"In the 1996-1997 admissions cycle, some 2,646 white applicants placed in the top 7.7 percent of LSAT-takers and had college GPAs of 3.5 or better. These credentials are very good, obviously, but not phenomenal; the average student who was admitted to Boalt Hall this year had an LSAT in the 97.7 percentile and a GPA of 3.74. And yet a mere 16 African Americans in the United States and 45 Hispanics had records that strong! In this elite group of applicants, whites outnumbered blacks 165 to one. If we relax the standard substantially and look at students in the top sixth (83.5 percentile or better) on the LSATs and a GPA of only 3.25, 7,715 whites and just 103 blacks qualify. In this broader elite, there were 75 whites for every African American." (See Diversity and Meritocracy in Legal education: A Critical Evaluation of Linda F. Wightmans's "The Threat to Diversity in Legal Education."; Response to Linda F. Wightman, New York University Law Review, vol.72, p. 1, 1997 Constitutional Commentary March 22, 1998).
The differences are even more astounding when you consider that many self-identified blacks and hispanics are actually mixed race (like Obama). Adjusted for the difference in the sizes of the white and black populations in America that 165 to 1 still comes to 30 to 1. If there were as many blacks in America as whites, there would still be 30 times as many highly qualified whites as blacks. This is obviously not a convenient fact socially. Inequality creates tensions and affirmative action is imposed by the elites to keep the people pacified. They will use this strategy for so long as it produces pacification. Of course, if it ever ceases to work, instead of treating the symptoms of pathology (racial inequalities in achievement levels), they might then adopt measures to treat the pathology itself (lower black IQ).
I've drawn up a rough statistical comparison of LSAT scores by ethnicity extrapolated from numbers from this report: Law School Admission Council, LSAT Technical Report 98-04, March 1998. The most recent scores are virtually identical to those in this report. The left hand column is LSAT scores, the numbers in the next 4 columns represent the percentage of each ethnicity who equalled or exceeded the matching score.
| median | 152 | 151 | 146 | 142 |
| SD | 9 | 10 | 9 | 8 |
| LSAT Scores | White | Asian | Hispanic | Black |
| 142 | 86 | 84 | 70 | 50 |
| 143 | 84 | 82 | 65 | 45 |
| 144 | 82 | 79 | 60 | 40 |
| 145 | 79 | 76 | 55 | 35 |
| 146 | 76 | 73 | 50 | 30 |
| 147 | 73 | 69 | 45 | 26 |
| 148 | 69 | 65 | 40 | 22 |
| 149 | 65 | 60 | 35 | 19 |
| 150 | 60 | 55 | 31 | 16 |
| 151 | 55 | 50 | 28 | 13 |
| 152 | 50 | 46 | 25 | 10 |
| 153 | 45 | 42 | 22 | 8 |
| 154 | 40 | 38 | 19 | 6 |
| 155 | 35 | 34 | 16 | 5 |
| 156 | 31 | 31 | 13 | 4 |
| 157 | 28 | 28 | 10 | 3 |
| 158 | 25 | 25 | 8 | 2 |
| 159 | 22 | 22 | 7 | 1.7 |
| 160 | 19 | 19 | 6 | 1.4 |
| 161 | 16 | 16 | 5 | 1.1 |
| 162 | 13 | 13 | 4 | 0.9 |
| 163 | 10 | 11 | 3 | 0.7 |
| 164 | 8 | 9 | 2 | 0.5 |
| 165 | 7 | 8 | 1.7 | 0.3 |
| 166 | 6 | 7 | 1.4 | 0.2 |
| 167 | 5 | 6 | 1.1 | 0.1 |
| 168 | 4 | 5 | 0.9 | |
| 169 | 3 | 4 | 0.7 | |
| 170 | 2 | 3 | 0.5 | |
| 171 | 1.7 | 2 | 0.3 | |
| 172 | 1.4 | 1.7 | 0.2 | |
| 173 | 1.1 | 1.4 | 0.1 | |
| 174 | 0.9 | 1.1 | ||
| 175 | 0.7 | 0.9 | ||
| 176 | 0.5 | 0.7 | ||
| 177 | 0.3 | 0.5 | ||
| 178 | 0.2 | 0.4 | ||
| 179 | 0.1 | 0.3 | ||
| 180 | 0.2 |
Only half of LSAT test takers actually go to law school and these attendees tend to be in the top half of each ethnicity's scoring distribution. This means that a 163 will place a white in the top 10 percent of test takers, but only about the top 18-20 percent of law students. Given that 152 is the median white score, it's fair to assume that few whites with sub-152 scores attend law school. So 152 is effectively the minimum expected from white applicants as far as admissions officers are concerned. If this minimum were imposed on blacks, just 10% would make the cut at any law school. About 15% of law students attend top 20 schools, which are those with nationwide reputations and prestige. This corresponds to a minimum LSAT of 165 for whites and asians. Only 0.6% of black law students score that high. On merit, they would represent less than 0.1% of the student populations at top 20 schools--in reality they are closer to 7%. A 70 to 1 ratio of charity cases to merit cases is impressive. This is a religion forbidding doubts. The black minimum score, then, is 156 at these schools--a full standard deviation below the white minimum. Of course, this makes sense insofar as the median black score on the LSAT is a standard deviation lower than the white median. Coincidentally, the black-white IQ gap has stood at a standard deviation for generations. The education system in this way achieves a superficial unity at the expense of fundamental inequality, de facto and de jure.
On merit, blacks would run to 0.25% of students at the top 50 law schools (which combined graduate a third of law students). And the study quoted at the top of the post estimated that, depending on the predictive model used, between 2% and 3.4% of all law students would be black if admissions were color-blind. Her methodology is quite generous to blacks. To estimate race-blind admissions rates she looked only at LSAT scores and undergraduate GPA. The thing is, white students at any given college have higher IQs than their black peers (due to undergrad affirmative action). I guarantee that people with higher IQs are more likely to take difficult classes than lower IQ people. Result: white and black GPAs are not comparable. An "A" in precalculus hardly carries the same signification as an "A" in multivariable calculus. She certainly overestimates the black representation that would be expected from a more thorough merit-based admissions process. But, even if you accept an average of her numbers, 2.7% is much lower than the 13% of the American population blacks represent. It's a 5 to 1 shortfall. This exactly matches my statistical estimates in the table above: if blacks had to meet the white minimum of 152 to get into any law school, only 10% would make it, compared to 50% of whites. This is slightly misleading, though, since only 10% of test takers are black and my table is based on test takers, not population numbers. By my numbers, 10% of the 10% of test takers who are black would make it--which means they would compose 2% of law students, a 6.5 to 1 shortfall.
The analysis I've done here showing the effect of affirmative action on law school admissions applies in similar fashion to undergrad admissions and admissions to any competitive graduate or professional programs. They are all based, to a significant degree, on pure (eg, SAT, LSAT) and applied (GPA, courseload) cognitive ability. The IQ differences between groups predicts to a high degree of accuracy, as shown above, the academic qualifications that characterize each ethnic group. I do not think most Americans have any notion of the magnitude of the assistance given to affirmative action cases. Even supporters of this racial strategy might be privately grieved by it. My purpose here was to pull out a variety of numbers to give some perspective on the issue.
No comments:
Post a Comment