The top 7% rule is not going anywhere anytime soon according to people I know. Of course, I do not like it because of the fact that I think lots of kids who are in more competitive schools who are well-rounded do not have a chance. In my son’s school, it is almost impossible to be on the football team and take all AP classes and be in the top 10% for example. I see so many great kids who have good ACT scores going out of state. But, on the other hand, it has really helped kids in the Valley and small town kids in West Texas.
@coase - I spent about 20 minutes on the train skimming the two papers you reference. I should caveat my opinion by making clear that I do not know the relevant literature at all, nor have I had the benefit of an expert guiding me through the work that’s been done. Also, I only quickly skimmed the 100 pages.
Nevertheless, I agree with you that Arcidiacono et al.'s papers seem to be careful and good quality analyses and should be taken seriously and not dismissed.
However, I would be cautious about drawing broad conclusions from the results of their study. They find evidence for the “mismatch hypothesis” in the UC schools for the case of the graduation rates of prospective science majors whose academics fall in the lower tier. To be honest, I’d say their estimates indicate that the net effect isn’t that large. Also, they don’t find evidence for the mismatch hypothesis in other cases that they examined. (The third study you mention, about convergence in GPAs, doesn’t seem to directly address these mismatch issues.)
One more general comment - these types of counterfactual empirical analyses can be very tricky. It’s almost never the case that one study is the definitive word on the subject. There are almost never any good natural experiments and important data that we would like to have isn’t available. Researchers have to rely on creative assumptions or invent imperfect proxies to get around these problems. We need to see multiple studies, critiquing and improving upon prior work and examining different data sets with different techniques, before we know what the true picture is. And my impression is still that - because of the controversy that surrounds this subject - a robust literature simply isn’t developing.
How many students change majors at UT? If a student does change from one major to another, how do you determine the cause was “mismatch” vs. student growth? I don’t think most high school students know enough about the world to determine the best major for them in the next century. I don’t think most adults know enough about the world to give them good advice.
If any of the Supreme Court justices were relatives, I wouldn’t be turning to them for career advice for my children, except if my children wanted to become judges.
If a student changes his major from petrochemical engineering to elementary education, is that a bad thing? Vice versa? (As the price of a barrel of oil trades at a 7-year low, http://www.bloomberg.com/news/articles/2015-12-14/never-mind-35-the-world-s-cheapest-oil-is-already-close-to-20).
If the average student changes majors 3 times in four years, and only 20% remain in their original major, how in the world does one detect the effect of “mismatch?” By the way, unlike many people, I consider a career in STEM fields not to make the most financial sense for most people. It’s fine if you have a family willing to support years of postgrad study. It’s not fine if you are your family’s best chance at financial support.
The reason this is so hard is that Grutter contained inherent logical problems that are now coming back to haunt the court. (Note: this is my opinion, and I’m a legal realist, which means I focus on what the court was trying to do, and not so much on its justifications.) What O’Connor did in Grutter was say that affirmative action could continue (at least for 25 years) as long as the decisions were made in a black box labeled “Holistic” and “Diversity.” The idea of diversity as the sole recognizable state interest is problematic though, because of the question, “How much diversity?” When the schools get deference for their decisions, that question doesn’t come up, and there is less reason to try to explain what “critical mass” is and how it’s different from a quota. All these problems come back, though, if true strict scrutiny is required. “I know critical mass when I see it” is probably not going to cut it.
In my view, this happened because O’Connor’s Grutter decision was a compromise. The real reason for affirmative action (again, in my opinion) is that a lot of people (including me) believe that it will be to the long-term benefit of society to perform social engineering that results in more URMs attending college and entering the work force at higher levels. I assume this was not an acceptable state interest for a majority of the Supreme Court, which is why diversity was chosen as its stand-in.
One comment on mismatch theory: surely we all believe in it, at the extreme–such as, for example, when applied to recruited athletes who cannot read and who don’t graduate? It seems to me that the only question is how wide the mismatch has to be to make it likely to harm the student. That becomes difficult, because we have to consider things like whether being a sociologist from Harvard is better than being an engineer from a state university.
As for my prediction, I think the court will find a way to punt.
