The numbers for 2005 are in, and the Dutch out-of-wedlock birthrate has done it again, shooting up a striking 2.5 percentage points. That makes nine consecutive years of average two-percentage-point increases in the Dutch out-of-wedlock birthrate, a rise unmatched by any country in Western Europe during the same period. Ever since the Dutch passed registered partnerships in 1997, followed by formal same-sex marriage in 2000, their out-of-wedlock birthrate has been moving up at a striking clip. That fact has created a serious problem for advocates of same-sex marriage. (For a visual on this, see the chart in “Going Dutch?” and imagine two further years of two-percentage point increases in 2004 and 2005.
In the last decade, only Eastern Europe has seen an increase in out-of-wedlock birthrates comparable to the Netherlands (even there, only Bulgaria’s rates are rising faster than the Netherlands’). Demographers explain the stunning increase in Eastern Europe’s out-of-wedlock birthrates by pointing to the economic and cultural traumas set off by the collapse of Communism. How striking that a prosperous country like the Netherlands should experience a spike in out-of-wedlock birthrates matched only by a region recovering from the collapse of its entire social system.
In several previous pieces (“Going Dutch?” “No Explanation,” “Dutch Debate,” and “Standing Out”) I’ve argued that the Dutch example provides us with a best-case scenario for isolating the negative causal impact of same-sex marriage on marriage itself. Increasingly, same-sex marriage advocates are running out of ways to explain the Dutch data away.
In their book, Gay Marriage: For Better or for Worse?, William Eskridge and Darren Spedale say it’s too early to draw conclusions from the Dutch experience. Too early? We’ve had a continuous nine-year spike in out-of-wedlock birthrates since the passage of registered partnerships, and a five-year continuous spike since formal gay marriage went into effect. Scholars recognized the first nine years of demographic upheaval in Eastern Europe following the collapse of Communism as significant. So why aren’t nine years of comparable change in the Dutch out-of-wedlock birthrate also significant (especially since there’s nothing else like it in Western Europe)?
I’ve argued that the long Dutch campaign for same-sex marriage (which began around 1990) helped set the stage for the big continuous spike in out-of-wedlock birthrates that began in 1997. Gay-marriage advocates rejected the idea that marriage is intrinsically connected to parenthood, and the Dutch public bought that argument. Once marriage stops being about binding mothers and fathers together for the sake of the children they create, the need to get married gradually disappears. That’s why I’ve argued that the successful campaign for same-sex marriage led to the spike in Dutch out-of-wedlock birthrates. A preliminary spike between 1994 and 1995 was likely influenced by this long public debate, even before formal passage of registered partnerships in 1997.
Eskridge and Spedale dismiss the notion that the long and successful battle for same-sex marriage could have had any effect on the Dutch public. According to Eskridge and Spedale, the Dutch campaign for same-sex marriage merely “consisted of a few lawsuits brought by a handful of lesbian and gay activists.” The campaign “had virtually no public visibility,” say Eskridge and Spedale: “Only a tiny number of straight people, and a minority of gays, would even have been aware that lesbians and gay men were seeking the right to marry….”
A Public Campaign
The Dutch gay community’s own history of the campaign for same-sex marriage shows otherwise. (See “No gay marriage in the Netherlands.”) This history details an extensive and highly visible public effort throughout the early 1990′s and beyond. Around 100 Dutch municipalities set up symbolic gay-marriage registries (much like the mayor of San Francisco and other municipalities here in the United States after the Massachusetts Goodridge decision), with ceremonies sometimes attended by hundreds of people demonstrating their support for gay marriage. These ceremonies, along with public petitions, and pressure for partnership recognition on prominent Dutch businesses, provoked a national debate on same-sex marriage that forced reluctant politicians to act. As the Dutch gay community’s own history emphasizes, public opinion polls throughout the early 1990s show that the campaign worked. Ever-larger percentages of the Dutch public came to favor same-sex marriage: it was at 73 percent by 1995.
What is the purpose of a movement if not to change fundamentally the way the public thinks? In the Dutch case, we know the movement for same-sex marriage succeeded. When they aren’t denying the obvious truth that public campaigns matter, Eskridge and Spedale make the point themselves. In their account of the Danish struggle for same-sex unions, Eskridge and Spedale describe polls showing growing public support for registered partnerships as the debate proceeded, but before the law was actually passed. Is it surprising that the same thing happened in the Netherlands?
