The Volokh Conspiracy
Mostly law professors | Sometimes contrarian | Often libertarian | Always independent
A Ridiculous Study on Asian American "Advantage"
Sociologist Jennifer Lee writes in the New York Times:
In "The Asian American Achievement Paradox," which I wrote with Min Zhou and is based on 162 interviews of Asian, Hispanic, Black and white adults in Los Angeles, we found that Asian American precollege students benefit from "stereotype promise": Teachers assume they are smart, hard-working, high-achieving and morally deserving, which can boost the grades of academically mediocre Asian American students.
Let's stop right there. The coauthors attempt to explain the average educational success of "Asian Americans," a classification that includes dozens of ethnic/national subgroups that have varying average degrees of educational success (including some that are below average), who live all over the United States, based on *162 interviews* with adults of various "racial" groups in one city, Los Angeles.
The book won various awards. Go figure.
UPDATE: Perhaps the book is much more nuanced? But in any event, to give you some idea about the extent to which Asian American subgroups vary in educational success, let's take a look at undergraduate matriculants to UC Berkeley, which breaks the classification down by subgroup.
California is about 1.5% Indian American. 12.7% of Berkeley's class is "South Asian," primarily Indian.
California is about 3.5% Chinese American. 15.3% of Berkeley's class is Chinese.
California is about 2.5% Vietnamese American. 3.9% of Berkeley's class is Vietnamese.
California is about 3.2% Filipino American. 3.8% of Berkeley's class is Filipino.
California is about 1.2% Korean American. 4.6% of Berkeley's class is Korean.
California is about .7% Japanese American. 1.4% of Berkeley's class is Japanese.
California is about .8% Pacific Islander. Pacific Islanders are often lumped together into an AAPI category, and assumedly would benefit from at least some of the positive stereotypes that Asian Americans get. Berkeley's class is 1/10 of 1% Pacific Islander.
As you can see, while the larger Asian American subgroups in California almost all do better than average, the stereotype of remarkable Asian American educational success, at least in California, is driven primarily by Indians, Chinese, and Koreans who are "overrepresented" by approximately 8, 4.5, and 4 times their populations, respectively.
Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of Reason.com or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Comments may only be edited within 5 minutes of posting. Report abuses.
Please
to post comments
Also, " is perceived as smarter, therefore also receives artificially inflated grades" is a really stupid premise.
Well, it does present a chicken and egg problem; it could just as easily be, "do better are tests, get better grades, and are therefore perceived as smarter."
Yup it’s harder to inflate math grades than (say) English, and even harder to inflate standardized test scores, so if perception-induced grade inflation were much of a thing, you’d expect to see pretty much the opposite of the pattern we in fact see.
Likely why mathematics scores are being dropped as considerations for graduations or acceptance by some schools, and 'equitable math' and 'ethnomathematics' are embraced by some.
So ... "Asian privilege?"
It's all racism -- all the time with these people.
Asian privilege is just as real as white privilege. 🙂
So ... a total fiction -- GOTCHA!
Indian, Chinese, and Vietnamese are catchall descriptions, which often cover several very disparate groups. The same claim is also true but to a lesser extent for Japanese and for Filipinos.
A lot of Hmong where I live. Not so much a college crowd, but their grandparents lived In Vietnam.
Ethnically, I can confirm that you'll find Filipinos who look like anything from Spanish, Thai, Chinese, Japanese, you name it. That island chain is a real melting pot.
One cannot end racism by focusing on race. One ends it by rising above it.
I feel like you should read the book before complaining about it.
Perhaps. But the op-ed quote is ... something. If the author isn't embarrassed to make to make a claim in NY Times, that 162 interviews of people of various "racial" groups in one city is enough to explain the educational success of "Asian Americans," which isn't anything like a uniform group to begin with, throughout the United States, I can't say I hold out much hope for the book's methodological rigor.
Indeed.
I am just reviewing an medical paper about China's top tier hospitals. What is the paper based on? A short survey of 23 hospitals in all of a country of 1.45 billion people.
Seem a lot like surveying 162 adults in LA
Not quite, Don.
These were interviews, not a short survey. And 162 is about 7 times 23.
Note too that they were confirming some earlier work.
I'm not going to claim that this is a wonderfully rigorous piece of work, but I think Bernstein is going a little overboard.
The authors state their primary source of data (aside from the anecdotal interviews) is the Immigration and Intergenerational Mobility in Metropolitan Los Angeles study, a 5 year program that surveyed some 5000 men in the ages of 20-39 in LA. The interviewees were selected from amongst those that participated in the survey.
I can't find the raw results, although several summaries are available by the original group that did the project. The study focuses heavily on the differences between Mexican and 'Asian' groups (even though more detailed groups are sometimes used).
The survey itself, as a representation of facts about Los Angeles immigrants (legal and otherwise) appears OK. However, it only covered 20-39 year old men in Los Angeles. You can't reliably take any results, no matter how sound, and then directly apply them to other groups or places.
The authors of the book here do exactly that, and then double down by creating data for people that were not sampled in any way. They also ignore the illegal/legal immigrant split (an artifact of the survey, which does so deliberately). This, all in addition to the racial grouping problem Prof. Bernstein has pointed out.
The racial grouping problem is especially bad, because the authors will at time split the groups out to make points, but other time smoosh them together to make a different point.
