News for hackers. Stuff that matters.
Re:Megan McArdle: Finding What You're Looking For by jtbigwoo on Tuesday September 07, 2010 @12:11PM (#1669207)
Huh?

So rather than confirm economic theories by using statistical correlations to economic growth or general well-being we should...um...not "hijack economics." Or something.

What's the point of indicting pseudo-science and confirmation bias? Is she saying that economics is a pseudo-science and should not be taken seriously? Or did we decide that the existence of confirmation bias invalidated all scientific inquiry? Should economists give up on back-testing theories and go back to proposing what ought to happen when perfectly rational consumers meet perfectly rational producers?

Re:Megan McArdle: Finding What You're Looking For by Symmetry on Tuesday September 07, 2010 @02:05PM (#1669566)
I think your post is actually an excellent example of the thing being warned against. If you say "Doing X will maximize growth" that is perfectly scientific, but if you say "Doing X will maximize growth and therefore science says we must do X" is not, because the audience might give some weighting to things besides maximizing growth, changing the outcome. After all "an ought cannot be derived from as is" and people claiming to do so deserve to be criticized.

And if all you've got is a statistical correlation without any good causal reasoning it always pays to remember Goodhart's law.

Re:Megan McArdle: Finding What You're Looking For by kenjackson on Tuesday September 07, 2010 @01:36PM (#1669483)
Remember she's not an economist. She just writes about it, but from all indications she's not an expert, but a good "pop-econ" writer.