Status and Interracial Marriages

PEW has a new study (pdf) out on the characteristics of interracial marriages in the U.S.

As you can see, not only is there a clear racial hierarchy in terms of education and earnings, but the interracial marriages have a similar hierarchy. Interesting findings: while asian-asian marriages have higher levels of both spouses having college education than white-asian marriages, white-asian marriages earn more; also, white male/black female marriages make slightly more than the white-white average, but white female/black male, make significantly less

Here is graph showing that those whites marrying asians are significantly better educated – whereas, among white women, there is a trend for those marrying blacks to be less educated:

Among blacks and hispanics, both men and women marrying whites tended to be better educated than those who married within their race, but by varying amounts (black women were 7% more likely to have a college degree, but black men only 2%; hispanic women were 20% more likely to have a college degree, and hispanic men, 12%). For asians, there was very little difference in education whether they married whites or asians.

Final note: given the prevalence of cohabitation, and declining marriage rates, this data is probably somewhat misleading. I already went over the differences in rates of interracial cohabitation vs marriage. If those relationships were counted, it would probably widen the gap found here. For example, asian-male/white-female marriages were more likely than black-male/white-female marriages, given their respective proportions of the population. However, that was strongly reversed in cohabitation rates. Since marriage vs. cohabitation is partly a class issue, this fits into the data here, and would probably make it stronger.

Interracial relationships – who is attracted to who?

[See also my post ‘Racial preferences and online dating‘ and also ‘Status and interracial marriages‘]

Johann Happolati responded to my post on racial preferences in online dating by noting that there are important sex differences in who is crossing what racial boundaries. His recounting of the unpleasant situation has moved me to hope that by playing around with numbers all of this stuff would go away. I was not entirely successful, but I did make many interesting discoveries along the way.

So, anyways, all this talk of racial preferences inevitably brings up the question: are some races in general more attractive to the men or women of certain other races?

First we will remind the reader of the evidence that all races have a general preference for their own, at least in terms of dating and marrying, and women more so than men. And not just in terms of what they claim to want, but as demonstrated in online dating sites by the likelihood of sending a message, or replying to a message.

But, even given that tendency, there is still plenty of room for varying degrees of attraction to other races depending on one’s sex and race.

The data isn’t really conclusive in many cases:

Some studies report (source: 1,2,3,4,5) that black women are considered less attractive by some other ethnicities. The well-known Kanazawa study, which claimed that the nationally representative ADD Health survey found African women were less attractive, has been much criticized, most notably by Kaufman and Wicherts.  Kaufman and Wicherts also found little to suggest differences between men of different races.

Michael Lewis has done a series of three studies in the U.K., where the same trend in interracial marriages exists as in the U.S. All three show higher ratings of attractiveness  for black men compared to white, and in the third, both over asian men. In all three as well, white women are rated more attractive than black women, and, in the last study, in which asian women are included, they are rated the most attractive. In one study mixed race male faces were rated intermediate to black and white male faces, but in another, black males were rated more highly (sources: 1,2,3). The studies, however, involve fairly small numbers of students, between 10 and 20 women rating men, and a similar number for men rating women.

A survey of over 2,500 whites asked which race they found most physically attractive, and found that overall 62.7% of women and 48.2% of men were most physically attracted to other whites, much of the rest claimed no preference, 12% of men and 3% of women found asians most attractive, and 1.5% of men and 1.9% of women found blacks most attractive. Political view had a very noticeable effect.

In another study, asian males were rated as least attractive, including by asian women, although asian men also rated asian women as less attractive than whites and hispanics. Black women were rated as the least attractive. White and asian women gave lower ratings of attractiveness to black men.

The data is not conclusive, but given the high sex-disparity in interracial marriages, it is plausible that there might be some difference.


OK, now let’s look at what people actually do:

Data from online dating sites shows that black men and women are responded to less than one would expect. Asian males actually receive a slightly higher response rate, and white males, much higher (among women in general – there are differences depending on race).

Update: Here’s data taken from Facebook app ‘Are You Interested’ which shows the highest and lowest positive response rate by race and gender when notified that someone has clicked ‘yes’ to the prompt ‘I like what I see’ when viewing their profile.


