They Might Be Giants


10 min read

This is part of IEEE Spectrum's special R&D report: They Might Be Giants: Seeds of a Tech Turnaround.

Free market economies cycle up and down. The bulls and bears come and go. But high-tech companies that plan to be around for a while must invest steadily in research and development, or risk being swept away in the next wave of innovation. That wave is rolling in partly on the strength of the five blockbuster technologies explored in this special report, each of which IEEE Spectrum expects will give birth to lucrative new markets or accelerate the growth of old ones within the next five years.

"It's a life-and-death struggle in the high-tech industry," said iconoclastic economist and perennial Nobel Prize contender William J. Baumol in a recent interview with IEEE Spectrum. To survive, companies must spend ever-larger sums of money developing materials, devices, systems, methods, and prototypes in light of the knowledge gained through basic research primarily funded by governments and, increasingly, by firms themselves.

In other words: innovate or capitulate. That Hobson's choice has not only driven the world's Top 100 R&D Spenders to pump more than US $215 billion into research and development departments last year—it is the engine driving the growth miracle of capitalism, as Baumol calls it in the most recent of his 40 books, The Free-Market Innovation Machine.

According to Spectrum 's first annual R&D survey [see table, PDF], the world's Top 100 Spenders increased R&D expenditures by an average of 5.25 percent in 2001, despite massive turmoil in the telecommunications, computer, and semiconductor sectors. Compared to the year before, 43 of the top 100 increased R&D spending by an average of almost 22 percent, while 53 cut spending an average of 8 percent (and of those, 12 trimmed R&D outlays by less than 2 percent). Predictably, most of the big drops were in the telecom sector, including beleaguered Lucent Technologies [ranked (17) among the top 100] and Nortel Networks (19), which slashed R&D outlays by 30 and 35 percent, respectively [see graph "R&D Bulls and Bears,"below].

R&D Bulls and Bears The 10 companies in the Top 100 R&D Spenders that increased spending the most last year versus the 10 that cut the most.

But even in depressed sectors, companies went on investing, with some posting surprisingly aggressive increases. JDS Uniphase (91) bucked the R&D slide in the telecommunications equipment sector by boosting its R&D spending over 50 percent. Chipmakers LSI Logic (96), Sun Microsystems (42), and IBM (5) all raised their R&D spending by more than 20 percent over the prior year. Several companies tied to the corporate networking and database markets also increased spending substantially, including EMC (77), which upped R&D spending 18.6 percent; Cisco Systems (8), up 17.2 percent; Microsoft (12), up 16 percent; and Oracle (66), up 12.7 percent. (Note: figures provided by Standard and Poor's are historical and do not included restated data.)

The continuous rise in corporate R&D spending through boom and bust cycles reflects a novel observation of Baumol's, which he describes in his book as a ratchet effect. He notes that R&D spending can be expected to surge from time to time, as it did during the mid-1990s, but once the new level is reached, "the ratchet—enforced by the competitive market—prevents a retreat to the previous lower level." Amid a serious economic downturn, the ratchet might slip a bit, but the mechanism nonetheless "accounts for the extraordinary growth record of free-enterprise economies and differentiates them from all other known economic forms, with ever-growing R&D expenditure norms leading to ever more rapid growth."

Competition not only forces companies to "run as fast as they can in the innovation race just to keep up with the others," it also compels them to integrate their research, production, marketing, and sales units into a single innovation machine, says Baumol. And when marketing, sales, product development, and research departments join in setting and executing the R&D agenda, market potential, as opposed to scientific curiosity, is the guiding light.

Five to thrive

For this report, Spectrum picked five electrotechnologies favored by some of the world's biggest R&D spenders—favorites they hope will pay off big within the next five years. An exploration of each of these new technologies reveals that the same competitive pressures that foster what Baumol calls an R&D "arms race" among firms also drive them into each others' arms, resulting in joint research projects, licensing arrangements, and standards-setting efforts that help disseminate new technologies.

How and when these alliances arise is often a function of the nature of the innovations being worked on—whether they are what economists call complementary or more radical, substitute innovations. If companies are striving to improve an existing category of computer system, for instance, they might soon decide to collaborate on standards to facilitate software interoperability across enterprise networks. Conversely, if companies are working on a basic departure—a substitute technology—that will change the way people light their homes, say, rivals will race each other to the patent office before forming alliances to advance the technology.

The Semantic Web effort is a textbook example of a joint attack on fundamental impediments to the further development of an existing technology. Too often, the billions of pages on today's Web frustrate searches or yield only inconclusive results. Better search mechanisms and personal agents that could automatically perform searches for you would be a boon even to successful e-commerce sectors like travel—currently a $20 billion-plus category that could exceed $63 billion by 2006, according to Jupiter Media Metrix (New York City). Described in "Weaving A Web of Ideas", Semantic Web research is spearheaded by the World Wide Web Consortium and its 500 university and corporate members (35 of which are also to be found in the Top 100 R&D Spenders).

