For the sake of reducing the general level of snarkiness in the world, the pursuit of truth to its innermost thingys, and of course the children, I’ve looked a bit further at the question of May-December marriages and what that tells us about revealed preferences. As is often the case, the data tell us both more and less than you might think. The amateur demography continues below the fold, at Holbovian length.

You’ll remember that David Bernstein used the example of May-December weddings as an example of revealed preferences in the mate market:[1]

I do wonder whether the authers considered the revealed preferences that seem blatantly obvious to those of us who merely observe human behavior (and maybe even look at the statistics, e.g., on which sex the Mays and Decembers tend to be in May-December relationships).

His thought was that, whatever people may say, their actions will tend to reveal that men go for looks and women for financial success. Not all the time, of course, but that’s the way to bet them. It’s easy to imagine an economic model of the same behavior, and that’s what’s suggested by the “revealed preferences” idea. We shouldn’t just discount the idea of revealed preferences, of course. We use it to explain people’s actions all the time in everyday life. But we should also remember that sometimes people have much less in the way of choice than we’d like, sometimes they just act in ways that seemed like a good idea at the time, and other times they are subject to what Harrison White once called (in a review of James Coleman’s Rat-Choice masterpiece, Foundations of Social Theory) “the awful grip of chance in human affairs.” It’s very tempting to assume that whatever we see people doing must reflect their preferences because otherwise, given they are rational utility maximizers, they wouldn’t be doing it. It certainly makes research much easier. But the danger is obvious: you’ll tend to assume what you need to prove —- i.e., that people are rational utility maximizers -— or infer panglossian conclusions about everything being for the best (rational choice sociodicy) or as nature intended (evolutionary psychology biodicy).[2]

The picture David conjures up is the familiar one of the disgustingly rich, ugly old guy with the beautiful young wife. This is common enough—much moreso, as David says, than the pattern of older women marrying younger men. I’d hazard a guess that in the (few) cases where women have attained a real measure of wealth or power, they have little trouble finding younger people to sleep with them, if that’s what they want. You might be surprised, for example, at the number of Margaret Thatcher’s underlings who had the hots for her. But it’s certainly true that men are everywhere much more likely to be older than their spouses. How much older varies cross-nationally and also historically, but typically there’s about a 2-3 year gap when both partners are marrying for the first time. This is partly conditioned on the economic and social position of women. The more opportunities they have (in the labor market, for example) the more likely it is that we’ll see homogamy (similarity) on partners’ age and education. This raises a nice question for Revealed Preferencers. If freeing up the choice-set for women results in a change in their decisionmaking, then it is hard to see the prior choices by similar women as representing their true preferences. The RP approach works best if we assume the institutional environment is fixed and there are only minimal constraints on people. The opportunity to exercise free choice with respect to one’s marriage partners is a comparatively recent (and still somewhat restricted) practice. Just read Jane Austen. Saying that people’s decisions reflect the best choice available under the circumstances would be cold comfort for, say, all the Charlotte Lucases who choose to marry their Mr Collins.

A common conservative move at this point is to say that traditional institutions exist because they stabilize the social order, and that people who make unconventional choices will end unhappily, as will the societies that permit such choices. There is an EP twist on this, too: it is the ham-fisted social engineering efforts by politically-correct lefties that prevent the evolutionarily correct patterns (to coin a phrase) from doing their work. But this forgets that the institutions which created the dilemmas faced by Charlotte Lucas and Elizabeth Bennett—such as Entailment—were themselves historically specific bits of social structure, not inevitable facts of nature. In our own time, arguments that American girls weren’t cut out for (EP) or didn’t really want to (RP) play sports look rather threadbare given the explosive growth in participation in women’s sports following the passage of Title IX. That case reminds us that creating variety of choice by reducing constraints on particular decision-makers is not equivalent to just junking the institutional rules and letting people do whatever they want. Freedom of choice in marriage partners (or anything else) is itself a positive, social accomplishment. Contexts where choice is free—like proper markets—are possible only because many other things are highly constrained, such as the chance to freely cheat, renege or use violence in exchanges. Widespread, stable freedoms are not the same as the state of nature.

Back to marriages. Vera, Berardo and Berardo’s “Age Heterogamy in Marriage” (Journal of Marriage and the Family, 47, 1985: 553-566) presents evidence about marriages where the husband is significantly (11 more more years) older than his wife. They are interested in whether such marriages are concentrated amongst better-off households, and in whether these marriages are unhappier than usual. They find the first hypothesis “was not supported: age-discrepant unions are clearly more prevalent among lower classes.” Age-discrepant unions are also much more prevalent amongst blacks and Hispanics than whites. So heterogamous marriages are much more common amongst poorer groups, not richer ones! This should come as a surprise if our image of the typical May-December union is Donald Trump and whatever her name is. If you are a seasoned RP or (especially) EP thinker, you will no doubt even now be constructing a rationale to explain this pattern: Out there on the Savannah, life was hard and when the hunters were killed on the job, their widows became poor and had to marry the old guys who were left behind… As you please.

The paper’s second hypothesis is also of interest. May-December marriages are typically frowned on, often provoke disgust, and are usually thought to be less happy than marriages contracted for better reasons than money. But, in fact, it turns out that these marriages appear no less happy than average.[3] Socioeconomic position is the best predictor of happiness in marriage.

Bytheway’s “The Variation with Age of Age Differences in Marriage” (JMF 43, 1981: 923-927) discovers more subtle patterns of variation in age-differences within marriage:

This paper is based upon an analysis of all marriages in England and Wales in 1976. It shows that the pattern of variation with age of age differences in marriage is not simple but is characterized by a change of direction in middle age. In particular it is apparent that both sexes are more likely to marry an older person at the age of 50 than at the age of 30. The same pattern is found in a large number of other countries.

