Last weekend, representatives of Canada’s mathematical and statistical sciences community presented NSERC’s President, Suzanne Fortier, with a “Long Range Plan.” Entitled Solutions for a Complex Age, the report – commissioned by NSERC – is the result of two years of consultations and deliberations under the guidance of a prominent Canadian statistician, Nancy Reid. The aim was to establish a unified vision of priorities and directions for the development of research in the mathematical and statistical sciences in Canada. The writing is crisp, the presentation professional (Thanks Harriet Gorham), and the pages are glossy. On the surface, the report seems to timidly and cautiously embrace the status quo. There are however several new elements in the document that are worth discussing.
1. After describing, quite articulately I should say, how important and useful Maths and Stats (M&S) are, and how good their Canadian practitioners are becoming, the report gets to the heart of the matter: “Invest in Canada’s mathematical and statistical sciences research via Discovery Grant levels that reflect the importance of the research base to Canada, and that acknowledge that the costs of research are similar to those in several related science and engineering disciplines.”
And the numbers are clear on this one. M&S represent 9.2% of the Discovery Grants (DG), yet they receive 5.1% of the DG funds. The average grant in M&S in the 2010-11 competition (the year of the revolt) was almost half of the average in all fields ($16,816 to $33,691). The NSERC leadership that rightly challenged the status quo by decreeing “that the level of a grant should be commensurate with merit, regardless of the applicant’s granting history with NSERC” is far behind on the equally important issue of doing away with the bias regarding the “discipline’s granting history with NSERC.”
Indeed, almost a decade after the last community-driven re-allocation exercise, it looks like any recalibration among disciplines in the Discovery Grants program will have to wait till 2014 if not later. We have been hearing about such plans for almost three years now. The CCA Report on Science Performance and Research Funding was released on July 5, 2012. NSERC has now initiated another survey on the matter. In the meantime, the Council will carry on with the current outdated funding scheme.
2. The “thematic and collaborative resources” refer to the three mathematical science institutes, the Centre de Recherches Mathématiques in Montréal (CRM), the Fields institute in Toronto (Fields), and the Pacific Institute for the Mathematical Sciences in Western Canada (PIMS), as well as to the Banff International Research Station (BIRS) and the nascent Canadian Statistical Institute (CANSSI) (see below). These institutions are recognized as an integral and essential part of Canada’s research capacity in the mathematical sciences.
I find the description of the 3 institutes – obviously written by committee – to be too detailed and sometimes repetitive. It looks more like a grant application (a laundry list of what they have done) than a vision for the future. A bit too defensive to my taste, when they have no reason to be. The institutes’ contribution to the country’s research capacity is huge (see below) and it is hard to imagine Canada’s research landscape without them.
3. The description of BIRS is short and crisp. It surely reflects the Station’s focus, its justifiable confidence, and its status as an amazing “good deal” for Canada. After all, NSERC’s contribution to BIRS (25% of its budget) falls behind the ones by the US National Science Foundation (30%) and by the Government of Alberta (27%). We still hear some noise from certain quarters about how “luxurious the place is.” To which we answer: Yes, the mountain setting is absolutely glorious, but is not reflected in the cost of Corbett Hall, where even the most senior of scientists are willing to share bathrooms and otherwise, in exchange of a precious week of intense interactions and collaborations without the distractions and the prohibitive costs of downtown hotels.
4. The first real breakthrough of this exercise is the decision by the 3 institutes, CRM, Fields and PIMS to set aside some of their own funds to jumpstart the “Canadian Statistical Institute”. This is huge and unprecedented in so many ways. There is no doubt that the role of statistics and its applications is becoming an essential feature of modern societies, their economies, their policies, and their decision-making processes. Ask Nat Silver (and check on his detractors). Yet, NSERC had twice before denied funding for the National Institute for Complex Data Structures (NICDS), in spite of glowing reviews by peers. And in a commendable display of leadership, the 3 institutes have decided to do it on their own – at least to provide this hugely important and much needed Canadian initiative some seed funding from their own limited budget. The LRP was therefore able to announce the birth of the “Canadian Statistical Institute.” Not one minute too soon.
5. The only ask for new funding in the LRP is for the support of Innovation and Connections through the Mprime network. What are the arguments for this?
- Well, politicians, policy makers, university administrators, industry leaders and NSERC officials, all seem to agree that the math and stats communities should be as engaged in the country’s innovation agenda as any other scientific and engineering community.
- The MITACS (now Mprime) Network of Centres of Excellence (NCE) used to be the preferred vehicle for the math and stats communities to do just that. In 2011-12 alone, Mprime supported 377 Canadian academic scientists working with 739 students, in partnership with hundreds of sponsors in the private and public sectors. It also supported 32 two-year industrial postdoctoral fellowships across Canada. But the NCE funding for MITACS/Mprime (~$5.4M with ~$1.5M from industrial leveraging every year) ended as of March 31, 2012, having reached its 14-year horizon. So, what now? That is the question that the LRP tried to answer.
- NSERC’s Research Partnership Program (RPP) currently supports not one single Canadian mathematician or statistician, yet the RPP use up almost 45% of NSERC’s budget. One can argue about the reasons and the history behind this shortcoming, but no one can deny that the RP programs –old and new– have been designed to be ripe for the engineers’ picking and not for scientists. Mprime is offering a way to remedy this flaw in our current granting system. It is now up to the NSERC leadership to decide whether they want to fix it or not.
Now where are the other opportunities for growth in the research capacity of Canada’s mathematical and statistical communities? Besides the call to support Mprime through the RPP mentioned above, no additional funding is requested neither for the institutes nor for the Evaluation Group 1508 (for Math and Stats). NSERC had indeed decreed that no additional funding will be coming to “the envelope” anytime soon, but this shouldn’t have mattered when developing a long-term vision.
Suzanne Fortier dangled one carrot during the Montreal presentation. “Big Data” seems to be the theme of the next competition in the “Frontiers program”, and NSERC thinks that mathematicians and statisticians could be serious contenders in such a competition. Conflicting reports indicate however, that “this is all about genomics” and not about developing new mathematical and statistical tools for data mining in all kind of settings. Considering the history of collaboration –or lack of– between the statisticians and the –provincially and federally– well funded but allegedly insular genomics community, this carrot doesn’t look very appetizing.