Contraceptive failure: user error?

I recently became aware of a statistic from the National Survey of Family Growth. In a study on Contraceptive Failure Rates from the Alan Guttmacher Institute, the data show that about 1 in 5 teenage females using condoms as birth control will become pregnant within the first year of use. That failure rate (20%) is much higher than the clinical tests for condoms, and inconsistent condom use is usually blamed for the high rate.

Interestingly, the failure rate increases significantly for cohabiting couples, even within the same age category. For teenagers (age < 20) who live together and use condoms, 51-71% (depending on socioeconomic status) will become pregnant within the first year.

The jump in unintended pregnancy for those couples living together is startling. These teenagers are receiving the same education about condoms and contraception as their non-cohabiting peers, and are making the same choice to rely on condoms, but the difference in lifestyle choice has a dramatic effect on their ability to avoid teenage pregnancy. Apparently, knowing how and why to use condoms does not reduce unintended pregnancy as significantly as choosing not to cohabit before marriage. This says to me that lifestyle education and encouraging positive lifestyle choices (such as abstinence) is far more beneficial to reducing teenage pregnancy than simply educating teens about condoms.