Now, the Amazon and Google free samples only covered the first two chapters (100 or so pages), plus some random parts towards the end. Perhaps the authors do, in fact, spent chapters justifying their odd treatment of the data. But from the chapter titles, and the chapters presented so far, this is a pop social science book, light on math and data, and heavy on "interpretations" and "meanings". Early complaints about "flippant dismissal of the significant of race" and "neo-conservative policy paradigms" kind of give the game away.
My biggest problem in the first two chapters is that while culture is frequently mentioned and included as a factor in their model, not once is there any attempt to model the original cultural behaviors of the immigrants. Instead, it is assumed that the behaviors of the immigrants are the behaviors of the culture as a whole, without doing anything to support this assumption.
And also, while they are "confirming some earlier work", they are also directly contradicting some earlier work - in fact, the introduction specifically describes some of what they hope to change about the topic with their research.
It is not the case that this book is just another publication confirming a long-existing and non-controversial theory.
Nonetheless, Bernard, the sample size is too small to be meaningful and is therefore much more susceptible to selection bias.
I will note that when I taught at UCLA 30 years ago, the graduate enrollment in engineering was nearly two-thirds.
Don -
Its even worse, while the sample size is too low for the population, there is no indication that any of the other criteria needed to ensure a random sample selection is present. In summary it is a junk survey.
Just to be clear: The original survey of 5000 people, from whom the 162 interviewees were drawn, appears to be a valid random sample.
The problem seems to be with using these 162 people to represent the larger populations.
Two major reasons for the replication crisis are (i) small sample sizes and (ii) a publication bias in favor of results which are congenial to bien-pensant liberals. So literally every social science result published in the past 20 years should basically be ignored (i.e., it should not move your priors). But I expect the bitter clingers in tenured academic positions to continue as they are.
(ii) a publication bias in favor of results which are congenial to bien-pensant liberals.
There's a slightly more innocent variation on this theme. Which is that inconclusive results don't tend to be published. If there is a natural bias in testing propositions that confim lefty notions (both amongst researchers and funders) then the possible results of such a test are :
1. the lefty notion is confirmed
2. inconclusive
3. the lefty notion is refuted
But practically, the test design is unlikely to be three pronged. The test will usually be designed to identify yea or nay to 1. Thus nay covers 2 and 3, and so there will usually be a binary - confirmation of the lefty notion, or not. "Not" will struggle to be published, even if the publisher isn't biased, because it's a muddy result. It might mean the lefty notion is wrong, or it might mean we don't know.
This then leads on to another source of bias. Because researchers know their results are unlikely to get published if they produce an inconclusive result, and so funding for the next project will be harder to come by, some researchers will put their thumb on the scale.
The title of this post is not about an op ed.
I dunno about the study, but your own methodological rigor is not looking great this weekend.
Perhaps the book is much more nuanced?
If you are not going to read the book before criticizing it, you should at least read the comments by reviewers at the site you link to.
Those comments do in fact suggest that the book is much more nuanced and that the phenomenon described in your quote is only a small part of its story.
Well, and this is in response to Sarcastro, too, the quote is *her* description of what the book concluded.
And you're not in a position of judging how she reached her conclusion without reading it. The exception you take, after all, has way more to do with your own hobby-horse than the book itself.
“In “The Asian American Achievement Paradox,” which I wrote with Min Zhou and is based on 162 interviews of Asian, Hispanic, Black and white adults in Los Angeles, we found that Asian American precollege students benefit from “stereotype promise”: Teachers assume they are smart, hard-working, high-achieving and morally deserving, which can boost the grades of academically mediocre Asian American students.”
How many ways can the authors slice and dice this to come up with something "meaningful"?
They put more work into interviewing these 162 people than you did in reading their book before assessing it.
Do you take pride in your meaningless posts?
No more nor less pride than David Bernstein does.
Astrologers may put a great deal of effort into reading astrological signs, and regardless of effort they still can't predict the future.
And people can put loads of effort into talking about books they haven't read yet and still not have much to say about the actual book.
Obviously true but meaningless.
Many people spend lots of time to produce garbage. But you'd rather snark than think it through.
And some people put in absolutely no effort and produce garbage, such as opinions on books they haven’t read.
Excuses, excuses. 162 adults out of nearly 4 million Angelenos.
Probably plenty of ways to criticize the study.
Denying that sampling is a thing is not one of the.
162 from a large population is a pretty poor sample.
For a standard .05 alpha and .8 power, you'd need about 400 samples from a population of 4,000,000.
Of course, this is sociology, so that level of precision is probably not used.
Maybe it should be called what it is, story-time, not anything academically rigorous.
Maybe you should learn about sociology before deciding it’s not science.
Sociology isn't even as sinful in abuse of mathematics as it's sister field economics. And you should consider the research that shows that stories (i.e. narratives) have more impact on belief formation than any mathematics.
Sarcastr0 7 hours ago
Flag Comment Mute User
"Probably plenty of ways to criticize the study.
Denying that sampling is a thing is not one of the."
You might actually try understanding data and statistics. If you did understand stats, then you would know the survey is junk social science.
Ok, chief.
Sarcastr0 29 mins ago
Flag Comment Mute User
"Maybe you should learn about sociology before deciding it’s not science."
