<p>
[quote]
Sakky, the revealed preference survey is not a random sample. It works backwards. It looks at what it perceives to be the top colleges, finds out which high schools sends kids to these colleges and then sends out questionnaires to these high schools. Some schools respond, some don't.
The survey is also skewed to a wealthier student body than the student body of the US.
[/quote]
</p>
<p>Now, this, I agree, is a valid criticism of the paper. I agree (and the authors agree) that the paper would be better if they looked at a larger set of data.</p>
<p>But look, that criticism is true of practically every paper out there. Every paper could be improved with more data. Every model could be refined if the search space was larger. But the reality is, you almost never have all the data that you want. Every paper has to start with something. If you're constantly going to complain about incomplete data, then no papers would ever get written, because the data is almost always incomplete. </p>
<p>As for the criticisms that the data is skewed towards the wealthy, I believe the authors addressed this point. Wealthier students tend to be the status leaders in that other students tend to want to emulate the behavior of those wealthier students. That's why Harvard is such a preferred school - because people want to rub shoulders with wealth and privilege. </p>
<p>
[quote]
And the logic that if a student chooses A over B and another student chooses B over C; therefore, students would choose A over C is wrong (page 10).
[/quote]
</p>
<p>First off, I think you meant p. 12, at least in the latest version of the paper. </p>
<p><a href="http://www.economics.harvard.edu/faculty/hoxby/papers/revealedprefranking.pdf%5B/url%5D">http://www.economics.harvard.edu/faculty/hoxby/papers/revealedprefranking.pdf</a></p>
<p>The notion of transitivity lies at the very heart of the principle of revealed preferences - to question it is to question the very foundations of revealed preferences, and by extension, the entire body of literature that is based on RP. NUMEROUS papers in social science (Econ, poli-sci, sociology, psychology) use RP models where the principle of transitivity is assumed. </p>
<p>So hey, if you want to question revealed preferences, that's fine. You are free to do that. Just understand that by doing so, you are questioning one of the very basic and accepted cornerstones of modern social science. </p>
<p>
[quote]
Plus students make there choice about which schools they prefer when they apply to school, not just after they get their acceptances. The revealed preference survey tries to measure the latter, but does not measure the former where many of the decisions are made.
[/quote]
</p>
<p>Now come on, Dstark. You just made the same mistake that the others made. It DOES INDEED measure the former, and in fact, is precisely the value add of the paper. If it only measured the latter, it would be a very boring paper, because, as I said before, it would only be reporting on cross-admit data. For example, students get accepted to schools A and B and x% of them choose A. That's just journalistic reporting of the data. </p>
<p>What the paper adds is a model that fills in the gaps by attempting to ascertain what people WOULD have done had they been given the choice to also go to C (whether they applied to C or not, or whether they got into C or not). That's what much of the latter part of the model is all about. </p>
<p>Let me give you an example. I think it would surprise nobody to discover that Harvard is a more preferred school than, say, Michigan State This is true even though obviously very very few people are going to apply to both Harvard and to Michigan State. And the few that do are probably unusual For example, maybe there are some guys who are good enough to get into Harvard, but are also good enough athletes that they think they have a shot at the NFL or the NBA, so they might consider going to Michigan State to play ball so that they can take their best shot at the pros. Or maybe they just happen to live in East Lansing and are highly family oriented such that they don't really want to move, even if it's to go to Harvard. But the point is, I think we can all agree that those few people who apply to both schools are certainly not representative of most other students. Very few other people are going to seriously consider both schools enough to apply to both of them. </p>
<p>Michigan State might actually win the cross-admit battle with Harvard for the small and unusual body of cross-admits that exists (again, football players might prefer Michigan State). Or even if MSU doesn't win, maybe the cross-admit battle is not heavily tilted towards Harvard. But does that translate to the rest of the population? Of course not - because you said it yourself, most people who are Harvard-caliber won't even apply to Michigan State, and most people, if they were to be given the choice, would clearly choose Harvard. Hence, on an overall basis, there is no comparison between the two schools in terms of RP. </p>
<p>
[quote]
Do you really think more smart students would rather go to Williams than NYU or UCLA?
[/quote]
</p>
<p>Is that really so hard to believe? Williams is an elite LAC. I don't think it's so hard to believe that a lot of top students would prefer elite LAC's over large schools (and both NYU and UCLA are very large). As has been said before on other threads, the elite LAC's have excellent reputations for strong teaching and strong graduate school placement, and a lot of top students want that. </p>
<p>
[quote]
I disagree. Employers want quality employees above non-quality employees and they consistently judge whether they can find those at MIT or at some alternative
[/quote]
</p>
<p>But right there is the contradiction in your logic. To reinterpret what you just said, employers don't just want quality employees. They want quality employees * that they can actually retain *. </p>
<p>And that's entirely consonant with what I said above. Sure, at MIT, you can find plenty of high quality engineers. But what if you can't actually get any of them to accept one of your offers? Or even if you do, what if he quits in a short period of time because he is frustrated by the perceived lack of opportunities at your company and thinks he can do better elsewhere (i.e. perhaps because he sees all his old MIT classmates doing well and getting great career opportunities, and then he starts asking himself why isn't he also doing as well as them)? In such a case, the employer might as well not even bother to recruit at MIT. They'd be better off recruiting at a lesser school where more students are more willing to take their jobs and keep them. </p>
<p>
[quote]
The employers set the market price for wages and adjust accordingly to fill their needs. If the MIT grad is too expensive, then they move on to an alternative. What this exposes is that the employers don’t see a huge difference in the quality of SOME students coming out of an alternative, lower ranked, school and what is coming out of MIT. They don’t need to hire the whole school, they need to hire one or more graduates, but the reality is that employers know that student quality is widespread (if erratically so) and that much of the education anyway is going to come on the job, not from what the student has learned through their undergraduate years.