If you look at 2008 PEW survey data, whites, both men and women, marry other races at a low rate, and those interracial marriages are not evenly distributed, as well see below:

Using the U.S. census data, asians make up 13% of the non-white population, blacks make up 34%, hispanics 45%, which leaves 10% for ‘other’).

As you can figure out from the above information, white men are marrying asians and ‘other’ at about twice the rate they should be, hispanics at the expected rate, and blacks at less than 1/5 of the rate they should be. White women are marrying ‘others’ also at about twice the expected rate, however, they are marrying hispanics at a slightly higher rate than are white men, while they are marrying asians and blacks at lower than expected rates: asians at 3/4, and blacks at 4/7, of what would be expected.

In other words, in the marriage market, asian men are actually more popular with white women than are black men.


But, of course, these days we can’t just talk about the married people – what about cohabitating couples? This is somewhat older data, from the 2000 census.

The picture changes here somewhat:

Once again, taking into account how much each of the minority population makes up of the non-white population (Asian: 13%; Blacks: 34%; Hispanics: 45%; Others: 10%), we see that white men are cohabitating with black women at 1/4 expected rate (8.8% of those cohabiting with non-whites), hispanics at the expected rate (46%), asians at 1.5x expected (18%), and ‘others’ at 2.8x the expected rate (28%).

For white women, cohabitation rates with blacks are almost the expected rate (32.5% of those cohabiting with non-whites), Hispanics are also just slightly under expected (43%), Asian men are doing somewhat worse here than in marriage, with only about 2/5 the expected rate (5.3%), and, finally, the always popular ‘others’ at twice the expected rate (20%).

In other words, there is change, but it doesn’t radically alter the picture: white men and black women still aren’t having much to do with each other, and white men still really like asian and ‘other’ women; white women have changed slightly more: while Hispanics maintain their position at the point that they should be at, and ‘Others’ remain the most popular groups, cohabitation rates are now almost what one would expect for blacks – but asian men become  less popular.

The point is, though, that the only group that is actually ‘popular’ with white women is the ‘other’ category, whereas for white men, asians and ‘others’ are popular. White men and black women are consistently not even remotely close to expected, while for white women both blacks and asians have a combined (marriage and cohabitation) rate that is less than expected (and less drastic than the white-man/black-woman gap). As noted above, women marry asians more (relatively speaking – according to their share of the minority population) than they do blacks, but cohabitate with blacks more than asians. Hispanics are consistently about exactly as popular as they should be.

“Should be” itself needs to be remembered in context –  in both marriage and cohabitation stats, people are disproportionately likely to be with someone of their own race, especially whites, who never fall below 90% white-white relationships, even in cohabitation stats, despite only being 64% of the national population. And so it needs to be remembered that when I said whites were marrying or cohabiting inter-racially at the expected amount, they are in fact marrying minorities at much less than expected amounts for the population as a whole – the comparison above was simply in terms of how minorities compared against each other.

For more graphs see this post at Audacious Epigone.

One more note: it is clear that white men’s preferences are skewing the picture both in marriage and cohabitation, artificially creating for many the impression that women have an opposite preference, which does not seem to be the case – in fact, white women are less interested in both blacks and asians, and more interested in Hispanics and ‘Others’. But, of course, both white men and women are much more interested in other whites.

(Some have proposed black women’s racial loyalty as an explanation for why there aren’t more white-male/black-woman couples, but this is unsatisfactory. For one thing, women of all races generally state a higher preference for dating someone of their own race than men do. Secondly, if you see my post on racial preferences in online dating, you’ll see that black women actually respond to black men at the lowest rate relative to what one would expect, and they respond to men of all other races at a higher level.)


A 2005 Gallup poll also has information on interracial dating in the U.S., rather than cohabitation or marriage. Not surprisingly, more people have dated someone of another race than actually cohabit with or marry someone of another race. Below is the important table, which includes an odd result. According to this, 25% of white men and 24% of white women have dated a black person. Given the huge disparity in terms of white men and women who cohabit or marry blacks, this result seems somewhat implausible, especially since, for blacks, 56% of men and only 35% of women reported dating a white person. (Steve Sailer, in his article Is Love Colorblind? reports that the 1992 Sex in America study which polled 3,432 people, found that white women were 10 times more likely than white men to report that their most recent sexual partner was black. I haven’t seen the study myself.) The numbers of people who report having dated inter-racially is fairly high, overall 48% of American’s (45% of whites; 52% of blacks; 69% of hispanics)

Dated a Non-Hispanic white Dated a black Dated a Hispanic Dated an Asian





All non-Hispanic whites





White men





White women





All blacks





Black men





Black women





All Hispanics





Hispanic men





Hispanic women






The following is an analysis of the 2005 Natality Detail File, which provides information on the degree to which children are being born into mixed-families.