You just demonstrated my point
Toranth gave you a good explanation as to why the study was statistical crap. It was pretty easy to spot the multiple statistical crap in the study - yet you not only missed the obvious - you defended the study, sample size as if you understood the data and statistics.
No, he gave an ignorant example that shows he has never examined the practice of sociology even once.
He may have experience in math, but he is criticizing something he has too much pre judged contempt to actually learn about.
Awful practice.
Sarcastro - you just explained why you get so much basic science wrong and why you are easily fooled with every subject. Explains why you were fooled with most every covid mitigation study
No, I made a sound mathematical argument, backed up by a few centuries of sound math, that show the restrictions that a 162 sample size out of 4 million would impose on any study regardless of field.
Pretending that sociology is a mystical field that somehow other people don't get, or that math is different when it comes to sociology, is just dumb.
If I'm wrong, and you know it, then you can actually show me where I am wrong. Show me these mysterious statistical methods that use something other than math.
If you actually know anything on the topic, put up or shut up.
A sample of 162 *could* produce meaningful results if the effect was strong enough... but I really have no idea how you can know that people are telling the truth, or how their own perceptions might cloud things even if they think they're telling the truth. Cultural differences might make one group think the teacher likes them more or less than they actually do, for one. Selective memory could be another factor, especially considering that they're only interviewing adults, some of them apparently two decades out of school.
And keep in mind that this is 162 total in a survey which included *four* racial groups; if they have approximately equal amounts from each group that means only about 40-41 from each group.
With a sample size of 162, even if you DID get strong results, you'd rationally suspect that it meant nothing more than that your sample wasn't random. Because you wouldn't be researching something that was THAT drop dead obvious in the first place!
None of us have read the book - maybe it sucks.
But your baseline attack on sociology continues to come from a place of ignorant analogy to physical science.
Descriptive statistics are not required to predict any trend-lines, and this have a different threshold for significance. Think a particularly robust case study.
A great example was learning the Iraqis have a thing about pointing the bottom of your shoe at them, even accidentally. Was more than a single interview, and was a qualitative and descriptive but nonentheless useful and scientifically derived observation.
In this respect, sociology is like a clinical aspect of anthropology; same ends, different means.
Sarcastro, what you have revealed is that you have no understanding of statistics. You don't even know what it is.
Statistics is a field of mathematics. Every statistical method is math. If it isn't math, it isn't a statistical method.
What you are describing is not a statistical method. It has nothing to do with statistics. I made no comment about that ability to interview people and learn things about those people.
I made comment on the ability to translate those interviews into numbers, perform statistical tests on the numbers, and then apply the results to general populations. That is statistics, and all the related rules of math for the field apply.
And that math is what the author of the book claims to have done (and yes, you CAN read the first 2 of 9 chapters for free - obviously you haven't bothered, though). And I am well qualified to point out that the sample size they used is small, and thus the resulting analysis is low power.
I think you misunderstood.
To say that the students benefited from the perception that Asian-Americans "are smart, hard-working, high-achieving and morally deserving," is not to say that this benefit is the only reason for their academic success.
It is true that the essay focuses on this issue, but as the reviews make clear the book itself takes a more complex view:
Jennifer Lee and Min Zhou provide a theoretically rich and empirically based answer to this question that goes beyond easy stereotypes of Tiger Moms and Confucian values. Their nuanced, convincing argument points to the selectivity of immigrants, the nature of the ethnic community and the reception of Asian Americans by others. Drawing from both sociology and psychology, this smart book should change the national understanding of this important group.
So says Mary C. Waters, M.E. Zukerman Professor of Sociology at Harvard, anyway.
One bien-pensant tenured liberal scratches the back of another. Proves nothing.
Ah. Y81 knows better than anyone. Got it.
Is an n~160 sample size really projectable with any confidence to a disparate (emphasis on disparate) potential Asian college population numbering in the tens of thousands annually? Uh, no.
It's a theory. Get some quant data with methodological rigor, then we can talk Asian-American advantage, is how I would answer the professor.
Perhaps there is a simpler explanation. Among the ethnicities classified as Asian, which ethnicities represent the largest dollar donors to the universities? I would surmise Chinese, Korean, and Indians on the theory that 'money talks'.
You don’t know enough about the study to criticize it’s sample size.
There are sociological techniques for which 160 is more than enough.
You seem to be using physical or medical science as a proxy. Not a safe thing to do for this discipline!
"There are sociological techniques for which 160 is more than enough"
Yeah, there are reasons sociology is looked down on by the 'hard' sciences, and thinking 'technique' can substitute for data is one of them.
I came into understanding social science as a physicist with the same skepticism as you.
Unlike you I had an open mind, and took the time to learn what the practices they use are and why they use them.
And just like that I avoided being a gatekeeping hard sciences asshole.
The idea that qualitative or nonpredictive methods can’t be rigorous comes from your own unexamined assumptions.
Hi, I taught statistical methods to sociology undergrads, and worked with sociology professors to analyze their results before my short career as a professional statistician.
Sociology regularly uses some of the lowest levels of significance or accuracy for their studies of any field. Only psychology ("n=1 is significant!") was consistently worse.
While there are plenty of statistical methods that can be used with smaller than normal sample sizes, they all result in significant drops in the power and/or accuracy of the results.
