“Operating” versus “Training”

Short version

• Operating and training are the same thing in academic science.
• No matter what happens, the great majority of trainees will still be supported by operating grants.
• If we want to make arguments for the scientific or training benefits of fellowships, we need to be able to say why only the small fraction of trainees who are eligible for and receive these fellowships deserve those benefits.
• Making training better means integrating it into the budgeting, evaluation, and award terms of operating grants.
• Shifting money from “operating” to “training” is an ineffectual zero-sum exchange that does nothing to improve training.

There is evidently discussion going on at CIHR that reflects a belief that there are insufficient funds spent on “training,” and that the place to find money for training is to redirect funds that are currently spent in the open operating programs. This belief points to a persistent and deep misunderstanding in high policy circles about what grants are.

Thinking of “operating” and “training” funds separately makes no sense. (Note this is in the context of lab/team research environments and does not apply to all fields. But it applies to the majority of CIHR-funded research.) Almost all training is paid for by operating grants, and a high proportion — usually the majority — of every operating grant is dedicated to trainee support as stipends and research costs of their projects. This will be true whatever new “training” programs are developed. So: Operating is training.

That is as clear an expression of the academic research model as you will find. Hidden in this are serious concerns about how public science procures research labor, but we’ll put a pin in that for now.

Likewise, because stipends are the major operating expense of labs, training programs like fellowships function as de facto operating subsidies for the labs hosting the trainees: Training is operating.

So yes, funding for training is woefully insufficient, because CIHR — and all of the the Tricouncils — are massively underfunded relative to Canada’s research capacity.

Now, we can discuss whether having a particular pot labelled “fellowships” and attaching it by name to particular people is a good way to fund the salary part of training for a small proportion of trainees. There are good arguments that this can benefit trainees, but I think we need to seriously consider what those benefits are, and, if we agree they exist, why those benefits should be exclusively available to those who are eligible for and receive fellowship support.

Undoubtedly, fellowships have career benefits because of the prestige they confer. I don’t think this is a good thing. This falls under the general category of “picking winners in advance,” something that I think works against meritocracy and tends to reinforce existing privileges and unearned advantages. To the extent that biases influence awarding fellowships (and they do), they exacerbate inequities like pedigree and socioeconomic status and all manner of silly, un-scientific advantages that we should be trying to purge from our evaluatory processes.

Do fellowships confer a scientific or training benefit independent of this prestige benefit? I am open to this but suspicious. I would like to see evidence, for example, that students or postdocs supported from different sources (e.g. fellowship or grant) in the same lab differ in their scientific development or productivity.

A possible benefit of fellowships that I am very supportive of is that they give trainees more intellectual or professional independence from their advisor. I like this idea, but again I am wary of the assumption that it really does this. Yes, your salary is not paid for from a grant, but all the other operating expenses are. How is your research then less beholden to the grant funding of the PI and their control over it? Maybe it gives you leverage to “walk” and take your salary with you. I can see that, in theory. Where I did my PhD, everyone was a “fellow” and we weren’t paid by our PI or out of grants. People changed labs sometimes. Did that substantively change the PI/trainee power relations or confer more “independence” overall? No, it did not. Those relations played out according to the same personality dynamics that drive them everywhere.

But my main concern with this argument is this: if we think a higher degree of independence, or moderating power issues between trainees and PIs is good (and I do), why would we want that only for the small fraction of students and postdocs anointed in advance with fellowships? Because, again, “more training support” at any scale under discussion at CIHR would not change the fact that the great majority of trainees and their research are and will be supported by operating grants.

If it’s just that being paid by a fellowship instead of a grant gives you some advantage over someone without one… well, I don’t think making Star-Bellied Sneetches who will be preferred over otherwise equivalent Plain-Bellied Sneetches is a legitimate use of public funds.

A short break for some RealTalk(TM): You should absolutely be applying for fellowships, and there is nothing wrong with having one. We are all trying to make it in a difficult career. So take the opportunites that exist, but also don’t make the mistake of believing that just because something does (or would) benefit you that that makes it good policy.

But the problems we want fellowships to address remain. If we want to improve training experiences, we have to do it through 1) more research funding overall to provide the best training experiences and opportunities and to reduce career advancement barriers, and 2) by making the conditions of training part of the terms of the award. One possible middle ground would be to earmark fellowships to operating grants and make trainee stipends/salaries ineligible expenses otherwise. Post-award, salaries are paid directly to the trainee. This would allow funders to track trainees and their experiences, which could be incorporated as review criteria in future applications.

No matter what, the goal needs to be to integrate operating and training funds in a way that supports good research training environments and fair, equitable, competitive opportunities for career advancement based on your research, not on how your stipend was paid.

Those solutions are hard. It’s so, so, so, much easier to swap the labels on some existing funding in an ineffectual, zero-sum exchange and call it “training support” and pretend that’s going to make career progression in health research in Canada better. It’s not.

Increasing support for early-career development in CIHR-land amounts to nothing if we stay in the 15% success rate, 25% budget cut wilderness. A training pathway that ends at the edge of this cliff is pointless. Meaningfully supporting training and career development in health research requires a thriving health research funding ecosystem. We don’t have one right now, and unfortunately there are no short cuts…the Naylor Report tells us how big the gap is, and we are nowhere close to bridging it yet.