israel, India, and Ireland), most Japanese software companies are divisions of large computer makers and have traditionally focused on developing tailor-made programs, especially for the parent company’s products. But tailor-made products have come to occupy only a small niche of the world software market, which finds packaged products more cost-effective. The vertical integration of software and hardware in Japan also makes it very hard for software startups to establish themselves. As a result, Japanese programming lags behind the rest of the world.

The organization of Japanese business may undercut Japan’s ability to succeed in a period of rapid technological innovation. Japan spends more of its GDP on R & D than we do, almost all the spending by very large companies. When a successful innovator like Sony develops a new technology, the advance is applied to the company’s products. But generally that’s where it stays. In America, where it’s much easier to start new businesses, innovation as often comes from university labs, startups, or small firms. American innovators typically license their advances to others, who can then adapt them to a wide variety of products and services. An innovation that produces one new killer product in Japan may, in the United States, give birth to a whole new market.

The success of American IT cannot be separated from the willingness of companies across the economy to reorganize their operations and take advantage of these technologies. The major contributions to recent U.S. productivity gains have come not only from the semiconductor and computer industries, but also from wholesale trade, retail trade, and financial services—sectors that invested heavily in IT.

But almost everything about Japanese businesses discourages taking advantage of innovation. The fabled reluctance of Japanese firms to fire anyone makes it hard to substitute IT for workers. Service companies and domestic manufacturers still enjoy extensive regulatory protections that leave little room for new businesses to form and exert competitive pressure. As a result, most Japanese companies resist the strategies that work in the United States and elsewhere in Asia, such as more foreign outsourcing. Apart from the major export industries, the average Japanese worker is nearly 40 percent less productive than the average American worker.

The American IT industry is certainly not immune from strategic errors: Witness the meltdowns in the Internet and telecom sectors. And our current macroeconomic policies are much less hospitable than five years ago to strong investment and growth. But market pressures here will continue to force American businesses to adapt, making it very unlikely that the United States and Japan will trade places again in the next 10 years.