Germany and the UK compared
In comparing the DE and UK systems of research institutes, the most obvious quantitative differences are the overall cost to taxpayers, number of centers, & staff numbers, which are roughly seven times higher in DE than the UK: respectively, €15b p.a. vs. €2b p.a; 326 vs. 42; and 120k vs. 16k. Clearly, therefore, the UK has no chance whatsoever of performing at the same level as DE. (If we were to add in the numerous other German not-for-profit institutes (trade associations, Lander governments, etc.) – which I have not done here – the institutional discrepancy would be even more striking.)
But it is also worth noting these particular differences represent a contrast in mass, rather than necessarily a qualitative difference. If you buy more apples, you get more apples. The average cost per staff member (does not represent salary), and per center, is indeed similar in DE and UK (taking into account the very rough nature of the figures). The German numbers are: €46m per center; and €125k per staff member. The equivalent UK numbers are: €47m per center; and €118k per staff member.
One qualitative feature of the famous German Federal research system is that academic disciplines seem in principle mixed together within a series of institute ‘networks’, which are, generally-speaking, application oriented. It is important to realize these networks have evolved over time and not always in a top-down, planned manner. Leibniz, for example, emerged from the Arbeitsgemeinschaft Blaue Liste (blue list partnership) in the 1990s (East German origins). The end result, however, is considerable organizational legibility – readily graspable from outside (an important element in accountability).
This contrasts with the much more historically contingent structure of the UK ‘system’ of ‘public sector research establishments’ – whereby institutes are almost exclusively focused on research that is less application-oriented (some exceptions), and are generally grouped by academic topic or for historical reasons, with lower overall salience.
Put another way, there is no UK ‘system’ (except an imaginary one that can be reconstructed on paper, as below). The UK therefore perhaps has less scope for strategies that cross disciplinary boundaries, and institutions (c.f. the Fraunhofer Key Strategic Initiatives). If such strategies were developed in the UK, there would be no structure or mechanism to take them beyond talk. Experience shows this to be a weakness of the UK’s efforts, in the sense that institutions float in a legal void where integrated ‘national effort’ could occur (if it was sought).
The UK government has been primarily interested in funding a handful of famous universities, rather than institutes, as sites for research, having merged the majority of research funding streams for civilian research into one mechanism (UKRI) focused on higher education. This means the UK’s institutes admittedly get scant attention due to the absence of a powerful unifying brand, unlike the German equivalents, not to mention the lack of collaborative governance. Haphazard efforts to subordinate branding for some of the institutions, associated with the new, and relatively unknown and un-storied funding agency, UKRI, have started. This will do nothing for the visibility of the institutes.
Logically, the UK could have re-positioned its application-oriented labs (alongside what are really strategic national assets disguised as mini-universities, such as the School of Oriental & African Studies, the London and Liverpool schools of topical medicine, etc.) into an interdisciplinary, application-oriented research network in the manner of Helmholz or Leibniz, or perhaps the Swiss ETH-Bereich. But, realistically, this would never happen, or even be contemplated.
Experience shows that endogenous attempts to produce unified effort will probably tend to fail in the UK case; the number of bureaucratic reforms that have never quite ‘taken’ testifies to the difficulty – the Catapult Centres being a prime example. It is what it is; perhaps it would be good to invite German managers to help support the UK system, for example, via international partnership with Fraunhofer. The impact of overseas management practices has been spectacular in the private sector, citing the example of Nissan in the car industry (compared to the UK’s long-defunct indigenous effort, British Leyland). Why not also in the public sector?
One interesting detail is found in the various application-oriented R&D activities associated with UK agriculture. Herein, there was once quite a large number of government laboratories. Much of this capacity seems to remain but with quite different ownership models, notably, joint venture (Fera Science); transfer to university ownership; independent not-for-profit (NIAB), private firm (ADAS); taxpayer-funded (Rothamsted); + various industry-funded networking schemes. It is, overall, an assemblage that often gets forgotten because agriculture plays such a small part in the UK economy.
Information on the German Federal system obtained from the relevant official websites; institutions funded by the States were not included (in relation to the UK, establishments funded via the regional governments were also not included).
