*Remember you can vote once per day for the About.com Atheism Awards. I’m one of five nominees for Best Atheist Blog. More details here.*

The Bright Maidens, a trio of young Catholic women who try to dispel myths and misconceptions about their faith, frequently do themed blog carnivals, and this week their topic is “Instructing the Ignorant.” Elizabeth Hillgrove of Startling the Day covered a lot of topics in her post, but you can’t be surprised to find out it was this one that caught my eye:

I got to explain to her, no, NFP is not like the rhythm method anymore. I explained that the rhythm method was used back in the 1930s and that we know a lot more about how best to avoid or to achieve pregnancy while keeping God in our bedrooms.

I also used the opportunity to explain that most of the methods are at least 97% effective when used properly (which relies on married couples communicating — another benefit), which is far more promising than something like condoms. The error margin of condom use increases exponentially with every use (statisticians, help me learn how best to explain this, please!), while NFP methods remain effective as long as couples chart and communicate.

You call for statisticians and I appear!

I think I know what Elizabeth is referencing when she talks about the risk of condom failure growing with every use, so here comes the quickie explanation. Imagine you have a weighted coin that is really, *really* likely to come down heads every time you flip it. Let’s say 97% heads and 3% tails. Obviously, you’ll call heads every time you flip it and make a profit. But if your goal isn’t to be right most of the time, but to *never be wrong*, you’re going to run into a problem pretty fast.

How do you figure out how likely it is you’ll lose a coin flip if you keep calling heads. Well, the chance that you’re *never* wrong in N flips is .97^N. So, the chance that you miss at least one is 1-.97^N. So if we flip 20 times, the odds I’ll miss at least one call is 46%. Not good if failure is costly.

The coin flip isn’t exactly analogous to the failure rates of condoms. According to the Guttmacher Institute, with perfect use, condoms are 98% effective over a year of use. So the errors don’t mount up anywhere near as fast as they would if they were only 98% effective each time you used them. But the same problem exists — in five years of use at 98% efficacy, 10% of users will get pregnant.

But this problem is decidedly*not*unique to condom usage. It applies to any method of forestalling pregnancy where the chance of failure each time you use it is

*independent*. ‘Independent’ is a statistics word which means the chance of failure is the same in every trial — it doesn’t matter what the previous results have been. A coin flip is independent, but drawing cards from a deck is not, as long as you’re not replacing the cards you drew. Most contraceptives’ risk of failure is independent or at least has an independent component (putting a condom on correctly, taking the pill at the same time every day), but IUDs and sterilization are more binary — either the procedures were done right or not — the independent component is small. NFP presumably has an independent risk — the risk of making errors in charting.

I think the most likely difference between NFP and methods like the Pill and condoms has nothing to do with those ‘perfect use’ statistics. Those numbers tend to be irrelevant since a lot of people use contraception correctly. The ‘actual use’ success rates for condoms and the pill over a year are 91% and 83% respectively. At those rates, the five year failure rate for condoms shoots to 62%.

Guttmacher doesn’t have statistics on ‘typical use’ for NFP, just the perfect use rate of 98% effective. My hunch is that more people who sign on for NFP hit that mark than people who use condoms. NFP is visibly hard — people who commit to using it take classes. Condoms are perceived as easy to use, and abstinence only education doesn’t clue people in on common errors or even that their use is not necessarily intuitive. So, the people that opt for the ‘hard’ method, aware of how hard it is, are probably more likely to stick to the ‘perfect use’ regimen than people who think they have an easy option.

This difference might very well vanish if NFP became more common, so NFP users were less likely to be people who had made a carefully considered choice. Ideally, everyone trying to delay pregnancy would put as much effort into their regimen as NFPers do. Or they’d opt for a method that requires low user commitment like an IUD or a patch.