Escaping the Average

Imperial College, the science-oriented school in London, recently pulled out from the umbrella organization, the University of London.

Imperial graduates will no longer have University of London diplomas, but will now have diplomas issued by Imperial.

The reason for the pull-out is that the college administration apparently felt that the Imperial cachet was more valuable than the broader London label. Some of the other constituent colleges of the university are now considering declaring independence also.

What is the equilibrium in this process?

If a school is convinced it is the best in a group, then it will pull out, lowering the average quality, rather than having its graduates tarred with the average label. The next highest school will feel the same way, etc. The equilibrium is independence for all.

I noticed a similar thing when looking at pictures of job candidates from a certain graduate program. Some of them had their pictures on their Web sites, and those pictured were gorgeous. I’m sure the best-looking students among those without pictures realized this, and, fearing they would be taken for the ugly average of the non-pictured, quickly posted their pictures. The equilibrium here too is for everyone to have a picture posted (and unfortunately this blog posts my picture!).

(Hat tip: Pierre Weil)


The logic of degree certificates is flawed on practicality. A number of University of London colleges have recently invoked their right, which has always existed, to issue their own degree certificates. To my knowledge Kings and LSE are doing this, giving students already enrolled the choice of UL or their particular institution.
Most students applying to a university will be unaware of what name appears on their graduation certificate, and just glad of a prestigious institution, regardless of its affiliations. Most of the top UL institutions barely mention UL on their literature, as they have so much to offer themselves.
There is more to Imperials disaffiliation than degree certificates alone.


As a graduate of one of the more prominent in the University of London umbrella (London School of Economics), I think why Imperial and LSE, as I've heard through the grapevine, would want to leave the umbrella is more because of marginal cost.

Imperial and LSE are able to make a lot more money, why should they pool their resources into the greater umbrella? It simply costs them more to stay with the University of London than to opt out.

I really don't think it has do to with the reputation thing as everyone (at least I know I did) weighs the reputation of a school by it's individual standing, not by the overall University of London umbrella. In other words, I doubt people equate UC Riverside with UC Berkeley. (No offense meant here).


what if the job was a circus clown?- would the ugly people all post pics, thereby coercing the beautiful people to post?- I'm assuming that there are no hot bozo/bozettes

Myron W

The same dynamic drives up executive salaries. CEO of XYZ Corp negotiates a salary X% above the average, thereby raising the average for subsequent salary negotiations. And the spiral continues on and on driving out economic justifications like absolute performance or marginal product. It's the scam of executive compensation.

Kevin H

This type of aberrant positive feedback is similar to the half of the reasons why I don't support most school voucher programs.

The families who can afford to pay above the voucher price for schools are also most likely to be above the average performers. Once one of those kids leaves, it moves the average down, increasing the incentive for more above average families to pull their kids out as well, and in the end may result in a only poor, below average children left in public education.



Your proof is not correct. There may not be a least uninteresting number if you allow any real number. e.g. the set (1, 2) has no least element.

What you probably mean is the following proof that there are no uninteresting natural numbers:
Assume there are some uninteresting natural numbers. By the well ordering principle there is therefore a least uninteresting number. But hey! That's pretty interesting. Contradiction.

Nathan Whitehead

There are other forces keeping the university together, just like in the private sector there are forces keeping some firms large. There might be economies of scale for various activities. In addition, brand awareness is valuable. Having more people know of your less respected brand (ULondon) might be better than fewer people knowing of your better individual brand (Imperial).


Well there is finally an answer to the photo problem. The students can now IM grad school their picture using bigstring's new product and the picture will self destruct after a certain period of time. There is no fear that the picture will be left floating around on the internet simply because it can never be reproduced again.

Charles D

I think brand name comes into account heavily here. Simply being better holds little value when no one knows who you are.

I imagine most Ivy League students are more motivated to go to that school for the name and chance to network with other people likely to be rich rather than the education.


There may be a level of quality below which the demand for a degree falls off enormously. Schools which are close to this cutoff may prefer to stay grouped together with other similar schools (even if the others are a bit worse) so that a shock to their quality doesn't send them below the cutoff. And there goes your independent equilibrium.


There is not much point in Imperial's actions. The mediocre label that you mentioned does not affect Imperial's graduates since it will be quite clear where their bachelor or MSc is coming from.

An Imperial's graduate CV will mention Imperial. On the other hand, if you see a CV which mentions University of London, it usually implies a crappy London college which takes advantage of the "umbrella's" higher status.

