Wednesday, February 03, 2010

Affirmative Action: Costly and Counterproductive

In the debates and court cases on using affirmative action to increase minority enrollment at highly selective colleges, it has been a self-evident article of faith among supporters of affirmative action that the educational experience is deficient and inferior without sufficient minority representation, and can therefore be enhanced and improved by artificially increasing minority enrollment, usually by lowering admissions standards for targeted groups. These proponents usually suggest that all students benefit from affirmative action, including both the majority-race students at selective, elite colleges, and minority student groups at those colleges.

That position is now being challenged with contrary statistical evidence in a
forthcoming paper in Economic Inquiry "Does the River Spill Over? Estimating the Economic Returns to Attending a Racially Diverse College," by two Duke University professors Peter Arcidiacono and Jacob Vigdor. Here are some key excerpts:

Introduction: "For more than a quarter century, the belief that diversity contributes to the quality of undergraduate and graduate education has motivated court opinions and college policies regarding racial preferences in admissions. Surprisingly the social sciences have provided very little evidence to support or refute this claim. Such evidence would clearly be of great interest both to policy makers and to scholars conducting more general studies of the impacts of affirmative action on higher education."

Summary: "Do white and Asian students at elite schools benefit from the presence of Under- Represented Minority students on campus or in the college classroom? While not all the evidence in this article suggests that interracial exposure is uniformly negative, it strongly suggests that the predominant policy tool designed to increase the representation of minority groups, affirmative action, has a negative net impact on students not directly targeted by the program.

Using data on graduates of 30 selective universities, we find only weak evidence of any relationship between collegiate racial composition and the postgraduation outcomes of white or Asian students. Our empirical results cover a broad range of outcomes, including earnings, educational attainment, and satisfaction with both one’s life and one’s job. Across these varying specifications, we fail to find any significant evidence that white or Asian students who attend more diverse colleges do better later in life. Moreover, the strongest evidence we uncover suggests that increasing minority representation by lowering admission standards is unlikely to produce benefits and may in fact cause harm by reducing the representation of minority students on less selective campuses.

Further analysis suggests that affirmative action is actually counterproductive, if its goal is to improve the productivity of majority race students. Preferential admissions for certain groups may still have a role in higher education, but they should be understood for what they are: redistributive mechanisms that create benefits for the targeted racial groups but costs for others."


At 2/03/2010 10:21 AM, Blogger juandos said...

Well this is an argument that's hardly new...

Stanley Rothman, professor emeritus of government at Smith College, is director of the Center for the Study of Social and Political Change had this editorial back in '03 in the New York Times: Is Diversity Overrated?

Note these two sentences from his editorial: 'If diversity works as advertised, we surmised, then those at institutions with higher proportions of black enrollment should rate their educational and racial milieus more favorably than their peers at institutions with lower proportions.

The results contradict almost every benefit claimed for campus diversity. Students, faculty members and administrators all responded to increasing racial diversity by registering increased dissatisfaction with the quality of education and the work ethic of their peers. Students also increasingly complained about discrimination

La Shawn Barber who I find to be sharp and as a rule brutally honest has her Race Preferences which she seems to think is both shameful and wasteful...

Charles Murray noted We Can't All Make the Grade...

At 2/03/2010 10:27 AM, Anonymous CompEng said...

I absolutely agree: any evaluation of the concept of diversity ought to start from honesty and objectivity. These traits have been sorely missing from such discussions.

At 2/03/2010 11:17 AM, Anonymous Anonymous said...

Logic does not matter to the people who believe in affirmative action. AA and political correctness trump all other concerns to these people, and make no mistake, they are in charge right now. Trade AA and PC for the lives of a few soldiers at Ft Hood, or the lives of 300 or so people on an airplane, or who knows what for the next terrorist event. Never, AA and PC are more important than a few American lives (to these people). Final thought: AA most hurts the people it supposedly is intended to help.

At 2/03/2010 11:45 AM, Anonymous Rand said...

I can see how affirmative action (lowering admission standards) can benefit the non-affirmative action students also. It provides additional people to fill in the lower part of the distribution - especially when the professor grades on the curve.

All it does is make the Asian students look even better.

At 2/03/2010 12:16 PM, Blogger Jared Field said...

This is one of those situations in which manufacturing a premise doesn't serve the overarching goal. I'm not necessarily against Affirmative Action, but I'm roundly against making any case on bogus grounds.

Love your blog, Professor.

At 2/03/2010 12:42 PM, Blogger al fin said...

Affirmative Action is a dysfunctional policy. But so are US tort laws, US labour laws, US energy rules, regulations and taxes, US capital gains policy, and on and on.

The modern US government is a tragic conglomeration of misguided and destructive policy -- administered by a huge overbloated bureaucracy.

There is no reform in the world that can correct this situation.

Remember how communism in Eastern Europe fell? Something of a similar nature would have to happen for US dysfunction to be swept away.

But if it happened to the US, there would be no one standing on the sidelines to guarantee the peace or to offer a helping hand. The US is different, it is the only superpower, it is the world's pivot.

A dysfunctional pivot.

At 2/03/2010 12:45 PM, Blogger Hey-nonny Bosh said...

I will first list my (unoriginal) concerns with AA, then one possible benefit.

Preferential admissions merely highlight the differentials and PERPETUATE negative stereotypes. For example, by lowering standards for one group (let's call them, for lack of a better term, "blacks") one is forced, due to scarcity of slots, to run up the qualifications of other groups (let's call them "Asians"). That is, by artificially reducing the availability of a good, the "price" on the remaining good increases.

One unintended consequence: students in the non-preferred groups are able to maintain potentially unwarranted biases, e.g. "he only got in because of affirmative action." In other words, the "true" accomplishments of blacks can always be undermined or questioned.

Fear can lead the under-qualified into academic ghettos (e.g., Af-Am studies) where grading is easier. This, again, leaves true ability untested (see note above).

Also leads to higher dropout rates and non-integration, as well as eventual discovery of limitations in the workplace; cf., the "Peter Principle."

AA sucks students away from their "natural" educational progression. State schools, for example, which might be the perfect venue for a first-generation college student, must ALSO lower standards as elite colleges consume all available applicants, both qualified and less so (with, as we discussed, the distinction between the two unfortunately obscured).

All that said, I do note one real benefit: let's call it "cultural capital."

While some blacks may be under-served and overwhelmed IN an elite college, they nevertheless become... graduates. And let's face it: most (MOST) jobs are less rigorous than many college courses. A Harvard or Williams graduate, regardless of actual qualifications, has ceteris parabis a greater-than-otherwise access to OPPORTUNITY (e.g., the alumni network, general prestige, etc.). This can then funnel down, releasing the graduate and the graduate's offspring from "the cycle of poverty," one family at a time.

This is a very cynical view, and requires top colleges to view their "brand" as just that: a brand (or commodity), a tool to get somewhere else... And I think that is indeed what the administrations of top colleges do, in large part (MIT, William & Mary, and some other notables mostly excluded...).

Is this "benefit" worth all the detriments? Apparently so...

It is the same issue with "talented and gifted" programs in public secondary schools: why worry about "them?" "They" will do all right REGARDLESS; better to layer resources onto the, uh.... "less fortunate," "underprivileged," and (last bastion of affluent whites) those with "special needs."

One other unintended consequence: AA-advocates get all in a tizzy over the "legacy" issue (where alumni offspring have a perceived admissions advantage). They fail to note, however, that AA "graduates" beget... LEGACIES! (Perhaps they want to rely on the "preferred class" status indefinitely, I guess).

At 2/03/2010 3:20 PM, Anonymous Anonymous said...

Notional evidence that AA can create an unintended legacy: In an Air Force technical training course in 1979, I was paired up with a black 2nd Lt from Washington DC and an obvious AA graduate of the University of Maryland. This guy was so bad he couldn't tie his own shoes and was quickly washed out of the training course. Unfortunately for U of Md, he made a lasting impression on me, and probably the course administrators, of the U of Md.

At 2/03/2010 6:56 PM, Blogger Craig Howard said...

I wonder if the diversity-benefit-to-whites shibboleth isn't just an attempt to hide the politically-incorrect belief that minorities benefit from exposure to higher-performing, um, majorities.

At 2/04/2010 12:12 AM, Blogger Marko said...

My first reaction is to join in and say bad things about AA, but then I remember that my primary value is liberty. Shouldn't universities be allowed to do whatever they want? Why should I care if they give preference to underrepresented 'racial' groups, or to rich white people as long as there is a free market?

At 2/04/2010 1:28 AM, Anonymous Lyle said...

Just recall that until fairly recently top schools imposed quotas on Jewish students and more would have gotten in if religion blindness had been in play. Also I suspect there is a limitation on asians right now. Note also that the biggest unstated AA right now is for Men over women as Mark has pointed out Alberta schools which don't do gender AA show a 2/3 women 1/3 men ratio. US schools regard that lack of diversity as unhealthy. If you do away with AA you will see the gender ratios go up to 2/3 1/3 instead of 60 40. If we do away with AA we need to do away with it completely.

At 2/04/2010 6:18 PM, Anonymous Mork said...

Yes Lyle, the best plan would be to do away with AA entirely. AA is a policy of institutional injustice which sows the seeds of bitter acrimony for the future.

Remember, Marko, underrepresented minorities are underrepresented for a reason. It is far better to address the reason for underperformance than to thrust an underperformer into a situation where he is virtually certain to fail.

Liberty is for individuals, not for tax supported institutions. Institutions are meant to provide maximum value for constituents and customers, without regard to race, sex, religion, or age.

At 2/04/2010 11:22 PM, Anonymous Lee said...

I have seen research (sorry do not have link) that indicates that students that received preferential admissions tend to perform just as well as other students after graduation.

So while the process of admissions may be unfair, it does not appear that admissions criteria are particularly effective at identifying the "best and the brightest." On a related note, college grades also do not seem to be predictive of future success.

I would suspect that this would be most true at the most selective colleges, where the vast majority of applicants are able to do the work and the ones that are admitted are ultimately those who can best game the system, either through AA, athletics, access to a large number of AP classes, legacies, etc.

A better (and perhaps fairer) solution than AA would be for colleges to establish a minimum threshold for admittance, and then admit those students meeting these requirements by lottery.

At 2/05/2010 12:10 AM, Blogger fboness said...

The absolute best affirmative action program was the GI Bill. It broke down the pre WWII barriers to entry to elite colleges and gave a massive boost to the skilled workforce that built up this country in the years following WWII.


Post a Comment

<< Home