I’ve written about assortative mating and income inequality before, pointing out that the more educated tend to marry each other and therefore increase their economic earnings. Ronald Bailey at Reason weighs in on the discussion, adding to the mix evidence that shows assortative mating isn’t just about education, but intelligence. Quoting a 2015 study, he writes,
For example, if spouses mated randomly in relation to intelligence, highly intelligent women would be just as likely to mate with men of low as high intelligence. Offspring of the matings of women of high intelligence and men of low intelligence would generally be of average intelligence. However, because there is strong positive assortative mating, children with highly intelligent mothers are also likely to have highly intelligent fathers, and the offspring themselves are likely to be more intelligent than average. The same thing happens for less intelligent parents. In this way, assortative mating increases additive genetic variance in that the offspring differ more from the average than they would if mating were random. The increase in additive genetic variance can be substantial because its effects accumulate generation after generation until an equilibrium is reached.
He concludes, “To the extent that intelligence is correlated with socioeconomic status, assortative mating will further exacerbate trends to greater income inequality.”
University of Washington professor Tony Gill once shared a thought experiment he employs in his classes during a Facebook discussion:
Most students are for higher marginal taxation on the rich (defined as the dollar amount of people who have a wee bit more than them).
I propose centrally planned sorting by either IQ or socio-economic status (noting some studies that show how IQ might have a hereditary component and how IQ might be related to long-term income potential). I also note that we tend to marry people who are educationally and socially close to us (e.g., people meet at Harvard or in the same upscale neighborhood bars). Some of us use mail order catalogues, but we usually get a box on education to check.
Students freak out. First, they say that this has never been done. Then I note how arranged marriages are not an uncommon fixture in history. Then they say it isn’t possible because of data concerns, and I remind them about all those tests they took in 3rd, 7th and 11th grade and their “permanent record,” not to mention all the income data the IRS has on their parents.
Then they squeal that this isn’t right because it limits their freedom to do what they want. And then I say, “Oh, so now you’re worried about centrally-planned limits on freedom, eh?”
So, next time you get the social justice itch to redistribute wealth, ask yourself the following:
- Are the adjectives smart or intelligent used to describe your spouse? Are they some of the reasons given as to why you love them?
- Did you meet your spouse at college?
- Would it have a negative influence on your choice to date an individual if they were a waiter/waitress, barista, fast food employee, Walmart cashier? (And not one who is working there part-time while they go to school.)
- Would you date someone you thought was uneducated?
If you answered “yes” to the first three and “no” to the last, congrats: you’ve officially contributed to income inequality.
14 thoughts on “Education, IQ, and Mating”
Man, I hate the mic-drop. It’s just arrogant. It’s worse when the arrogance is unwarranted. For example, no one thinks they have no impacts on the world which increase inequality. What people aim for is to find ways to decrease their net contribution to inequality, not gross. Moreover, much of the resistance to inequality comes from the intuitive judgment that merit cannot possibly explain the inequalities we see, so inequality at current levels represents a corruption of the ideal of meritocracy by unequally distributing opportunity. But this piece is all about the unequal distribution of merit, not the fairness of the distribution of opportunities to develop, display, and be rewarded for merit.
You really want to pull out the mic drop after a piece which almost completely misses the point?
Walker’s point, I think, has more to do with those who like to signal how progressive and enlightened they are.
I’m not sure I follow you here. If intelligence is heritable (which seems very plausible) and if in a meritocratic society intelligence correlates strongly with IQ (which also seems very plausible) and if there is assortative mating (also very plausible) then the difference between a meritocracy and an aristocracy collapses. The end up being the same thing, because smart people marry other smart people and have smart children and those smart children (even if they have equal opportunities, other than genetics) will make more money than the dumb children of dumb parents. So you’ll get almost no social mobility (up or down) even though it’s a meritocratic society.
Now, if you does to make society unmeritocratic, you break this cycle. You could make people’s salaries random or the same (in theory). Well, on top of being unrealistic, those are probably unfair and certainly would destroy the incentives that economic activity depends on.
And you can’t really do anything at all about heritability of intelligence. That’s just genetics. It is what it is. (Unless you randomize people’s children, but that actually is addressing the third point, below.)
So the only thing you can do–if you want to avoid the meritocratic aristocracy–is discourage assortative mating. And that’s Walker’s point: if you embrace the idea of assortative mating, then you really are contributing to income inequality. And that’s definitely worth raising as a point, I think.