I have not included the industrial research association, Zuse-Gemeinschaft, created 2015, and comprised 75 centers (non-profit, private institutes conducting application-oriented research for medium-sized companies).
Nor did I include Deutsche Agrarforschungsallianz (DAFA), founded 2011, with the goal of improving the performance, transparency and international visibility of German agricultural research. Universities, Federal and Länder research institutes are members, including institutions affiliated to the other networks; as such it seems to be a research network ‘in formation’, much looser in structure and not really like the others.
| Name | Scale | Scope |
|---|---|---|
| Helmholtz-Gemeinschaft | 19 centers/>40,000 staff/€ 4.8b | Solving major societal challenges |
| Deutsches Zentrum für Luft- und Raumfahrt (DLR) | 51 centers/8,200 staff/€3.8b | R&D in aeronautics, space, energy, transport, security & digitization |
| Fraunhofer-Gesellschaft* | 74 centers/28,000 staff//€2.8b (2020) | Application-oriented research: health, security, communication, energy & environment |
| Leibniz-Gemeinschaft | 96 centers/20,000 staff/€1.8b (2019) | Addressing issues of social, economic & ecological relevance |
| Max-Planck-Gesellschaft | 86 centers/23,767 staff/€1.8b (2019) | Basic research in natural sciences, life sciences, & humanities |
Basic information about the UK system of research institutes is harder to obtain than the DE case, in such matters as mission, funding and staffing levels, thus making it less legible.
The main sources used were the UKRI and center websites and annual reports; as well as the 2013 GoScience list, a 2015 BIS report, and the EY Catapult Review.
| Name | Scale | Scope |
| Science and Technology Facilities Council (STFC) | 8 national laboratories/1,934 staff/£678m (2018) | Support research with equipment e.g. neutron scattering and light source facilities, laser facilities, accelerator research facilities and high performance computing |
| Defense Science and Technology Laboratory (Dstl) | 6 centers/4,000 staff/£359m (2019) | Science and technology contributing to UK defense and security |
| National Institutes of Bioscience (NIB) partnership | 8 institutes/2,400 staff†/£319m (2017) | Bioscience research and training…underpinning agriculture, bioenergy, biotechnology, food & drink, & pharmaceuticals |
| Catapult Network | 9 centers/4,000 staff/£180m p.a. (2014-2017) | Technology centers designed…to promote productivity and economic growth |
| Medical Research Council (MRC) | 5 institutes/2,500 staff†/£150m (2018)* | Research and training aimed at improving human health |
| Natural Environment Research Council (NERC) | 4 research centers/1,400 staff/£102m* | Research in the polar regions, geology, atmospheric science, & earth observation |
| National Oceanography Centre (NOC) | 2 centers/650 staff/£45m (2012) | Research & technology development to advance knowledge about the ocean |
France and the UK compared
A good deal of taxpayers’ money – but not astronomical amounts – is delivered in FR and UK to buy civilian R&D through the mechanisms discussed below; at a guess, probably a combined total of <€20b p.a. across both countries. In what follows I have tried to understand some of the basics of how this is achieved; quite a bit of tedious detail that makes it hard to see the wood from the trees. I must admit to remaining somewhat baffled by some of the detail and suspect I have missed out some key elements.
The French government organizes civilian research funding through various ‘operators’, notably, ANR and Agence de la transition écologique; these in turn, distribute funding to ‘executors’, i.e., universities and firms, as well as specialized government research agencies divided according to topic (e.g., Inserm, INRAE, CNRS); ‘executors’ can & do of course also receive funds directly from government. ANR dates from 2005, ADEME from 1991; Inserm, INRAE, & CNRS are rather more venerable organizations, tracing their history back to, respectively, 1964, 1946, & 1939.
The UK government decided in 2018 to merge almost all of its top-level apparatus for civilian R&D funding (as well as university funding) into a single agency (UKRI); beneath this agency, funds are directed by various topic-specific ‘research councils’ (dating back in one case to 1920) as well as other UKRI sub-divisions to universities, government labs & firms; these second-tier entities may, in turn, be directed to further distribute grants via sub-contracts (the latter often called such things as networks, focal points, etc. denoting a measure of national-level responsibility and often operating almost as mini-research councils themselves).
