MARRIAGE IS SLOWLY DYING IN SCANDINAVIA. A majority of children in Sweden and Norway are born out of wedlock. Sixty percent of first-born children in Denmark have unmarried parents. Not coincidentally, these countries have had something close to full gay marriage for a decade or more. Same-sex marriage has locked in and reinforced an existing Scandinavian trend toward the separation of marriage and parenthood. The Nordic family pattern--including gay marriage--is spreading across Europe. And by looking closely at it we can answer the key empirical question underlying the gay marriage debate. Will same-sex marriage undermine the institution of marriage? It already has.

More precisely, it has further undermined the institution. The separation of marriage from parenthood was increasing; gay marriage has widened the separation. Out-of-wedlock birthrates were rising; gay marriage has added to the factors pushing those rates higher. Instead of encouraging a society-wide return to marriage, Scandinavian gay marriage has driven home the message that marriage itself is outdated, and that virtually any family form, including out-of-wedlock parenthood, is acceptable.

This is not how the situation has been portrayed by prominent gay marriage advocates journalist Andrew Sullivan and Yale law professor William Eskridge Jr. Sullivan and Eskridge have made much of an unpublished study of Danish same-sex registered partnerships by Darren Spedale, an independent researcher with an undergraduate degree who visited Denmark in 1996 on a Fulbright scholarship. In 1989, Denmark had legalized de facto gay marriage (Norway followed in 1993 and Sweden in 1994). Drawing on Spedale, Sullivan and Eskridge cite evidence that since then, marriage has strengthened. Spedale reported that in the six years following the establishment of registered partnerships in Denmark (1990-1996), heterosexual marriage rates climbed by 10 percent, while heterosexual divorce rates declined by 12 percent. Writing in the McGeorge Law Review, Eskridge claimed that Spedale's study had exposed the "hysteria and irresponsibility" of those who predicted gay marriage would undermine marriage. Andrew Sullivan's Spedale-inspired piece was subtitled, "The case against same-sex marriage crumbles."

Yet the half-page statistical analysis of heterosexual marriage in Darren Spedale's unpublished paper doesn't begin to get at the truth about the decline of marriage in Scandinavia during the nineties. Scandinavian marriage is now so weak that statistics on marriage and divorce no longer mean what they used to.

Take divorce. It's true that in Denmark, as elsewhere in Scandinavia, divorce numbers looked better in the nineties. But that's because the pool of married people has been shrinking for some time. You can't divorce without first getting married. Moreover, a closer look at Danish divorce in the post-gay marriage decade reveals disturbing trends. Many Danes have stopped holding off divorce until their kids are grown. And Denmark in the nineties saw a 25 percent increase in cohabiting couples with children. With fewer parents marrying, what used to show up in statistical tables as early divorce is now the unrecorded breakup of a cohabiting couple with children.

What about Spedale's report that the Danish marriage rate increased 10 percent from 1990 to 1996? Again, the news only appears to be good. First, there is no trend. Eurostat's just-released marriage rates for 2001 show declines in Sweden and Denmark (Norway hasn't reported). Second, marriage statistics in societies with very low rates (Sweden registered the lowest marriage rate in recorded history in 1997) must be carefully parsed. In his study of the Norwegian family in the nineties, for example, Christer Hyggen shows that a small increase in Norway's marriage rate over the past decade has more to do with the institution's decline than with any renaissance. Much of the increase in Norway's marriage rate is driven by older couples "catching up." These couples belong to the first generation that accepts rearing the first born child out of wedlock. As they bear second children, some finally get married. (And even this tendency to marry at the birth of a second child is weakening.) As for the rest of the increase in the Norwegian marriage rate, it is largely attributable to remarriage among the large number of divorced.

Spedale's report of lower divorce rates and higher marriage rates in post-gay marriage Denmark is thus misleading. Marriage is now so weak in Scandinavia that shifts in these rates no longer mean what they would in America. In Scandinavian demography, what counts is the out-of-wedlock birthrate, and the family dissolution rate.