Assortative mating and income inequality

by on January 27, 2014 at 11:16 am in Economics | Permalink

Jeremy Greenwood, Nezih Guner, Georgi Kocharkov and Cezar Santos have a new NBER Working Paper, the abstract is this:

Has there been an increase in positive assortative mating? Does assortative mating contribute to household income inequality? Data from the United States Census Bureau suggests there has been a rise in assortative mating. Additionally, assortative mating affects household income inequality. In particular, if matching in 2005 between husbands and wives had been random, instead of the pattern observed in the data, then the Gini coefficient would have fallen from the observed 0.43 to 0.34, so that income inequality would be smaller. Thus, assortative mating is important for income inequality. The high level of married female labor-force participation in 2005 is important for this result.

That is quite a significant effect.  There are ungated versions of the paper here.  Here is a related post by Alex.

Addendum: Kevin Drum offers additional comment.

prior_approval January 27, 2014 at 11:40 am

So, this compares household income, and assortative mating here would seem to mean that two working people are more likely to marry than stay single – explained by this sentence ‘The high level of married female labor-force participation in 2005 is important for this result.’

Somehow, the fact that married women contribute more when they are in a dual income household seems to be the sort of significant effect which is really, really obvious, when considering that people of roughly similar career backgrounds (university professor marries government lawyer, for example, regardless of which nation they grew up in) tend to marry more than people of disparate backgrounds – not too many delivery drivers are married to surgeons, for example.

But it isn’t as if this effect is really all that hard to research or see in action over multiple decades (centuries encompass a bit too much social change) – just look at how Beamten marry each other in Germany.

Marian Kechlibar January 28, 2014 at 2:06 am

I think the worst case of assortative mating is relatively recent, and that is the permanent welfare-addicted underclass. There is such thing in every rich European nation, much less so in the Balkans.

No one really knows what to do with it. Some deny its very existence (which is hard to do, if you don’t live in a bubble), others hope to turn them into a permanent voting block (which is hard to do, because the prevailing mindset there does not care about elections), yet others want to get them to work using various Hartz-like incentives (which is hard to do, because, as of 2014, lots of people are underqualified for any jobs in demand).

This is the dark side of the utopia. I’ve seen it all with my own eyes as a child. Growing up in Ostrava was a life-long immunization against most political theories.

Marian Kechlibar January 28, 2014 at 2:13 am

Interestingly enough, back then, the ruling Communist class was rather endogeneous as well. By 1989 there was the third generation of apparatchiks coming of age – marrying mostly among themselves.

JosieB January 30, 2014 at 12:13 am

This is about more than two incomes. I know an unworking doctor married to a finance professional. I know an unworking engineer married to a law partner. I am an unemployed MBA married to a working MBA.

How many factory workers, with or without jobs, do you know who are married to spouses with professional degrees? Waiters or waitresses? Home health-care aides?

MG January 27, 2014 at 11:49 am

Although the joke will likely revolve around soon-to-come Progressive strategies to ban assortative mating, I do look forward to measures aimed at nudging the process back to randomness. I expect the marriage tax penalty will not ever be reduced, and in fact, it could even be beefed up.

Doug January 27, 2014 at 1:14 pm

Clearly, we just need to expand the reach of Section 8 housing. Rich, sorry I meant “privileged”, people with nice, big suburban houses will be required to house ghetto single mothers and their demon spawn in the extra guest bedrooms. By living with the privileged this will help close the thirty million word gap and assure that the Section 8 residents learn all the racist terms on the SAT like boating references. And the benefits don’t just go one way either! The privileged oppressor will bask in the richness of experience of the wise Latina woman eating his cereal.

Such a system is sure to spring a million romances, the inevitability of living under the same roof. After all opposites attract. A whole slew of Prescotts and Ladonnas, Madisons and Billy-Bobs. All, decidedly non-assoritve, mating to help control this awful awful inequality problem.

We live in interesting times January 27, 2014 at 1:57 pm

That’s the plan.

T. Shaw January 27, 2014 at 12:01 pm

After 35 years of wedded . . . .

Here’s my advice to young men: Find a rich, nymphomaniac that (or whose father) owns a liquor store.

Otherwise, opt out.

Thrity-five years . . . three of the best weeks of my life.

Someone from the other side January 27, 2014 at 12:19 pm

I always thought marriage was an opt in thing….

Z January 27, 2014 at 1:04 pm

Unless you want to have children, marriage is a bad choice for men. Maybe there’s some benefit for women, but for men there’s zero benefit. I suppose this is why monogamy is rare amongst homosexual males. Without reproduction, or even the threat of it, monogamy is pointless.

msgkings January 27, 2014 at 1:12 pm

Except that married men report themselves happier than single men. And live longer.

