Information Processing

Just another weblog

Archive for October 2004

Revealed Preferences and College Rankings

with 2 comments

I just read a fascinating “revealed preferences” paper on college rankings, in which the rankings are based on decisions of students admitted to multiple schools (e.g., Harvard beats Brown if an applicant is admitted to both but chooses to enroll at Harvard). They consider each decision as a “match” (as in chess) between pairs of schools, and use the mathematically sophisticated Elo system (invented by a physicist and used in chess) to determine a ranking. These revealed preference rankings are at odds with the U.S. News results in interesting ways.

Harvard is number one by a big margin – no surprise given its 70% admissions yield. The authors list Yale, Stanford and Caltech as the rest of the top four, although if you look carefully at their regional data it seems clear that Caltech is number two (their admissions yield is around 50% and no other schools do much better than 30%).

Another interesting piece of data is that at certain schools like Princeton the probability of admission is not monotonic with SAT score. Your probability of admission is higher at around 91 percentile than at 95 percentile, although then it goes up again. The authors claim this is evidence of Princeton “managing” its admissions to improve yield – they don’t want to fight over the candidates who will be admitted to multiple top schools, except the very exceptional ones. I’m not fully convinced, but the oddly shaped probability curve is very striking.


Written by infoproc

October 25, 2004 at 6:19 am

Posted in caltech, universities

Dumbass Voters – or is it just Tennessee?

leave a comment »

As discussed by social scientists here.

Are the voters stupid? It is not considered politically correct to point out that an awful lot of voters don’t have a clue what they are talking about. A recent poll from Middle Tennessee State University sheds some light on the subject. For example, when asked which candidate wants to roll back the tax cuts for people making over $200,000 a year, a quarter thought it was Bush and a quarter didn’t know. And it goes down hill from there. When asked which candidate supports specific positions on various issues, the results were no better than chance. While this poll was in Tennessee, I strongly suspect a similar poll in other states would get similar results. I find it dismaying that many people will vote for Bush because they want to tax the rich (which he opposes) or vote for Kerry because they want school vouchers for religious schools (which he opposes).

Written by infoproc

October 19, 2004 at 5:54 pm

Posted in Uncategorized

Faith-based Presidency

leave a comment »

Bruce Bartlett, a domestic policy adviser to Ronald Reagan and a treasury official for the first President Bush, told me recently that ”if Bush wins, there will be a civil war in the Republican Party starting on Nov. 3.” The nature of that conflict, as Bartlett sees it? Essentially, the same as the one raging across much of the world: a battle between modernists and fundamentalists, pragmatists and true believers, reason and religion.

”Just in the past few months,” Bartlett said, ”I think a light has gone off for people who’ve spent time up close to Bush: that this instinct he’s always talking about is this sort of weird, Messianic idea of what he thinks God has told him to do.” Bartlett, a 53-year-old columnist and self-described libertarian Republican who has lately been a champion for traditional Republicans concerned about Bush’s governance, went on to say: ”This is why George W. Bush is so clear-eyed about Al Qaeda and the Islamic fundamentalist enemy. He believes you have to kill them all. They can’t be persuaded, that they’re extremists, driven by a dark vision. He understands them, because he’s just like them. . . .

”This is why he dispenses with people who confront him with inconvenient facts,” Bartlett went on to say. ”He truly believes he’s on a mission from God. Absolute faith like that overwhelms a need for analysis. The whole thing about faith is to believe things for which there is no empirical evidence.” Bartlett paused, then said, ”But you can’t run the world on faith.”

NYT Magazine

Written by infoproc

October 16, 2004 at 5:49 pm

Posted in Uncategorized

FANNIE and OFHEO House testimony

leave a comment »

Well, after watching Falcon and Raines on CSPAN yesterday for several hours, I would say I am not convinced that OFHEO really has anything on them. They might, but it wasn’t clear from the testimony, and I would not be surprised if OFHEO was under pressure to deliver a damning report and went about it in an amateurish way.

The $200M deferral looks suspicious. The question of whether FNM’s hedge accounting for their derivatives was complaint was answered in contradictory fashion by OFHEO and FNM. OFHEO was definitely wrong at the beginning when they answered that the value of FNM’s retained mortgage portfolio could easily be described by a single number instead of a range. The inputs into any valuation model are uncertain, so a range is necessary and it is this range that can be manipulated for cookie jar purposes.

I must say Raines and Howard are more impressive figures than Falcon and co. (who had a really hard time explaining their report to the subcommittee). But on the other hand Andrew Fastow of Enron was also a smooth and impressive-sounding guy.

In any case, the public focus on the huge compensation of senior FNM execs (mostly from stock grants, I believe, not bonuses) will be a black eye.

I suppose we’ll have to leave it to the SEC and the fourth estate to sort this all out 🙂

Written by infoproc

October 7, 2004 at 6:12 pm

Posted in Uncategorized