Here’s the rise in unmarried birth rate by race:
The racial trends are somewhat in parallel, but they are exaggerated dramatically for black women relative to white.
One hypothesis would be that this is the result of a changing threshold for marriage across the society, that affects black women first.
We can try to fit this model- what would you predict the black nonmarried birth rate would be, if you just predicted it from the white rate and assumed two normal distributions one standard deviation apart?
It’s fascinating to me that this does an okay job in 1960, a worse and worse job in the 70s and 80s, and then is almost perfect in the last ten years. The big transition looks like it comes in the late 90s, which would make sense- this is when AFDC was canceled and replaced with TANF, in part with the hope of boosting black women’s marriage rates and lowering nonmarried birth rates. TANF is largely considered to be a success in work promotion and largely a failure in marriage promotion, but this little statistical exercise makes me wonder if we’ve been too quick to kick it to the curb.
It feels good when a rough and ready prediction turns out to fit the data pretty well.
LikeLiked by 1 person
Dude, I know. Why should it be 1 SD?
LikeLike
Damned if I know. It is a bit of a pattern though. http://isteve.blogspot.com/2009/07/why-is-fundamental-constant-of.html
LikeLiked by 1 person