The self-reported happiness ranking goes: married men happiest, then single women, then single men, then married women

Someone from the other side January 27, 2014 at 1:15 pm

You are missing divorced men in that list. And with approx 50% divorce rate, there is a significant chance that you migrate from married to divorced (unless of course they are hiding in single but that just sort of underlines the point about the data being bad).

msgkings January 27, 2014 at 1:26 pm

Not sure if divorced men are on there as ‘single’ or what….
But it seems like most divorced men get remarried pretty quickly. Less so with divorced women.

Norm Kurtz January 29, 2014 at 8:24 am

Correlation / causation?

msgkings January 29, 2014 at 11:37 am

Even if it’s not directly causal, the data clearly don’t support marriage = unhappiness for men.

Marie January 27, 2014 at 1:19 pm

Men usually die earlier, so it can be a benefit, having a caretaker. Of course, at that point it’s usually an old woman. . . .

Z January 27, 2014 at 3:44 pm

We die sooner than women because women nag us into the grave.

Marie January 27, 2014 at 6:25 pm

Well, if you’d just wash a dish now and then maybe your life expectancy could improve, but no, there it is, a full sink of dishes, and . . . (trails off into that teacher’s wah wah voice from Charlie Brown).

TMC January 27, 2014 at 8:44 pm

Marie, you’re a pretty good sport.

Errorr January 27, 2014 at 10:01 pm

Actually the life expectancy difference is highly skewed by the fact that men under 30 are significantly more likely to die. In reality by age 50 the life expectancy isn’t very skewed at all. Makes a lot of sense considering men are more likely to die violently, by suicide, or as the result of risky behaviors. Young men die and that effects life expectancy which is calculated at birth.

msgkings January 28, 2014 at 1:47 am

‘Errorr’ is right…
Go check out a nursing home…report back on the M/F ratio

Marian Kechlibar January 28, 2014 at 2:17 am

Errorr, you are in error.

At any age, males are more susceptible to cancer, heart problems etc. than females. The incidence of heart attacks in 50 y.o. male population roughly corresponds to those of 60 y.o. females.

Have you never heard of actuarial science? Insurance companies have every incentive to try to know the truth, and their results are pretty much the same around the developed world, regardless of culture, diet and climate. Women are much more tough than men when it comes to health.

Doug January 27, 2014 at 1:32 pm

The primary between marriage and long-term cohabitation is that separation is more expensive, consuming and difficult in the former. That is being married serves as a stronger commitment and is more likely to keep people together. Most divorces and breakups are initiated by the female partner, so making it harder to separate is statistically of more benefit to men than women.

You’re conflating marriage with monogamy. The vast majority of those who long-term cohabitate are also completely monogamous. So marriage isn’t the issue. You’re right that the female is almost always the strongest demander of monogamy (which is why male homosexuals are not monogamous). The vast majority of females will demand a monogamous commitment before entering a sexual relationship.

Unless you have enormous mating market value, as a heterosexual male you are pretty much stuck with serial monogamy. The primary exceptions are if you’re enormously wealthy, unbelievably attractive and/or willing to devote nearly all free time to chasing women. There simply aren’t that many women who are cool with you getting your rocks off with other women at the same time.

Rahul January 27, 2014 at 1:51 pm

Can you explain why it is pointless? In general, people seem to value stability and are risk averse. It’s like saying sticking with your business partner or joint venture or research collaboration is pointless. You carefully ponder before jumping in & then you try to stick barring extraordinary circumstances. The risk & transaction cost associated with search & switching makes non-commitment expensive.

Marian Kechlibar January 28, 2014 at 2:08 am

I think your advice misses an important development. If your true love’s father owns a MJ farm, I can see a happy marriage in front of you (and longer life – much more so than with the liquor store).

Axa January 27, 2014 at 12:04 pm

How strong is the correlation between marriage and mating? Authors talk about data on Page 8: “Marital status: married, never married males, never married females, divorced males, divorced females.”

I don’t see anywhere cohabitation and the US census said that approx 1 in 3 unmarried woman who give birth cohabits with the partner. Problem is, this statistic started in 2010.

JWatts January 27, 2014 at 12:10 pm

That’s a good question. But regardless of the answer, I would suspect that those who are “mated” (cohabitate) but not married tend to keep their finances distinct. At least that’s been my observation.

