Assortative mating in Bolivia

It is potent:

If people married each other more randomly, poverty levels would be considerably lower than they are now.  If we abandoned all current family arrangements and randomly grouped all Bolivians into new families of 5 persons, poverty levels would fall by about 15 percentage points (from the current level of 55% of all households to about 40% of all households).  The Gini coefficient measuring inequality would also fall from about 0.70 to 0.55.

But Bolivians do not mix much in marriage.  The correlation between partners’ education levels is extremely high at about 0.77, with no signs of falling.  For comparison, the corresponding number for Germany is 0.52 and for Britain it is 0.41.

But not all Bolivians are equally restricted in their marriage choices.  In the department of Santa Cruz the correlation is only 0.69 while in Potosi it is 0.82, with a corresponding difference in poverty rates.

That is from Lykke E. Andersen, Development from Within, an interesting and well-written collection of essays on Bolivian development, and sometimes on development policy more generally.  The cited piece was written in 2008.

Here is a good sentence from that book:

Just one little road block can disrupt an entire vacation.

Here is the author on Twitter.  Here is her blog.


So, redistribution works?

Define "works"?

Education levels are misleading. I know many women smart enough for graduate school who opted to raise their children instead.

My observation is people marry within a few IQ points of each other, and I have read (I don't know) of studies that say +5/-5 IQ points. Beyond that range there is not a lot of common ground for a marriage.

Lots of people are smart enough for graduate school, but don't do it. Graduate school is no more difficult than lots of private-sector work. In Bolivia, where graduate school is surely very rare, education levels more likely reflect access to upper-secondary and tertiary education, which is far more likely to depend on both access and intelligence.

anti-gnostic -possibly, that plus 5 minus 5 is only true if you use the better fit of M or V between the male and female, and is not true at all where the male is below 90 and above 125, or the female below 95 or above 120. Give or take half a standard deviation for each assertion.

+1.96 for good comment.

Would you please help us middling-IQ people understand your statement, by recasting it in simpler language?

Sure. There are many differences between men and women, but they tend to like people about as smart as themselves. (I am basing this on familiarity with the published statistics and with extensive personal anecdotal evidence). Often the difference between husband and wife is big when it comes to verbal intelligence, often it is big when it comes to math skills. Where one has higher verbal or math skills, often the other has higher math or verbal skills (in other words, the complementary or opposite skill from the spouse - in the US and Europe, that usually means the husband is higher M, the wife higher V, with the husband's high M within 5 points of the wife's v or the wife's low M within 5 points of the husband's low V). Happy people just like being around people who remind them of themselves (and of their friends). The letter M and the letter V just refer to relative genetic skills at (broadly speaking) mathematical or verbal skills. The comment I was responding to stated that people associate within 5 points on a scale (namely g, general intelligence as one number); I was trying to be more accurate to the real world by making the much more modest proposal, saying it was 5 points within one of two scales (M and V) (I was being honest, by the way - of the couple hundred couples I know, I guess that half fall within this 5 point difference if you use two scales). I don't think anyone thinks that 50 percent of couples contain 2 people within 5 points on general intelligence, but if you throw in the M and the V for each, somewhere there is likely to be a close 5 point or less match. The reason men over 125 and under 90 are not included is that there just are not that many available women at those numbers to pair up with, and, happy about it or not, they have to be willing to forego matching someone like themselves. The reason the range is smaller for women is because women vary less on these two scales (M and V) than men do, so a woman at 120 is just as unusual as a man at 125, and will have to make the same sacrifice in choosing someone not like herself as the man at 125 would. The bit about a half standard deviation was a (self-depreciating) joke; half a standard deviation is between 5 and 10 points difference for most people, so if I am off by a half standard deviation then I am just wrong about people matching up within 5 points at M or V. The 1.96 reply was .04 less than a 2 reply; meaning perhaps that I just made it within the .05 limit I had set myself; making fun of my comment, a little bit, because nobody can be that accurate, but also hopefully recognizing the humility in my description of my fellow humans. Maybe I did not get what the 1.96 meant exactly, though.

easily amused - I left out that I am really not sure ... exactly... what Anti-Gnostic's statement and Millian's statement meant, so they would have to speak for themselves, and I left out that the Borgnine reference is mostly to the movie Marty, a great film about assortative love and mating in 50s NYC....there is also some good assortative mating in the Poseidon Adventure, not to mention McHale's navy (both the E.B. version and the underrated time-defying Kelsey Granmer version)

Oh. I thought I was praising a facetiously complex satire of IQ-determinism. Turns out it was a sincerely complex example of IQ-determinism. My mistake. Perhaps if my IQ were just 4.2 points higher, I would have known the difference.

I believe a little incompatibility is the spice of life, particularly if he has income and she is pattable.
Ogden Nash

Millian - you may still be mistaken! But that cannot reflect on your IQ, as there is, to date, no sincerity verification test for internet comments. Maybe I think anti-gnostic's insight was valuable (as his insights often are - maybe yours usually are too, I don't read this site enough to know), or maybe I am not a "real statistician", an "academic statistician", a "folk statistician", or even a would-be "comical statistician", but at best some guy who for arcane reasons wanted to set up a Rube Goldberg-like slow-developing appreciation of Ernest Borgnine films in the quiet and mostly unvisited further reaches of an evanescent internet "community".

