New York Times explores the low SAT scores of poor children

“New SAT Data Highlights the Deep Inequality at the Heart of American Education” (New York Times, October 22, 2023):

One-third of the children of the very richest families scored a 1300 or higher, while less than 5 percent of middle-class students did, according to the data, from economists at Opportunity Insights, based at Harvard. Relatively few children in the poorest families scored that high; just one in five took the test at all.

The disparity highlights the inequality at the heart of American education: Starting very early, children from rich and poor families receive vastly different educations, in and out of school, driven by differences in the amount of money and time their parents are able to invest. And in the last five decades, as the country has become more unequal by income, the gap in children’s academic achievement, as measured by test scores throughout schooling, has widened.

What are readers supposed to do with this information? SAT scores are correlated with job performance. By highlighting the dismal scores of a subset of Americans on its front page, is the NYT trying to persuade readers to avoid hiring those who grew up in poverty?

The Newspaper of Truth says that helicopter parenting is the sure path to a smart kid:

Parents have embraced what researchers call intensive parenting — the idea that parents should immerse children in constant learning. Half a century ago, rich and poor parents spent about the same amount of time with their children. Now high-income parents spend more one-on-one time with them, doing activities like reading — what Robert Putnam, the political scientist who wrote “Our Kids: The American Dream in Crisis,” calls “‘Goodnight Moon’ time.”

If true, shouldn’t the SAT scores of children from high-income families be much higher today compared to in the 1970s? The NYT cites no evidence to suggest that “Goodnight Moon” time has helped the privileged brats of today compared to 1970s kids who were left with their toys while moms socialized over gin and tonics, read their own books, had sex with neighbors (“One woman who married at 20 started an affair within a year. ”I think it’s your way of asserting that you can still act independently,” said the woman, now in her mid-30’s.” (NYT 1987)), etc. Also, aren’t the poorest parents the ones who have the most time to spend with kids? Consider what used to be called a “welfare family” whose house, health care, food, smartphone, and broadband are all paid for by taxpayers slaving away at boring jobs. The adults in that family don’t need to suffer the indignity of wage labor in order to keep a roof over their head and food on the table. The NYT does not explain what the jobless poor are doing if not spending time with their children.

One explanation that the NYT does not explore in detail: SAT performance is heritable. If rich parents had high SAT scores and the ability to score well on the SAT is heritable, it would make sense that children of the rich also have high SAT scores. One sentence is devoted to this topic: “Although the heritability of cognitive ability appears to play some role on an individual level, there is also a lot of evidence that environment matters.” There is no explanation for why heritability couldn’t play the same role on a neighborhood or city-wide level. If a neighborhood is packed with low-income parents due to everyone with a higher income having moved out, and employers in our modern economy pay for higher cognitive ability, why wouldn’t the average cognitive ability in the low-income neighborhood be low?

In a study of supernerds, it turned out that a higher SAT math score did correlate with higher income. From Insider:

The chart below compares the top (Q4) and bottom quartile (Q1) of the top 1% of performers on the SAT math section. It shows a significant difference, even among those subsets, in performance later in life (participants were surveyed at around age 33). For example, men in Q4 from one study group earn 13 percent more than those in Q1.

Note that “bottom quartile” was not the “bottom quartile” of all Americans who took the SAT, but of the top 1% supernerds. (identified at age 13).

It is surprisingly tough to find a broad study of how SAT scores from, say, 1990, correlate to 2022 income. But it makes sense that there would be a correlation. People who do well on the SAT are good at sitting at a desk, following instructions, being consistent with procedures, etc. These are exactly the capabilities that many high-paying jobs require. Some high-paying jobs, such as physician, have been explicitly limited to those who score well on standardized tests (though that may change; see “Removing the MCAT Could Improve Diversity in Medicine” (Newsweek 2023)).

Circling back to the NYT article, I find it interesting that the possibility of SAT score being heritable was not considered, even for long enough to dismiss it. Let’s also look at the solution:

The solution, researchers say, is addressing achievement gaps much earlier, through things like universal pre-K, increased funding for schools in low-income neighborhoods, and reduced residential segregation.

It could benefit all parents and students, even wealthier ones. Parenting in highly unequal societies is intense and competitive, driven by fear of the increasing risk that children will be worse off than their parents. Parenting in places with less income inequality and more public investment in families is more playful and relaxed, research shows. When the risk of falling is smaller, a college admissions test becomes less fraught.

The “increased funding for schools in low-income neighborhoods” idea seems inconsistent with a note earlier in the article that the typical state is already spending “more for students in low-income schools”. For example, Baltimore, one of the nation’s worst-performing public school systems, was spending over $17,000 per student in pre-Biden money (Fox), above the state average. Was the money effective? “At 13 Baltimore City high schools, zero students tested proficient on 2023 state math exam” (Fox).

