First published in Between the Lines, April 5, 2007
Opposition to homosexuality has long been marked by bad science. In the past, that usually meant bad psychology or even bad physiology. Today, the more common problem is bad social science, usually involving cherry-picked data about alarming social trends followed by breathtaking leaps of logic connecting these trends to same-sex marriage.
David Blankenhorn positions himself as an exception. In his new book The Future of Marriage, and in a recent Weekly Standard article entitled “Defining Marriage Down…Is No Way to Save It,” Blankenhorn makes the familiar argument that supporting same-sex marriage weakens marriage as a valuable social institution. But he claims to do so in way that avoids some of the simplistic analyses common in the debate, including those made by his conservative allies.
In particular, Blankenhorn criticizes Stanley Kurtz’s argument that same-sex marriage in the Netherlands and Scandinavia has caused the erosion of traditional marriage there. Blankenhorn rightly recognizes Kurtz’s causal claims to be unsupported: “Neither Kurtz nor anyone else can scientifically prove that allowing gay marriage causes the institution of marriage to get weaker,” Blankenhorn writes. “Correlation does not imply causation.” This is a refreshing concession.
But having made that concession, Blankenhorn proceeds as if it makes no difference: “Scholars and commentators have expended much effort trying in vain to wring proof of causation from the data, all the while ignoring the meaning of some simple correlations that the numbers do indubitably show.” But what can these correlations mean, if not that same-sex marriage is causally responsible for the alleged problems? What do the numbers “indubitably show”? Blankenhorn’s answer provides a textbook example of a circular argument:
Certain trends in values and attitudes tend to cluster with each other and with certain trends in behavior…The legal endorsement of gay marriage occurs where the belief prevails that marriage itself should be redefined as a private personal relationship. And all of these marriage-weakening attitudes and behaviors are linked. Around the world, the surveys show, these things go together.
In other words, what the correlations show is that these things are correlated. Not very helpful.
From there, Blankenhorn argues that if things “go together,” opposition to one is good reason for opposition to all. He attempts to illustrate by analogy:
“Find some teenagers who smoke, and you can confidently predict that they are more likely to drink than their nonsmoking peers. Why? Because teen smoking and drinking tend to hang together.” So if you oppose teenage drinking, you ought to oppose teenage smoking, because of the correlation between the two. In a similar way, if you oppose nonmarital cohabitation, single-parent parenting, or other “marriage-weakening behaviors,” you ought to oppose same-sex marriage, since they, too, “tend to hang together.”
This is breathtakingly bad logic. The analogy sounds initially plausible because teen drinking and teen smoking are both bad things. But the things that correlate with bad things are not necessarily bad. Find some teenagers who have tried cocaine, and you can confidently predict that they are more likely to have gone to top-notch public schools than their non-cocaine-using peers. It’s not because superior education causes cocaine use. It’s because cocaine is an expensive drug, and expensive drugs tend to show up in affluent communities, which tend to have better public schools than their poor counterparts. Yet it would be ridiculous to conclude that, if you oppose teen cocaine use, you ought to oppose top-notch public education.
The whole point of noting that “correlation does not equal cause” is to acknowledge that things that “tend to hang together” are not necessarily mutually reinforcing. They are sometimes both the result of third-party causes, and even more often the result of a complex web of causes that we haven’t quite figured out yet. In any case, when babies correlate with dirty bathwater, we don’t take that as a reason for throwing out babies.
Which brings me to another significant flaw in Blankenhorn’s analysis. Even if we grant that support for same-sex marriage correlates with negative factors such as higher divorce rates, it also seems to correlate with positive factors such as higher education, greater support for religious freedom, and greater respect for women’s rights. On Blankenhorn’s logic, we ought to oppose those things as well, since they “tend to hang together” with the negative trends.
I don’t often find myself agreeing with Stanley Kurtz. But at least he seems to understand that, without the causal connections, the “negative marriage trends” argument gets no traction.