Roughly equivalent government research agencies in France & UK*
| FR | UK |
| CNRS; Décret-loi du 19 octobre 1939 | EPSRC + STFC (PPARC, CCLRC, SERC, SRC, DSIR); Science & Technology Act 1965† |
| Inserm (INH); Décret n°64-627 du 18 juillet 1964 | MRC; Royal Charter 1920† |
| Agence de la transition écologique (ADEME, AQA, AFME, ANRED, AFME, AEE, COMES, etc.); Loi n°90-1130 du 19 décembre 1990 | NERC; Science & Technology Act 1965† |
| ANR; Décret n°2006-963 du 1 août 2006 | UKRI; Higher Education & Research Act 2017‡ |
| INRAE (INRA, IRSTEA, Cemagref, CTGREF, CERAFER, etc.); Décret n°2019-1046 du 10 octobre 2019 | BBSRC (AFRC, ARC); Royal Charter 1931† |
In big picture terms, I guess there is not in the end much to distinguish the two systems; there are only a certain number of possible bureaucratic combinations to deliver cash. In both places, but especially in FR, there has been a good deal of administrative twitching, renaming, merging, and the like. As noted elsewhere on this website & as a very rough approximation, the UK agencies were & are organized around academic disciplines; the French agencies were & are more application-oriented.
Let’s now turn to the way the distribution of cash is structured, i.e., funding instruments. As far as I can understand (and my understanding is not complete), the terms listed beneath are in play; FR appears to have more formally identified instruments than UK.
| FR | UK |
| Unités propres de recherche (UPR); unités de service et de recherche (USR); unités mixte de recherche (UMR); groupements de recherche (GDR); fédérations de recherche (FR); unité sous contrat avec un organisme de recherche (USC); unité de recherche (UR); international research laboratory (IRL) | Strategically-funded institutes; institutes; units; centres; networks; grants |
The unité mixte de recherche (UMR) is relatively celebrated because it is one of the commonest ways of organizing research in France, and a bit seems to get written about it. Its legal basis, known initially as associated research unit, was created in 1966; the UMR terminology emerged in the 1980s on a relatively small scale in the CNRS, but was eventually generalized across other French research agencies starting in 1998. The principle underlying the UMR is that it is a partnership between the research agency and another entity (typically a university), and possibly other partners, to deliver particular kinds of R&D and advisory capacity; it seems not as such a project grant but an investment in relationships (on a four year assessment cycle).
The basic structure for the UMR is agreed in principle between the relevant ministry, the research agency, and the university presidents’ conference (CPU), e.g.,: the following framework agreement (accord-cadre) made 16 November 1998 between ministry, INSERM & CPU:
The UMR is an unique structure, placed under the authority of a director who ensures the activity and management of the pooled resources to lead, in their continuum, all the actions of the unit while respecting the missions of each institution. Any mixed unit must be subject to contributions from each of the partners even if their nature and level differ from case to case.
One does not just become an UMR; there are various qualification stages. There are now probably over 1000 UMRs; they vary in scale and ambition. They include a mix of application-oriented and basic science activities; the UMR is a funding structure, not a programmatic initiative (UMR are organized into collective forms, such as the nine discipline-specific ‘institutes’ in the CNRS).
Various comments have been made about UMRs – based on my very cursory search on French websites, & noting that they emanate from parliamentary committees which often seem pretty supine to me and therefore would not level heavy criticism (here, here & here):
- Some institutions host multiple UMRs associated with different research agencies; it is not clear if they are managed in a unified manner.
- The choice of the UMR leadership is important and cannot be undertaken lightly; such factors as relations with the local economic environment, national networks, etc. are crucial.
- Complex budgeting arrangements, purchasing policies etc. of a university hosting the UMR make financial management hard as the systems conflict with those of the tutelle (oversight agency).
Some of the contractual structures noted above are directly translatable across the Channel. The UK’s ‘strategically-funded institutes’, for example, seem perhaps parallel to FR’s ‘unités propres de recherche‘. But, in the UK, perhaps there is no such entity as an UMR, by which I mean a structured, legally-binding, R&D performance framework, made between relevant government departments, the funding agency, and Universities UK.