162 out of 4,000,000 is a poor sample size, even for sociology. The margin of error is large, and power small. Any conclusions you attempt to draw from such a sample are necessarily limited by those factors.
Unlike you, I know what I'm talking about and actually understand the math behind determinations of good sample sizes. Making attacks on other people for accurate describing the problems with the chosen interview sample size is quite revealing of your bias and prejudice.
Sociology is not limited to the quantitative nor the predictive qualities you judge or by.
You can appeal to your own authority all you want but if you attack sociology by misunderstanding sociology you just look like an officious idiot.
Also attacking case studies as a method in psychology.
...
Sarcastro, this may be difficult to understand, but if someone is using statistical methods in any field - sociology, psychology, golf - then the statistical methods they use are subject to the conditions that the mathematical field known as statistics requires.
If a sociologist is misusing a statistical method, then it is no different than if a physicist is misusing one - the output of the method is unreliable.
You have now claimed several times that there are statistical methods, in use by sociologists, that surpass the field of mathematical statistics (however little sense that makes) and are "not limited to the quantitative nor the predictive qualities" of the field.
So go ahead - tell me some of them. Since you claim to know what they are, I'm sure you can do it.
Once again, the standards you are citing are assuming a practice of sociology that is wrong,
You seem quite well up on stats for medicine or physics, but kinda narrow as to the practice of social science.
No shame in that, so was I. But you think anything you haven’t learned about is invalid.
Because you make it very clear you haven’t tried to learn about it. I guess you’ve learned all the statistical practices there are to learn, and see, eager to invalidate entire disciplines based in your own pridefulness on the subject.
Sarcastro, I taught statistical methods to sociology majors.
I understand you don't read well, but this should not be unclear.
When it comes to statistics, then yes I am quite familiar with what sociologists do with it.
If you think there are statistical methods that are "not limited to the quantitative nor the predictive qualities" then show them. You claim to be familiar with such things.
Vague assertions that you know there is something out there that would prove me wrong, but you can't show them (maybe they go to a different high school, eh) just show that you don't know what you are talking about.
Sociology in a nut shell:
Sociologist #1: it is possible eating competitions are sexist (actual thesis I read)
Sociologist #2: based on my research, and others, it is possible eating competitions are sexist
Sociologist #3: based on my research, and others, eating competitions are likely sexist
Sociologist #4: based on an overwhelming consensus of research, eating competitions ARE sexist
Two years later media parrot the finding. Seven years later high school textbooks state unequivocally that eating competitions are sexist. Eleven years later the next gender studies grad eyeing a masters has to go even further in finding sexism to get published.
Really telling on yourself here.
Lots of that isn’t even sociology.
What an ignorant post!
Really? Is that a fact? = You don’t know enough about the study to criticize it’s sample size.
Sarcastr0, I make my living on the private side with applied stats. You cannot project (with any meaningful confidence interval) to a universe of tens of thousands (annually) of different demographic and cultural backgrounds (e.g. asian-american) from a sample of n~160. Sorry, math is math. Sarcastr0, you tell us you have a physics degree; I accept that at face value. Do the math. That aside, apply some reasoning. How many distinct ethnicities are captured within the legal descriptions (in existing federal law) of 'asian-american'...answer: literally dozens. The n~160 sample size doesn't work there, just on that basis alone: it is not truly representative of the universe called 'asian-american'. I should not have to explain this, but Ok, whatever. That's fine.
Now, as to Zhou and Lee's general hypothesis...well, I personally have not ruled it out and I do not reflexively reject it. Their interpretation might well be true (a sort of 'halo-effect' if I had to dumb down the explaining of what they posit); I cannot say. I don't know that answer and nobody apparently does. But I can say, "Professor, you need to test your hypothesis with more rigor. Much more. Because I also want to know if it is true." We might actually agree about that Sarcastr0 (and others)...I really do want to know that answer about modern American society (independent of these cases).
You make your living in physical sciences.
Again, I come from a place being as myopic as you are. But I didn't cling to that when I was shown that different disciplines are different.
None of us should be talking about the book itself; I'm not. None of us have read it. I'm talking about the general discipline of sociology, which too many on here are arguing isn't science because it isn't physics. Including someone who supposedly taught sociologists, while I guess hating their field.
Sarcastr0, I said nothing about sociology. I spoke about how a sample of n~160 cannot be projected to a universe called 'asian-american' because it is not representative. That is objectively true.
Personally, I think the hypothesis is worth pursuing (and answering); that will require more rigor, though. I don't see that as myopic.
This all but guarantees multiple follow-up posts of "I read the book and it's even worse than I thought." These authors are going to get Nancy MacLean'd. The O/U for Bernstein posts on their book is set at 4.5 for those wagering.
Unless the authors simply made up stuff that’s obviously inaccurate to people who know the subject-matter with either no footnotes or footnotes that say something that doesn’t support the text, they are in no danger of attracting Nancy MacLean-like criticism.
That said, being wlling to write something as inane as "I interviewed 162 adults in Los Angeles and figured out that "Asians" have an advantage due to perceived intelligence and that's why they do well in school is hardly to her credit, especially given that, as noted, "Asian" subgroups have wide variations in average success academically.
So, over or under 4.5 follow-up posts after you read the book?
