For more than a generation, astronomers have been waiting for the James Webb Space Telescope, NASA’s successor to the iconic Hubble Space Telescope. On Tuesday they learned they will have to wait even longer, as agency officials revealed Webb’s launch date has slipped from spring of 2019 to approximately May 2020—a delay that could cost hundreds of millions of dollars, breaching the $8.8-billion telescope’s Congressionally mandated cost cap and requiring legislators to either provide more funding or abandon the project.

The delay also jeopardizes the next decadal survey, a once-every-10-years strategic plan produced by the National Academies that assesses the research landscape and makes recommendations to federal agencies and Congress about developing and funding future generations of ground- and space-based telescopes. Originally scheduled for completion after Webb had launched and begun science operations, the survey’s next iteration may instead now occur while Webb, arguably its most important consideration, is still on the ground.

The slip is not exactly surprising, even though construction and testing of Webb’s primary mirror and scientific instruments—its riskiest, most expensive elements—is already complete. These components were delivered in early February to Webb’s prime contractor, the aerospace company Northrop Grumman, for further testing and integration with the rest of the telescope. But later that month a report from the Government Accountability Office warned the company had fallen behind schedule on the supposedly easier parts of the observatory. Valves on the spacecraft’s thrusters had sprung leaks after being improperly cleaned, and replacing them had taken the better part of a year. Webb’s tennis-court-size, five-layer folding “sunshield” had also been torn during a test as it unfurled, requiring time-consuming failure analyses and repairs.

“Webb is the highest priority project for the agency’s science mission directorate, and the largest international space science project in U.S. history,” said acting NASA Administrator Robert Lightfoot, who is retiring from the agency next month. “All the observatory’s flight hardware is now complete—however, the issues brought to light with the spacecraft element are prompting us to take the necessary steps to refocus our efforts on the completion of this ambitious and complex observatory.”

Tim Paynter, a spokesman for Northrop Grumman, says the company “remains steadfast in its commitment to NASA and ensuring successful integration, launch and deployment of the James Webb Space Telescope, the world’s most advanced space telescope.”

Northrop Grumman now reportedly has teams working on the telescope 24 hours a day, and they will soon be joined by additional personnel from NASA Goddard Space Flight Center, which manages the Webb project. Still, the setbacks have deferred Northrop Grumman’s most crucial final tests, which expose Webb’s components to a harsh vacuum like that of space and jostle the fully assembled telescope in a simulated rocket launch. “It is taking longer than anybody thought a year ago,” says John Mather, an astrophysicist and Webb’s senior project scientist at NASA Goddard. “But we have to do it right. We have built just about everything. We just have to make sure it will work way up there. No taking chances.”

Conceived nearly three decades ago in the aftermath of Hubble’s debut, Webb (like its predecessor) has long been intended as a technological tour de force that would make the next giant leap in our understanding of the universe. Unlike Hubble, which resides in low Earth orbit and has been kept alive by repeated servicing missions, Webb would be stationed beyond the orbit of the moon and cooled nearly to absolute zero by its sunshield. Its mission would last until it runs out of fuel in five to 10 years. From its lofty perch Webb’s sophisticated infrared instruments and giant 6.5-meter mirror would see what Hubble never could: the universe’s very first stars and galaxies, embryonic star systems mid-formation in cosmic wombs of gas and dust, and the atmospheres of planets orbiting other stars. For astronomers these horizon-expanding views are so vital for progress that Webb has become the keystone for their field’s long-term future: Take it away, and all plans built around it collapse.

“The dominance of a single mission like Webb can be a bad thing,” says astrophysicist Martin Elvis, a noted critic of budget-busting space missions at the Harvard–Smithsonian Center for Astrophysics. “It fosters a ‘too big to fail’ syndrome, where because it mustn’t fail you can’t realistically threaten it with cancellation, and people err on the side of caution to ensure it will succeed. More caution means more testing and more money, which you must provide to avoid failure, and so you get a feedback loop of inflating costs.” And Webb could still end up blown to smithereens or at the bottom of the ocean due to a launch failure, or drifting uselessly in space due to snags in its carefully choreographed deployment sequence. Rocketry may be a science, but it remains decidedly imperfect.

As early as the 2000 decadal survey, U.S. astronomers ranked Webb as their highest priority for the first 10 years of the new millennium. The telescope’s projected cost was then roughly a billion dollars. Yet by the next survey in 2010 the project had already spent $3.5 billion on developing a host of necessary and expensive new technologies. Webb had effectively become too big to fail, and the 2010 Decadal Survey had no choice but to embrace it. It also recommended several other projects—chief among them the Wide-Field Infrared Survey Telescope (WFIRST)—to overlap with and complement Webb, the field’s crown jewel, which was then planned to launch around 2014.

