TaxProf Blog

Editor: Paul L. Caron
Pepperdine University School of Law

Monday, April 20, 2009

Affirmative Action "Mismatch" at Duke

Inside Higher Ed:  Testing for "Mismatch", by Scott Jaschik:

If members of some minority groups are admitted to elite colleges because of affirmative action -- and don't perform as well as they expected -- does this show a serious flaw in efforts to diversify student bodies?

Critics of affirmative action answer in the affirmative, and this is the basis of the controversial "mismatch" theory -- namely that affirmative action doesn't actually help its intended beneficiaries because they may struggle academically where admitted instead of enrolling at less competitive institutions where they might excel. Mismatch is heatedly debated ....

In a paper released Friday [Does Affirmative Action Lead to Mismatch?], four scholars at Duke University (three in economics and one in sociology) propose a new way to test for mismatch.  ... They propose a test in which applicants admitted to an elite university are asked to predict their first-year grades and are then told the average grades earned by members of similar ethnic and racial groups admitted under similar circumstances. In this situation, they argue, students admitted under affirmative action could make an informed judgment on whether they were being mismatched.

Duke Admissions and Academic Performance Race and Ethnicity






Admissions office evaluations




















--Personal qualities










--Test scores





SAT average





Family income





--Less than $50,000










--$100,000 and higher





Academic performance





--Students' expected GPA





--Students' actual GPA






Legal Education | Permalink

TrackBack URL for this entry:

Listed below are links to weblogs that reference Affirmative Action "Mismatch" at Duke:


It looks like a controversial "mismatch" theory.
It needs to be revised and as per the opinion of Students good changes can be brought.
Online Uk Jobs Resource

Posted by: Sam | Apr 21, 2009 4:58:46 AM

I wonder if one problem with a.a. studies is that they are done so frequently at "top" schools. At Rutgers, diversity is also a factor, but there is a strong sense among the faculty that the gap between minority and nonminority performance has narrowed--we are seeing many more A and A+ performances from minority students and it is no longer unusual for all or most of the best students in a class to come from minority backgrounds. Whether anyone has quantified this, or whether it might result from Rutgers-specific factors like minority scholarships/recruiting, I have no idea. But I can't help wondering if the data in these studies isn't skewed by the concentration on a relatively small number of schools. In the real world, so to speak, I think it is a diminishing issue.

Posted by: mike livingston | Apr 21, 2009 7:38:07 AM

Academic performance is the most interesting category for me.

For every group of students, actual academic performance underperforms expected performance.

Does that say something about the the students? Or does that say something about the predictive skills of the admissions department?


Posted by: Phil Hodgen | Apr 21, 2009 11:50:04 AM

On the surface, it would appear that Admissions Departments do a less effective job than test scores in predicting "student's actual GPA".

Posted by: Jim,MtnViewCA,USA | Apr 21, 2009 11:59:55 AM

A couple of things, from above: "They propose a test in which applicants admitted to an elite university are asked to predict their first-year grades and are then told the average grades earned by members of similar ethnic and racial groups admitted under similar circumstances. In this situation, they argue, students admitted under affirmative action could make an informed judgment on whether they were being mismatched."

1. If the students are asked to predict and THEN told about others, how are they making an informed decision?

2. I teach graduate courses and everyone thinks they are "A" material. Many are only "B" material and argue with you that they deserve the "A" because...
The number of actual “A’ students are few with most probably B+. The B students tend not to last.

From many years of looking at student predictions of their grades both overall and in groups, for undergrad and grad are consistently high and higher than their actual grades.

Posted by: Rich | Apr 21, 2009 12:11:30 PM

This also demonstrates something that has been noted in the California university system - the biggest victim of affirmative action policies is not white people - it's asians.

Posted by: Just Me | Apr 21, 2009 12:14:59 PM

the chart doesn't mean much without evaluating within each block.
So each block in the family income and Admissions office evaluations sections needs a separate X/Y graph to be of any use.
Also add graduation rates and class rankings.
With separate charts an applicant could then look at their own ratings against the earned GPAs and graduation rates and class rankings, ect. of similarly qualified applicants.

Posted by: J'hn1 | Apr 21, 2009 12:16:44 PM

Hmmm... GPA tends to correlate to SAT scores, no matter what their socio-ethni-economic background. Interesting.