I’m not sure how anyone can hate the mic drop, but I suppose you’re forgiven.
Nathaniel already did a good job of addressing some of the complaints I had with your response. Assortative mating has been increasing over the last several decades and correlates nicely with increasing income inequality. Is it the only contributing factor? Of course not, but the clustering of highly-educated, affluent people through marriage and zip codes certainly plays a major role.
But another point of the piece–as I hoped to illustrate via the professor’s thought experiment–is that centrally planned attempts to social engineer the outcomes of inherently different people tend to be ill advised, whether it be through the redistribution of wealth or state arranged marriages via IQ.
And as Allen rightly pointed out, it was a slight jab at those who are more concerned with looking enlightened by “raising awareness” about income inequality and suggesting unimaginative policies to address it.
This is false, for at least three reasons. First, intelligence is not completely heritable–even in adulthood (when the effect is, surprisingly, greatest), there’s a lot of variance between generations. You’re absolutely right that there would be some expected continuity, but it would fall far short of aristocracy or your claimed “almost no social mobility”. Second, no measure of merit is perfect. Heavily biased systems are not only poor at measuring merit, they’re also poor at detecting the biases which keep them from improving, so one of the goals of those who find existing inequality so troubling is often to reduce those biases to the point where the system can usefully be used to diagnose its own problems. Third, while intelligence is correlated with merit, it isn’t the only relevant factor. A better meritocracy would more effectively identify sources of merit other than intelligence. Given that the cited article names intelligence as the most heritable behavioral trait, these other sources of merit would be even less heritable than intelligence, further increasing inter-generational economic mobility.
Even if it were true that the sources of merit are so perfectly heritable that perfect meritocracy would collapse into aristocracy, which several points from the article linked in the OP refute, this is still a massive leap. You can’t just consider the alternative of random or equal salaries and claim to have exhausted the options.
But keep in mind, ALL of this misses the point people are making when they complain about inequality. We’re not worried about aristocracy primarily, we’re worried about oligarchy (which, given peoples’ sentiments toward their children, does tend toward resembling aristocracy). Even a perfectly meritocratic society could be oligarchic if merit happened to be extremely unevenly distributed. What liberals tend to believe (largely for ideological reasons, so I’m not claiming perfect epistemology here) is that merit is much less unevenly distributed than the inequality in our income and wealth would suggest, and that equality of opportunity to develop merit would make this even more true. If that’s right, it strongly suggests that our methods of determining merit are badly biased, probably favoring those who already hold lots of power.
Yeah–that was foolish. If you want to make the point that central planning is usually ill-advised, introducing an ill-advised policy as the only topic of discussion simply assumes what you set out to demonstrate. It’s absolutely transparent straw person argumentation.
If your real point is just that those who make arguments to signal their values rather than to illuminate problems or explore promising solutions are bad, what exactly do you think you’re doing in this piece?
“If you want to make the point that central planning is usually ill-advised, introducing an ill-advised policy as the only topic of discussion simply assumes what you set out to demonstrate. It’s absolutely transparent straw person argumentation.”
But in the case of the professor, it *isn’t* introduced as a ill-advised policy. He introduces it quite seriously as a contender for an inequality-reducing policy. He first offers empirical evidence by citing studies of assortative mating, heredity, and IQ. He then counters objections by explaining the cultural and historical norms of arranged marriage (thus attacking the Western interpretation of “marriage for love”) and the plausibility of financial and educational data collection. The final objection given by his students is lack of freedom; an imposition that contributes to the deterioration of their personal flourishing. And *this* is what he focuses on. Now he has common ground to discuss the use of centralized power and social engineering in all its various forms. That’s hardly a straw man. That’s getting to the root of the problem.
“If your real point is just that those who make arguments to signal their values rather than to illuminate problems or explore promising solutions are bad, what exactly do you think you’re doing in this piece?”
Being a bit of an ass, admittedly. I get as much of a thrill as the next man having my biases confirmed. Nonetheless, I hardly think it is a bad thing to point out that many political beliefs are more about cultural signaling than empirical evidence: https://difficultrun.nathanielgivens.com/2014/06/09/evolution-scientific-literacy-and-culture-again/
“Nonetheless, I hardly think it is a bad thing to point out that many political beliefs are more about cultural signaling than empirical evidence”
And not such cultural signaling, but *virtue signaling*.
