Maggie Gallagher has said that civilizations that lose the marriage idea die out.
I doubt whether that’s ever literally happened. I’d actually kind of giggle at the prospect of it, but that would be unkind.
Still, good or bad things can still happen on the margin when traditional, procreative marriages are encouraged or discouraged: Raise the cost of marriage, and fewer people will get married. Make a less-procreative life more rewarding, and you will get more people living that life.
But if you do that, you’ll also get fewer babies.
This plausibly constitutes a harm, at least in the abstract: If Jonathan V. Last is right – which he could be – Americans’ failure to have enough kids spells demographic disaster in the near future. And every extra baby will mitigate that disaster, at least a little bit.
In this post, I’m going to try to outline how we might estimate one alleged demographic harm of same-sex marriage, namely the excess births that our society will forego if same-sex marriage is universally recognized.
I will not consider any of the downsides of prohibiting same-sex marriage. Those downsides are real; I do not deny them. But this post will be complex, shoddy, and incomplete enough as it is. I don’t have all the answers, and I welcome your help in the comments.
To clarify, I propose the following two interrelated questions:
How many people, P, have gone from having an average of M kids to having an average of N kids as a result of banning same-sex marriage?
How big are M and N?
Once we know P, M, and N, we have a very simple formula for number of surplus births, S, that come from prohibiting same-sex marriage and routing people into heterosexual marriage:
S = P*N – P*M
Let’s say a bit about each of the variables.
We call population P the persuadables. These people would get married heterosexually, and they would (arguably) enjoy heterosexual-married lifetime fertility rates, but only if same-sex marriage were foreclosed to them.
We should note several things about the persuadables:
1. The persuadables must exclude anyone who would never marry heterosexually, not even if same-sex marriage were forbidden. My great-uncle lived in upstate New York in the 1950s. He wasn’t the marrying type, as they used to say. if he ever married anyone, it would have to be the other gentleman he spent so much of his reclusive life with. He’s not a persuadable. Neither am I, for that matter.
2. The persuadables must exclude all those bisexuals who would have married heterosexually anyway, and who would therefore have enjoyed similar lifetime fertility rates: the law isn’t changing their behavior.
3. The persuadables almost certainly must consist entirely of people whose orientation is less than 100% heterosexual. The existence of same-sex marriage isn’t going to entice anyone who is 100% heterosexual, and I think it is absurd that any 100% heterosexuals will change their marriage status, or procreational behavior within marriage, as a result either of same-sex marriage’s existence, or of what the persuadables are doing within same-sex marriages.
Some people may disagree, and I welcome their input in the comments. Note that this means some chunking out of P, M, and N, but that’s not by any means intractable as a problem. (Note that heterosexuals can and do change their marriage or procreational behavior as a result of other trends in society (income effects, no-fault divorce, and the birth control pill all come to mind), but these are all matters that touch their everyday lives in a way that same-sex marriage doesn’t so clearly do.)
4. I think that a small but non-negligible number of 100% homosexuals may marry heterosexually in some regimes where same-sex marriage is forbidden, particularly if the life option of being openly gay and unmarried is met with significant social stigma. Some persuadables, then, may be entirely gay, but they’re still persuadable, in our sense. (Will they enjoy heterosexual-married lifetime fertility rates? A difficult question.)
How many persuadables are there, given the conditions likely to prevail in the United States? How many people will get heterosexually married, but only if the law forbids same-sex marriage? And what are their fertility rates in either case?
I’m curious how we might put numbers to all of these factors, but my gut tells me that P is relatively small, and that, while there is a relatively large difference between M and N, P being small means that there won’t be very many excess births as a benefit of prohibiting same-sex marriage.
I welcome attempts to estimate P, M, and N in the comments, but I have to admit that I have found the problem nearly intractable. Recall that we want lifetime fertility rates for people who marry heterosexually and for people who marry homosexually. (I can only presume that the prospect of marrying has some impact on out of wedlock births. I can’t say exactly what it is, but if I can get these numbers, I don’t have to.)
This is all very difficult, except for N. Alone among the variables, N is (I think) easy: In 2003, the fertility rate among married women was 88.1 births per 1,000 women per year, ages 15-44. Because 2003 saw no same-sex marriages apart from those performed in Canada, Belgium, and the Netherlands, any influence they may have had on fertility rates would be virtually nil.
In 2003, essentially all persuadables were behaving as if same-sex marriage were prohibited. After all, it was. Assuming that they enjoyed fertility rates akin to the general run of heterosexuals, we have our N. Or perhaps something a bit higher than N.
The rest? It’s largely a lot of guesswork, as far as I can tell. Can anyone help?
 If I may say so myself, I am raising an awesome four-year-old daughter. She loves Star Wars, Lego, and My Little Pony. She already reads as well as many first graders. She loves spinach and broccoli. Her favorite composer is Prokofiev. But while I’m raising one awesome daughter, why isn’t it two awesome daughters? Or five or six? And why did I adopt, rather than personally creating all these awesome human beings?
I suppose if I felt sufficiently guilty about it I could always try to become a sperm donor. But various unmentioned genetic defects suggest that I shouldn’t go this route. I wouldn’t wish my genetics on anyone. And anyway, what I like about my daughter isn’t necessarily her genes either. She seems healthier in general than I am, but I don’t value health per se, but only as a means to the Prokofiev-and-Lego cultural superstructure.
 In short, I’m only considering social harms. I’m not going to get into Bryan Caplan’s utilitarian argument for more kids. Bryan believes that in hedonic terms, it’s vastly better to be alive than never to be born. I think he makes a category error to compare the two, but if he’s right, then gays and lesbians are hardly the only derelicts to duty. Everyone should be having way, way, way more kids. And perhaps we gays and lesbians could then specialize our parenting labor — and adopt the kids who, in the process, would end up in otherwise difficult homes. It may be our greatest marginal benefit, if everyone else is way out at the fertility production frontier.
 This suggests the existence of a second population of persuadables, P2. P2 folks would take a gay marriage if they could get one, but they would never take a straight marriage for any reason. P2s in gay marriages may become more likely to procreate than P2s who can’t get married: Lesbians may become marginally more likely to seek artificial insemination; gay men may become marginally more likely to have surrogacies. Likelihood of adopting a kid may be a social benefit, but it’s not the benefit we’re looking for here.