New Study: Gay marriage in Holland may have reduced marriage rates among … liberals?

For decades, conservative opponents of same-sex marriage have been making a brazen claim. Extending marriage rights to gay couples, they say, will weaken the marriage institution.

In the near-absence of reliable data, their arguments have relied almost entirely on hypothesis. However, a new study, published earlier this year in the social science journal Demography, could offer same-sex marriage opponents some of the first empirical evidence in support of their theories. If the study’s findings are correct, same-sex marriage in the Netherlands decreased the opposite-sex marriage rate in all but the most conservative groups.

The Theory

It’s not unanimous, but growing majorities in developed nations are now on board with the idea that same-sex marriage is a good thing. As of 2013, around 80% of Scandinavians, 63% of Canadians and 55% of Britons support same-sex marriage, according to a global Ipsos survey. Recently, the tide of public opinion has even turned in the United States, where a recent poll has support as high as 59%.

For supporters, it’s not difficult to see why the public has come around. The philosophical and moral case for same-sex marriage is compelling, as this conservative will admit. Defenses can be made in the name of freedom (why should we force people who love each other not to marry?), equality (traditional marriage laws turn gays into second class citizens) and the public good (who would be harmed by gay couples marrying?).

Opponents of same-sex marriage have been caught flat-footed, perhaps even dumbfounded, as marriage laws move past them in ways that would have been inconceivable just thirty years ago. They are often unable to explain exactly why they oppose same-sex marriage but aver that it remains bad policy at best, and morally wrong at worst. The phenomenon of “moral dumbfounding”, or an inability to explain one’s moral intuitions, is not unique to same-sex marriage opponents. The most common example: most Westerners (but not all) rightly oppose contracepted sex between siblings but are unable to articulate why they do. Getting stumped, then, does not make a person wrong, but it does mean that they will need to dig deeper to justify their beliefs.

Some advocates for the traditional definition of matrimony say they have done just that, catching a subtler vision of the marriage institution and promoting a corollary new case against same-sex marriage laws. Sherif Girgis, PhD candidate in philosophy and principal co-author of the 2012 book What is Marriage?: Man and Woman: A Defense along with Robert P. George and Ryan T. Anderson, argues that there are two public visions of marriage. The “conjugal view” sees marriage as oriented toward procreation. The “revisionist view” calls marriage a union oriented toward love and commitment, with procreation and childrearing an elective option. (See here for a lengthier explanation.)

Girgis argues that only the former view explains why marriage should be a public institution, because it enforces norms that arise from sexual reproduction. The latter view, then, would make marriage functionally indistinguishable from non-marital romantic unions. He contends that in order to coherently support same-sex marriage, advocates have had to adopt for themselves and promote to the public the revisionist view. This, he thinks, constitutes a removal from the public square of the only institution geared toward procreation. The implication is that the old (to some, outdated) norms that historically sprung up around procreation, as well as the motive to marry, will fade along with the old view, more than they have already done.

Despite Girgis’ and his coauthors’ new defense of the traditional marriage definition and a small renaissance of secular opposition to same-sex marriage, many conservatives seem to have given up fighting in the face of a near-monolithic public consensus. And while Girgis’ reasoning is not unsound, there has been little hard evidence to support his broader conclusion that same-sex marriage laws would harm marriage itself.

What everyone, on both sides of the issue, should admit is that the same-sex marriage debate has taken place in an empirical vacuum. For better or worse, the conversation has been about same-sex marriage the moral question, not same-sex marriage the public policy. While no empirical study can or should settle the question on its own, everyone interested should take careful consideration of the data that do exist.

The new study

The paper in question, “The Effect of Same-Sex Marriage Laws on Different-Sex Marriage: Evidence From the Netherlands“, by Mircea Trandafir, was published in February of this year in the journal Demography. The author attempted to find the effect on the marriage rate of a 1998 domestic partnership law and 2001 same-sex marriage law in the Netherlands. Its conclusions have mostly escaped attention—partly because of a vague abstract.

The paper contains two statistical analyses. The first is a regression on aggregate (country-level) data that compares the Netherlands to a control group of OECD countries over a number of years, while the second is a time series analysis of individual-level data. In the abstract, the author writes that according to the first model, “neither law had significant effects on either the overall or different-sex marriage rate”. This is true at the 5% level, but it’s worth noting (in the context of the entire paper) that the effect of the same-sex marriage law on the marriage rate was significant at the 10% level—implying that there is only a 1 in 10 chance that Dutch marriage rates in the absence of the same-sex marriage law would have fallen as much as they did in reality. The overall trend from 1988 until 2005 is described in the paper:

As expected, the actual rates are relatively close to the synthetic marriage rate [control group rate] between 1988 and 1997, the period used to construct the synthetic control. After the introduction of registered partnership, the three rates are all higher than the synthetic marriage rate, but they all fall below the synthetic rate at some point after 2001, the year in which same-sex marriage was legalized.

According to this first, aggregate-level analysis, while there is not enough statistical power to conclude with certainty, it appears that marriage rates rose slightly as a result of the 1998 domestic partnership law but were depressed by the 2001 same-sex marriage law.

