Congressional hearings can sometimes hide more than they reveal. So it was yesterday, when the research panel of the U.S. House of Representatives science committee held its first public airing of a bill that would make some controversial changes to peer review at the National Science Foundation (NSF).
One key change would require the NSF director to certify that every grant will achieve one or more of six national goals—including strengthening the U.S. economy, bolstering national defense, increasing partnerships between academia and industry, and training the next generation of scientists. The director would also have to post a description of each pending award before it is made, along with the names of the relevant program managers who made the decision.
A casual observer might easily have come away from the 95-minute hearing with the impression that such language, drafted by the committee’s Republican majority, is in line with current NSF policy. The witnesses didn’t dwell on the language, and Representative Larry Bucshon (R-IN), who chairs the subcommittee, said that the proposed changes to NSF’s merit review system were “consistent with steps the NSF is already considering to improve accountability, which have been approved by the National Science Board.”
That’s not exactly true, says Dan Arvizu, chairman of the 24-member board, which sets policy for the $7 billion agency. The presidentially appointed board has yet to review the 96-page bill, Arvizu told ScienceInsider, much less sign off on any of its provisions.
Arvizu explained that he and acting NSF Director Cora Marrett did meet in September with Bucshon and Representative Lamar Smith (R-TX), who leads the full committee. “We embraced the idea of transparency, accountability, and continuous improvement,” says Arvizu, adding that Smith told him those concepts were also guiding principles for the committee. But Arvizu says that none of the specific language in the bill, dubbed the FIRST (Frontiers in Innovation, Research, Science, and Technology) Act, was even discussed.
Two weeks ago, when science lobbyists were given a 2-page description of the draft legislation, committee staffers repeated the message that the science board supported the changes. And after today’s briefing, a senior aide said that the board had told Smith at the September meeting that it had adopted new guidelines that were consistent with the FIRST Act.
But Arvizu takes issue with that description. What actually happened, he says, is that the board heard a presentation last month in closed session from Marrett and her aides about “NSF’s approach to implementing” the principles that had been discussed at the September meeting with Smith. The board has asked NSF officials “to implement those principles independent of any legislation” the committee might adopt, Arvizu added.
Although Arvizu declined to give details of NSF’s plan, he said it consisted of reinforcing existing procedures at the agency “both before and after an award is made.” Such steps are likely to include better documentation of how each award satisfies the agency’s two funding criteria—scientific merit and broader impacts—as well as a clearer description for the public of what the researcher hopes to accomplish.
A second false assumption from the hearing would be that the gentle criticism of those provisions from two former NSF senior officials means only minor tweaks are needed to satisfy the scientific community. In reality, the testimony of Richard Buckius, vice president for research at Purdue University in West Lafayette, Indiana, and Tim Killeen, vice chancellor for research for the State University of New York (SUNY) system, pointed to a fundamental misunderstanding in the bill of how NSF operates and, more broadly, the purpose of federally funded research.
Oddly enough, Bucshon hinted at the first problem in his opening question to Buckius. “Yes or no, does the proposed FIRST bill have what you might call any congressional interference in the peer review mechanism for evaluating grants at NSF?” he wondered.
A professor of mechanical engineering who led NSF’s engineering directorate for 3 years before going to Purdue in 2008, Buckius immediately asked for some wiggle room. “Can I do yes and no?” he replied. But then he gamely plunged ahead.
“My first point is who would actually affirm the awards that go out,” Buckius began. “The bill says the director should. But I don’t believe anyone is all-knowing enough to affirm the 11,000+ awards made every year. So I’d change that.”
Buckius then offered up a more substantive criticism, suggesting that posting pending grant winners would spur widespread second-guessing of reviewers. “The bill talks about the prior announcement of awards before they are made,” he began. “In some of our competitions, we only fund a few out of a hundred proposals. So that means there will be 90+ folks who, prior to the award, can energize the system and create chaos. So the system could become extremely bogged down.”
After the hearing, Killeen also amplified Buckius’s concerns about posting advance information about an award. “It could lead to the worse type of grandstanding,” he told ScienceInsider, “with scientists making their case in public for why NSF should have funded their research. It could become a real free-for-all.”
Arvizu was also disturbed by the provision. “I don’t believe NSF has ever published information about an award before it was made,” he said. “I’m not even sure how that would work. It certainly was not part of our discussion [with Smith].”
For Killeen, who served as head of NSF’s geosciences directorate for 4 years before joining SUNY in 2012, a bigger problem with the legislation is how it regards basic research. “If I have a concern, it’s mostly the message that this bill will send out to the world, in fact. As my testimony indicated, I hope it’s a vibrant, enthusiastic, ‘let’s take on the 21st century,’ United States can-do kind of [legislation] rather than one that seeks to find the constraints and stiffen the sinews. My personal experience with NSF is that it is a magnificent national asset. We don’t want to throttle it back.”
The list of goals in the legislation also troubles him. “I do worry about every proposal having to conform to a specified set of criteria,” Killeen said. “That’s why I used the word vibrant in my testimony. Scientific inquiry includes following leads that may not take you anywhere, and setting out hypotheses. But yes, the overall portfolio should address all of those statements.”
What’s next? Bucshon said in his opening remarks that “members and staff on both sides of the aisle are continuing to work on the final legislation.” But he did not say when the subcommittee planned to mark up the bill, and he handed off the gavel to Representative David Schweikert (R-AZ) and left before the hearing ended.
The status of the bill could be affected by a private sitdown next week between the science board and Smith, which will take place during the board’s regular 2-day meeting at the foundation. Arvizu said it would be a chance for the board and the science committee’s chairman “to understand one another, for the board to say, ‘Here’s what we are planning to do,’ and to ensure him that we are moving in a productive manner to address his concerns.”
In a statement this morning, Smith said that “I look forward to continued discussions with the National Science Board next week on this draft legislation and ways to boost R&D funding.” The statement clarifies Bucshon’s comments by explaining that the draft legislation “is consistent with the plan approved by the National Science Board that requires NSF staff to provide clear justifications for why grants are worthy of American taxpayer funds.”