Stratification in tertiary education

When people call for a university system more like that of the US, they commonly have in mind the idea that Australia should have institutions like Harvard and Princeton, and a belief that more competition in tertiary education would bring this about. There are a couple of obvious problems with this.

First, high-status universities like this provide undergraduate education only a tiny proportion of young Americans. Around 1 per cent of the college age cohort attends high-status private institutions like the Ivy League unis, Chicago and Stanford, and this proportion has been declining steadily over time. Most of the Ivies enrol no more undergrads than they did in the 1950s. Adjusting for population, an Australian Ivy League would consist of a single institution enrolling perhaps a thousand students a year.

Second, the US experience shows that the idea of competition between universities is a nonsense. Harvard, Princeton and the rest were the leading universities in North America before the US even existed, and they are still the leaders today. The newest of the really high status universities is probably Stanford, founded in 1885. Competition between universities is pretty much the same as the competition between the Harlem Globetrotters and the Washington General.s

The reality of US education is a highly stratified system. Below the high-status private universities are the “flagship” state universities, which educate around 10 per cent of the college age cohort (again, a proportion that is declining, or at best stable).

After that, there are lower-tier state universities, two-year community colleges and, worst of all, for-profit degree mills like the University of Phoenix which exist largely to lure low-income students into debt and extract Federal grant money, with only a minority ever completing their courses.

Australia has always had a stratified system, but to a much lesser extent. (More on the history when I get a chance). The big question facing policy is whether to increase stratification, by widening the gap between the “Group of 8” and the rest, or to treat tertiary education like other public services, available to all who can benefit from it, at the best quality we can provide for everyone.

University education systems mirror and recreate the society to which they belong. A highly stratified system, like that in the US and UK, reflects and reinforces a class-bound society in which the best thing you can do in life is to choose the right parents. We should be aiming at less stratification, not more.

Update Just by chance, one of the lead articles in the NY Times advises that, thanks to increased international intakes, the number of places for domestic undergraduates at the Ivies has fallen sharply

39 thoughts on “Stratification in tertiary education

  1. I think you are all reading way to much into this. Christopher Pyne is just upset because, after attending a private school in Adelaide, he only got a lowly law degree from a university that doesn’t make the top 200 in the world rankings.

  2. @tgs

    Thanks for kind words about my lecture.

    The US situation is much worse than Oz. The high-status researchers don’t lecture undergrads much, but their replacements (mostly) aren’t specialised teaching staff in the sense of teaching-focused academics at Oz unis. They are poorly paid adjuncts, who often have only a week’s notice that they will be teaching a course, have to work at multiple institutions to stitch together a living etc.

  3. Dear John,

    Ivy League, flagship state universities, lesser state universities, community colleges, rip-off degree mills. But what about the many (hundreds at least) of private liberal arts colleges across the US? In my (admittedly brief) experience of a US graduate school many of the best graduate students came from these small institutions, typically devoted to good undergraduate teaching rather than research.

  4. @Andrew Gleeson

    I don’t know as much about liberal arts colleges as I should. But my understanding is that they are a pretty small segment of the sector. This story says that only 130 colleges now qualify for the traditional definition of “liberal arts”.

    My impression is that the good ones typically have 1-2000 students, which would make them collectively account for maybe 2 per cent of the age cohort.

  5. @Andrew Gleeson
    It’s a bit misleading to say these institutions are devoted to undergraduate teaching and not research. In fact, staff are expected to produce a very significant research output for tenure and often maintain it after tenure.

  6. @John Quiggin
    This is certainly a worthy topic, and end-to-end good educational systems are rather crucial for any wealthy, well-developed country that wants to stay that way.

    1) I make no claims as to the extent any of the following applies to Australia: I have lectured at ANU, U of Sydney, U of Queensland, UWA, James Cook, and if a recall aright UNSW and U of Melbourne, but that was in the 1990s, so easily out of date.

    2) When evaluating the nature of competition, one has to look carefully at the structure and time-constants involved. General Electric has been very big for a long time. AT&T / Bell System was huge and dominant for a long time, and for many decades, Bell Labs was likely the premier industrial research lab anywhere. ExxonMobil has been big for a long time. In some industries, the power of incumbency lasts a long time, in others, the competition is so savage and fast-moving that that quite powerful companies can fall quickly, and if skilled employees are mobile, they move.
    IBM was long completely dominant in the computer industry, and still powerful, but nothing like as strong as they were in the 1960s. Digital Equipment Corporation was the superstar of minicomputers in the 1980s, quite profitable in 1988 … and gone within 10 years, acquired by Compaq.
    In the 1990s, Microsoft was incredibly powerful, and still large … but by now, Apple, Google, etc lead, and smartphones mostly run some UNIX descendant, not Windows.

