Gauging the Randomness of Contraceptive Failures

Jane Leber Herr, University of Chicago

This paper considers whether contraceptive failures are random occurrences. If unintended pregnancies are more costly for high-ability women, one might expect contraceptive failure rates to drop as ability rises. Using data from the NLSY, I show that the opposite holds for women with less than a college degree, and that no relationship exists for women with a college degree. The proximate determinants of a contraceptive failure are frequency of sexual intercourse, the use of contraception given the intent to prevent pregnancy, the quality of contraception used, and the length of time between the onset of sexual activity and the point when a woman wants to conceive. Assuming a negative relationship between ability and contraceptive failures focuses on the positive link between ability and contraceptive quality. I find that the most important link is instead between ability and intended fertility delay. I also find that contraceptive failures are clearly random for college graduates.