Frederic Mari January 27, 2014 at 1:02 pm

but the effect on poverty and inequality should still be present as they would share household expenses and cohabitation still occurs along social lines (i.e. it’s assorted)…

Hopaulius January 28, 2014 at 1:02 am

A lot of this data comes from the IRS, where there is a sharp divide between married and cohabiting partners. Let’s say the couple makes $20k and $30K respectively. If they are married, this will show up as household income of $50k. If they are unmarried but cohabiting, it will show up as two households, one making $20k and one $30k. Assuming the woman makes the $20k and has two or three kids, she’ll show up in poverty, but not if they are married.

Frederic Mari January 28, 2014 at 5:36 am

I understand but, in which case, it’s more a problem of relying on IRS data and definitions than ‘true’ poverty.

Matt Yglesias was making a good point when, to conservatives who say that marriage is the answer to fight poverty, he replied that cohabitation and the sharing of household expenses (as practised by students in many expensive cities across the world) works just as well…

Marian Kechlibar January 28, 2014 at 2:22 am

If there is a welfare state present, there is an incentive to keep the finances distinct at least officially, because being in the right poverty bracket means to get more money. A quasi-family with double income would not qualify for as much.

There is a trend my home country (the Czech republic) in the welfare class to proclaim paternity of a newborn as unknown, even though the father is really known and often lives in the same household for years. That is because fatherless children get child support directly from the state. So far, there was no effort on the part of the state to crack down on this outright abuse.

Dollared January 27, 2014 at 12:42 pm

It’s pretty easy to lower the effect, you just need to slow the effects of sorting-for-like-kind in the critical 16-30 age range for mating:

1. Affirmative action in college and graduate school enrollment – it always was a sound policy, now we have another reason.
2. Elimination of legacy admissions – talk about sound policy!
3. Universal service requirement – bring all citizens together in a common project and environment at the optimal age.

Of course you can never eliminate it, but these are all sound policies that support national meritocratic advancement and better selection of talent for advancement, while reducing the sorting effect of private schools, expensive hobbies, expensive camps, expensive post-college neighborhoods, fraternities, sororities, racially sorted religions, etc.

Yancey Ward January 27, 2014 at 12:49 pm

I just love how supposedly liberty loving liberals support coerced national service, and the funny thing is that almost to a man/woman, they are already too old to be a part of it.

mulp January 27, 2014 at 2:37 pm

So, you don’t support coerced work to get welfare?

Or even coerced work to have housing and food?

JWatts January 27, 2014 at 3:14 pm

The stated argument was a coerced “Universal service requirement”.

Marie January 28, 2014 at 12:34 pm

Also, in neither of those situations are you forced to work for the government.

Yet. ;)

David S. January 27, 2014 at 12:53 pm

To enhance the effectiveness of the universal service requirement, clothing should be banned and drinking should be mandatory.

dearieme January 27, 2014 at 1:07 pm

I think you’re confusing it with going to university.

dollared January 27, 2014 at 1:08 pm

Agreed! Although the Disney movie about this will take a great deal of strategic shooting….

Mark January 27, 2014 at 1:05 pm

Arrow proved that to achieve perfect income inequality it is necessary to eliminate individual liberty completely. I see the ” issue” of assortative mating as a perfect illustration of that principle. I put issue in quotes to emphasize that from a non-progressive pov it isn’t an issue at all becuase it is no one else’s business whom one marries.

dollared January 27, 2014 at 1:16 pm

“proved.” Arrow is truly the equal of Pythagoras….

Mark January 27, 2014 at 5:26 pm

It is in fact a proof worthy of Euclid or Pythagoras. But let me remind you that neither of them won the Nobel.

Yancey Ward January 27, 2014 at 12:52 pm

This a trend that is essentially irreversible at this point of the information age. It will be come increasingly easy for assortative mating to occur and intensify.

Mark January 27, 2014 at 1:07 pm

And conversely – if a progressive wanted to search singles databases for a non-assortative mate, s/he could easily do so.

Yancey Ward January 27, 2014 at 1:12 pm

You see, you should have put quotation marks around “wanted”.

msgkings January 27, 2014 at 1:04 pm

How much mating was ever really non-‘assortative’?

Did rich folks mary poor ones a lot before? Seems like it was pretty assortative then too. Women didn’t work nearly as much outside the home but men and women still met and married within their class pretty much…

dollared January 27, 2014 at 1:11 pm

There was the golden age – I understand that from about 1940 to 1980 princesses were freed from the requirement that they marry princes.

