Along with my colleague Marcos Ramos, I’ve been working recently on some analysis of “world-class universities” and how they have been getting on in the world.  Over the next couple of days I’m going to just lay out some of this research.  Buckle in:

We define “world-class universities” as any institution in the top 200 of the 2017 Shanghai Academic Rankings of World Universities (ARWU).   Our measure of financial clout is expenditures per student, converted to $US at purchasing-power-parity, using the Economist’s “Big Mac index” (yes there are more sophisticated tools out there, but they are usually 2-3 years behind).  Data on both expenditures and student numbers are either from national statistical sources (mainly US, Canada, Australia) or from institutional websites (most of Europe and Asia) or a mix of the two (Japan and the UK).  Of the 200 ARWU universities, we were able to gather data on 174 of the institutions for the years 2015 or 2016.

One main caveat in this data is that neither the term “institutional expenditures” nor the term “student” mean the same thing across countries.  American universities, for instance, pay huge amounts for employee health insurance, whereas in most of the rest of the world this is not the case. They also include the expenditures of their teaching hospitals, whereas in most of the world (Canada included) these are excluded.  Also, they have more ancillary businesses and in many cases operate what amount to professional sports teams. These activities do (usually) generate extra funds for teaching and research, but reporting gross expenditure rather than the net income from these activities overstates the financial clout of American universities compared to those of other countries. Conversely, universities in some European countries do not provide housing to students, as that function has been delegated to a national organization such as Deutsche Studentenwerk (Germany) or CROUS (France). This tends to understate expenditures in these institutions relative to those in other countries.  As for student numbers, both the US and Canada have ways of converting publicly-available student numbers to “full-time equivalencies” while the UK and Australia (the only other countries which have large numbers of part-time students) do not, which means expenditures at the latter will tend to look smaller than at the former.

All of which is to say that when comparing institutional expenditures per student across country, figures need to be treated as indicative rather than definitive. We can see broadly what kind of differences exist, but it is best not to focus on small differences because of data incompatibility. Nevertheless, one can still use these per-student expenditure figures to come up with a broad categorization of ARWU top-200 institutions, as follows:

 Super-specialized and super-rich. Seven institutions have expenditures over $500,000 per student, all of which are in the United States. Five of these are medical schools: UC San Francisco, the Icahn School of Medicine, the Baylor University School of Medicine, the UT MD Anderson Cancer Centre, and the University of Massachusetts – Worcester Medical School. The UCSF’s expenditures are an eye-watering $1.7 million per student, but evidently most of this money is not actually destined for instruction costs. Indeed, the Mayo clinic, which closely resembles these institutions, only claims about $84,000 in per-student expenditures, so there is presumably considerable leeway with respect to how institutions report expenditure under US reporting criteria. The two non-medical institutions are CalTech and Rockefeller, both of which have graduate-student focussed mandates (exclusively so in Rockefeller’s case), and receive a tremendous amount of research funding in extremely expensive areas of study. For instance, well over half of CalTech funds are tied up in just one program, the Jet Propulsion Laboratory.

The Rich. Another 35 institutions have expenditures between USD $100,000 and $500,000 per student. Thirty of the institutions are American, two-thirds of which are private. The five non-American universities that make this grouping are the Weizmann Institute of Science in Israel (basically an Israeli Rockefeller University), Cambridge University in the UK plus Tokyo, Kyoto and Tohoku universities in Japan.

The Well-off. There are 38 institutions with expenditures between USD $50,000 and $100,000 per student. This category remains mostly the preserve of American universities – 20 of them fall into this range (8 private, 12 public) – but it includes an expanded set of non-American ones as well. Twelve of the 18 non-American universities come from Asia: Nagoya, Osaka and Hokkaido Universities and Tokyo Tech in Japan, the mainland Chinese powerhouses of Tsinghua, Peking, Zhejiang, Shanghai Jiao Tong and Fudan plus the National University of Singapore, Hong Kong University, and the Chinese University of Hong Kong. No Canadian institutions crack this group.

The Middle Class. The next 35 institutions have annual expenditures $35,000 and $50,000. This band is mostly where you find the higher-ranked Canadian (e.g., McGill, University of British Columbia), European (e.g., Basle, LSE, Technical University of Munich) and Australian institutions (e.g. Australian National, Western Australia), as well as the poorest among the US publics (e.g. UC Riverside, Purdue).

The Stretched. The next-to-last grouping contains 39 institutions that must face significant challenges staying among the world’s best on budgets of $20,000 to $35,000 per student. European universities comprise over two-thirds of this grouping, which also includes many institutions from Australia (e.g. Melbourne, Monash) and Canada (e.g. Toronto, McMaster), as well as Israel’s Technion.  Most of Canada’s U-15 would fit in this category, though only 8 of them actually make the ARWU-200.

The Over-Stretched.  The final grouping, containing those 16 institutions with expenditures per student of under USD $20,000, is nearly entirely European. Only Israel’s Hebrew University and Australia’s Macquarie and Curtin universities – the only Australian institutions outside the Group of 8 to break into the top 200 – break up the continental monopoly. Two institutions – Kiel and Vienna – have per-student expenditures below $10,000 per annum (ie. financially comparable to Canadian community colleges)

As this brief description makes clear, the distribution of the financial capacity of ARWU top-200 universities is enormous.

Figure 1: Distribution of Annual Expenditures per Student, in USD at PPP, ARWU top-200 institutions

In fact, the financial gap between the top and bottom institutions is so big that picturing them all on a single graph makes most of them look extremely titchy.  Figure 2 shows the distribution of institutions with expenditures of $100,000 or less (that is, it excludes the mainly American and private rich and super-rich institutions that make Figure 1 difficult to read properly). What is striking is how little clustering there is in expenditures per student. The pattern is closer to a power law than it is to a normal distribution.

Figure 2: Distribution of Annual Expenditures per Student, in USD at PPP, ARWU top-200 institutions with expenditures below $100,000 only

In sum: yes, some world-class universities – and most of the top 50 – are very rich indeed.  But below that, it’s quite possible for institutions to make the top 200 with expenditures per student which are at or below the range at which Canadian universities operate.  In other words: world-classness isn’t just a matter of money.

More tomorrow.