But this conclusion is misguided. It seems likely that women who don't want to be pregnant are more likely both to use contraception and to seek abortion than women who actively want to be pregnant or women who are ambivalent about pregnancy. If that's the case, we have a correlation/causation problem.
|Which means I get to use an XKCD comic in my blog post!|
Scientific controls allow an investigator to make a claim like "Two situations were identical until factor X occurred. Since factor X is the only difference between the two situations, the new outcome was caused by factor X."In the case of contraception and abortion, ideally we would have research that compares a control group of women to an experimental group of women. The two groups would be identical in terms of relationship stability, financial status, beliefs about abortion, feelings about pregnancy, and any other relevant factors. The control group of women would use no contraception, and the experimental group of women would use contraception. Then we could compare the rates of unplanned pregnancy and abortion for each group, and we could more reasonably talk about how contraception affects abortion rates.
A big part of the contraception debate is whether contraception actually decreases unplanned pregnancies. One side cites research showing the rate of unintended pregnancies is much lower for sexually active people who use contraception than for those who don’t.
The other side counters that not everyone would necessarily be as sexually active if contraception wasn’t so widely available. This side talks about risk compensation – the idea that if people believe contraception makes sex less risky, those people will make riskier sexual decisions. For example, they may choose to have sex more frequently, in less committed relationships, or with less careful attention to the woman’s cycle.
The fact that many women who have abortions were using contraception when they got pregnant doesn’t tell us anything about how contraception affects pregnancy rates. To illustrate the problem, here is a hypothetical situation using entirely made up numbers:
Suppose we have two groups of 100 women each. The first 100 women, called the Nope group, really don't want to be pregnant. 90% of them use contraception, and 100% of them will choose abortion if pregnant. The second 100 women, called the Meh group, are either open to or ambivalent about pregnancy. 10% of them use contraception, and 17% of them will choose abortion if pregnant.
|Each symbol represents 10 women.|
|If 90% of Nope and 10% of Meh used contraception.|
Yet, in this scenario, contraception did greatly decrease pregnancy rates. The Meh women were three times as likely to get pregnant as the Nope women. If, like the Meh women, only 10% of the Nope women used contraception and 60 of the Nope women got pregnant, all 60 of those Nope women would have chosen abortion. Then there’d be 70 abortions instead of 30.
|If 10% of Nope and 10% of Meh used contraception (scenario 1).|
|If 10% of Nope and 10% of Meh used contraception (scenario 2).|
|404 Error: Insufficient Data|
But there's a big difference between saying "contraception won't fully eliminate abortion" and "contraception makes no difference in abortion rates" or even "contraception increases abortion rates." The point about abortive women using contraception only speaks to the first statement, not the second two statements. For those statements, you need more information.