    Even areas change: in the 1940s, California was still a bit of a remote backwater, (even though UC Berkeley was already pretty strong), and Stanford was a good school, but not a serious challenge to places like Harvard or MIT. Stanford Dean of Engineering Frederick was long frustrated that his students would go back East (Boston, New York, NJ especially) for the real high-tech jobs, hence encouraged two of his students (Hewlett and Packard) to start a local business instead.
    Later, he got William Shockley to come back from NJ to Palo Alto to start a semiconductor business … from which most chip companies are descended. But still, even in the 1960s and 1970s, in much of high-tech, the places to be were (NJ-Bell labs, NY – IBM, and especially Boston, given MIT and Harvard, and Rt 128).

    Boston is still strong, but Silicon Valley overtook it, for which there is a well-done history by Anna Lee Saxenian, Regional Advantage: Culture and Competition in Silicon Valley and Route 128. (She’s very sharp.)

    3) How about universities? Incumbency in high rank tends to perpetuate itself even more than in companies, given a few factors:
    a) A strong reputation can draw good people, sometimes more than a fair share.

    b) Tenure tends to make people less mobile than in industry, and 2-academic couples are absolutely nightmarish to move, especially to areas where there is only one major university. The SF Bay Area has enough universities to make this easier, but it is a real rarity to see a couple both get hired at the same school, in the same department. Stanford got a pair like that of brilliant young researchers who were bicoastal at the time, one at MIT and one at Caltech, but I hear it was not easy getting that to happen.

    c) More-established schools are more-practiced at getting grants and some get good income from technology licensing or stock from startups.

    d) In places where alumni donations really matter, incumbency is a terrific advantage, as successful alumni naturally give more to their own schools. See US list, sort in descending order by 2013).
    Of course, Harvard, Yale and Princeton rank highly, but many other, newer schools compete well with the other, older Ivy League schools.

    e) Let us try another area, with a much longer history (which they always remind us of).
    In UK, Cambridge recently celebrated 800 years, and one of its colleges. Trinity is reputed to be no lower than the 4th wealthiest landowner in UK. Of course, Oxford is even older and not shy about saying so. All that is serious incumbency advantage. Still, if you look at rankings, a relative newcomer, about 100-150 years old, Imperial College gives them strong competition in some fields.

    Interestingly, as pressure on funding as increased, a few years ago, Cambridge hired a terrific scholar and administrator Allison Richard as Vice-Chancellor, somewhat to the dismay of the old boy network, who harrumphed a lot. She’d been Provost @ Yale, whose endowment is #2 in US, which had something to do with the choice. Cambridge fund-raising changed dramatically … and other UK schools took note.
    (My wife was undergrad @ Cambridge, PhD @ Imperial, and we attend the local Oxford-Cambridge Boat Race Dinners every year, and have often met Cambridge Vice-Chancellors and Imperial College leaders when they visit. My wife introduced Allison at an alumni event, noting that she was exceptionally well-qualified to run Cambridge, among other things, because she’d studied howler monkeys :-))

    f) Anyway, I’d say there is intense competition at various levels, and (I think) at least some of it is good, but I also think the time-constant for changes of rank in universities is much longer than those in at least some businesses.

    3) In any case, I’d say tertiary education may be in for a lot of turmoil. MOOCs are interesting (no panacea), but it is fascinating to watch the competition among schools in that turf. Having been an instructor for several years while doing PhD, I sure think replacing huge lectures by MOOCs is usually a win, especially if it saves more time for direct faculty-student interaction in small groups.
    My wife took a Coursera class on optimization by University of Melbourne and thought it was very well done.

    4) Finally, in many ways, some of the most exciting possibilities for change may not be at the top, but in the middle. A few years back, I was on an external review team forRepublic Polytechnic in Singapore, and I was quite impressed by their approach, and intense version of Problem Based Learning. (Of course, Singapore is not exactly typical, but there are good lessons.) This basically offers 3-year courses for 16-19-year-olds. The first week, one lesson is how to use the Internet to search and assess the credibility of what they find. I sat in on several classes, including one for 16-year olds on statistics (Poisson arrival processes) and another on Internet website systems design for 19-year-olds. I saw students I’d easily have imagined hiring as junior team members who could be productive quickly.

