February 02, 2004,
In "The End of Marriage in Scandinavia," I show that gay marriage has helped hasten the decline of marriage. Andrew Sullivan dismisses my argument, claiming I fail to show causality, and draw impermissible inferences about gay marriage from Scandinavian registered partnerships. Trouble is, when Sullivan thought he could prove that marriage is not undermined by registered partnerships, he was happy to argue causality, and eager to equate registered partnerships with gay marriage. Now that we see that Scandinavian marriage is in a state of collapse, Sullivan pretends that Scandinavia has no relevance to the gay-marriage debate. In the meantime, Sullivan ignores one of the key points of my piece that Scandinavian gays themselves have rejected the "conservative case" for gay marriage. To see why Sullivan is wrong, let's take a look at marriage in Norway.
Consider "Church flies gay flag," a story from the English-language edition of Aftenposten, Norway's premiere newspaper. Two parish councils in northern Norway recently voted to fly rainbow flags on their churches. The flags signal that no one in these churches priests included may speak or preach against homosexual behavior. The flags also welcome gay clergy, including those who live in "registered partnerships" (i.e. de facto gay marriage).
Obviously, in the county of Nordland, where these two parishes are located, gay marriage has achieved a high degree of acceptance. After all, the Lutheran church has long led the opposition to gay marriage in Norway. One of the few things distinguishing same-sex registered partnerships from marriage is that they cannot be celebrated in the Norwegian state church. And the ordination of clergy in registered gay partnerships is the most divisive question in the church. So when two parishes in the same county fly the rainbow flag to welcome partnered gay clergy, gay marriage has obviously achieved an extraordinary degree of popular acceptance.
That acceptance isn't total. Many are unhappy with the flags and the silencing of conservative congregants and priests that the flags symbolize. And as the original news accounts make clear, these parish councils are acting in defiance of their bishop. Clearly, though, Nordland is a socially liberal county in which gay marriage has achieved a high degree of acceptance. So what's the state of marriage in Nordland?
Marriage in Nordland is in severe decline. In 2002, an extraordinary 82.27 percent of first-born children in Nordland were born out-of-wedlock. A "mere" 67.29 percent of all children born in Nordland in 2002 were born out-of-wedlock. As I explained in "The End of Marriage in Scandinavia," many of these births are to unmarried, but cohabiting, couples. Yet cohabiting couples in Scandinavia break up at two to three times the rate of married couples. Since the Norwegian tendency to marry after the second child is gradually giving way, it is likely that the 67-percent figure for all out-of-wedlock births will someday catch up to the 82-percent figure for first-born out-of-wedlock births. At that point, marriage in Nordland will be effectively dead.
Now consider the county of Nord-Troendelag, which is bordered by NTNU (Norwegian University of Science and Technology). NTNU is where Kari Moxnes and Kari Melby teach two radical pro-gay marriage social scientists. Nord-Troendelag is like Massachusetts a socially liberal state influenced by left-leaning institutions of higher learning. In Nord-Troendelag in 2002, the out-of-wedlock birthrate for first-born children was 83.27 percent. The out-of-wedlock birthrate for all children was 66.85 percent. These rates are far higher than the rates for Norway as a whole.
When we look at Nordland and Nord-Troendelag the Vermont and Massachusetts of Norway we are peering as far as we can into the future of marriage in a world where gay marriage is almost totally accepted. What we see is a place where marriage itself has almost totally disappeared.
The story of the rainbow flag in Nordland embodies one of the causal mechanisms I outlined in "The End of Marriage in Scandinavia." There I showed that gay marriage had split the Norwegian church and weakened the position of those clergy most likely to speak out against the trend toward unmarried parenthood among heterosexuals. In Norway, the clergy most accepting of gay marriage are the clergy least likely to criticize unmarried parenthood. With priests who see homosexuality as sinful effectively banned from churches, their criticisms of out-of-wedlock parenthood will be lost as well. Since traditional religion is one of the strongest barriers to out-of-wedlock births (conservative religious districts in Norway have by far the lowest rates), it's obvious that the flag movement will help remove a key counterforce to the decline of marriage. And it is very unlikely that conservative priests would have been so thoroughly and effectively banned if the issue were only unmarried heterosexual parenthood. It took the question of homosexuality to produce what amounts to a near total purge of conservative clergy from Nordland's churches.
The deeper point is that, contrary to the "conservative case," those who favor gay marriage tend to favor or condone unmarried parenthood. The connection between gay marriage and unmarried parenthood extends to all sectors of Scandinavian society religious or not. So when professors from NTNU use the example of gay marriage to argue that marriage is unnecessary for parenthood they have just as much effect on their secular "congregations" as Lutheran clergy have on theirs.
Although Andrew Sullivan has challenged my causal analysis, the causal mechanisms I've described here are of the same type social scientists use to explain trends in marriage. Scholars agree that, when it comes to the out-of-wedlock birthrate, ideas and values are key variables. They establish causal links by noting broad correlations (like the low rate of out-of-wedlock births in religiously conservative districts of Norway), and then connecting those correlations to a cultural analysis. If religious districts have low out-of-wedlock birthrates, and if clergy preach against unmarried parenthood, it's reasonable to conclude that religion contributes to low out-of-wedlock birthrates.
The causal mechanisms I've outlined are of just this sort. One district bans clergy who oppose gay marriage (and these same clergy are the ones who criticize unmarried parenthood). Another district lionizes leftist professors who cite gay unions to prove that marriage has no intrinsic connection to parenthood. If both districts have high out-of-wedlock birthrates, it's reasonable to conclude that gay marriage contributes to those rates. Andrew Sullivan can reject that sort of analysis if he likes, but why does he accept the idea that secularism has an influence on marriage? The causal mechanism in the case of secularism is no different in kind than the mechanism I use in my own analysis. The truth is, Sullivan doesn't object to the causal analysis. He objects to what I've found.