Of the children born to non-Hispanic White women, the fathers were:

89.23% Non-Hispanic White
5.46% Hispanic
3.18% Non-Hispanic Black
2.13% Non-Hispanic Other Races

Of the children born to non-Hispanic White men, the mothers were:

92.76% Non-Hispanic White
4.12% Hispanic
0.96% Non-Hispanic Black
2.17% Non-Hispanic Other Races


Global data:

In the Netherlands, intermarriage rates are quite similar among Surinamese and Antillean men and women. 22% of Surinamese men marry native Dutch, while 26% of women marry native Dutch. Among the Antillean immigrants, 48% of men marry native Dutch, compared to 40% of women.  (source)

In Brazil, data collected on marriage between whites, mestizos, and blacks can be seen in these tables (numbers are %s of marriages):

Men Women Total
White Brown Black
White 84.2 14.4 1.4 100%
Brown 25.4 72.2 2.3 100%
Black 19.7 25.0 55.4 100%
Women Men Total
White Brown Black
White 82.2 15.8 2.0 100%
Brown 22.8 73.8 4.1 100%
Black 16.9 16.9 66.2 100%

As you can see, it is less common for black women to marry brown or white people as compared to black men (combined total of 44.7% for black men, and 33.8% for black women). Of whites, 15.8% percent of white men marry black or brown women vs. 17.8% of white women (or 1.4% vs 2.0% for blacks only). Among both whites and browns more women marry out, but among blacks, more men marry out. Among browns, also, more women marry blacks than do men (4.1% vs 2.3%), and more men marry whites than do women (25.4% vs 22.8%).

In Australia, like the U.S., there is a trend for North Asian women to marry outside their race at a higher rate than men, whereas the opposite was the case for immigrants from, for example, Lebanon. (source)

The U.K. shows the same gender asymmetries in interracial marriages as the U.S., as can be seen in the below chart – it is interesting, though, that the asymmetries are smaller in the UK.

In Canada, we find the same pattern. From a 2006 Statistics Canada report: “In 2006, there were more than twice as many Arab or West Asian married or partnered men who were paired outside their group (19%) as there were women (9%). Similarly, three in ten Black men in couples were in mixed unions as were two out of ten Black women. Filipino, Korean, Southeast Asian, Japanese, Chinese or Latin American women in couples accounted for a higher proportion of spouses or partners in mixed unions than did men from these visible minority groups. There were more than three times as many married or partnered Filipino women in mixed unions (28%) as there were Filipino men (9%). For Japanese, nearly two-thirds of Japanese women in couples were in mixed unions while this was the case for over one-half (52%) of men from this visible minority group.”

It’s interesting to note that, like the UK, the discrepancies are still obvious, but not as large as in the US.

In Sweden, if we look at the global sources of marriage migrants (those who migrate from elsewhere in the world and marry a Swedish spouse), compared to immigration to Sweden in general, we see different patterns between the genders:

To summarize some of the international gender discrepancies among asians and blacks in marrying into native European populations:

Country Bias toward asian women (asian women marrying whites / asian men marrying whites
Canada 1.6
U.S. 2.64
UK* 2.45
Sweden** 7.25
Australia*** 4.65
Average 3.718 (median 2.64)
*Chinese only
**includes south asia
***foreign born, only Japanese & Koreans
Country Bias toward black men (black men marrying whites / black women marrying whites
Canada 1.5
U.S. 2.32
UK 1.32
Sweden 1.8
Netherlands* 1.2
Brazil 1.17
Average 1.55 (median 1.41)
* from the Dutch Antilles, which is 85% mixed black. Immigrants from Surinam not included because it is only 40% mixed black

Take into account that these numbers are based on studies that are not all using the same methods, I haven’t made any attempt to correct for gender bias in immigration, etc.

I suspect some of the very high biases have to do with smaller numbers and unusual migration patterns. But there is certainly a pattern here.