Within the UK system of grants, one can though identify a small number of long-running programmatic grants given to universities, known variously as units or centers & (occasionally) institutes. I think the common feature of all these interventions is that, despite certain appearances & naming, none are legal entities; just long-standing projects funded within a university. Some of these presumably feature policy advisory services as well as R&D, e.g., the MRC Centre for Environment & Health, co-supported by a public health agency (PHE). One could also point to the EPSRC Framework agreements with individual universities, which apply to the 13 universities that absorb 60% of EPSRC funding. Perhaps herein are elements of the UMR, operating in different places at different intensities. Hard to say.
Conclusion
Let’s wrap up this discussion. Ideally one wants a delivery system for taxpayers’ cash that is legible, i.e., one can quickly understand where the money is going, and how it is being spent (accountability). One also needs the legal ability to intervene as needed (such as to address mismanagement), and to make fair evaluations of performance. Given the total sum is relatively small, as a % of the total government expenditure, one must have modest expectations; reorganizations might not be worth the candle with efficiency gains being inevitably small relative to effort. Perhaps the most important thing one needs is the ability to observe what is going on in sufficient detail to spot intriguing opportunities as they emerge.
One aspect to note on both sides of the Channel is the apparent centralization. In the UK this has been taken to the maximum with the creation of a single agency (UKRI); indeed, the UK is now more centralized than the alleged centralizer’s centralizer, the Fifth Republic. Centralization presumably creates managerial headaches which take years to resolve. But perhaps the problem is more fundamental; it generates management complexity; it is simply not possible to follow the lines of inquiry in thousands of units & determine where there are intriguing developments.
Computer systems often reliant on extensive (and labor-intensive) reporting from projects, are being deployed; pertinent examples are Labintel/RESEDA (FR) & Researchfish (UK). But I wonder if there is just too much noise generated by these systems (most reported activities are irrelevant for all practical purposes). One could say, why not recruit a substantial number of the highest quality of program managers; the reality is that this has certainly not been done. Even if one could recruit an army of geniuses.
At least in the UK, one tacit solution has been ‘fund & forget’. Another has been to avoid co-dependency between projects, i.e., projects run in parallel, not in sequence (thereby abrogating the need for much project management intervention). A third strategy has been to give out cash in large lumps; this has the side-effect of cutting down the number of small and medium sized grants. Recent UK initiatives such as the Faraday Institute (research on batteries) have sought to introduce more interventionist approaches to program management, but these have not really taken root in the UK, partly because there is no strong culture of contract monitoring & enforcement. The French cases would be useful to study; problem is, I lack the information.
One final question worth asking is whether previous R&D agencies such as MRC were seen, bureaucratically-speaking, as sector specific, i.e., research on medical topics was not germane to agricultural research, for example (& each responded to policy initiatives in its own domain & perhaps even under a different ministry). Certainly, their officials might well have perceived ‘a collective responsibility to the development of science’, and often there were shared activities, such as joint memberships of supervisory boards and so forth, but there was not an over-arching superstructure.
The traditional UK division of research councils into academic disciplines mirroring the structure of university faculties (medical, engineering & physics, bio-sciences, etc.) did not perhaps map all that well onto the interests of society or the economy (which might often be more application-oriented). But it is not quite clear to me if centralization into UKRI has addressed that particular issue as the cash is still given out in research council streams.
One could also argue that a unified structure runs the risk of breaking hard-won links between R&D activity and the individuals and groups interested in that activity (the clients). UKRI is indeed separated bureaucratically from the needs of its potential clients, except at the highest levels of government (where the volume of information is evidently too vast to allow any meaningful integration of ends and means).
The historical context in which the past glories of taxpayer-funded R&D played out (e.g., post-WWII national modernization of agriculture & manufacturing) are long gone; new constituencies for taxpayer-funded R&D therefore need to be constantly found. The extent to which lumping R&D together into big, and often quite illegible agencies with little or no name recognition such as ANR or UKRI, helps or hinders this process remains to be seen. Perhaps fashion will turn the other way at some point – the glorious return of the silo.
Science Think Tank 