A lot of it has to do with the administrative hoops that you have to jump through with the FDA. Regardless of your stance on governmental oversight, there is a ton of red tape you have to get through to even BEGIN to study a drug.

In fact, my whole company makes hundreds of millions every year from pharmas paying us to deal with it FOR them.

My billable rate is up near $300/hr now, over the course of your standard phase II trial, I'll spend no less than 100 hours on a trial and I'm just a small drop in the bucket. If the trial is a bust, then thats just money up in smoke.

As for the quoted efficacy rates, I don't know if I believe that or not. There are things called "End Points" which are statistical milestones to guage the "efficacy" of a drug. Depending on the trial/compound/phase, you'll have number of different end points and whether you make them or not determines the efficacy. Reading a trial protocol alone can give you an amazing sense of what the companies are trying to do with the drug.

For instance, on an oncology trial, one of the exclusion criterias is commonly "Subject life expectancy is greater than 6 months." Meaning that if the subject is expected to live more than 6 months, they cannot be included in the trial. Rating the expected end points on that drug is quite depressing.