Some L&P readers may be following the Justin Wolfers guest-blog over at Marginal Revolution, where he's been discussing the Census Bureau data on divorce. See here and here. Justin, who has done some excellent work on the economics of marriage and divorce, is arguing that the US divorce rate, contrary to popular perception, has flattened, if not declined. There's some discussion of the details of the underlying data, but in the second post, Justin provides a link to an Excel file with the 2004 data.

Here's a few tidbits: Of men's first marriages begun in 1965-69, the percentage that reached a 10th anniversary was 76.9. The percentage of men's first marriages begun in 1985-89 that reached a 10th anniversary was 76.4. For 15th anniversaries, the percentage of the 65-69 group was 67.4, which compares to men's first marriages begun in 1980-84's rate of 65.3.

If you want to look at the 5 year numbers, those marriages begun in 1990-94 saw 89.4% reach a five-year anniversary. The comparable numbers for each of the five-year periods of first marriages by men preceding it are 89.3, 88.7, 88.1, 88.5. Not until marriages begun in 1965-69, of which 90.5 reached a five-year anniversary, do we see a number higher than the 90-94 number. So the five-year rate of divorce is down over the last 20 years (i.e., a higher percentage of marriages are reaching their fifth anniversary).

So despite the popular belief in many places that divorce rates are up significantly, there is no evidence to support that claim. Some rates are down slightly, some are up slightly, but most are pretty flat over the last twenty years.

It's is worth noting that saying that divorce rates are not rising need not mean that "marriage as an institution" hasn't changed. We do know that marriage rates are down and marriage is taking place later, so one likely explanation is that we are seeing fewer but better marriages. That would be consistent with the general decline in the benefits of marriage as the gains from specialization have dramatically decreased with the increase in the female labor force participation rate and the closing of the gender wage gap. The reasons to marry have much more to do with personal happiness than broadly economic considerations. Without the stronger economic incentive to marry, it may be that people are getting "pickier" and thus entering better marriages (even given that divorce is easier than it used to be).

Let me also throw out another hypothesis that Peter Lewin and I put forward in a forthcoming paper in the Review of Austrian Economics that takes an Austrian perspective on the economics of marriage and divorce. With longer life expectancies, it might be the case that few marriages can last 50, 60, or even 70 years as more and more marriages would have to be if people marry in their 20s and don't get divorced. At some point, the benefits from marriage might exhaust themselves as either socially or in some evolutionary way, humans can't keep a pairing together that long. If so, shifting marriage to later in life might make sense as it would keep the total years of marriage lower, perhaps staving off the challenge of keeping together for decades, and it would move marriage from the early years when both financial needs and companionship needs might be lower to later in life when they might be higher. Specifically, given the declining health associated with older age, having a spouse to care for you and having two incomes to draw on might be more valuable later in life than earlier. Issues of the discount rate will matter here too.

No doubt, there are a lot of objections one might raise to this hypothesis, but Peter and I think it has enough plausibility to be worth thinking about.