What Happens When You Get Rid of Affirmative Action?

A new working paper (abstract; PDF) by economists Peter Arcidiacono, Esteban Aucejo, Patrick Coate, and V. Joseph Hotz looks at the effects of California’s Proposition 209 on university matching:

Proposition 209 banned using racial preferences in admissions at California’s public colleges. We analyze unique data for all applicants and enrollees within the University of California (UC) system before and after Prop 209. After Prop 209, graduation rates of minorities increased by 4.4%. We characterize conditions required for better matching of students to campuses to account for this increase. We find that Prop 209 did improve matching and this improvement was important for the graduation gains experienced by less-prepared students. At the same time, better matching only explains about 20% of the overall graduation rate increase. Changes after Prop 209 in the selectivity of enrolled students explains 34-50% of the increase. Finally, it appears UC campuses responded to Prop 209 by doing more to help retain and graduate its students, which explains between 30-46% of the post-Prop 209 improvement in the graduation rate of minorities.

One caveat: the study doesn’t address outcomes for students who didn’t attend University of California schools as a result of the change.

Leave A Comment

Comments are moderated and generally will be posted if they are on-topic and not abusive.

 

COMMENTS: 32

View All Comments »
  1. Eric M. Jones says:

    The concept of “minority” or “mixed-ethnicity” or “non-white” will become harder to define, especially when there is some benefit from being so-defined. The key reason for affirmative action is that many believe that a better society will result. Alternatively one could simply grease the skids for young people who show intelligence and talent and not consider ethnicity at all. Doing both is not a mistake either.

    ps: There are many ethnicities and skin colors. Even so, many “blacks” are of many different ethnicities…Papuan and Austronesian for example. But there is only one race…human.

    Well-loved. Like or Dislike: Thumb up 20 Thumb down 5
    • George says:

      Hidden due to low comment rating. Click here to see.

      Disliked! Like or Dislike: Thumb up 2 Thumb down 7
    • Geroge says:

      Hidden due to low comment rating. Click here to see.

      Disliked! Like or Dislike: Thumb up 1 Thumb down 6
      • Vince says:

        African-American is an ethnicity, black is a skin color. There are countless numbers of ethnic groups that share a black skin color throughout the world.

        Thumb up 6 Thumb down 3
      • James says:

        Tell that to the Tutsi and Hutu. There are in fact an uncounted number of ethnicities among Africans, some of which carry over to the US, along with new ones that have developed. Or consider the politically correct types who dump all American Indian tribes in their catchall “Native American” category, ignoring the great social, cultural, and ethnic differences seen by members of the tribes themselves. Or even try, as an American, to distinguish between a Serb and a Croat, or between Protestant & Catholic in Northern Ireland.

        Well-loved. Like or Dislike: Thumb up 6 Thumb down 0
  2. Chad says:

    It’s nice to finally see a study take a hypothesis that is capable of providing conclusions that dispute the value of affirmative action.

    “The way to stop discrimination on the basis of race is to stop discriminating on the basis of race”

    Well-loved. Like or Dislike: Thumb up 60 Thumb down 6
  3. Malcolm says:

    Hidden due to low comment rating. Click here to see.

    Disliked! Like or Dislike: Thumb up 33 Thumb down 39
    • JAM says:

      “…Affirmative Action is about helping disadvantage minorities who are disproportionately in poor neighborhoods and under-performing schools…”

      It seems that you could accomplish a very similar effect by simply focusing on helping those who are disadvantaged regardless of minority status. If a minority is disproportionately disadvantaged, then they will be disproportionately benefited by a program that is colorblind. Further, you would also avoid the moral dilemma of ignoring others who are disadvantaged but are not the right color to receive your help.

      Well-loved. Like or Dislike: Thumb up 48 Thumb down 2
      • Enter your name... says:

        That’s what Malcom recommended: “My opinion is that the program needs reform rather than elimination. There should be family income caps on eligibility to target the students based on disadvantaged status rather than simply ethnicity. ”

        Perhaps you didn’t read the whole post?

        Thumb up 8 Thumb down 6
      • Nil says:

        “It seems that you could accomplish a very similar effect by simply focusing on helping those who are disadvantaged regardless of minority status. If a minority is disproportionately disadvantaged, then they will be disproportionately benefited by a program that is colorblind. ”

        You cannot get anything close to similar results due to the significantly higher success rates for Asian students who by every measure except for race would appear to be disadvantaged. Asian students have been the biggest beneficiaries of getting rid of Affirmative Action, yet many of those same students grew up in poor California neighborhoods with poor parents and attended lousy inner city schools. Statistically middle class Asians only slightly outperform their middle class peers, but at the bottom of the economic scale there is a much larger gulf between the test scores of poor Asian children vs other poor children.

        Well-loved. Like or Dislike: Thumb up 9 Thumb down 2
      • JAM says:

        In response to Nil

        My casual observation on what you are describing in the lower income Asian group performance success is that most of these are likely 1st and 2nd generation immigrants coming from societies that place a high value on education. Maybe some of this is due to rationing of education in those places.