All good. And I’d add, Hunt, as one generation attends college, they are able to positively influence the next. Their influence can run both deep and broad. We see this now, in mentoring programs.
If I were advising a college, I’d tell them to announce that “race” will no longer be considered as an element of holistic admissions–but that “culture” would be considered.
@al2simon, my impression, by skimming some of the cites mentioned in the various Supreme Court briefs, and also by Googling around, is that there seems to be a good amount of literature from authors with plausible affiliations. Aricidiano is at Duke, and other authors are from Brown, Princeton, Stanford and other highly ranked colleges. I’m surprised that you say there is a paucity of literature.
Aricidiano’s 2015 paper is odd to me. He concocts an elaborate theory as to why black students would choose a higher-ranked institution than a lower-ranked one, even though they’d have a better chance of success at a lower ranked one: because cynical colleges and universities are hiding the truth from them. If only the colleges told the truth, black students would turn down their acceptances to Berkeley and go to UC Santa Cruz instead, we are supposed to conclude. He writes
Does this guy teach undergraduates? Does he have any kids? Has he ever met another person who is not an economist? He may be right about the mismatch theory, but explaining why young people overestimate their ability to succeed does not require a conspiracy theory about colleges hiding information. Anyone can easily discover the number of college football players and the number of NFL rookies, yet students still act on rosy beliefs about their chance of becoming professional football players. It’s no secret that physics is hard, yet students still think they can pass their physics courses by studying an hour a week. Sports fans think they can make money on FanDuel and DraftKings, for heavens sake. People just don’t believe averages apply to themselves.
This made me have the following thought: some state universities have the reputation for admitting students with a wide range of stats, and then culling out those who can’t perform (“look to the right of you…”) More selective schools supposedly don’t do that–they only admit students that they believe can succeed. But what if some selective schools in fact, do that, but only for a subset of admits?
Elaborate, please, @Hunt
For the most part, this is based on where the school falls in the selectivity scale. More selective schools will enroll students who mostly can handle college work, while less selective schools will have more students who have trouble handling college work.
Yes, there may be deliberate policy decisions in public schools (particularly open admission community colleges) to give everyone a chance to go to college, with the recognition that not everyone will succeed. I.e. the idea is that graduating high school with a 1.5 GPA, or doing a GED instead, or not even graduating or doing a GED, should not close the door forever on going to college, but such a student better shape up on academic commitment to succeed if s/he enrolls in college.
@al2simon, when I Googled “melguizo mismatch”, I found a link to the text of a book called *Affirmative Action and Racial Equity: Considering the Fisher Case to Forge the Path Ahead *. Google supplied me with a lot of pages of it, including a discussion of the literature about the mismatch theory. I call your attention to the length of that discussion-- many pages. Whether you agree with the authors’ conclusions or not, they bring forward a lot of research bearing on the issue. I don’t think you’re right about the paucity of the research in that area.
Culture isn’t a protected class. But I can achieve whatever kind of diversity I want by considering culture. That would be my new black box.
The mismatch theory folks are making specific claims, namely that black students who are the beneficiaries of affirmative action at more selective institutions would be better off enrolling at less selective institutions. I’m interesting in knowing whether that specific claim is true, or partly true in the case of some students (e.g. students who hope to get science degrees), or in general not true at all. Nobody disputes that an illiterate is going to have trouble graduating from Harvard.
@“Cardinal Fang” - I definitely understand your reaction, but this is an example of how professionals would read this literature through very different eyes. Economists have a very strong aversion to using “people are stupid and/or irrational” as an explanation. This aversion permeates the entire discipline. To them, it would be like a biologist, who when asked to explain why a given species of bird evolved to have a certain characteristic, explains it by saying “because it was God’s will”. It’s just not considered valid science even though it’s a great “explanation” for lots and lots of observations.
I think it’s fair to say that economists have a strong presumption that seemingly irrational behavior generally has some rational basis (for example, it might be a response to a perverse incentive structure), and if you don’t understand what’s driving people to do something it’s because you don’t really understand what’s going on. Relying on the “people are stupid” explanation is their last resort. I actually think economists are often right in this belief, but it would be too long a digression to give a full explanation or to talk about behavioral economics.
