Saturday, May 28, 2011

No Assortative Mating?

Here's the abstract to a paper, posted on Barking up the wrong tree, which outlines a study that shows that whom people date and, presumably, eventually marry, is much more influenced by availability than by preferences (via Roissy):


Marriage data show a strong degree of positive assortative mating along a variety of attributes. But since marriage is an equilibrium outcome, it is unclear whether positive sorting is the result of preferences rather than opportunities. We assess the relative importance of preferences and opportunities in dating behaviour, using unique data from a large commercial speed dating agency. While the speed dating design gives us a direct observation of individual preferences, the random allocation of participants across events generates an exogenous source of variation in opportunities and allows us to identify the role of opportunities separately from that of preferences. We find that both women and men equally value physical attributes, such as age and weight, and that there is positive sorting along age, height, and education. The role of individual preferences, however, is outplayed by that of opportunities. Along some attributes (such as occupation, height and smoking) opportunities explain almost all the estimated variation in demand. Along other attributes (such as age), the role of preferences is more substantial, but never dominant. Despite this, preferences have a part when we observe a match, i.e., when two individuals propose to one another.
Source: "Can Anyone Be “The” One? Evidence on Mate Selection from Speed Dating" from IZA Discussion Papers, number 2377.

If this study is more generally applicable, then assortative mating is a much weaker phenomenon than heretofore thought. Another excerpt on the blog quotes the author of the study, who says that preferences account for only 2% of mate selection, with market availability accounting for the other 98%. In other words, we date and eventually marry from those who are around us, not necessarily those with whom we have a lot in common.

Any given social class or ethnic group will be more likely to live and work among people similar to themselves, so availability will be greater. Knowledge workers will work among other knowledge workers, blue collar among blue collar, which would account for the apparent illusion of assortative mating.

Herrnstein and Murray's The Bell Curve predicted that IQ stratification would become greater as the society becomes more meritocratic and social mobility increases. It would seem that this would also cause the smart and ambitious to be together more and hence they will marry each other. Because of the heritability of intelligence, this could produce an elite social class that, because of proximity, will tend to perpetuate itself. Perhaps this result isn't any different than if preferences rather than availability were the cause.

The most interesting result of this paper is that a phenomenon widely assumed to be an important explanation of trends in society appears barely to exist.

Update: hbd chick calls total b.s.

27 comments:

  1. This post sounds like a good argument for freedom of association at the very minimum (the power for a business or individuals to legally discriminate based on race or other attributes in all non-governmental entities) or perhaps even Jim Crow. By the way, I am for freedom of association, not Jim Crow.

    ReplyDelete
  2. If there is no one around you with your preferred profile in a mate, of course you would date whoever is available. But if someone of your preferred type is there, then you would be more likely to date them. How exactly does this study show that individual preferences aren't the main criteria for who you date if those people are available?

    ReplyDelete
  3. I think you have misunderstood what Murray and Hernstein (M&H) were saying, and there is one other factor.

    I believe that M&H were saying that formerly, ie, before the great upsurge in people going to college, and especially before so many females went, most people found spouses/partners from the communities they lived in.

    Now, with so many people spending time in college during the years when they would be selecting partners, there are now many more opportunities for people to select from among people more like them intellectually.

    Sure, when it comes down to the actual decision, I am sure that availability is very important, but then, these days there are many pretty intelligent people at college.

    Finally, for the other factor, assortative mating has always been practised by the upper classes. They rarely, if ever, marry down.

    ReplyDelete
  4. From the Abstract: "We assess the relative importance of preferences and opportunities in dating behaviour, using unique data from a large commercial speed dating agency. While the speed dating design gives us a direct observation of individual preferences, the random allocation of participants across events generates an exogenous source of variation in opportunities and allows us to identify the role of opportunities separately from that of preferences. We find that both women and men equally value physical attributes, such as age and weight, and that there is positive sorting along age, height, and education."

    Key sentence: using unique data from a large commercial speed dating agency.

    Key concept: range restriction.

    Modification: "We find that [among a population of people who use speed dating] both women and men equally value physical attributes, such as age and weight, and that there is positive sorting along age, height, and education."

    I wondered whether the authors know much about the large amount of good work on assortative mating - done by people who understand biology and psychology?

    Lets take a look at the reference list:

    http://ideas.repec.org/p/cpr/ceprdp/5926.html

    Nope - not one single ev psych reference.

    My impression: This is just a bunch of dumb economists crunching numbers. I don't suppose they even understand what assortative mating actually is...

    ReplyDelete
  5. i call b.s.

    bruce is exactly right -- the speed daters in this study were already self-sorted. (hint: they don't resemble the general population at all.)

    ReplyDelete
  6. This is just a bunch of dumb economists crunching numbers. I don't suppose they even understand what assortative mating actually is...

    Exactly right. Anyone who gets results disproving assortative mating has to think thrice before publishing them. Clearly the authors could not be bothered with thinking or familiarizing themselves with existing literature (or an outside real world, in this case).

