How UT manages its relationship with influential people has been a very big controversy the last few years. A whole other thread would need to be devoted to the Kroll report, the Wallace Hall lawsuit, Wallace Hall censure, the departure of Bill Powers and the eventual change in the President’s ability to influence the admissions office.
@al2simon. That is yet another thoughtful post. I agree with this interpretation. One criticism of the mismatch theory is that rational students know what is best for them so there cannot be a problem. If they matriculate at a school, doing so must be (ex ante) optimal. A behavioral economist might say that the students are not rational, so there may yet be a problem. Arcidiacono et al. take a different (and harder) tack, and argue that there may be a problem even if the students are rational. A necessary condition is that the university has superior information. This is not spell-binding stuff, but it is a point worth making. This paper is certainly the least important of the three that I cited.
Again, I agree with the general thrust. The anecdote I gave previously is in line with this interpretation. It is easy to absolve oneself of all responsibility, say “I did my part” and walk away self-satisfied. If we are going to admit someone, we should be ready to give him or her the opportunity to succeed. Texas has one very promising program itself. http://www.nytimes.com/2014/05/18/magazine/who-gets-to-graduate.html?_r=0
I want to conclude with a general point about the whole peer-review process. The Arcidiacono et al. paper concerning student outcomes in CA was presented by one of the authors in faculty seminars at Berkeley, Duke, Columbia, Cornell, NYU, Oxford, Penn, USC and WashU, as well as at Brookings, the National Bureau of Economic Research, and the Institute for Research on Poverty. It was accepted for publication by the American Economic Review, which accepts fewer than 10% of manuscripts. The typical review process would involve an editor and three or four (anonymous) referees going through two or three iterations. That is no guarantee of a pristine final product, but one should expect the published version to be carefully written, especially given the contentious nature of the topic.
I think this is unknowable, except at the extremes. What the students and/or the schools may know, though, is that the decision to enroll in the selective school is more risky for some students than others. This is actually something that gets discussed here on CC all the time.
A sports analogy: If I go to XYZ College, I’ll probably start as a freshman. If I go to ABC, with a much stronger program, I have a much higher risk of sitting on the bench. Where should I go? It’s hardly a no-brainer–there are many considerations. In my opinion, what’s crucial is the degree of apparent mismatch.
How do you read that table @coase? and the comment he makes which states that the actual results do not match his model? His thesis was that if URM students went to a “lesser” UC they would be MORE likely to graduate with a science major and yet they aren’t. Where does he show data that actually proves his theory in that paper?
Aren’t you then sacrificing the students who ‘don’t’ succeed – but may have a a lesser school – for the ‘benefit’ of the rest of the students?
btw: don’t forget, that Sander’s thesis is all about LS admissions, where GPA+LSAT trump all at every school not named Stanford and Yale. So even a bad Frosh year and all A’s later, hurts significantly in admissions chances and in merit money. (Law School is the only grad/prof school that I know of that is so grade-grubbing!)
Doesn’t medical school also highly encourage grade-grubbing to at least be in the running, though once the pre-med gets a high enough GPA and MCAT, there are numerous other (often subjective) admission criteria that s/he needs to fulfill? Though perhaps the pre-med grade-grubbing is more obvious among undergraduates, since pre-meds tend to be concentrated in those courses required by medical school (biology, general and organic chemistry, etc.).
Yes, that’s how the case would go.
I’m sketching out the argument, not committing myself to it. IF mismatch is true, which I regard as unproven, then we could nevertheless make the case that the benefit of having more Hispanics and blacks get degrees from UT exceeds the cost. In Texas, Hispanics and blacks are the majority of the college age population. If a UT degree has value in Texas over and above a TAMU degree, then (this argument goes) it behooves us to let blacks and Hispanics in on this power.
Remember, nobody is forcing anyone to go to UT if they don’t want to. The black and Hispanic admitted students go there voluntarily.
I wasn’t saying that Aricidiano 2015 was wrong. I just don’t think it contributes much to our understanding.
Let’s say you are admitted to UT. You find stats that say people with your academic credentials have a 69% chance of graduating, as opposed to an average of 84% for all students. What do you do with this information? It seems to me that it depends on how risk-averse you are, assuming you see a UT degree as a strong benefit.
I doubt if very many people would change their mind about UT on knowing people with their credentials have a lowered chance of graduation, even if they were risk-averse, because they would rationalize that the statistics didn’t apply to them. That’s what people do: they think up all the reasons why they are more likely to graduate than the average person with their credentials, and ignore all the reasons why they are even less likely to graduate than the average person with their credentials.
I think it’s clear many would like predictability, in some absolute yea/nay, sense: the kid belongs here, will do well, (even to the point of predicting gpa,) will graduate on time, with the exact major he set at 17, (before he even experienced more than high school courses,) and go on to be successful in life, based on that snapshot at 17.
I suspect that, if we look at our own stories or our own kids, we wouldn’t want this same narrowness applied to us or to them. We do know life is not one straight path, that what we wanted at 17 was just a beginning.
