Evidence-based policy

I’m all for basing social policy on good social science evidence. But as Dean Dad writes:

We have anecdotal evidence that suggests that students who actually take math for all four years of high school do better in math here than those who don’t. We also have anecdotal evidence that bears crap in the woods. Why the hell do the high schools only require two years of math?

I say we can bypass the regression analysis on this one.