Here's a helluva chart:¹
Both elite private universities (Harvard, Stanford, etc.) and elite public universities (Berkeley, U Michigan, etc.) got most of their students from affluent families up until World War II. After that the GI Bill identified smart but poor teens who could benefit from an elite education, and the number of kids from upper income families declined.
But by the end of the '70s, those kids were all high income too and started sending their kids to elite colleges. By now we've done such a good job of finding and rewarding smart kids that there are hardly any left among poorish families. So we're back where we started.
¹It comes from a new paper, "The G.I. Bill, Standardized Testing, and Socioeconomic Origins of the U.S. Educational Elite over a Century." The authors performed a truly stunning bit of data collection to get the results they did.
Kevin, their paper explicitly rejects the GI Bill effect you claim here! From their abstract: "... two major policy changes in the history of American higher education, namely the G.I. Bill after World War II and the introduction of standardized tests for admissions, had little success in increasing the representation of lower- and middle-income students at elite colleges."
Seriously, not only does the paper explicitly not say what Kevin says, the idea that "there are hardly any smart poor kids anymore" is a really weird takeaway.
This sure looks a lot like an inverse graph of the middle class and widespread economic opportunity to me.
...the number of poor but good students isn't something the runs out. Children are a renewable resource.
But if poor families are getting poorer and environment (and diet) plays a part in intelligence, maybe that is the driving force. Also, the quality of the education at school was declining that could also play a role. Kevin is ignoring that. That inflection point in the 80s sure seems to come up a lot!
The poor are not getting poorer. They are getting slightly less poor very, very slowly.
The poor are not getting poorer on an absolute scale. OTOH, they are quite possibly finding their disadvantage is increasing relative to the non-poor. In other words, while the scale is shifting to the right (by some amount), the trichotomy separating poor, middle, and rich continues to expand.
To the extent that differentiation disadvantages the poor educationally, the problem of identifying--and then educating--the capable children of poor families continues.
Seems worth noting how few kids from the middle 60% are getting in, especially at the private colleges. Looks to be about 25% of the student body at this point. That seems extremely low to me.
What strikes me about these charts is how the inflection point aligns with so many charts showing the decline in worker’s share of productivity increases. Something serious happened in 1978.
And at the same time, capital’s share of national income began rising, management’s share of income exploded, and the national debt skyrocketed.
Must be a coincidence…
“ we've done such a good job of finding and rewarding smart kids that there are hardly any left among poorish families”
Kevin just said all the poor kids today are stupid.
WTAF!?
Look in the mirror to see stupidity today, sir, because that was a whopper.
I'm in a more generous mood this morning, so I'll just say I think Kevin was using "smart" somewhat glibly here to refer to "skills and motivation necessary to gain admission to a top college or university these days," not raw intelligence (whatever that means).
Alternatively, John Reed, Kevin pointed out that "stupid" (your word, John) families likely produce a significantly higher percentage of stupid kids, due to both heredity AND environment. This strongly mitigates against those poor kids getting into Harvard and Stanford, thus continuing to shrink the percentage of poor kids who can get into the top schools.
Also, for further explication, see the jte21 comment, just below.
Look in the mirror to see failed logic today, John, because yours was a whopper.
Sure, Ken. Kevin said "By now we've done such a good job of finding and rewarding smart kids that there are hardly any left among poorish families."
There are hardly any smart kids left among poorish families. That's not interpretation, that's what he wrote. You can go back to the top of the page and read it. If there are hardly any smart kids left in poorish families then what type of kids ARE in poorish families?
So Kevin said there are hardly any smart kids left in poorish families today but he didn't mean "smart" as in "not stupid", what he meant was that there are hardly any "skilled and motivated" kids in poorish families. Got it.
The paper's authors explicitly say that the GI Bill and standardized testing had little success in expanding the number of lower income students at elite institutions. Could growing inequality since the 70's (which Kevin has discussed extensively) be maybe a more logical takeaway than "poor families producing a higher percentage of not smart kids due to heredity and environment"?
cistg: You wrote: "If there are hardly any smart kids left in poorish families then what type of kids ARE in poorish families?"
The answer to that simple question is also simple: It most definitely doesn't mean the only possibility is either "smart" or "stupid." A large percentage of the kids of poor families have been lifted, by virtue of improvements in childhood education across the board, into the large middle category--"average."
