The glass-and-concrete headquarters of the National Science Foundation in Arlington, Virginia, normally hosts scientists who decide the fate of fellow researchers' grant proposals. But in a nondescript spare office on the 12th floor, new players have set up shop: congressional aides reviewing the merits of scientific studies conducted with government funding.
The two aides are evaluating the scientific merit of research proposals submitted to the the $7-billion-per-year agency, the nation's biggest funder of basic science initiatives. They've selected several dozen federal science grants for special scrutiny, in a move that critics say reflects a conservative political agenda at work. Among these are a climate change education project, archaeology studies in Ethiopia, anthropology work in Argentina, and others dating back to 2005.
The aides, who have been at the NSF since August, have begun a review process that critics say threatens to topple a long-standing wall at the agency between science and politics. The new process reflects an escalating debate between scientists and politicians on Capitol Hill over how much of a say Congress should have in the scientific enterprise.
In recent years, that debate has included skirmishes over appointments at the U.S. Department of Agriculture, the role of science at the Environmental Protection Agency under various administrations, and, indeed, the conduct of the committee that's investigating the NSF—the House Committee on Science, Space, and Technology.
Representative Lamar Smith, the Texas Republican who chairs that panel, says the review of NSF activity is routine "congressional oversight," done to make sure the federal government is not wasting money. He would not say why his committee has targeted several dozen specific grants out of 12,000 projects the NSF supports annually.
"The NSF must be held accountable for its funding decisions," Smith told National Geographic last week, two months into an investigation that he says will take a year. His mission, he says, is to ensure the agency requires every grant it gives is in the "national interest."
In May his committee passed a bill that would dictate a 40 percent cut to social science research at the NSF—think sociology, anthropology, and psychology, among other fields—so that more can be spent on engineering, math, and other so-called "hard sciences." Smith calls those areas "the highest priority research."
Congressional oversight of the NSF's inner workings may seem like the nerdiest of inside-the-Beltway disputes, and some speculate that Smith is simply trying to score points in the media by highlighting research that sounds silly on paper. ("Oppression and Mental Health in Nepal" is one project he's mocked.)
But since the National Science Foundation opened in 1950—and even before—politicians have attached strings to science funding. Early legislation stipulated officials avoid "undue" geographic concentration as they gave out money.
Now, though, lawmakers are getting involved in deciding the merits of individual grants to scientists. And so an enduring question has gained newfound importance: How should political considerations affect decisions around basic research?
Rules dictating the terms of the closely monitored visits to the Arlington office were forged after months of wrangling between the agency and the Republican lawmakers who control the House science committee.
"You want a balance between political and pure scientific influence in making decisions about science policy—you want a tension between the two," says David Goldston, former chief of staff of the House science committee. But depending on whom you ask, Smith's efforts to steer NSF policy are either an appropriate exercise of congressional prerogative or an overreach that will politicize a process that should be left to the science community.
"Chairman Smith wants to prioritize areas of science he personally prefers," says Wendy Naus, who leads a coalition of social-science groups in Washington opposing Smith's plan. Instead, she says, the NSF's expert staff and the scientists who volunteer to review grant proposals should decide the priorities for basic science.
At Issue: "National Interest"
Smith's plan would strengthen congressional control over the NSF in two ways. First, lawmakers have until now usually funded the agency with what is more or less a single appropriation for the whole foundation, allowing the foundation to decide how to divvy up funds among various "directorates" that fund the geosciences, social sciences, and so on. Smith's bill would authorize specific funding levels for the directorates, which Congress has done only a handful of times in the 64-year history of the foundation. Opponents say that will tie the foundation's hands.
Possibly more intrusive, say critics, is Smith's proposal to require the NSF to show, in writing, that every research grant it funds is in the "national interest." That requirement, Smith wrote in a recent op-ed, will help the NSF "cut out wasteful spending and fund high-quality research."
Many scientific organizations have opposed this measure, include the NSF's own governing body, the National Science Board.
Defining just what constitutes the national interest is tricky business. The bill says it means that the research would increase "economic competitiveness," support health or defense, or promote "the progress of science."
The NSF already requires grantees to show both the "intellectual merit" and the "broader impact" of their proposed research. Policy expert Daniel Sarewitz of Arizona State University testified last year that the "national interest" requirement would just add a new "meaningless level of rubber-stamping to the grant."
The NSF's governing science board, composed mostly of academics, published a letter in April that broke with its long-standing tradition of not commenting on pending legislation. "We ... do not see a need to impose new, more inflexible, legislated requirements on NSF... We are concerned that the proposed new legislative requirements might discourage visionary proposals or transformative science," they wrote.
One thing is clear, says Goldston: Tensions over the role of politics in science decisions have "been baked in from the start at NSF."
On Basic Science, Fundamental Differences
The NSF's grantmakers have had to contend with politics for decades. In the 1950s, at the height of McCarthyism, the young foundation announced it would not support avowed communists. And beginning in 1975, Senator William Proxmire, a Wisconsin Democrat, ridiculed funny-sounding research projects with his "Golden Fleece" awards, adding to pressure on the NSF to focus on applied problems, like urban pollution or energy production, at the expense of basic science.
Before the NSF was founded, the government funded science by awarding contracts to researchers who agreed to deliver certain results. The NSF, instead, was to give out open-ended grants, lest bureaucrats restrict scientific flights of fancy. Yet from day one, officials had to consider more than pure scientific merit, including the wide geographic distribution of funds stipulated by Congress.
The NSF's procedure for evaluating proposed basic research projects has become the global gold standard. At its core are scientist-bureaucrat program managers who convene panels of peer reviewers to examine and rate grant proposals from the community.
"The best and brightest ideas, according to the best and brightest experts" is how NSF describes its review system. "No free passes. No good-old-boy network." And, it notes, the NSF merit review process is confidential between reviewers and the agency, allowing colleagues to rate each others' proposals anonymously.
But some good old boys have sought to have their say. In 1975, House lawmakers, led by conservative Republicans, passed legislation that would allow any individual grant proposal, once approved by the NSF's process, to be vetoed via House or Senate vote.
"It raised serious challenges to the whole process of making funding decisions through peer review," says NSF historian Marc Rothenberg. The measure was eventually removed during legislative negotiations.
The latest NSF investigation will set a dangerous new precedent, some academics say, as applicants and reviewers will question whether the process will remain confidential.
"The U.S. is the standard bearer when it comes to peer review," says Harvard University sociologist Michèle Lamont, who has studied that process. "This kind of interference is clearly putting us on a slippery slope." She notes that "it is not unusual" in France for high-level politicians to try to sway decisions on basic funding with a phone call.
It's not a fight that shows any signs of resolution. Assuming Republicans maintain control of the House in midterm elections next week—a virtual certainty—the scrutiny of specific NSF grants will "continue until NSF agrees to only award grants that are in the national interest," Chairman Smith told Science magazine in September.
And if Republicans take the Senate—an increasingly likely proposition—this latest round of skirmishes between lawmakers and scientists is likely to escalate.
Correction: The original version of the article stated that no authorization bill has ever specified how the NSF should divvy up its funding among directorates. Several such bills have in fact been passed. It also said that legislation giving Congress new control of the NSF passed in 1976. It passed in 1975.