Dec 13, 2017 - Estates cost no longer part of indirect cost, but 'direct' charge to project .... User-driven CoE for HPC
The Future of Scientific Research in U.K. Universities Peter Coveney Centre for Computational Science, University College London United Kingdom 13 December 2017 Durham Institute of Advanced Study
Overview
• • • • • • • • •
What is academic freedom? The state of UK research Problems in the system Funding research in the UK The full economic costing model Where does the money go? Industrial collaboration Brexit Conclusions 2
Scientists, freely making their own choice of problems and pursuing them in the light of their own personal judgment, are in fact co-operating as members of a closely knit organization. Such self-co-ordination of independent initiatives leads to a joint result which is unpremeditated by any of those who bring it about. Any attempt to organize the group ... would, in effect, paralyse their cooperation. Michael Polanyi, The Republic of Science, Its Political and Economic Theory: A Lecture Delivered at Roosevelt University, January 11, 1962. Roosevelt University, 1962.
3
The State of UK Research: The Good News ‘The UK punches above its weight as a research nation’ In 2014, the UK represented: • 0.9% of global population • 2.7% of R&D expenditure • 4.1% of researchers The UK ranks first amongst its comparator countries by fieldweighted citation impact, an indicator of research impact and quality
Share of world articles for the UK and comparators, 2010-2014 excluding the US and China for clarity
International Comparative Performance of the UK Research Base 2016, BEIS, https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/ 651174/uk-research-base-international-comparison-2016.pdf
4
But Are We Really World Leading? QoS 2018 World Rankings: 57 of the 76 UK institutions received lower ratings than last year We only spend 1.7% of GDP on research... ...and generate 2% of world patents
R&D intensity - Gross Expenditure on Research and Development (GERD) as a share of GDP - for UK and comparators, 2005-2014.
Relative share of key input and output indicators per researcher. All data are expressed as world share divided by world share of researchers, giving a relative index where a value of 1.0 implies that, per researcher, the indicator is equal to the world average.
5
It’s all about money...
6
Where does the money come from? The percentage of income coming from funding body grants is falling, and the percentage from research grants is static
UK institutions are relying more and more on tuition fees as their main income source, with the obvious risk that a sudden drop in demand (or cut in tuition fees) would have a devastating effect on university finance.
Tuition fees as a portion of income
7
Research Grant Numbers Falling We used FOI requests to establish that EPSRC use multistage application processes to filter out bids, which aren’t reported in these official submission figures, and demand management to penalise “repeatedly unsuccessful applicants”
Reduction in the number of grants available limits research creativity NERC introduced demand management on an institutional basis in 2015
8
Full Economic Costing for Research Dearing Report (1997) highlighted lack of sustainability in HEI funding FEC Background • 2002 - £120m p.a. extra for Research Councils to better support existing research in universities & so institutions are aware of the full economic cost of all research proposals FEC Changes • Indirect Costs on a £ rate per FTE, not % of salary • PI time (salary) will be a charge on project • Estates cost no longer part of indirect cost, but ‘direct’ charge to project • More sophisticated estates costs (at least 4 levels) • Research Councils fund a fixed percentage of FEC, with universities having to identify where the balance of funds is 9 coming from
Calculating FEC: FEC costs are calculated by the “Transparent Approach to Costing” In return for the increased level of public investment in the HE sector, the Government required assurance that the investment it was making would tackle the backlog of investment Calculations are based on:
2012
• Annual TRAC return: Each HEI reports annually on the full economic costs of activities analysed under three types of activity – Teaching, Research and Other. • TRAC FEC for research project costing: This methodology forecasts the full economic cost (FEC) of research projects à FEC model is applied to UK Research Council grants but not accepted by other agencies (e.g. charities, EU, industry, … ). http://www.hefce.ac.uk/media/hefce/content/pubs/2012/201228/ Review%20of%20TRAC%20-%20Consultation.pdf
10
How Much Does Research Cost? Who Knows? • UK Research Councils pay ‘80% FEC’ for projects they fund – includes an cost of academic staff time, institution's facilities, estates & indirect costs.
