A “Successful” NSERC Discovery Grant applicant replies to Isabelle Blain

Here is the reply of fellow mathematician, Greg Martin, to Isabelle Blain’s letter to the 2011 Discovery Grants applicants in mathematics and statistics (See below). There is also the open letter of Frithjof Lutscher.

Dear Ms. Blain:

I received a letter from NSERC last month. It said, in huge block letters: “SHUT DOWN ALL TRAINING OF HIGHLY QUALIFIED PERSONNEL IMMEDIATELY.”

A few weeks later I received another letter with a little more information. It said: “Your application has been judged to be very strong in all aspects. In particular, your very strong existing programme of training highly qualified personnel was an important aspect of your application. NOW QUICK FOR THE LOVE OF GOD SHUT DOWN ALL TRAINING OF HIGHLY QUALIFIED PERSONNEL IMMEDIATELY.”

You might be surprised to learn that this is what your letters said, but the message was as clear as it could be. Corresponding to the expectations of my profession, and in line with the evaluation criteria stated by NSERC, I have built a thriving programme of graduate students and undergraduate research students. In addition to my three graduate students who will be continuing next year, I have accepted two new MSc students; moreover, I finally have the chance to be the primary supervisor of a postdoctoral fellow.

Suddenly, I received the news of my miserably low Discovery Grant. Allow me to do the math for you.
* Discovery Grant: $15k for five years, 2011-2016. Total funding available: $75k.
* Three graduate students, RA support for two years each: $48k.
* Two graduate students, RA support for one year each: $16k.
* One postdoctoral fellow, salary support: about $20k plus benefits per year. Support from the postdoc’s secondary supervisors might cover half that; let us say $11k for me as the primary supervisor.
* Total committed funds, based only on the HQP I have right now: $48k+$16k+$11k = $75k.

In other words, NSERC’s decision has the following inexorable message:
– I should not accept any HQP under my supervision, not a single one, for the next five years. –

(Let’s not even discuss the fact that NSERC believes it has adequately supported my personal research program. The remaining $0k over five years for my own research won’t get me to many conferences to disseminate my research. And HQP themselves should be able to attend conferences as part of their training – no funding for that either. Oh dear, we haven’t even begun to consider undergraduate research….)

At this point you will be tempted to think that this email is simply “sour grapes” for not doing as well in the Discovery Grant competition as I would have liked. On the contrary, my application was rated VVV, which is every bit as good as I could have realistically hoped. In fact, last year’s VVV applicants received grants of $24k per year, which is in fact a reasonable amount on which to base a programme of training HQP.

You will also be tempted to quibble about the exact numbers in the above calculation, to equivocate about the exact parameters of my expectation for an HQP programme … in short, you will be tempted to dismiss this one case as an anomaly. I assure you that, while I am communicating my own situation to you in detail, my story is playing out over and over in mathematics departments all over Canada. I am not the only one who received these letters from NSERC last month. Dozens of deeply productive mathematicians, in departments in every province, opened their letters only to receive the message “SHUT DOWN ALL TRAINING OF HIGHLY QUALIFIED PERSONNEL IMMEDIATELY.”

You spoke multiple times about “fairness” in your email. By this I can only assume you mean that applicants with the same ratings received the same amount of money. Pardon me, I must amend this: I can only assume you mean that applicants, in the same discipline, with the same ratings, this year only, received the same amount of money; despite the fact that costs for HQP are extremely similar across many scientific disciplines, statistics and other disciplines were given substantially more funding than mathematics for the same ratings. (Let us also not dwell on the reality that for many other disciplines, Discovery Grants are merely supplements to their other funding sources; in mathematics, if you can’t pay for HQP with a Discovery Grant, you can’t pay for them all all.)