There are many, many bad sociology papers trying to explain either why "Asian Americans" do so well, or, contrariwise, why "Asian American" success is a myth. My "favorite" that I came across is the one that argued that white people "let" Japanese and Chinese Americans succeed because they are light-complexioned, but draw the line at Cambodians and Vietnamese because they are darker. Life is too short to spend a lot of time on this stuff. Give me some data, and by "data" I don't mean 162 interviews in Los Angeles, and I'm interested.
Has anyone ever considered that enough Asians are sufficently rich to ensure a better chance of academic success for their children to create the stereotype that all Asians do well academically, a function of attributing characterstics to a broad category based on a sort of self-selected sample combined with some other pre-existing national and racial stereotypes, and that stereotype has been creating positive expectations which may benefit those within and around that sample in various ways? Of course this might be seen as a variation of the dynamic that has benefitted white people of a certain class for a very long time...
If you want a valid stereotype, it would be the Asian "tiger mom". Being married to a Filipina, I know a LOT of Asian/American couples, and this stereotype is right on the nose.
What you're looking at is simply the result of culturally reinforced diligence. Hours of study and practice enforced by hard nosed parents who treat a "B" as under performing.
Now, this truth makes other groups that don't share that cultural emphasis on academic diligence look bad. Why, one might even suspect it's their own fault they under-perform relative to Asians!
That's a conclusion that's inadmissible, viewed as a species of 'blaming the victim', so there's a strong incentive to find some other reason, ANY other reason, for Asian success.
What does a valid stereotype mean, operationally?
A "valid stereotype" is just a stereotype that's an accurate statistical generalization.
I mean, you are aware that one of the most replicable results in sociology is stereotype accuracy; The fact that most stereotypes are actually statistically valid generalizations, that get abandoned as soon as people have individualized information about a person? It's been discussed here before.
Culturally enforced or otherwise, diligence is no more unique to ‘Asians’ (inverted commas there as a nod to our host’s legitimate concerns about the vast diversity hidden by that category) than ‘tiger’ parenting – the stereotypes accrue to ‘Asians’ for cultural and other reasons, and I suppose some may benefit from them while others resent being judged using then as a standard while others still find them a struggle to live up to. For a certain type of non-Asian suburban parent, they might be dismissed as ‘helicoptering’ or ‘hothousing.’
Who the hell said anything about it being “unique”?
It’s a product of cultural values that work, just as under performing groups have less productive values.
But you can’t say that a group succeeds because of hard work, because this automatically leads to the inference that other groups are failing because they DON’T work hard. And as that inference is unthinkable, success by hard work, too, becomes unthinkable.
And so Asians must be succeeding for some outside, non-merit reasons. Thus the theory we’re discussing.
Not anyone actually familiar with the immigration patterns of various Asian groups which, with the exception of Indians, do not suggest that they arrived in the US with any wealth. Eg, Filipino Americans had the lower average incomes of any ethnic group in California, and now have higher average earnings than whites... Etc.
As long as there are always more applicants than available openings at these schools, I don't know how the Court can craft a workable solution (which the schools are likely to ignore by crafting some new workaround much like some governments have done after Bruen).
Same way they did it after, Bakke, same way after Grutter, same way after fischer , etc
Just development new variations/ new metrics that achieves the same de facto quota.
My wife is Cambodian, and grew up in Phnom Penh just a few years after it was repopulated following Pol Pot's genocide.
I interact with a lot of Khmer here, and I see mostly working class people, not a lot of graduate degrees or tech careers.
Someone should tell them they have an inherent advantage.
It's bullshit, of course. Some people work hard and do well. Others don't. New York Times et al., being good leftists, want to equalize outcomes. So, they come up with stupid (and racist!) stuff like "white privilege" or "Asian privilege."
This post helps explain why David's recent book got rejected by so many academic publishers.
This comment helps explain why QuantumBoxCat has no recent book to be rejected by any academic publishers.
FWIW I have absolutely run into a number of people who assume "Asians are smart" and seen individual colleagues benefit from this where another person would have been called out for mediocre work much earlier. That's an anecdote, not data, but the phenomenon exists.
I won’t comment on the book unless and until I’ve read it, but as long as the subject of Asian-American achievement has been broached:
Let’s say a teacher thinks Asian-Americans are hard-working model minorities. Then an Asian-American student turns in work which is merely average. Would the teacher give that work a higher grade than usual, or a lower grade based on disappointment that the student didn’t live up to the “stereotype”?
“This is A *minus* work – you’re lucky that corporal punishment is out of fashion, or I’d give you a paddling.”
Yup. In my school the coveted grade was A- . ie success without trying. The ultimate insult was C+ . ie well done for trying really hard, but you're still a dolt.
The teachers in my school seemed to be pretty good at divining when I was just going through the motions - which was almost always - and it didn't amuse them.
Recast the discussion to bypass the race/ethnicity baggage. How many think professor types at highly competitive schools always look past athletic prowess, if a student apparently has it, and evaluate without prejudice liberal arts academic work by that student?
Another point about Asian American heritage and academic performance. There may be a sociological disconnect which needs to be applied before you go ahead with race/ethnicity classifications.
Do we know what a family history of recent immigration from various Asian nations implies in terms of social class and educational background? Is there any reason to suppose that if Korean American applicants to Harvard excel, so would all Koreans? Or could it be that Korean immigrants' kids excel because their parents got elite educations before immigrating?