Why has so much heat been generated over the seemingly uncontroversial point that the long Dutch campaign for gay marriage had an effect on the public’s view of marriage? Gay-marriage proponents care about this issue because of that preliminary spike in the Dutch out-of-wedlock birthrate in 1995. Since this small spike occurred before the passage of registered partnerships in 1997, gay-marriage proponents argue that the big uptick in Dutch out-of-wedlock birthrates began even before new laws were passed. Supposedly, if Dutch out-of-wedlock birthrates began to rise even before the passage of registered partnerships in 1997, this would somehow prove that gay marriage had nothing to do with the huge, continuous nine-year spike in out-of-wedlock birthrates after the passage of registered partnerships in 1997, and formal gay marriage in 2000. But the critics are wrong on several counts.
To begin with, it’s a mistake to pretend that only legal changes matter. If a campaign convinces nearly three quarters of the Dutch public that marriage has little to do with parenthood, this can easily have an effect on behavior well before any new laws get passed. But for the sake of argument, let’s say the two-percentage point rise in the Dutch out-of-wedlock birthrate in 1995 had absolutely nothing to do with the previous five-year-long debate over same-sex marriage. That still wouldn’t begin to explain away the continuous nine-year spike in out-of-wedlock birthrates following the passage of registered partnerships in 1997, or formal gay marriage in 2000. You can’t extrapolate from a rate spike in a single year to the longest continuous Western European rate spike of the last decade.
No matter what happened in 1995, there was nothing inevitable about the extraordinary and continuous nine-year acceleration in the Dutch out-of-wedlock birth-rate from 1997 to 2005. In 2002, for example, demographers David Coleman and Joop Garssen published, “The Netherlands: paradigm or exception in Western Europe’s demography?” (I discuss that article in “Going Dutch?”) Coleman and Garssen concentrated on data through 1998, so they were well aware of the small spike in the Dutch out-of-wedlock birthrate in 1995, and of the consecutive rises in 1997 and 1998. Although Coleman and Garssen knew that the Netherlands’ famously low out-of-wedlock birthrate was on the rise, they did not assume (nor should they have assumed) that the rate spike would continue for nine consecutive years. On the contrary, at the time, it still seemed as though the Netherlands represented a relatively moderate and stable alternative to the Scandinavian pattern of large-scale parental cohabitation. Coleman and Garssen were interested in the Dutch case because it still looked as though Holland was not going the way of Scandinavia. Yet in just a few short years, because of the ongoing and unexpected acceleration in the out-of-wedlock birthrate, that prognosis has changed. As the more recent report by Dutch demographer Jan Latten confirms, the Netherlands seems to be going the way of Scandinavia after all. (See “Trends in Cohabiting and Marriage” and my discussion of Latten’s report in “Standing Out.”)
So in the unlikely event that the highly successful campaign for same-sex marriage did nothing to shift Dutch views of marriage prior to the passage of registered partnerships, that would still leave same-sex marriage advocates with the problem of explaining the continuous nine year spike in the Dutch out-of-wedlock birthrate, beginning with the passage of registered partnerships in 1997, and continuing through the five years of experience with formal gay marriage between 2001 and 2005.
Recognizing this, Eskridge and Spedale make a perfunctory attempt to explain away the data. They suggest that the accelerating Dutch out-of-wedlock birthrate is being produced by the entry of women into the workforce, much as we saw in Scandinavia. Yet in “No Explanation,” I’ve shown that the work patterns of Dutch women are very different from the work practices of Scandinavian women. In marked contrast to Scandinavia, Holland is the land of the “mommy track,” where women work part-time, and where the childcare sector is still small and largely private.
Eskridge and Spedale further suggest that the spike in out-of-wedlock birthrates may be explained by the fact that Dutch registered partnerships are accessible to heterosexual as well as homosexual couples. Yet in “No Explanation,” I show that this is not the case. The number of registered heterosexual partnerships is too small to explain the huge surge in the Dutch out-of-wedlock birthrate. Nor do Eskridge and Spedale address my points on women’s work or registered partnerships.
Eskridge and Spedale also claim that, contrary to my own account of the debate, Dutch advocates put forward a “conservative case” for gay marriage. The official history of the movement for Dutch same-sex marriage (linked above) tells a different tale. Nor do Eskridge and Spedale address my detailed account of the Dutch legislative debate over same-sex marriage, in which this reform was treated by legislators on all sides as anything but conservative (see “Going Dutch?”).
So despite efforts to explain away the continuous nine-year spike in the Dutch out-of-wedlock birthrate since the passage of registered partnerships in 1997, and formal gay marriage in 2000 (an increase unmatched in Western Europe, and virtually unmatched even in Eastern Europe), the case of the Netherlands constitutes powerful evidence for the negative causal effect of same-sex marriage.
Back to Scandinavia
I’ve already responded to Eskridge and Spedale on the subject of Scandinavian marriage (see “No Nordic Bliss”). Yet the new Dutch data returns us to the question of Scandinavia. We’ve seen that the continuation of statistical trends cannot be taken for granted. Although out-of-wedlock birthrates bumped up a bit in the Netherlands in the mid-nineties, it was neither inevitable nor predictable that they would accelerate for so long. Nor was it inevitable that the decline of marriage in Scandinavia would continue after the advent of registered partnerships. Eskridge and Spedale dismiss post-registered-partnership increases in Scandinavian out-of-wedlock birthrates as the mere continuation of pre-existing trends. In contradiction to this, Eskridge and Spedale tout small resurgences in Scandinavian marriage rates, as if they prove that gay marriage strengthens marriage. So for Eskridge and Spedale, Scandinavian martial decline is inevitable…until it isn’t. In fact, post-registered- partnership increases in Swedish and Norwegian out-of-wedlock birthrates were not inevitable, and the slight recent resurgence in Scandinavian marriage rates has nothing to do with registered partnerships.
Talk about setting the bar high–Eskridge and Spedale say that the only acceptable proof of gay partnerships’ negative effect on Scandinavian marriage would be a 100 percent out-of-wedlock birthrate! That is an unserious claim, which merely shows that Eskridge and Spedale are ignoring key parts of my argument. I’ve said from the beginning that the ongoing deterioration of Scandinavian marriage has been partially offset and disguised (especially in Denmark) by “catching up” (delayed childbirth by older working women), and by remarriage among the large pool of divorced.
I’ve also argued that sharp initial accelerations in Scandinavian out-of-wedlock birthrates were bound to slow down. That’s because Scandinavians tend to use the first out-of-wedlock child as a test for a possible marriage. Only gradually are Scandinavians giving up on marriage after the birth of even the second child. Given factors like “catching up,” remarriage among the divorced, and the tendency to marry before the birth of the second child, there is no way I would expect to see an acceleration of Scandinavian out-of-wedlock birthrates to 100 percent in the wake of registered partnerships. The ongoing growth in Scandinavian out-of-wedlock birthrates, especially as they move through the “tougher” territory of second and third births, is itself quite significant, and can in no way be taken for granted.
Eskridge and Spedale would be more convincing if they acknowledged and addressed my points about catching up, remarriage among divorced, and the greater likelihood of marriage before the birth of the second child. Yet they ignore these points, and instead demand absurd proofs, like 100 percent out-of-wedlock birthrates.
Eskridge and Spedale also fail to address the regional comparison so critical to my treatment of Norway. Marriage has deteriorated far more markedly in Norway’s socially liberal, gay-marriage-accepting north than in its more conservative, religious south. Eskridge and Spedale effectively ignore this comparison, falsely characterizing my regional case as merely anecdotal. (For more on the regional issue in Norway, and on the birth-order issue, see “Unhealthy Half-Truths.”)
This brings us back to the Netherlands. The Netherlands is so important because it provides evidence for an actual acceleration of out-of-wedlock birthrates following the passage of same-sex unions–proof demanded by Eskridge and Spedale. We see this in the Netherlands because, in contrast to Scandinavia, out-of-wedlock birthrates were relatively low in Holland prior to the advent of registered partnerships and gay marriage. So in the Netherlands, we aren’t comparing apples and oranges (out-of-wedlock births for second-born children and above, as opposed to first-born children). In the Netherlands, we are still largely dealing with an increase in out-of-wedlock births for first children. And the Dutch out-of-wedlock birthrate has in fact accelerated dramatically following the introduction of legal same-sex unions.
All indications are that the Dutch case is a causal smoking gun for gay marriage’s negative effects. Although the matter is fair game for continued debate, no one has yet offered a convincing alternative explanation, or even fully confronted the arguments already on the table. At a bare minimum, the rapid and ongoing deterioration of Dutch marriage shows that the “conservative case” for same-sex marriage has been proven wrong in the Netherlands. Convince the public that marriage is not about parenthood, and increasingly parents simply stop getting married.