Dept. of Uh-Oh, college and mobility edition

Intergenerational mobility is higher among college graduates than among people with lower levels of education. In light of this finding, researchers have characterized a college degree as a great equalizer leveling the playing field, and proposed that expanding higher education would promote mobility. This line of reasoning rests on the implicit assumption that the relatively high mobility observed among college graduates reflects a causal effect of college completion on intergenerational mobility, an assumption that has rarely been rigorously evaluated. This article bridges this gap. Using a novel reweighting technique, I estimate the degree of intergenerational income mobility among college graduates purged of selection processes that may drive up observed mobility in this subpopulation. Analyzing data from the National Longitudinal Survey of Youth 1979, I find that once selection processes are adjusted for, intergenerational income mobility among college graduates is very close to that among non-graduates. This finding suggests that expanding the pool of college graduates per se is unlikely to boost intergenerational income mobility in the United States. To promote mobility, public investments in higher education (e.g., federal and state student aid programs) should be targeted at low-income youth.

That is from new research by Xiang Zhou, via the excellent Kevin Lewis.


This definition helped me, it might help you: "Intergenerational income mobility refers to the extent to which income levels are able to change across generations. If there were no intergenerational mobility at all (that is, the intergenerational income elasticity was equal to 1), all poor children would become poor adults and all rich children would become rich adults.". Once again the Scandinavian (socialist?) countries have very good low elasticity, meaning poor kids don't stay poor. The UK, Italy and US are about 4 times higher (worse) .

Most Scandinavian countries are highly capitalistic. They balance that with generous social safety nets.

Definition of socialism may help you:
any of various economic and political theories advocating collective or governmental ownership and administration of the means of production and distribution of goods.

@Armin @EdR - in the USA, 'collective bargaining' and "strong social nets" are synonyms for "socialism", rightly or wrongly. And do the math: if 10% of Denmark are immigrants, in the USA, it's not much different, around 13%. In Switzerland it's much higher, about 25%. - RL ("In their paper "The Scandinavian Fantasy: The Sources of Intergeneration Mobility in Denmark and the U.S.", Rasmus Landersøn and James J. Heckman compared American and Danish social mobility and found that social mobility is not as high as figures might suggest in the Nordic countries")

The issue is the nature of the 10%. You can have 10% milk in your coffee or 10% gasoline.

They are not the same.

"Once again the Scandinavian (socialist?) countries have very good low elasticity, meaning poor kids don't stay poor. The UK, Italy and US are about 4 times higher (worse) ."

Most Scandinavian countries don't pump millions of poor, illiterate people into their country every year. They also are starting from an advantaged position.

Consider Denmark, a socialist favorite. The population of Denmark is about 6 million - roughly the size of the SF Bay Area. The population is about 86% English/Danish bilingual. About 10% of the population are immigrants.

From Wikipedia:

"Denmark is a historically homogeneous nation.[165] However, as with its Scandinavian neighbours, Denmark has recently transformed from a nation of net emigration, up until World War II, to a nation of net immigration. Today, residence permits are issued mostly to immigrants from other EU countries (54% of all non-Scandinavian immigrants in 2017). Another 31% of residence permits were study- or work-related, 4% were issued to asylum seekers and 10% to persons who arrive as family dependants.[166] Overall, the net migration rate in 2017 was 2.1 migrant(s)/1,000 population, somewhat lower than the United Kingdom and the other Nordic countries."

"There are no official statistics on ethnic groups, but according to 2018 figures from Statistics Denmark, 86.7% of the population was of Danish descent, defined as having at least one parent who was born in Denmark and has Danish citizenship.[11][N 5] The remaining 13.3% were of foreign background, defined as immigrants or descendants of recent immigrants. With the same definition, the most common countries of origin were Turkey, Poland, Syria, Germany, Iraq, Romania, Lebanon, Pakistan, Bosnia and Hercegovina, and Somalia."

Like the other Scandinavian countries , Denmark has begun an immigration experiment. Let's watch that experiment run for awhile and then examine the outcome.

One thing Denmark does not have, just like Sweden and Norway, is a long defenseless border with a third world CONTINENT.

Let's see what happens.

And the Inuit do not do so well in Denmark.

I think the Sami also lags a bit in Norway and Sweden and Finland.