Sullivan says there are too many independent variables to separate out gay marriage as a cause of marital decline. I've just explained how gay marriage can be separated out as a cause. But think about what Sullivan is saying. Sullivan is really saying he'll never accept any claim that gay marriage harms marriage. If the mere existence of prior causes of marital decline makes it impossible to isolate new factors, then the offer of state-by-state "experiments" in gay marriage is bogus. No matter how bad things get and no matter how clearly we show a cultural connection between attitudes toward gay marriage and marital decline Sullivan will deny that gay marriage makes any contribution to the problem.
Of course, when Sullivan thought he had statistical proof that heterosexual marriage was doing well in post-gay marriage Scandinavia, he was eager to play social scientist. Take a look at "Unveiled," the piece where Sullivan relies on an unpublished study by a kid barely out of college to prove his "conservative case" for gay marriage. When Sullivan thought he had proof that heterosexual marriage was not undermined by gay marriage, he was more than happy to tout the Scandinavian example. If it's really impossible to disentangle the gay-marriage variable, why did Sullivan introduce data in the first place?
But now, after I've exploded his use of the Spedale study, Sullivan claims that Scandinavian registered partnerships "have no relevance" to the gay marriage debate. Sullivan sure thought registered partnerships had relevance to gay marriage in 2001. But after having seen the collapse of marriage in Scandinavia, Sullivan says registered partnerships "have no relevance" to marriage.
As for Sullivan's complaint about my use of the terms "de facto gay marriage" or "gay marriage" for Scandinavian registered partnerships, I've simply adopted Sullivan's own language. In "Unveiled," Sullivan himself calls registered partnerships "de facto gay marriage" and "gay marriage." And by the way, in "Unveiled," Sullivan used data on Vermont's civil unions to draw conclusions about "gay marriage." Yet now Sullivan is attacking me for doing exactly what he did three years ago.
Sullivan is wrong to say that Scandinavian registered partnerships are open to heterosexuals. They're not. Sullivan wants to claim that registered partnerships are a "marriage lite" that attracts large numbers of heterosexuals and thus weaken conventional marriage. This is how Sullivan wants to explain the decline of Scandinavian marriage. But Scandinavian heterosexuals do not enter into registered partnerships, so Sullivan's way of explaining the decline of marriage in Scandinavia is wrong. (I see Sullivan has now corrected his error. But he's avoided acknowledging that his mistake sinks his explanation for the link between gay marriage and the decline of marriage in Scandinavia.)
While we're at it, where is Sullivan's causal warrant for the "conservative case" for gay marriage? How can Sullivan proclaim with such confidence that gay marriage will strengthen marriage when (according to his new position, anyway) formal gay marriage has existed only for a couple of years in the Netherlands, and no other evidence has any bearing on the question? If Sullivan is such an empiricist, why doesn't he express more uncertainty about the effects of gay marriage? Given the fact that marriage is fast disappearing in the very places most hospitable to gay marriage, you'd think Sullivan might at least consider the possibility that his totally ungrounded predictions about the future are wrong.
And note that "The End of Marriage in Scandinavia" refutes the "conservative case" for gay marriage on several matters that have nothing to do with the causal question. Scandinavian gays have not taken to monogamous marriage, and they openly reject the "conservative case" for gay marriage. Sullivan says nothing in response to these points.
The mechanism by which gay marriage undermines marriage is easy to grasp. We see it at work in Sullivan's own writings including his reply to me. Sullivan claims that "coupling not procreation is what civil marriage now is." That is false. Just because we can find cases in which infertile couples marry, Sullivan thinks he's proven that marriage has nothing to do with parenthood. But marriage and parenthood are still deeply linked. That is why Scandinavia's practice of unmarried parenthood shocks us.
Scholars treat the connection between marriage and parenthood as something that erodes gradually. That is why Sullivan is mistaken to say that American marriage is about coupling, not procreation. The connection between American marriage and parenting may have diminished, but it is far from gone as is quickly revealed by the European comparison.
But every time Andrew Sullivan claims that marriage is about coupling, not procreation, he helps weaken the connection between marriage and parenting in America. The gay-marriage debate is eroding the cultural connection between marriage and parenthood. Despite all the changes in marriage since the Sixties, Americans have a long way to go before marriage and parenthood are decoupled to the degree that they are now in Nordland and Nord-Troendelag. There is more than enough scope for a new factor to intervene and heighten that separation. This is exactly what gay marriage has done in Scandinavia and is doing right now in America, especially through the work of Andrew Sullivan.
I don't mean to deny Sullivan the right to advocate for gay marriage. He has every right. But the fact is, Andrew Sullivan himself is the causal mechanism by which gay marriage undermines marriage. His persistent belittling of the connection between marriage and parenthood and his attempts to elevate infertile exceptions into the rule for a transformed understanding of marriage are laying the cultural groundwork for a Scandinavian-style disappearance of marriage in the United States.
I end with three questions for Andrew Sullivan. 1) Is it mere coincidence that in districts of Norway where de facto gay marriage (your phrase) is most accepted, marriage itself is virtually dead? 2) If this is not pure coincidence, how would you explain the connection? (Remember, your marriage-lite theory doesn't work.) 3) Would it be possible for gay marriage to be an effect of the decline of marriage, without also becoming a contributing cause?