• It is effectively defined to be loss making! • Is this a meaningful calculation of the cost of doing research? • How is it calculated? Estates Costs
Indirect Costs
PI Time
Direct costs (RA time)
11
FEC Review 2009 • Universities UK and RCUK commissioned a review of FEC • Panel chaired by Professor Alan Alexander • The review made 18 recommendations • No evidence most of these recommendations have been acted on Key recommendation: Institutions tighten audit mechanisms to ensure extra money is used for intended purposes à Our FOI requests reveal institutions have no idea where their money is spent (more later) 12
Student Fees Subsidising Research? UK Research Income Sources 2014/15
HEPI Report Findings: – 37% of research income
• Surplus from non-publicly funded fees of £1.3 billion - 28% of non-publicly-funded teaching income
• Surplus from teaching funds 13% of UK university research – around £1 in £7
• Each international student contributes (on average) £8,000 to British research à Teaching surplus must include UK student fees
‘
’
£1.1B £0.6B
£1.6B
• Research ‘deficit’ of £3.3 billion
Source: Higher Education Funding Council for England, TRAC income and costs by activity 2014-15, 2016: http:// www.hefce.ac.uk/data/year/2016/ tracincome/
13 2017, http://www.hepi.ac.uk/wp-content/uploads/2017/11/HEPI-How-much-is-too-much-Report-100-FINAL.pdf
Investigating University Finances In compiling this talk, we submitted four questions via FOI to EPSRC, MRC, STFC and BBSRC None were fully answered. Half were partially answered. We went on to ask 12 different questions to subsets of 38 different HEIs relating to their finances Just 13% of these questions were answered. The majority were refused because the HEI didn’t hold the data, or couldn’t access it in the available time.
à Underlining the fact the HEIs do not have a good handle on how their money is spent!
14
Are UK Fees Funding Research Too? The universities don’t know The University does not hold data that shows the percentage of fee income used to directly support research activities, as we do not routinely hypothecate our income streams to specific activities WARWICK
FOI Request For each of the last five years, please list the total income of the institution from student fees, and the percentage of that income used to directly support research activities. YORK The University does not allocate a specific proportion of tuition fee income to support research activities, nor is it possible to give a notional figure of how much tuition fee income is spent on 'research'.
NEWCASTLE Our financial systems do not record information that would allow us to identify whether tuition fee income was used to support research activities. CAMBRIDGE
Income credited to the Chest is pooled and used to fund general University expenditure, including research; a fixed percentage therefore does not exist. LEEDS I am afraid that we are not able to provide a breakdown of the percentage of tuition fee income that is used to directly support research activities...income streams are just not accounted for in a way that would allow us to track the tuition fee income through in this way. DURHAM We do not hold information regarding the percentage of that income used directly to support research activities. 15
And where does the money go? The overall budget surplus of the sector is growing at a remarkable rate Surplus = Total Income – Total Expenditure
Source: HESA - https://www.hesa.ac.uk
16
Institutional surpluses The surplus at research active HEIs is far higher than at teaching led HEIs
HEI’s annual reports
17
Paying Senior Staff
Sir Michael Barber, new Office for Students Chair, vows to clamp down on this Plans to publish ratios of vice chancellor's pay to average sector pay Office for Students can require institutions to explain any pay packet over £150,000.
18
Expanded Administrative Staffing Academic and professional staff in UK HE 2015/16 Academic staff
Academics in a minority at 2/3 of universities
Professional staff
Managerial, professional and clerical staff salary bands
Salary range (2016)
The percentage of admin and clerical staff moving to higher salary ranges is increasing
à Draining HEI funds but not doing much to generate revenue Source: HESA, www.hesa.ac.uk
19
Measuring Research Excellence • The Research Excellence Framework aims to measure the impact of UK research • Performance in RAE / REF is deemed to be vital to HEIs à it brings in HEFCE funding • Runs every 6 years; takes 3 years to prepare Staff covered in submissions: Review
No of Staff
REF 2014
52,061
RAE 2008
52,400
RAE 2001
48,022
RAE 1996
55,893
RAE 1992
42,000 20
REF Measures Impact of Research
Average number of submissions per department, by year of publication, from fast and slow-paced fields.
“Either measure what you need to become or be what you’re obliged to measure” – Prof G. David Price
By conducting this process, we are damaging the quality of the research we are performing in UK Groen-Xu et al., Short-termism in Science: Evidence from the UK Research Excellence 21 Framework, latest version available at www.moqixu.com.