        However, I don’t see how educational assistance programs that are colorblind would hurt this group relative to any other.

        Thumb up 1 Thumb down 2
      • Malcolm says:

        There should absolutely be programs to benefit all low-income students to recognize the circumstances around their economic achievement and consider that along with standardized test scores. But in the real world I haven’t seen anything close to the reform or financial investment in low-income schools that would actually accomplish this.

        Affirmative Action has always been a band-aid fix. If we could fix the root problem of under-achieving primary schools for all poor students, then I’d absolutely agree that we should go that route instead. But that hasn’t happened, and I don’t see it happening any time soon. Instead, AA is a pragmatic approach that aims lower than comprehensive education reform and targets money at under-performing minority groups to try and balance out existing gaps between ethnicities.

        Without a doubt, the racial attainment gap and generational income / educational mobility are both important social problems. Affirmative Action is essentially a policy that puts racial equality ahead of class mobility. While the later is a bigger problem, and one that is at the root of the former, I’m not convinced that society is ready to take it on. I’d rather have a focused and well funded campaign targeting the smaller issue rather than a weak and ineffectual effort at the larger.

        Thumb up 8 Thumb down 5
      • Vince says:

        @Malcolm: The root problem isn’t under-achieving schools, it is poverty. Without addressing the underlying problem of poverty, which itself is not colorblind, you will never be able to solve the problem of under-achieving schools.

        Thumb up 4 Thumb down 4
      • Vince says:

        Low-income children who are not black or Latino are more likely to attend better schools and much of this is caused by discrimination in the housing and housing finance markets. I’m not going to go into this in length, but if you’re interested, look into “geographic and racial steering”.

        Thumb up 5 Thumb down 3
    • Rob says:

      I clicked “Like” on this, not because I agree with Malcom’s stance on affirmative action, but because it is a common and reasonable take on the issue that deserves to be discussed. This comment shouldn’t be hidden.

      Thumb up 5 Thumb down 2
      • JAM says:

        In response to Rob

        I agree with your take and did not “dislike” Malcolm’s comment. I merely commented on it to drive home the fact that racial distinction should be removed from these types of programs. I did not feel that he went that far in his critique.

        Groups that are truly disadvantaged will benefit disproportionately from colorblind programs without having to divide us by race.

        Thumb up 4 Thumb down 1
      • Rob says:

        JAM,

        I hope you don’t think my comment was directed at you, or anyone in particular. I thought that Malcolm’s comment was worthy of discussion and had been unfairly hidden simply because people disagreed with it, not because it was inappropriate or unworthy of notice. It seemed wrong, so I said something.

        Thumb up 1 Thumb down 1
      • Malcolm says:

        Thanks, I know this is a controversial issue and I appreciate people to disagree and debate rather than disregard. I also understand how unappetizing the idea of any discrimination, even “reverse-discrimination” feels.

        Thumb up 1 Thumb down 1
    • mannyv says:

      If the goal is to help the disadvantaged poor, why not just give them an extra $20k a year in cash? Direct payments may be cheaper than trying to shoehorn people into a system that they can’t compete in.

      In any case, starting Affirmative Action in college is too late. Interventions work best when you’re young, not when you’re that old.

      Thumb up 3 Thumb down 5
      • Travis says:

        I can only assume you got the 3 downvotes because of your (tongue in cheek?) comment about just giving poor people an extra $20k a year.

        But the last sentence of your statement is 100% true. Sending kids to college who didn’t have good teachers, didn’t pick up good study habits, and generally didn’t learn enough in their primary and secondary education because inner city schools are horrible does little to solve “discrimination” or to improve the status of minorities.

        There’s an argument it could hurt it, because when given a “free ride” they are prone to failure and dropping out. They can also drag down a school’s ranking, which is particularly problematic in post graduate education – and is even worse for the image problem of both the ‘minority’ and the school.

        Thumb up 2 Thumb down 0
    • Mr. Econotarian says:

      The data in the paper suggests that post-Prop 209, 707 fewer minorities enrolled in UC, however 92 more minorities graduated from UC.

      One has to ask, is the world a worse place because 707 minority students decided not to enroll in UC (and could have enrolled in another school, we don’t really know), or a better place because 92 more minority students graduated from UC.

      Thumb up 3 Thumb down 0
      • Malcolm says:

        But check table 8. Only 20.5% of the increase (18 graduates) is attributed to “improved matching” ie the elimination of affirmative action. Student selectivity and university response are both greater factors. So the study is hardly damning evidence for the value of Affirmative Action. This looks like a well-done study into the side-effects of AA policies. The matching problem is a completely legitimate side-effect, but that isn’t to say it can’t be mitigated.

        For example, this could be addressed by having the less competitive UCs, those that are better at graduating lower performance students per the study, take on a greater share of AA minorities than the top-tier UCs. That is a policy change that would replicate student matching improvements.

