In part 1 of this series I promised to go through the statistic shared in this book in a later post. The key statistic used to justify the arguments in the book is shared upfront in Chapter I (emphasis mine):
Here’s the key: Don’t Marry Young. In fact, don’t get married until you’re thirty. According to the National Center for Health Statistics, your chances of staying married more than double if you get married after the age of twenty-five. That’s right, the old “50 percent of all marriages end in divorce” statistic is literally cut in half for those who marry for the first time after twenty-five. And after thirty combined years of working with women on the verge of divorce, we’re taking it a step further and saying that you’ll have even better odds the closer you are to thirty.
Wow. That is a pretty powerful stat if it is true. But I’ve learned not to trust the headlines and to look at the source before taking this kind of statement on face value. They don’t reference which NCHS report they pulled their stats from, and so far they haven’t replied to my email Thursday asking them to point me to the specific study they used*. However, I was able to find a 2002 report from NCHS which covered the topic. Here is what it shows:
The only way I can torture these results to match the chances of divorce are cut in half if you marry after age 25** claim is if I compare the divorce rate (all races) of those who marry at age 25 or over with the rate of those who marry before age 18. But their book is specifically targeted to women in their 20s. From the Product Description on Amazon:
getting a ring on your finger is the last thing you should be thinking about when you’re in your twenties
To show how dogmatic they are on this point and how they use the statistics, take a look at this transcript at the Washington Post*** of a discussion they had with a young woman while promoting the book. One woman asks:
Suburbs, VA: I’ve been dating my boyfriend since we were 17 (we are 23 now). We went to college apart, have grown as our own people, and still have a great relationship that we know will result in marriage some day. While neither of us are in a rush to get married yet, it seems sort of silly to just wait until we are an arbitrary age (say, 27) before tying the knot, when we are already with the person we’ll marry. Do you see exceptions in this case?
Their answer is that “there’s nothing arbitrary about our recommendation to wait until your late twenties to marry”, and they then repeat the reference to the statistic from the NCHS. This woman wasn’t asking why she should avoid marrying when 17 or younger, she was asking what was wrong with marrying at age 23. Not only did they not address her actual question, but they used scare tactics to frighten her away from marrying at her current age. This makes me suspect that they are either incredibly dishonest or just plain don’t know how to read basic statistics. Who in the world is counseling women to marry at 17 or younger? And why would finding that marriage at such a young age often results in divorce cause them to suggest women wait until they are 30, when the data suggests that early 20s has a very similar divorce rate as later 20s?
In considering the 23 year old woman’s actual question, what we are left with is a mild correlation between lower divorce and and a woman marrying in her later 20s vs early 20s. But correlation doesn’t prove causation. Just because people who marry later tend to divorce less doesn’t mean if everyone waited to marry there would be less divorce.
The key question is who marries later, and does this group tend to have lower divorce rates not explained by age at first marriage?
Historically those who complete college tend to marry a little later, for fairly obvious reasons. This gap has narrowed very recently, but any data analyzing divorce rates is backward looking so the gaps of the past are what matter. The 2002 report referenced above used data from a 1995 study (it is the same NCHS report we looked at here for remarriage rates), so what matters in this case is the trends in place around the 1980s. This is important because education correlates strongly with lower divorce rates, as Steven P. Martin shows in his research presentation Education and Marital Dissolution Rates in the United States. He not only found that women with a college degree are far less likely to divorce (pages 10-13), but that the difference in divorce rates isn’t explained by their later average age at marriage (p 14). The same connection with education and lower divorce rates was observed in the PEW research publication The Reversal of the College Marriage Gap.
* It is possible they were referencing a different report which I couldn’t find. If they respond to my request I will do a follow on post.
** They actually make the claim that a woman’s chance of staying married will more than double, not that her chance of divorce will be cut in half. Since the lowest chance of (all races) staying married for 10 years in the chart is 52%, I’m assuming they actually meant it would cut the likelihood of divorce in half. I’m not sure where the “more than” came from unless they were only looking at the data for white women. They could of course make a case for this but they should have clarified the distinction when they made the claim.
*** Also see how they respond to another woman’s concerns about fertility in the same transcript.