The answer: half as many children/others(and almost twice as many couples)
Observed “Children” (plus others)
3.8 = 1930 w/1930 marriage rates
2.2 real difference in favor of 1930
Counterfactual “Children” (plus others)
2.0 = 1930 w/1540 marriage rates
0.4 difference if 1930 had 1540 rates