> There was a meta-study analysis in 2015 that pointed at a decline in its effectiveness between 1977 and 2014. I wonder whether it's because studying anything to do with a human brain is hard and we can't trust the numbers or if it's because of changes in society that caused people's personality to shift in a direction which makes CBT less effective.

Another theory is that methodology doesn’t matter and the result mostly depends on personal qualities of the psychiatrist. And CBT was effective because the best specialists adopted it. When the rest of the profession caught up effectiveness went down.