For those who have had enough election coverage, here’s the alternate story of the day: “Sex on TV Increases Teen Pregnancy, Says Report” (Time, 11/3/08). “Teen Pregnancies Tied to Viewing Sexy TV Shows” (AP, 11/3/08). “Study First to Link TV Sex To Real Teen Pregnancies” (Washington Post, 11/3/08).
Stories like these are catnip for the news media, which loves a chance to moralize about entertainment media’s obsession with teenage sex while indulging its own prurient fascination with the subject. The trouble is that the study the story is based on is pretty obviously bogus.
The study, published in Pediatrics (11/5/08), found that teenage girls who get pregnant watch sexier TV shows than their non-pregnant counterparts. The interesting question, though, is whether watching sexy TV makes you more likely to get pregnant–and that seems unlikely, despite the tendency of journalists writing up the study to assume that correlation proves causality. (Time‘s Alice Park fell particularly hard for this fallacy, writing that the study’s findings “may explain in part why the U.S. teen pregnancy rate is double that of other industrialized nations.”)
The study’s lead researcher, Anita Chandra, is quoted in the Washington Post story: “Sexual content on television has doubled in the last few years, especially during the period of our research”–the study period being 2001-04. So what was happening to teen pregnancy rates while sex on TV doubled?
Well, they dropped every year–the birthrate for teens 15-17 was 27.4 in 2000, 25.2 in 2001, 23.2 in 2002, 22.4 in 2003, 22.1 in 2004 and 21.4 in 2005. If that’s what happens when you double the sex on TV, then it’s very unlikely that sex on TV has much influence on teen pregnancy rates. Or, if you work from the assumption that correlation means causality, then sexy TV might be a very effective form of birth control.