Primate social behaviour and genes

Some have stressed an environmental determination of primate social behaviour, but a recent paper suggests that social behaviour in primates is genetic, not environmental: “Stepwise evolution of stable sociality in primates” (2011). Also see Nature’s commentary, and Nicholas Wade’s article in the NYT.

By studying many different primate lineages (see image below), the important conclusion they reach is that closely related primate groups display similar social organizations even when in different environments:

Caption from the study: "Branches and tips are coloured for solitary (purple), uni-male (orange), multi-male (red), pair-living (pink) where the combined probability of the state and the branch is greater than or equal to 0.7. Where the combined probability is less than 0.7, the branch is grey. Histograms represent the posterior probability distribution of each social state at the nodes indicated (a, primate root; b, anthropoid root, c, catarrhine root; d, great ape root; e, Pan–Homo split; f, Old World monkey root)."

From the Nature commentary: “The existence of a strong phylogenetic signal spells trouble for socioecological models that aim to explain the evolution of primate social organization. The models hypothesize that food distribution shapes competitive regimes, and that these, in turn, shape dispersal patterns and the nature of relationships within groups. These models generally assume that phylogeny does not impose notable constraints on social organization, and that changes from one form of social organization to another are all equally likely. But there is a growing realization that history does have a role and the new results strengthen that view.”

Racial preferences in online dating

Update: See also my post ‘Interracial relationships – who is attracted to who?‘ and also ‘Status and Interracial Marriages

OKCupid put up a blog post a while ago in which they analyzed how race affects rates of replies.

The below images show how the actual responses stack up against ‘expected’ responses (ie. corrected for all other factors). Redder = lower than expected response, greener = higher than expected response.

OKCupid also did a survey of who would ‘strongly’ prefer to date someone of their own race:

So, minority groups claimed to be much more open to interracial relationships – what’s driving this? Clearly these things are highly flexible. European men often took non-European wives over the last few hundred years as they traveled the world. But that could be driven by scarcity without changing underlying preference – even strong inclinations are often overriden by environmental factors.

However, OKCupid followed up with another analysis, asking the question ‘what if there were less white people?‘ how would the interracial messaging pan out then?

What they find is that the popularity of whites is distorted by the also overwhelming numbers of whites. When they recalulated a world with equivalent populations, but the current messaging habits, they find that Asians would be the most popular, followed by latinos, with whites being third, and even declining to last as age increases:

And here we see how many messages per month an individual would receive in this brave new world, compared to today’s world (lighter shade):

Finally, we see what the situation would look like if the racial bias were the same as it was today, but the numbers of people are equal. As you can see, all groups have a stronger preference for their own (especially Asians), and all find certain other ethnicities more or less appealing.

It seems as if, despite their claims to the contrary, minority groups are actually more racially selective than whites, although all groups have some preference for their own race.

Good luck post-racial world.

Monogamy and Western Civilization

There was a case recently, in British Columbia, which ended in retaining the illegality of polygamous marriage. It revolved around a breakaway Mormon cult in Bountiful, B.C. As part of the proceedings, three academics provided briefs, which are both very interesting. Walter Scheidel (pdf) of Stanford, and Joseph Henrich (pdf), of UBC. Henrich has just published a paper on the same subject. Another paper by Scheidel on the origins of western monogamy is here. See also the affidavit by John Witte Jr. on the history of western ethical and legal outlooks toward polygamy (pdf).

What I’m interested in is the connection between monogamy in western civilization and the evolution of perceptions towards marriage and women.

Ancient Greece and Rome both had, as Scheidel describes it, ‘socially imposed universal monogamy’, and likewise Christianity (unlike Judaism), seems to have implicitly rejected polygamy from the beginning, and explicitly within a short amount of time.

So the earliest form of socially imposed universal monogamy dates to the ancient Greeks. And most interestingly, it seems to be connected to the democratizing and egalitarian ethos of the period – (see the paper “Solon and the Institution of the ‘Democratic’ Family Form”; as well see Scheidel’s papers). The Greeks themselves considered this to distinguish them from the barbarians. The exact roots of socially imposed universal monogamy are unfortunately unclear, so it is hard to be certain about the social forces driving it in ancient Greece. But the cultural memory of polygamous marriages was still present, and represented in epics like the Iliad. So it came after that, presumably. We find in the laws of Solon an apparent intention to re-organize family law in order to equalize Athenian male citizens, suggesting the socially imposed aspect of it came into play in connection with egalitarian (among men) ideals.

