How new British PhD funding damages UK research and undergraduate education

In the UK (like most places), science PhDs are mostly paid for off of government grants.  Sometimes this is part of a normal research grant, or sometimes it is a direct allocation of funding for PhD studentships from a research council.  However, recently at least two British research councils (the two I work with most) EPSRC and BBSRC, have decided
  1. that you can no longer have PhD studentships on grants, and 
  2. to allocate PhD studentships based on the amount of income a university gets from them.  
This latter unfortunately depends not only on the quality of the researchers there, but also on the size of the university (the number of people who can apply for grants).  The upshot is that Bath, recently ranked the fifth best university in the UK, has trouble getting any funding to pay PhD students.

Small universities, or at least universities with small class sizes, are better for undergraduates more generally.  As with all learning, having more attention from your teachers helps.  But if small universities can't have PhD students, then we also can't have well-qualified teaching assistants to work in small groups with the undergraduates.  Worse, no serious researcher in science or engineering would work somewhere where they can't have PhD students.  PhD students are an essential part of university research.  Since professors are both researchers and teachers, students at small universities are no longer going to have good teachers either.  So the UK is seriously compromising both its research and its undergraduate education with this policy.

The excuse the councils give for this is concern that PhD students do better when they work in decent-sized cohorts and attend classes aimed at them together.  Small universities can't have enough PhD students to really do either of these.  Historically, this has been a serious problem when you compare postgraduate education in Europe and the USA, particularly for the UK.  The problem for the UK is that they will sometimes accept students into their PhD programs directly after their undergraduate degree, so these students  may only have had 3 years of taught courses.  In most of Europe, students have always done 5 years – currently this is split into an undergraduate 3 years and a 2 year Masters program, but historically it was just called a 5-year diploma.  Nevertheless, Europe has also come up with a brilliant solution to this problem: specialist taught postgraduate schools for two weeks which let students from across Europe and even around the world learn from the leaders of their field.  Will teaches at one of these for quanitative social sciences in Ljubljana every year.

I attended a NATO-funded Advanced Studies Institute on Autonomous Agents in 1993 (organised by Luc Steels).  Among my classmates were Sebastian Thrun, who is now directing the research for Google cars and teaching AI at Stanford, which is now being broadcast online to over 100,000 students.  Another was Giulio Sandini, now the Director of Research at the Italian Institute of Technology.  Needless to say I learned a lot – more than I learned in three of the courses I had to talke at MIT, and about the same as the two MIT courses I learned something useful I didn't already know in (theory of computation and principles of computing systems (which was mostly about concurrency).  I got out of taking one course at MIT because they decided the 8 courses I took at Edinburgh were good for one MIT course; I only needed to take six to graduate.)  But back to the NATO ASI, I made friends that I kept in contact with for years, they were all like classmates to me.

So do I worry that my PhD students at a small University don't get enough education?  Well, I do worry, so I generally only accept students who have a good Masters degree (or European diploma), and then I make a point of sending them to specialist courses and conferences.  Recently I've realised the most important thing is for them to work side-by-side with great researchers full time for a while, so I'm making more of an effort to hire postdocs and to send them to larger laboratories than my own for a while.  But having taken these precautions, I'm confident they are getting awesome educations.  Even though Bath is a small university.

The bottom line then is that this is bad policy.  The UK is compromising their undergraduate teaching, their capacity for research, and their competitive standing in global academia by choosing research universities on the basis of size rather than exclusively by their quality.  They get no real advantage out of this that would not be better won by just increasing the standard of educations required for their PhD degrees, which incidentally they are already doing.  This is a shame.


Simon Cutler said…
The way that BBSRC and EPSRC assess and support Training Grants is different; as I manage the Studentships and Fellowships Competitions at BBSRC, my comments relate to BBSRC alone.

Following renewed advice from its BSC Bioscience Skills & Careers Strategy Panel (made up of academics, industrialists, etc), BBSRC ‘s position in not supporting PhD students directly on standard responsive mode grants remains. The clear recommendation is that “individual” Training Grants (such as those awarded through Quota DTG, CASE competitions in the past and DTP Doctoral Training Partnership – BBSRC’s new academic block award – in the future) are the most effective way of ensuring high quality training of bioscience researchers.

You also suggest that BBSRC will “allocate PhD studentships based on the amount of income a university gets from them.” However, if you read carefully the extensive FAQs (Section 4) at it is very clear that whilst research income is a factor, full assessment considers elements such as strategic alignment, quality of research training programme, facilities available to students, student support and monitoring, broader scientific and professional development, governance of the partnership and institutional commitment to bioscience research training. Following more advice from BSC, BBSRC will not use algorithm-based methodology alone to assess DTPs, instead opting for a more holistic competition-based approach which challenges institutions to constantly evolve and improve their postgraduate training, and enables BBSRC to invest in excellent training environments whilst avoiding funding institutions with poor records in student training and support.

Dr Simon Cutler, BBSRC Senior Innovation & Skills Programme Manager
Joanna said…
Dr. Cutler,

Thanks for your attention to my blog and to this problem, and for your more detailed clarification of the process. Nevertheless, I don't see how your comment addresses the basic problem, which is why a strong research-lead university like Bath no longer has direct access to BBSRC PhD studentships. We now have to enter bids for DTCs with other universities as a the smaller & less influential partner and thus cannot necessarily direct PhD student resources to our own institution's strengths. The consequences for research (I think more than teaching for Biology, but I'm not in that department & am less directly involved in the outcomes) are as I described.

In my own department (Computer Science) we have already won a DTC from the EPSRC despite our small size & relatively low research income (and again with another university). However, here the vast majority of places go to EngDs who spend 3 of their 4 years in industry, leading to a short-fall in availability of tutors. Also, all of the students are focused on research only 1/4 of our department conducts.

However fairly or complexly-calculated the process of centralising PhD studentships is, the essential consequences are the same: a good-quality researcher's access to this important resource is determined by the size of their group, not just their own merit or the importance and complementarity of their research to others in their institution. This means there will be less distribution of good researchers & consequently certain types of good teaching across UK universities. But I do appreciate your effort to make it possible with the BBSRC for a smaller number of good groups to bring money into a smaller department and/or university.
Joanna said…
Other comments & corrections on this posting (most of the discussion has taken place on twitter, not here):

First, the best researchers are by no means necessarily the best teachers. Some people are good at both, more are good at one or the other. But in general top researchers are good for a certain kind of teaching & role modelling, and as I hope I made fairly clear, my main claim is that PhD students are important not only for research but also for tutoring small groups and individuals as teaching assistants.

Second, a DTC participant on twitter observed that filling up a large class on one particular topic in one particular university is difficult and results in quite a spread of abilities at that university. So perhaps the problem is not only limited choice for PhD positions within most institutions, but also limited choice of institutions for PhD students. Given that part of the criteria for the awarding of a PhD is to make a unique contribution to a field, diversity of educational experience may be very important to producing both good, innovative scientists and other professionals, and for making them employable.
Joanna said…
Speaking of twitter I should make clear here what I've said there many times before – I don't blame the research councils for their predicament. I've never been anywhere that had 33% cuts in staff and I have nothing but sympathy for people who are. This policy is a consequence of government cuts, but it's hurting the UK in ways that aren't obvious to the government.d