Reverse discrimination by the numbers

Academic Questions

Volumn 13, Issue 3, 2000, Pages 71-84

  Lerner, Robert ^a   Nagai, Althea K ^a  

^a Academic Questions   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 77953616728     PISSN: 08954852     EISSN: 19364709     Source Type: Journal    
DOI: 10.1007/s12129-000-1039-3     Document Type: Article

References (35)

1. The term racial is used as a synonym for racial and ethnic and he for he or she.
2. Discrimination is not to be confused with prejudice, societal discrimination, or institutional racism. Prejudice is an attitude, discrimination is a kind of behavior (for a classic sociological treatment of the distinction, see Robert K. Merton, Discrimination and the American Creed, in Sociological Ambivalence and Other Essays, (New York: Free Press, 1976), 189-216. Societal racism and institutional racism are vague pejorative terms without any precise operational meaning.
3. One might define racial preference as a synonym on the flip side for racial discrimination. A situation of no preference could also be considered a neutral standard (e.g., raceneutral admissions).
4. See, for example, Stephan Thernstrom and Abigail Thernstrom, America in Black and White (New York: Simon and Schuster, 1997).
5. As far as we are aware, there has been little systematic study of those who engage in reverse discrimination.
6. The Supreme Court ruled illegal the California system favoring minorities for medical school admissions over whites in Regents of the University of California v. Bakke, (1978). In Podberesky v. Kirwan (1995), the Fourth Circuit Court of Appeals ruled that a blacks-only state-funded scholarship program for college students was illegal. In Hopwood v. University of Texas (1996), race preferences in the form of separate admissions pools based on different admissions criteria for underrepresented minorities and whites in law school admissions were ruled illegal. In Wygant v. Jackson Board of Education (1986), the Supreme Court declared unconstitutional racial preferences through the forced layoffs of white teachers with greater seniority in favor of minority teachers with less seniority. Two Supreme Court cases arose in the past decade, throwing into doubt many federal and state contracts set aside for minorities. In City of Richmond v.J.A. Croson Co. (1989), the Supreme Court declared that strict constitutional scrutiny would be applied to racebased classifications by the state government. In Adarand Constructors Inc. v. Pena (1995), the same was applied to the federal government.
7. See Carl Cohen, Naked Racial Preference: The Case Against Affirmative Action (Lanham, MD: Madison Books, 1995)
8. on the general topic and for an overview of the cases. On contracting, see George R. La Noue, "To the 'Disadvantaged' Go the Spoils," Public Interest 138 (Winter 2000), 91-98, especially 92-93.
9. See such classics in social science research design and methods as Thomas D. Cook and Donald T. Campbell, Quasi-Experimentation: Design and Analysis Issues for Field Settings (Chicago: Rand-McNally, 1979)
10. James A. Davis, The Logic of Causal Order (Beverly Hills, CA: Sage Publications, 1985)
11. Travis Hirschi and Hanan Selvin, The Principles of Survey Analysis (New York: The Free Press, 1973)
12. Morris Rosenberg, The Logic of Survey Analysis (New York: Basic Books, 1968).
13. The correlation should also be statistically significant to insure that the relationship cannot he plausibly attributed to chance. Statistical significance is conventionally established at p < 0.05, meaning that there is a 5 percent or 5 in 100 chance that the relationship is merely a function of statistical chance. The distinction is taken from Davis, The Logic of Causal Order. Statistical nonsignificance means the calculated p-value is greater than 0.05. In other words, we cannot reject the null hypothesis that the relationship is due to chance.
14. The Bell Curve makes a similar argument about race and income when controlling for IQ score.
15. See Richard Herrnstein and Charles Murray, The Bell Curve (NewYork: Free Press, 1994), 321-22.
16. Cook and Campbell, Quasi-Experimentation: Design and Analysis Issues for Field Settings.
17. How large is large enough? The question cannot be definitely answered. Statistician Jacob Cohen however defines a small correlation as one of 0.10, a medium-sized correlation as 0.30, and a large correlation as 0.50.
18. nd Edition, 1988), 79-80.
19. While we describe the strength of a relationship in correlational terms for generality, the logic is applicable to a difference of means or proportions, mutatis mutandis.
20. The articles in Christopher Jencks and Meredith Phillips's book The Black-White Test Score Gap show that these gaps are large and not the result of test bias.
21. See Christopher Jencks and Meredith Phillips, The Black-White Test Score Gap (Washington, DC: Brookings, 1998).
22. Worthless, however, only applies to the logic of scientific causal inference. It does not apply to the logic of the law, where many claims of discrimination are made and many cases are won, though lacking in even the most rudimentary statistical standards. Most notorious are those arising from the disparate impact theory of employment discrimination, whereby any action having a disproportionate effect on some group can be put forth as a type of illegal discrimination.
23. See Roger Clegg, The Bad Law of "Disparate Impact," Public Interest 138 (Winter 2000): 79-90.
24. Cook and Campbell, Quasi-Experimentation: Design and Analysis Issues for Field Settings, 97-98.
25. We are not making a legal argument about signature evidence and statistical disparity, however. With the hope that discovery will turn up hitherto unknown evidence of an actual intent to discriminate, filing suit on the basis of simple statistical disparity is one thing--winning is something else entirely. On legal matters, one ought to seek legal counsel.
26. Robert Lerner and Althea K. Nagai, Racial Preferences in Virginia Higher Education (Washington, DC: Center for Equal Opportunity, 1999). Available at www.ceousa.org.
27. Robert Lerner and Ahhea K. Nagai, Preferences at the University of Virginia (Washington, DC: Center for Equal Opportunity. 1999). Available at www.ceousa.org.
28. Foreign students and students identified by UVA as Missing, Other, Native American, and Unknown in terms of their racial-ethnic group were dropped from the analyses.
29. Whites made up 78 percent of applicants, 72 percent of admittees, and 75 percent of enrollees.
30. The number of admittees in each group is in parentheses and preceded by the phrase, -'n='
31. Correlations assume homoscedasticity, or equal variance among groups, a binary dependent variable such as admission status (reject versus admit) is inherently heteroscedastistic.
32. For a more complete discussion of odds ratios, see David E. Lilienfeld and Paul D. Stolley, bbundations of Epidemiology, Third Edition (New York: Oxford University Press, 1994), 226-228; 316-317.
33. See also Alan Agresti, An Introduction to CategoricalData Analysis (New York: John Wiley and Sons, Inc., 1996).
34. William G. Bowen and Derek Bok, The Shape of the River: Long-Term Consequences of Considering Race in College and University Admissions (Princeton, NJ: Princeton University Press, 1998), 72.
35. Alexander Astin, What Matters in College? Four Critical Years Revisited (San Francisco:Jossey-Bass, 1993).