After 59 years of service, Jason, the famed science advisory group, was being fired, and it didn't know why. On 29 March, the exclusive and shadowy group of some 65 scientists received a letter from the Department of Defense (DOD) saying it had just over a month to pack up its files and wind down its affairs. "It was a total shock," said Ellen Williams, Jason's vice chair and a physicist at the University of Maryland in College Park. "I had no idea what the heck was going on."
The letter terminated Jason's contract with DOD's Office of the Under Secretary of Defense for Research and Engineering (USDR&E) in Arlington, Virginia, which was Jason's contractual home—the conduit through which it was paid for all of its government work. So, in effect, the letter killed off all of Jason's work for defense and nondefense agencies alike.
Just days away was the group's spring meeting in Washington, D.C., where members and government sponsors would refine the dozen or so problems Jason would tackle in San Diego, California, during members' summer leave from their campuses and labs. Jason had to keep functioning, even as it prepared to die. It told sponsors it was still planning to do the studies, and advised members to keep their calendars open but not sign summer leases. It made plans for an attenuated spring meeting reception: not the usual dinner, but meatball and spinach-feta appetizers and plastic cups at the cash bar.
Meanwhile, members hurriedly wrote emails and made urgent phone calls, looking for other contractual homes. Then, on 25 April, the night before the reception, came a reprieve. Williams and Jason's chair, Russell Hemley, a materials chemist at the University of Illinois in Chicago, heard from the Department of Energy's (DOE's) National Nuclear Security Administration (NNSA), which for decades had commissioned Jason to study the health of the U.S. nuclear weapons stockpile. Now, NNSA said it couldn't afford a gap in its studies and pledged to pick up the Jason contract, at least until January 2020.
At the reception, in an auditorium at MITRE Corporation, Jason's administrator in McLean, Virginia, Jason members appeared relieved by NNSA's decision, although what went wrong at DOD was unclear. "The department remains committed to seeking independent technical advice and review," a USDR&E spokesperson said in a statement. "This change is in keeping with this commitment while making the most economic sense for the department." Mike Griffin, a former NASA administrator who heads USDR&E, declined to speak to Science about the dismissal.
The reprieve leaves Jason without a long-term home—and still facing an existential question: Can a group created during the Cold War's nuclear and missile races, when the U.S. government was keenly aware it needed scientific advice, survive today? Times and national problems have changed. The government employs many more of its own scientists and has many options for getting scientists' advice. "If Jason didn't exist, who would create it? Maybe nobody," says Henry Abarbanel, a physicist at the University of California (UC), San Diego, and a Jason since 1975.
But the group has always had a plan for survival. It actively self-renews—between two and five young scientists join Jason every year—and it is diversifying its customer base. Traditionally, Jason did national security studies for DOD, DOE, and the intelligence agencies. In the past 5 years, though, it has ramped up its nondefense studies and now works with the Department of Health and Human Services (HHS), the Census Bureau, the National Science Foundation, and the U.S. Department of Agriculture (USDA).
Catherine Woteki was USDA's chief scientist in 2016 when she commissioned a Jason study on ways to remotely detect crop yields. She thinks the group is as relevant as ever. "Science agencies need external advice," she says. "And especially in times like this, where science advice is perhaps undervalued by the public, it's more important that the science agencies get things right."
Many of Jason's customers seem to agree. Even after its near-death experience, Jason had 13 studies to work on, starting the morning after the reception. Pragmatic as always, its members left the reception a half-hour before it was set to end.
A mythical beginning
Jason was created in 1960 by a group of physicists who had summers off and were familiar with government consulting. They also had prestige: Eleven early Jasons—including Charles Townes, Murray Gell-Mann, and Burton Richter—eventually won Nobel Prizes. Their main customer was DOD's Defense Advanced Research Projects Agency (DARPA), which originally dubbed them Project Sunrise—a name that seemed presumptuous to them. So, inspired by Mildred Goldberger, wife of one of the founding members, they renamed themselves in honor of the mythical Jason, leader of the Argonauts.
