> But switching to new measures to validate your preconceived gut feeling is fraught with peril. This was explored, but most of the article was about enhancing

> But switching to new measures to validate your preconceived gut feeling is fraught with peril. This was explored, but most of the article was about enhancing existing stats: > She argues — rightly, I think — that the government should not focus on creating wholly new statistics. It should instead change and expand the ones that are already followed closely. This is basically Anscombe's Quartet -- https://en.wikipedia.org/wiki/Anscombe%27s_quartet https://en.wikipedia.org/wiki/Anscombe%27s_quartet -- four completely different sets of points that have the same mean, variance, linear regression, and other stats but plotted out represent drastically different realities. But instead of abstract datasets, we're saying drastically different underlying economic circumstances for different populations may be reported with the same GDP number. So a single summary statistic "GDP is up!" without more detailed breakdowns can mean either "everyone is doing better!" or "the investor-class is making bank while everyone else is stagnant or falling behind". The point of the article is one of those seems to explain the populist anger, and our focus on the single summary statistic creates cognitive dissonance and makes the problem worse when people say to themselves "hey, this shit doesn't match what I'm struggling with everyday."