And on the third day…

So here I am, brain the size of a planet, stuck in a corridor in Polaris House in  Swindon for while the rest of the Astronomy Grants Panel of the Science and Technology Facilities Council considers applications on which I have a conflict of interest. We’ve had two very busy days so far, hence no time to post yesterday, but we’re on track to get through the order of business by the end of today as scheduled. Now I’m at a bit of a loose end I’ve been catching up on emails and other stuff I have had to ignore for the past couple of days.

And now there’s even time for a brief blogette.

It’s a stressful business being on these panels, not just because it’s a lot of work but that everyone involved knows how important the outcome is, for science in general and in terms of the consequences of success or failure in obtaining funding for individual researchers.    Under the current system of “Consolidated Grants”, anyone unsuccessful in this round will effectively be locked out of STFC funding for 3 years. That seems very harsh to me. However, we have to work with the system we’ve got and make the best we can of it.

Anyway, bearing in mind that this is a personal blog and not an official mouthpiece for the AGP, if anyone out there has any comments about the system please feel free to vent your spleen via the comments box. As long as you keep it reasonably polite.

23 Responses to “And on the third day…”

  1. Anton Garrett Says:

    My favourite letter…

    Dear Sir

    I wish to complain most strongly about everything.

    Yours sincerely

  2. That someone who is unsuccessful is essentially locked out for 3 years has been something that I have been concerned about. In fairness, it was also true for Rolling Grants, so it’s not exactly new. If you believe that it is the Astronomy programme that is important, then the 3-year lock out could be regarded as unimportant. It’s not as if were not funding an excellent programme of research throughout the UK. On the other, hand, if you think the system should – in some sense – support individuals, then the 3-year lock out can be very damaging. Being essentially unfunded for 3 years can be very difficult for people who are mid- (or even late-) career and can then disadvantage them in the future. Assessing these grants is not easy and the outcome is not going to be perfect (not in any way suggesting that panels aren’t doing the best they can). Those who are unsuccessful this round may well have the ability to do fantastic work in the future, but it may be very hard from them to do so if those who are successful this time have an advantage in the next round. I do understand, however, that we are in difficult times and the current system may be doing the best it can do given the constraint that exist.

  3. peter:

    i’m now out in the corridor for precisely the same reasons…

    in an unofficial (non-AGP) capacity i agree that the new consolidated/consortium grants rules are far from perfect.

    locking researchers into a 3-year application cycle stifles innovation and biases against younger researchers.

    the rules baring membership of multiple grants produces artificial divisions between researchers and barriers against collaboration.

    the structure of the consolidated grants, which are becoming very large for some institutions, are also making them harder for the referees to review.

    i think some of us supported consolidated grants in the hope that we could retain some of the benefits of the old rolling grants (ie slow variations in support). however, the new process isn’t providing that.

    so, i would argue that we should go back to a “standard-grant” only system, with people proposing individual stand-along projects. these would be far easier to assess and wouldn’t require any arbitrary rules about collaborations/networks/etc.

    to reduce volume there could be rules to restrict how frequently projects can be proposed or re-proposed. similarly, there would need to be a separate process for instrumentation groups (at the cost of perhaps decoupling their technical support from the assessment of exploitation projects).


  4. As chair I am declining from commenting.

  5. The process of “locking out” actually begins at Departmental level. There is considerable pressure not to “over-bid” and this leads to decisions being made locally on how many PDRA cases will be put forward, and who can bid… I am not at all comfortable with this, and very much inclined to agree with Ian Smail – go back to standard grants only (including collaborative teams bidding together, and separate provision for instrumentation groups, post-launch support etc)

    • That’s intentional, of course: reduces the number of cases that the panel and referees have to consider by shifting the work on to departments. A consequence of going to a standard-grants-only system would be a steep rise in the peer review workload. Not saying that’s necessarily a bad thing…

      • from what we’ve seen on AGP there is little real reduction (~10’s%) in the total number of “projects” being proposed. its just that they are bundled into fewer (larger) grants.

        most of the reduction in numbers has come on the standard grant side and that could have just as easily been achieved by only allowing people to apply every second or third year.

