It’s 2017, which means that I have to start thinking about submitting an NSERC Discovery Grant proposal in the fall. For Canadian astronomers, this is a pretty high-stakes operation. These grants are our research-funding bread-and-butter since there aren’t many alternative sources of funding: there are no regular sources of funding from our space agency, for example. It’s a big source of anxiety (for me at least) because a Discovery Grant is more like a hybrid of three different things: a research grant proposal, a job application, and a tenure application. As I described this situation to a US-based postdoc recently, her eyes kept getting wider and wider; compared to a lot of other grants, DGs are unusual. As I start on the process, some reflections are therefore in order.

Discovery Grants are like other research grants in that they describe proposed research and ask for funding to carry it out. A panel of researchers evaluates the proposals and ranks them in order to determine which ones get funded. Sounds pretty standard so far, right? But here are the differences between these grants and others with which I’m familiar:

  • The proposal is for a program of research, not a specific project. So it can contain multiple projects in different sub-fields, as long as they can be linked together. There is no requirement for progress reports during the grant, and what an investigator spends the funds on might not be exactly what’s described in the proposal.
  • The grants are for individual researchers (no co-investigators, no co-principal investigators) and an individual researcher can hold only one DG at a time. Discovery Grants don’t pay investigator salaries, but in the Canadian system, nearly everyone who is eligible to apply already has a 12-month salary.
  • While the proposal has to contain a budget, the budget has very little bearing on the amount of funding actually awarded. The proposals a given panel sees are sorted into bins, and the bin determines the funding level. These levels vary by field — chemistry is more expensive than mathematics, for example — but are not directly related to the proposed research.

Discovery Grants are like job applications in that there is emphasis placed on both the proposed research and the individual applicant’s track record. The evaluation is based on three criteria: merit of the proposal, excellence of the researcher, and contributions to the training of highly-qualified personnel (HQP). It’s not only what you say you’ll do, but what you’ve already done (overall, not just in relation to the present proposal). This could be considered the “reporting requirement” for previous Discovery Grants: if you didn’t do anything with your previous grant, it’s hard to argue for continued funding.

There is an entire section in the proposal in which the proposer must describe his or her “most significant contributions” to research, including how this research has contributed to the training of students and postdocs. It used to be that the marker of success in training was the fraction of trainees who went on in academia: prestigious postdocs, faculty positions, etc. This is problematic for many reasons and I’m told that at least some panelists interpret “successful” training more broadly. The importance of track record, combined with the “ranking determines budget” business, means that it’s tempting to interpret the funding as a direct metric of competence: if my Discovery Grant is half as big as someone else’s, I must be half as good a researcher.

One way in which Discovery Grants are more like tenure applications than job applications is the relatively high success rate. Many researchers in Canada argue that the high success rate is a good thing: it assures relatively stable funding for a large fraction of researchers and avoids some of the randomness and time-wasting associated with very low success rates. Like tenure applications, though, the cost of failure is high. Not holding a Discovery Grant, particularly in fields where there aren’t many other funding sources, makes it difficult to do the research and training needed in order to successfully apply again.

There is a direct relation between Discovery Grants and tenure as well. The first DG renewal application is often due the year before a tenure application, and failure to renew would likely not help a tenure case. Like tenure applications, DG applications are all about the individual researcher, so demonstrating independence and leadership are critical. You could worry that a focus on funding people rather than projects results in an “old boys’ network” determining who gets funded. This may be less of an issue in Canada compared to other countries because the research communities are smaller and few researchers would be complete strangers to the panel members; I’m not sure how you would test this hypothesis, though. Ironically, the connectedness of Canadian research communities can also be a problem: for several years the Physics panel had no Canadian astronomers on it, because NSERC said that we all worked together too closely and had too many conflicts of interest!

The Discovery Grant system has aspects that are good (high success rate), not-so-good (small grants; of course related to the success rate!) and downright odd (the requirement for a budget which is then ignored). Like it or not, it is the system we have, and there don’t seem to be indications that any major changes are coming. Time to get thinking about my future research and finishing up those languishing papers!