Grants: Load, Deadlines, And Gamesmanship

The NSF conducted a very interesting experiment which halved the number of grant proposals (boldface mine):

Annual or semiannual grant deadlines lead to enormous spikes in submissions, which in turn cause headaches for the program managers who have to organize merit review panels. Now, one piece of the agency has found a potentially powerful new tool to flatten the spikes and cut the number of proposals: It can simply eliminate deadlines.

This week, at an NSF geosciences advisory committee meeting, Assistant Director for Geosciences Roger Wakimoto revealed the preliminary results from a pilot program that got rid of grant proposal deadlines in favor of an anytime submission. The numbers were staggering. Across four grant programs, proposals dropped by 59% after deadlines were eliminated. “We’ve found something that many programs around the foundation can use,” Wakimoto told the advisory committee on 13 April….

She eliminated the twice-a-year deadlines for four of her grant programs, in geobiology and low-temperature geochemistry, geomorphology and land-use dynamics, hydrological sciences, and sedimentary geology and paleobiology. NSF sent out a notice about the change at the beginning of 2015, and after a 3-month proposal hiatus, the no-deadline approach began in April 2015. The number of proposals plummeted, from 804 in 2014 to just 327 in the 11 months from April 2015 to March

Feedback from scientists has been good so far, Isern adds. In a field where many scientists do field work, having no deadline makes it easier for collaborators to schedule time when they can work on a proposal. “I think they like the flexibility,” she says. “They’re able to be more thoughtful about it.”

But I think this comment inadvertently gets at why proposals dropped:

Paul Bierman, a geologist at the University of Vermont in Burlington, says the move is an “incredibly good idea” and expects success rates to go up. In October 2015, he and two collaborators resubmitted a previously rejected proposal to the geomorphology program: a $265,000, 3-year request to study the thinning of glaciers that retreated from New England within the last 20,000 years. Bierman thought it would only take the three of them a month or so to revise their proposal, but the lack of a deadline allowed them to buff the proposal to a shine over the course of several months. The extra polish apparently paid off: He received a notice of recommendation for funding this week.

Often, a grant is submitted to ‘get it in the queue’ and to receive feedback. When a researcher is faced with semi-annual or annual deadlines, you have to submit something, no matter how good it is at the time. The pressure to submit is all the more critical when it will likely take two funding deadlines to be competitive. Depending on the agency, there’s an additional problem: feedback from the previous submission often arrives with very little time left before you have to resubmit. Removing this time pressure means researchers can submit ‘better’ proposals on a schedule that works for them.

There is, of course, a potential downside:

When proposals go down, and success rates go up, programs could be punished for having higher success rates than their peers. “One of the arguments that has been made for increasing budgets has been, ‘Look, we have such proposal pressure, give us more money,’” she says. The experiment provides evidence that proposal pressure can be easily manipulated, she says. “It’s not a good metric to use to decide whether a certain program deserves to have an augmented budget.”

Still, the NIH needs to try this, and sooner rather than later.

This entry was posted in Funding, NIH, NSF. Bookmark the permalink.