Likewise with any other categories. Perhaps immigrants from some Asian regions or nations tend to be under-educated victims of social disorder, who left behind an educated social elite when they fled. Perhaps immigrants from other Asian regions or nations have been principally members of social elites who fled persecution targeted at elites.
And of course, that kind of historical variability applies potentially to every group of immigrants. How should SCOTUS take that factor into account while probing group academic statistics to discern prohibited motives used by admissions committees?
The easiest way for SCOTUS to handle that would be to ban admissions officers from considering race/ethnicity, and the easiest way to do that, in turn, would be to prohibit schools from giving information about race/ethnicity, beyond what what be implied from activities, eg “member of the Latino students club,” to admissions officers. In particular, ban admissions officers from keeping daily track of who has been admitted from which groups. Whether or not that’s the “Best” way of going about things, it’s obviously “a” way. I can’t be prejudiced when I grade exams because all I have is an exam number, so I couldn’t favor X or Y group even if I wanted to. Indeed, one thing I learned in the course of my book research is that it used to be illegal in many states to ask an applicant their race or religion, because it was thought to be per se discriminatory to ask (or care).
In order for someone from India to get to the US, they often need to obtain an advanced degree at home and/or get into good US college. This often means they are from the Brahmin class, have higher IQ, and have an educational orientation. These immigrants are NOT a random selection of the home population. When Chinese were brought in to build railroads etc 100 yrs ago, perhaps they were, but not now. IQ is highly heritable. Why is it surprising that they do well in school?
Actually, the authors apparently discuss selectivity in immigration as a factor.
The bit you quoted suggests something else, that the authors are investigating the perception of Asian academic superiority, its origins, and to what extent it is self-fulfilling.
You seem to be triggered by references to racial categories, but if you even just scan the preview pages on Google Books you may find more common ground with the authors than you think.
If it's self-fulfilling, why do Cambodian Americans have lower college graduation rates than Americans in general? Why do Chinese Americans do so much better, on average, than Filipinos? And why would you think you can learn the answers to such questions by interviewing 162 adults in Los Angeles?
I mean, I'd be find with the claim, "we interviewed 162 people of different ethnic groups in Los Angeles and among those interviewed we found a widespread perception that teachers perceive students they consider to be Asian to be of higher average academic ability than non-Asians. Further research is needed to (a) see if this holds true in other parts of the US; (b) is a cause of, or is caused by, Asian American students doing well in school, or both; (c) determine why some "Asian" subgroups, regardless of this factor, do better than other subgroups, including some that do below average academically, and to what extent this fact undermines any causal story about Asian stereotyping leading to academic successs.
Perhaps the book gets into all that. If so, then the op-ed itself is tendentious, an attempt to push a perspective relevant to a current Supreme Court case that really isn't backed by the researach. If not, then the book has the deficits noted above.
The op-ed, by Lee, does not just rely on her work with Zhou that is described in the book.
And if we started listing "tendentious" op-eds, and shaming their authors, we'd have a big job to do.
I once read that, even at three meals a day, it is impossible to eat in every restaurant in NYC, because they open and close too fast. I suspect that our task would alsd be impossible for similar reasons.
Hey, it isn't my book so I am reluctant to getting roped into defending it -- I am generally skeptical of the squishy sciences -- but I don't understand the authors to be claiming that the positive feedback dominates all other inputs, only that there is a definite effect.
And, although that bias is the main point of the NYT article, it does not appear to be the main point of the book. For a broader view (without actually buying the thing) you can scan this 2016 summary article by the authors.
The entire article proves it isn't about diversity so much as diversity as crypto affirmative action. Which is fine, just quit lying.
Wtf does success matter if it's about diversity, the benefits to the organization of different viewpoints. But if it's about affirmative action, the proactive amelioration of past racial injustice, then success matters as you don't aid success.
Wow that’s a lot of proving for an article about such a ridiculous study!
The whole legal racialist spoils system is just BS, I don’t understand why we don’t just come out and say it. Pretending that in order to eradicate racism we have to in fact be racist is a contradiction, full stop.
Why are we treating these legal claims as sincere and worthy of serious consideration? The folks who want to divide us by race have a religious-like devotion to racist ideology. You can’t compromise with ideologues, they are thoroughly fundamentalist in this regard. Their minds are utterly captured and enclosed inside the boundaries of that ideology.
I’m not advocating we cancel racialists, I’m just advocating that we treat their beliefs the same way we’d treat any religious belief. We recognize that people may believe in racialism, we acknowledge that those beliefs are deeply held, but we refuse to grant those particular beliefs themselves any special weight in society, and certainly never in the law. What we grant them is the right to freely express and exercise their religion, the same way we grant that to everyone.
Why don’t we just quit with legal process and never listen to whomever DaveM declares is too radical!
That’s not authoritarianism at all!
I love legal process. I don’t like mixing religious fervor with it, that’s all, because it makes what ought to be formal and well-defined into something malleable and deformed. And I say this as a deeply religious person myself. The separation of church and state is an excellent doctrine and we need to return to it.
You don’t get to decide what policies are wise and which are religious nuttery. Neither do government institutions.
While I’m still not convinced of your critique of a book you haven’t read, I am pleased that your response – that actually ‘Asians’ is a category that represents an incredibly broad array of geographical, cultural and racial diversity that does not deserve to be erased – is extremely woke.