      • This process of offloading some level of selection to department level also applies to fellowships, with negative consequences for cohesion and morale.

    • PLS is covered by UKSA grants, not STFC, so that is already dealt with…

  6. We now have two members of the panel saying they don’t support the current system, and yet, by working within it you are, in both fact and deed, supporting it.

    Unless people put their money where their mouth is the system will not change.

    • telescoper Says:

      The money belongs to STFC, not individual panel members! Any system is bound to have flaws, and I think it’s better to have a panel do the best it can within the system rather than not give out any grants at all! And I’m not sure it would work to the benefit of the astronomical community if the AGP consisted only of people who thought Consolidated Grants were a marvellous idea.

      • This does actually beg the question… Does anybody outside STFC think consolidated grants as they stand are a good idea? After al, it was not close to any of the options we were offered.

        And if somebody gave a grants system and nobody agreed to review it, or be on the panel, because they thought the system was so poor, what would happen?

      • Dave, it depends, but if not done carefully, it could lead to a period of no grants awarded whilst others ran out meaning more job losses for post-docs than there otherwise might be.

  7. After the NAM meeting I vented my spleen about the consolidated grants on my blog, including a discussion of some of the unwise words spoken in the community forum. It was only a couple of months ago I noticed I had not hit the publish button. To summarise: consolidated grants (and associated rules) = very bad idea.

    In the spirit of the new STFC I thought I might be more constructive:

    Why not consider a variation on the current NERC model (though this may be about to change)?

    There are responsive mode standard grants assessed via a peer review college (two rounds per year). These can involve PIs and CoIs from multiple institutes. For larger projects there are large grants (I think >£1.5M) which used to be called consortium grants. And for new, freshly minted academics there are smaller, new investigator grants. The latter two (NI and Large grants) are restricted to one grant round per year.

    This allows for large scale projects as well as more modest ones, I also think I am right in saying they are not limited to a 3 year period. This also ploughs the fertile land that is cross-department collaboration .

    I know the concerns with multiple rounds of too much workload for panels but maybe it is time to do away with the panels and introduce a peer-review college. This, of course, has it’s own weaknesses but at the moment NERC is attempting to address some of these (such as drafting successful PI bidders onto the college). There is also a pre-sift where college members reject in the region of 50% of grant applications before external review and full consideration. This may seem draconian but my experience of the system so far has shown it to be pretty fair (as long as NERC does not start insisting on a fixed target of 50% – which would be unwise). Indeed given the way that consolidated grants seem to work it doesn’t seem as if the workload on academics has been much reduced – each consolidated grant contains seperate themes or whatever, which might as well have been submitted individually as standard grants.

    By moving to standard grants again STFC might be able to re-inject the element of responsiveness into its science; this has been missing for some time. Reducing applications by limiting to once every three years may seem like a good idea in terms of administration load but in terms of serving science it is a poor idea. It may lead to savings in a financial sense but it damages scientific output and thus could damage financial return in the long run (bit of an extreme argument I know).

    Aside from that there are other streams:
    national capability which supports the research centres (though core funding is being reduced) and large kit (radars, planes, etc). Of course this tends towards long-term projects since NERC recognises the importance of long-term monitoring being an environmental (in it’s broadest sense) science funder. STFC already does this in a sense due to its national facilities but maybe a retweak is in order. Several of the NERC research centres are spread across groups of universities. This could be a mechanism for providing ongoing support for technical staff in a more concrete manner than the old rollers and certainly the new consolidated grants. Certainly support the existing STFC labs but consider the benefits of closer partnerships with universities and enabling them to retain key expertise without the danger of over consolidation in a few locations.

