Apparently a growing number of researchers are saying so, even alleging it partly explains the severe black-white mortality gap. To a certain degree they’re right — some infant deaths are stress-related, racism causes stress, and blacks face more racism than whites.

John Lott needed little more than a Google search to prove it’s not that big a cause, though. He cites one report:

Since racism improved for blacks but not so much for whites in that time span, if it’s a major cause we’d expect black infant mortality to have improved more than white mortality.

However, there are a bunch of different ways to look at the data, and I’m not sure which is right for the question at hand. Overall infant mortality fell from 29.2 to 7.1 per 1,000 births between 1950 and 1999. For blacks the numbers are 43.9 and 14.6, whites 26.8 and 5.8.

(Because of the different ways the data were collected in different years, I had to conflate “white/black children” with “children born to white/black mothers.” For the years where the numbers overlap, they’re pretty similar.)

One might point out that the black mortality rate improved by about 30 deaths per 1,000, whereas whites only improved by 20. Maybe racism was causing 10 or more deaths for every 1,000 black infants, and now it’s causing 10 fewer. Much of the remaining gap has to do with current racism. But then again, whites had fewer than 30 lives to improve by to begin with, so there’s no way to test this.

Alternately, the changes between 1950 and 1999 dropped the white rate by 78.4 percent, the black rate by 66.8 percent. The overall rate fell by 75.3 percent. This — blacks’ rates improving less than whites’ when they should have improved more — fits with Lott’s thesis.

And we should bear in mind that stress, nevermind specifically racism-related stress, is but one of many factors that could impact the disparity:

poverty, poor nutrition, inadequate prenatal care, teen pregnancy, heredity, high blood pressure, stress, obesity, low birth weights and prematurity.

In trying to narrow the gap, the government could work on any one of those and have more success than in trying to ameliorate subtle, modern racism.

Finally, one interesting thing I saw in the data is that overall infant mortality fell from 20 to 12.6 between 1970 and 1980. There were significant improvements at other times (26 to 20 in the 1960s, for example), but I would guess that’s largely a Roe v. Wade effect. I’d imagine mothers who wish they could abort their kids, but can’t, don’t work too hard to stave off birth complications. Not to go all Steven Levitt or anything.

Blog: Robert VerBruggen