The Census Bureau is More Than a Little Wrong on Divorce

The Census Bureau recent released an alarming report that most marriages from the late '70s on did not make it to their 25th anniversary. More careful examination by Justin Wolpers and Betsey Stevenson of the Wharton School reveal an elementary error. The Bureau was including marriages made in the latter half of 1979, but counted their longevity from mid-2004 -- before 25 years had actually elapsed. When those last few marriages were included, most of the late '70s marriages -- 53% -- did indeed celebrate their silver anniversary.

There is some further wrong, though. It was not just the Baby Boomers who showed a high divorce rate. Every generation of marriages showed a sizable jump in divorces in the five years preceding 2004 over the five-year period just before that. For example, for men married between 1975 and 1979, 5.6% divorces in the five years between 1995 and 1999, yet the Census Bureau shows nearly twice as many, 10.1%, divorced between 1999 and 2004. The late '70s were a bad era for marriage, but the marriage cohorts around this one, from the early '60s to the late '80s, show nearly identical huge leaps.

It is possible that divorce became much more likely for everyone in the early '00s. What this looks like to me, though, is that the Census Bureau changed the way it measured some aspect of divorce, causing an apparent leap in the divorce rate, without a real change in the underlying facts. More as this story develops.