Countries with lower inequality will also tend to have higher measured mobility. The smaller the difference between the bottom quintile and the top quintile, the easier it is to move between them.

This leads to some strange results. For one thing, Europe includes rich countries and poor countries so inequality in Europe as a whole is higher than in most individual countries, and as a result mobility in Europe as a whole is lower than in most individual countries. The U.S. may have lower measured mobility than most European countries, but it may very well have higher mobility than Europe as a whole.

Okay, now control for those who graduate with a degree that's worth a damn vs those who don't. Any analysis of college education is totally useless if all majors are treated as equal.

I think this is relevant because many people think that universities have economic value that's not just signaling and not just specific to the useful majors.

+1 on this. It kills me when I see a first generation college kid getting a crap degree.

President Captain Bolsonaro has ordered a 30% cut across the board federal universities funding. He has ordered the Minister of Education to direct money for useful courses, such as STEM courses and Medicine. Brazil is expected to double its blockchain capabilities and successfully launch an artificial satellite before next decade ends.

An artificial satellite? Instead of a real one?

I wouldn't be surprised if Thiago credits Brazil for the existence of the moon.

Artificial satellites are objects intentionally placed into orbit. They have many military, scientific and economical applications.

I get this, but also a lot of first-generation students aren't super-qualified for college and most likely would not have gotten in barely more than a generation ago. They're not going to do well in pre-med or engineering courses; if they tried those paths, they would likely drop out a lot sooner for failing grades. You can say that that's how it should be, because they shouldn't have been there in the first place and now maybe they won't waste several more years and debt getting a worthless degree, but when they have the message rammed down their throat from elementary school onward and colleges are under pressure to retain first-gen students, they will take the type of class they think they can do well in.

High school guidance counselors are a big problem. They are hyper focused on getting kids into college while ignoring trade schools and the military.

The military is a great option for some people. I had great success hiring ex military for very technical high tech jobs and those folks have had fine careers as software engineers. As a bonus, not one was a whiny entitled crybaby, and I was stuck with a lot of those out of the University CS departments.

Maybe true but anecdotal.

Start at the beginning. What sense does it make to learn how to read if all that's going to be read is the National Enquirer and People? Any serious effort must be directed at an increase in income. Achievement is measured only in $.

They can read about wokeness.

In the US, though, how much of "the dupe factor" (and its propagation) is in play with education?

What sense does it make to impart a working level of literacy . . . if all that's expected of you is your transformation into a couch potato for video fare consumption and musical extravaganza exposure, or if your prized literacy limits you to pecking out terse texts on small keypads?

If literacy in the US is conceived and executed so well that graduates of any schooling are the ready dupes for anyone clever enough not to've actually needed formal education to become literate, arguably, US education efforts are at least as ill-conceived and poorly executed as they readily appear, since over 20% of US adults qualify as sub-literate or illiterate (metrics most Americans seem happy not to contest, yielding no demand for any remedy from responsible educators, policymakers, or cognitively elite personages or institutions).

Casting currency as sole metric of literary achievement is a dubious proposition, too, not because it's so unpopular a metric but because currency can never tell the entire story of what "achievement" consists. "Bread alone" is the commencement of many a human tale, but seldom does "bread alone" render any tale credible or complete.

I have been curious about your literacy for quite a while now. You have interesting ideas, so why obfuscate them in impenetrable prose?

Thank you, with due apologies.

Residing in Colorado, I'm usually still waking when I post anything here at MR in the a. m.

I was reading by age four without benefit of instruction (and without my parents' knowledge). I attribute the relative lack of discipline in my thinking to my idiosyncratic auto-didactic capacity for following my curiosity where it's led me. My love-hate relationship with formal education remains visceral and sincere.

My prose is not impenetrable, but I am afflicted with a large vocabulary and a vivid imagination (that I compose absurdist tales, science satires, horror and murder comedies, and verse may not help).

I prefer to think that I merely challenge readers, though some (many?) may find the effort not so very rewarding.

The study was about income mobility, should we not measure that in money?

Check out the other papers as well. Good stuff. I liked the results.

What are the "selection processes" that were purged in the study? [I didn't buy the paper - $36.] Since the author concludes that public investment in higher education should be targeted at low income youth, one might guess that the "selection processes" he purged were those that distinguished the poor from the not so poor (middle class), that by doing so he found that those with the most to gain (i.e., the poor) achieved the greatest relative income mobility - when one is poor to begin with, "success" is measured in small sums, but relatively large sums from the point of reference (i.e., poor). In other words, direct public financial aid for higher education to the poor not the middle class. Maybe other readers who purchased the paper can enlighten us as to the "selection processes".

