> The problem is, if you do everything right, and there's no economic collapse, people wonder why you need to exist. If there is, you're clearly useless.
Interesting. Can we come up with some general method of answering questions like "does institution X really play a part in keeping parameter Y stable?"
The problem is: the only way to detect this, as the article points out, is for them to stop doing their job well intentionally. You either wreck the perfectly good situation you have locally, or you find some other place where everything is terrible, and you compare the two. This is one of the reasons why scientific methods are hard to apply in non-experimental situations: you can't necessarily destroy the national economy 'to see if it would really work'. It's the same as informed consent in medical experiments, just on a larger scale.
Interesting. Can we come up with some general method of answering questions like "does institution X really play a part in keeping parameter Y stable?"