Four times this past summer, in a spare room on the top floor of the headquarters of the National Science Foundation (NSF) outside of Washington, D.C., two congressional staffers spent hours poring over material relating to 20 research projects that NSF has funded over the past decade. Each folder contained confidential information that included the initial application, reviewer comments on its merit, correspondence between program officers and principal investigators, and any other information that had helped NSF decide to fund the project.

The visits from the staffers, who work for the U.S. House of Representatives committee that oversees NSF, were an unprecedented—and some say bizarre—intrusion into the much admired process that NSF has used for more than 60 years to award research grants. Unlike the experts who have made that system work so well, however, the congressional staffers weren’t really there to judge the scientific merits of each proposal. But that wasn’t their intent.

The Republican aides were looking for anything that Representative Lamar Smith (R–TX), their boss as chair of the House Committee on Science, Space, and Technology, could use to support his ongoing campaign to demonstrate how the $7 billion research agency is “wasting” taxpayer dollars on frivolous or low-priority projects, particularly in the social sciences. The Democratic staffers wanted to make sure that their boss, Representative Eddie Bernice Johnson (D–TX), the panel’s senior Democrat, knew enough about each grant to rebut any criticism that Smith might levy against the research.

The peculiar exercise is part of a long-running and bitter battle that is pitting Smith and many of his panel’s Republican members against Johnson and the panel’s Democrats, NSF’s leadership, and the academic research community. There’s no end in sight: The visits are expected to continue into the fall, because NSF has acceded—after some resistance—to Smith’s request to make available information on an additional 30 awards. (Click here to see a spreadsheet of the requested grants.)

And the feud appears to be escalating. This week, Johnson wrote to Smith accusing him “of go[ing] after specific peer-reviewed grants simply because the Chairman personally does not believe them to be of high value.”  (Click here to see a PDF of Johnson’s letter and related correspondence from Smith and NSF.)

Smith, however, argues he is simply taking seriously Congress’s oversight responsibility. And he promises to stay the course: “Our efforts will continue until NSF agrees to only award grants that are in the national interest,” he wrote in a 2 October e-mail to ScienceInsider.

Ask, answered

How did things get to this point? For the past 18 months, Smith has waged a very public assault on NSF’s storied peer-review system. He’s issued a barrage of press releases that ridicule specific awards, championed legislation that would alter NSF’s peer-review system and slash funding for the social science programs that have supported much of the research he has questioned, and berated NSF officials for providing what he considers to be inadequate explanations of their funding decisions.

NSF has defended itself at congressional hearings, in personal meetings with committee staff and the chair, and with a stream of letters and e-mails. White House officials, university leaders, and Democratic legislators have joined the fray, roundly criticizing Smith for what they see as an attempt to impose his political judgment on a process that draws upon the wisdom of scientific experts. But that nearly universal condemnation hasn’t stopped Smith, who was first elected to Congress in 1986 and last year was named chairman of the science committee.

Smith describes his growing frustration with NSF in a 27 August letter to NSF Director France Córdova. (The committee made this and another letter available to ScienceInsider.) Smith notes that he first asked for materials relating to several grants in the spring of 2013, soon after Cora Marrett became acting NSF director following Subra Suresh’s resignation to become president of Carnegie Mellon University.

But after being rebuffed by Marrett, Smith writes that he “set aside the request … until a permanent NSF director was installed.” Córdova was confirmed by the Senate this past March, and on 7 April Smith wrote her a letter containing a list of 20 grants that he wanted to examine.

Smith’s request created a major dilemma for NSF. On the one hand, Córdova knew that Congress has the authority to obtain information as part of its job to oversee the actions of federal agencies, a right that federal courts have repeatedly upheld. On the other hand, NSF constantly assures scientists that every aspect of the peer-review process will remain confidential. (NSF’s website contains abstracts of projects it has funded, and the public can obtain a copy of a successful application. NSF does not share any information about, or even acknowledge the existence of, proposals that have been rejected.)

Smith wanted the material shipped to his offices on Capitol Hill. But Córdova made a counteroffer that the Texas legislator grudgingly accepted. First, the committee staff could see everything related to the grant except for the names of the reviewers, which would be redacted. Second, the material would remain at NSF headquarters in Arlington, Virginia. Third, the staff could take copious notes, but none of the information could be photocopied or otherwise reproduced.

Judy Gan, head of public and legislators affairs at NSF, says the arrangement “preserves the integrity of the merit review process.” Even so, NSF officials have sent letters to the president of each university with a grant on Smith’s hit list, hoping to reassure them that everything is under control. NSF had no choice but to comply with the committee’s request, the letters explain. But NSF chose to tell each institution about the request “so that you may take appropriate action to inform your principal investigator and other potentially impacted parties about this production of documents.”

In many cases, NSF staffers had already sounded the alarm. Steven Folmar, a cultural anthropologist at Wake Forest University in Winston-Salem, North Carolina, recalls getting a call from his program manager last month alerting him to the science committee’s pending review of his 2012 grant, titled “Oppression and Mental Health in Nepal.” The 3-year, $160,000 award supported him and two colleagues in a study of how social status affects the mental health of Nepalese adolescents. Folmar has worked on and off in Nepal since 1979, and he says the country’s economic and cultural divisions are so striking that it’s an ideal place to measure the impact of discrimination on those in the lowest caste.

