How then do we account for lower abortion rates, driven by a steep decline among teens?
John Santelli et al. looked at the 27% decline in pregnancy rates for girls 15-19 between 1991 and 2000, and concluded that improved contraceptive use accounted for 86% of the pregnancy risk decline (attributing only 14% to increased abstinence). Among 15-17 year-olds, they attributed 77% of the decline in pregnancy rates to contraception and 23% to abstinence. These conclusions are plausible only if one assumes a high level of effectiveness for contraception and ignores the implausible premise of a drop-off in contraceptive effectiveness or use among women over age 20. Where did Santelli and his colleagues go wrong? They estimated pregnancies prevented by contraception using the contraceptive effectiveness rates applicable as a whole to women 15-44: an 8% failure (pregnancy) rate among women who take oral contraceptive pills (OCs) and a 15% failure (pregnancy) rate among condom users over a 12-month period. That error is inexcusable, as evidenced by the research of Haishan Fu et al. o
