By the Numbers

Just saw this post on the Daily Kos, a listing of teenage birth rates per "red" and "blue" states, purportedly to point out that "red" states have a bigger social problem than "blue" states (or that "red" state girls are sluts, take your pick). I can see a couple of problems with jumping to that conclusion, just by focusing on good old Iowa.



First: If you track the table back to the original CDC source, the statistics cited are birth rates, not pregnancy rates. So what the numbers show is not how many teenage girls are getting pregnant, nor the number of teenage girls raising kids, but the number of teenage girls giving birth. A simple hypothetical will show the problem with this: Suppose you have State A, in which the teenage pregnancy rate is very high, say 40 teenage pregnancies out of every thousand teenage girls that year. In State A there is also a highly accessible medical system with abortion and adoption referrals common. So the actual birth rate is only 20 per thousand, and the adoption rate is 5 per thousand, leaving only 15 teenage mothers per thousand teenage girls. The number on the CDC chart listed would be 25 per thousand, significantly higher than the number of teenage parents, but significantly lower than the number of girls who got pregnant. State B has a lower teen pregnancy rate, say 35 girls per thousand. But the residents of State B are far less pro-choice than State A, and thus abortion referrals are less often given and less often used even if provided. Adoption referrals are a more common option. So the birth rate is 30 per thousand, but the adoption rate is 15 per thousand, so the net result is also 15 teenage mothers per thousand teenage girls. The number on the CDC chart for State B would be 30 births per thousand, significantly higher than State A. But the actual pregnancy rate would be lower, and the number of teenage mothers raising kids would stand the same. The upshot: If "red" teenage girls share the "pro-life" stance commonly attributed to their parents by those who feel GBWII won only on the moral vote, you would expect them to have a higher birth rate, even if they have a lower pregnancy rate. We simply don't have enough information to draw the type of conclusion that's being advocated.



Second, the CDC birth rate chart used was from 2002. Iowa is cited as an example of a "red" state. Which it was - in 2004. It was not in 2002, not in 1996, not in 1992 or 1988. So it's simplistic to say that birth rates in Iowa in 2002 are reflective of the "red" values of this state. Particularly when the CDC chart you're citing shows that the comparative birth rate in Iowa during 1991 - a solidly "blue" year - were 42.5 per thousand, significantly higher than the 32.5 in 2002. So did becoming more "red" actually drop our birth rate? If so, how does that fit with Kos' theory that "red" states have more problems with teenage pregnancy rates?



It's a pretty chart, with interesting numbers. It just doesn't add up.