Friday, January 16, 2009

Brainpower and globalization

Below is a partial breakdown by nationality of applicants to our PhD program in physics. I chose 2007 because that is the last year I was on the admissions committee. Here is a post I wrote then, which characterizes the applicant pool: Graduate admissions, human capital and globalization. Our target for each year is about 15 incoming students.

At bottom are some figures I first blogged about in 2004 (if you know of more recent data, please tell me). They indicate that already by the late 1990s there were more Asian students staying at home to do their graduate work than coming to the US. Yet, as the 2007 numbers show, Asia is our largest source of applicants. Related posts.


china 158
usa 75
india 19
nepal 2
sri lanka 2
taiwan 3
korea 4
indian 3
new zealand 1
germany 1
colombia 1
egypt 2
bangladesh 1
japan 2
iraq 1
pakistani 1
eritrea 1

More up to date numbers here:
...Overall, the U.S. share of world S&E PhDs will fall to about 15% by 2010. Within the US, moreover, international students have come to earn an increasing proportion of S&E PhDs. In 1966, US-born males accounted for 71% of science and engineering PhDs awarded; 6% were awarded to US-born females; and 23% were awarded to the foreign-born. In 2000, 36% of S&E PhDs went to U.S.-born males, 25% to U.S.-born females and 39% to the foreign-born. 8 Looking among the S&E fields, in 2002, international students received 19.5% of all doctorates awarded in the social and behavioral sciences, 18.0% in the life sciences, 35.4% in the physical sciences, and 58.7% in engineering.


Anonymous said...

As long as the US continues to use what the British call "internal assessment", i.e. grades, in place of the uniform and objective entrance exams that China, Korea, and Taiwan use, it will continue to decline, in relative terms, economically, intellectually, and in every other way. It has exhausted its enormous advantage in natural resources.

Nicolas said...

The sheer number of phd seems to me a rater rough indicator of 'brainpower'.

However, the trends regime change clearly represents something.

Anonymous said...

I couldn't believe UO was so competitive.

According to Petreson's the UO physics department had 10 applicants and all were accepted.

Where'd they get that from?

Steve Hsu said...

Uh, how would Petersen's have any good data on our graduate admissions? What a joke.

I was on the admissions committee two years in a row (most recently in 2007) and I would say the typical admitted US applicant has a high GPA (3.5--4.0) in their sci/math courses and good GRE scores, and comes either from a good state university or liberal arts college. The foreign students accepted are stronger, at least on paper.

We accepted a small fraction of the US applicants and a much smaller fraction of the foreign applicants.

Anonymous said...

US News ranks UO 56. Perhaps UO's is a small department, but still, why should admissions be so competitive?

Perhaps there are far more physics bachelors than there is use for?

Steve Hsu said...

I'm not sure what you were expecting. For a US student to get into a top 5 or 10 department they would need nearly perfect grades and very high GRE scores. For foreign students you need those credentials to get into *our* program.

The top few undergrads in a particular year from a decent state school like Ohio State or UCLA can get into a top 5-10 program, but students below that level cannot. UG schools like Caltech/MIT/Harvard/Princeton etc. will send more to top 5-10 PhD programs, but they are the exception.

It's a PhD in physics -- most people can't do it.

BTW, the rankings from places like US News are somewhat size-dependent. Smaller departments can have the same average quality level of faculty and students as a larger and higher ranked department, but just less comprehensive coverage of all fields. In terms of average faculty or student quality I doubt there is much difference between UO and (say) the 20-40 ranked programs. Once you are out of the top 20 the quality of the advisor/group in a particular subfield matters more than the school.

Also, don't forget Eugene is a desirable place to live :-)

Anonymous said...

Once you are out of the top 20 you're really just killing time for 5-6 years before you move on to another field, so why not pick a nice place to live?

Anonymous said...

As competitive as HBS, but
MUCH less money, and usually none. Poor sods who take up physics.

Anonymous said...

Quality and numbers entering physics grad schools aren't good measures of 'brainpower and gloablization'. The programs aren't as selective as you're saying, and the students at even the best schools aren't (with a few exceptions) all that brilliant. I'm basing that on Cornell and MIT a decade ago. The cream of the US student crop is not going into these areas. My hedge fund experience post-school is that the quality of people is uniformly better (with the stars in both groups comparable). As a related aside - when I give brainteaser interviews to potential hedge fund hires, the asian-born students regularly destroy them - much better than the natives. But on hiring, my sample space of ~10-20, the Asians who destroyed the interview process seem to have less intellectual horsepower than expected. So far no stars from that group and some real duds. I've heard 'less creative' comments which I think are bs - but I wonder if there is something to the higher level of diligence and focus on presentation that helps these students on tests and interviews but leads to 'overachiever' performance? (And if I read one more resume from a student having done undergrad in China which says 'number one student in (some big chinese province)' i'm going to flip - how many number ones can there be??? and given the size of the population, why aren't you a genius? (understanding the challenge of conveying performance across academic systems))). Great blog by the way.

Steve Hsu said...