    But most striking was the RP explanation: “Our goal is to take second and third quartile students and help them be as good as they can be, and ready to go into jobs and be productive quickly. Some will catch fire academically and go on to university, but most will go get jobs.” The other interesting thing was the use of a permanent cadre/faculty who developed course materials and taught some classes, combined with a large number of mostly part-timers who needed to understand the material and be good coaches for the students, but who might not have been able to create the courses themselves. The woman who taught the 19-year-olds website design was actually a realtor.

    Now, this was Singapore, and this might not work everywhere, but I cannot but help think that approaches like this (or the apprenticeship style in Germany) do a better job for lots of students than traditional tertiary education. These students were used to working in ever-changing teams, working on projects …every day … and presenting their results and getting critiqued by teacher and the other students, and getting a grade every day.

    I think at least some competition is pretty useful, as well as systems that allow for different experiments and priorities, as one size does not fit all. There is also room for a wide range of funding mixes between private and public.

  7. JQ @27:

    The US situation is much worse than Oz. The high-status researchers don’t lecture undergrads much, but their replacements (mostly) aren’t specialised teaching staff in the sense of teaching-focused academics at Oz unis. They are poorly paid adjuncts, who often have only a week’s notice that they will be teaching a course, have to work at multiple institutions to stitch together a living etc.

    The situation of sessional academics in Australian universities is not all that far from that of the “poorly paid adjuncts” in the US, with perhaps the main difference being that we’re not yet as poorly paid, but the short notice, multi-institution employment, etc., is something I have direct experience of, as the growing reliance on sessionals for teaching gigs that go well beyond the tutoring and guest lectures done by PhD candidates as part of what was once considered their academic apprenticeship.

  8. @Paul Norton
    And every year the NTEU promises to do something about the adjunctification of higher ed (as well as the erosion of conditions more generally). And every year the NTEU settles for a pay rise for the full-time academics instead.

  9. Some data inputs on where Australia stands in higher education. First, in the new Times Higher Ed Supp table on ratings among 100 new (under 50 years old) universities, Australia has 14 – as many as the UK, more than America; . Obviously, that’s partly because the US and the USA got many universities earlier, but you’d still have to say that it supports ProfQ’s position fairly strongly – legacy effects are powerful.
    Looking at the THES list of the top 200 unis ( Australia has only 7, equal fifth with Switzerland and Canada. If you weight for both score and population size (yes, yes, rough as guts, but quick and dirty) Australia comes 10th out of 24, three times as bad as Switzerland and well below the UK but just ahead of the USA and well ahead of Canada.
    Out of the THES top 400 world universities, 19 are Australian – equal fourth with Canada, behind USA, UK and Germany; per population (can’t do a score weighting, THES doesn’t score 201-400) behind New Zealand, Eire, Sweden, Switz, Denmark, and HK, but just ahead of Norway, Netherlands, and the UK, solidly ahead of Canada and the USA.
    If you want 8 out of the top 10, USA; if you want a top-400 (passable) uni for every 1,235,170 people, Australia.
    That said, an all caps complaint about THE WAY THE COMMENTS HERE DON’T START AT THE BEGINNING – at 1 rather than modulo 50. I often read halfway down the column before realising that everybody’s been thrashing things out for 100 posts already.

  10. Whether or not Oz universities would be better off if they were more “like” US universities, I’d be really wary of over-generalizing, because I’m not sure what “like” really means. The US tertiary system is huge and very diverse, especially given 50 states with differing worldviews, economies, and priorities. (Google: Nine Nations of North America for some useful insight: while Oz us hardly uniform, the regions of US are really different).

    Teaching mixes vary all over the map, and I still do guest lectures in classes, and the last few I’ve done (Penn State, Princeton, Stanford) were for classes taught by senior Professors, as in one a few months ago at Stanford for sophomores.
    I do think the MOOCs ate shaking things up.

    In US, most states there is usually one university started as an agricultural and engineering school, usually the home of the agricultural extension service for the state, and such ate often quite strong in life sciences and/or engineering. (For instance, in CA, UC Davis does that, famed for its wine research and education.)

    Anyway, for US comparisons, I might be tempted to narrow the scope from the whole US.
    CA might be closer in population, it does have an extensive public tertiary architecture (University of CA (undergrad-grad-research), California State schools (primarily undergrad), 2-year junior colleges), plus mix of private schools. But CA has enough oddities that one must be careful.

