The (false) headline conveys the sporting analog of NSERC’s new policy on Postdoctoral Fellowship Competitions:

Effective as of the 2013 competition, you can only apply once to the NSERC Postdoctoral Fellowships (PDF) Program; however, applicants whose first PDF application was submitted prior to the 2013 competition may submit a second application provided they are within the eligibility window.

What’s going on? Why would Canada choose to limit the pool of participants competing for advanced training opportunities in science and engineering? A recent letter to the Canadian Association of Postdoctoral Scholars by NSERC’s Director (Scholarships and Fellowships Division) Serge Villemure gives the following reasons:

In recent years, NSERC has seen a growing disparity between the number of applications submitted to the Postdoctoral Fellowships (PDF) program and the number of awards available. As a result, NSERC has decided to reduce the maximum number of applications an individual may submit in a lifetime to its PDF program from two to one.

This change to the eligibility rules will contribute to a better alignment between both the number of applications submitted and the awards available, thereby streamlining the application and review processes. Limiting the number of applications an individual may submit to the program will not impact the the current budget projections or the number of anticipated awards available.

The success rate for the postdoctoral fellowships competition in 2011 was 9.3% and in 2012 the rate was 7.8%. (The tables and visualizations are appended below.) Another strategy to confront the “growing disparity” is to invest more money into Canadian human capacity for research and development by expanding the number of awards. However, changes in NSERC policy over the past decade have transferred investment away from its mission supporting discovery and the training of highly qualified personnel into many new programs aimed at commercialization of research. Restricting Canada’s young scientists to one postdoctoral fellowship competition per lifetime has “better alignment” with the transfer of funds toward commercialization, but it is a bad policy change.

Funding support for graduate programs from Ontario (and likely from other provinces too?) is frequently limited to four years. This means that faculty and departments are under pressure to have their graduate students complete their PhD in four years. Unfortunately, many students do not meet this timeline. Funds to pay for extensions of PhD studies into a fifth and sometimes a sixth year must come from other sources and are often uncertain, conditional upon adequate progress, and may involve expanded teaching responsibilities. Graduate students know all this.

Consider the point of view of a graduate student. Suppose the key advances for the student’s thesis are completed during the summer between the third and fourth year of studies and the student starts writing the thesis during the Fall of the fourth year. The funding uncertainty for the fifth year motivates the student to want to finish the PhD in the fourth year. To maintain a career in science, the student needs to also spend that Fall preparing job and fellowship applications, a process that can take up a lot of time and mental energy. The student’s application materials (research statement, letters of recommendation, thesis abstract) will be not as strong as they would be if the thesis were entirely nailed down. Nevertheless, the funding uncertainty for the fifth motivates the student to submit postdoc applications in the fourth year.

What happens next? In this situation, students sometimes get a postdoc but more frequently don’t. When they don’t, they stay on for another year and often make substantial advances. Their science comes together during the fourth year and the summer thereafter. They have a working draft of their thesis at the start of the fifth year and can concentrate on applications. Instead of merely talking about the student’s potential, the letters of recommendation can reference accomplishments. Students who fail to land a postdoc offer in their fourth year often emerge as extremely strong candidates in the next year.

PhD students will soon be asking graduate advisers for advice: should I apply for an NSERC postdoc now or should I wait until next year? The right answer was both. Under the new policy, the answer is not clear. A certain outcome: some excellent candidates will be forbidden to enter the competition because they applied the year before.

NSERC’s new one-postdoc-competition-per-lifetime rule combined with the funding uncertainties around fifth and sixth year support are a lethal combination. The victim is Canada’s scientific research capacity.


I’ve set up a “hive” on BuzzData (an open social media platform for discussions around data) focused on NSERC. My view is that there is a need for respectful discussion about Canada’s research and development policy driven by transparent data. So far, there are four public Datarooms devoted to the following topics:

Others are welcome to join the hive.

NSERC Funded PDFs data (Thanks David Kent.)

  • Awards/Applicants (Year)
  • 250 / 1169 (08)
  • 254 / 1220 (09)
  • 286 / 1341 (10)
  • 133 / 1431 (11)
  • 98 / 1254 (12)

(Extracted from NSERC’s Scholarships and Fellowships Competition Results.)

The acronyms appearing in the tables are defined as follows:

  • CGS M/PGS M (one-year scholarship for the first or second year of graduate studies);
  • CGS D2/PGS D2 (two-year scholarship tenable during the first five years of doctoral studies);
  • CGS D3/PGS D3 (three-year scholarship tenable during the first five years of doctoral studies); and
  • PDF (two-year postdoctoral fellowship).

Visualizations by Brent Pym:


2012-08-21 Addendum (New visualizations by Brent Pym.)


10 comments as of now

  1. Serge Villemure, the director of NSERC’s scholarships & fellowships division, gives a single reason for this change in policy:

    “This change to the eligibility rules will contribute to a better alignment between both the number of applications submitted and the awards available, thereby streamlining the application and review processes.”