Of course, you want to know if there’s a risk of not graduating. But you also need to look at the attributes that can put one into the (hypothetical, above) 69%, not the 31% who fizzle out. And that’s much more than the numbers you plug into some predictive program.
Parents, however, may want to keep in mind that the risk of needing an extra semester or year or few (possibly after financial aid and scholarships run out) may affect budget plans. Parents may feel comfortable making an 8 semester budget for their 4.0/2300/35 student, but may want to consider that their 3.0/1700/25 student is more likely to need more than 8 semesters to graduate.
At the University of Texas, it’s not just 3.0/1700 students who don’t graduate in four years. Almost half of the students don’t graduate in 8 semesters. UT only has a 52% four year graduation rate.
It really depends. How do you predict a kid will need extra time at one college versus another?
And I don’t think we “know” that the grad issue at UT isn’t coming from the 10% Law.
Wouldn’t you look at whether kids like him tend to need more time at those colleges? I’m not saying this is easy to do, but it’s pretty much the same sort of analysis as the athlete deciding whether he’s likely to start at one school as opposed to another.
Here’s one issue, I think: we tend to tell kids here, at least with respect to highly selective schools: “Don’t worry–they wouldn’t admit you if they didn’t think you can succeed there.” But what if this isn’t entirely true for some groups of kids? (It’s certainly not true for recruited athletes at some schools.) What if the truth is that the school thinks that most of the kids in that group can succeed, but not as many as the average of all students? I think this is something that students should think about if their stats are below average.
I also think it matters how much below average a person’s stats are. I think at highly selective colleges, there isn’t really all that much difference in academic ability in the top three or four quintiles–after all, the fourth quintile may be populated by kids with three Bs in high school. But how long is the “tail” of the fifth quintile? If you’re at the end of that tail, you might have a very difficult time.
I think I misunderstood some of your previous comments. I thought you were commenting on Aricidiano 2011 (Does affirmative action lead to mismatch? A new test and evidence). The 2011 article is the one I was commenting on above.
Aricidiano 2015 (Affirmative Action and Quality-Fit Tradeoff) is not original research. It’s a survey of a branch of the mismatch literature; mainly on how affirmative action affects the education market and labor market results in the undergraduate and law school settings. It appears to be reasonably thorough and objective. I would recommend this article as a guide to this swath of the economics literature and to the results that have been obtained.
Aricidiano 2011 (Does affirmative action lead to mismatch? A new test and evidence) is the research article that explores the existence and role of schools’ “private information” in ex ante mismatch. It didn’t interest me very much, nor did I find the evidence they presented very compelling (but that’s just my opinion).
@Hunt - I wasn’t using the word “know” in the sense of an omniscient knowledge. I meant “know” in a statistical sense as in “Tobacco companies ‘knew’ ahead of time that cigarettes kill people”. Of course, it’s empirically false that the vast majority of smokers die of smoking related diseases, so the fatality of smoking is “unknowable, except at extremes”. But it’s empirically true that, on average, smoking a fair amount kills people, and harms them more than the pleasure any rational person should derive from smoking.
It seems to me that this is more of a legal question, so it’s more up your alley. Here’s how I might pose the questions -
If empirical studies show that admissions preferences as practiced by UT actually cause net harm, on average, to a segment of their intended beneficiaries, then is this enough for the courts to conclude that these particular preferences fail “strict scrutiny” and/or the harm caused overrides any diversity benefit? Will any of the justices latch onto this? Maybe it means that the preferences can be so many SAT point equivalents, but no more than this because then they start causing net harm to students?
Personally, the mismatch research I’ve read in the last 2 days doesn’t compel me greatly, with the exception of the “STEM mismatch” effect. Even where there is evidence for mismatch, there are many value-laden societal tradeoffs to make. In my very uninformed opinion, we shouldn’t base any policy decision primarily on this research.
Not to turn this into an Affirmative Action thread, but this was a good read on the racism on campus issue.
I think the challenge would be in defining what goes into “net” harm or benefit. Let’s assume, just for the sake of argument, that affirmative action creates a class of students at UT who will graduate at a lower rate than the UT average AND that that class would graduate at a higher rate at a “lesser” school. For the excess students who don’t graduate, that’s harm–but you’d have to balance against that the benefit to those members of the class who get a UT degree instead of a degree from the “lesser” school, as well as the benefit to all students of the added diversity. I don’t know how you calculate that, unless you’re just going to defer to the college’s balancing of the costs and benefits.
With a 52% graduation rate, it’s coming from everywhere. But I’d also guess that a student who is in the top 7% at a bad school (admitted to UT on the 7% rule) is more likely to struggle academically than a student who is at the bottom of the top 15%-20% at a good suburban school. And further, affirmative action admits at UT are more likely to be students of color who are good students but not top students at good suburban schools than students from bad rural or urban schools. That suggests that the students most likely to struggle are students admitted from the top 7% of bad schools.
@zinhead the author of your good read was the republican who ran in Oregon for the senate. He is a tea party favorite and he got all As from the NRA. He calls Hillary Clinton a liar in the daily caller. What is next for him? Talking about the lies regarding climate change? People like Jim Huffman are making racism worse in America not better