Are Marriage Statistics Divorced from Reality?

A new book takes aim at the notion that 50% of newlyweds are destined for Splitsville

  • Laughing Stock / Corbis

    Do half of all marriages really end in divorce? It's probably the most often quoted statistic about modern love, and it's a total buzz kill, in line with saying that half of all new shoes will give you hammertoes or that 50% of babies will grow up to be ugly. Now the divorce stat is coming under scrutiny — and not just because of its unromanticity.

    "It's a very murky statistic," says Jennifer Baker, director of the marriage- and family-therapy programs at Forest Institute, a postgraduate psychology school in Springfield, Mo. She's often erroneously credited with arriving at the 50% figure; it was around long before she used it. Figuring out divorce rates is tricky. Not all states collect marital data, and the numbers change dramatically depending on the methods and sources that are used. In the end, the best that researchers can do is look for trends within a specific group or cohort (say, all people who married in the 1980s) and project what will happen. As Baker says, "It's very difficult to know, if a couple gets married today, whether they'll still be married in 40 years."

    But in an upbeat new guide to marriage, For Better , Tara Parker-Pope, a New York Times reporter (and divorcée), devotes a chapter to debunking the 50% stat, at least among the subset of the population that reads books like hers. Since the 1970s, when more women started going to college and delaying marriage, "marital stability appears to be improving each decade," she writes. For example, about 23% of college graduates who married in the '70s split within 10 years. For those who wed in the '90s, the rate dropped to 16%.

    According to research at the University of Pennsylvania's Wharton School, one of the clearest predictors of whether wedding vows will stick is the age of the people saying them. Take the '80s: a full 81% of college graduates who got hitched in that decade at age 26 or older were still married 20 years later. Only 65% of college grads who said I do before their 26th birthday made it that far.

    But just 49% of those who married young and did so without a degree lasted 20 years, a cohort that Parker-Pope spends little time discussing. Instead she contends that the 50% stat is a myth that persists because it's something of a political Swiss Army knife, handy for any number of agendas. Social conservatives use it to call for more marriage-friendly policies, while liberals find it handy to press for funding for programs that help single moms.

    Moreover, Parker-Pope argues, all the talk about grim marriage stats becomes a self-fulfilling prophecy. "It makes us ambivalent and more vulnerable to giving up when problems occur," she writes.

    Perhaps, but there may still be truth to it. Penn State sociologist Paul Amato, in a thorough new report on interpreting divorce data, writes that the half-of-all-marriages-end-badly figure still "appears to be reasonably accurate."

    What seems most clear is that less-educated, lower-income couples split up more often than college grads and may be doing so in higher numbers than before. "The people who are most likely to get divorced have the least resources to deal with its impact, particularly on children," says Amato.

    Talk about buzz kill.