What to make of a pattern like that? Bytheway thinks that the answer is found in changes in the meaning of marriage with age. And there’s also the question of where the criteria for a meaningful marriage come from to begin with. There seem to be a lot of open questions here.

A relaxed assumption of Revealed Preferences won’t help us find the answers. Sloppy application of ideas from evolutionary psychology makes it even easier to just Make Shit Up (MSU). Because David Bernstein’s original post didn’t make any claims about Evolutionary Psychology, I was surprised to see the comments thread to my post quickly become home to a bunch of accusations that we here at CT are all blank-slater Marxists from the ’70s who just can’t face up to a supposedly emerging scientific consensus about the hard-wired nature of sex roles. That’s just rubbish. No-one around here is denying that our evolved biological nature plays some role in everything we do and a large role in some things we do. But the Devil is in the details. Popularizers like Steven Pinker are notorious for spending their time knocking down straw Blank-Slaters and then cheerfully MSU-ing an evolutionary rationale for the contemporary behavior of their choice. Here, for instance, is Jerry Fodor (no Blank-Slater he!) giving Pinker the treatment he deserves for this:

The literature of Psychological Darwinism is full of what appear to be fallacies of rationalisation: arguments where the evidence offered that an interest in Y is the motive for a creature’s behaviour is primarily that an interest in Y would rationalise the behaviour if it were the creature’s motive. Pinker’s book provides so many examples that one hardly knows where to start. … [H]ere he is on why we like to read fiction: ‘Fictional narratives supply us with a mental catalogue of the fatal conundrums we might face someday and the outcomes of strategies we could deploy in them. What are the options if I were to suspect that my uncle killed my father, took his position, and married my mother?’ Good question. Or what if it turns out that, having just used the ring that I got by kidnapping a dwarf to pay off the giants who built me my new castle, I should discover that it is the very ring that I need in order to continue to be immortal and rule the world? It’s important to think out the options betimes, because a thing like that could happen to anyone and you can never have too much insurance.

When pushed in this way, the better EP people emphasize much more circumscribed claims. They will point to very carefully designed experiments with interesting results, on the one hand, and the generalized promise of evolutionary thinking for constructive social theory in general. These are reasonable points. It’s the difference, to put it in terms of the popular literature, between Pinker’s The Language Instinct, where he makes an accessible case about a literature he knows very well, and How the Mind Works, which is larded with EP howlers like the one just cited. Sloppiness like that doesn’t do any justice to those in the field who are trying to make it work properly.

Relations between the sexes provide the most fertile soil for the proliferation of the MSU branch of EP. Gender roles are deeply institutionalized—that is, they are highly scripted and chronically reproduced—and we like nothing better than to think of our institutions as inevitable or natural.[4] I can see how very widespread trends—such as men being slightly older than women at first marriage, for instance—might be traced back to very ancient social arrangements, though even here there’s enough variation to make it a difficult sell. Neither am I opposed to the idea that there are very basic drives or predispositions that go back very far which might reliably generate patterns of social organization or culture. But it also seems obvious to me that ideas about the appropriate relations between the sexes—or races, classes, nations, or whatever you like—thrive best if they appear to be emanations from the mind of God or the structure of DNA. This has consequences on the input and output sides of EP arguments. On the output side, it’s no surprise that Sunday-Supplement versions of EP gender theory have been taken up by those who would benefit if things were organized the way those arguments suggest. On the input side, it’s striking how the natural gender division of labor theorized to have evolved once-and-for-all on the Savannah often bears a strong resemblance to middle-class life on the Eastern seaboard of the United States around 1960. The Darwinian mode of argument is very powerful, and it’s easy to go crazy with them when all you have to constrain you is the complete body of surviving evidence about how social groups were organized in the Environment of Evolutionary Adaptiveness—which is to say, bugger all.

Getting at the truth the relationship between biology, social organization and individual behavior is a lot harder than the likes of Steven Pinker make it seem. It’s clear that society as we know it wouldn’t exist if it weren’t for our evolved capacity for language. Or that things would be very different if, like many other mammals, bits of us started broadcasting “Time To Reproduce!” to the world every few weeks. Or that Presidential debates would be trickier to manage if women laid eggs and men had to incubate them for a year.[5] Patriarchy would be harder to organize if females were 25,000 times larger than the colony of tiny males that lived inside them. But as we move down this scale towards differences like “If kinship relations were everywhere fully matrilineal”, or “If females were allowed to own property and vote” or “If men were capable of accurately assessing how much housework they do”—well, clearly there must be a sweet-spot along the way where EP explanations are both non-trivial and supportable by the evidence. It’s just really hard to establish where that sweet-spot is. Taking that view doesn’t make you science-phobic, it makes you a better scientist.

Notes

fn1. It’s not really a market, of course. This is just a metaphor. But insofar as it is, it’s a mate market and not a marriage market. You can’t buy a marriage. Why would you even want two extra people in your house, anyway?

fn2. As Jon Elster complains, there are “inverted sociodicies”, too, where all is for the worst in the worst of all possible worlds. That’s a vice as well.

fn3. This need not be particularly happy.

fn4. To pick an article at random, see Karen Martin, “Becoming a Gendered Body: Practices of Preschools” (American Sociological Review 63, 1998: 494-511) for some data on a tiny piece of this effort, directed at little kids.

fn5. It would explain the bulge in Dubya’s suit, though.