Aricidiano et al.'s “private information” explanation is completely within the mainstream of the discipline. It actually encompasses a wider variety of “stories” that would appeal to a lay audience when properly packaged, but Aricidiano doesn’t bother writing out these soothing explanations in an academic paper. It’s unfair to accuse them of “concocting” an elaborate theory. Of course, that still doesn’t mean he’s right.
P.S. Based on my experience, I’ll bet that you still don’t buy this, but that’s why people spend years studying this sort of stuff 
P.P.S. You may be right about the literature being more robust than I’m giving it credit for. I’m the first person to admit my ignorance of the literature. Really, we need to get an expert’s view.
@tiger1307 @GMTplus7 Very few people attend posh boarding schools. My children attend a public school that is 40 percent black. Many people want the government and institutions to operate on the assumption that all whites are wealthy and all blacks are poor and deprived. Well, it ain’t so. Should you need it, I will be glad to draw you a map so you can make a White Poverty Tour of the United States. Very few programs are designed to address economic inequality; it’s all about race and racial preferences, and poor white people are the ones who are harmed the most by racial discrimination.
I was not aware of the claim that black law students had grades that were identical to those of whites after the first year. That certainly wasn’t the case where I attended. William Raspberry wrote a column once suggesting that law schools stop using a blind grading system so professors could just give blacks passing grades, specifically because too many blacks were flunking out of law school. The problem shows up on the bar exam, where blacks fail their first attempt at a rate two to five times greater than whites. You can Google this.
“Mismatch” tells the story of George Mason Law School, which did not employ racial preferences. Although blacks were underrepresented at the school, those who did attend had the same bar passage rate as the white students. The ABA threatened to withhold the schol’s accreditation unless it began using preferences, and the school’s black bar passage rate immediately dropped. What’s interesting is that blacks with similar grades and LSAT scores who attended “lesser” law schools had a higher bar passage rate than those who attended Mason by virtue of preferences. It seems that students do better when they attend a “slower track” school where the instruction is designed to help a person of their ability pass the bar.
However, irrational behavior is a topic of research in behavioral economics, in terms of understanding how and why people act seemingly irrationally (but often predictably).
I skimmed the science major graduation by UC campus paper, and it does not seem that the actual data support the hypothesis. In Table 2 it seems to report that 26.9% of initial science majors at UCLA and UC Berkeley graduate with science majors but only 19.1% do at UC Riverside.
In fact he says:
And then goes on to propose reasons why. Thus, this paper does not at all support the mismatch theory.
@mom2and: That is not my reading of the table. I believe the number you quote is the percentage of matriculates who ultimately graduate in the sciences (“second row is percentage of enrollees who started in a particular initial major”).
@al2simon: I agree with these points. It is an empirical question how much mismatch matters and whether it trumps the potential benefits of attending a better school. But it is a question worth studying. Alas, Arcidiacono got so much grief (even from Duke’s president) when he wrote the paper on GPA differences that few would dare to tread these waters.
By the way, the latter paper may well be relevant to the mismatch debate. You could imagine two groups that have the identical distributions of planned majors, identical graduation rates and identical distributions of GPAs, yet members of one group are much more likely to switch majors. To the extent that the switchers disproportionately switch to an easier major, that is relevant. Now, if a student enters Duke as a math major and ends up a sociology major, is that worse than being a math major at a lower-ranked school? Perhaps not, but we should at least be clear about what is happening when students matriculate. I served as the chair of our department’s graduate admissions committee for two years when I taught at a large university in the Midwest. I know in my heart that we admitted a couple of students who had no chance of finishing the Ph.D. Sure enough, both dropped out. I am the last person to take away someone’s agency, but it is not clear to me that we did them a favor by giving them a chance.
Anyway, “culture” would be a better word and from my perspective, private college, it is one of the considerations for UG. With respect to this convo, there’s a deep commitment to academic and other support, then finding kids able and willing to use that, should they need it. The goal is a meaningful graduation. There’s an understanding any kid can change majors (when he hits the grand buffet of choices of major.) In the way a class is put together, there’s accounting for that. In this respect, different from grad school.