Rome, too, had universal monogamy for as long as we know. What caused this? It’s even harder to say, but after all Rome was a republic, had heavy influence from Greek society, etc. – but, as Scheidel points out, in the end, we basically just know that Greece and Rome were monogamous. No compelling explanation for why both of them specifically has yet been forthcoming.

I should note that at this point that Laura Fortunato argues that monogamous marriage systems date back much further among Indo-Europeans. However, given the evidence of polygamy from numerous ancient sources both in their own prehistory, in their discussions of neighbouring Indo-Europeans, etc., it is clear it could not have been a universal monogamy, even if it was overwhelmingly prevalent, and perhaps served as a precursor to the Greco-Roman variety.

But what interests me is the connection between monogamy and egalitarianism. In the same way that democratic ideals eventually expanded beyond the realm of male citizens, to women and slaves, I wonder if the egalitarianism inherent in monogamy eventually had an impact on the view of women in western civilization. Even from the time of Socrates and Plato, to the period, of say, 1st century imperial Rome and Greece, there seems to me to have been a shift in attitudes towards women. Compare for example Socrates attitude towards his wife at his death, compared to that of Seneca towards his wife – despite the death scene of Seneca being modeled on that of Socrates.

Or, as an example of how these things change meaning over time – consider Plutarch’s interpretation of Solon’s marriage laws – ie. he throws in ‘pure love’ as a motivation, which may be a reflection on the recursive nature of the institution of monogamy, because Solon certainly never emphasized that, and neither would his contemporaries:

In all other marriages he forbade dowries to be given; the wife was to have three suits of clothes, a little inconsiderable household stuff, and that was all; for he would not have marriages contracted for gain or an estate, but for pure love, kind affection, and birth of children. When the mother of Dionysius desired him to marry her to one of his citizens, “Indeed,” said he, “by my tyranny I have broken my country’s laws, but cannot put a violence upon those of nature by an unseasonable marriage.” Such disorder is never to be suffered in a commonwealth, nor such unseasonable and unloving and unperforming marriages, which attain no due end or fruit; any provident governor or lawgiver might say to an old man that takes a young wife what is said to Philoctetes in the tragedy-

“Truly, in a fit state thou to marry!”

and if he find a young man, with a rich and elderly wife, growing fat in his place, like the partridges, remove him to a young woman of proper age. And of this enough.

Musonius Rufus, the ‘Roman Socrates’ and one of the most prominent of Roman philosophers, taught:

The husband and wife, he [Musonius] used to say, should come together for the purpose of making a life in common and of procreating children, and furthermore of regarding all things in common between them, and nothing peculiar or private to one or the other, not even their own bodies. The birth of a human being which results from such a union is to be sure something marvelous, but it is not yet enough for the relation of husband and wife, inasmuch as quite apart from marriage it could result from any other sexual union, just as in the case of animals. But in marriage there must be above all perfect companionship and mutual love of husband and wife, both in health and in sickness and under all conditions, since it was with desire for this as well as for having children that both entered upon marriage. Where, then, this love for each other is perfect and the two share it completely, each striving to outdo the other in devotion, the marriage is ideal and worthy of envy, for such a union is beautiful. But where each looks only to his own interests and neglects the other, or, what is worse, when one is so minded and lives in the same house but fixes his attention elsewhere and is not willing to pull together with his yoke-mate nor to agree, then the union is doomed to disaster and though they live together, yet their common interests fare badly; eventually they separate entirely or they remain together and suffer what is worse than loneliness.

And here is Hierocles (early 2nd century):

[T]he beauty of a household consists in the yoking together of a husband and wife who are united to each other by fate, are consecrated to the gods who preside over weddings, births, and houses, agree with each other and have all things in common, including their bodies, or rather their souls, and who exercise appropriate rule over their household and servants, take care in rearing their children, and pay an attention to the necessities of life which is neither intense nor slack, but moderate and fitting.

My general point here is that classical attitudes did change over time, and it is possible socially imposed universal monogamy may have had the effect of promoting the idea of marriage as a partnership in which love was important. Not just in the ancient world, but perhaps served to encourage that line of thinking throughout until now.