Those average kids can frequently go to a pretty decent community college with low (or even zero) tuition, or even a decent public 4-year college, and can learn what they need to survive and succeed in the modern world. But there sure as hell ain't many average kids from poor families getting into Harvard or Stanford, or Michigan or Cal Berkeley.
David Brooks has an interesting essay in the Atlantic (blind squirrels do find nuts occasionally) this month where he makes a related point about higher ed, accessibility, and the impact of technocratic/meritocratic admissions standards. What higher ed came to value after WWII (led by Harvard's president James Conant) was raw intelligence, as measured by IQ and standardized tests, grades, etc. This replaced blue blood as the currency that got you into the Ivy League and other elite colleges, as Kevin's article notes. The problem is that measuring just a certain kind of analytical cognitive ability, rather than making things more egalitarian and accessible, just created a new kind of elite -- and one that tends to reproduce itself in ever-narrowing iterations as they become wealthier and more connected and more influential. As we all know, kids from highly-educated, wealthier households tend to be inculcated early on with the cognitive and social skills necessary to excel in this "meritocratic" filtering system.
Brooks asks what would happen if we stepped back and universities and Fortune 500 companies focused less on someone's big brain credentials -- especially those test scores and IQ and stuff -- and began valuing other metrics of intelligence that don't necessarily rely on being born into an upper middle-class home with two engineers for parents. The problem is how you actually fairly measure that stuff (back to metrics again!) and decide which skills to prioritize, but interesting to consider. The larger problem, IMHO, is just teaching kids to love learning, regardless of what kind it is. That's the real key to success.
+1 for David Brooks making this point.
+2 for jte distilling it down to two easy to understand paragraphs.
What would "other metrics of intelligence" even be? This is some kind of intelligence that doesn't help you on test scores in school, but somehow you are supposed to do as well as higher IQ people in school and jobs. Perhaps there are aspects of intelligence not captured in IQ tests that help people in jobs specifically, but as you note, fairly measuring this is hard.
And often it’s more personality traits helping people do well in their jobs rather than some kind of unmeasured intelligence. Simply being organized, likable, and able to work well with others gives you a big boost in the workplace.
This is a very impressive study overall. There is a complicating factor that does not seem to be addressed from my quick reading of the article, especially for the private elite institutions. Basically, they are much more selective now , and account for a much smaller percentage of the overall student population, and the US population, than 100 years ago. Harvard accepted over 90% or applicants in the 1920s, 30% in 1960, and 3.5% now. Same for many others. This should significantly skew results over time.
Another limitation is that there is little attention to immigrant background. This is understandable as it would be a lot of effort to do this. But we see that some institutions like Cal and SUNY do much better at enrolling kids from the bottom 20% income than others -- are they actually doing something right, or is it just that they are in states with a lot of immigrants, or children of immigrants, which in my experience make up a large fraction of the not-from-the-top-20% students at elite institutions?
Kevin says "After that the GI Bill identified smart but poor teens who could benefit from an elite education", but the graphs do not show this - the increase in bottom 20% is at most only a couple of percent. It was actually the middle 60% which increased temporarily at the expense of the upper 20%. And the abstract of the paper agrees. It says "despite a large increase in the share of lower-income students in the overall college-going population, the representation of these students at elite private or public colleges has remained at similarly low levels throughout the last century."
So we're back where we started with respect to the top vs middle incomes, and the lowest income fraction never changed much. This is a poor record overall with respect to elite admissions. As the paper says the share of non-whites has increased, but apparently these came from the middle-income fractions.
Most people filling most jobs can get a perfectly good education from non-elite colleges, so this doesn't tell the whole story. The GI bill (for example) was probably not aimed at changing the nature of enrollees at Harvard.
Maybe Kevin thinks 79th percentile is "poorish" but that still doesn't justify the interpretation that the more talented students have been identified.
"By now we've done such a good job of finding and rewarding smart kids that there are hardly any left among poorish families."
JFC, that had better be satire.
If we believe that intelligence is hereditary or that a big component is cultural factors that can be transmitted without too much money (or prioritizing academics) then we should expect this.
When you give the smart poor access to higher education then the second generation is still smart but no longer poor. We are now three to four generations into this sorting process. It is reasonably complete except among first and second generation immigrants.
Add in associative mating and the effect becomes even more extreme.
This is probably also a factor in why income mobility is down. Children born in each quintile are NOT the same. Higher quintile children are genetically and culturally better suited to financial success in a modern technological society.