Shortly after the 2010 Decadal Survey, however, independent reviews found the Webb project in crisis. It was burning through its budget so fast that crucial work to build the telescope was being pushed back, causing future costs to balloon for lack of immediate funds. NASA and Congress intervened, resuscitating the project with a “re-plan” that gave it a larger budget capped at $8 billion (plus $837 million in operational costs once in space) and a launch date of October 2018. But in September 2017 NASA announced Webb’s launch would be delayed until spring 2019, due to Northrop Grumman’s mounting woes and scheduling problems with European launch provider Arianespace.

Prompted by Webb’s damaged thrusters and sunshield, the latest delay to May 2020 is based on the findings of another independent review helmed by NASA’s Paul McConnaughey, an associate director of the agency’s Marshall Space Flight Center. Officials involved with the project believe that if this delay breaches Webb’s cost cap, it will likely only do so by at most a few percentage points. Even so, the pressure for Webb to meet its new launch date while incurring minimal additional costs is so great that NASA has taken the unusual step of forming a second independent review, led by former NASA Goddard head Thomas Young, to determine if Webb can meet its latest launch date without breaching its cost cap.

Testifying to Congress in December 2017, Young offered what some insiders see as a ringing endorsement of the project—and what others less charitably view as a blank check for Northrop Grumman and “too big to fail” syndrome: “[Webb] is at a point in its development where the only criterion that is important is mission success,” Young said. “Every appropriate thing that can be done to maximize the probability of success should be done. At this stage of the project, a few extra days or weeks or even months of schedule delay or the expenditure of some additional dollars is a small price to pay to assure success of a mission as important as [Webb].”

Meanwhile, WFIRST—NASA’s planned successor to Webb and the highest priority of the 2010 Decadal Survey—has experienced tumult of its own, suffering moderate delays and cost growth. It drew the ire of the Trump administration, which proposed canceling it entirely in the White House’s FY 2019 budget; Congress and the space science community have so far stood firm against that proposal. Anticipating budgets that will be effectively flat for the foreseeable future, NASA officials have stated any overruns on Webb are likely to be paid for out of WFIRST’s coffers. That in turn would lead to further delays and cost growth for the latter mission, potentially preventing it from operating contemporaneously with Webb—one of the original justifications for its existence.

David Spergel, an astrophysicist at Princeton University and co-chair of the WFIRST science team, acknowledges the trade is probably worth making. “Webb is one of the most complex engineering projects ever done and the most complicated science mission ever built by NASA,” he says. “We are all disappointed; so many of us are eagerly awaiting Webb’s science. However, it is better to have a modest delay and a modest overrun (as a fraction of Webb’s cost) and have a successful mission.”

Other influential astronomers begrudgingly agree. “You’re never happy when any mission slips, because it means the next thing will slip inevitably, too,” says Marcia Rieke, an astrophysicist at the University of Arizona. Rieke is the principal investigator of Webb’s NIRCam instrument, as well as the co-chair of the National Academies’ Committee on Astronomy and Astrophysics, a body that along with NASA, the National Science Foundation and the National Academies’ Space Studies Board helps plan the decadal surveys.

The troubles surrounding Webb and WFIRST, Rieke says, could delay the next such survey, which would at minimum throw off the timing for those of other fields—particularly the next decadal survey planned for planetary science. Or the survey for astronomy might continue on schedule without knowing whether Webb will be successful, particularly if it treats the troubled telescopes as first and foremost a cautionary tale against similarly ambitious future missions. “The decadal is meant to be a measure of what’s actually achievable,” Rieke says. “And so it walks a fine line between pushing the envelope and being feasible.”

In other words, the easiest solution might be for the next decadal survey to simply proceed as planned and to recommend that NASA pursue a portfolio of smaller, cheaper space telescopes rather than more multibillion-dollar behemoth in the 2020s. That prospect sends chills down the spines of some astronomers, who hope to build even bigger space telescopes using the new technologies developed at such great cost for Webb. They hope these bigger projects might be able to directly image Earth-like planets around nearby stars in search of signs of life.

“My fear now is that the community will be so frightened of cost that they won’t recommend any large telescope in the next decadal,” says one senior astronomer, who asked not to be named due to the politically sensitive nature of the situation. “That would delay starting anything big until 2030. It would probably be 2050 before something ambitious could be built and launched. This idea that money is transferable between mission sizes is just wrong. If NASA doesn’t pursue another big strategic mission after Webb and WFIRST, there’s no guarantee Congress will put all that ‘saved’ money into smaller missions. More likely it will go into rockets to put people on the moon or to some program outside of NASA entirely. If we’re not careful, this could lead to the end of the golden age of U.S. space astronomy.”