UK Funding Up To 2020 The overall budget is growing...
...but the budget to research councils is falling year on year...
BEIS claim that allocating more budget via HEFCE based on research assessment will provide more stability... ... by reducing individual project funding and making more funding available via the institution (but likely to lead to “100% FEC" being levied)
22
Where does all this lead? An HEI case study • The imposition on faculties of a requirement to produce a high surplus (between 5% and 10%) • An increase in faculties pushing Voluntary Redundancies • Academics in leadership positions forced out of positions if they do not accept quotas and potential redundancies • An unprecedented six figure loan to be paid for by generating higher surplus in faculties • Aim towards 60,000 students 23
Industrial Collaboration • UK HEI-Industry Interaction is very poor • Industry warned FEC would make UK research much too expensive • Universities aim for at least 100% FEC and protect IP as first priority Imperial College data:
UK lags behind in average industry income per academic by country
Source: FOI request.
https://www.timeshighereducation.com/world-university-rankings/funding-for-innovation-ranking-2016
24
Industrial Collaboration Case Study University policy is an active barrier to industrial research collaborations • Collaboration result of 2 years negotiation • Industrial partner to fund research project within HEI, possibly up to five years • Withdrew due to shambolic way they were dealt with by HEI’s research services – whole process took too long, and matters were just left to drift – attempt by the department and institution to maximise financial return the sole and paramount concern of the institution
• Additionally, the same partner withdrew from participation in an EPRSC grant
25
EU Funding: Supercomputing & Science EU grants fund different kinds of research than UK agencies, allowing us to achieve things we otherwise couldn’t SuperMUC, Munich “Scientists can work out the way that a candidate drug will act on a target in the body – a protein – and in a matter of a few hours.” 11-13 June 2016 36 hours on entire machine. LRZ Press Release:
~250,000 cores 36 hours
~100 target binding affinities
Generating high impact science:
26
Computational Biomedicine http://www.compbiomed.eu/
• User-driven CoE for HPC in biomedical modelling – Academic biomedical science across all scales – 15 core/30 Associate Partners
Molecular Systems
Organisms (The virtual human)
Organs
Targeting new and emerging biomedical research areas
Cardiovascular medicine
Molecularly-based modelling
Neuro-muscoskeletal medicine
This project has received funding from the European Union’s Horizon 2020 research and innovation programme.
Research Funding & Brexit Sources of research and development funding received by UK Higher Education
Data from: http:// www.universitiesuk.ac. uk/facts-and-stats/ data-and-analysis/ Documents/facts-andfigures-2016.pdf
Losing this funding will have serious consequences for research led institutions 28
https://www.timeshighereducation.com/features/universityfinancial-health-check-2017-future-prospects
BREXIT Consequences The top 4 recipients of EU funding are all in the UK Organisation Name
Total
EC funding (€)
Ranking
University of Oxford University of Cambridge University College London Imperial College London University of Edinburgh University of Manchester London School of Economics University of Southampton University of Bristol University of Birmingham
267 285 249 198 145 123 19 83 103 118
186,087,964 179,926,792 172,404,776 127,662,205 104,373,087 75,224,479 70,580,607 63,297,290 61,198,164 52,909,457
1st 2nd 3rd 4th 8th 10th 13th 20th 24th 33th
• This is already causing massive reputational damage • We will no longer be able to set the research agenda in FP9 29
Conclusions • The number of funded Research Council grants has fallen dramatically over the period from 2000-2016. Individual and collective punishment is inflicted by Research Councils to manage demand. • The mantra that research in UK universities is loss-making is not true. The assertion that only research income at “100% FEC” breaks even has no justification • There is no mechanism available to investigate how FEC is being spent in HEIs. • Universities say they don't know if UK fees are subsidising research because they don’t hypothecate funding separately—they simply pool all their income and spend as they see fit. • HEI engagement with industry is poor. One reason for this is their demand of 100%FEC; desire for IP cpntrol is another. • Brexit is degrading UK science in many ways, including reach, influence and impact of our research, and in terms of the reputational 30 damage caused to UK
Acknowledgements
Stefan Zasada
Chris Greenwell
Dave Delpy
Doug Kell
Graeme Reid
Vicky Olive
Jamie MacIntosh
Tom McLeish
Moqi Xu
Derek Groen
