More marriage predictions gone awry

If the nuts and bolts of demography aren’t for you, may as well save yourself and move on to something else. On the other hand, if you do like this maybe you should consider this whole demography thing for yourself.

Following up my previous rumination on marriage projections, I’ve noticed that another landmark prediction from the past underestimated the eventual number of marriages among Black women Baby Boomers. Paul Glick, a famous Census demographer, predicted in 1984, based on 1980 data, that 75% of Black women born in the early 1950s would eventually marry. Among those who were still alive in March 2010, however, 83% were ever-married. Since they are now only in their 50s and 60s, another 5% or so may still marry.

A 1992 Census report by Arthur Norton and Louisa Miller (which someone thankfully took the trouble to preserve on the website), reiterated Glick’s projections for the same group of women. Noting that, as of 1990, 75% of Black women in their late 30s had ever married, they wrote:

“Assuming that the small amount of first marriages that have taken place after age 40, both among Black and White women, will remain the pattern for the near future, less than 3 out of 4 Black women will eventually marry, compared with at least 9 out of 10 White women.”

In fact, that 75% mark has now been reached by Black women born in the first half of the 1960s, who were the next cohort coming along as Norton and Miller wrote.

That might not seem like such a big error — some 10 percentage points — but  the alarm that was sounded at the time was quite extreme, with the likes of David Popenoe (in 1993) referring to a crisis on the scale of “‘end-of-the-line’ family change.” Revising that number from 75% up to 85% might mean they should have cut the alarm volume in half.

Incidentally, that Glick article has been cited 117 times, and no one appears to have pointed out that the Black projection didn’t quite pan out. It’s asking a lot to hold people to their predictions from 1984 all the way forward to 2010. So this isn’t about blame, but about correcting the past and maybe improving predictions for the future.

The Norton and Miller report suggests that the problem lies in underestimating later-life first marriages, especially among Black women. Another, grimmer possibility is that never-married Black women are substantially more likely to die young than those who are married, which would lead to an increase in ever-married women living at older ages, even if there are on more marriages.

So, my two questions — for someone who knows, or for someone who wants to do some new research — are: what’s up with later-life marriage among Black women? And, does mortality vary a lot by marital status among middle-aged Black women? And, since this is a blog post rather than traditional, boring, peer-reviewed academia, we’ll need that answer within a couple hours.