alumni speech - 5 July 2007

Ivy League

The magazine US News and World Report has a set of rankings for American universities, for the main purpose of helping various groups of students to select where to apply for admission. The most important criteria used in the ranking relate to how popular a college is with the top students. This popularity is based on several measures. The first measure is selectivity: the ratio of applicants to acceptances; the second measure is yield: the percentage of accepted students who consider a particular offer the preferred choice and actually enrol. The percentage of students returning after their first year is also used as an indicator.

High selectivity could be due to either a large number of applicants, or a small class, and it shows nothing about how good the students are; similarly, a high yield could be due to the students having no other choice. Hence, the numbers alone do not show how attractive a college is. But if there is a standard way to measure how good the students are, then the two numbers become much more meaningful. The class rankings of the students and their school grade point averages provide some information, but the former say nothing about how strong the competition in the school is, and the latter depend on how strict or lenient the grading might be.

This is where the SAT scores come in. The average SAT scores of students from each school provide a usable measure of the student quality and course standard. With this information, college admission officers are able to judge the transcript of each student in the context of the school, and measure the overall popularity of colleges against each other. In other words, the main use of SAT scores is collective, not individual. However, SAT scores do play a part for individual applicants too, varying from college to college, but generally the very top few colleges get so many high score applicants that they do not need to consciously choose on the basis of the scores, and are much more interested in special qualities that would make a student a promising alumni: so and so appears on TV and someone would say "he/she went to University X". They can accept sportsmen, movie stars, children of congressmen or tycoons, etc, with mediocre scores and still produce an impressive class average. In contrast, the slightly lower rank colleges are much more interested in taking in high score students in order to show that "we are as good as the best". While a score of 1600 in SAT1 is not a guarantee of admission, it at least arouses attention and puts one into the file for further consideration.

In addition to these "preference of top students" figures, the financial and staff resources of a college and its reputation among other academics are also important ranking figures. Here the teaching colleges that emphasize undergraduate studies rather than research do not come out so well, even though they attract good students: research universities are able to skim off 50% of all grants received for infrastructure expenses, and business, law, medicine and other graduate school alumni donate generously towards the university endowment funds. Hence, liberal arts colleges like Williams or Wellesley are separately ranked from the top research universities like Harvard and Stanford.

The percentage of alumni who make donations to their alma mata is also included as a factor, as an indication of satisfaction with their experience. For professional postgraduate programmes like MD, LLD and MBA, a curious measure included is the average debt of a school's students at graduation: if this is large, it shows students find the school worth the investment and therefore a large debt is good! Results for specific areas like branches of Engineering and other academic disciplines are also available based on some other formulae.

Several attempts to rank universities world wide were known. One was from Times (of London) Higher Education Supplement, based mainly on the evaluation of a group of people working in universities. The results seems to be highly dependent on name recognition, e.g., Beijing University was ranked much higher than Qinghua (including even in Engineering), whereas they ought to be comparable - presumably people are more likely to recognize the name Beijing because they have heard of the city. The other was from a research group in Shanghai Jiaotong University, based on number of publications, citations, etc. taken from sources like Science Citation Index, a methodology partially adopted by Newsweek when it went into the ranking busines for the first time. and shared by a similar report produced by a Wuhan University research centre. This results in a very different ranking list, e.g, NUS is around 20 in the Times Higher Education list, around 120 in the Jiaotong and Wuhan lists. It was 36 in the Newsweek list, which seems to have done a kind of merger. To some extent, it benefited from the large number of foreign undergraduate students attending with Singapore government scholarships (scholar immigrants) as percentage of foreign students is a factor in the computation.

Since quite different things are being measured, there is no point in comparing the different rankings and arguing about which is right or wrong. What we do know it that universities take rankings very seriously and would try to find ways to increase their own numbers for those measures used in the ranking formulae.

I myself have a very simple measure for whether university A is better than B: if B's bachelor graduates go to A for PhD, and B hires A's PhD graduates as faculty members, but not the reverse, then A is better than B. I will leave you to judge under this scheme whether NUS is better than, say, Texas Austin or Wisconsin Madison.

this is an old article transferred from NUS SOC site

--------------------------------

NUS, your Alma Mater, has often been urged to make itself as good as Stanford and Ivy League universities. To know what this means, we must first explain the term Ivy League and its model of higher education.

