Rerouting the Pork Barrel
Forbes ASAP , May 30, 1999
Silicon Valley's newly politicized executives stopped playing defense this year and went looking for government subsidies. But if high tech companies get their way in Washington, they may exacerbate the industry's biggest problem -- the lack of talented people to fill technical jobs.
Doubling federal research spending for basic research is a top goal of the Technology Network, the high-powered lobbying group better known as TechNet. Federal R&D spending has declined sharply since the end of the Cold War, when between 50% and 70% of it went for Pentagon projects.
TechNet is eyeing the anticipated budget surplus to boost R&D funding.
This time around, however, the goal won't be national security but industrial policy. "Research is the fuel that fires technology, and powers the engine of the New Economy," says Eric Benhamou, chairman and CEO of 3Com, in a TechNet press release.
Leave aside the issue of whether Benhamou and company can predict the proper composition of the "new economy," whose 21st-century form may have more to do with entertainment and aesthetics than electronics. Even from TechNet's own myopic viewpoint, a general hike in federal R&D spending looks like a bad idea. Economic relationships aren't as simple as they first appear.
It's true, of course, that if you subsidize something you'll get more of it. The question in this case is what "it" is. Research isn't a product but a process with several components. What do R&D subsidies produce more of? Austan Goolsbee, an economist at the University of Chicago Business School, has taken a careful look at that question. Analyzing the effects of federal R&D spending from 1968 to 1994, he found that most of the money went not to more inventive activity (an increase in "quantity") but to higher salaries for scientists and engineers.
Technical training takes a lot of time and specialized effort.
Journalists and lawyers can't easily switch to electrical engineering just because the money looks good. As the economists say, the supply of technologists is "inelastic." In the short run, it changes little in response to higher pay.
So to attract people to subsidized R&D projects, the subsidies must go to bid up wages. In the old days, that meant higher pay for people working on defense projects: Goolsbee finds a major impact on aeronautical, mechanical, metallurgical, and electrical engineers.
Mining, civil, industrial, and chemical engineers, whose work had little to do with defense, didn't enjoy the pay hikes. Under the TechNet plan, which calls for more money for universities, professors and their research teams would reap the higher wages.
Regardless of who gets the government money directly, the result is greater competition for employees and, thus, higher salaries throughout the field. That's bad for anyone who doesn't have Uncle Sam footing the bill. "Government R&D directly crowds out private inventive activity," writes Goolsbee. "By increasing the wages of electrical engineers, for example, government R&D spending increases R&D costs for computer manufacturers who do not receive federal funding but are forced to pay their scientists higher wages." This distortion tends to decrease the amount of R&D that unsubsidized companies can afford to do.
Goolsbee acknowledges that R&D subsidies can encourage more people to pursue scientific careers, with eventual benefits to the innovative process. But, he suggests, increasing everyone's salary is an awfully roundabout way to accomplish that goal.
That R&D subsidies might make life harder for high tech companies is counterintuitive. But the pattern is well established: A similar effect holds true for farmland, which goes up in price to reflect agricultural subsidies; college tuition, which tends to rise with federal student aid; and commercial and residential real estate, whose prices fluctuate with the tax code. And the law of unintended consequences applies even in Silicon Valley.