Long-range plan in the works

The latest Discovery Grants competition by the Natural Science and Engineering Research Council (NSERC) has sparked an outpouring of criticism from the mathematics research community that contends the system used to select grant recipients is broken. An April 26th letter signed by 140 researchers in mathematics and statistical sciences states that recent changes to the program threaten to reverse gains made over the past decade.

The letter calls on NSERC and Industry Canada to restore confidence in the peer review system and undertake "an immediate review of the newly implemented review system by an independent international panel to consider its impact on the Canadian mathematical sciences".

This isn't the first time the mathematics community has expressed concern over the new two-stage conference model of reviewing DG applications. The new system assesses scientific merit by placing applicants in so-called bins determined by how they score against three merit-based criteria, followed by the assigning of funding levels.

The math community expressed concern the first time the conference model and binning system was used in 2009 (R$, April 30/09). But this year those concerns have been amplified and disseminated throughout the blogosphere, underscored by the realization that years of modest increases to the NSERC funding envelope are starting to have an impact across the whole research community.

Complicating the situation is the inability of NSERC to release the full results of the 2011 competition, leaving only anecdotal information on the latest success rates and the success of specific disciplines. NSERC has informed researchers that will receive funding but full competition results won't be released until the election is over. NSERC officials declined to be interviewed for this story.

select results released

In a letter sent out to the broad research community, NSERC released some details including the number of applicants reviewed (3,482), the success rates and average funding levels of various applicant types: early career researchers receiving a first grant (54%/$22,481), established researchers renewing a grant (74%/$33.045) and established researchers who did not hold a grant (33%/$28,082).

Nevertheless, word is spreading that overall success rates continue to fall, leading to an increase in the number of researchers who received nothing or substantially less than in the past.

"We've been too polite in the past but this year there were several striking examples of internationally leading mathematicians who have seen their grants cut by up to 60%. We are now realizing how entrenched the problems are and how it needs to change," says Dr James Colliander, a researcher in the Univ of Toronto's department of mathematics and its associate chair of research. "We need to rebuild our peer review system. The first step is a recognition that there is a problem from (NSERC president) Suzanne Fortier and (Industry minister) Tony Clement and a review has to be done in consultation with the entire community."

While there is anecdotal evidence of concern in other disciplines such as chemistry and computer science, the loudest criticism is coming from the mathematics research community. Math has been particularly hard hit in recent DG competitions, with lower success rates and far smaller grant sizes than the average, leading to charges that math is "grossly underfunded by NSERC compared to other disciplines".

The letter from the mathematics community says the average DG for mathematics researchers is less than $20,000 compared to the overall average of $34,000, although both are used for the training of junior scholars. It recommends a one-time investment in grants awarded to mathematics and statistical sciences researchers "in order to restore fairness compared with the 2009 and 2010 results".

Average success rates for the DG program have also been in decline in recent years, dropping from a high of 87% in 2002 to 58% in 2010.

"It's a great shame. NSERC was once the envy of the world for research funding," says Dr Jeremy Richards, a mathematics research in the Univ of Alberta's department of earth and atmospheric sciences. "They're changing the tone of the Discovery Grants program. It used to be everybody got something. It was a low amount but it was something. Now the success rate is falling close to 60% … The conference model and related outcomes has confused things. No one knows how it works yet (and) no one has figured out how to live in a system where you get nothing. SSHRC (Social Sciences and Humanities Research Council) has been like this for a long time with success rates of less than 25%."

Compounding concern within the math research community is the rise of research funding outside of the peer review system. Many have cast a suspicious eye at the generous funding received by the Perimeter Institute for Theoretical Physics (PI) outside of peer review, with some alleging political motivation (R$, March 31 & April 18/11).

U of T's Colliander says it's disconcerting to see the average Canadian scientist receiving "12 times less research funding that the average scientist at the Perimeter Institute" when awards to PI scientists are not peer reviewed.

"People who are not scientists are making research funding decisions," says Colliander, whose current DG funding expires next year. "The health of universities needs to be maintained so there is a place where kids can come in 15 years ... We need an ecosystem for innovation to emerge and the ecosystem is not healthy."

U of A's Richards concurs: "The Discovery Grants program was never intended to produce Nobel Prize winners but to support researchers across the country. It's intended to fund a program of research-enabling grants but it's transitioning into something that is not helpful. There are lots of programs out there with lots of money for star researchers."

long-range math plan in development

Perhaps in response to sustained criticism, NSERC has asked the mathematics and statistical sciences research community in June/10 to come up with a five- to 10-year long-range plan. A steering committee chaired by U of T's Dr Nancy Reid has been established and submissions were due April 18th. "It's a good idea but the activity needs to be empowered with some authority," says Colliander. "How will the plan inform NSERC policy?"

R$