Your rebuttal to my point has a pretty significant flaw: you added the word “completely” to my argument (a word I did not include) and then rebutted something I didn’t say.
From my post (which you quoted)
A portion of your response:
My argument doesn’t rely on intelligence being perfectly, completely, etc. heritable. Just on it being largely heritable. Keep in mind that aristocracies were not perfect either. Even in an aristocracy, social status had some flexibility. You could be disowned and lose your high status, or you could be given a title and move up to achieve higher status.
The point is simply that to a great degree social status was passed down in an aristocracy. And if IQ is heritable and leads to social status and mating is assortative, a similar function will emerge. What’s more, this is only the kernel of the problem. Because then the high-class will cluster in similar neighborhoods, go to similar schools, etc. and so the effect is magnified and a genuine aristocracy will result in practice.
As you note, the two seem interchangable to a great degree. I’m not sure what the distinction is.
The really important thing, though, is that this argument about meritocracy –> aristocracy is largely of independent of details like the exact method by which merit is measured.
Assume–for the sake of argument–that we measure merit perfectly (somehow). IQ–and a great many traits like it–are heritable. And with assortative mating, they will all begin to cluster. Your argument that if other factors are less heritable that the effect will weaken is not true, because the effects will be cumulative. Suppose there’s also EQ (emotional intelligence, I don’t know if it’s really a thing, but let’s say it is) and suppose that merit is composed of IQ and EQ.
All this means is that people will mating on both IQ and EQ. The top performers (who have high IQ and high EQ) will marry other top performers, and they will pass down high IQ and high EQ to their kids at a greater rate than the overall populace. Of course it won’t be perfect. It doesn’t have to be. The point is that all the genetic factors of merit will cluster together, and that over time the populations will become more and more distinct.
The problem is actually quite serious precisely because it is robust. Even if you measure merit correctly, it doesn’t go away.
And that’s why this is such a compelling piece.
Um–right. That’s the point he’s after, the objection. He isn’t introducing the policy as an example which, in his view, is the best possible version of a centrally-planned method of addressing inequality. He chose a policy which violates freedom far more than necessary to achieve that goal, in order to cause students to recognize that there are other values than equality. And that’s fine–with students, sometimes a highly memorable method of making a point is productive. But the role that proposal HAS to play in order for the lesson to work as intended is to be so awful that students reject it. You recognize that, right?
Eh, okay. If I tell someone he’s being kind of an ass, and his response is to admit it, I can commend that. It just seemed weird to me that you’d write a post which isn’t a deep argument, but merely cultural signaling, and defend it as a critique of political writings which are more about cultural signaling than deep argument.
That’s a much more promising argument–you’re now claiming that functional equivalence to aristocracy simply requires a certain degree of inherited privilege. Here’s why I tossed in “completely”: the argument you initially presented made absolutely no attempt to support the claim that assortative mating would meet any particular threshold. It only worked if intelligence were completely heritable. All you wrote was that assortative mating happens, so smart people have smart children who make more money. If smart people only sometimes have smart children, the heritability of the privileges which accrue to intelligence may or may not rise to whatever the threshold of “effective aristocracy” is.
You simply can’t get there from “smart people have smart kids, and smart kids make more money”. This is particularly true because that statement is also accurate, on average, even without assortative mating! A smart person who marries an average person also tends to have smarter-than-average kids.
Not interchangeable; oligarchy leads to aristocracy, but not the reverse. A large class of privileged people will both tend to divide the benefits of privilege among them (thus leaving the Eloi closer to the Morlocks) and also better reflects the intuition that merit is much more flatly distributed than income currently is. Society pulls apart less the larger the aristocracy, and is more plausibly based on merit.
Not necessarily their sum. It depends on the bases people find important in their sorting. If it’s merit, then yes, this is absolutely a problem. But if it’s intelligence, then a high-IQ but low-EQ person (who might be high-merit) would tend to be more attracted to a low-EQ mate (who might be low-merit) than a high-EQ one. Moreover, there could well be entirely merit-irrelevant qualities which matter greatly for mate selection, which could introduce enough noise into the system to keep the heritability of privilege relatively low. Finally, what counts as merit changes as societies face different challenges. So long as the time scale on the development of this aristocracy is longer than a few generations, the drift in society’s needs is likely to be so great as to disrupt its formation entirely. Think about how different are the qualities which make one a success today compared with 100 years ago.