    Theme action plans: relatively small amounts of money in directed strategic science areas (developed – I believe – with the community, rather than imposed from above). These could be adapted for areas of community-wide interest for dedicated research projects. Maybe a new type of measurement or instrument development. The programme would be refreshed every few years so that it isn’t just a fixed pot for a few academics pet interests. Sometimes it will be relevant to your work, sometimes it won’t. And there will always be the responsive mode.

    I’m not satying that a funding model based on this would be a bed of roses and lead to singing and laughter, but it might actually benefit the wider astronomy community rather than the current system which clearly does not.

  8. Dave, it is possible to work for the panel, doing the best for the community in the circumstances that exist, whilst at the same time working to change the system to make it fairer and more sensible. For example one could write a blogpost offering their view and encouraging discussion on the topic…

    I agree that sometimes one has to fall on the sword and (for example) resign from panels or committees (or not join them in the first place) I’m just far from convinced that this is that time. I can think of other examples not long ago when the surrounding environment and context was vastly different and such action would certainly have been appropriate.

    If the consolidated grants are reviewed and it is decided that they are brilliant and the only way forward, then maybe I would agree.
    In the meantime perhaps we should all be pushing for such a review (maybe some already are – I’m somewhat out of the loop these days), using the astronomy forum to get the message across and convince our colleagues who might not see the flaws in the system.

    • I appreciate that there are two sides to this, and that sometimes that working inside the system can help. But there is no sign of the review you’re suggesting here (unless I’m completely misinformed!) in spite of this being the second year of consol grant operation. Maybe our inside men can let us know if such a review is being considered. If not, something more radical than working inside for change is needed.

      In contrast, when EPSRC scientists got unhappy with their grant system, they delivered a coffin to BIS.

  9. Monica Grady Says:

    Following up on Matt’s comment – our consolidated grant is due in next Jan/Feb, whenever the deadline turns out to be. We have 26 potential co-investigators, all of whom are full of brilliant ideas for scientifically excellent, internationally-leading etc etc projects. How to select which are submitted? I suspect that AGP would balk at a submission from one institution for almost as many PDRA as are available for the entire grant round….

    I am more than ever convinced that my scheme for sorting the funding is the fairest. Take the pot of money. Divide by the number of ref-able or RAE-able academics. Give each academic that sum of money. Then allow the academics to self-organize into consortia to study problems. Do this every year for 4 years. In year three, everyone writes a short, 1 page report of what they have achieved with the money. This will then liberate a similar sum for years 5 to 7. And so on. New people can enter into the lottery each year, assuming they meet the criteria set by the RAE/REF.

    Think how much extra time available for research – no more writing or reviewing of proposals, except 1 page every 3 years. No more pacing the corridors of Polaris House as your proposal is savaged by a panel of ignorant and unappreciative, sorry, interested and informed fellow academics.


  10. Phil Uttley Says:

    Having moved to the Netherlands last year I’m now trying my hand at the Dutch system. One major funding agency across science and social science (NWO) but a diverse range of different funding pots to match the needs of different divisions of science. Astronomy is lumped in with mathematicians and computer scientists for many grants. The equivalent to standard grants (funding for 3 years postdoc or 4 years PhD which also comes with 100k Phd production bonus from the govt when they graduate) is the “free competition” – deadlines twice a year and not a huge pot, but your chances are about one every 10 years, on average. To cover the rest of the time, there are other types of larger grants, and also the NOVA research school in astronomy, funded separately by govt every 5 years and worth ~1 PhD or postdoc per 2 academics during that time. It seems to work okay and the main thing is that there is diversity of opportunities, which is what seems to be really missing from the STFC system now. But of course there is the ever-looming threat of budget cuts, and these types of grants are usually the first to go, as happened with STFC…