    ReplyDelete
  7. I'm not sure that Murray etc. are correct in asserting a self-perpetuating genetic aristocracy.

    For example, AA pushes a lot of folks (Herman Cain, Obama) into senior positions where they have no ability to function. Yet remain part of the aristocracy, the elites. Further, regression to the mean implies a falling of cognitive ability in an elite (and ultimately in-marrying Hapsburg like genetic defects) if there is not constant selective pressure on the elites (make dumb procreation choices and your line dies out).

    To put it this way, if you work for NPR as an exec, your kids are likely to work for NBC or CBS or CNN, as long as they utterly conform. IQ, drive, ability, and other things never come into play. They'll marry someone whose parents works for the Times, or WSJ, or works at the State Dept. And so on. Its just a self-perpetuating aristocracy indistinguishable from that of say, the Hapsburgs.

    ReplyDelete
  8. To put it another way, without selection pressures on the elites, their destiny is somewhat that of Ferdinand the First of Austria. Severely retarded, deformed, subject to massive seizures, his commands were "I am the Emperor, and I want noodles!"

    ReplyDelete
  9. Lets get real, the market for mates is a market for reproduction. Intelligence (particularly among women, less so among men), physical fitness, symmetry, and so on make those qualities the most important. As well as compatible immune systems, and so on (bad for the baby to die because of incompatible immune systems). As elites become self-perpetuating and not G-loaded, meritocratic, (i.e. cronyistic, not by examination) the temptation to improve the gene pool by judicious marrying out is likely to increase.

    The only problem is finding a method of introduction, and so on, which the various online dating services purport to offer.

    ReplyDelete
  10. Being an economist of sorts, I readily admit the primary fault of many economist studies trying to show something supposedly surprising. First, let’s take two extremes from the male perspective. First, for example, think of the first Spanish conquistadors in Central and South America. Those that stayed in certain recently conquered areas there was no opportunity to breed with non-Indian women (i.e., there were no other women, i.e., until wealthier men were able to bring in brides from Spain). Therefore, in that case we have 0% assortative mating, because it wasn’t even possible to have an alternative. Now let’s take cases where you have an almost unlimited choice. Imagine you are a wealthy, good looking, and famous athlete. In that case you tend to pick light skinned, blond, blue eyed women. Therefore, if you are blond and blue eyed male you have 100% assortative mating, otherwise 0% (i.e., based on our small set of narrow criteria – hair and eye color especially). Thus, either way the design will skew results toward finding 0% assortative mating, even when the opportunity set is apparently unconstrained and choice dominates. Clearly in most circumstances it is a combination of opportunity and choice (especially, for the average case, female choice). My guess is this study had a result in mind and pushed the design in that direction and even surprised themselves at the apparent lopsided result. Be wary, very wary of economists of the “Freakonomics” variety bearing “surprising” results. In this case I suspect that this result is largely driven by design, not by a reasonable experiment design.

    ReplyDelete
  11. "In that case you tend to pick light skinned, blond, blue eyed women." I'm sure that's why so many Europeans secretly fancy German women rather than Italian.

    ReplyDelete
  12. "Lets get real, the market for mates is a market for reproduction."

    You are projecting easterner's conception of mating onto western folks. In the eastern cultures (including Judaics), marriage has only two purposes: 1) promote endogamy to enhance tribalistic/ethnic ties and 2) to reproduce. Such cultures scorn romantic love and fear the havoc it could cause if romance were to supersede culturally mandated reasons for marriage.

    In the Western world on the other hand, there is a deep and long tradition of courtly love, which is seen as the primary reason for marriage as opposed to a side effect of it.

    ReplyDelete
  13. anon said:

    In the Western world on the other hand, there is a deep and long tradition of courtly love, which is seen as the primary reason for marriage as opposed to a side effect of it.


    I guess you have just proven that a good percentage of anonymous posters are ignorant.

    In the Western world, women, especially upper-middle class women, gauge reproduction partners by how well they can follow the cultural norms and mask their intentions.

    Read things like "The Mating Mind" to get a clue.

    ReplyDelete
  14. Regardless of the value of this particular study, it does raise the question of whether men are picking their mates based on what they truly prefer or merely on what is available to them. You have to actually meet the women you date and marry before you date and marry them, and higher IQ people can be pretty isolated from the rest of the population these days.

    This doesn't mean that a more intelligent man would prefer to date a total bimbo, but there is still the open question of whether he would trade some intelligence for some other quality, particularly looks.

    ReplyDelete
  15. anon said,

    "In the Western world on the other hand, there is a deep and long tradition of courtly love"

    beyond anon,

    "In the Western world, women, especially upper-middle class women, gauge reproduction partners by how well they can follow the cultural norms and mask their intentions."

    Surplus and non-surplus.

    If a group of people evolve over a long period of time in a subsistence environment where only monogamy was viable i'd assume that could lead to various side-effects. Some of the things that seem plausible are
    - women being more valued
    - selection on physical looks
    - selection on depth of affection

    If that group then move into a different environment i.e. they conquer an area with a surplus and become the elite, then the selection pressures would start to operate in a different direction.