In portable electronics and flat-panel displays, upstart organic light-emitting diode (OLED) displays soon will give liquid-crystal displays (LCDs) a run for their money. As discussed in "Just One Word—Plastics", over 50 companies from a variety of industries have sunk resources into OLED R&D, trying to get in on the ground floor of a market seen as growing from $84 million today to $1.6 billion in 2007. Cross-industry R&D alliances are sprouting to meet a vast array of technical challenges. Giants in electronics research such as Motorola (13), Xerox (73), and Lucent are collaborating with chemical powerhouses E.I. DuPont de Nemours (54), Dow Chemical (65), Bayer AG (40), and even paper companies such as International Paper.

Enabling a computer system to diagnose and optimize its performance and allocate its compute and storage resources automatically should increase efficiency and boost an already mammoth market, as discussed in "Helping Computers Help Themselves". According to Gartner Dataquest (Stamford, Conn.), the revenues worldwide from information technology (IT) services, which cover hardware and software support and professional services, will reach $557 billion in 2002 and rise to $696 billion by 2005.

Yet despite this rosy market outlook, IT is facing a crisis: barring a revolution in the way computers are maintained, upward of 200 million IT workers will be needed in the next decade to run the world's computer systems. No wonder Big Blue is pushing for open standards and marshalling its research organization and marketing teams to generate a host of new labor- and cost-saving products. But this 800-lb gorilla is on everyone's R&D radar screen, and IBM's competitors, among them Hewlett-Packard (30), Sun, and Microsoft, all have similar initiatives, a follow-the-leader strategy that Baumol says is exactly how competitors wind up continuously upping the R&D ante.

If nurturing innovations complementary to existing technologies spurs companies to collaborate, the promise of raising a new market by overthrowing an entrenched technology can tempt firms to go it alone, at least at the initial stages. As described in "Let There Be Light" gallium nitride is the semiconductor basic to white-light LEDs. These are poised to claim a large chunk of the $12 billion-a-year market for sources of white light, now the sole property of makers of incandescent light bulbs and fluorescent tubes.

In the emerging market, the leader, Nichia Corp., is taking on the leading lights of conventional lighting, including General Electric Co. (43), Osram-Sylvania, and Philips Lighting. But these "old-timers" are not standing pat. Each has engaged in strategic joint ventures or alliances in solid-state lighting: Philips Lighting with Agilent Technologies (58) to form Lumileds Lighting LLC; GE with Emcore Corp., to form GELcore; Germany's Osram Opto Semiconductors working with Osram Sylvania in the United States; and Tokyo's Toshiba Corp. (31) working with Toyoda Gosei Co.

So far, few industrial collaborations have coalesced around speech recognition, the fifth technology that Spectrum has chosen to highlight in "Talk to the Machine". Replacing keyboards, pushbuttons, and knobs with speech recognition could transform human-machine interaction. You could dictate memos into your PDA instead of scratching illegible notes with a clumsy stylus, or adjust the air conditioning in your car without taking your eyes off the road.

But even after 50 years of basic research, much of it conducted in government and university laboratories, this is just now a market in the making. Consequently, companies trying to commercialize the technology guard their intellectual property jealousy, and rarely choose to collaborate. The global market for embedded speech applications like those described is projected to grow from a measly $8.4 million in 2002 to $277 million in 2006, according to market research firm the Kelsey Group (Princeton, N.J.). The prospect has lured big fish and little, from IBM and Royal Philips Electronics (24) to Voice Signal Technologies Inc. (Woburn, Mass.) and Sensory Inc. (Santa Clara, Calif.).

Where ideas are born

New technologies such as those described in this report develop out of advances in basic research. It is this process of"creative destruction," as Austrian economist Joseph Schumpeter called it, that catalyzes growth in capitalist economies.

But should we buyers of the end products of basic research—be it the slickest OLED-equipped PDA or a missile defense system—also fund the research with our taxes and essentially pay twice?

The short answer, according to Baumol, is yes. Government-funded basic breakthroughs are the foundation for growth. The building blocks are the incremental improvements made to basic advances by companies, which, compounded over time, outdo the initial discovery in significance. So it was with the transistor, the Internet, and, as Baumol points out, ENIAC, a government- and university-funded project that resulted in the first general-purpose electronic computer in 1945.

"You need the individual inventor, the government, or university lab for the breakthrough," says Baumol, referring to the Electronic Numerical Integrator and Computer (ENIAC). "Without the breakthrough, the corporations would have nothing to improve. But without the improvement, we would have, as was originally expected, maybe 200 computers in the U.S., each of them now costing $20 million."

Firms feast off the fruits of basic research and spend most of their own R&D money on development (71 percent or $128 billion of the $181 billion spent by U.S. industry on R&D in 2000) because they have a huge disincentive to fund much of it themselves. Economists even have a name for it: spillover effects. The benefits of basic research tend to trickle down to businesses that didn't fund it, as well as to society as a whole. The ultimate applications, markets, and beneficiaries of basic research are unpredictable, so society can't expect corporations to pay for it, Baumol contends.