    My one gripe about the Dutch system and the way ERC money is going too, is that there is a lot of funding with age/experience eligibility criteria, and if you miss those you lose out on a lot of potential funding. I’m all in favour of funding excellence, but I think the current balance is split too much towards the starters and the “godlike” (vis a vis ERC advanced grants), with big pots of money but leaving the people in the middle (i.e. those who are a few years into tenure but not yet heading a group) to struggle to find things they can apply for. I would be in favour of there being fewer of these huge payouts and the money directed to a ‘standard’ project grant of order a few hundred K to pay for postdocs and PhD students, without setting up a huge team. This is important because we all know that to get one of the big grants you have to be good but also lucky, i.e. with the panel and referees – I’ve been on the right and wrong side of these decisions, and the hope is that if you’re on the wrong side (but your proposal wasn’t trashed) you can dust yourself off, improve your proposal or pursue something else that is promising and have another go next year. With these time-restricted grants you can’t do that, and of course then you are disadvantaged for when you do finally become eligible for the next level up, since you haven’t had so much funding in between. I agree with the need for time-restricted grants to help starters, but I think the overall balance of these big grants may be wrong. Perhaps the problem is that these grants are a ‘one-size fits all’ strategy in order to provide sufficient funding for people who want to set up labs, so that if you just need to hire people the amounts awarded are simply too high.

  11. John Peacock Says:

    For a change, I don’t feel the new system is necessarily a big change for the worse. There wasn’t a whole lot wrong with the old system of rolling grants plus some standard grants. The idea of a rolling grant as a pot meant that a group had the flexibility in principle to support work they felt was high priority for them, or to start up new things – you didn’t need to do exactly what you bid for. So the 3-year lockout wasn’t total, although of course deciding as a group to take money from something the AGP liked in order to pursue something they didn’t like raises difficult internal politics. But you nevertheless could do it – and still can, with CGs.

    The big problem with rollers, which I think CGs have cured, was multiple bites at the cherry. Many places had several rollers which, while nominally covering distinct science, actually allowed a lot of interface science to be bid for multiple times. I used to hate reviewing a roller where I was pretty sure the place already had STFC funding for similar work – but where you didn’t see the complete picture. Now there is no way to play this divide-and-rule game, and I think that’s a good thing.

    The big potential problem with CGs has not yet been encountered in practice, which is what happens over continuity at renewal. The 2-year runout wedge of a roller allowed you to give long contracts to key postdocs, so that they didn’t get frightened and leave while the grant was getting renewed. Now we will find out only about 6 months before the end of the grant whether it will be renewed. STFC allows the money to be spent over 4 years, which helps a bit – but not enough. You can’t hold back a lot of money, otherwise you will only have 3/4 of the number of postdocs you could have had for the first 3 years, producing less science to show off at the renewal bid, so in practice you need to spend >90% of your cash in the first 3 years.

    And in any case, a contract extending even a whole year beyond the renewal date isn’t long enough. Suppose you are a PDRA on a grant being reviewed by the AGP right now, and let’s pretend the existing grants seeking renewal are already CGs. The existing grant would cease in April 2013, and money could be spent up to April 2014. But the postdocs being advertised now typically start in Autumn 2013; if I was a PDRA, the only reason I wouldn’t be applying for one would be if I knew I’d have another chance next cycle, meaning I’d need to have a 2-year contract extending to Autumn 2014. With rollers, you could offer this 2-year cushion to the people you really can’t afford to lose – but with CGs you can’t. I predict this will be a big problem when the first CGs come up for renewal in April 2015.

    The only solution will be for universities to underwrite the necessary longer-term contracts themselves. Rationally, this should be easy: what’s needed is a contract that goes 18 months beyond the renewal date. To do this for all your postdocs requires half the money of a 3-year CG, and there is zero chance in reality that a CG could get chopped by more than a factor 2. But university HR is so conservative that I can see them being unwilling to take the risk. All that would be needed is a letter from STFC on CG award guaranteeing that funding would not fall by more than a factor 2 on renewal and the problem would be solved. Without some such mechanism, a lot of groups will be in danger of losing their best people, who will have to start applying for jobs elsewhere in Autumn 2014. We have two years to fix this.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: