Friday, July 26, 2019

How economics can raise its game — Tim Harford

Another one on methodology.

I, my view, Tim Harford is basically correct in thinking that the optimal methodological approach in economics, as in the social sciences, is to "let a hundred flowers bloom," to borrow a metaphor from Chairman Mao. But from the POV of the "hard" sciences, "soft science" is not "real" science. That is drawing lines arbitrarily. It basically says that the other "sciences" are not physics or chemistry. Well, doh.

Anyway, the post is worth a read. It takes off from a recent paper by George Akerlof.

Tim Harford — The Undercover Economist
How economics can raise its game
Tim Harford, FT columinst


edzimmer said...

I believe that's a distortion of the "hard science" POV. IMO, the more realistic POV is that if the variables in a relationship can be precisely defined and accurately measured, it can be called a "science"; otherwise it's "philosophy".

Tom Hickey said...

Ed, that's only half of it. It's why the soft sciences are not just philosophy. But hard science also has to have a framework ("normal paradigm") from which theories can be generated. The hard sciences have this while the soft one's don't, so they look more like philosophy with some numbers than the hard sciences. Physicist Jason Smith pointed this out and I posted links.

Samuelson recognized that a key assumption of hard science is ergodicity, and so he tried to assume ergodicity for econ to distinguish it from soft science. Didn't work out so well, as the heterodox folks have been pointing out.

Ryan Harris said...

Maybe have econ for skilled people and econ forsoft skilled people. Orthodox, quantitative and empirical studies along with a rag tag group of heterodoxy and plenty of overlap. Sort of like exactly what exists now.