How 'Mathiness' Made Me Jaded About Economics
bloombergview.com
bloombergview.com
Of course, you can use statistics to aggrgate past behavior, but all you'll end up with is a model that can describe past behavior, which you'll then misapply when trying to predict future behavior.
If only you could use statistics, or math, or tea leaves, or whatever to predict long tail events; then you might know when people were about to change their behavior...
And don't believe anyone, no matter how smart, who says they can. Look at guys like LTCM that followed their models blindly to financial ruin and almost ruined the economy too.
And even that isn't half as bad as someone trying to use models to draft social policy. Anyone wonder why we're financially overextended after 40+ years of applying Keyenesian policies on the downside and Classical policies on the upside?