However, that doesn't mean bad FDA rules are not also an issue. From the blog post I linked to:
Any pharma company that wants their medication approved in both the US and Europe has to spend a billion or so dollars getting it approved by the FDA, and then another billion or so dollars getting it approved by the Europeans. A lot of pharma companies don’t want to bother, with the end result that Europe has many good medications that America doesn’t, and vice versa. Just in my own field, amisulpride, one of the antipsychotics with the best safety/efficacy balance, has been used successfully in Europe for twenty years and is totally unavailable here despite a real need for better antipsychotic drugs.
For genuine public health emergencies (anyone remember AIDS?) there's fast-track approval, the first antiretrovirals were at market remarkably quickly.
Let me play devil's advocate here and suggest that entrenched market participants love high-barrier-to-entry regulatory regimes. The FDA is "people in the drug discovery business'" moat.
Still, I'd be willing to say EU approval should drop the cost in some way by reusing studies or other things.
The hypothesis Balaji and others here suggest is that the FDA requires drugmakers to pass useless onerous regulations and expensive studies before it'll approve a drug. In so doing, the drugs become more expensive to market.
Having worked a little in the drug discovery area, I'd advance a different hypothesis: the low hanging fruit are gone.
Balaji in his post explicitly mentions insulin coming to market in 2 years. That's a pretty easy thing to approve: you have a molecule that is identical to one produced by your own body, that has known or easily understood physiology and pharmacokinetics. All you have to show is that your preparation isn't toxic, so of course it's going to take nearly no time to get approved.
Contrast with a cancer drug. There you have a really complicated disease, not very good model systems before you get to test the molecule in a human subject, and when you get there most of your patients die anyway because diseases are hard/they're too sick/it doesn't work well enough.
I think the information/silicon technology side of silicon valley doesn't quite understand how hard biochemistry and biotech actually is. In traditional engineering you know a lot about the inputs to your system (code, silicon, plastic, etc.) and you know a lot about how they work. In Biology/Biochemistry/Physiology none of that is true. Your inputs, underlying system and outputs are hugely complex and in many ways completely non-understood.
That's why it takes so long to "innovate" or develop new drugs. Bureaucracy may not help, but it's not the main cause either.
An essential blog to read if interested in drug discovery:
http://blogs.sciencemag.org/pipeline/
topic relevant post:
http://blogs.sciencemag.org/pipeline/archives/2015/08/07/on-...
Full disclosure: I once met with Balaji for an informal job interview/get to know the company at Counsyl