Professor Bernstein, should SCOTUS make a list of every kind of admissions criterion which is not meritocratic, and ban them all? Or is it okay to single out race and ethnicity—to distinguish them from other more-favored types of non-meritocratic admissions factors, such as legacy admissions, faculty progeny, geographic diversity, or prospects for donations from wealthy parents?
Is it possible to reason that because race has been such a fraught topic throughout American history, that means it must go uniquely unconsidered among a list of other historically less-troublesome academic irrelevancies? If anyone did conclude that, and singled out race, and only race, as forbidden from consideration, what other than the usual rhetorical tics would defend that against a charge of de jure racism?
If, alternatively, the demand is for a switch to thoroughgoing academic meritocracy—with every other consideration excluded—on what legal basis do you propose to define that, and enforce it? What would empower congress to define merit, and establish meritocracy as law?
If meritocracy is not law, and cannot be law, what justifies a SCOTUS decision to consider whether a private university admissions process must consider any merit-related factor at all? Even if racial discrimination is legally banned, if merit is not legally relevant, what except merit-related issues passes for evidence of racial discrimination in academic admissions? Do you propose to make the standard, "The admissions committee noticed she was black?"
The thing about Lathrop is that no matter how many times you respond to his arguments, refuting the mistaken assumptions on which they are based, he will just longwindedly post them over and over and over again.
If you do a short post mocking his position, he may come back with a 20 paragraph screed about how you're ignoring his points because you aren't a newspaper publisher with a history degree who does geology through grainy satellite photos and also has experience with firearms. But if you write a long comment rebutting his points, he will just run and hide and disappear from the thread, never acknowledging the rebuttal, and then come back posting the same thing in another thread.
He won't shut up about the courts mandating meritocracy, even though that is a complete strawman as I have explained to him repeatedly. Under federal law, Harvard can use any criteria it wants as long as it doesn't consider race (color, national origin). It can pick people with the highest SAT scores. It can pick people with the lowest SAT scores. It can pick great tuba players, or children of government officials, or people who played on the lacrosse team in high school, or it can people at random. But the law says that race can't be a criterion.
Sure, in Bakke, the Court decided that even though the CRA said that, it didn't really mean it, exactly. But it still didn't say anything about this "meritocracy" strawman. And none of this is some brilliant insight of yours; these are well-established policies that courts have implemented for decades without struggling as you are with the idea that discrimination can't be measured without a pre-existing government definition of merit.
Nieporent, as you surely know, what is demanded now is not continuation of decades-long policies, but the abolition of them.
"Is it possible to reason that because race has been such a fraught topic throughout American history, that means it must go uniquely unconsidered among a list of other historically less-troublesome academic irrelevancies? If anyone did conclude that, and singled out race, and only race, as forbidden from consideration, what other than the usual rhetorical tics would defend that against a charge of de jure racism?"
For one thing, the 1964 Civil Rights Act bans discrimination based on race in admissions, and doesn't ban the other things you are alluding to. So, the law is on the side of race neutrality.
Second, you've got it exactly backwards. The history of the US, as well as many other countries, provides evidence that race, as you say, is fraught. More than fraught, it causes wars, division, horrific violence when the government favors one group or another. Religion is also "fraught." So we prohibit the government from favoring one over the other. That's worked out pretty well.
Third, there is a reason that race is subject to so-called strict scrutiny, and other things are not. Because the hallmark of the EP clause is arbitrariness, and singling out people based on their race is presumptively extremely arbitrary. The Court has allowed that presumption to be overcome if universities engage is a narrow, nuanced use of race, and instead, for 44 years universities have taken it to mean, "we can do whatever we want, so long as we don't use explicit quotas."
Finally, my own contribution to the debate is that classifications used are not even "racial." Asians, ranging from Austronesians to East Asians to Caucasian South Asians are in no meaningful way connected to each other, and Hispanic isn't even a race officially, but an ethnicity, the only one to get it's own classification. Native Americans aren't a race, either, as most Americans who so identify are majority "white" in ancestry and come from a variety of tribes who historically ranged from allies to enemies to no contact at all with each other. "Whites" are also not meaingfully any kind of "race," ranging from Icelander to Armenians to Greeks to Arabs to Berbers. Nor were the classifications established to provide "educational diversity," as they are purportedly being used.
Any of these four justifications are entirely good reasons for eliminating race from consideration. You may not like them, but none of them are "racist."
And to add a fifth, "race" isn't even an objective term in the first place. It's a rough social proxy, at best, and even at that it is a ridiculously bad one.
Bernstein, I disagree, respectfully.
I do not for a moment suspect you intend racism. But to permit any and every axis for arbitrary discrimination, but rule out any chance to favor race among the others, is on its face a racist policy.
Whatever the law has been, so long as it did not go so far as that, it has avoided racism. Now the plaintiffs demand that the law go exactly that far, and you support them.
The, “ism,” on the end of the word, “racism,” refers not to inward corruptions of the heart, but instead to systematization. To use the law to set aside race—and only race—as a standard for preferment is systematization aimed to eliminate a claim for preferment which some groups can justifiably claim, and have claimed. To retain legally in force various preferments used to disadvantage those same groups is to turn policy into a taunt; not many blacks are positioned to benefit from legacy admissions in the Ivy League, or to impress the admissions committee with the prospect of an endowed building. That completes in the most authentic Jim Crow tradition a picture of systematic racial discrimination, by adding to an outward show of rational basis, a barely suppressed display of hostility.