The name change was a small but telling example of the group's independence. "I used to tell sponsors from the get-go," says Roy Schwitters, a physicist at the University of Texas in Austin (UT Austin) and Jason's head from 2005 to 2011, "that we tell people things they might not want to know."
That independence sometimes gets the group in trouble. In 2002, the long relationship with DARPA dissolved when the agency tried to nominate members for Jason. Within months, the group used its connections to find a new home within DOD: USDR&E. Such dust-ups aren't surprising. Political appointees don't always want independent advice, says Albert Teich, a science policy expert at George Washington University in Washington, D.C., and former head of science policy at AAAS (publisher of Science). "They want the advice that supports the positions they've already carved out for themselves," he says. As a result, Teich says, study requests come mostly from civil servants, who are more likely to know their agencies' technical problems.
In Jason's early decades those problems were physics-related defense questions, like how to detect the infrared signals of an enemy's missile launch or decipher the seismic signals of an underground nuclear weapon test. In an early study for the Navy, Jason devised a communications system for nuclear submarines, first called Bassoon, that bounced low-frequency radio signals off the ionosphere and into the oceans. It operated from 1989 until 2004, when the Navy declared it an unnecessary Cold War system.
During the Vietnam War, Jason designed a forerunner to the electronic battlefield: an anti-infiltration barrier that linked hidden acoustic and seismic sensors on the ground to bombers and artillery. In the mid-1980s, the group invented a way for telescopes to detect and compensate for the jitters caused by atmospheric turbulence, by using a laser to create an artificial guide "star"—a glowing spot high in the atmosphere. The technology, intended for tracking satellites and missiles, remained classified until 1991, when lobbying by Jasons helped convince the Air Force to open it up to astronomers. In 1989, the group reviewed the Star Wars antimissile program called Brilliant Pebbles, judging it technologically unsound; the program was canceled in 1993. In 1995, Jason's study on what could be learned from small nuclear tests—not much—helped convince then–DOD Secretary William Perry to recommend that the United States sign the Comprehensive Nuclear Test Ban Treaty. (The Senate, however, refused to ratify it.)
With the end of the Vietnam and Cold wars, Jason members began to branch out from physics and engineering. In 1977, they did their first assessment of global climate models and later advised DOE on which atmospheric measurements were most critical for the models. Since the mid-1990s, Jason has studied biotechnologies, including techniques for detecting biological weapons.
Membership has shifted along with the subjects under study. Many first generation Jasons have died or retired. (Notable exceptions are Freeman Dyson, 95, emeritus faculty at the Institute for Advanced Study in Princeton, New Jersey, and Richard Garwin, 91, retired from IBM, both still regularly involved as senior advisers.)
Jasons are, as they always have been, selected only by other Jasons. The criteria for selection include intellectual brilliance—maybe a third are members of the National Academies of Sciences, Engineering, and Medicine—in fields that are more or less matched to the current range of studies. As a result, though half of the Jasons are still physicists, the group now includes statisticians, mathematicians, computer scientists, chemists, oceanographers, geologists, atmospheric scientists, materials scientists, aeronautical engineers, and electrical engineers, along with what William Press of UT Austin, a former Jason chair, calls a "critical mass with esprit" of biologists and biochemists.
Contributing to Jason is one of the most important things I do.
It is also no longer just a boys' club. Jason took 23 years to invite its first female scientist: Claire Max, now the director of the UC Observatories in Santa Cruz. Today, women are being invited to join at higher rates: Nine of the 23 Jasons who have joined since 2010 are women. Many take on the leadership of studies. "Contributing to Jason is one of the most important things I do," says Sallie Keller, a statistician at the University of Virginia in Charlottesville who joined in 2007 and led Census Bureau and HHS studies.
They're reportedly paid $1200 per day—a goodly amount but less than what many of them could make as industry consultants. Upward of half of their studies are classified—perhaps explaining why their membership list is not public and why they prefer not to name other Jasons. They're generally allergic to publicity: Six of the 17 Jasons approached for this story refused interviews.