An Onyx Mousse January 27, 2014 at 1:18 pm

In the past, there was more non-assortative mating of a certain sort. Men often married “down’ in terms of human capital, if we exclude beauty/charm from the human capital assesment. (e.g. executives married pretty secretaries). Women’s earning power was not considered as much of a marital asset for folks marrying prior to the high levels of female workforce participation we have seen recently. There has been much more sorting of young people by human capitall in the college & work systems than in the past due to more efficient labor markets and college admissions. Also the U.S is tending more and more to group into geographies sorted by human capital (think Silicon Valley vs. Rust Belt). And the earning power of the top 5% is much more separated from the pack than the past, so marrying another top 5%-er has much larger effects on income distribution than in the past. A lot of “Average is Over” type analysis is relevant here.

dollared January 27, 2014 at 1:23 pm

all excellent points.

msgkings January 27, 2014 at 1:30 pm

‘Of a certain sort’ implies to me it wasn’t that common. And most men were not executives. So middle class men married middle class women, etc.

Granite26 January 27, 2014 at 2:14 pm

I guess the point is that you can’t have assortive mating if only 1 member of the pair is generally sortable by the metric you’ve chosen.

More importantly, see the point about geographic sorting. The rise of uber-elite enclaves means that it is far more possible for upper middle class to find partners of comparable status/income. In the past, if you were generally well-to-do in a mid to small town (think lawyer or doctor or dentist or whatnot), your dating pool was ‘the people in the town you were in’, which was unlikely to include a large number of equal status’ed people. Most importantly, it wasn’t going to include enough equal status’ed people for you to isolate yourself socially.

With the increasing urbanization of the country, and especially the increasing concentration of upper middle class wealth, those rules changed.

This is all very Charles Murray

albatross January 27, 2014 at 5:00 pm

Approximately summarizing the argument in The Bell Curve:

Being rich is correlated with being smart (it’s easier to get and stay rich if you’re smart, and having rich parents gives you whatever environmental advantages are available toward becoming smart). But the correlation isn’t all that strong, whereas there’s a much stronger correlation between getting really good SAT scores and being smart, or getting really good grades and being smart.

During the 20th century, we had two huge transitions:

Early in the 20th century, going to college (especially an elite college like Harvard) was something you did because you were a rich young man from the right kind of family. There were also super bright people who got to those top colleges despite being from a modest background, but the main drivers were money and family. As the 20th century went on, that became much more meritocratic, until eventually, almost everyone at an elite college is there because they got really good SAT scores and had really good grades. There are legacy and affirmative action admits who aren’t quite so impressive, but they’re not the majority, and even they tend to have pretty good grades and scores, just not quite as good as the average students.

Early in the 20th century, few women went to college or into intellectually demanding professions with many men in them. Again, this wasn’t 100%–there were extremely driven and smart women who went into law and medicine and science and economics, but they were the exception. Over the course of the century, this also changed, and now women go to college and go into intellectually demanding professions all the time.

The result of these two is a great setup for associative mating: From high school on, you are increasingly put together with other people who are about as smart and driven as you are. You’re in the same classes, circles of friends, colleges, occupations, companies, etc. That just wasn’t present to anything like the same degree in 1910.

Granite26 January 27, 2014 at 2:14 pm

I guess the point is that you can’t have assortive mating if only 1 member of the pair is generally sortable by the metric you’ve chosen.

More importantly, see the point about geographic sorting. The rise of uber-elite enclaves means that it is far more possible for upper middle class to find partners of comparable status/income. In the past, if you were generally well-to-do in a mid to small town (think lawyer or doctor or dentist or whatnot), your dating pool was ‘the people in the town you were in’, which was unlikely to include a large number of equal status’ed people. Most importantly, it wasn’t going to include enough equal status’ed people for you to isolate yourself socially.

With the increasing urbanization of the country, and especially the increasing concentration of upper middle class wealth, those rules changed.

This is all very Charles Murray

JWatts January 27, 2014 at 4:17 pm

‘Of a certain sort’ implies to me it wasn’t that common. And most men were not executives. So middle class men married middle class women, etc”

“Class” is the wrong terminology for this and is clouding the issue. This isn’t about class, it’s about IQ and education.

Previously, with small amounts of the population attending college, most people married someone living in their local area. IQ was somewhat randomly distributed and you ended up with certain percentage of high-high IQ, high-low IQ and low-low IQ marriage. By high-high IQ, I mean someone with greater than 100 IQ marrying someone else with greater than 100 IQ.

Currently, 40-50% of students attend college. Attending college is highly correlated with IQ and people who attend college are more likely to marry someone who also attended college, than someone who did not. Therefore, you have both more high-high IQ and more low-low IQ marriages.