You have labored mightily, while collecting for your efforts cheers from the worst kinds of racists. If on no other basis, you would be wise to reconsider advocacy which draws such corrupt support.
That said, please note that your reply mostly ignored my various questions, in favor of restating your own positions. I challenge you to provide answers to each question, to show you can do it without entangling yourself in contradictions.
"The, “ism,” on the end of the word, “racism,” refers not to inward corruptions of the heart, but instead to systematization." That's a recent corruption of the term "racism." Even now, if you google racism and definition, the first two definitions are the traditional ones: (1) prejudice, discrimination, or antagonism directed against a person or people on the basis of their membership in a particular racial or ethnic group, typically one that is a minority or marginalized; (2) the belief that different races possess distinct characteristics, abilities, or qualities, especially so as to distinguish them as inferior or superior to one another.
That said, the notion that if you allow government and major social institutions to classify by race that in the long run this will redound to the benefit of minorities it quite heroically optimistic.
Bernstein — So no luck getting my questions answered. Oh, well.
What I love is that despite your complete ignorance on the subject, you think that you have the capacity to socratically ask questions to give other people understanding.
But I'll play along: no; yes; yes; familiarity with American history, law, and the English language; strawman; the constitution; false assumption, and the constitution; false assumption, examination of the evidence, and statistical analysis; and no.
The simple solution is this.
Title VI bans all racial discrimination period.
No exceptions.
No. Why should they? That's not what the law says.
It's Congress, not SCOTUS, that singled out race. You can't discriminate based on race.
I mean, if you want to argue that the Civil Rights Act is de jure racist...
Good thing nobody is proposing that this be law, I guess, considering that there would indeed be some practical problems.
I mean... in many cases you have the university itself saying they base admissions on race. You may have a case where something is going on that you can't prove, but that's true of everything; I'm sure that someone is getting away with embezzlement at some university somewhere as we speak. The details of whether a particular case can be proven should usually be left to a jury, not SCOTUS.
Too bad the edit function can't edit this into a reply...
No person in the United States shall, on the ground of race, color, or national origin, be excluded from participation in, be denied the benefits of, or be subjected to discrimination under any program or activity receiving Federal financial assistance.
To read the Civil Rights Act in its entirety is to understand its purpose was inclusivity, not exclusion. The fate of Black people in particular were the focus. Everything in the CRA must be read and interpreted in that light.
There is thus no legitimate way to read the text of the CRA to mean, "Anti-racism is the real racism." Yet that perverse rhetorical twist underlies replies above from Nieporent and Davy C. They are the people who in the enduring depiction of Stephen Colbert, say, "I don't see race"—to justify a preposterous attempt to exclude racial issues from discussion, and thus to impose continuing exclusion on Blacks.
The CRA has not changed. The work it was meant to accomplish remains partly incomplete, still notably short of its goals. CRA interpretations and policies followed previously are as justified today as they were when the CRA was enacted.
Other changes now bring forward the affirmative action in admissions issue. What changed were only the names of the Justices on the Court. In short, there appears now—by a malign combination of happenstance and ignominious foresight—to be a Supreme Court majority possibly willing to reinstate Jim Crow in college admissions.
Everyone understands that if the Court overturns affirmative action in college admissions, college attendance by Blacks at elite institutions will decline. It may in fact decline to levels comparable to the early 20th century under Jim Crow.
As has always been true since 1865, the nation includes among its population a substantial fraction still pleased to degrade the status of Black people, if they can, and even, if possible, to strip them of progress won during the decades after the 1960s. Reconstruction was lost following the Civil War. It has never been regained.
Whether or not they have the historical perspective to understand it, advocates like Nieporent and Davy C make the case to accept that lost struggle over reconstruction, and transform it by judicial fiat into a continuing status quo. They hope to find votes on the Supreme Court to back that objective. If their motives come from sources remote from racism, it makes no difference with regard to the outcome. If it comes to pass, what Nieporent and Davy C demand will deliver a shocking racist result.
The Court has capacity to avoid that. It would be wise to craft a limited opinion, leaving affirmative action in college admissions in place.
Yeah, but you also have to read the text. An anti-lynching law might have been passed to protect Black people, but that doesn't mean it shouldn't be used to protect anyone else. And frankly, the law does not and cannot only apply to discrimination against some races and not others, or else it would run afoul of equal protection and be unconstitutional.
And the passage of decades. Pretty much nobody seeking college admission today was ever personally subject to Jim Crow. I would even hazard that most of the students applying now have *parents* who were never subject to Jim Crow. Which means the rationale of correcting specific injustices by a particular institution should pretty much be gone by now.
You think the *exact same colleges* who are currently (and arguably illegally) using race to allow more Black students in, are going to turn around and exclude Black students the moment the Supreme Court lets them? It's not like these colleges are currently *forced* to use affirmative action; they could end it themselves on a whim.
Also, no. SCOTUS is not going to allow Jim Crow. (Removing affirmative action is not Jim Crow, to be clear.)
I guess this is a different Jennifer Lee than New York Times writer Jennifer 8. Lee, author of the Fortune Cookie Chronicles.