3 ms·

Indeed. Hard to design a double blinded, randomised, placebo controlled trial for suicide prevention.

• dmix 7y ago
  I was thinking the same for economics. How can you really measure all of the side effects of a government policy? For ex: if the conditions in a marketplace are so bad (for ex: because of a monopoly, or gov-back entrenched businesses, or other useless barriers to entry) that no one ever even tries to start a business, so there's no one ever to complains and nothing to measure. At most you have some increasingly out-of-date information from before the policy was implemented, when the wider market, technology, access to capital, etc has changed. Another example is minimum wage laws, it's very difficult to measure what costs there were if the jobs simply weren't created, not simply going declining in existing businesses or measurable marketplace. Combined with the countless other factors which influence hiring and cyclical trends. I'm not sure what the solution is here but I'm skeptical of a lot of economic research.