Age at Marriage and Happiness

The title of this post is a bit misleading because there is no actual data that I can find answering the question, "Is happiness in marriage related to age at marriage?". But, you might object, every now and then there's an article in a young, liberal publication that tries to get readers to forward it to each other in anger by touching a nerve about age of heterosexual marriage; here's a recent one that seems to be doing its best to piss of their usual audience by contradicting their belief that assumes young marriage = uneducated, traditional, unhappy marriage. News flash: in the U.S., age of marriage has been going up, and every year we set a new record. (As I write this it's men 29, women 27. Note also that this article focuses on heterosexual marriage. Data from gay marriages as compared to straight will eventually unmask a lot of the patterns in how this institution works, but that's not what this post is about.)

The perennial question is whether later marriages are better ones. It's worth asking whether the answer matters, since it's not clear that we're really "deciding" to marry later and really deliberating the pros and cons, rather than just a being products of more complex professions requiring longer education. And the way the question is discussed further clouds the issue, because the outcome these debates focus on is divorce - when divorce is only a proxy for happiness, and probably not a good one.

I've seen data that the older you are when you're married, the less chance of divorce - and also, conflicting data that suggest there is no relationship at all. For the moment let's make the assumption (subject to revision with further data of course) that the consensus is correct, and the older you are, the less the risk of divorce. Is this necessarily because the marriage is better, or just because you're older, and you have fewer mate-finding options once the marriage ends? If you're on the "later marriage is better" side, and you think that because later marriages are (probably) less likely to end in divorce, the answer must obviously be that later marriages are happier - then what about arranged marriages? They seem to have a far lower rate of divorce, although again it's hard to make a direct comparison not confounded by other factors. For what it's worth, globally, arranged marriages have a 6% divorced rate. The United States has a 54% divorce rate (the vast majority of which marriages are not arranged). So are these more "successful" arranged marriages lasting because they're happier? Or because there's pressure forcing two miserable people to stay together to avoid greater social and familial censure? (That would equate to a much thicker "marriage cushion" as referenced here.) In any event, it's fairly easy to find divorce statistics for first marriage at age X, but I wasn't able to find anything (recent) for happiness or satisfaction in marriage with first marriage at age X - that would have to be collected a lot more actively. Until some awesome psychologist does such a study we'll have to keep using divorce as a proxy indicator.

A separate concern is that by making informed decisions to maximize personal happiness, we're actually working against our own reproduction. That is to say, it's possible that by thinking about what makes us happy and making good decisions with our frontal lobes to pursue that, we don't get married and we don't reproduce, at least as effectively. (This is in contrast with the way we've reproduced up until basically the past century: from a very small pool of potential mates with very poor information and the females more or less forced into the deal.) That having the freedom and ability to actual ponder our options could even be stated as a concern admittedly cuts at the foundation of the Enlightenment, but there are strange realities: for example, that having kids makes you less happy. Well think about it, of course it does! So if everyone is making decisions to maximize happiness, why would we last beyond this generation? (Answer: we wouldn't.) And what do we see in liberal democracies in the West? A decreasing birth rate - especially where women have an increasing say in work and education and therefore how the family grows and runs. (Although this is probably also partly a result of lower average fertility of later marriages.) And marriage can be hard too - so might not the same argument apply to delaying marriage, that people are better informed and able to make more rational decisions to maximize their happiness?



From the Onion of course.
It turns out that people are more satisfied on average after they get married than if they don't (or at least Germans are, per Zimmerman and Easterlin 2006), so happiness-maximizers won't avoid marriage completely - although men seem to be the winners in terms of both income and life span in the deal, whereas women with large age differences relative to their husbands can actually lose life span. So if anyone would be rational to avoid marriage altogether, it would be women. Of course if we made only decisions that made us happy in the near-term even when they stopped us from reproducing, we wouldn't be here, which is exactly my point - inherited biological imperatives in humans lead us to make choices that are not predicted by maximizing utility, but when we are actually rationally maximizing utility, we get into uncharted territory.

The "Marriage Light" From Sex and the City

The characters from Sex and the City once discussed the "marriage light": their observation that men inexplicably seem to marry lower-quality women than the ones they dated, and this is best explained by something internal to the man's decision-making process rather than any observable characteristics of the woman who've dated him. It's as if, like a taxi, his marriage light goes on, and he walks down the aisle with the next one.

It's possible that it's not really a marriage light, but that there is something completely internal to the man's decision process that influences it; what looks like a marriage light is actually an interaction between the women he's dating, and this mysterious internal state. A quick-and-dirty explanation might be as follows. (Of course it generalizes about both genders, but about the calculation that men are making, based on what we seem to be maximizing; woman are free to chime in with their guess as to the variables underlying their own decision process.) Heterosexual men value sex with multiple partners, but they also want stability and children. Most women will not remain in a relationship with a man who has sex with other partners. At the same time, men do want a high-quality mate when they do commit. Therefore, to a man, the decision to marry is a trade-off: he gains a stable mate who will bear and help raise his children, but he will lose sex with multiple partners. He must balance the quality of the current sexual partner on one hand, against the prospect of number and quality of future sexual partners. As he gets older, and that future gets shorter, and the potential mates decrease in number and quality, his commitment threshold will drop. His current sexual partners' quality will be somewhat stochastic, but his own threshold will change (probably decrease) much more predictably over time. This is how a low-quality mate at 35 can receive a marriage proposal when the same or even higher-quality mate would not have received one at 25.

