French funding panel quits in protest

Blaise Pascal, the famous French mathematician from the 17th century.

French funding panel quits in protest

The entire 20-member panel tasked with evaluating mathematics and theoretical informatics proposals for the French National Research Agency (ANR) in Paris has resigned to protest the low success rate in the discipline. The group has declined to release its verdicts on ANR's mathematics proposals for 2016; unless a solution is found soon, this could mean the agency won't fund any fundamental mathematical research at all this year.

Before 2014, thanks to the so-called Programme Blanc, an ANR program that tended to reward more theoretical disciplines, the panel had more leeway in selecting proposals, with its success rate typically reaching 30%. But after the government merged most ANR programs into one generic call for proposals, the success rate has been standardized across all disciplines, including mathematics. In a context of funding cuts, this has meant that the success rate has fallen below 10% for the last 2 years.

The panel, which has the broad support of the mathematics community, contends that the low rate has increasingly discouraged mathematicians from applying and has hurt their field. ANR counters that mathematics had a slight but unfair advantage over the more applied sciences earlier, and says it is balancing the scales.

The panel “steps down en masse to protest against the confiscation of scientific choices by an entirely administrative management process,” the panel wrote in a public statement on 6 June. Christophe Besse, a mathematician at Paul Sabatier University in Toulouse, France, who chairs the panel, argues that mathematics and theoretical informatics are different from the more applied sciences, and the panel should have more freedom to choose how many proposals to fund within a predetermined budget.

Previously, with the Programme Blanc, the panel could reduce the size of individual awards so it could “select a higher number” of proposals for the same money “and thus apply a much higher success rate,” Besse explains. The mathematical community tends to submit small proposals, he adds, as “we can make very good fundamental science with small projects.” But for 2016, the panel has received about 100 prescreened proposals, down from last year by 20%. Assuming a success rate of 10%, that means only 10 proposals will be funded, down from 40 in 2010—a "ludicrous situation," Besse says.

On 6 June, participants to the first congress of the Mathematics Society of France (SMF) in Tours issued a statement supporting the panel’s decision. A few days later SMF, the French Society of Statistics, the Society of Applied and Industrial Mathematics, and the Informatics Society of France wrote a joint letter in which they deplored “a vicious cycle that endangers the vitality of our communities.” The four societies want ANR to be more transparent about how it sets its success rates and fully involve the panels in such decisions.

ANR President Michael Matlosz held a meeting with the panel on 6 June, says Nicolas Ehrbar, ANR’s director for information and communication. Besse’s conclusion of the meeting is that ANR was “more or less in agreement with our assessment of the situation, but they can’t do anything for us.” The panel, which is now trying to take its plight to the ministry responsible for research, has until early July to release its classification of the proposals to the agency so funds can still be awarded on time.

Ehrbar, who currently estimates a success rate of 11% across all disciplines, expects a swift resolution of the conflict, because receiving no funding at all would be an even worse outcome for French mathematics. The injection of an extra €65 million into the ANR budget this month, which the government promised in May in a pledge to increase the number of supported projects in all disciplines, may help ease the tension. The government’s desire to level the playing field by replacing the Programme Blanc “has slightly eroded the advantage [mathematicians] had in comparison to others,” Ehrbar concedes, but going back to the old system is not on the table.