Chris H

This is similar to the proof that there are no numbers that aren't interesting.

Given that we know some numbers are interesting (0, 1, pi, all primes) then we can create a set of numbers that aren't interesting. But the smallest of these numbers is interesting precisely because it's the least of the interesting numbers. So it instantly cannot be a member of that set. Repeat.

In practice it implies that the colleges, schools and constituent parts can order themselves and for many of them in UL this isn't possible as their specialist areas are non-overlapping. For the generalist colleges (while IC is science focussed, I'll assume it as a generalist in this argument) leaving makes sense, but for the School of Oriental and African Studies or the London School of Hygiene and Tropical Medicine?

- chris (an Imperial grad twice over, who has both Imperial College and University of London on both certificates and is proud of it)


Not Mel Brooks

Everybody and every organization tries to differentiate themselves. Equilibrium will come when everyone is special.


I wonder what type of correlation there is between looks and money earned or employer hiring preference.

I imagine some such study has to have been done, although it would be difficult to empirically demonstrate what "good looking" is.

San Diegan

Follow the money sounds right. Imperial charges their students fees and is apparently favoring the highest paying students or so I heard. Imperial's new independence may be freeing up their financial resources which no doubt needs freeing up.


I don't think your equilibrium point is where you think it is. I can easily imagine someone who thinks they're so ugly as to prefer someone picture in their mind the "ugly average" than to get a glimpse of their real face. This is true on dating sites, which either require a photo, or there will be several people with no photo. The equilibrium comes when you start hitting the people who think they're uglier than they are actually perceived.

Similarly, cache is an elusive thing. It's like branding. When there's a buyout, does the buyer keep the brand of the buyee on the product, or use it's own branding? It's all a matter of the perceived cache.

By your logic, many unions wouldn't exist. Every worker, who thought he was better than the average union member, would leave, until there were none left. But sometimes the union, or conglomerate, or consortium is greater than the sum of it's parts. So being a member is still better than being an average member.


Clement Chung

Although this distinction is symbolic, Imperial has long regarded itself as a separate institution. I'm an IC alumnus, and ever since applying for university through the UCAS system, which classified other University of London colleges (including UCL, Kings, LSE, SOAS etc.) alongside fully-fledged universities, my feeling is that no-one had ever promoted to me that IC was an integral part of the University. The only connection I remember was that we were allowed to use ULU facilities, and that our degree certificate was awarded by the University. Even then (in 2000), we were awarded a complementary certificate by one of the four historic constituent colleges (which roughly fell in line with the engineering, science, mining and medical faculties) to make it clear that although it was a London degree, the graduate was clearly a product of IC schooling. The practice of embellishing a otherwise "ordinary" degree title is nothing new nor unique to IC - c.f. (Oxon.) and (Cantab.) for Oxbridge graduates.


Clement Chung

Footnote to my previous post: Living and working in the US, the reaction of most Americans (even those working in my engineering field) to the mention of either IC or University of London is more often than not a blank stare, and I don't think this is going to change that! I would think that there would be more general recognition of MIT in the UK (if only through "Good Will Hunting"), but don't quote me on that!


Another aside: Most university rankings list several London colleges within their top 10 or top 25 "universities" and not a single applicant I knew who applied, considered or mentioned their application as being to the "University of London". They would always say Kings, SOAS, Imperial, UCL or whichever college it was to which they applied. The same is not true of Oxford or Cambridge, where "which college?" is a follow-up question.

Carlos Manta Oliveira

There is another force that unbalances that equilibrium (at least one). Funding! (Another one is lobbying).

In most european universities state funding is proportional to the perception of quality, and that perception is modelled by the number of publications. This means that sometimes it is useful being under a bigger umbrella.

Consider that the EU offers funding for research. Several institutions will apply to receive, and they will be ranked by the perception of quality. Several parameters can apply, such as to compensate newly joined countries, projects that involve industrial partners, projects that involve knowledge exchange between countries, but usually at least one of the parameters is the number of publications by the hosting institution.

Lobbying is another force that aggregates institutions with similar interests, even sports and medicine faculties for example. Or Economics and Law.

And sometimes just the plain old exchange of knowledge. This can be as simple as purchasing books for a library. If more institutions share services such as a library, less book purchases and subscriptions to publications need to be made.

Thus I am not sure if independence for all is really the equilibrium. Some time dependence brings benefits as well.