I think the relevant number is the fraction of PhDs that go one to a position that mints new PhDs. (That's where the multiplier comes from.) So that almost ne

I think the relevant number is the fraction of PhDs that go one to a position that mints new PhDs. (That's where the multiplier comes from.) So that almost never includes non-tenure-track positions, but does include assistant professors who don't go on to make tenure.