Some states have higher education programs based on need, not the financial need of the students but the financial need of the state to keep the best students in-state. In these programs, the highest achieving high school students can attend the flagship state university tuition-free or almost tuition-free. The programs have been a great success, both in keeping the targeted students in-state and in public support (or the absence of public opposition) for the programs. Grades in high school are the only condition to qualify for the programs (i.e., financial need is irrelevant). One might see where this is headed: public financial aid for higher education will be directed (limited) to the highest achieving students and the poorest students, omitting the great whale (the middle class).

The whale can vote.

Yeah, and most of the rich kids aren’t at public school anyway and therefore aren’t top public school students.

Those higher achieving public school students are the middle class.

Not that I don’t agree that things are screwy, but this isn’t one of the problems.

My friend, a successful professional, with two very smart children who attended a private prep school, when asked if he would send them to the very expensive elite college he attended, said no, because he could send them to the flagship state university for free. Yes, those in the 1% are looking for handouts too, probably as much if nor more than those at the bottom.


What state is that?

The university here in CA, both UC and CS, used to be virtually free, but not anymore.

"purged of selection processes that may drive up observed mobility in this subpopulation"

I agree that someone should explain this to us. Is it like looking at Olympic medals while purging the selection effects of qualifying rounds?

The paper costs $0 at Sci-Hub, similarly to all paywalled papers:

The link triggers a security warning.

"This interpretation, however, rests on the implicit assumption that the college premium in mobility reflects a genuine, causal effect of college completion on equality of opportunity. This assumption, as discussed earlier, might not hold if college graduates from low- and moderate-income families are highly selected on such individual attributes as ability and motivation, attributes that by themselves may have substantial returns in the labor and marriage markets. In this case, the influence of family background might persist even though mobility appears to be higher among college graduates.

In fact, the selection problem has long been recognized in previous research (e.g., Mare 1981; Torche 2011). For example, Torche (2011:800) wrote,

[G]iven the substantial economic and cultural barriers that lower-class students face in attaining postsecondary education, those who make it to college are positively selected on unobserved attributes such as motivation and ability. To the extent that these attributes are rewarded in the labor market, lower-class college graduates will likely experience upward intergenerational mobility.

More recently, Witteveen and Attewell (2017: 1569) speculated about the direction of bias associated with conditioning on education:

Based on a vast amount of empirical findings, we suspect that our sample of BA holders is likely to favor high-income students after two important previously stratifying selections: students from lower-class backgrounds are less likely to attend college (1) and less likely to complete college (2). Unobserved differences are therefore likely to underestimate the effect of household income on earnings in this specific sample.

Aside from these ad hoc speculations, however, no previous study has systematically assessed the validity of the two competing explanations—equalization and selection—for the college premium widely observed in intergenerational mobility research. This study aims to bridge this gap.5

...The left panel shows that parental income (X) affects the adult child’s income (Y) both directly and through educational attainment, which, for our purpose, is coded as a dummy variable denoting college graduation (C).6 A range of individual attributes (Z) other than parental income—such as family structure, cognitive ability, and motivation—may affect both the likelihood of college graduation and adult income. From this DAG, we can see that college graduation is not only an intervening variable in intergenerational income reproduction, but also a collider variable—a common consequence of parental income (X) and other individual attributes (Z) (see Elwert and Winship 2014).

Therefore, as shown in the right panel, when we condition on college graduation, the associations between parental income and other individual attributes are artificially distorted. For example, if parental income is positively associated with cognitive ability in the population, this association will likely be attenuated or even reversed among college graduates. Such a distortion in the association between parental income (X) and other individual attributes (Z) can translate into a change in the marginal association between parental income and adult income—which we use to measure the degree of intergenerational immobility—even if neither the direct influence of parental income (X → Y) nor that of other attributes (Z → Y) differs between college graduates and non-graduates. That is, observed mobility may differ between college graduates and non-graduates even if a college degree does not modify any pathways of intergenerational reproduction.

