Friday 25 April 2008

UK Elite Universities

Which are the elite universities in the UK? And why is the number declining?

Bruce Charlton

Oxford Magazine. 2008; 275: 22-23.

***

How many elite universities are there currently in the UK? And which are they?

If ‘elite’ is defined in terms of the intellectual quality of their students, then the number of elite UK universities has declined very substantially from about 35 to about 12.

I suggest that the main reason for this decline is the expansion of the undergraduate intake in the most-selective universities.

My suggestion is be that the current elite UK undergraduate universities are: Oxford, Cambridge, LSE, Imperial, Warwick, St Andrew's, UCL, York, Bristol, Edinburgh, Bath and Durham.

***

Introduction

There were about 50 UK universities pre-1992 (when the former polytechnics were re-christened). The current ‘elite’ of these pre-1992 institutions are usually considered to be those thirty-eight research-orientated universities who are members of either the Russell Group (larger institutions) or the 1994 Group (smaller institutions).

Among the Russell and 1994 Groups, according to the Sunday Times University Guide, the top-twenty most-selective UK universities are, in order: Oxford, Cambridge, LSE, Imperial, Warwick, St Andrew's, UCL, York, Bristol, Edinburgh, Bath, Durham, Nottingham, Manchester, King's, Glasgow, Birmingham, Sheffield, Southampton and Newcastle.

But how many UK universities are elite? Are all of the Russell and 1994 Group universities elite, or just the Sunday Times top-20, or more, or fewer? The answer depends on how terms are defined.


Defining the cognitive elite of students

I will define elite universities as those recruiting mostly from the top 10 percent of the population in terms of IQ. Since IQ in the UK has an average of 100 with a standard deviation of 15, the top 10 percent of the UK population have an IQ of about 120 plus.

IQ mainly measures rapidity of learning and ability at abstract logical thinking. It is highly predictive of a wide range of successful outcomes in modern societies such as educational attainment, salary, life expectancy and social class. But IQ does not measure all valuable attributes – for example a ‘conscientious’ personality capable of sustained and methodical work also predicts success in many domains. (For a clear and balanced discussion of IQ see Intelligence: a very short introduction, by Ian J Deary from OUP.)

My definition of the cognitive elite derives from the work of IQ scholars such as Linda Gottfredson or Richard Herrnstein and Charles Murray (authors of The Bell Curve). They note that US data suggest that relatively few ‘high-IQ’ professions have an average entry standard of 120 plus and absorb about half the cognitive elite.

These professions include accountants, architects, scientists, computer scientists, social scientists, university teachers, mathematicians, engineers, lawyers, dentists and physicians. Leading Chief Executives and senior managers make up the other main high-IQ group.

The suggestion is that the great majority of the national elite in societies such as the US and the UK are drawn from the top ten percent of people with an IQ of at least 120; since in modern developed societies (although less-so in less-complex societies) almost all leadership positions require a high level of those attributes such as rapid learning and abstract thinking which are measured by IQ.


Defining an elite university – a majority of elite students

Using the IQ 120 threshold, I will define an elite university as an institution that has a majority of students in the top ten percent, with an IQ at or above 120.

There are currently approximately 800,000 eighteen year olds in the UK population in any given year. This means there are about 80,000 potential undergraduates per year in the cognitive elite group having an IQ above 120 (ignoring undergraduates from abroad).

I roughly estimated the numbers of first year undergraduates in the Sunday Times guide top-20 most selective UK universities by looking at the number of undergraduates listed in Wikipedia and dividing the number by three (this will somewhat overestimate the number of first years because some undergraduate degrees last for longer than three years – for example MAs in the Scottish universities and also several professional and vocational degrees).

In round numbers it turns-out that there are around 80,000 undergraduate first year places at the top-20 most selective UK universities – i.e. about the same number of first year places at top-20 universities as there are 120-IQ 18 year olds. I will assume that virtually all of the top ten percent of 18 years olds by IQ will go into higher education – and this seems to be largely correct.

So, if there was a perfect system and assortment of students by IQ, there would be enough 120-IQ students completely to fill the top twenty universities with none left over, or else to provide between 20 and 40 universities with a slim majority of cognitive elite students.

