“All models are wrong, but some are useful,” the British statistician George Box liked to say.
As aphorisms go, it’s hard to beat. As well as neatly describing the scientific method, it skewers the notion that humans can use data and mathematics to ever replicate a world of near-infinite variability. In particular, the quip serves as a subtle rebuttal of economists’ faith in their ability to determine optimal preferences within a complex system, such as an economy or a market.
Another way to look at this is that by choosing what to measure, you implicitly choose what not to measure. And these choices determine your results, which at best can only ever hope to show a version of ‘reality’. However, as Box points out, even imperfect methodologies can matter.