Under the Microscope
WASHINGTON -- In times of fiscal crisis, scientific research is a ready target for budget cutters. Members of Congress grandly and publicly trash programs they see as wasteful, like government-funded research into the sex life of screwworms and the nervous systems of fruit flies, while advocates point out that such research sometimes turns out to be incredibly useful.
But there may be another, less obvious way to cut costs. The process the government uses to determine which projects are worthy of funding is itself expensive, drawn-out, and -- as some politicians argue and many researchers agree -- not as effective as it could be. As federal funding tightens, Congress needs to increase its oversight of the review process, Rep. Mo Brooks (R-Ala.), chairman of the House Science, Space and Technology Subcommittee on Research and Science Education, argued at a hearing Tuesday. "Given that those limited dollars should go to the very best scientific research," the National Science Foundation and other agencies "must maintain a robust and transparent merit review process," Brooks said.
The discussion at Tuesday's focused on ways to streamline the process of evaluating the tens of thousands of grant applications submitted each year. In 2010, for example, NSF received more than 55,000 grant proposals, which were whittled down to 13,000. About half of the NSF's $7 billion budget was used for those grants, as well as continuing grants approved in previous years, according to the subcommittee's charter for Tuesday's hearing. (Note: This article has been updated from an earlier version to correct some facts.)
The foundation and other federal agencies that support scientific research have a lot of money at stake when evaluating proposals, so a great deal of time and money goes into evaluation efforts. After being submitted to the NSF, grant proposals go through a peer review process, drawing on a community of more than 300,000 researchers to review the projects via mail correspondence and discussion panels.
Lawmakers on the subcommittee invited a panel of scientists and research administrators to discuss proposals for streamlining the review process. Those proposals include conducting some panels online (saving travel costs), weeding out weaker applications early in the process, and asking researchers conducting peer reviews to measure a project's contribution to achieving certain "national goals" outlined by the foundation. Each of these measures would save money, the witnesses agreed, but researchers worry they could stifle important research projects -- especially the projects whose utility isn't immediately obvious.
"For any research project, it could be impossible to see how it's related to national goals," said Keith Yamamoto, vice chancellor for research at the University of California at San Francisco and a witness at the hearing.
Yamamoto was speaking out against proposals by the NSF and the National Science Board to change the peer review criteria to measure a proposed project's contribution to national goals mentioned in the America COMPETES Act of 2010. Those goals include improving the United States' economic competitiveness, increasing national security, and increasing public scientific literacy.
He argued that the "national goal" criteria would ask researchers to make evaluations outside of their fields of expertise. More important, he said, many important discoveries are stumbled upon during smaller research projects.
But not all of the witnesses agreed with him. Nancy Jackson, president of the American Chemical Society, said that researchers in the peer review process often have to make tough decisions. In cases where the proposal is on the border, looking at it from a broader perspective -- how could this benefit society? -- would be a good tiebreaker.
"There are clearly far more proposals than can be funded," said Jackson, manager of the International Chemical Threat Reduction Department in the Global Security Center at Sandia National Laboratories.
Jackson was there to advocate reforms that would, in her words, "triage some of the weaker proposals." For every 100 applications, she said, about 40 are poorly written and ill-conceived, with little chance of being approved for funding. Those applications are still given the peer-review treatment, but Jackson said they need to be pre-screened by the agency to save resources.
"They all take the same amount of time in the review process," she said. "But the real agonizing choices are made with the remaining 40 or so."
Three of the four panelists warmed to the idea of holding more review panels online, which Rep. Daniel Lipinski (D-Ill.), the committee's senior Democrat, said "intrigued" him. Cora Marrett, the deputy director of NSF, said she's concerned about security, but having panel discussions online could save on travel expenses and allow more members to participate.
"We're looking into the idea and the possibility of expanding the field," she said.