However, this cannot be the case; because in practice cognitive elite students are spread across a much larger number of institutions. This happens due to personal choice (students who choose to attend a less-selective institution than their qualifications would allow), constraints on personal mobility (eg. students’ need to attend a local institution), centres of excellence located in lower-ranked and less-selective institutions (such as medical schools and law schools – which may be attracting 120-IQ students to institutions that are considerably less selective than this on average) – and of course the inevitable imperfections of national examinations and institutional selection procedures.

My guesstimate, therefore, is that less than half of the age cohort of 80 000 elite – not more than 35,000 students per year - will find their way into the top 20 most-selective UK universities.

It is worth focussing on this number for a moment. My proposition is that there are at-most just 35,000 IQ-120 university students for whom all the best universities are competing. It does not take very many universities to absorb 35,000 UK students per year.

This analysis implies that at most twenty UK universities can be regarded as truly elite in the defined meaning of it being possible for them to have a majority of students from the top 10 percent of IQ.


Fewer than twenty elite UK universities

However, twenty elite UK universities is an upper limit, and in practice the number of elite universities must be lower than twenty.

A further down-grading of this estimate is required because there will be large differences in the proportion of the cognitive elite even among elite universities defined in this fashion.

If US data on the Ivy League are taken as a guide, a university such as Oxford or Cambridge will probably have students with an average IQ more like 145; which is three standard deviations above average – or roughly the top 0.1 percent of IQ, or roughly the top thousandth of the UK population. So that we should assume that virtually all Oxbridge students will have an IQ above 120. This would mean that more than six thousand of the best of the top ten percent students in each year cohort will go to Oxbridge alone.

Recall that there are only about 35,000 potential elite undergraduates. If the top-two universities pretty-much fill-up with elite students, then the same applies – to a decreasing extent – as we descend the selectivity league table. Each decrement in university selectivity will take a lower proportion of the elite among their n-thousand first year entrants; nonetheless the threshold at which there is less than a majority of IQ-120 undergraduates in an institution will be reached considerably before the twentieth university.

The conclusion is that there is currently something between ten and fifteen elite universities in the UK.

But if we go back forty-something years, the average intake of a UK university was less than half, more-like a third of what it is today. In those days, even the largest of the most selective universities took just a few thousand new undergraduates per year, and some took less than a thousand. Inevitably this meant that the cognitive elite was spread thickly across a much larger number of institutions.

My hunch is that forty years ago, instead of about ten to fifteen elite universities there would have been more like thirty to forty elite universities. In other words, a couple of generations ago most UK institutions with the title of ‘university’ could legitimately have been considered ‘elite’.

This means that twenty-something previously elite UK universities have declined to non-elite status over a fairly short period of time – mostly during the past twenty or so years of rapid university expansion .


Who are the current elite among UK universities?

This analysis suggests that there has been a rapid decline from elite status in more than half of the less-selective pre-1992 universities as the most-selective universities have expanded their intake; because relatively few top universities can now hoover-up almost all of the top ten percent of students available for selection.

My point is that a major but neglected cause of the average students’ cognitive decline, which has been noticed in many of the UK’s most prestigious universities, must surely have been the several-fold expansion in the size of the most selective universities. As the annual undergraduate intake of the top UK universities doubled, then trebled in size; they became able to mop-up almost all of the limited supply of circa 35,000 students per year who constitute the UK cognitive elite.

There must therefore have been a very-significant decline in average cognitive ability of undergraduate students at most (but not all) of the Russell and 1994 Group universities – especially a decline of IQ-related abilities such as rapidity of learning and capacity for abstract logical thinking.

The outcome is that the student intake at the minority of most-selective Russell/ 1994 Group universities is bigger in numbers and has largely retained the same high levels of average IQ as before the massive UK university expansion; while among the lower-ranked majority of the Russell/ 1994 universities the post-expansion intakes are bigger in numbers but also the average students are significantly lower in terms of IQ. So that most of the Russell and 1994 Group universities are now non-elite.

In conclusion, I suggest that there are now likely to be only between ten and fifteen elite universities in the UK; where an elite university is defined as one in which the majority of the undergraduates have an IQ in the top ten percent of the population.

Assuming that the Sunday Times data are correct, my tentative suggestion is be that the only current elite UK undergraduate universities are: Oxford, Cambridge, LSE, Imperial, Warwick, St Andrew's, UCL, York, Bristol, Edinburgh, Bath and Durham.

***

Second thoughts: 18 March 2009

I would now consider that in the modern educational system, the personality trait of Conscientiousness counts for as much, or more than, IQ in determining examination results.

http://medicalhypotheses.blogspot.com/2009/02/why-are-modern-scientists-so-dull.html.