Posted by: gmcraff | Apr 21, 2009 12:20:23 PM

Interesting that the white students were the most pessimistic regarding their grades and the black students the most optimistic.

Oh! Sorry! Silly me, I meant to say "Interesting how the grades given to white students matched their expectations, while the black students were cruelly marked far below what they'd no doubt reasonably expected."

What was I thinking? Must check into re-education camp soon.

Posted by: Carl Pham | Apr 21, 2009 1:17:42 PM

Boy, the main thing I notice is that SAT averages seem to have shot up since I took it 30 years ago. The average is about as high as anybody scored back in my day.

Clearly, the most effective educational institution in America is Kaplan....

Posted by: Mike G | Apr 21, 2009 1:19:44 PM

Wondering to what extent the family-income disparities are due to affirmative-action allocation of need-based scholarship rather than demographics.

Posted by: wumhenry | Apr 21, 2009 3:03:35 PM

Interesting, but the raw GPA is only part of the story. They need to normalize by the difficulty of the classes.

Posted by: jon | Apr 21, 2009 3:09:21 PM

Maybe I'm reading this wrong, but it looks to me that even after over-estimating their first year grades by half a point or so, both Black and Latino students are getting around a B. I wouldn't think that would be so bad in an elite school?

Given the students I have who will never get such an opportunity, I'm just glad when they learn enough to earn a C and realize they can do college level work.

Posted by: JorgXMcKie | Apr 21, 2009 3:52:42 PM

I'm having trouble typing this because I'm laughing so hard, but it's obvious that the problem is racial discrimination in the grading of their work while at Duke.

At least, that's the analysis academics and lawyers usually seek to impose on the rest of us.

Posted by: Anon | Apr 21, 2009 4:19:09 PM

So obviously the data shows that there is a much bigger gap between expectations and reality for blacks and latinos. Including information about GPAs for each racial group in admissions letters might help to reduce that gap, but what happens if some prospective students decline offers on the basis of this new information? Will Duke lower standards even more to keep up the numbers of black and latino admissions? Won't the result just be that Duke will start recruiting students who are even weaker academically and, presumably, less able to make good life choices?

Posted by: Dr. Strangelove | Apr 21, 2009 4:28:52 PM

Giving traditionally disadvantaged minorities the opportunity to flunk out of or be at the bottom of the class of America's finest institutions of higher learning doesn't help anyone.

I am sure these students would do quite well in the schools that their academic records qualified them to attend.

Posted by: JimBeam | Apr 21, 2009 4:30:20 PM

It is more humiliating to get a 2.9 at an Ivy League school than to get a 3.6 at a local state schools.

Affirmative action is merely a left-wing mechanism to tear down the self-esteem of blacks (and now Hispanics) to make them more of a left-wing voter block.

Asians present a problem for leftists.

Posted by: Toads | Apr 21, 2009 4:40:12 PM

The authors of the study are obviously racist, but at least they were smart enough to not further subdivide by gender.

Posted by: MMS | Apr 21, 2009 6:04:52 PM

The most horrifying thing on this graph is that the actual GPA sets in somewhere between a B+ and an A-. The worst GPA listed is the Black Actual GPA, a B-. It appears that you have to be an idiot to get a C and perhaps commit a felony to get an F. If failure is not permitted, success is meaningless.

Posted by: Ken Hahn | Apr 21, 2009 10:48:54 PM

What does the students' expected results say about our obsession with their having 'high
self-esteem?' Seems to have instilled unrealistic expectations. Perhaps this is the
first occasion of their receiving actual grades rather than 'gifts' or not.

Posted by: kiwikit | Apr 22, 2009 4:19:31 AM

Controversial? I forgot - reality is controversial for the social engineers of racebaiting.

Posted by: Bandit | Apr 22, 2009 6:27:21 AM

Something is strange here. The average Black SAT score is over 1200. That should be high enough for better than B-/C+ grades at Duke. I wonder what the median Black SAT score is? I'll bet it's one of those cases where there is a significant average/median variation. The mismatch could be worse than suggested by average test scores.

Posted by: David | Apr 22, 2009 1:38:02 PM