        Thumb up 0 Thumb down 0
      • Malcolm says:

        As a former UC student I realized that an additional factor is overall UC expansion. During the period of this study white enrollment was up 7% while under-represened minority enrollment was down 5%. Had URM enrollment increased at the same rate as that of white students at the same graduation rate as before then we’d see an increase of 667 graduates instead of just 92.

        (The graduation rate over this period for whites was up 2.5%, “others/undefined” it was 1.5%, asians 3.7%)

        The lack of that control applied to the results of this study completely skews the result.

        It can be said that in the absence of AA poorer performing minority students were better matched resulting in higher graduation rates but reduced attendance. In effect the system was made more efficient, but more minority students were NOT graduated than would otherwise have been without AA.

        Thumb up 1 Thumb down 0
  4. Derek P says:

    What about the number of minorities enrolled in CA universities? Has it increased or decreased?
    Without this info, the analysis incomplete – not only is it incomplete, it has barely begun.

    Well-loved. Like or Dislike: Thumb up 14 Thumb down 0
    • Malcolm says:

      Minority enrollment in the UC system is down since Prop 209, significantly in some cases.

      See the difference in the graph in this article:
      http://californiawatch.org/dailyreport/despite-diversity-efforts-uc-minority-enrollment-down-prop-209-15031

      Thumb up 1 Thumb down 2
      • Joe Dokes says:

        The reality of Prop 209 is that while it decreased enrollment of Black and Hispanic students at the most competitive UC schools like UCLA, UC Berkely, and UCSD it lead to an overall increase of minorities system wide. Thus, enrollment increased at universities like UCR.

        Well-loved. Like or Dislike: Thumb up 9 Thumb down 0
      • Enter your name... says:

        It depends on how you define “minorities”. The proportion of white students has stayed about the same. The proportion of Asian students went up overall. The proportion of African-American, Latino, and Native American students went down overall. So the “number of minorities” is the same, and the “number of historically disadvantaged minorities” is down.

        Black students in particular have also self-selected into certain campuses, so there are individual schools with many more, and schools with many fewer, Black students.

        The result is that you can write a paper that “proves” whatever you want, by careful choice of your terms.

        Thumb up 4 Thumb down 1
  5. Pat says:

    If affirmative action is to be utilized with college enrollment, perhaps a better option is to base it on whether or not an applicant’s parents graduated college. Students with college educated parents are more likely to attend college, so perhaps utilize AA for those whose parents didn’t.

    Likewise, if there is a group with a historical under-representation, this method naturally migrates to students in that group (and others) and also allows for a consistent focus as racial/ethnic demographics change and mix.

    Of course, look for an increase in students who claim their parent(s) didn’t graduate, for bureaucracies to struggle with defining “parents,” and for college educated parent(s) to feel their children are being penalized. These all point to common issues of AA, but I still think it would be a better method than simply basing it on race.

    Thumb up 3 Thumb down 0
    • Enter your name... says:

      Another approach is to make it based on income. If 10% of the high school graduates qualify for a full Pell grant, then schools should aim for 10% of their students qualifying for a full Pell grant. Ideally, you’d back that up not just with an offer of admission, but also with a scholarship that covers dorm costs or other non-tuition expenses.

      Thumb up 2 Thumb down 0
      • RGJ says:

        I hate when I mean to hit the reply button and hit the green thumb by mistake.

        Hey, here’s a wild idea. Let’s keep the experimental social out of our higher education, which should be a meritocracy the way it is in the countries that are kicking our butts in….everything.

        The white kid who loses a slot to someone based on his color has a lot bigger gripe than the black kid who didn’t make the grade. and don’t think this generation isn’t watching it and picking up the same prejudices their father’s had — that blacks are less than equal, and need the game slanted their way in order to win.

        We elected the junior senator from Illinois president, for god’s sake. Do you think that would have happened if he looked like Opie? Discrimination is dead, and the reverse kind is thriving.

        Thumb up 5 Thumb down 5
  6. RGJ says:

    Affirmative Action drives me nuts. It is something that is so prima facie wrong that it should bear no need for debate or discussion. It is discrimination on the basis of skin color (generally).

    What are we teaching our kids?

    Discrimination against something should be illegal. But discrimination FOR something under the law is EXACTLY THAT. WHAT IS THAT YOU SAY? NO….SHUT UP!….THINK IT OUT.

    When the Census came around I signed my kids in as 1/16th Carukite Indian. Let the scholarship money roll!

    Thumb up 5 Thumb down 2
  7. Sam says:

    Has anyone been able to document any net value to affirmative action? That is, does giving preference to certain racial groups produce more (or less) economic value than a more random selection — e.g. giving preference to those who last names begin with a vowel? If I gave preference to white men in the NBA, that might look good esthetically, but I’d wager it wouldn’t produce better basketball teams.

    Thumb up 3 Thumb down 0