Eight private universities located in the north-eastern part of USA: Dartmouth College (New Hampshire), Harvard (Massachusetts), Brown (Rhode Island), Yale (Conneticut), Columbia (New York City), Cornell
(Upper New York State), Princeton (New Jersey), and Pennsylvania (Philadelphia), together comprise the Ivy League. Harvard was established in the 1600s, Cornell in the 1800s, while the other six were established in the 1700s. The origin of the name "Ivy League" is somewhat obscure, but probably arose from a now-defunct football league formed by the eight universities. Stanford is not a member of the League, and also has a shorter history, but has over the years risen to a standing comparable to, and in some disciplines higher than, that of the League members.

There is a College Guide published by US News and World Report (a weekly news magazine similar to but less well known than Time and Newsweek) which ranks US universities and provides other information that
helps applicants to select their future alma mater. (For an explanation of the ranking system see above).

Year after year the Ivy League universities all rank within the top 15, with only a handful of state universities (usually UC Berkeley, UCLA, Virginia, North Carolina, Michigan and Illinois) appearing near the top. The other top private universities (usually MIT, Johns Hopkins, Northwestern, Duke, Rice, Caltech, Carnegie Mellon, etc), though not members of Ivy League, share many of the characteristics of the League members:

1. They are small in numbers of undergraduate students - the largest, Cornell and Pennsylvania, have only around 10,000 each, while the other six Ivy League members each has 4-5000. Stanford has about 7000
undergraduates. Compared with UT Austin, Ohio State, Penn State, Texas A&M, Arizona State (which are just some of the Big Ten State Us), which have over 30,000 undergraduates each, these are very small numbers. In fact, NUS, and soon NTU too, has over 15,000 undergraduates, more than any member
of the League. However, most of these US universities have major postgraduate professional schools and research establishments. For example Harvard has more than 10,000 postgraduate students, while Stanford's Linear Accelerator Laboratory and CalTech's Jet Propulsion Laboratory each has thousands of employees.

2. They are generally well endowed with land, buildings and investments, and receive generous support from their alumni, as well as government and company research grants. (Take the example of Rice University, with $2.5B of assets and just 2500 undergrad students, without expensive postgrad schools of law or medicine to run, it can use for each student the income from US$1M for his/her educational need, theoretically at least.

Those students who can afford it have to pay the regular tuition of >$15K per year, which gives the University additional means to provide for educational needs) Yet, despite all this,

3. They are very expensive to go to, with annual tuition fees of twenty thousand US dollars, and sometimes more. Though generous scholarships are given to poor, especially minority, students and part time work is often available to senior students in the research laboratories, parents without deep pockets would usually not think of sending their children there. For comparison, the tuition fees for Singaporean students at NUS average less than US$4000, higher for foreign students.

4. They are at the top of a large higher education industry (which in total consists of some 2000 universities and colleges around USA) by virtue of their academic reputation, facilities, staffing and snob value, and are highly selective. Students admitted are, usually, either very bright, or very wealthy. (Remember Edward Kennedy being expelled from Harvard for cheating on a Spanish examination?) Consequently, they are less worried about their own job prospects, and can take a more relaxed and exploratory attitude towards their education. Hence,

5. The main role of the staff members is to provide guidance and set an example, rather than to actually "instruct". In fact, because most of them are busy with outside consulting, running their own companies, lobbying for research grants, etc., faculty members spend a great deal of time away from their offices and are often hard to find. In NUS, teaching duties are taken far more seriously, and any staff member that persistently behaves in this way would in the past probably get dismissed, though we are increasingly more like the US universities these days.

6. Similarly, students engage in many non-academic activities, and depending on the prevailing social fashion, campuses become centres of movements for minority rights, anti-war, feminism, environment, business ethics, etc., besides the numerous sports and social activities one usually finds.

7. They do not regard undergraduate education as vocational training, and do not usually offer majors in Hotel Management or Journalism to undergraduate students, though some would have postgraduate schools for such purposes. Undergraduate education is based on the liberal arts model, with students taking courses from a variety of disciplines that are meant improve their minds and make them better citizens and better gentlemen/ladies. It is the large state universities, and latecomers like Stanford and Carnegie Mellon, that took the liberal arts undergraduate degree structure, but filled the structure with vocationally oriented courses, producing the more familiar American model of higher education: a flexible, credit unit degree requirement that can be completed quickly by good students and over longer periods by weak students, and allows a wide combination of vocational interests, such as double majors in Accounting with Computer
Science, Food Technology with Music, etc. Useful as that model might be, it is actually very different from the Ivy League model.

