Printer friendly version

April 27, 2004

Bad Marriage Math

Its absense was palpable, so I won't flatter myself that it was in response to me that Darren Spedale added, to his piece about Scandinavian marriage, some of the hard data that I faulted him for leaving out. Now that we've got some objective numbers to address, it would seem that he's continued to be selective in those that he presents, sometimes attempting to spin the statistics past the breaking point. To start with the mild:

In 1989, at the outset of the partnership law's existence, there were 6.0 heterosexual marriages per 1,000 persons in Denmark, according to Danmarks Statistik, the national statistical organization. By the mid-1990's that number had climbed to 6.8 marriages per 1,000 population, or an increase of just over 10% from 1990. As of 2002, the latest year for which statistics are available, the number of marriages per 1,000 population has increased to 6.93.

What's new is the extension out to 2002. The mildly conspicuous omission is the degree of fluctuation during this time. Furthermore, as I offered for perspective last time, with miniscule native population growth, increases in the per-1,000 marriage rate aren't as significant as they would be in a nation with a more rapidly expanding total population. These are both relatively nitpicking points; that becomes less the case with Spedale's newly offered information about divorce:

Furthermore, the number of heterosexual divorces in 1989 stood at 2.95 per 1,000 population. By the mid-1990's, it was at 2.4 per 1,000, or an approximate 12% decrease in the number of divorces. While that number has increased in recent years to slightly below pre-gay marriage levels (2.72 in 2001 and 2.85 in 2002), the fact that the number of divorces fell during the years following the passage of gay marriage in Denmark demonstrate that heterosexual couples didn't abandon the institution when it was opened up to gays and lesbians, as many on the right predicted.

Admittedly, I wasn't paying attention back then, but I still have no idea who those people "on the right" were who predicted an immediate exodus of straights from the institution of marriage; that sounds a lot like the strawmen being thrown into the current debate. Nonetheless, there is a limited sense — but significant to Spedale's analysis — in which that prediction was realized.

Spedale is correct that the divorce rate in 2002 was "slightly below pre-gay marriage levels" — 2.81 in 1987 and 2.87 in 1988. However, the only data he provides is for 1989, the year SSM was introduced, when the rate was 2.95 per 1,000 of the population. What's especially interesting about this is that, as the by-month data shows, the 435 divorce increase from 1988 to 1989 is more than covered by a surge in November 1989 — the month after the first same-sex marriage. That month, divorces increased by 553 from October and were 479 above the average for the year.

That isn't a factor that I'd bother to include, unprovoked, in my own analysis, but it is interesting that, in using data to mitigate the recent increase in the divorce rate, it may very well be that Spedale relies upon the very phenomenon that he intends to dismiss. (The line graph that the statistics Web site provides for this data shows that the jump was much more than any seasonal boost that might have played a role.)

Spedale goes beyond all of these subtleties, as significant as they may be, when he tries to justify his claim that "divorce rates among gay and lesbian couples is so much lower than rates of divorce among their heterosexual counterparts":

In January of 2004, there were approximately 2468 gay and lesbian couples, or 4936 individuals, in registered partnerships in Denmark, according to Danmarks Statistik. Danmarks Statistik also records 1169 individuals as divorced from these partnerships. (The organization does not carry statistics on the annual number of same-sex divorces.) This means that, at a minimum, 585 same-sex couples have divorced since the gay marriage came into being 15 years ago. This would equate to a divorce rate of approximately 19% of all same-sex couples.

However, to be fair, we must assume that a percentage of such 'divorcees' have remarried, taking themselves out of this pool. Nevertheless, even if we assume that 50% of all divorced same-sex couples have remarried in this period of time (unlikely, but possible), then we would calculate that a total of approximately 26% of all same-sex marriages have ended in divorce. Compared to the divorce rate for heterosexual married couples in Denmark, which in 2002 was 41% (or to the U.S. rate, which is closer to 50%), gay and lesbian marriages are indeed more stable than those of heterosexual couples.

It took me longer than I'd like to admit to figure out Spedale's math here. The problem, I ultimately concluded, is that he doesn't "assume that 50%" of divorced homosexuals remarried. That would mean that the same number of same-sex couples had remarried as were still divorced, for a total of 1,170 couples, or 32% of all same-sex marriages. Instead, he got the 26% by multiplying 585 by 1.5, which actually ends up assuming that one-quarter of divorcees remarried. Even 32%, though, is obviously less than 41%; unfortunately, the two percentages measure the divorce rate in incomparable ways. Spedale got the 41% by dividing the number of opposite-sex divorces in 2002 by the number of new marriages in that year. That's a very different measure than the number who have ever been married or divorced.

Before I homogenize the data, I should confess that I have no idea where Spedale got 4,936 SSM individuals in 2004. According to my source, which I believe is his, the number should be 5,577. Thus, his 19% divorce rate should be 17.3%, and his adjusted 26% divorce rate should be 23.9%. From Spedale's perspective, that's even better. Still, applying the same calculations to the numbers of heterosexuals married and divorced as of 2004 (2,154,117 and 405,198), the corresponding divorce rates are 15.8% and 22.0%. Whatever these numbers show, it certainly isn't that "gay and lesbian marriages are indeed more stable than those of heterosexual couples."

And cut the numbers however we may, it is still ridiculous to declare the door closed. We'll see which way Scandinavian marriage and family statistics go, and I can only guess that, deep down, Spedale realizes that it doesn't look good for the argument that he's attempting to make.

Posted by Justin Katz at April 27, 2004 12:51 PM
Marriage & Family

Or many he's just bad at math.....

Posted by: lucia at April 30, 2004 7:30 PM