Evident to any observer, however, are multiple ways in which the results of the recent competition are anything but “fair”.
* Depending on what year you apply for a grant (a factor not under your control), your funding could differ up to 40%. For example, last year’s OVV applicants received commensurate grants of $30k annually; this year’s OVV applicants received inadequate grants of $18k. A difference of $60,000 in grant money, to applicants rated exactly the same by the panel and who differed only by applying one year apart, cannot possibly be called “fair”.
* Despite NSERC’s stated commitment to evaluating applications independently of past awards, you implemented a policy of “protecting the purchasing power” of the grants of applicants who received an E in the research criterion. In other words, for applicants in this category, funding was decided solely on the basis of their existing grants – not exactly a “fair” process.
* The professional researchers who volunteered their time as members of NSERC Evaluation Committees slaved for hours and hours, painstakingly debating whether applicants should receive S’s or V’s in the various categories. Once they had completed their exhausting work – work they donated generously to the granting system – they found that bin-to-funding decisions made by non-academics rendered their effort almost completely moot. All those hours deciding on ratings, only to find that the entire set of bins between SSS and VVV were all given grants that fell beneath last year’s VSS bin amount of $15k? All those hours deciding on ratings, and the end result might be a difference between $10k (SSS) and $11k (VSS) – in other words, a difference of 1/8 of a graduate student? Certainly not “fair” to these professionals’ time either.

In your email you retreat to the statement “[T]he Executive Committee and NSERC believe that the 2011 competition outcomes were the fairest possible under very difficult circumstances.” You compare success rates in this year’s competition to last year’s, as if that were the point of the criticism to which you are obviously responding. I propose that you need to focus your attention upon the outcome of this year’s Discovery Grants competition:

* increased inequality between mathematics and other scientific disciplines, and increased inequality between mathematicians with comparably rated applications; and

In short, the outcome of this year’s NSERC Discovery Grants competition is a failure: failure to achieve the stated goals of the programme. All the reported percentages and protestations of “fairness” in the world cannot disguise this clear appraisal.

In fact, I only wish we could deem the competition a mere failure, due to its mortal blow against the training of HQP in mathematics departments across Canada. But professional mathematical researchers are expected to be able to travel to conferences, invite research visitors, supervise graduate students – all activities for which their Discovery Grants are rapidly becoming inadequate – as part of their research programmes. It is no stretch at all to say that an additional result of this year’s competition is to cause researchers in Canadian mathematics departments to wonder whether their only chance of being supported in these expected activities is to find employment in another country. In this respect, the message to shut down HQP training, combined with the reality of driving mathematicians (present and potential) away from Canada, makes this competition not a mere failure, but a crushing failure.

It is possible that you are aware of this, and that your email to which I am responding is a dissembling effort to support your employer’s decisions as a matter of loyalty. It is possible that your attempt to describe this year’s competition as anything other than a crushing failure for Canadian mathematics is, in your eyes, a necessary evil.

However, I cannot help but wonder whether it is possible that you are not aware that the actual outcome, independent of NSERC’s intentions or aspirations – that the actual outcome of this year’s NSERC Discovery Grants competition was indeed a crushing failure. If it were indeed the case that you were not aware of this crushing failure, then the only adjective I could think to describe that state of mind is “deluded”.

Sincerely yours,

Greg Martin
“Successful” NSERC Discovery Grant applicant
Associate Professor
Department of Mathematics
University of British Columbia

On May 6, 2011, at 11:32 AM, Z-COORD wrote:

(français ci-dessous)


Dear Applicant:

NSERC has just published the preliminary statistics for the 2011 Discovery Grants competition. This note will provide you with some additional context on how decisions on funding were reached.

Globally, the success rate in the Discovery Grants program remains strong: the success rate for established researchers applying to renew a grant increased from 72% in 2010 to 74% in 2011. The success rate for researchers not holding a grant also increased, from 29% in 2010 to 33% in 2011.

In the mathematics and statistics group this year, the success rate for established researchers was 79 %, for early career researchers, 68%, and for returning established ‘unfunded’ researchers, 38%. The average grant for each of these categories of researchers is lower than in the 2009 and 2010 competitions.

There were two significant challenges facing the Mathematics and Statistics Evaluation Group’s Executive Committee in making its funding recommendation to NSERC.

1. The Executive Committee had to balance success rate and grant values. It aimed to fund all applications to the quality category (bin) J for established researchers and K for Early Career Researchers (Note: an application in category J requires three ratings of “Strong” or their equivalent). To realize this, the values associated with most of the bins had to be set at lower levels than in previous years. The alternative, which was not recommended by the Executive Committee, was to have maintained grant levels relative to 2009 and 2010 at the expense of the number of applications funded.