Graphically, it would look like this, with marriage occuring where the two functions are equal:

The straight line is his more-or-less predictable commitment threshold. At the start, Angelina Jolie could fling herself at him but he would hold out because his future is wide open. At the end, he knows time is running out. (It can be more complicated than this; more later.)

The curvy line is the women who've actually been available and mutually interested in him. Maybe the first peak is his senior year in college when he's big man on campus; then suddenly he graduates and drifts for a little while. The second peak came when he moved to a new city and got a real job. Then, another trough, as he found that he hated his job, got run down and grouchy and generally not pleasant. Finally he hits his personal global maximum. He has a new job that he loves, he finally figured out how to dress like a grown-up, he's a gym-rat in better shape than he was in college, and he meets a fantastic girl who, friends whisper, might be The One. But he's still looking for something more, and Seinfeld-like he focuses on her minor flaws. He lets her get away, focuses on the new business. Next thing he knows, he's 39 and overweight, and the game of romantic musical chairs that took place while he was working has left him with few choices. At his wedding, his female friends wonder what he sees in this one versus the ex from his global max period, who hiked across the Kamchatka Peninsula with him and was a well-known local artist and was beautiful. But hey - his friends are nice so they avoid mentioning her to his new wife. Sound familiar?

It's likely that divorce is not merely the reverse of commitment. Men tell their wives they'd marry them all over again. It's doubtful that this is always true. The social pressure and administrative and emotional baggage that comes with marriage may have the effect of lowering the threshold to which the wife's quality has to drop to reverse the commitment. If she drops below the thick black line, he wouldn't have married her in the first place but he won't divorce her. But if she drops below the gray line it's over. In some cultures with arranged marriages and low divorce rates, the gap between the thick black line and "marriage cushion" line might be quite large because of family interrelationships.

While I've written this as a heterosexual male, from the standpoint of a heterosexual male, it's interesting that professional women in large cities in the developed world have become more explicit about having their own commitment threshold curves. (In contrast, the idea of a commitment threshold is meaningless for women in pre-demographic-transition cultures that force them to sit on the sidelines pining to be chosen, their value almost entirely determined by qualities they have no control over developing, usually fertility indicators.) While the commitment threshold in women is also determined by age and quality, differences remain; namely, the importance that members of each gender assign on average to earning power, sex, and fertility in potential partners.

There are variations on the shape of the commitment threshold line for men, as follow:

- It might not start out at maximum as I have depicted here. That is, an 18 year-old might actually settle down with Angelina Jolie because he might not really believe that he does have future prospects; he doesn't know what else is out there. As he gets a little older, and so does she, he gets confident, and he realizes he would actually have future prospects, were he single; his newly-positive commitment threshold curve climbs far enough above his wife's quality that even the social-pressure cushion of marriage can't close the gap, and he dumps her.

- If a man's future prospects change dramatically - he gets in shape, he gets a great job - then if his commitment threshold curve is a reflection of his future prospects, it will jump as well. (The American philosopher Christopher Rock reminds us that a man is only as faithful as his options.) On the other hand, if the commitment curve were at some constant set point, we should expect that (for example) an actor in a surprise hit film who can suddenly date a much higher quality woman would marry the first high quality woman right away. This is not what happens.

- The commitment threshold line might not reflect reality. If the man's estimation of future amount and quality of sex partners is unrealistically high (unbelievable, I know!) he will inexplicably hold out even though his current partner might be, as his exasperated friends try to convince him, "the best thing that ever happened to you". In my experience, this is resolved in one of two ways: the male remains uncommitted for life to avoid the injury to his inflated self-image, or the slope of his commitment threshold gets very steeply negative later in life as his estimate of his prospects comes crashing back to Earth. (A game theory note: if all males in a dating community collude to hold out, either from unrealistic self-assessment or another reason, females who are interested in an earlier marriage will have to bargain more aggressively. Kate Bolick's outstanding Atlantic article might make one think that men in New York are doing exactly this.)


[The sort-of related figure above added later, credit Reddit user u/Smart_Ass_Pawn. I'm not going to bother with data on on how old Depp was when he was dating these people, but if you assume a) he has always had pick of whoever he wants, i.e. the only constraint is his preference, and that b) his preference in women has remained constant, including their age REGARDLESS of his age, this figure is interesting, and just looking at the two extremes, he divorced his first wife when he was 23 and she was 30, and most recent one when he was 54 and she was 31. Tina and Amy illustrate this pretty well here.]