But leave aside all the ways it could be false, and assume your conclusion is true: that assortative mating and meritocracy leads to effective aristocracy. Isn’t it preferable to have a genuinely better, more useful set of families getting a little more power than a tiny number of families getting radically more power?
“He chose a policy which violates freedom far more than necessary to achieve that goal…”
Actually this would have to be demonstrated. You would need to show what policies that violate freedom less that are actually more effective at reducing inequality. But yes, of course there is shock value involved. But just because there is shock value involved doesn’t mean that one can’t make a serious argument in favor of it.
“Eh, okay. If I tell someone he’s being kind of an ass, and his response is to admit it, I can commend that.”
To be clear, this mainly has to do with my last bit where I say, “congrats: you’ve officially contributed to income inequality.” It’s a slight jab at those who care passionately about inequality from their educated bubble of affluence. I actually debated about whether or not to include the last part (Nathaniel and Allen can confirm that) because I was afraid the snarkiness might betray the larger point.
“…and defend it as a critique of political writings which are more about cultural signaling than deep argument.”
I suppose I could say that you pointing out that my virtue signaling via pointing out others’ virtue signaling is likely virtue signaling on your part. And you could respond that you just found the hypocrisy ironic, indicating that your virtue is still intact (and thus better than mine). But then I could counter by saying that I was basically doing the same thing. And it would continue ad infinitum.
But that would be boring and pointless. I’ll take the hit that the piece wasn’t as dispassionate as it could’ve been. Happens.
“Not interchangeable; oligarchy leads to aristocracy, but not the reverse. A large class of privileged people will both tend to divide the benefits of privilege among them (thus leaving the Eloi closer to the Morlocks) and also better reflects the intuition that merit is much more flatly distributed than income currently is. Society pulls apart less the larger the aristocracy, and is more plausibly based on merit.”
This argument isn’t making much sense to me. You don’t think that members of monied and connected elites have ever become part of an oligarchy when the balance of power and society have shifted to the latter?
“Finally, what counts as merit changes as societies face different challenges. So long as the time scale on the development of this aristocracy is longer than a few generations, the drift in society’s needs is likely to be so great as to disrupt its formation entirely. Think about how different are the qualities which make one a success today compared with 100 years ago.”
I feel entirely confident in saying that the same qualities making success likely today are the same qualities that would have made one a success one hundred years ago. The specific skills no, but things like education, connections, and funds?
Absolutely, but Professor Gill isn’t out to persuade people that this is even a legitimate option. I’m not saying he’s not giving the idea as generous a hearing as it can be given, but that’s my point–he’s chosen an idea so bad that, even presented as well as it could be, it sounds ludicrous to even those most deferential to centralized action.
I appreciate that–it’s a classy move. I also must own up to approaching everything on this site as though it’s written by Nathaniel. He’s professed a desire to welcome other views and treat them with respect, and I’ve been absolutely delighted to find someone who’s so effective and earnest about defending conservative values without belittling my liberal ones. That’s a great set of virtues, but it’s not the set every piece everyone writes needs to aspire to, and I registered such disappointment with this piece particularly because I was holding it to a standard I now understand you weren’t really reaching for.
Of course they have. If an aristocracy collapses into an oligarchy, members of the former aristocracy would have a huge head start in jockeying for a place among the oligarchs. But large aristocracies aren’t especially likely to shrink. Conversely, oligarchies seem to me likely to adopt features which make them resemble aristocracies.
None of those are heritable character traits, but you might have instead chosen grit, intelligence, and charisma and made a similar point which would have played the same role of defending Nathaniel’s claim that meritocracies tend to collapse into aristocracies given selection of similar mates.
There are a few nice counterexamples, though. For one, it turns out that attractiveness correlates with success, but which qualities count as attractive have changed a good deal over time. Another is particularly notable for women: the quality of being reserved. This is certainly less beneficial now, and possibly even harmful. Even intelligence was arguably as much a handicap as a benefit for many highly intelligent women of the time. Certainly, in any rotating scheme of qualities, some will remain common between any two points, but so long as there are substantial differences, the very definition of “merit” will change enough over time that it’ll be impossible for as slow a process as selective breeding in humans to converge on it in the way Nathaniel’s view seems to require.
Comments are closed.