    When visiting Oz every year in the 1990s, at universities I was repeatedly asked in effect “how do we duplicate Silicon Valley”, meaning at least, how could they get the effect of Stanford, et Alvin generating a cluster of high-tech businesses related to the schools.
    (That of course is a different question than where a school gets on world rankings, especially on whole-school ranks.). My usual advice was to look for strength on a few departments and build on that to create narrow centers of excellence rather than trying to do too much. There is a tricky balancing act for government-influenced collections of schools, as creating centers if excellence contends with spreading the wealth around. I am pretty sure that in many disciplines, there is some minimal critical mass for impact, and it is all too easy to fall below that.

    Finally, as implied by my Singapore example, successful societies not only need to create a few excellent institutions, they need to do a good job for a much larger fraction of the population. People worldwide know Stanford, but likely not San Jose State or Foothill College, but lots of people come through those and do just fine.

  11. Taking it another way, 44.2% of Australia’s universities are in the top 400, 30.2% of the UK’s (actually, after trimming off some single-faculty minnows that’s 34.5%), and 0.39% of the USA’s.

  12. @faust
    I don’t think we could could have a prolonged row at dinner even after the third bottle (each) but I don’t actually have clear ideas about stratifications by IQ: solving the Kashmir problem would be easier than prescribing for Australia’s tertiary education or just universities after taking proper account of 1. What every citizen needs to know. No, that’s for schools so start again; 1. The very even distribution of abilities and relevant attributes generally; 2. the country’s future economic needs, assesses with all due modesty about predictions; 3.scarcities, particularly of talented teschers willing to teach and especially to teach those who really need teaching; etc.

    Your point about IQs touches on valid points but isn’t focused and precise enough to be useful. Sure there are people with comparatively low IQs (or reasonably presumed to be so) who are exceptionally successful in important competitive positions – some PMs and Premiers come to mind – and some of their high IQ school contemporaries end up looking like holy fools or drunken clowns. But the fact is that, strictly on average, higher IQ on modern well designed roughly culture fair tests correlates well with success in just about everything where the brain’s cognitive activity matters right down to the performance in certain key positions in American football. Moreover all supposedly important tests of mental, including emotional abilities correlate highly with IQ’s g. And there is the practical discovery by the American army when it gave up conscription that they had better not take recruits with IQs under 90 because at 80 the rate of loss – of almost everything – from accidents was insupportably expensive.

    True, there is evidence supporting your remarks in Chinese-Americans tending to have better SAT scores (themselves not a bad surrogate for an IQ test) than non-Asian students with similar IQ scores. (Maybe the Chinese Tiger mother has taken kver from the Jewish mother of 50+ years ago – though Jewish IQs have been demonstrably higher than those of any other large ethnic group).

    You say that clever but lazy students have trouble getting in to the top institutions. I have no doubt that their later performance, if lazy, may be poor, but they are unlikely to suffer much in America because the SAT is so close to an IQ test. Not so ?

    Would you agree that the elements of cognitive ability that IQ tests measure are important in varying degrees in heavily g-loaded activities? Of course: no well liquored belligerence could make us fight over that. So…
    Would you agree that there are thresholds for different high g activities? E.g could you imagine a professor of physics with an IQ under 125? A Vice-Chancellor of any but Hicksville Uni under 135? If one competing bank’s senior execs were known to have average IQs of 130 (nothing else being known about them) and the other’s 120 wouldn’t you prefer to invest in the former cet. par.? And wouldn’t you have trouble recruiting teachers who could maintain the minimum needed respect of the 2-3 per cent of >130 IQ students who will lagely create the future if you spread clever teachers evenly amingst students of all abilities?

    I repeat for those who didn’t notice or have forgotten that I only started off challenging JQ’s consistency. I have few macro solutions to the world’s problems but a perhaps excessive irritation and reaction to bad arguments and unclear thinking at least from those who should do better. I pay JQ’s bloggers the compliment of being no less capable than my family and make no apology forva comparable intolerance in response to them.

    BTW I like your nom de guerre. As I trust this life is good to you I hope you will be with us for many years.

  13. @ChrisB
    And what that means (or ratings of the top 50 or 100 by more thsne publisher of ratings) is open to discussion as you know doubt would say yourself.

    I am reminded of a survey that reported equal or greater material success by those offered places by the Ivies and their peer institutions but went to the local state college instead than by those who accepted the prestigious places. (They probably didn’t marry mates as rich as they might have though – at least not first time round).

  14. @yuri @ faust
    Oops! for “even distribution” read “uneven d….” Obvious but I apolgise for impeding easy reading.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s