Computing Research Policy Blog

NSF Funding Outlook Grim But Cyberinfrastructure's a Priority, says NSF Director

On Tuesday, National Science Foundation Director Arden Bement met with the Coalition for National Science Funding (of which CRA is a member) and warned the science community CNSF represents to lower expectations of increased funding for the agency in the near-term, saying the expectation of budget-doubling, as authorized by Congress and the President in 2002, "isn't reasonable."

"The NSF budget is correlated with the other non-defense research agencies," Bement said, "and those are not scheduled to grow very much [in the President's future budgets]." The Administration's priorities are very clear, Bement said: secure the homeland, fight the war on terror, and reduce the budget deficit.

In light of the tough times ahead, Bement said the 3 percent increase (to $5.745 billion for FY05) in NSF funding requested by the President for should be seen as a symbolic show of support for the agency, especially as other agencies are seeing their budgets stay flat or decline in relation to inflation.

Given the relatively bleak outlook, Bement said the agency would pay special attention to three areas in the FY05 and FY06 budgets: improving productivity of researchers by increasing average grant size and duration; strengthening the national research infrastructure by investing in cyberinfrastructure research and development; and strengthening NSF management.

It was encouraging to hear Bement talk specifically of cyberinfrastructure in his remarks, especially as there was/is some concern in the computing community after the departure of Rita Colwell from NSF that the new NSF regime just might not "get" cyberinfrastructure. Bement, I think, is saying the right things, noting that research today is increasingly complex and "network-oriented" and that a crucial part of the enterprise is reliant on cyberinfrastructure. For FY 05, Bement said, NSF would spend ~$400 million on cyberinfrastructure related R&D foundation-wide and that funding would go to research beyond just "boxes and network" -- research into algorithm development, the architecture of the net, software, etc.

The other two priority areas -- increasing grant size and duration, and strengthening NSF management -- are not particularly new challenges for the agency. Bement says he hopes to see average grant size grow beyond $142,000 (as it will be in FY05) to $160,000, and hopes that increasing grant duration will lead to proposals with more risk. He also noted that he's growing concerned with the success rate of proposals dropping significantly, citing the increased number of proposals the agency receives (over 40,000 last year, compared to an average of 22-33k previously) and theorizing that other agencies decline in budgets and changes in funding regimes (see here and here for more) may be to blame for the increased pressure on NSF.

Bement also indicated he's seeking $20 million in FY 05 for more staff at NSF. "We need more program officers," he said, noting that proposals are now more complex than ever before, the administration of the process of peer-reviewing them is more complex, and there are a whole lot more of them to consider. "While the opportunities have never been greater," he said, "the number of claimants has never been larger."

It's not clear how long Bement will remain in the driver's seat at NSF. His appointment lasts only until the end of the year, so his job status is probably heavily dependent upon what happens in the elections in November. The grapevine, which had been chattering pretty loudly in March about a possible candidate for the permanent job (well, the six year term), seems to have quieted down considerably. It seems as though Bement will remain in charge at least through November.