...Now, we can define two measures of intergenerational mobility among college graduates (and non-graduates): conditional mobility and controlled mobility. Conditional mobility, as captured by Equation 3, reflects the degree of mobility we observe among actual college graduates (i.e., by “conditioning on” college graduation). Controlled mobility, as captured by Equation 4, reflects the degree of mobility we would observe among college graduates if, given parental income, college graduation did not depend on other predictors of adult income (i.e., after we control for selection processes that may confound the causal effect of a college degree on intergenerational mobility).

Thus, controlled mobility reflects the degree of mobility we would observe if, at each level of parental income, college graduates were a representative sample of the general population. By extension, it might also reflect the degree of intergenerational mobility we would observe in a utopian world of “college for all.” The distinction between these two concepts is crucial not only in theory but also in terms of policy implications. Whereas conditional mobility is a descriptive measure that gauges intergenerational mobility within the subpopulation of college graduates, controlled mobility has a prescriptive value because it better informs whether we can promote equality of opportunity by inducing more people into this subpopulation. If, for example, controlled mobility is no higher among college graduates than among non-graduates, an expansion of higher education per se may not be an effective tool for promoting equality of opportunity—unless it simultaneously equalizes access to higher education (by weakening the arrow X → C or Z → C)."

I assume this came from the study. If policy is to follow from the results of this study (which I do not question), then it follows that public aid should be directed based on conditional mobility not controlled mobility. This would be rational, but it ignores what I will refer to as fairness: the high cost of higher education will penalize (discourage) those whose income mobility is determined according to controlled mobility. And will piss off their parents too.

Public aid for higher education was essentially non-existent when I attended college. But the cost of higher education back then was a mere fraction of the cost today. One could attend a flagship public university for a few hundred dollars a semester. How does one shape public policy when the cost of higher education is out of reach for the vast majority of families/students absent some form of public aid.

Does this result surprise anyone outside of the liberal bubble and self-regarding academia? I know its a mantra amongst the elite that "Educashion" will fix human capital problems and we can go on pretending that the meritocracy is open to all. But it just ain't so.

The ZMP worker world is coming. Prepare.

I didn't read the paywalled paper and the abstract is, to be polite, turgid.

I think the conclusion is that getting a degree doesn't move the needle, except for the very poor. That should not surprise anyone, because when you are at the bottom, the only direction is up.

Wow! Stunning! Ultimately, one way or another, we pay for this sh*t.
Linking some of Tyler's posts, like open borders and this one, I guess we should import 100,000,000 poor people from Latin America and pay for their college education even though we - the middle class - can't pay for our own kids.

That ain't gonna fly.

We should cut the budgets of the wokeness factories and tax endowments.

Isn't this just rehashing what Charles Murray laid out in The Bell Curve, using the same dataset?


We get these papers purporting to demonstrate that higher education is ill-equipped to accomplish things that few people expected it to accomplish (but which twits employed in higher education have been touting as a benefit of higher education).

We might just be sensible and make use of tertiary schooling to impart vocational training which was (for one reason or another) not imparted in the course of secondary schooling. We can do that without requiring youth to stay in school for four years and meander around taking a grab bag of courses in academic subjects in order to get their credentials.

Within my lifetime, Britain managed to maintain a productive industrial economy with just 4% of each cohort obtaining a university degree in an academic subject. Let's go there.

People usually point to the GI Bill, and the high levels of economic egalitarianism experienced from the mid 50s through the 70s, as proof that college is a leveller.

The GI Bill did expand college access from where it had been in the 1920s and earlier, but federal policy and those who support expanding it further have taken the logic of it to absurd conclusions.

President Captain Bolsonaro has ordered a 30% cut across the board of federal universities funding.

"To promote mobility, public investments in higher education (e.g., federal and state student aid programs) should be targeted at low-income youth."


Given the results: "I find that once selection processes are adjusted for, intergenerational income mobility among college graduates is very close to that among non-graduates. "

Why would you recommend targeting the poor to promote mobility when you've just concluded that it doesn't have significant results?

The only way to square that is such that his "selection process" currently favors the affluent. That is, you can bring up the low-income mobility.

I could see some value providing scholarships - based solely on merit established by competitive exam - which would allow the smart/hardworking poor access to further education.

This is more or less the way the handful of elite academic high schools in NYC work - admission by competitive exam. IIRC, they have quite a large fraction of students from poor, indeed often poor immigrant, families. Its been that way for generations, although the ethnicity of the poor has changed over time.