Hmm.... That 'group into 5' statistic seems rather forced. If we grouped all people in the United States randomly into groups of 5 people, wouldn't the poverty level pretty much go to zero?

How much of Bolivia's assortive mating is actually race-based mating? Many Latin American countries are split along class/race lines, with a wealthy white, urban class, and the rest darker-skinned and poorer.

Some of the assortative mating is due to evolutionary adaptations to Bolivia's altitude extremes. White women have increasing problems carrying babies to term at above 10,000 feet. Conversely, people from the Altiplano are said to find the lowlands oppressive.

Isn't that the plot of "The Giver"? Equality achieved through random assignment of children to families?

I don't know, but unfortunately that would not result in equality, but plenty of child abuse.

Make everyone reproduce by IVF. Randomize the redistribution of zygotes, adjust for race, and people would never know which, if any, of the children were their biological descendants.

Probably wouldn't matter- they would know with a high probability that the child they were raising was someone else's. What you describe is essentially mandated adoption, and I think the results would be catastrophic.

Give it eugenic spin: a rancher seeks out the best sperm for his stock; how odd that he not give the same consideration to his sons and daughters.

Down on the farm, genetic diversity is also a consideration.

how odd that he not give the same consideration to his sons and daughters.

A lot of this is burned into our primal brains. Men aren't aroused by apple-shaped women; women aren't aroused by pear-shaped men. Facial asymmetry is a turn-off. Bad skin, bad hair, obesity, etc.


For much of history plumpness (though, no, not morbid obesity) was considered a sign of good health and prosperity, and wide hips in women was a sign that they would have an easier time in child birth. The modern preference for the extremely svelte is a creation of modern circumstances-- thinness is now a sign of prosperity since it means a person has the leisure to exercise. (Something similar has happened with suntans: once a sign that a person worked outdoors and hence was of low status, now it's a sign that someone has the leisure to sun themselves.)
For most of history the ideal woman's shape was a lot closer to Ruben's paunchy nudes than to today's "Heroin chic"

Yes, we hear that often ("In the 17th century, I'd be a diva!"). Actually, Rubens' paintings are rococo grotesqueries. Here's his "Pumping Iron" version of Christ's Resurrection, for example.

We also hear about "heroin chic," which actually exists only in homosexual-dominated industries. In all other contexts, beautiful women have an hour-glass shape.

The hourglass shape however involves wide hips.
The fact remains that today's preferences in body size are shaped by today's culture: thin is a sign of health and prosperity (implying the leisure to exercise and devote plenty of time to careful food prep) while any excess weight is seen as a sign that one is a "wage slave" and does not have the resources for a more effete diet. If you look at women who were acclaimed beauties in earlier eras of history they generally had some heft on them (again, not to the point of obesity!) and the men too sported a look that said "well fed, not starving". Women's fashions until the 1900s allowed for plumpness, the sole exception being the Empire gowns of the early 1800s, and that did not last very long as they made full-figured women look dreadful (Mme Recamier, a virtual wraith, could pull it off; her best friend buxom Mme de Stael looked like an grotesquely overstuffed sausage).
I'm not sure why you associate the heroin chic look with gays, unless this is just lazy stereotyping (many fashion designers are gay, hence gays must like the look).

Well, I've done my bit for equality: My wife outearns me by a factor of three or four!

What kind of pediatrician is she?

Isn't this evidence in favor of assortative mating?

In my experience, men put more thought into buying a car than choosing the mother of their children.

We said the same thing, but you were a thousand times more eloquent. Well done.

When I see a potential mate, thought isn't what um thinking of putting into her, if it can eve be called thinking.

>If people married each other more randomly, poverty levels would be considerably lower

I think you mean "women."

Men already do marry quite randomly, socio-economically speaking. Much, much more so than women, anyway.

Leaving aside same-sex marriages, how can the rate of marriage across income or education levels be different for men than women?

If the distribution of the never-married is different between sexes, then the rate of exogamous marriage could also be different between sexes. Imagine a society where the women who never marry are the least physically attractive, while the men who never marry are the poorest.

LOL. Congratulations on your logic, you wins the thread. Don't expect to change any minds, thoguh - dem vacuous wimmins have to be to blame somehow, even if their fault is mathematically impossible!

I see very little marriage "outside one's station," as the saying goes, by men or women. The archetypal exception is the heiress bringing home the prole galoot, much to the family's mortification.

I'd have enjoyed marrying a high-earning doctoress or Old Money who would allow me to indulge my passion for expensive hobbies and self-directed research, but haven't managed it in this lifetime.

Louis CK: 'Nobody fΠ¢%$ down, except when a man tricks a women into thinking it was up."

In what sense do men marry randomly? Most people marry within their social circle, which is usually limited to their social class (somewhat broadly defined)

I don't mean to be cynical but how is this potent ? Y in some completely fictional world things would be different. What an insight. Has any civilisation existed without assortative mating ?

Maybe small hunter gatherer bands where there's no social classes because the group is too small. But a full-fledged civilization-- no.

It is weird to put it in terms of "randomly." Presumably to come up with their poverty numbers and Gini, they selected one random draw. Wouldn't some random draws increase poverty and inequality? If they initially drew such an outcome did they then throw it out and draw again? Should be discussed in terms of a probabalistic range of random outcomes.

Comments for this post are closed