It is also notable that the operating expenditure per student of the 25 top universities range from US$20K per year (Virginia, a large state university) to US$100K (CalTech, with a small student population but large research establishments). The 8 Ivy League universities average around US$40,000. In comparison, the NUS operating expenditure per student is approximately US$10,000, with SOC student expenditure being quite close to the NUS average. Further, many of the NUS degrees, including your degree, can be completed in 3 years, not 4, so that the total educational investment for a graduate is only around US$30,000, no more than the per student expenditure of the cheapest Ivy League University for one year!

So surprised as you may well be, it is useful for you to know that in some ways NUS is already better than Stanford: in terms of trained manpower useful to the industry, it already delivers more per dollar spent than any of the great US universities. But achieving other forms of excellence is a lot harder and the expenditure will be great, as both the US examples and our own past experience would serve to illustrate. Further, many of our old habits will have to be changed. In fact, the changes up to now have done no more than scratching the surface.

INSTITUTIONS RANKED BY FISCAL YEAR 1998 MARKET VALUE OF ENDOWMENT ASSET
(During the dotcom boom the endowments rose dramatically, but fell back after 2000 though there has been some recovery.)

Rank Institution Endowment Assets ($000s) Life Income Fund Assets

1 Harvard University 13,019,736 664,770
2 Texas System, University of 7,647,309 26,526
3 Yale University 6,624,449 31,692
4 Princeton University 5,582,800 158,195
5 Emory University 5,104,801 14,561
6 Stanford University 4,559,066 185,822 August
7 California, University of 3,787,884 100,398
8 Massachusetts Institute of Technology 3,678,127 158,895
9 The Texas A&M University System and Foundations 3,531,517 20,600
10 Washington University 3,445,743 61,600
11 Columbia University 3,425,992 NA
12 Pennsylvania, University of 3,059,401 58,216
13 Rice University 2,790,627 158,594
14 Cornell University 2,527,871 146,120
15 Northwestern University 2,397,715 227,558
16 Chicago, University of 2,359,358 88,822
17 Michigan, University of 2,303,054 68,385
18 Notre Dame, University of 1,766,176 45,292
20 Dartmouth College 1,519,708 78,758
21 Southern California, University of 1,432,786 215,627
22 Johns Hopkins University 1,373,155 90,827
23 Duke University 1,359,992 72,711
24 Case Western Reserve University 1,328,800 36,056
25 Virginia, University of 1,227,880 23,734
26 California Institute of Technology 1,164,183 140,869
28 Brown University 1,111,760 34,476
29 Rochester, University of 1,069,641 19,371
30 Purdue University 1,052,614 103,365
53 Illinois 654M
54 Carnegie Mellon University 653,919 27,588
59 Georgetown University 624,980 2,951

31 Grinnell College 1,019,048 16,804
37 Swarthmore College 833,659 39,134
39 Smith College 793,214 25,553
42 Wellesley College 780,872 66,851
50 Williams College 724,354 53,142
52 Pomona College 675,137 160,811
65 Middlebury College 580,597 35,756
70 Vassar College 554,974 42,000
73 Amherst College 539,800 61,347
78 Wesleyan University 484,654 15,594
81 Carleton College 461,648 18,129
83 Macalester College 460,163 16,316
85 Oberlin College 434,855 19,549
95 Bryn Mawr College 387,486 7,388
100 Bowdoin College 373,300 14,847
109 Bucknell University 339,380 15,276
113 Mount Holyoke College 327,124 14,733
114 Claremont McKenna College 321,260 72,605
121 Pepperdine University 313,520 45,345
134 Reed College 247,560 9,578
138 Haverford College 240,388 11,222
192 Harvey Mudd College 160,956 8,179
198 Scripps College 148,813 13,295
203 Bates College 145,062 NA

(The list is divided into (a) major research universities; and (b) well known liberal arts colleges, whose staff spend more time with students. Note the numbers are very old but the ideas remain same.)