An employment rate without a precarity measure attached is a bit like a statistical claim without an effect size attached. Something fundamentally dishonest about it. There’s jobs that go away when the boss simply has a bad mood, and jobs that are robust even to pandemics.

