I know this is going to sound crazy controversial, but the reason that healthcare companies innovate is to make a profit. And those profits are the first thing that politicians target when they aim to keep costs down. Sadly, so far there's little recorded success with things like drug and medical equipment development outside of the private sector.
Dean Baker is proposing that the government should set up a parallel system, to prove how awesome it can be at drug development. Perhaps surprisingly, I agree--we should have empirical validation of the notion that the market is better at drug development. But the metric has to be the same as for private companies: an actual drug that people take. Drug development is not, as the activists screaming that the NIH "really" invents all the drugs, a simple matter of finding a target that might have some effect on a disease. Once you've found a target, you need a molecule that will hit it. And not just any molecule. It has to be small enough to dose orally, unless you're developing a short-term treatment for something really gnarly like cancer. It has to make it into the blood at detectable levels without gettng chewed up by the liver. Once in the blood, it has to do what you expect, which it often doesn't, and not do anything you don't expect, like kill the patients. It has to be cost-effective to manufacture, which means not only finding reasonably cheap ingredients and a short process, but also something that scales up to produce in industrial quantities--it's no good having a great molecule that can only be produced in .5 milligram lots by a team of devoted chemists. And oh, it has to improve the lives of enough people to make it worth all the research you've invested. Once you've nailed all that, and a few things I've forgotten, you have a drug. Until then, you have a maybe interesting chemical.
Unsurprisingly, I doubt that the government will turn out to be so good at this. The record of governments at inventing consumer goods is, she said with characteristic understatement, somewhat spotty.
This article available online at: