Non-bayesian thought
Grouchy convective thought because I'm in a grouchy mood this morning.
A basic rule of metrology is to be sure your instruments are measuring ALL of the phenomenon you want to understand, not just the visible part. If part of the country lacks thermometers, your temperature readings are meaningless. If part of the country lacks plumbing, your sewage output measurements are meaningless.
Workforce participation and election participation keep dropping. All of our stats are based on the half of the population that is still involved. "Elections" and "unemployment" figures are valid in a Bayesian way, but we don't
use them in a Bayesian way. We use them as representing the whole population.
We're not asking what the unmeasured people are doing.
Articles about low workforce
guess and speculate but never seek real data. Are most of these folks doing unregistered casual work or 'gig' work? Are most on welfare as the
guessers always guess? Are most engaged in crime?
Politics isn't comparable. Everybody needs an income one way or another, but nobody NEEDS politics.
In both cases, the non-participants are
more realistic than the participants. The uncounteds have learned by long and painful experience that participating isn't worth the trouble.
So we are counting and valuing the part of the population that DOESN'T learn from experience. No wonder we're drowning in delusion.
EXPERIENCE SURVIVES. THEORY KILLS.
= = = = =
Later thought: This also applies to parenting. The MOST IMPORTANT skill of all, in ANY civilization, is birthing and raising kids. If you aren't reproducing, the whole species dies. And we're there. In the Sorosian lands, raising kids properly is hugely expensive and mostly illegal. You can't discipline them, you can't give them useful work at home, you can't pull them out of pointless school and place them in apprenticeships. So it's no surprise that the majority of would-be parents have opted out of parentforce participation.
Labels: Blinded by Stats, Experiential education, Metrology