    Suppose that we take Mr. Villemure at his word: NSERC needs to streamline the application and review process. Given all the new indusrial-partnership programs, I don’t doubt that the NSERC staff are stretched to the limit. Incidentally, this restriction does not apply to the Industrial R&D Fellowships Program.

    One is led to the question: what is the current application and review process?
    The only description I could find is here:

    I think this process can easily be drastically streamlined. There is multiple redundancy in the process already. They can reduce the number of meetings and committee members, take more seriously (and require) the internal rankings of the applicant’s current institution, etc.

  2. Karel Casteels @ 2012-08-17 12:10

    Well I’m glad the NSERC PDF issue seems to finally be gaining some attention and, hopefully, traction.

    To repeat a comment I made on The Black Hole blog, it should be noted that NSERC’s reason for the drop this year (and partially the reason for the drop last year) is that they had to move money to fund this year’s new CREATE programs. On the other hand, the lifecycle of a CREATE program is 6 years, and the first ones are only 4 years in. Therefore, there are 2 more years of new CREATE programs that will need to find funding from somewhere. Since the IRDF program is evidently untouchable, it seems that unless there is an increase in the total S&F budget, we can look forward to two more years of huge drops in the “traditional” S&Fs.

    Also, James, considering how negatively these “reallocations” affect pure mathematics, is the CMS going to at least try to address the issue this year?

  3. Thanks Marco and Karel for your comments.

    Karel: what would you like CMS to do to address this situation? There was a considerable amount of discussion related to NSERC and Math/Stats at the 2011 Winter meeting. My report is posted here:

    The transfer of funding away from basic research toward commercialization affects all disciplines, not just math/stats. Different academic societies have expressed concerns but I am unaware of any leadership on these issues from the Royal Society of Canada.

    Karel, are you aware of any data demonstrating the transfer of funds you mentioned? If so, can you share it with me via email or join and share it publicly on the data sharing site:

  4. Charles Mire @ 2012-08-19 12:08

    It should also be noted that the CREATE programs are for non-postdoc students (although a postdoc may be involved and get limited funds through CREATE — but the limited role must be “explicitly defined”). This is classic “rob Peter to pay Paul.” The politicians directing these funding changes seem to have a very poor understanding of exactly how broad STEM is, how the STEM fields all depend and build on one another, and the degrees of research maturity among students (undergraduate, graduate, and postdoc).

  5. Charles Mire @ 2012-08-19 12:31

    And if I wanted to be more cynical as to why the PDF funding will be limited to one application/lifetime, it could be argued that the motive is to cut funding even further. The one-time-only clause will certainly reduce the number of annual applications. Then the Conservatives can say “well, fewer people are applying for postdocs, so there’s no need to keep funding up in that area.”

  6. Thanks Charles. As I asked Karel above, can you point to data showing the transfer of funds from Fellowships and Scholarships to CREATE? It can likely be extracted from the files here:!/data

    NSERC administrative staff (who are implementing the recent changes) are appointees, not politicians. My view is that NSERC staff should be strong scientists with long term vision. Canadian research policy needs to run smoothly over time scales that are longer than a typical government.

    Minister Goodyear seems to understand the importance of investment in science, and also speaks well of “blue sky” research. However, the recent policy changes by NSERC staff targeting short term commercialization don’t appear to be guided by that understanding.

  7. […] research and post-secondary education. Jim Colliander wonders what would happen if “Canada restricts athlete participation to one Olympic games per lifetime.” A rightfully exasperated David Kent exclaims from his academic exile, “Come on […]

  8. Karel Casteels @ 2012-08-20 13:45

    Hi James,

    Well, I realize the CMS has no actual power to change NSERC’s direction, especially these days. But it does seem like image is everything to this government, and so I was thinking that something even as simple as an official public statement of “concern” from the CMS (and hopefully other societies) could be effective. In any case, the CMS is in a much better position to have a dialogue with NSERC (and the universities themselves) than I can via blog comments.

    As for CREATE, all I (think I) know comes from NSERC releases. For example, in the preamble of the following document, they explain that there were less traditional awards this year due to reallocation to CREATE:

    I have been meaning to try and predict next year’s results by extrapolating data from that document and this one:

    I will try to add anything I (think I) figured out to Buzzdata if I get around to actually doing this.

    By the way, Charles, are you sure about a postdoc only having a “limited role” with CREATE? Everything I remember seeing about CREATE seemed to indicate postdocs funded through that program are postdocs in the usual sense. In my opinion CREATE has some fundamental flaws, but if you’re right, then it’s even crummier than I had thought!

  9. […] What if Canada restricted athletes to only participating in one Olympics in their lifetime? Because that’s essentially what NSERC has done to Post-Doctoral researchers. […]

  10. The Canadian Association of Postdoctoral Scholars has responded to NSERC regarding the recent changes to the postdocs competition. Here is the link:

Add a comment