Folmar says that his first reaction after hearing that his grant had been singled out was to hunker down and keep quiet. “I felt like somebody in a war movie, with bullets whizzing over my head.” But after further reflection, he thinks that speaking up may not be such a bad idea.

“I’d tell [Smith] that our work has a great deal of relevance to this country,” he says. Measuring how social inequality can cause depression and anxiety is valuable information for U.S. public health officials, too, he explains, noting that some Nepalese victims display symptoms akin to post-traumatic stress disorder.

The project was a bargain, he adds. The grant covered several months of field work by three senior researchers and their graduate students, he notes, “all for about $50,000 a year. That’s pretty cheap science.”

Parsing the list

The scientific community is scratching its head over how Smith compiled his list of questionable grants. Many have also been flagged by other legislators, notably Senator Tom Coburn (R–OK), who issue annual lists of what they consider to be wasteful government spending. Research grants often appear on such lists. Decades ago, former Senator William Proxmire (D–WI) created what he called the Golden Fleece Awards to poke fun at such supposed boondoggles. In fact, the practice has become so widespread that 3 years ago a coalition of scientific organizations created a counterpoint, called the Golden Goose Awards, which honors federally funded basic research that later turned out to have huge societal benefits.

But Proxmire’s awards were never meant to fundamentally alter NSF’s peer-review system, according to Folmar. “This sounds like Golden Fleece with a much more dangerous twist,” he says.

Smith so far has asked to take a look at 50 grants. (Note: ScienceInsider was able to identify just 47 unique awards.) And the list is hard to characterize. One grant goes back to 2005, and 13 appear to have expired. The total amount of money awarded is about $26 million. The smallest grant, awarded in 2005, is $19,684 for a doctoral dissertation on “culture, change & chronic stress in lowland Bolivia.” The largest, for $5.65 million, is for a project that aims to use innovative education methods to educate Arctic communities about climate change and related issues.

More than half of the grants appear to involve work outside the United States. The largest number—29—were funded through NSF’s social, behavioral, and economic (SBE) sciences directorate. Of those, 21 came from SBE’s behavioral and cognitive sciences division, including a number of grants in archaeology and anthropology. But six of NSF’s seven directorates also funded grants on Smith’s hit list.

What the science committee expects to learn from its investigation is a burning question from scientists. A committee representative declined to answer repeated queries about the criteria used to select the grants. In his written statement to ScienceInsider, Smith said only that “there are many grants that no taxpayer would consider in the national interest, or worthy of how their hard-earned dollars should be spent. … The public deserves an explanation for why the NSF has spent hundreds of thousands of dollars on musicals about climate change, bicycle designs, and a video game that allows users to relive prom night.”

Mont Hubbard is the “bicycle designs” grantee on Smith’s intended list of shame. An emeritus professor of mechanical engineering at the University of California, Davis, Hubbard received $300,000 in 2009 to study the feedback system that allows humans to control a vehicle, in this case a bicycle. And Hubbard has a ready answer to Smith’s question about how his research could possibly serve the national interest.

“It’s easy to learn to ride a bicycle, but it’s hard to explain how we do it,” Hubbard says. His broader research into operator control of mechanical systems has applications across many areas, he explains. Substitute “pilot” for “rider” and “airplane” for “bicycle,” he says, and it’s clear that helping humans do a better job of manipulating machines has the potential to greatly improve performance, reduce safety risks, and promote economic growth.

Present stalemate

What’s next? So far, neither side has shown any signs of backing down. In his 27 August letter to Córdova, Smith declares that “the current review work is 5% complete, which implies that this oversight initiative will span at least 12 months.” He accuses her of reneging on a promise to provide the committee with everything it requested and speculates that she “may be banking on a cumbersome, time-consuming federal court process” to back her up. That approach puts NSF “in an indefensible position,” he says, predicting that such tactics will ultimately fail and that NSF will be forced to give in to his demands.

In her reply 2 weeks later, Córdova denies withholding any pertinent information. “To the contrary,” she writes, “NSF has provided the Committee full and complete access to our files for each of the grants of interest.” She disagrees with his assertion that “NSF does not trust the Committee.” But she acknowledges that “we are balancing this access with the need to preserve the trust of the scientific community, whose participation in the merit review process occurs in a confidential environment.”

With such strong rhetoric on both sides, it’s hard to see a quick or quiet ending to this confrontation. Johnson certainly seems prepared to continue defending NSF and, in particular, its funding of the social sciences. “This campaign against NSF’s merit-review system is indefensible absent some compelling explanation of what you are trying to accomplish,” she tells Smith in her 30 September letter. “If your ultimate goal is to cut funding for social and behavioral sciences …I respect your right to try to make that case as Chairman. But please do not compromise the integrity of NSF’s merit review system as part of this campaign.”

WIth reporting by David Shultz.

Correction 3 October, 8:05 a.m.: Steven Folmar studies how social inequality can contribute to, not treat, depression.