    I would have thought it would still take a long time before there was no longer a difference (on average) between that group and people who had never had the initial selection pressures.

    .
    "women, especially upper-middle class women"

    In a way, resistance to love, would be highly selective for women in traditional societies where there is enough of a surplus for an elite to form. Wannabe gold-diggers without that resistance will fall for their plumber or accountant boyfriends before they get the chance to claw their way to the top.

    .

    ReplyDelete
  16. Some of the things that seem plausible are
    - women being more valued
    - selection on physical looks
    - selection on depth of affection


    Apropos of Wandrin's first and third suggestions, Jew Among You has a post on this very subject (with a link to a video).

    Rapists have been cast out of Norwegian society for a thousand years, and coincidentally, Norwegians never rape anybody. All hail the psychic unity of mankind!

    A Norwegian's idea of being nasty to a woman, in the 19th Century, was not letting her help plan the family budget.

    ReplyDelete
  17. Thank god Yan Shen is in Jail. I was getting tired of him spamming this blog with his nonsense.

    Lets hope the same happens to Whiskey.

    ReplyDelete
  18. In the Western world on the other hand, there is a deep and long tradition of courtly love, which is seen as the primary reason for marriage as opposed to a side effect of it.

    This is the single most important and probably the biggest difference between White European cultures and everyone else.

    This is might also be the reason why Jewish psychoanalysts like Freud, or Jewish feminist political activists, etc, all misinterpreted so many things about relationships between men and women in the West. I bet they couldn't really grasp it. For example, Freud, who was coming from a Jewish background, had probably never experienced love as defined by White European standards. To him marriage was about reproducing, and he couldn't help but to diagnose everyone around him with neurosis, where in reality they were simply suffering from side-effects of love or lack of it.

    ReplyDelete
  19. haha

    I doubt that is the same Yan Shen, though it would be funny if it was.

    ReplyDelete
  20. "In the Western world on the other hand, there is a deep and long tradition of courtly love, which is seen as the primary reason for marriage as opposed to a side effect of it."

    Correction: courtly love was (at least originally) adulterous. Think Launcelot and Guenevire. See The Allegory of Love, by C.S. Lewis.

    *

    Re 'assortative mating' versus availability.

    People choose availability as well as choosing from the available. Where you go looking for a partner is, of course, itself a choice - even before you choose among the people.

    If men seek their partners from women at work, do they choose them from among co-workers of similar status; or from the cleaners?

    If they are at college, do they seek partners among undergraduates; or from among the staff who serve their food? But what about the women in the factory next door?

    In other words, most assortative mating happens at the level of big and general choices of where is a suitable place to seek partners.

    In other words, partners are usually selected from the broad strata of social class.

    Also, partners are selected from whom is available after the most desirable men and women have paired-off.

    So, first the high status men and beuatiful women pair off, then the moderates, and so on down - with choice narrowing as it goes.

    Also, (within gene pools, although not between them) desirable traits tend to correlate, at a broad level - status, beauty, wealth, athleticism, health and so on all cluster (loosely) together - providing significant correlations.

    But of course there are numerous individual exceptions to these general observations. The statistical correlations may be significant, but they are modest.

    ReplyDelete
  21. In case anyone was wondering why Yan Shen has not been commenting lately.

    http://www.reuters.com/article/2011/05/27/banker-gun-idUSL3E7GR0UW20110527

    ReplyDelete
  22. Dregs says:

    In case anyone was wondering why Yan Shen has not been commenting lately.


    Bit late to the party ...

    Even if it is not the same Yan Shen, it is interesting that a supposedly highly intelligent Chinese guy (a hedge fund sort of guy) would take a gun to the airport ... maybe he has been watching too many Hong Kong gangster movies.

    ReplyDelete
  23. bgc says:

    So, first the high status men and beuatiful women pair off, then the moderates, and so on down - with choice narrowing as it goes.


    This is correct, with one addition interesting point, IMO.

    Usually it is women who do the choosing, except for the very highest status me. The top one or two percent of men (perhaps more) choose from among the available women and have the power to choose more than one woman ...

    ReplyDelete
  24. "I doubt that is the same Yan Shen, though it would be funny if it was."

    Aye, fun indeed.

    I highly doubt it, though, since Yannie thinks like I did at 14.

    The preponderance of the evidence is -- he's a 14-year-old, pimply, scrawny, smart-aleck kid.

    ReplyDelete
  25. People choose availability as well as choosing from the available. Where you go looking for a partner is, of course, itself a choice - even before you choose among the people.

    This is where you go wrong. I don't think men ever choose their social class/social circle because they'd like to meet the women that are in it. They choose their social class/social circle for other reasons (money, interest, status) and often have to take what they can get matewise.

    ReplyDelete

Please post a civil and intelligent comment, preferably using a screen name other than "anonymous". Comments are not currently moderated and their publication does not imply the agreement of the blog's author.