"?" Isn't there more science done by deciding what to measure on purely theoretical grounds, and then pushing technology to make those measurements possible?
That's somewhat different to deciding that because it's easy to measure X, Y or Z, everything else should be ignored.
Part of the problem is that economics students will be exposed to some version of Keynes' theories, or to some other purely monetarized formulation of economics, without a discussion of scope or, even worse, with claims of universal applicability.
Yes, exactly. Which is why from the outside it looks as if economists are taught stock answers, instead of being taught to ask deep questions.
I know doodley squat about econometrics or its history, but I can tell you Keynes does discuss the issue of what to measure, what data are available to him, how they are inadequate for his purposes, etc. Maybe I should write a diary about it. The contemporary source that he talks about is Kuznets' econometric data for the US.
As a swedish kind of death points out [and Keynes also discusses] a prime source of economic data and economic concepts (as in, what can be measured) is government tax offices. Think of Inland Revenue and National Statistics, as analogous to CERN: "big science" econometrics. Nothing is 'mere'. — Richard P. Feynman
That's an issue of scope. As long as you don't claim universal applicability, you're ok. Nothing is 'mere'. — Richard P. Feynman