2. The Executive Committee noted that the distribution of applications to the bins varied significantly resulting from differences in the interpretation of merit indicators between Mathematics and Statistics. Consequently, it recommended that measures be taken to ensure an equitable treatment for all applications. After in-depth discussions, it was agreed to implement tailored funding levels for Mathematics and Statistics.

In reaching these conclusions, the Executive Committee and NSERC believe that the 2011 competition outcomes were the fairest possible under very difficult circumstances.

NSERC is committed to working with all Evaluation Groups to keep improving the peer review process and to ensure that the communities understand how it works. Adjustments have already been made in response to the first two competitions that used the processes. NSERC will continue to refine the system.

NSERC is aware that the mathematics and statistics communities have for some time indicated that the amount of funding dedicated to mathematics and statistics is too low.Current levels are based largely on historical values, and were influenced by past reallocation exercises. We have requested an assessment to be conducted by the Council of Canadian Academies (CCA) on performance indicators for basic research. NSERC intends to adopt a new methodology once the assessment report provides indicators, metrics and suggestions on how to do so. The CCA report is expected to be released in mid 2012. NSERC will incorporate feedback from the research community and other sources when identifying the best mechanisms and indicators to use in future decisions about budget allocations.

For additional information please consult the question and answer section of the Discovery Grants Information Centre.

Isabelle Blain
Vice President, Research Grants and Scholarships

This entry was posted in Op-eds. Bookmark the permalink.

4 Responses to A “Successful” NSERC Discovery Grant applicant replies to Isabelle Blain

  1. Claire says:

    I agree with everything that Canada’s mathematicians are saying about the NSERC Discovery grant – with one exception. Despite the smaller numbers of dollars that mathematicians get they are no worse off financially than the lab-based scientists. We have to spend money on the supplies and equipment that our students need to do their experiments. Our slightly larger grants don’t begin to cover this. NSERC has said for years that our Discovery grants are “grants-in-aid” and encouraged us to apply elsewhere. But it is no secret that the minute one gets a grant from another agency NSERC funding decreases or disappears at the next competition.

    NSERC is a disgrace. The only upside of the current mess is that those who have always been NSERC apologists will finally change their tune as they, too, suffer under the new system. I say this as someone who has held NSERC funding since 1989.

  2. Claire, thanks for contributing to this conversation! Do you have any data (or suggestions on how to collect it) which supports your claim that funding outside of the DG program eventually causes a decrease in DG funding level?

    The report of the international review claims:

    “Award holders typically obtain 40% to 60% further funding from other NSERC programs to supplement their Discovery Grants – and the larger the grant, the larger tends to be the ‘leverage ratio’. Thus excellence is amplified.” (p2; see also figure 13 on p30)

    “Discovery Grants are also often supplemented substantially by non-NSERC sources – e.g., a recent on-line survey conducted by NSERC showed that DGP awards are leveraged overall about 2:1. It should be noted that different disciplines have substantially different abilities to access non-DGP and (especially) non-NSERC funding. For example; researchers in the biological sciences often have access to significant additional funding from the CIHR and from foundations that support research on individual diseases. For other disciplines (e.g., pure mathematics), such additional access is difficult or not available at all. For these disciplines the Discovery Grant is not a “grant-in-aid” – it is the only grant.”

    • Claire says:

      Unfortunately most of my evidence is anecdotal. Colleagues of mine who have gone up for NSERC renewal after getting grants from MRC (as was), CIHR, NIH, Genome Canada, Gates Foundation, etc invariably lost their NSERC grants – even when they made it absolutely clear that their NSERC proposal was on a separate topic. Panels simply assume that the investigator “doesn’t need the money”. This is understandable when there are limited funds to go around. However, grants from other agencies are usually one-off, and often short duration.
      During my time as a department chair, and later a dean, I have always advised new faculty to get their NSERC grants first before applying to other agencies, otherwise getting the NSERC grant will be difficult to obtain.
      Until recently at least, NSERC funding represented a stable source of funding for a lab over the course of a career, allowing continuous training of graduate students, undergraduate research students, co-op students etc. Now NSERC grants are just as sporadic as the rest.

  3. Pingback: Term limits and the integrity of the peer review process | Piece of Mind

Leave a Reply

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

WordPress.com Logo

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

Facebook photo

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

Connecting to %s