Apple Research: An Immodest Proposal

on September 20, 2013

A recent column by Brian S. Hall and a followup conversation in the comments got me thinking about one of his “wild and crazy” ideas for Apple: That the company should fund a research lab in the tradition of Bell Labs. And the more I think about it, the better an idea it seems. I don’t believe for a minute that Apple would do this; it just doesn’t seem to be their style. But that is a big loss for the company and the world of innovation.

Apple is sitting on well over $100 billion in cash that it has no real ideas about using. It’s giving some back to shareholders a dividends, is using some to buy back its own stock, and is letting the rest sit, doing no one much good. Using, say, $10 billion to endow Apple Research would be a gift of which Apple itself might ultimately be the greatest beneficiary.

The great industrial research labs arose out of special circumstances that made certain companies very rich. Bell Labs was a product of AT&T’s government-sanctioned monopoly. (Anyone who believes that lack of competition is necessarily the enemy of innovation should study the history of research at AT&T from the 1920s through the 60s.) IBM built IBM Research into a powerhouse during a period when it had an effective monopoly in computers. Xerox build its Palo Alto Research Center at a time when its dominance of the office copier business brought in big profits. Apple is nowhere close to being a monopoly, but it does enjoy monopoly-like profit margins.[pullquote]The great industrial research labs arose out of special circumstances that made certain companies very rich. [/pullquote]

Bell Labs in its heyday may be the best model. I have been looking into the history of mathematical research there for a project and have come to an interesting realization. Most of the esoteric math research at Bell Labs was actually closely tied to AT&Ts business needs. AT&T hired Claude E. Shannon, one of its most important scientists, because of his MIT master’s thesis, which created a mathematical model for a telephone switching system (and later became a model for digital computers.) His work on the mathematical theory of communications was very relevant to AT&T’s business as was mathematician Richard Hamming’s work on error-detecting and error-correcting codes.

Bell Labs researchers occasionally wandered far afield. But the Nobel Prize-winning work of physicists Arno Penzias and Robert Wilson on cosmic background radiation was the product of AT&T’s version of Google’s 20% time. Penzias and Wilson were allowed to experiment with an huge unused microwave antenna (photo above) that had been built for an early satellite communications project. ((For an excellent account of the experiment, see Jeremy Bernstein’s Three Degrees Above Zero.  And for a general history of Bell Labs, see Jon Gertner’s The Idea Factory.))

IBM Research, too, did basic research that ended up making major contributions to the company. It’s groundbreaking breakthroughs include hard-drive storage and solid-state memory, both cornerstones of modern computing. Though IBM had to rein in many of the more freewheeling aspects of Research when the company fell upon hard times in the 1990s and most research projects today are closely tied to business units, IBM Research still manages to do a fair amount of basic science. ((In the interest of full disclosure–and parental pride–my son Jon is a research staff member at IBM’s Thomas J. Watson Research Center.))

Not all industrial labs are great successes for their sponsors. Xerox PARC famously developed such breakthroughs as the graphical user interface and Ethernet that the company failed to exploit. And Microsoft’s massive research effort remains something of a mystery in its failure to produce successful innovations for the company.

Still, by sponsoring a real lab, Apple could do well by doing good. It’s not that Apple doesn’t do innovative research; its products a patent portfolio attest to that. But the company has a tendency to acquire key technologies, such as low-power semiconductors from PA Semi or biometric sensors from AuthenTec, through strategic investment rather than develop them in-house. And the disadvantage of a too-focused R&D effort is that you are denied the benefits or serendipity that a more wide-ranging effort can produce.

If I were running the U.S. government, I’d offer Apple a deal. Most of its cash represents foreign earnings that avoid U.S. taxes as long as they remain in foreign accounts. I would offer tax-free repatriation of the funds Apple uses to set up a research endowment. The U.S. badly needs to increase research spending. But a federal government on a starvation budget can’t do it and private industry is not picking up the slack.