2/23/2012: That’s the title of the lead post by UO physics professor Stephen Hsu, in a NYT sponsored discussion of racial preferences in college admissions, prompted by the SCOTUS decision to revisit their constitutionality.
2/23/2012: That’s the title of the lead post by UO physics professor Stephen Hsu, in a NYT sponsored discussion of racial preferences in college admissions, prompted by the SCOTUS decision to revisit their constitutionality.
Why the times asked Hsu to comment on this is beyond me. Peter Arcidiacono (Duke) has a phone in his office, doesn’t he?
Hsu’s post is ridiculous and dangerous for this statement alone:
“Would students admitted through race-based preference be better off at somewhat less elite universities where their abilities are better matched to those of their classroom peers?”
The key word in that statement is “abilities”. The assumption is that those scores are a reflection of “ability”. In many cases, according to the research, that is just wrong.
See Claude Steele’s work on the effects of stereotype threat on performance. His work demonstrates an effect that is about negative stereotypes and NOT race (it persists in any situation where negative stereotypes influence performance…women and math for instance).
The argument of merit v. race presents a false dichotomy and ignores the complexity of the variables at play.
I think I should be shocked that we aren’t further along in this conversation. Sadly, I am not.
Good question – why is Hsu commenting on this? What is his expertise?
James Flynn, who knows a thing or two about testing and achievement, has the following to say about meritocracy: “The case against meritocracy can also be put sociologically: (a) Allocating rewards irrespective of merit is a prerequisite for meritocracy, otherwise environments cannot be equalized; (b) allocating rewards according to merit is a prerequisite for meritocracy, otherwise people cannot be stratified by wealth and status; (c) therefore, a class-stratified meritocracy is impossible.” (http://www.iapsych.com/iqmr/fe/LinkedDocuments/flynn1999.pdf)
Or in less lofty terms: differences among 17-year-olds in things like test scores and grades result, in some part, from social inequalities. As a public institution, our mission is to educate people who will contribute to society. Selection of people who are adequately prepared to benefit from a college education is one of several important things that need to be coordinated in order to best serve that goal. Selection is not itself an outcome to be maximized at the cost of perpetuating inequality and privilege.
Ancillary point: Hsu’s argument is just as much an argument against considering class as it is against race.
Re: stereotype threat, etc.
The main point is that SAT validity (predictive power for class rank or graduation rate or STEM success) is about the same regardless of race or SES. (Actually, URM performance is slightly over-predicted.) That is what Arcidiacono et al. find at Duke, what Kuncel and ETS find using UC data, etc. So we know when we admit low scoring candidates through preference (whether racial or legacy), we are setting them up to fail academically.
Below are the results of prop 209, which reduced significantly the amount of racial preference in UC admissions. The overall outcome for URMs is better than under the (earlier) racial preference system which pitted URMs against more able competition at stronger UC campuses.
“4) Graduation trends. URM graduation rates have improved sharply since Prop 209 went into effect. For the six cohorts of black freshmen who started at UC campuses before Prop 209 went into effect (the matriculating years of 1992 through 1997), the average 4-year graduation rate was only 22.2%. For the years since 1998 (matriculating years 1998 through 2005), the black 4-year graduation rate across the UC system is 39.4% — a near doubling. For Hispanics the 4-year graduation numbers are 27.2% for 1992-97, and 41.8% for 1998-2005. Six year graduation rates have risen as well, though less dramatically. Combined with the matriculation trends described above, the number of black and Hispanics graduating from the UC system has been rising steadily and remarkably. For example, the number of black students who matriculated at UC campuses in 2005 and graduated in 2009 was over two-and-a-half times higher than the number of blacks who earned 4-year degrees annually in the early 1990s.”
http://seaphe.org/pdf/analysisoftheeffectsofproposition209.pdf
The commenter above is quick to appeal to credentials as equivalent to expertise. I assume that person is already familiar with the literature I cited, and with the (strongly validated) mismatch hypothesis. For other readers, perhaps these results are enlightening.
My office is on campus just like yours. If you are truly interested in these issues I am happy to talk to you. Don’t hide behind blog comments.
