News By Tag
News By Location
Follow on Google News
Harvard, Like Princeton, May Use Tobacco Defense Against Racism Charge
But Better Investigation Under New Administration Likely to Find Discrimination
As just reported in Inside Higher Ed [IHE], the stated reason why the government found in a prior investigation that Princeton doesn't discriminate against Asians, despite the overwhelming statistical evidence to the contrary, sounds very much like the deceptive arguments the tobacco industry made in the past to try to convince the gullible that smoking doesn't cause lung cancer.
Their spokesmen would point out that, because most smokers do not get lung cancer (which is true), and that some people who never smoked do get lung cancer (which is also true), smoking doesn't cause lung cancer.
But if that type of argument were valid, it could be used to prove lots of things: e.g., drunk driving doesn't cause automobile accidents. Because most drunk drivers don't have automobile accidents, and because many accidents occur even where the drivers are not drunk, it should be obvious, by the same reasoning, that drunk driving doesn't cause accidents.
In this light consider that IHE says Princeton's successful defense can be summarized this way: many non-Asians with great academic records are also rejected by Princeton, and some Asians "could also be found among some of the less than perfect applicants" who were admitted, so obviously, by this reasoning, there is no discrimination against Asians.
But this no more proves that Princeton was not discriminating against Asians applicants than the evidence cited above proves that smoking doesn't cause lung cancer, and that neither drunk driving nor jaywalking causes automobile accidents, says Banzhaf.
In many discrimination cases, the overwhelming strong statistical evidence against Princeton (now Harvard) would have shifted the burden of proof to the respondent, and forced it to show that there were one of more other factors which accounted for this vast discrepancy.
Only if the university could then prove that these other possible factors, what the article calls "qualities and the overall strength of their applications,"
Sociologist Thomas Espenshade is quoted as suggesting that these factors might include things like "teacher and high school counselor recommendations, essays, and lists of extracurricular activities."
But in virtually all cases these three factors - and others which might be suggested by the respondent - could be evaluated on a double-blind basis (e.g., outsiders evaluate them with any racial references removed), and the correlation between each, and the decision to admit, could be calculated.
If this were done, it is possible if not likely that discrimination - not one or more so-called holistic factors - would be the cause, since - as with smoking - the evidence is overwhelming.
That, by the way, is what researchers did with regard to smoking. Factors other than smoking - e.g., a family history of lung cancer, industrial exposure to certain toxins, exposure to radiation while flying, living with a smoker, etc. - all can play some role in causing lung cancer. But since none of them come anywhere near the effect of smoking, it was fair to conclude that smoking is the major cause of lung cancer.
For example, in studying the odds of admission of various groups, Banzhaf cites reports that Asian Americans have to score on average about 140 points higher on the SAT test [8.7% more on a scale of 1600] than white students; a very large and well known gap sometimes called "the Asian tax."
Even stronger evidence of discrimination is that Asians had to achieve an astonishing 450 points [28%] higher SAT score than black students to equal their chances of gaining admission to Harvard, even after controlling for other factors known to affect admission decisions.
While the large advantage in admission odds of blacks compared to whites [310 points or 19%] might arguably be justified because blacks are a minority which has historically been oppressed, especially by whites, the much larger negative gap in admissions percentages of blacks compared with Asians [450 points, 28%] cannot even arguably be justified by the same arguments, since Asians are a minority group which has likewise been oppressed historically in this country, and not by African Americans.
In other words, even if discriminating against white applicants might be justified because whites discriminated against blacks in the past, and thereby acquired what some have called "white privilege," that would not justify discriminating against a minority which itself has suffered discrimination.
Another very striking indication of unfair discrimination occurs when one compares the percentage of various minorities at top schools in states where racial discrimination is prohibited with those where it is permitted and appears to be practiced. For example, Harvard's 2013 Asian-American enrollment was about 18%, and other Ivy League colleges such as Brown, Columbia, Cornell, Princeton and Yale had similar percentages.
In stark contrast, the percentage of Asians during that period in top California schools was so much higher, and could hardly be based upon a statistical fluke, or on a tendency of high scoring Asians to apply in California where they are protected by a law banning any racial preferences:
Good and careful investigators would require Harvard itself to identify other factors which cause these huge disparities, and to see to what extent they correlate with the admission rates for Asians. This could be done by using correlation analysis and/or multiple regression analysis which, Banzhaf predicts, will show that discrimination, not other so-called holistic factors, are responsible for these shocking disparities.