I won’t bother going much into Christianity and all the intermediate history, but we know that Christianity strongly adopted and enforced Greco-Roman monogamy, and spread it throughout Europe, with an even more strict sexual morality. In fact the motivations of the Catholic church have been argued to be very similar to the Greeks: by insisting on monogamy and making all children born from other women bastards without rights, one means of accruing power was removed (and in the case of no legitimate heirs, property went to the church). Attitudes towards women were also, of course, influenced by Christian teachings, and also the attitudes of Germanic and Celtic societes. But I will skip all that, and will quote from one of the earliest Japanese visitors to Europe in the mid-19th century (from “The Japanese Discovery of Victorian Britain: Early Travel Encounters in the Far West”):

“Kawaji … had ample opportunity to observe the European tradition of chivalry in practice. It seemed to him that ‘the ladies on this ship possess great authority and assume an air of importance equivalent to that of an imperial princess in our land’. He was curious to discover ‘the British custom of paying inordinate respect to ladies; they take their seats before their husbands and sit in the best places at mealtimes as well’. Later in the voyage he concluded that, ‘of all the countries in the West, Britain has the most pronounced custom of paying respect to ladies. From what I have seen on this ship, it seems that, when talking to a lady, you take off your hat and treat her most politely. This is the reverse of the ranking in our country, and I find it most astonishing’.

“Kawaji felt moved to ask for some clarification on the subject of chivalry, but was disappointed to learn that ‘this is an old custom, and there is no particular explanation for it’. He was, nevertheless, impressed by the fact that ‘the custom of holding ladies in such respect enables them to travel thousands of miles overseas on their own without coming to any harm’.

I also remember reading a similar quote by an Ottoman official visiting a western European in the 19th century in one of Bernard Lewis’ books; Lewis took the view that the strict enforcement of monogamy was one of the reasons women in western Europe were better off than women elsewhere.

Of course in many places monogamy was overwhelmingly common, even though polygamy was accepted. This distinction could nevertheless leave room for quite a bit of difference in perception, even if the numbers of people actually in a polygamous marriage are not large.

One problem with my general interpretation is that, like a lot of other developments generally attributed to Western Civilization, or Christianity, it in reality seems to be more of a (north-)western European thing. Many other factors could have played into it as well (‘it’ being women’s rights) – the Western European Marriage system, for one.

Still, monogamy seems to have a connection to various forms of egalitarianism, whether between men, or between men and women. It could be a coincidence of course, but then it was born in the move towards Greek democracy, and Roman Republicanism. Perhaps rather than being a direct cause, it is part of a western package of beliefs that mutually reinforce each other. If that is the case, the dangers in isolating one aspect, like socially imposed universal monogamy, and asking whether it specifically can be relinquished, is risking a lot for no apparent benefit. But given the recent enthusiasm for abolishing western civilization, I guess it’s unlikely to be maintained for long.

Hadza and height preference

Amongst the Hadza of Tanzania height doesn’t appear to be a factor in choosing a partner (How universal are human mate choices? Size doesn’t matter when Hadza foragers are choosing a mate, 2009). For example, the graph below shows the percentage of marriages one would expect to find in which the woman is taller than the man, if mating were random, vs. what you actually find. As you can see, the Hadza (and Gambians) have the rates of taller wives-shorter husbands one would expect if it were not an important factor, whereas in the UK, that pairing is significantly smaller than expected.

A 2012 paper on the Baka Pygmy (Short stature in African pygmies is not explained by sexual selection), however, includes as a comparison the neighbouring non-Pygmy Nzimé and find that in both populations, the rate of female-taller pairs is less than one would expect by random pairings – but not nearly so much less as in western samples:

Observed (white bar) vs. Expected (striped bar) female-taller pairings among several populations.

Observed (white bar) vs. Expected (striped bar) female-taller pairings among several populations.

As for preferences, the authors write:

When asked about the importance of stature in the Baka society, men and women stand together on a very egalitarian discourse: “tall or short, this makes no difference.” When asked their personal preferences, 21 out of 29 of our male informants openly declared that they would not marry a woman taller than themselves. Women were less definitive on the question: they never said that tall stature is a standard of desirability in men, and only 10 out of 38 female informants said that “a woman should not be taller than her husband.” Interestingly, all the informants mentioned that Baka men propose a love relationship to women, never the reverse, which may be an explanation for different opinions expressed by men and women. If men mate choice is highly constrained by gendered representations, as we recorded, then a man would only very rarely propose to a taller woman, and women will thus very rarely be in the position to consider the opportunity of marrying a man shorter than themselves. It is therefore unclear whether, when faced with a choice between several men, a woman would preferentially mate with the tallest.