I find tis recent piece in the NYT by david brooks insightful. my apologies to more liberal folk who may take offense at the reference to brooks, as well as to more conservative folk who may take offense at the reference to the NYT! class, race, and issues of merit are ingredients for a witches brew that can become poisonous if the proportion of any one is too dominant. the tough issue is the right balance.
http://www.nytimes.com/2012/02/14/opinion/brooks-the-materialist-fallacy.html
I am very pleased to see Steve Hsu being asked to comment by the NY Times. The data in the Duke study show that test scores are pretty good predictors of academic success, without regard to racial or ethnic identity. Furthermore, students who supposedly are beneficiaries actually are damaged when given admissions preferences. This is tremendously striking in STEM fields where the dropout/failure rate is very high when students “get in over their heads” by being placed with classmates who, for whatever reason, are more academically qualified. The Duke study shows that those same students have much higher rates of success in STEM fields when they attend schools where they are more competitive. People can pretend that these data don’t exist, but that doesn’t make them any less true or relevant.
This effect of misplacement and appropriate placement is reflected in an immensely striking way at the University of California. The voter-imposed elimination of racial preferences in admissions has actually led to much higher rates of minority success in STEM fields. This seems to be a real-world application of the principle described above: get into Berkeley through racial preferences, find yourself outclassed, become demoralized and drop out or fail. Or, get into one of the less competitive UC schools solely on your merits, rise, shine, and succeed. That is what is happening – the racially neutral policy is actually helping the minorities supposedly being helped by the now-illegal racial preferences! Or – better to be in the top 10% at UO than the bottom quarter at Harvard!
Another aspect of this which Hsu doesn’t mention, but which I will, is the gross discrimination carried out against students of Asian background, associated with the use of racial preferences, most notably in the Ivy League. This is not all that much different from the very ugly discrimination against Jews in the Ivy League in the bad old days, which only ended in the 1950’s or later, greatly facilitated by the accession of SAT and other testing. I only wish that Asian-Americans were more assertive in protesting this abominable discrimination, the way Jews were back in the day. (There were other things the American Jews should have been protesting against, more than they did, i.e. the exclusion of their European family members from entry into America when they were being threatened and then murdered by the Nazis. But that is another unpleasant story in American racial and ethnic relations.)
In short, racial preferences are damaging to everyone – underrepresented minorities, Asians, whites, everyone, whether the supposed beneficiary or the victim of the discriminatory policies. Discrimination was wrong when it was used against blacks and Jews and others, and it’s wrong when it’s used now supposedly to benefit historically oppressed groups. It’s destructive, it’s wrong, it hurts everyone, let’s hope the Supreme Court finally puts a stop to it – and let us praise the people of California and other states who put a stop to it by referendum.
As for whether Hsu is worthy to comment in the New York Times – I’m glad they think so! I don’t see why the issue should be left to lawyers, journalists, and sociologists. Hsu is probably better-versed in the relevant statistics than all of them put together. He has been blogging about these issues for quite a while. I’m very happy that people are paying attention to his sensible commentary.
Socko work, Hsu and eloquent Anonymous!
I may be a little late to the party, but isn’t there a problem with praising SAT scores as predictive measures for college achievement when that’s precisely what they were designed to do? It seems like that behavior creates a closed system where the test created to tell us who will succeed in a pre-constructed learning environment is used to determine who gets to be in that learning environment, and then the process repeats ad nauseam. But what happens if the values and structures of the environment are unjust? Then the test would actually be predicting how well students will succeed in a biased environment, which seems like it would then invariably tell us that Black and Latin@/Chican@ students can’t succeed…
I don’t see that succeeding at doing what it is intended to do — predict college achievement — is a reason to fault the SAT! I’d say that it’s up to the critics to demonstrate that “the values and structures of the [academic] environment are unjust” and “biased.” But I don’t see any reason whatsoever to think that academic physics or chemistry or electrical engineering (to pick a few STEM fields) is based on unjust values or bias. Nor that the tests tell us that the groups mention “invariably can’t succeed.” For whatever reason, the tests predict that certain groups will perform much better than others. And that is what happens, on average (of course — lots of variations among individuals.)
Using the tests WITHOUT bias apparently also results in much better outcomes for the underperforming minority groups, as discussed in some of the previous posting. That seems like a pretty good argument in favor of using the tests!