The headline finding here is that, among those who grew up poor, black women are the only group showing a marked difference between the risk of being in the bottom quintile of the
individual earnings distribution (for each gender), and the risk of being in the bottom quintile of the
family income distribution (for the whole age cohort). Whites do well on both counts; black men do poorly on both counts. Black women do reasonably well on the first and very poorly on the second.
This result is probably driven by the fact that black women tend to create families with black men who do poorly on both counts and thus bring down the family income results for black women.