+1 to this...
The original source article says
"In the early years, drugs easily beat the placebo: They were, on average, 4.5 times as effective, where effectiveness means how well they lowered blood pressure, vanquished tumors, lifted depression or did whatever else they were intended to.
* * *
Experts disagree on why that should be, but suspicions range from the U.S. regulatory process to basic biology."
The standard wisdom in the industry, according to those who work in the field, is that the FDA is not terribly concerned with efficacy. Any new drug needs to merely be more effective that nothing at all. It cares first and foremost about safety- that's what politicians hear about ("My great aunt Hilda was on drug X and it killed her!!"), so that's what the FDA looks at. New drugs are not removed from the market because of failures of efficacy, but safety issues. Think of Vioxx- an effective pain drug that worked just like ibuprofen, but with better GI safety. It was pulled because of increases in the risk of adverse cardiac issues, especially in patients at high risk for cardiac events, at high doses of the drug. (to be fair, part of the issue was that the company looked like it had covered up the cardiac data... but "safety" killed that product.)
This is different in other countries- efficacy is (relatively) more important in e.g., Japan, and drugs that are equally effective as older drugs have harder times being approved, and safety is not the end-all, be-all.