The progressives in NYC don't like the results, because their favored minority groups don't earn merit admission at population quota rates, so they are currently trying to "adjust" the admissions process to achieve a "fairer" result. We all know where that will end up.

Maybe a few of us would be ready for a post-racial society, one that didn't even ask people to self- identify by "race," but your progressives aren't the only ones throwing sand in the works.

Actually, he is a multiethnic progressive. He just has opinions that differ from your opinions, mouse.

Ethnicity can be rationally defined. A politics that requires you to mash Swedes and Egyptians into one "race" cannot.

If I had a nickel for every time "race" was defended by changing the subject to non-equivalent concepts with non-equivalent implications ..

Swedes and Egyptians are very different people with very different cultures stuck together to satisfy a racist identity politics of the moment.

"stuck together to satisfy a racist identity politics of the moment"

To the extent they are "stuck together" at all, it's because of classification difficulties and the very small number of Egyptians in the U.S. Just because a measure is coarse doesn't mean it is meaningless- as of course we already know, given the highly significant correlations between coarse racial groups and all manner of effects, including in medicine.

Let me get this straight, because some (but not all) "racial" groupings are useful for medical pre-screening, the same question should be asked as you take the SAT?

I think you can sit back and see that you are relying on tenuous connections, especially from "race" to identity politics.

Do Swedes and Egyptians agree on politics?


You are obsessed with race. Why?

One note samba ...

I guess I should really be satisfied. No one is arguing that "race" is a fully correct biological categorization anymore. At best some argue that it is "close enough" to use for .. mostly bad purposes.

And in terms of why I really do care (not cool in some circles), it's those mostly bad purposes.

Just what I said a couple days ago! Race doesn't exist except when I want to invoke it to suit my policy preferences.

I don't see that as a very interesting troll. As if "policy preferences" is all you need to say to prove malicious intent.

Suck a fart, racist.

Oh, I see. Worse than just boring, this is "people who oppose racism are the real racists," episode million forty two. Never original, never well meaning.

Maybe you are projecting.

I remember the words of the late great Chief Justice of the Supreme Court, Antonin Scalia, "
If you want to stop discrimination on the basis of race, then stop discriminating on the basis of race."

So, just stop.

It's all tribal with him. You don't see him hounding the threads with how racist Harvard capping Asians or diversity quotas are, do you? Nope, it's always "this is a terrible wrong" caused by "you guys".

That's hilarious. Why would I, California guy who has never been to Harvard, have to own their mistakes?

Because *your* tribalism.

I am happy that the UCs place our kids, and nieces and nephews, whatever degree of Asian they might be :-)

The mouse can hide, but not from God.

Federico has interesting observations on the same.

I'm less completely down with Goodman, but as a historian of immigration, she raises some good points.

Specifically, the choice to exclude some immigrants from your in-group is a choice.

Yep. I want to exclude low skilled immigrants. My in group is high skilled tax payers.

If you are a highly skilled taxpayer, I am sure Canada and Australia would be happy to give you work visa. To get into the US you need someone to marry you and then wait 5 years.

There is nothing wrong with a points system, but it kind of leaves out the joys, and the entrepreneurship, of Tyler Cowen's ethnic dining guide, doesn't it?

Man does not live by an elitist definition of skill, alone.

the book "Paying for the Party" explores this to some extent. The book has many flaws: being and ethnographic sociology study, its population is limited, and the conclusions drawn reach much farther than they have a right to.

But if viewed within the restricted lens of a midwestern university with a strong Greek element, it has some saddening insights on the myth of the college education being a great equalizer.

That one looks good, thanks.

Paging Bryan Caplan...

I wonder to what extent social mobility is a function of intelligence or work ethic, and those are being imperfectly captured in educational attainment.

You can't ask that question.

Another point for Bryan Caplan.

Jane Wilson
Jeg var nødt til at skrive for at sige tak. Du har leveret den bedste spilleservice, jeg har betalt for! Du hjalp med at nedbryde min angst natten over. Jeg har været en stave castingservice bruger i flere år nu, og ingen af ​​de andre spielhjulere har jeg brugt cancompare til din amazingservice! Mange tak for at være der, da jeg neededyou "Kontakt Maleek-Ihama på WhatsApp

Comments for this post are closed