Another paper (Variable Preferences for Sexual Dimorphism in Stature (SDS) Might Not Be Universal: Data From a Semi-Nomad Population (Himba) in Namibia, 2011) looked at stated preferences without looking looking at actual pairings. The results were shifted away from western preferences. While 50% of the Himba preferred a male-taller coupling, 30% preferred equal-height partners, and 20% preferred female-taller couples.

In Western studies in general the vast majority of people state a preference for a male-taller partner, with men having a generally less pronounced preference; as well, there is almost no one who prefers a female-taller pairs (sources: 1,2).

Compare the Himba preferences in sexual dimorphism:

With that of three European nations (using the same ratios) (source):

A study of the Yali tribe in Papua New Guinea (Judgments of Sexual Attractiveness: A Study of the Yali Tribe in Papua, 2012) found that they similarly do not show the same preference for male-taller pairs. As you can see in the below graph, preferences for different types of pairings are no different than what would be predicted by chance. As the authors note, this could be either because they have no real preference, or because of diverse preferences. While still about 60% chose a male-taller pair, that is significantly lower than European populations which all chose male-taller pairs at over 90%.

Another 2012 study, of the Datoga people of Tanzania (Height preferences in humans may not be universal: Evidence from the Datoga people of Tanzania) found yet another pattern. As you can see in the below graph, the Datoga expressed a preference for the extreme pairs, and, once again, a significantly larger amount than in western samples (like in the Polish sample, left) preferred female-taller or equal-tallness pairs.

Left graph shows SDS preference of Polish sample, while right shows Datoga.

Now, I would like to see more comprehensive studies done in horticultural and hunter-gatherer societies, as well as more non-western societies, however the above evidence suggests that there is certainly less conscious emphasis on height as an attractive male quality, backed up by evidence suggesting female-taller pairings are nearly (or least closer than one finds in industrial societies) to what one would expect by purely random pairings.

Is it possible that height as an important feature of male attractiveness is cultural? If so, what would be the cause? I was reading Leopold Pospisil’s ethnography of the Kapauku Papuans – he writes:

The quantitative orientation of the people leads them into placing value upon higher numbers and larger volume. Accordingly, a tall individual is admired and a weak or small one is regarded as peu, bad… Most of the objects that are small are bad, or at least not so good as larger ones.

On the other hand, perhaps these different populations have undergone different selection pressures to mold these preferences.

Family structure and inequality

“Income Inequality: New Trends and Research Directions” (2010):

By 2007, four in ten births were to unmarried mothers (Ventura 2009), and although the characteristics of unmarried mothers changed somewhat from the 1980s to the 1990s and 2000s, the nonmarital birth rate climbed at a fairly steady rate over this period, with the 2007 rate about 80% higher than that for 1980. This increase in the share of families headed by single mothers is hypothesized to increase inequality by increasing the number of families with very low incomes, as most single mother families work fewer hours and receive less pay than other families. Studies of the impact of increasing single motherhood on family income inequality vary in their estimates, with a range from 11% to 41%.

Notably, several studies that simultaneously consider the effects of
single motherhood (or female headship) and women’s employment on income inequality find that these trends had largely offsetting effects…
…relatively few published studies have considered how increases in homogamy have affected income inequality… both sets of authors
conclude that increasing marital homogamy
cannot account for much of the increase in income inequality in the periods that they examined… Likewise, Western
et al. (2008) find that changes in educational homogamy do not alter their estimates of changes in family income inequality… However, three recent working papers reexamine this question, and two conclude that increases in homogamy have made the distribution of family incomes substantially more unequal.

There is strong support for the hypothesis that increases in single mother families and decreases in married couple families have increased income inequality and fairly strong evidence that increased women’s employment and earnings have reduced inequality, at least through the 1990s. In contrast, there is little consensus about the impact of marital homogamy or income sorting on income inequality, suggesting that more research is needed.