The other day, I wrote that it might be time to socialize drug research, or at least take a kangaroo hop in that direction. (Granted, the Pharma lobby would never let this happen, but let's keep things unrealistic for now.) In comments, serial catowner pointed out that the drug industry isn't in any way innovative these days, which I definitely agree with, and JimPortlandOR pointed to the Bayh-Dole Act of 1980 as a culprit mucking things up. Good point. So here's a suggestion.

To recap: The Bayh-Dole Act, in essence, transferred the patents for all pharmaceutical inventions made with the help of federal research grants to the universities and small businesses where they were made. No longer would taxpayers own the research that the government had paid for; it would be in private hands from now on. Many people credit this act with spawning the multi-billion dollar biotech industry, since in 1979, only 5 percent of government-held patents had ever been developed—because companies didn't want to risk commercializing them if they didn't own the patents. Bayh-Dole fixed that, in theory. But it also made academic institutions much more unwilling to share their research with other scientists, instead spending their time seeking out licenses with private business in order to earn millions. Out with the altar of Hermes, in with Mammon.

Whether or not Bayh-Dole was warranted at the time—if nothing else, it helped many research universities reap windfalls—it's certainly not having a positive effect on drug innovation today. Between 2000 and 2003, the average number of "new molecular entities," or genuinely new drugs (as opposed to "me-too" drugs) dropped to eight a year, and few of them were by the major corporations. More tellingly, drug companies have made often made relatively little progress on any number of important diseases in the past few decades. There's virtually nothing out there to treat MS, or Parkinson's, or Alzheimer's. Diabetes treatments have stalled. Cancer medications aren't really going anywhere. Perhaps that's just because these things are intrinsically difficult. But one theory is that, because government-funded research institutions now worry more about cashing in on their inventions, they spend more time hoarding their research, groveling for contracts, and litigating over patents than they do collaborating fruitfully with other scientists. Plus, Bayh-Dole inflates the price of drugs—drugs researched with taxpayer money.

The thing is, as Clifton Leaf pointed out in a recent Forbes article, there's another "high-technology, university-incubated industry" that's doing perfectly fine without anything like Bayh-Dole: the computer industry. The IT industry still has patents, of course, but companies and research institutions are much more generous in licensing their technology, and inter-company sharing is much more widespread. In part because of all this sharing, computer prices keep going down, Moore's law is awesome, and innovation after innovation keeps cropping up. Meanwhile, entrepreneurs and researchers at universities don't need restrictive patent rules as incentive to innovate: Leaf points out that the "$50K Competition" at MIT, which offers a mere $50,000 in seed money for innovative business plans, "has showcased some notable winners—and losers" over the years, including Ask Jeeves and Akamai. Smart people will always find ways to bring good ideas to the market, and, the more widely ideas are shared, the more stuff they'll probably invent. No reason the pharmaceutical industry should be any different.

Now here's the kicker: technically the Bayh-Dole Act empowers federal agencies to ensure that new technologies—gene analysis, cell lines, research techniques—are being shared as widely as possible. But the NIH has never once used this power. As well, Bayh-Dole technically allows the government power to use its taxpayer-funded research royalty-free, but it's never done that. One wonders: What the hell? Government reticence on both these measures essentially acts as a taxpayer subsidy to Pfizer and Bayer, and hinders innovation. I guess that's the point, but it sucks. I honestly don't know whether it's time to repeal Bayh-Dole altogether. As a "compromise" an amendment could be passed, though, that forces the government to do the above two things—and require scientists to license their patents as widely as possible—at minimum.