JWatts January 27, 2014 at 4:18 pm

See Charles Murray.

Mo January 27, 2014 at 3:06 pm

Were there many 23 year old executives? Because that was the median age of marriage for men back in the 50s and 60s.

zbicyclist January 27, 2014 at 9:57 pm

I’m not sure I agree. Sure, the income of the female may have been less in that stereotype period in which husbands worked, wives didn’t. But it seems to me that most marriages were within the same class. A woman’s earning power might not have been as much of a marital asset, but her charm and ability to run a household in the style the man would be accustomed to (i.e. the way he was raised) were still assets, albeit harder to measure in economic terms.

So is this at least partly just an effect of the fact that in a given social class, men and women are now both likely to have similar income prospects?

Steve Sailer January 27, 2014 at 2:59 pm

My mother was a secretary at Lockheed during WWII. She married my dad, who was a pretty average guy. My mom’s best friend, another secretary, married a brilliant engineer, “an irascible genius” according to Ben Rich’s “Skunk Works,” who became the chief designer of the Mach 3 SR-71.

Granite26 January 27, 2014 at 1:06 pm

I would be interested in looking at assortive mating compared with gini coefficients across countries (and cultures). Even better if there’s a third variable for wage-differential between males and females in general.

Is this an American trend? (I doubt it). Is it Western? First World?

Fun fact: According to Wikipedia, .34 is lower than ANY of the OECD countries

David Autor January 27, 2014 at 1:09 pm

“If matching in 2005 between husbands and wives had been random, instead of the pattern observed in the data, then the Gini coefficient would have fallen from the observed 0.43 to 0.34, so that income inequality would be smaller.”

In what sense is this a relevant counterfactual? I got married in 1995, for example, and I don’t recall that matching was random at the time.

Z January 27, 2014 at 1:12 pm

Women are wired to seek out and attract high status males as mates. Humans learned that just after Grog invented the wheel and started getting all the hot cave tail. Males will select females that are most attractive to them based on the cultural norms of their society. To the shock of only university researchers, the result is the best of most successful of each sex pair off as mates.

Dan S January 27, 2014 at 2:04 pm

I think you misspelled Gronk. Still a caveman as far as we can tell.

Humor gets me through Monday.

Mo January 27, 2014 at 3:08 pm

That’s an insult to the intelligence and creativity of cavemen.

JWatts January 27, 2014 at 4:21 pm

Indeed, Grog invented the boiled grain party mash.

dollared January 27, 2014 at 1:21 pm

The other obvious factor here is the differential in pay that has widened in the last 3 decades. With 40% of the jobs within 50% of the minimum wage, two top 5% incomes are now far, far apart from two median incomes.

Marie January 27, 2014 at 1:23 pm

I don’t think this is really about economics or income, maybe not even class, in themselves.

In today’s narcissistic world, tons of people naturally want to come as close as they can to marrying themselves (I could say mating with themselves, maybe more accurately).

Z January 27, 2014 at 1:29 pm

+1

I’ve been told to bleep myself many times. I used to take offense, but now I see it is just the thing to do.

The Anti-Gnostic January 27, 2014 at 4:21 pm

I doubt there’s ever been a period in history when any more than a small percentage of people married outside their family’s social class or outside 5+/- IQ points.

Careless January 27, 2014 at 5:18 pm

So you think it was rare for a 97 and 103 IQ to marry? I’m going to bet against that.

The Anti-Gnostic January 27, 2014 at 8:46 pm

I’m going to bet it’s a Gaussian distribution with the majority of marriages falling somewhere in a + or – 5 point spread either way, which would be the implied context for my comment anywhere but here at Sperg Central.

Cliff January 27, 2014 at 11:03 pm

Laughably wrong and absurd (both this comment and your original one)

The Anti-Gnostic January 28, 2014 at 11:34 am

How many smart men do you know who married dumb women and vice versa? Do you really think this is a new phenomenon? I think you’re conflating female workforce participation with female IQ.

Cliff January 28, 2014 at 3:58 pm

Five points of IQ is a smaller difference than you realize if you think no more than a “small percentage” of marriages have involved partners differing in IQ by more than 5 points throughout all of human history. I would bet the majority of marriages involve partners differing by more than 5 points of IQ right now.

The Anti-Gnostic January 29, 2014 at 1:41 am

Okay. How about six points? Seven?

Did you take your meds today?

Marian Kechlibar January 28, 2014 at 2:27 am

That is a pretty big “OR”, because the classes are / were huge (peasants formed some 90 per cent of Euro population until the late 1800s), so the statement is too wide to be useful.