[Note that these per-pupil spending numbers are substantially fraudulent. They don’t count capital costs, which are enormous. When $154 million is spent on a new high school (see ), that isn’t “spending”. Nor is the cost of the real estate considered. Baltimore official spending is up to about 22,000 Bidies per year per student, but it would perhaps be over 30,000 Bidies per year if these off-books costs were folded in. ]

Given that the number of spaces at elite colleges is held fixed while the population expands, I would like to see an explanation for how the rich will “benefit” if their kids are out-competed for elite college admissions by the children of the poor, whose schools have been turbocharged with extra money (on top of the existing extra money mentioned in the article). Why didn’t Asian-Americans realize how much better off they were when Harvard rejected them in favor of non-Asians? (see Students for Fair Admissions v. Harvard) Are Jewish families better off now that their kids can’t get into elite schools? (“Harvard has gone from being 25% Jewish in the 1990s and 2000s to under 10% today. … Penn’s Jewish population declined from 26% in 2015 to 17% in 2021”; Tablet)


24 thoughts on “New York Times explores the low SAT scores of poor children

  1. Not to forget, the Baltimore Public Schools CEO gets paid half a million dollars a year for her sterling achievements!

  2. It seems obvious that IQ is heritable — why wouldn’t it be? The Son Also Rises indicates that it is as well as ordinary experience — how many high IQ parents with natural children do you know who have low IQ children and vice versa? You would not think that would be such a big issue since a lot of abilities are not spread evenly across the population. I suppose where the problems are is that there is some evidence that there are differences in IQ among races and ethnicities and that cuts against the American ideals of equality. In the past IQ was probably of lesser importance than it is now to enjoy the American dream of a solid ;middle class life since there was more availability of high paying industrial jobs that depended more on abilities other than IQ intelligence. I don’t know that we know how to deal with these two issues other than wealth transfers in the form of government benefits, school funding, etc. which the evidence seems to indicate does not accomplish much in terms of reducing disparities. These issues are probably not discussed honestly for a lot of reasons including that there is a large bureaucracy that administers the transfer payments and earns a nice living doing so and it is far from clear that discussing the issues honestly would accomplish more than generate envy and dissatisfaction.

  3. At the last minute and with no prep, I took the SAT in 1980 and scored at the 60th percentile. In 1987, I took the GMAT and scored at the 60th percentile. In 1997, I again took the GMAT and scored at the 98th percentile. That score didn’t get me into Stanford but it did get me a half tuition scholarship to the MBA program at the University of Georgia.

    • I thought that for MBA lower score signified higher chances for success moving up corporate ladder. How did you progress in corporate world, top 2% smarty? I recall that when I was looking at it college GPA (out of 4) admission requirement for MBA was a point below STEM MS program requirement, at same universities.

    • I attribute my big jump in GMAT score to four months of intensive prep, 4 – 6 hours a day prior to the taking the test for the second time. Mostly taking practice tests over and over and studying the correct answers and problem-solving tips and methods.

      I would have earned a higher salary had a foregone the MBA and remained a software developer. A couple of job layoffs and sub-optimal career decisions, and dwindling ambition stalled my career progression and salary growth (never cracking six figures).

  4. The article does mention heritability briefly and the sentence quoted below includes three links to academic papers.

    “Although the heritability of cognitive ability appears to play some role on an individual level, there is also a lot of evidence that environment matters.”

  5. Rich kids would benefit from not living in constant fear of being robbed, carjacked, shot… by poor kids who dropped out of shit school for having nothing to look forward to, which just perpetuates the downward spiral of the entire system. Kind of like israel-palestine situation you’re also completely blind to.

    • knch: Why would kids in the poorest neighborhoods be afraid of being carjacked? U.S. taxpayers provide those who don’t work with housing, health care, food, smartphone, and home broadband. Not on this list: car, car insurance, gasoline. And why would they not have access to a decent school? The poorest neighborhoods are in cities run by Democrats and the teachers are members of the union run by Randi Weingarten. I hope that you’re not suggesting that these well-meaning Democrats aren’t serving the poor, the very people whom Democrats say they care the most about.

      The New York Times doesn’t consider crime as a possibility (the word “crime” is not in the article), though a linked-to article does. We can test your theory that crime explains low SAT scores by looking at neighborhoods with the same income, but different crime rates, and seeing if the SAT scores track the crime rates. (We can also look at the same neighborhoods/cities over time. Since the crime rate was much higher in the 1970s-80s, urban kids today should be doing much better (immigration is a big confounder, though; the population of a neighborhood or city may have been replaced with a new one).)