One has to wonder how far this will go. Could we actually get speciation? Or at least rich/smart and poor/not smart becoming different races? Will we have Eloi and Morlocks?
I wonder if assortative mating is playing a role. Educated, intelligent people are increasingly marrying each other, which in turn reduces the number of couples with one particular intelligent partner.
I am on an advisory committee at, a so called, elite university (sub ten percent acceptance rate). The discussion on this topic includes::
Low income domestic students, with elite test scores, tend to come from first generation Asian families or international students.
Low income, elite scoring, student are broadly recruited/ get offers from multiple elite schools
Students without elite test scores tend to struggle in STEM majors. Further, even in non STEM majors, you get much lower graduation rates and worst job placement.
You'd think the line to watch would be the top 20% line. The bottom 20% line doesn't change much which is not surprising. Since we're talking about the elite schools, any interesting action is going to happen in the middle. Also, the focus should be on male enrollment since social norms changed during this period. Many elite schools didn't admit women until the 1970s. What one sees in the charts is roughly 60-70% in the top 20%. There was a downward trend in the 1950s followed by a data blackout in 1970 and a return to the high end, around 70%, from 2000. Yawn!
This tells us something about the elite schools. They are still largely the province of the economic elite. (I'm an MIT alum, and I know that MIT has long had a different economic mix than the rest of the schools in Table A2. It still does, but it's an engineering school, so it actually has to qualify its students. MIT tried dropping the SAT requirement during COVID but quickly reinstated it. The other schools don't have this problem, so they can have legacy, athletic and paid admits.)
If one is assessing societal structure, it isn't about poor students getting into elite schools. It's about non-elite students attending college at all. All those exploitative non-union shops now have college as a safety valve for avoiding or getting rid of their brighter workers who might become labor organizers. It's worse if you think intelligence is inherited. Aside from immigrants, the low end jobs have a much better way of filtering out bright workers and potential union problems.
It just doesn't make sense to assess the state of baseball as a sport by counting major league baseball players.
The extent to which this is worth caring about depends on what you want from a university — the blessings of having joined “the elite” or an actual education.
There are rumblings right now, it will probably take a few years for the phenomenon to become large enough to seep into general consciousness, that the actually brightest students at the best high schools in the country (ie “prep” schools) are mostly abandoning the expected destinations for the South and religious colleges (think eg Georgia Tech rather than Stanford, or BYU rather than Yale). I think it’s worth keeping tabs on this, to see how big it is, and how it plays out.
As for why, well we all have our theories, but I think Yeonmi Park probably has something useful to say…
Park? She’s a crank conservative grifter.
That's literally the definition of living in a bubble - dismissing everyone who presents anything you dislike...
Do you also claim Bari Weiss is a "crank conservative grifter"? Heck, maybe you do, since she doesn't toe the party line on any and every occasion?
She's not quite as extreme as Park in her denunciation of her time at Columbia, but it's not like she says that she has no idea where Park is coming from, that Weiss' experience at Columbia in no way matched Park's. She sees Park as an exemplar of the experience, not a wild and crazy exaggeration.
"By now we've done such a good job of finding and rewarding smart kids that there are hardly any left among poorish families."
You sure that this is a valid conclusion? Dumb kids of rich people go to college all the time. Ever heard of one George Walker Bush?
While he should not be sure of the conclusion, your comment is a non-sequitor. Rich indifferent students (dumb, lazy, whatever) going to [Elite] colleges does not comment on Drum's observation - they are separate issues not necessarily related
I would a significant part of the issue here is Diminishing Returns - and an over-focus on elite education versus broader and non-college education. Sending more poorer students to elite universities as a continuous objective will naturally run into diminishing returns on effort and equally speaking sending more and more students to college will
I do not think the quality of the data and quality of information on intelligence (from a very strict scientificate PoV) permits anyone to confidently say if the students are smarter (for whatever smarter means - whatever soup of genetic, household, environment they have for college academic performance) etc.
But the economic return on such an effort has clearly gone sideways - if only from diminishing returns. (and it rather appears broadly - not just in USA case - that the return on "more college for all" has at minimum diminished in high and middle income countries, while the physical economy needs of green transition and infra-industry electrification (and dare one say equally AIification) rather suggest less emphasis on the Anglo-American model of the four year university and more broader and less devalued technical. (a change from 20 yrs ago when the Americans would have everyone become a coder.)
Widening one's reflexion in goals would also be probably political as well as economically wise.