[/quote]
</p>
<p>Ah, but this is not quite a fair comparison. You've admitted it to be the case. I agree that SOME of the students at a lower-quality school are just as good as those coming out of MIT. But that's the key word - "some". The problem is, how do you know who they are? Economically speaking, you have to pay a search cost to ferret out those students at the lower-ranked school who are high quality from those that aren't. </p>
<p>MIT, on the other hand, provides you with a presorted selection set. Their admissions are tough, and the curricula is tough, such that those who make it through are, on average, of very high quality. That means that you can afford to pay lower search costs. You don't have to engage in such an intensive and expensive search because much of the sorting has been done for you. </p>
<p>To use more econ-speak, the major problem with any labor market is adverse selection. If you choose to pay X for a particular job, you are always going to tend to attract those people for whom X is the best they can do (because other employers will pay them less than X). On the other hand, those people who have choices that are better than X are going to turn your job down (or quit the job if they mistakenly choose it). Note, X doesn't have to be just money, it can be a combination of things (benefits, career opportunities, etc. - hence, X can be thought of as "utility"). </p>
<p>
[quote]
For the large majority of students picking colleges, finding a job after college is a critical factor. But the reality is that most job markets are local and the value of the overpriced brands can get marked down.
[/quote]
</p>
<p>They "can" be marked down, but they usually aren't. Let's face it. MIT grads (and grads of HYPS) aren't exactly hurting for jobs. MIT, in particular, is swamped with consulting and banking firms. I don't think that MIT students are unhappy with their choices of jobs. Far from it, in fact. It's become something of a running joke at MIT that many of the top MIT engineering students will never work as engineers, instead opting to become consultants or bankers instead. But that does seem to be what they want. Hence, I hardly see any evidence of the MIT 'brand' being "marked down". Rather, it's that the engineering firms who want to hire MIT students now have to compete against the world's top consulting and banking firms, and many of those engineering firms just don't want to do that. Either that, or they have to compete against Google, which is the one engineering firm that does seem to have sucked much of the oxygen out of the room at MIT, and again, most engineering firms just don't want to compete against Google for talent. I can immediately think of one large software firm that shall remain unnamed that was quite ticked off by the MIT hiring process, because every one of their offers got turned down because the students would rather do other things (many of them turned down the company to take offers at Google, some turned it down to go to grad school, and the rest turned it down to become consultants/bankers). I have a feeling that they won't come back next year to recruit, for fear of the same thing happening. </p>
<p>
[quote]
As for how this compares to the current PA system, my suggestion is that employers, as a stakeholder in the process, would be a valued addition to the evaluation of colleges. Academics would hate it, but I’m guessing that students and families would love it
[/quote]
</p>
<p>Again, no, I have to disagree, because, again, employers are not an unbiased stakeholder. Far from it, in fact. Employers have their own agenda, as has been stated above.</p>
<p>Let me give you another example. Let's use Harvard. Harvard is quite infamous for ticking off many recruiters. This is especially true of HBS, but is also endemic in many of Harvard's other programs. HBS recruiters in particular often times tend to give HBS extremely low marks. And the complaints tend to be the same - that the students are arrogant, that they demand too high pay, that they demand too much power, that they're difficult to negotiate with, etc. etc. But here's the rub - * they keep coming back anyway *. No matter how annoyed the recruiters are, the top recruiters keep coming back to HBS. If they really didn't like it, they wouldn't bother to show up. And many of the less desirable recruiters indeed don't show up, because they found it so difficult to recruit people in the past. But the top recruiters keep coming back, no matter how annoyed they are. </p>
<p>The point is this. I don't think employers should be "satisfied" with a particular school. In fact, I would argue that they should actually be DISsatisfied, as long as they keep coming back anyway. After all, think of it this way. A truly satisfied employer is one that is getting all the people they want at low cost. In other words, a school can greatly increase employer satisfaction by just convincing all the students to just take lower salaries. That would certainly increase employer satisfaction. But would it really increase student satisfaction?</p>
<p>The point is, by bringing in employers as a stakeholder, you are bringing in a party that doesn't really have the students best interests at heart. Like I said, we all know that employers would love to get away with paying slave wages to everybody if they could. No employer WANTS to pay high wages to anybody. They do so only because market forces dictate that they do so - i.e. that if they don't, people will choose to work elsewhere. </p>
<p>Hence, instead of using employer satisfaction, I would key solely on STUDENT satisfaction. After all, at the end of day, what ultimately matters is how satisfied the STUDENTS are. They are the ones paying, after all. They are the ones who actually attend the school. Hence, they are the ones who should be satisfied. If schools can figure out a way to increase student satisfaction, even at the expense of losing employer satisfaction, I would say that's a good trade. Again, at MIT, at Harvard, sure, a lot of employers are ticked off. But who cares? Like I said, you SHOULD be ticking off the employers, because if they're not ticked off, that just means that aren't negotiating hard with them to extract as much money as you possibly can. The STUDENTS at MIT and Harvard are satisfied with their career choices, and that's all that really matters. You want your students to get what they are worth, and if that means ticking off employers by forcing them to pay higher salaries and perks than they want to, that's too bad for them. You want your students to have negotiating power, not the employers.</p>