I agree with your conclusion. It's worth thinking about a few complicating details however. First of all, discipline. Implicitly, you are talking about STEM institutes and there's quite a variety of types - the Crick, the Turing, CERN, the MRC Laboratory of Molecular Biology, the Wellcome Sanger Institute have interesting differences (not to mention the various CNRS and Max Planck Institutes). I wonder how the argument might vary if you consider humanities institutes, such as the Warburg Institute or the much-missed Wellcome Institute for the History of Medicine. Second, mission: problem led institutes (like the LMB) are different from policy led institutes (like the Turing). It's a hard distinction to make precise, but its a testable hypothesis that if you start with a problem and think what resources do we need to tackle it, and how will scale help, you get better results than if you just build an institute at scale and think that something interesting will happen if you bring people together and stand back. (This is, to a degree, the same debate as "do we need an industrial policy", only on a smaller scale). Third: history. Maybe it's because I trained in history and philosophy of science, but I think historical contingencies matter at least as much as institutional form: so, Bell Labs were amazingly successful for a couple of generations, and then they weren't. I don't think this was because it "completed its mission" so what was the origin of its decline? (Jon Gertner's book The Ideas Factory is brilliant on this). Finally, universities are - in concept at least - forever institutions. They are intended to outlive their (temporally) local strategies and even missions, and are deliberately loose and flexible, porous and open to change. Whereas institutes are usually tightly specified, and come and go. That said, I don't know where the Princeton Institute for Advanced Study fits into this!
The acid test is whether the Institute achieves things that could not be achievable with a largish grant at a collabotative university or two. Too often though the public funding within institutes simply supports more funadmental research and the key players duck the hard yards of creating impact, and especially economic impact (products, servces, income and jobs). Often the leaders of Institutes find ways to avoid those responsibiities and aspirations, instead supporting the effective buy-outs of their reserachers, and hoping. So Instututes with long term goals are good, but they should still be tensioned. And they make life hard for competing (excluded) HEIs, We dont want the Institute and consortium founders creating the "European Super League for Soccer" (with no relegation). Tension and competition are forces for good within science.
I think this underplays the challenges which institutes, particularly government institutes - or public sector institutes have and which universities don’t so much. Salary constraints have always been a problem (adding bureaucracy to appointments at the very least). When funded outside DSIT the budget is tensioned against other departmental proiorities and short-term pressures can fall on the institute funding. Govt Depts find it really challenging to display policy continuity, particularly as ministers change. UKRI-funded institutes are tensioned against response mode funding which can be pressured by industrial strategy. Universities can be - to be blunt - more ruthless than govt if rationalisation is needed.
All of those are in some way manageable but in essence only exceptional institutes prosper and sector-focussed institutes can struggle to demonstrate their quality. It is tough going for them.
Some of the discussion about research institutes - which it is important to have - doesn’t engage with those complexities and LMB and Crick (and some others) show that - with generous sustained funding - institutes can be truly great. However they are not a model which can be readily deployed and that is a weakness in Paul Nurse’s Landscape review. If only everyone were as good as Paul and as well-funded the Crick (which in any case has had substantial university support).
Orchestration is the key word... Lots of contributions from different players... But common mission. Some more directed than others to achieve desired outcomes. Try choreography as a metaphor too.. Subtly different
I agree with your conclusion. It's worth thinking about a few complicating details however. First of all, discipline. Implicitly, you are talking about STEM institutes and there's quite a variety of types - the Crick, the Turing, CERN, the MRC Laboratory of Molecular Biology, the Wellcome Sanger Institute have interesting differences (not to mention the various CNRS and Max Planck Institutes). I wonder how the argument might vary if you consider humanities institutes, such as the Warburg Institute or the much-missed Wellcome Institute for the History of Medicine. Second, mission: problem led institutes (like the LMB) are different from policy led institutes (like the Turing). It's a hard distinction to make precise, but its a testable hypothesis that if you start with a problem and think what resources do we need to tackle it, and how will scale help, you get better results than if you just build an institute at scale and think that something interesting will happen if you bring people together and stand back. (This is, to a degree, the same debate as "do we need an industrial policy", only on a smaller scale). Third: history. Maybe it's because I trained in history and philosophy of science, but I think historical contingencies matter at least as much as institutional form: so, Bell Labs were amazingly successful for a couple of generations, and then they weren't. I don't think this was because it "completed its mission" so what was the origin of its decline? (Jon Gertner's book The Ideas Factory is brilliant on this). Finally, universities are - in concept at least - forever institutions. They are intended to outlive their (temporally) local strategies and even missions, and are deliberately loose and flexible, porous and open to change. Whereas institutes are usually tightly specified, and come and go. That said, I don't know where the Princeton Institute for Advanced Study fits into this!
The acid test is whether the Institute achieves things that could not be achievable with a largish grant at a collabotative university or two. Too often though the public funding within institutes simply supports more funadmental research and the key players duck the hard yards of creating impact, and especially economic impact (products, servces, income and jobs). Often the leaders of Institutes find ways to avoid those responsibiities and aspirations, instead supporting the effective buy-outs of their reserachers, and hoping. So Instututes with long term goals are good, but they should still be tensioned. And they make life hard for competing (excluded) HEIs, We dont want the Institute and consortium founders creating the "European Super League for Soccer" (with no relegation). Tension and competition are forces for good within science.
I think this underplays the challenges which institutes, particularly government institutes - or public sector institutes have and which universities don’t so much. Salary constraints have always been a problem (adding bureaucracy to appointments at the very least). When funded outside DSIT the budget is tensioned against other departmental proiorities and short-term pressures can fall on the institute funding. Govt Depts find it really challenging to display policy continuity, particularly as ministers change. UKRI-funded institutes are tensioned against response mode funding which can be pressured by industrial strategy. Universities can be - to be blunt - more ruthless than govt if rationalisation is needed.
All of those are in some way manageable but in essence only exceptional institutes prosper and sector-focussed institutes can struggle to demonstrate their quality. It is tough going for them.
Some of the discussion about research institutes - which it is important to have - doesn’t engage with those complexities and LMB and Crick (and some others) show that - with generous sustained funding - institutes can be truly great. However they are not a model which can be readily deployed and that is a weakness in Paul Nurse’s Landscape review. If only everyone were as good as Paul and as well-funded the Crick (which in any case has had substantial university support).
Orchestration is the key word... Lots of contributions from different players... But common mission. Some more directed than others to achieve desired outcomes. Try choreography as a metaphor too.. Subtly different