One thing that stuck with me was understanding race in America through labor market segmentation, that if they’re “last hired and first fired” and occupy distinct social universes unemployment is experienced as different rates, like in a 20% black, 80% white city at booming 4% unemployment might mean that a black 2% of the workers – 1 in 10 in their segregated world – are unemployed, plus a white 2%, or 1-in-40 (for a white unemployment rate of 2.5%)

But 8% unemployment might leave 5% of workers as idled blacks - black unemployment of 25%. While the 3% who are idled whites leave the white rate below 5%.

Now push to 25% unemployment, which could happen in the boom-and-bust cycles of the 19th century. Say 15% of workers are idled blacks – that’s 75% black unemployment! Even 10% as idled whites only leave the rate at 12%

Hell, reverse the 15/10 distribution and you’ve still got 50% black unemployment vs only 19% white.

And that means when the white population, living in the white community, evaluates economic policy – inflation vs. employment – not only are they the ones with wealth at risk of being inflated away but trading X% inflation for 1% less unemployment will make less than 1% employment worth of observed improvement in the community and thus seem a less rewarding move to white decision-makers, given the inflationary cost, even as blacks might favor an even more employment-first policy than under equality of employment