Most of the data I gave in the post are for science and engineering PhDs in general, not specifically physics.

"The programs aren't as selective as you're saying, and the students at even the best schools aren't (with a few exceptions) all that brilliant."

I'm not sure what you mean by that. Are you saying the GPA and GRE ranges I gave are off, or that you just don't think that the actual people turn out to be that great? If it's the latter I would just chalk it up to imperfect filters. I'm pretty sure I gave the right ranges for people we actually admitted to our program.

I doubt that, e.g., the physics PhD program at Harvard or Caltech is admitting more than the top one or two grads from, e.g., Ohio State each year. Whether you think that person is any good is another story.

re: PRC students, read my earlier post from 2007. Despite the fact that they tend to be far superior on paper, the longer term performance (i.e., in actual research) is similar to American students with less impressive records, at least according to anecdotal evidence.

"My hedge fund experience post-school is that the quality of people is uniformly better (with the stars in both groups comparable)."

That is an interesting comment, could you elaborate? At least in my age group I am pretty sure that is not true, based on a pretty big sample.

Anonymous said...

"The cream of the US student crop is not going into these areas."

And they never have. Science and engineering are not acceptable occupations for the children of the econmic/cognitive elite. For these people science is for lower class uncouth nerds.

"As a related aside ...the Asians who destroyed the interview process..."

The best you can do for predicting job performance is give an IQ test for which your prospects could not have prepped. For the "Asians" make sure the test includes a verbal component.

Of the three smartest guys I've known, one is an engineer and the other two majored in some silly social science and then went to business school.

Anonymous said...

The University of Oregon physics department has an accepatnce rate lower than Yale Law School.

What a joke.

Steve Hsu said...

How would you know what the acceptance rate is? All you know is that our entering class size target is 15. But you don't know what our *yield* rate is (ratio of accepted to enrolled applicants).

If you think about this a bit it will resolve the paradox for you. If not, ask someone a bit smarter.

Anonymous said...

You're right. My bad. I thought you aimed to admit 15, not have a class of 15. But how low can yield be? For one thing it can be "dangerous" to accept too many.

Steve Hsu said...

We and other programs like ours have a real problem with fluctuations in size of each year's grad class. While I've been here I think we've seen as small as 6 and as large >20. This is not due to our changing the target, but just variation in yield.

The problem is that our yield is low -- many students apply to and are accepted at multiple programs. This is particularly the case for the PRC students.

But the application numbers I posted are accurate -- when I was on the committee we really did have to read through that many files. I'd say it's because there are many comparable PhD programs in the US (i.e., below the top 20 or so). Think of it this way -- most states have a flagship campus with a pretty strong graduate program. For larger states, perhaps several (or in the case of California, 10!). Then add in the good private universities. You end up with well over 50 pretty good programs.

Anonymous said...

I didn't mean to suggest your figures weren't correct.

BUT the level of competition is an example of market failure.

The number of science majors in general should be limited by law.

Being a professor partly dependent on tuition you're part of the problem. Your cushy life comes at a cost to all your students who have wasted their time taking physics courses.

The promotion of physics (in mass media) is second only to the much more useful promotion of biochemistry/molecular biology. The result is too many wanna be physicists.

Steve Hsu said...

"The promotion of physics (in mass media) is second only to the much more useful promotion of biochemistry/molecular biology. The result is too many wanna be physicists. "

Search on this blog for the many posts I've written with the words "Don't become a scientist!" or "Tale of two geeks" :-)

I would argue it's the opposite of a market failure (in a certain sense): many talented people are willing to accept modest salaries (relative to what they could otherwise earn) in order to do research in physics. Similarly talented people could make much more money as doctors, lawyers, engineers, etc.

Anonymous said...

Still the level of competion is too great. The result is that at a certain point going up the bell curve, you realize however smart you are you probably won't make it AND you'll have to work 90 hrs a week befor tenure, therefore the very smartest will not pursue physics or science of any sort.

An anecdote in this regard:

There was a member of the Forbes 400 a few years ago. I forget his name. He was an Irish American from Boston and went to Boston college. He told his priest advisor he wanted to major in physics. The priest told him, "Physics is only for serious students." So he majored in business, went on to HBS, and made a billion when he arranged the LBO of the company he went to work for. He said, "In retrospect that was the best decison of my life." That is, not to major in physics.

Steve Hsu said...

"...therefore the very smartest will not pursue physics or science of any sort."

That is your contention but I doubt it is true. The type of intelligence that is good for science does not automatically translate into the ability to make money. Also, many people are more interested in intellectual pursuits than in money. I'm a good case in point -- on several occasions I turned down chances to work in finance. But I'm not a monk; I also started a successful startup company.

I do agree with you in the following sense: most people bright enough to become scientists will probably end up happier by doing something else!

Anonymous said...

I'd be curious to what extent (perceptions of) US immigration policies affect the numbers of students coming to the US for graduate school.

Another interesting thing to pay attention to is the flux of PhD-holding academics & professionals in/out of the US, especially from/to countries now producing many PhDs than before.

Blog Archive