But they are not allergic to work. They spend some or all of 6 weeks, every summer, at General Atomics, a defense and energy contractor in San Diego. They take over the second floor of a secure building, a hallway lined with small offices, two or three people per office. "It looks cheap and it is," Schwitters says. "We bump into each other when we roll back in our chairs." Jasons say the forced proximity favors interdisciplinary cooperation.
The working culture is no-nonsense. There's no recognition or prizes for the work. It's get the job done, says Douglas Finkbeiner (no relation to the author), an astrophysicist at Harvard University who joined in 2014. "It's this brutal efficiency," he says, "like, ‘We need this info, I'll email’—‘No don't email, get them on the phone, now.’" A draft report is critiqued in real-time by other Jasons and sent to the sponsor by 1 October.
In the past 5 years, the range of studies Jason has done for nondefense agencies has broadened. HHS, for instance, has sponsored Jason only since 2013. The first of its three studies for the agency proposed an information systems architecture that would allow electronic health records to be operable across all health systems. In response, HHS formed a Jason Task Force that helped implement the report's recommendations through something called the Argonaut Project. "The health community has a unique sense of humor," says Teresa Zayas-Cabán, chief scientist at the Office of the National Coordinator for Health Information Technology at HHS in Washington, D.C.
The next HHS studies, in 2014 and 2017, were broader. One was about how data not in electronic health records—environmental data, data from health apps and fitness devices, social media data—could be used to improve personal health without threatening privacy. The other, Keller says, studied how to apply artificial intelligence (AI) to health, given the problems of uneven data quality and opaque, irreproducible AI models. Zayas-Cabán says one reason she likes Jason is the group's independence. The field of health care has "many powerful and entrenched interests," she says, "so independent and expert study of our issues can be extremely valuable."
The Census Bureau commissioned Jason after John Thompson, the bureau's director at the time, happened to sit on a national committee with Keller. "You got these brainy people to look at our problems," Thompson says, "and the price wasn't that much, it was a buy." The 2020 census will be the first to be conducted mainly online, which could open up new avenues for fraud. Jason's first report, in 2015, said the threat to the integrity of the census would come less from individual mischief than from large, organized attempts at fraud. Jason's second report, on the 2030 census, suggested that 90% of the population could be located simply by combining, for instance, Social Security and Internal Revenue Service data.
In 2016, USDA asked Jason about using data from remote sensors for midsummer estimates of fall crop yields, instead of relying on farmers' reports. Woteki, now at Iowa State University in Ames, says one reason she commissioned Jason was that it knew about DOD's remote-sensing capabilities.
The impact of Jason reports is sometimes hard to assess. Many sponsors talk in generalities about their value: For instance, John Abowd, the Census Bureau's chief scientist in Suitland, Maryland, says only that Jason's report on the 2030 census "has been incorporated into our thinking."
Part of the reason for that fog is that a finished report belongs to the sponsor, who can implement all, some, or none of it, and can publish it or keep it private. For sponsors, owning the report is self-protective. Jason can say your program "is stupid," says Gerald Epstein, who is now at the National Defense University in Washington, D.C., but worked loosely with Jason when he was at the Department of Homeland Security, "but they don't have to tell your boss or the rest of the world."
What happens when Jason's contract with NNSA expires in 2020 is unclear. One possibility is yet another home within DOD: This month, the U.S. House of Representatives added a line to DOD's preliminary budget directing the Office of the Under Secretary of Defense for Acquisition and Sustainment to pick up Jason's contract.
Many people say the government ought to find a way to tap into Jason's blunt advice. "Yes, there's a continuing need," says John Holdren, a physicist at Harvard University's Kennedy School of Government who was former President Barack Obama's science adviser. "And yes, Jason should be kept alive to help meet that need." Abowd thinks Jason has lasted this long "because of its reputation for not being manipulable," he says. "You can't stack the deck."
But if nobody picks up the contract, Williams says, "is Jason revocable or irrevocable?"
"I don't want to go there," she says. "For now, I'm feeling optimistic. But I really don't know."