      Separately, I’m not blind to the disadvantages of growing up in Gaza. The adults who live there chose (1) war with the new state of Israel in 1948, (2) to reject peace proposals over the past 75 years, (3) to have more babies than humans anywhere else on the planet, (4) to vote for the Islamic Resistance Movement (“Hamas”) whose charter promises to eradicate Israel via military action (an example of which occurred on October 7, 2023), and (5) to put all of their resources into their military (made possible by US and EU taxpayers funding the basics), including the Palestinian Islamic Jihad whose locally produced rockets have been in the news lately. These choices guarantee that there will be more children, as a percentage of the population, than anywhere else on earth but also that each individual child won’t enjoy the advantages of open space, economic prosperity, etc. Since you have sympathy for these kids and say that I don’t, what are you doing for them? Do you live in a studio apartment and send all of your excess income to the Palestinians? Do you give all of your money to politicians, such as AOC, who promise to stop the Israeli military response to the October 7, 2023 attacks? Do you spend all of your excess time and money to open the borders of Arab countries, EU countries, and the U.S. so that people from Gaza who wish to enjoy the more comfortable lifestyle elsewhere can do so?

    • Formidable mental gymnastics you have… Why do you deny palestinians your shared anti immigration logic? Why would they peacefully accept forceful eviction, massive immigration into an area the size of a small US state and a constant stream of (even UN recognized) war crimes against them?

      Also, to turn it around, what have you done to help israelis move to a part of the world where they wouldn’t be either the hostiles or victims of hostilities?

      PS read again, the comment starts with rich kids, not kids in poor neighborhoods. Or do you suppose poor kids can’t find a way into a better neighborhood where they can help themselves to everything on display?

    • knch: I did not mean to suggest that the Arabs and Muslims, including the Palestinians and Iranians, have acted irrationally. I always start from the assumption that each person acts in his/her/zir/their best interest. If the long game of eliminating the Zionist entity is what matters to them then attacking Israel on October 7, 2023 may have been a smart move (just as the rejection of the UN’s partition plan in 1947 was a smart move and the declaration of war on Israel in 1948 was a smart move). But in the short run, they also have to expect Israel to engage in military action aimed at eliminating their ability to conduct similar attacks. (I personally don’t think that Israel will be successful. As long as Palestinians have unlimited food, health care, and education from EU and US taxpayers, it wouldn’t make sense for them to give up on their goal of eliminating Israel. Maybe the Islamic Resistance Movement (“Hamas”) and the Palestinian Islamic Jihad are somehow disbanded, but the majority of Palestinians want an Islamist government and, therefore, a new political party called Palestinian Islamic Movement or Islamic Palestinian Jihad can be formed.)

      What have I done to help Israelis? I’m not sure why you ask. As far as I am aware, I haven’t posted anything about Israelis suffering or being victims of war crimes, etc. I would like to see Israel prevail over Hamas and Palestinian Islamic Jihad, but that isn’t because I consider the Israelis to be the world’s most oppressed or unfortunate people (they’re not getting into boats and crossing the Mediterranean into the EU, for example).

      (If you believe that the Israelis are committing war crimes, as recognized by the impartial folks at the UN who’ve been taking care of all basic needs for Palestinians since the late 1940s, why not advocate for NATO and the US to attack Israel and eliminate the country’s ability to fight any wars? If the Israeli military is eliminated then the Jews cannot commit any war crimes. We did this to Yugoslavia in 1999. See )

    • I think that knch is on something. I personally totally disagree with his assumptions of war crimes, NATO officer experts with experience in former Yugoslavia, Iraq, Afghanistan confirm that Israel response does much more to protect hostile civilians then what NATO did in those countries. But Philip, this blog is rationalist and totally ignores God of Bible and supreme moral postulates. Why would not you rationally advocate to sacrifice state of NJ or Massachusetts, that are comparable in size to modern Israel, clear them from those who expelled Native Americans and resettle either Israelis or Arab population opposing Israel there? Either one of them? To atone for the Holocaust? Or White privilege? Hypothetically, as there are no wrong assumptions on this blog.

    • perplexed: That’s an excellent point. The people of Massachusetts admit that they’re on stolen land. They could give it back to the Native Americans and then it could be rented for use as a Jewish state.

      I think that you’re also right about the U.S. military being far more destructive to civilians than the Israeli military. Germany and Japan during World War II are obvious examples, but if we say that the U.S. military has reformed itself, let’s consider the Iraq war. Iraq never declared war on the U.S., as the Arabs did on Israel. Hamas is the elected government of the Palestinian people, both in Gaza and the West Bank (though the West Bank was stolen from them). Saddam Hussein did not come to power via a popular election. So we have no reason to assume that the average Iraqi civilian backed Saddam in the way that Palestinians back Hamas (the 2006 election and, more recently, opinion polls). Yet we caused about 600,000 civilian deaths (Lancet estimate) in Iraq. By contrast, the UN says that about 5,000 people have died in Gaza (see ; they say that “children” have been killed, but I wonder if that includes teenagers who fight on behalf of Hamas; here in the U.S. we are exposed to media reports of “teens” killed by guns and it turns out that this includes 19-year-old drug dealers).

