, , , , , ,

Spoiler alert: not very.

Research costs money.  Whether it’s lab analyses, field work, or just paying people, research costs money.  It’s a lesson that every nascent B.Sc. graduate learns when they start asking potential supervisors.  “Do you have any funding?” the supervisors ask, “No”, the student replies. The chances that they’ll be able to join the lab are therefore severely reduced.

If we assume a 2-year MSc, and 4-year PhD, funded at minimum NSERC rates, that’s about $119,000 over 6 years.  Now, some of that could come from a variety of sources, including NSERC post-graduate scholarships (PGS).

Then the search for post-doc funding begins, and as I’ve pointed out, the odds aren’t good.  And most folks will do 2, 3, or more postdocs before they land that most fabled of academic jobs: the tenure track position.

But the struggle doesn’t end there – research groups rarely fund themselves (unless you run analyses for others, but that’s another story).  In Canada, the main funding mechanism in ecology & evolution is the NSERC Discovery Grant (which run ~5 years, and average just under $30,000/year).

Lather, rinse, repeat.

Some back-of-the-envelope calculations

So what are the chances that a MSc student will go on to land a Discovery Grant & get their first renewal?  And is there a bottleneck in the system somewhere?

First, some assumptions:

  • Grad students, and to a lesser extent, postdocs, are funded by a variety of sources, not just NSERC.  But there aren’t any data for those sources.
  • NSERC only funds 1 year of a MSc, and 2-3 of a PhD so that any single grad student receives no more than 3 years of NSERC support (which is odd, when the minimum time is 6 years for a MSc + PhD, but that’s a tale for another day)
  • NSERC also operates industry-  and government-funded postdoc programs for which there are no numbers (update: see comments for discussion of the industrial postdoc), and which are largely (if not entirely) funded by the industrial or governmental partners.
  • For the purposes of this post, I’ll assume that postdocs that receive an NSERC PDF are either able to find another postdoc, or move into a faculty job.
  • We have no idea what proportion of PDFs move into faculty jobs.  It’s probably > 50% (PDF, see table 7.1), so let’s be generous and assume that half of the postdocs get faculty jobs where they could apply for Discovery Grants.
  • Yes, funding rates vary from year to year, but I’m going to use the most recent (2012 or 2013 depending on the program).
  • Lastly, NSERC funding rates for the PGS-M (post-graduate scholarship – masters), and PGS-D (post-graduate scholarship – doctoral) are overestimates because it only reports on those applications that are sent to NSERC by universities. Unless someone from university admin cares to chime in with actual numbers, we’ll work with what we’ve got.

Ready? Hold on to those mortar boards, boys and girls – it’s going to get rocky from here on out.

We’ll start with a class of 1000 B.Sc. graduates who are all admitted to a MSc program to keep things simple (the transition probability is more likely ~25%, but higher for those who do a research-based honours thesis).

The success rate of the combined CGS-M and PGS-M programs is 53%.  Right there, that takes us down to 530.  We’ll assume that all of these funded students love research so much that they’ll go on to do a PhD.

These 530 prospective PhD students face a slightly tougher field, and only 44% of applicants are awarded a PGS-D.  That takes our theoretical group down to 233 people, or only 23% of our starting population.

Now, these remaining 233 newly-minted doctors all need to do a post-doc, and not surprisingly, this is the major choke-point.  The success rate of NSERCs PDF program was 7.8% last year, which translates to just 18 fellowships.  We’re left with under 2% of our starting budding professors.

Let’s assume that half of these are able to navigate the cut-throat academic job search, and land a tenure-track job – that’s 9 wide-eyed (and likely exhausted) faculty applying for their first Discovery Grant (DG).

Early-career researchers (ECRs) are judged a little more leniently in their DG applications, and the ecology/evolution evaluation committee (committee 1503 if you’re keeping track) funded 52% of ECRs (PDF, Table 7). That translates to 5 Discovery Grants in our population.

If you get a DG, you’re more likely to get it renewed in 5 years’ time, and Committee 1503 renews about 82% of applicants (PDF, Table 7).  So after 15-20 years, from the start of a MSc to submitting a tenure dossier, there’s a 0.39% chance of being funded successfully the whole way through.  That’s basically 4/1000.

Program Success Rate Number
Starting population 1000
PGS-M 0.53 530
PGS-D 0.44 233
PDF 0.078 18
Tenure Track 0.5 9
DG-ECR 0.52 5
DG-Renewal 0.82 4

As I said above, these are very rough numbers, but they’re based on what’s available.  There are other sources of funding (provincial government, federal government departments, private foundations, etc), but when one thinks of research funding in Canada, one thinks of the TriCouncil (the collective noun for NSERC, SSHRC, and CIHR).

At my only in-person academic job interview, I was asked by the department head and faculty dean what sources of funding I would use to support my research.  My default answer was NSERC.

What’s more, these numbers are generally going down.  NSERC-wide, the proportion of Discovery Grants to ECRs dropped from 77% in 2002 to 62% in 2012.  Renewals are down from 95% in 2002 (!!!) to 77% in 2012 (summary here; Table 2, PDF)

But what’s most important, I think, is that it’s obvious where the bottleneck is: postdoc funding.  NSERC rewards the training of “highly-qualified personnel” (HQP; grad students, postdocs, and technicians) in the Discovery Grant application process. But the postdoc funding available is in high demand and low supply.  I suspect another bottleneck occurs at the hiring stage, but there aren’t many data for that transition.

What we need is a mark-recapture study to generate a population viability analysis (PVA) where we can estimate the “survival” of each “age class” (career stage), and estimate the “transition probability” (success rates) between career stages.

But until that happens, we can at least be honest with the young researchers we interact with.  As a grad student, I always assumed that it would be tough, but not impossible to land a faculty job, and get my own research group off the ground.  Now, I’m not so sure.

UPDATE: as others have pointed out on Twitter, Discovery Grants aren’t the be-all and end-all.  There are other sources out there, and we need to make grad students and postdocs aware of them.  But NSERC is often used to leverage funds from other granting agencies, and is more likely to be unfettered (i.e., not tied to a specific project).