So I would now refer to these dozen elite universities as having students in the top 10 percent of examination results, instead of the top 10 percent of IQ.

9 comments:

cantillonblog said...

From personal experience, I find it extemely unlikely that the following is true:-
"a university such as Oxford or Cambridge will probably have students with an average IQ more like 145; which is three standard deviations above average – or roughly the top 0.1 percent of IQ, or roughly the top thousandth of the UK population".

As a primary strategy, they must be satisficing on IQ (whilst secondarily letting in a few very bright people who fall short on other aspects such as potential contribution to college life).

Given that the selection comes at interview stage (since many students who are not accepted get good enough A levels), and interviews are not primarily focused on assessing sheer ability, perhaps this is not a surprise.

Bruce Charlton said...

@C - I certainly agree. At the time I wrote the above I had not extended my analysis to include Conscientiousness, which is heavily selected for at the expense of IQ - esepcialy by an ever increasing emphasis on 'course work (often copied-work).

Furthermore UK universities openly and explicitly favour women, low SE Class, and non-natives; which (with the mixed exception of increased admission of Chinese) reduces peak, and probably average, IQ. Furthermore there is so much cheating allowed in UK school examinations that the system (for the past 20 years or so) selects for undetected cheating.

Average IQ probably reached its peak in elite UK universities c. 1950 and has been dwindling since, and declining rapidly for a couple of decades.

Valentine Cawley said...

Your remarks regarding Oxbridge remind me of an unfortunately anonymous comment I read in the Varsity newspaper (from Cambridge University) some twenty five years ago.

This unknown wit was reported to have said: "The search for extraterrestrial intelligence? I would settle for intelligent life at Cambridge."

That about sums the place up for me as it was in the late eighties. I wasn't impressed.

Perhaps your conscientiousness theory explains my experiences there, somewhat.

Gypsy said...

Assuming there are 800,000 18 year olds, and 6,300 Oxbridge places(3,200 Oxford, 3,100 Cambridge), and that all of the most intelligent students go to Oxbridge and that there's a perfect correlation between intelligence and academic achievement the minimum rarity of an Oxbridge undergraduate is 1:127 implying an IQ of 137, by making that twice as rare, 1:254 we can get a rough implied median of 140.

Since the correlation is not perfect, and some studies based on the CAT seem to indicate a correlation of .80 of achievement to intelligence in the UK we can expect a true minimum of 129.6 and a true median of 132, but assuming these are mathematics students we can give even further generosity because the correlation between achievements in mathematics and IQ is closer to .85 leaving us with a minimum of 131.45 and a median of 134.

This is much more in line with empirical findings the average IQ of Cambridges faculty, which is presumably mostly composed of Cambridge graduates.

It would be very useful if we had actual empirical research on Cambridges Undergraduate population though

Gypsy said...

I hate to be obnoxious Bruce but are you in a position, as an academic to instigate a study of the average IQ of the undergraduate population at Oxford or Cambridge?

Predictions like mine are frustrating because they mean nothing without empirical data and rest on a number of assumptions.

Bruce Charlton said...

@G - I have published several pieces on this general theme archived at

http://iqpersonalitygenius.blogspot.co.uk/

which you might discover by word searching.

What is clear is that the elite universities are not aiming to maximise intelligence among their students - that can be seen from the explicit policies (de facto quotas on the high IQ groups), the exams (evaluations that favour Conscientiousness over intelligence) and by increasingly aggressive affirmative action for lower intelligence groups and individuals.

The average intelligence of university leadership has collapsed over the past three decades - probably down by a full standard deviation At Least.

Add in the culture of pervasive dishonesty in universities, and the situation is not really very rewarding to spend time on. Also, I don't actively work on intelligence any more; I have probably said what I want to.

Gypsy said...

@Bruce

I have my doubts about that, average undergraduate IQ has probably marginally gone up since the introduction of the A* grade, especially in mathematics.

Gypsy said...

Scratch that, Oxbridge entrance examinations and the former S-level would have weeded out the weak just as well, so to speak.

There's probably no difference between the academic elite of yesteryear and today, then.

Bruce Charlton said...

If you look at my online book The Genius Famine you can see an account of the evidence for decline in intelligence ongoing, also the social selection causes of decline in elite intelligence. I have not the slightest doubt if it!