Monday, April 25, 2011

The Typical Use Failure Rate of Abstinence

Last Friday I wrote a post about birth control effectiveness. I wished to demonstrate that even a careful user of birth control has a fairly high probability of becoming pregnant, and the assumption that all persons seeking abortions are pregnant because they screwed up their birth control is just plain wrong.

If you missed the last few paragraphs, though, you might think I was building up to some sort of abstinence-only argument. Even if its failure rate were zero, I would not argue that abstinence is a good birth control method. If you want to have sex, I am sure as hell not going to tell you not to.

But the typical use failure rate for abstinence is definitely not zero. Good statistics are hard to come by, but most attempts at quantifying the failure rate of abstinence rank it worse than condoms.

The most optimistic non-zero value cited for the failure rate is 26%. Time for another plot!

With these numbers (which are admittedly much more poorly constrained than the failure rates I discussed on Friday), a person using abstinence as their main form of birth control has a roughly 70% chance of getting pregnant within 5 years. Definitely not a great method for preventing pregnancy. I cannot emphasize enough how useless I find abstinence-only sex ed.


  1. Isn't this apples and oranges, however? Your (sobering, sobering) Friday post--correct me if I misread--had more to do with applying a 15% failure rate out over a period of thirty or more years. Whereas you're argument here seems to be more geared toward a failure to apply abstinence properly. Ie: eventually, these people are going to do it, and then get pregnant. I'm not about to go preaching abstinence to *anyone,* but it works as a technique if not as an ethos. There would seem to be a fairly low rate of pregnancy among nuns, for example.

  2. No, actually, I think this is comparing apples and apples! All the failure rates I quoted Friday were typical use failure rates. If you use condoms perfectly, the failure rate drops to a few percent, and if you use abstinence properly, the failure rate drops to zero. But I'm not interested in perfect use; as I said Friday, "are you really going to choose your birth control assuming that you're going to be better at it than everyone else?"

    I can't really comment on pregnancy rates among different abstinent groups, as I don't have that kind of data.

  3. I see, I see. I misread a key component of your earlier column, which throws in to doubt the fact that, like everything else I do, I *do* pick my birth control methods based on the assumption that I'm better than everyone else.