There’s a Lot More R&D; Going on Than Figures Indicate

Like Chicken Littles of global competitiveness, technology pundits are running around the country shrieking, “The R&D; budget is falling! The R&D; budget is falling!” Japan now spends more of its gross national product on commercial research and development than we do, these gurus whine, and that means America is losing its innovation edge.

While it’s true that the R&D; numbers have long served as a surrogate index for innovation, that only confirms Mark Twain’s nasty little epigram that there are lies, damned lies and statistics. America’s R&D; stats hit a very large nail not quite on the head.

Yes, industrial giants such as General Motors, IBM, DuPont and Merck spend billions of dollars annually on R&D; (with wildly varying results). But what about our post-industrial giants such as Citicorp, Federal Express, Goldman Sachs and American Express? What does research and development mean when you’re in a service industry?

- Federal Express is “a very R&D-intensive; company,” says Ed Clark, who handles investor relations for the innovative shipping firm whose annual revenue is approaching $6 billion. “We are so oriented towards developing state-of-the-art technological solutions, that maybe we should consider capturing that expense.” In other words, Federal Express has no R&D; budget.

- Citicorp, America’s biggest bank, spends hundreds of millions of dollars each year developing and modifying its state-of-the-art telecommunications networks and computer software to provide a host of innovative services for its institutional and retail customers. The bank has consistently been a pioneer in the application of technology to financial services.

- Dow Jones & Co., publisher of the Wall Street Journal and a top provider of data services, spent more than $5 million to buy two Connection Machines--state-of-the-art supercomputers--as platforms for a next generation of on-line information.

- Goldman, Sachs & Co. hires physicists and mathematicians to dream up new financial instruments and build the software to implement them.

- American Express and its Shearson Lehman Hutton unit have pioneered voice recognition technology on the trading room floor.

None of these market leaders have R&D; budgets. In post-industrial fact, there are numerous examples of multibillion-dollar service sectors--ranging from publishing to movies to restaurants--where research and development are definitely taking place but in forms invisible to the statisticians.

Maybe a McDonald’s McCaviar wouldn’t be what most electrical engineers would call a technological innovation, but it would certainly be a competitive innovation in the right marketplace. The challenge is measuring it in a meaningful way.

This new reality didn’t completely escape the National Science Foundation and the U.S. Census Bureau, the government agencies that track R&D; spending. “We clearly weren’t doing a good job” of tracking service sector spending, acknowledges Melissa Pollak, a science resources analyst at the NSF. “So, we increased the coverage of that group. It caused a flurry of phone calls from those companies to the Census Bureau last year. They said, ‘Why are you surveying us? We’re not industrial companies.’ ”

Of the 470 non-manufacturing companies surveyed, only 14 had R&D; data. “Basically, their response was, ‘We don’t do R&D;,’ ” Pollak says.

Now this may be accurate, but it’s also meaningless. Of course more than 3% of these companies do R&D;, it’s just that it may not be research and development in the way the NSF traditionally thinks of research and development. Clearly, crafting new software, whether it is an expert system or a process simulation algorithm, involves both research and development. It also costs money and manpower. But where do you draw the line? What qualifies as innovation? New software? New packaging? New designs? New networks?

These numbers conceal more than they reveal. The latest NSF figures say that there was $64.9 billion of research and development spending in the U.S. in 1987. The four largest companies accounted for 19% of all R&D; spending, and the top 25 companies accounted for 40% of the total. Yes, the top 25 are virtually all industrial companies.

NSF’s Pollak notes that where non-manufacturing companies once accounted for 3.4% of R&D; spending in 1981, they now account for 9%. But “most of that’s reported by the emerging software companies,” such as Lotus Development and Microsoft. The software companies, like their industrial cousins, do have formal R&D; budgets. That makes them easy to measure.

The point isn’t to pick on the NSF and Census Bureau for taking the easy way out. On the contrary, the NSF approaches the challenge of measuring R&D; spending from a completely understandable science-and-technology-based perspective. It makes sense for people who still view global markets in the context of technologically flavored industrial goods.

And yet, while these R&D; numbers may work for traditional science and technology investments, do they adequately capture what’s going on in some of the most critical segments of America’s economy? If we’re really concerned about research and development, if we really believe that innovation is key to global competitiveness, we have to redefine what those words mean. We have to recalibrate the tools we use to measure them.

I don’t want to diminish the importance of the more traditional industrial R&D; in any way. But it’s time that policy-makers understand that innovation can be made of more than microchips and restriction enzymes. It’s also time for people who work in service sector companies to recognize that, whether their budgets formally measure it or not, they also do research and development. Managing that R&D; well will be key to their success in global markets.