A Definite Beta Guy January 27, 2014 at 10:26 pm

True, Marie, but unfortunately both God and Sheldon Cooper are fictional characters.

Marian Kechlibar January 28, 2014 at 2:32 am

I don’t believe that, Marie; such a pairing would be much less stable and much more prone to conflicts than one in which the partners have complementary qualities.

But people do want to understand their significant other a little, and being from similar milieus helps a lot – shared experience etc.

A thought experiment: imagine that your husband/boyfriend was the same person as it is today, but with one big difference: he would be a time traveller from 1614. Do you think you would find it much harder to resonate with each other? Finding common themes and values? After all, the culture back then was much more alien to today than China is. Burning witches at stake was almost as normal as getting a parking ticket.

Marie January 28, 2014 at 9:29 am

I was being a bit tongue in cheek, but I do think might be one trend, Marian. And no, I don’t think it’s one that leads to stability. I think it’s a product of extended immaturity and social dysfunction. I think it’s responsible a bit for the “we just grew apart” divorce or breakup themes — I think sometimes folks don’t grow apart so much as discover that the spouse is more of a whole other person than was previously thought.

To a degree, I think young people are taught that a spouse (and children, for that matter) is an accessory, another form of branding. And you brand with your preferences, to show your preferences to the world, it’s all about you, there is no other as an actual person.

I also think the trend to celebrate conformity prevents checks on the trend — conformity is good, we’re all conforming to this one (or maybe these two or three) ideal (s), so if I’m doing it well and my potential spouse is doing it well we’re pretty much going to be close to being the same person.

I think both these things would factor into the idea of assortive mating, since if you are picking a brand or co-conforming to a social ideal you will select your mate differently than you will if you are trying to find a true other whole human being to have a relationship with.

But I do kind of sit in my living room chair and make things up, so I could be way off. Maybe I don’t even believe myself. It also could just be that we’re all, and always have been, prone to self-centeredness and the older couples I see are the ones who have overcome that and survived, while the younger couples are still getting there, and some have dropped out.

Your example is funny to me, though, since my husband and I are both pretty much throw-backs, though in different ways, and probably different centuries!

The Anti-Gnostic January 28, 2014 at 11:39 am

I think sometimes we give ourselves a little too much credit. We just choose different targets for our ire and express our brutality in more antiseptic ways. Affluence enables a lot of gentility.

Marie January 28, 2014 at 11:57 am

I read that as referring to my fairly coldly brutal and probably unfair critique of modern young Americans, then realized you may have meant the witch hunt comment.

Which I think proves your point. Which I agree with.

Turkey Vulture January 27, 2014 at 2:15 pm

I got married in 2005, so this is my fault.

Fred January 27, 2014 at 2:26 pm

Assortative mating is probably largely irrelevant to the recent gains made by the top .1 percent, but it clearly plays a big role in household income distributions for the rest of us. Two facts here are especially relevant: 50 years ago, wives in higher income families were far less likely to be employed than wives in low income families: that has been reversed; low-income households are now much more likely to have an unmarried head. Those two simple facts go a long way to accounting for changes in household income distributions over this period for 99.9 percent of the population. For the top .1 percent, one can account for a lot of the change reported from PIT figures on the basis of increased pass-through income on the part of proprietors and financial professionals and the timing of income recognition. Nevertheless, over the past 10 years we have seen a change in wealth distributions at the very top, which does suggest a real and significant change. Unfortunately, we don’t really have very good measures of wealth. Still changes in the ratio of wage income to capital income, which had been constant for a very long time and which tend to confirm that the distribution of wealth has become more unequal, hence the renewed interest in wealth taxation.

Bill January 27, 2014 at 2:27 pm

If you are really wealthy on your own, you don’t need your wife to work. In fact, that your wife doesn’t work is a sign of status. So, I would be interested in looking at how, at the top income levels, labor force participation by females compares to middle class households.

Jmo January 27, 2014 at 6:47 pm

In fact, that your wife doesn’t work is a sign of status.

Is the status of the Chief of Surgery at UCSF Medical Center lower if his wife is a partner at Kleiner Perkins? If say it’s higher than if she was a SAHM.

Bill January 27, 2014 at 7:07 pm

You obviously work for a living and submit W-2s. The folks I know do better than that.

Cliff January 27, 2014 at 11:05 pm

Well, now you are talking abut a really tiny segment of the population

Jmo January 27, 2014 at 11:23 pm

But, I think you’ve hit on a source of income inequality. It’s no longer socially acceptable for the rich not to work. For thousands of years, the goal was to be able to live in luxury without working.