  6. Heritability comes up if you read far. Steve Sailer mentioned it in his fisking:

    “Strikingly, the Times reporter eventually gets around to inserting a one sentence admission against interest that it’s not all that the rich can afford better nurture, but that nature matters too:

    “ ‘Although the heritability of cognitive ability appears to play some role on an individual level,’

    “I.e., not on the racial level. (This Chetty study prudently doesn’t break out any data by race.)

    “ ‘there is also a lot of evidence that environment matters.’”

    • Sandy: Thanks for that. I think I searched for “heritable” and, therefore, Google Chrome didn’t find “heritability”. Let me update the original post.

      (It is interesting that the NYT says that “cognitive ability” (not really the same thing as SAT score, which can be increased substantially via grit) is heritable on an individual level, but that a group of low-ability people gathered in a low-income neighborhood could not produce a group of low-ability children via the same genetic method.)

  7. This subject is silly and a waste of time. Every rational person knows what’s going on here. The data go back decades and decades. Charles Murray has been writing about it forever. But when even suggesting the obvious is prohibited, then all manner of wacky theories spring up to take its place.

  8. Animal breeder here (as a job, not a hobby, industrial settings). Heritability does not work the way you think it does.

    • Federico: Instead of merely telling us that we’re wrong, why not educate us? Why is the NYT article wrong in saying “the heritability of cognitive ability appears to play some role on an individual level”? (i.e., why is it wrong to say biological children of parents who scored well on the SAT will tend to score higher on the SAT than biological children of parents who scored poorly on the SAT? (let’s assume that all children are adopted to separate out the nurture aspect))

      See also “What Happens When We Randomly Assign Children to Families?”

      in which the educational attainment and income of adopted Korean children had almost no correlation with the education/income of the families in which they grew up.

    • Phil, people pay good cash for the intel, or they bother to go get a degree in animal breeding, etc.

      You cannot meaningfully talk about ‘heritability’ without a pedigree or a genetic kinship matrix, from actual data, not general ideas such as ‘child 1 is adopted, child 2 is not’. If child 1 has an unknown genetic kinship with the rest of the family, we cannot infer how much genetic material is shared, so no heritability can be estimated. *Anything* goes. This kind of example are amateur pseudoscience. The only meaningful comparisons between siblings are between homozygote twins (that is, clones) and fraternal twins (i.e. people sharing 50% of their genetic material, and also sharing the same uterine environment).

      Also, heritability is a *sample* property, not a population property, and it is not stable across generations — why do you think plant and animal breeders keep measuring the damn thing at every generation they have data for? I hint it is not for fun.

      The fact that adopted kids might do worse is due to a heaps of reasons: I personally know adoptees that have good jobs and PhDs, and others with drug problems stemming from childhood trauma and attachment issues. These latter issues are not genetic, but I bet you a signed 10€ note they are more common in adoptive children.

      Finally, it is trivial to provide a minimum of 3 hypotheses explaining why a child is doing poorly in school: 1) the child is stupid, 2) the teachers are inept, 3) the child is not doing any work irrespective of the intelligence (a symmetrical approach can be used for academically successful kids). How do you verify and disentangle these in the real world, especially at the population level?

    • “ The only meaningful comparisons between siblings are between homozygote twins (that is, clones)”
      There are tons of studies on twins that show IQ is hereditary. All one has to do is be willing to look.

    • Anonymous, there are lots of twin studies showing that clones have a less variance in IQ than fraternal twins, which is hardly surprising — charging head first through a wide open door in fact.

      Once more, either you understand what you are talking about, or you don’t — you folks clearly do NOT. The fact that adoptive kids have a different life outcome than natural children can be explained by gagillions of reasons that are not genetics. The fact that human clones are very similar is also not exactly surprising — they are clones.

      Proving that genes cause X mean that we know what genetic variants are present in a sample before X can happen altogether, and then assess how many individuals with variant A get to X, how many fail to get to X, and how many individuals with any other variant that is not A get to X, and conversely fail to get to X. Then the stats begin, and there is NEVER any clear clear cut answer, just some wishy washy answer.

  9. Last two comments and Philip’s response strike me as schizophrenic. Nobody states their opinion and everyone is sure what they are discussing. I myself do not want to add to the discussion because this is a rationalist blog and I think what we call intelligence relates to morality and spirituality, not rational things. Because extra-rational is the easiest explanation why some people who excel at complex skills, abstractions and behaviors such as for example driving, speech and other communications and ceremony (such as modern fashion) are not mastering other mental, operational or behavior tasks.
    But I am expecting rational discussion on this blog, and do not see it for this topic.

    • “But I am expecting rational discussion on this blog” this bird brain doesn’t think that’s very…. rational!

Comments are closed.