But companies will spend the money if they have no choice. The numbers confirm that they will pick up a bigger share of the basic research tab where government spending (with the exception of the biomedical sector) has lagged industry's need for it. In the United States, the federal government has always footed the bill for most of the basic research performed in the country and in the year 2000 still funded almost 50 percent of it. But that's a far cry from the 70.5 percent of basic research funds shelled out by the federal government in 1980. According to the latest available figures, taken from the National Science Board's Science and Engineering Indicators 2002, "this decline in the Federal share of basic research support does not reflect a decline in the actual amount of Federal support, which, in fact, grew 3.5 percent per year in real terms between 1980 and 2000. Rather," the report continues, "it reflects a growing tendency for the funding of basic research to come from other sectors. From 1980 to 2000, industry's self-reported support for basic research grew at the rate of 10 percent per year in real terms."

The story is much the same in other countries in the Organization for Economic Cooperation anbd Development (OECD), which includes the Group of Seven (G-7). They have all seen a fairly steady rise in total R&D expenditures over the past two decades. Even though governments have provided a declining share over the last decade in almost all OECD countries, industry has accelerated its expenditures to make up for the shortfalls and then some [see graph].

graph, R&D expenditures 1985-99

R&D Expenditures 1985–1999 R&D spending rose steadily in the world's most developed countries during the 1980s and 1990s. Barring an economic disaster, that trend should continue, according to William J. Baumol's theory of innovation.

Spreading the wealth

Not coincidentally, OECD countries have also enjoyed the fastest and most sustained economic growth in the world since World War II. All the Top 100 R&D Spenders are from members of the OECD, with 45 hailing from the United States, 22 from Japan, and 11 from Germany.

Taking a broader view, it's clear that the R&D wealth is spreading farther, particularly in Asia, as countries like China, Taiwan, Singapore, and South Korea recognize the vital part R&D plays in promoting prosperity. Beyond their hefty direct investments in research, they've also been luring multinationals to their shores with enticements ranging from liberal tax breaks to government-supplied research labs.

Taiwan is a truly a tiger when it comes to creative government incentives. Witness its offer to exempt multinational companies from income taxes for five years if they set up shop next to the Industrial Technology Research Institute in Hsinchu. The kicker: continuation of the preferential tax treatment even after the first five years, including the tax-free export of goods manufactured there.

Since 1993, at least 67 non-Taiwanese companies have inked strategic agreements with Taiwan's government. One of the first to set up there was Motorola. Speaking at a recent roundtable on corporate R&D, the company's chief technology officer, Dennis Roberson, said the dearth of top researchers in the United States heavily influenced Motorola's decision to move more of its operations offshore.

The large supply of trained scientists and engineers in mainland China is certainly a draw to companies. For example, the northern port city of Tiajin boasts 21 institutes of higher education and 750 000 scientists and engineers, according to the government. Since 1991, 550 outside companies in the computer hardware and software, advanced materials, and biotechnology sectors have settled in the TEDA (Tiajin Economic-Technological Industrial) High-Tech Industrial Park. Incentives offered to non-Chinese firms include exemption from income and sales taxes; free water, sewage, and gas transmission; and discounted electricity. In addition, the government will waive its charge for the land if an approved company stays on for at least three years.

By creating incentives for R&D investment and providing market conditions favorable to business and the dissemination of technology, Taiwan and China are hitching a ride on the magic beanstalk of capitalist growth. The two have enjoyed remarkable increases in gross domestic product (GDP) over the last decade, with Taiwan boosting its GDP by an average of 6.3 percent annually between 1990 and 2000 and China by an astounding 9.6 percent annually over the same period, according to World Bank figures.

These success stories are still exceptions to the rule. While growth seems to be spreading slowly in the wake of R&D investment, much of Africa, Latin America, and the less prosperous parts of Asia have been left far behind.

"In my view, that is because [countries in these regions] haven't learned the lesson that China has learned and that Russia is learning," laments Baumol, "that only with that incentive system can you expect this incredible growth rate that has never been approached before in human history."

With research and insights provided by IEEE Spectrum researcher Alan Gardner and reporting from David Kramer, editor of Frost and Sullivan's Science and Government Report.

To Probe Further

New York University economics professor William J. Baumol outlines a novel theory of industrial innovation in his latest book, The Free-Market Innovation Machine: Analyzing the Growth Miracle of Capitalism (Princeton University Press, 2002).

Statistics cited in the article and table are obtained for the most part from the U.S. National Science Foundation's Science and Engineering Indicators 2002. See the foundation's Web site at

This article is for IEEE members only. Join IEEE to access our full archive.

Join the world’s largest professional organization devoted to engineering and applied sciences and get access to all of Spectrum’s articles, podcasts, and special reports. Learn more →

If you're already an IEEE member, please sign in to continue reading.

Membership includes:

  • Get unlimited access to IEEE Spectrum content
  • Follow your favorite topics to create a personalized feed of IEEE Spectrum content
  • Save Spectrum articles to read later
  • Network with other technology professionals
  • Establish a professional profile
  • Create a group to share and collaborate on projects
  • Discover IEEE events and activities
  • Join and participate in discussions