[scottjad]

> Finally, regarding funding, yes, NIH spends about $31B per year, which is a lot. However, drug companies spend $4B per drug approved[1], which is an incredible amount of money when multiplied across all drugs.

I think this $4B number is the Research & Development line off a financial report. I am not an accountant but I believe this number can include tons of things that people don't normally think of as R&D but just regular cost of doing business, and what we do think of as research is often a very small percent of it. Perhaps tax laws incentivize companies to put as many expenses in this category (of like 3 categories) as possible.

On a related point, people used to quote a number like it cost $500 million to bring a new drug to market, often as a justification for patents. I remember reading an academic article that showed how the real cost was almost always 1/20 of that, but drug companies were including the construction of optional new plants that were used for existing drugs too, 20 year leases on specialty tree farms and other supplies, and huge "present value" calculations that would take $2 million spent by a university 20 years ago, use the companies cost of capital (maybe 14%), and even though the company only paid $500k for the rights a few years ago they would calcuate some number in the tens or hundreds of millions that was never paid by anyone.

[ramanujan]

  I remember reading an academic article that showed how the 
  real cost was almost always 1/20 of that

This is the Light and Warburton study. The short answer is that it's kind of like a freshman saying "I could build Facebook overnight" by wgetting Facebook's CSS/JS. But for anyone in engineering who knows what FB's backend infrastructure is like, they know it would be nontrivial to clone FB.

The best response to Light and Warburton is that if it really took only $43 million to ship a drug, then they should raise the capital and start a drug company. It would be by far the most capital-efficient and successful drug company of the last 50 years. If they have really figured out how to cut all the fat, they would be hailed throughout the industry.

But I hope to persuade you that when someone outside the industry is off by two orders of magnitude ($43M vs. $4B), it is likely that they are the ones who have missed something important. I encourage you to read Derek Lowe's more detailed critique here:

http://pipeline.corante.com/archives/2011/03/07/the_costs_of...

http://pipeline.corante.com/archives/2011/03/08/that_43_mill...