You’ll recall the Dowager Countess’s comment, “What’s a weekend?” It was at one time rude to mention the weekend, as it implied you had something to do during the week.

Marian Kechlibar January 28, 2014 at 2:46 am

Your view is limited to the USA. It is still pretty common for the rich not to work across much of the world.

Finch January 28, 2014 at 12:54 pm

No, it’s lower. Those careers pretty much imply the marriage is for appearances, as the couple will never see each other.

Bill’s sort of right.

Steven Kopits January 27, 2014 at 2:31 pm

Do any of you actually like your significant others?

We live in interesting times January 27, 2014 at 5:30 pm

25 years.

zbicyclist January 27, 2014 at 10:02 pm

30 happy years (38 in total).

anne January 27, 2014 at 2:31 pm

Pretty sure Charles Murray covered this in The Bell.Curve a couple decades ago. And yes as information has become.more prevalent and mobility has increased of course people mate with their equals/level. Oh and those hot chicks in the 40s/50s/60s who married up? Just because they may not have had a college degree doesn’t mean they weren’t smart.

Steve Sailer January 27, 2014 at 2:54 pm

When the first wife of Kelly Johnson, the genius aircraft designer who founded Lockheed’s Skunk Works, was dying, she instructed Kelly to marry his secretary. When, in turn, his second wife was dying, she instructed Kelly to marry her best friend, another secretary.

Nowadays, Kelly would have remarried a Harvard MBAette he met at Davos. Well, maybe Kelly Johnson wouldn’t have been caught dead at Davos, but you get the general picture.

David Wright January 27, 2014 at 4:24 pm

Drum makes a reasonable point that, in terms of explaining some part of the rise in income inequality over the past 50 years, the relevant counter-factual isn’t random mating but mating patterns 50 years ago. It would be nice to go a bit further and actually compute the fraction of the change in inequality attributable to these changes in mating patterns. (Contrary to Drum’s suggestion, it doesn’t just depend on the change in the diagonals.)

I do find it a bit weird that Drum says that “the real drivers of income inequality” are those that “have driven up the incomes of the top one percent so stratospherically”. In Gini coefficient terms, changes in incomes of the last 1% aren’t all that important — there just isn’t enough area there to change the coefficient much. To account for Gini changes over the last decades, you really need to include the increased incomes of the top 20% at least, i.e. people like Kevin Drum. And from a leftist social justice standpoint, shouldn’t the real worry be about stagnating incomes of the bottom 20% — do we need massive policy interventions to prevent the top 1% from pulling away from the top 2%?

Urso January 27, 2014 at 5:23 pm

I thought the same thing about Drum’s article, which seemed a little dashed-off as opposed to thoughtfully considered.

You make an interesting point about the 1%. It seems that the prevailing Mother Jones counterfactual is that, if the 1% hadn’t made such stratospheric gains, those gains would’ve been equitably distributed across the spectrum. But what if the counterfactual is that those gains would’ve been distributed among the 2-5%. Is that a big difference?

albatross January 27, 2014 at 5:05 pm

I wonder how associative mating affects the building of informal networks of people, whatever the modern form of the old boys’ network is. My intuition is that we should see the networks of friendships made at top colleges shifting over the years, from networks of rich people from good families, to networks of smart people related to other smart people.

clayr January 27, 2014 at 6:08 pm

I would posit an additional factor in this is that the needs of a marriage have changed. With fewer children and more job opportunities, women are less likely to demand a provider and more eager to find someone they get along with well — which is apt to be someone close to them in terms of socioeconomic status. By the same token, men are more likely to give more weight to factors other than homemaking, since many don’t expect children (at least not for a few years down the road), and thus tend to look more for someone they find to be a good companion — which also leads them to someone of similar status.

Turkey Vulture January 27, 2014 at 8:19 pm

What if we suppose, instead, that assortative mating has been going on to essentially the same extent for the past 100 years – i.e., men and women with similar theoretical earnings potential have been pairing off. But, until a few decades ago, there was a significant impediment to one half of that paring – the female – reaching her theoretical potential wage. Now, John and Jane are married just like they would’ve been in 1950, but instead of John being able to work as a lawyer and Jane capped at a nursing or teaching wage (which I would further expect to induce labor force exit at a higher rate than among two low-wage-earners who need both incomes to get by), both can work as lawyers. Those couples with women who were previously below their theoretical potential will now earn more. Household income (and wealth) inequality will increase – the top, say, 20% will pull away from the rest.

This is certainly a part of the story. I have no idea how much.