The second regression of the paper is perhaps more valuable—it uses individual-level data, which allows for greater analytic nuance. In the abstract, the author concluded: “The effects of the two laws are heterogeneous, with presumably more-liberal individuals (as defined by their residence or ethnicity) marrying less after passage of both laws and potentially more-conservative individuals marrying more after passage of each law.”

This finding is true, but represents only a part of the main results of the analysis—furthermore, important context is left out. “Potentially more-conservative individuals”, as defined by the author, represent less than 10% of the Dutch population, while “more liberal individuals” make up more than 80%.

Trandafir also claimed in his abstract that the results of the individual-level analysis “confirm the findings in the aggregate analysis”. It’s unclear what he means by this, since the individual level-analysis did not estimate the overall effect of the laws (only by gender), and only measured the overall marriage rate (as opposed to the different-sex marriage rate).

Trandafir made the decision not to perform a regression for men and women combined, because women tend to marry at a younger age than men. The findings of his analysis are that the same-sex marriage law had essentially zero effect on the male first marriage rate overall, but a statistically significant negative effect on the female first marriage rate (i.e. the age-specific rate of first marriages in a person’s lifetime).

But the story becomes more interesting. In the conservative Dutch Bible Belt, which represents around 4% of the Dutch population, the regional effect of the law was to strongly increase the marriage rate for both men and women. The same goes for the 3% of Dutch who are Turkish or Moroccan. But for the vast majority of the population—that is to say, for native Dutch and especially for residents of the four largest cities—the effect of marriage laws was significantly negative for both men and women.

The Dutch royal family in April 2013

The Dutch royal family in April 2013

The regression is not directly interpretable quantitatively because of the complexity of the model; but for comparison, the effect on marriage rates of the law for women in the four largest cities, for example, was three times greater than the effect of actually living in the four largest cities. This suggests that the adjusted effect is large, since the marriage rate in the four largest cities is around 20% lower than in Holland overall (this is not to suggest that the law had a negative 60% effect, because the raw statistics are not controlled for demographic variables). The effect on the marriage rate for native Dutch men was a rate twice that of the year-to-year downward trend.

Somehow, it appears that the demographic groups most supportive of same-sex marriage—urbanites and native Dutch—are precisely the groups whose marriage rates declined in connection with the law. On the other hand, the conservative subcultures who were likely to oppose the 2001 law–Bible Belt residents, Turks and Moroccans–seem to have experienced a marriage rate renaissance as a result of it. Somewhere, Sherif Girgis is wide-eyed; his theory that the revisionist view of marriage undermines marriage seems to fit the data almost uncannily.

But the fact that makes these findings even more shocking is a bias in the data that causes the negative effect of the law to be underestimated, as the author admits: “identifying the spouse of all individuals is virtually impossible, and I am unable to distinguish between same-sex marriages and different-sex marriages. This induces a small upward bias in the estimate of the different-sex marriage rate after 2001.”

In other words, the gay marriage law did not just reduce the opposite-sex marriage rate, but the marriage rate overall. More explicitly, since we know that no same-sex marriages occurred prior to 2001, this means that, apart from the small conservative minority, the same-sex marriage law was associated with a drop in opposite-sex marriages that was larger than the rise in same-sex marriages.

The implications of these findings are stark. If this study is correct–and for the robustness of the methodology, it seems unlikely to be far off–then it is appropriate to suggest that same-sex marriage had the effect of decreasing the mainstream marriage rate among heterosexuals in the Netherlands, possibly by changing the way marriage was perceived.

There is a more hopeful story for advocates of civil unions, or marriages-in-all-but-name. In both the aggregate and individual analyses, domestic partnership laws had a neutral or small positive impact on the marriage rate, suggesting that it is not gay unions that have the potential to disrupt marriage culture, but the redefinition of marriage itself.

Conclusion

Caveats are in order after such claims. First, cultural differences between the Netherlands and other countries mean that the results cannot necessarily be extrapolated. Indeed, in a 2009 state by state analysis in the US, the authors found that a state’s gay union policy had no significant effect on marriage rates (it should be noted that the study did not separate civil unions from same-sex marriage).

However, Girgis’ theory might well explain this particular discrepancy. American religious conservatives are not the tiny minority they are in Holland. In the Netherlands, there has long been massive majority support for same-sex marriage, which may help explain the significant effects found in the Trandafir paper. It could be the case that the more people support same-sex marriage, the greater downward effect the law has on marriage rates–a troubling policy conundrum.

The larger caveat: it bears reiteration that these findings do not themselves justify a particular position on same-sex marriage. There is much more to be said, and at the end of the day, public policy is not only an empirical question; it is a social, philosophical and moral one as well. However, science, philosophy and morality are often found to intersect in the same place once the dust kicked up in a whirlwind of controversy has settled. Time will tell what conclusion future generations will draw. When it comes to culture wars, history has been known to vindicate both winners and losers.

Photo credit: “King of the Netherlands” by Tom Jutte, licensed under CC BY-NC-ND 2.0

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s