Jay January 27, 2014 at 9:00 pm

Is there a human dumber than Kevin Drum? Just one. And you can’t use Ezra Klein.

CBBB January 28, 2014 at 4:49 am

Actually they’re both orders of magnitude smarter than you. So I name you.

David+ January 28, 2014 at 12:57 am

Who was that musician who said “getting married for sex is like buying 747 for the free peanuts”?

The marriage thing is nuts but men can’t compete with females when it comes to getting what they want. Females are smarter, period.

David+ January 28, 2014 at 12:58 am

Buying a 747…..(as in the Boeing jumbo jet)

Panskeptic January 28, 2014 at 2:12 am

What a lot of borscht!

Increasing income inequality is because those with friends at court have juggled the tax rates to their advantage, fired as many underlings as they can, then give themselves bonuses for their financial acumen.

Go assortative mate yourself!

Marian Kechlibar January 28, 2014 at 2:45 am

I’ve seen the effect of assortative mating even in nerdy fields such as pure mathematics, where there are no financial incentives to do so. Plenty of algebra postdocs marry each other, while their friends shower them with paper Möbius strips on the church stairs. And there is no serious movement to claim that the resulting disproportions of mathematical abilities in the population are a result of insidious manipulating of the exams.

But, well, let us have a mental experiment: let us presume for a moment that what you say is truth and the complete truth with no other confounding factors, such as existence of global trade or the Internet.

Now the question is: do you rule out the hypothesis that mental abilities necessary to “juggle tax rates / get to the business top / getting away with large bonuses” in long term are inheritable?

Because if they are, you’ll still have the assortative mating effect.

Marie January 28, 2014 at 9:46 am

Ah, this is one of the big catches on the whole income inequality debate, isn’t it?

Do people get drastically richer because they are smarter and more industrious (so the rich are rich because they enrich, they create value so fairly have a greater share of it) or because the game is rigged (and the rich are rich because they are unproductive but clever plutocrats)?

And the answer, of course, can certainly be, that in a society with absolutely no moral or ethical underpinnings, the smarter people are going to be the ones who can and do rig the game; that you would be by definition foolish if you spend your resources adding value you will not share well in (because the other clever people are spending their resources making sure the value you add comes their way) instead of working to rake what value is already out there in your direction.

Are we there? Hope it’s still a mixed bag, at least.

albatross January 28, 2014 at 11:06 am

Also, if you’re smarter and more industrious mainly because of choosing your grandparents wisely, it’s not clear to me why massive social inequality caused by smarter, more industrious people doing better and rigging the game in favor of the smart and industrious is in any way morally superior to massive social inequality caused by people with inherited titles of nobility rigging the system to ensure that other people with inherited titles of nobility do well.

Meritocracy justifies inequality in functional terms if it works well, and perhaps this does work pretty well. But in moral terms, meritocracy can only justify inequality if the merit involved came from some moral superiority of the people who succeed, not if it’s just good genes.

Marie January 28, 2014 at 11:28 am

Good point.

Jay January 29, 2014 at 6:20 am

Why don’t we abduct all children from their biological parents and put them in homogenous households so their opportunities are “=”.

Marie January 29, 2014 at 9:55 am

Jay, I think you’re missing my point.

I don’t mind people of different abilities having different opportunities, or providing them for their kids. Rising tide and all.

But If the tide is slamming us into a cliff, that I mind. And the solution is not to sort mates for people or kidnap kids, it’s to turn back in the direction where our smart and able people spend their smarts and ability on doing stuff, not on strengthening and solidifying their position and their family position in a Byzantine plutocracy.

This article sums up for me what we’re doing with our merit class. As opposed to other times and places, when our merit class was, I don’t know, walking on the moon and inventing light bulbs.

http://thefederalist.com/2014/01/28/the-clueless-self-importance-of-dc-captured-nicely-in-one-simple-tweet/

Floccina January 29, 2014 at 9:56 am

Men only care about looks so is the problem pretty women not caring about getting a high earner?

MBA Data Guru February 11, 2014 at 9:21 am

Intelligence is only one component of a successful person who will earn a lot of money. Social skills (emotional intelligence) is just as important or more important. Who do you think earns more? The math phd or the partner at a consulting firm who knows office politics and can sell the company’s services to clients? Who do you think has a higher IQ?

If you think the people who are rich and successful add no value then you are an idiot. People always call out investment bankers as people who add no value but without them there would be no money to start new businesses like Facebook and companies couldn’t raise money to expand and hire more employees.

Comments on this entry are closed.

Previous post:

Next post: