Every ten years, the United States astronomy community does something remarkable and brutal: it votes on itself. The National Academies of Sciences, Engineering, and Medicine convenes hundreds of researchers, agency officials, and instrument builders to produce a ranked priority list for the next decade of astronomy funding. The result — the Astronomy and Astrophysics Decadal Survey — is the closest thing Big Science has to a democratic constitution. It is also one of the most contentious documents in American research policy, a battlefield where careers are made, telescopes are killed, and billion-dollar bets are placed on science that won’t pay off for a generation.
The 2021 survey, Pathways to Discovery in Astronomy and Astrophysics for the 2020s — known in the community as “Astro2020” — made headlines when it recommended a new large infrared/optical/ultraviolet space telescope as the first flagship mission in a Great Observatories program, a concept NASA later framed as the Habitable Worlds Observatory (HWO), with costs expected in the roughly $10 billion to $11 billion class and substantial uncertainty. But the headline obscured the real story: a years-long institutional fight over who controls American astronomy’s agenda, how NASA and the National Science Foundation respond when a community hands them a wish list, and what happens when the community’s wishes collide with fiscal reality.
The Mechanism: Panels, White Papers, and the Politics of Consensus
The Decadal Survey process begins roughly two years before publication. A steering committee is appointed by the National Academies, and the broader community is invited to submit “science white papers” — short advocacy documents arguing for particular research priorities. In the 2020 cycle leading to Astro2020, more than 900 white papers were submitted, a record. The sheer volume is a measure of the stakes: a project that makes it into the survey’s top tier has a reasonable shot at funding. A project that doesn’t appear may wait another decade.
The survey’s steering committee for Astro2020 was co-chaired by Fiona Harrison of Caltech and Robert Kennicutt of the University of Arizona. Their job was to synthesize competing scientific visions — exoplanet atmospheres versus gravitational wave follow-up, space telescopes versus ground-based arrays — into a coherent ranked list that NASA, NSF, and the Department of Energy would actually use. “The survey is not a wish list,” Harrison said at a 2021 National Academies briefing. “It is a prioritized program that we believe is executable given realistic budget assumptions.”
That word — “realistic” — does a lot of work. The survey’s budget assumptions have historically been optimistic, a fact that has caused cascading problems for every major project that followed.
The JWST Shadow
No Decadal Survey in the modern era can be understood without reference to the James Webb Space Telescope. JWST was the top priority of the 2001 survey, Astronomy and Astrophysics in the New Millennium, which estimated its cost at roughly $1 billion. By the time Astro2020 was being written, JWST had consumed more than $10 billion and had not yet launched. The overrun — documented exhaustively in NASA Inspector General reports and a scathing 2010 independent review chaired by John Casani — was the original sin haunting every subsequent discussion of flagship missions.
The Casani review, delivered to NASA in November 2010, found that JWST was “not executable” at its then-projected budget of $5 billion and schedule of a 2014 launch. The project was restructured, the launch date pushed to 2021 (it ultimately launched December 25, 2021), and the total cost ballooned to $9.7 billion. Congress nearly cancelled it in 2011, when the House Appropriations subcommittee on Commerce, Justice, and Science voted to terminate the program outright. It survived only after an intense lobbying campaign by the astronomical community and the intervention of senators from Maryland, home to the Space Telescope Science Institute.
The lesson the community drew from JWST — or should have drawn — was that flagship missions require independent cost validation before they are formally prioritized. Astro2020 attempted to institutionalize this lesson by recommending that NASA commission independent cost and technical evaluations (ICTEs) for any flagship mission before Phase A begins. Whether NASA follows that recommendation is a different question.
The Habitable Worlds Observatory: A $10 Billion-Class Bet on Life
The centerpiece of Astro2020 is the large ultraviolet/optical/infrared space telescope that NASA later framed as the Habitable Worlds Observatory, designed primarily to directly image Earth-like planets in the habitable zones of nearby stars and search for biosignatures in their atmospheres. The science case is compelling. The political case is complicated.
HWO emerged from the ashes of two competing concept studies: LUVOIR (Large UV Optical IR Surveyor), a NASA-commissioned study that produced two variants — LUVOIR-A at 15 meters and LUVOIR-B at 8 meters — and HabEx (Habitable Exoplanet Observatory), a 4-meter telescope with a separate starshade. Both studies were completed in 2019 and submitted to Astro2020’s panel on large space missions. The survey ultimately recommended neither directly, instead synthesizing elements of both into a large space telescope in the 6-meter class with a technology development roadmap that would need to run through the 2020s before a formal mission start.
The estimated cost — in the roughly $10 billion to $11 billion class, with significant uncertainty — immediately drew skepticism. “We’ve heard ‘$1 billion’ before,” one senior NASA official told me privately in 2022, declining to be named because the agency had not yet formally responded to the survey. “The community knows what happened with Webb. The question is whether they’ve actually internalized it.”
The survey’s answer was to build in what it called a “Great Observatories Mission and Technology Maturation Program” — a pre-Phase A funding line of roughly $500 million to $1 billion over the decade to mature the technologies (coronagraphs, ultra-stable structures, detector systems) that the future observatory would require. This was a genuine institutional innovation: rather than committing to a mission before the technology is ready, the survey recommended funding the technology first. Whether NASA’s budget environment in the mid-2020s allows for that kind of patient investment remains, as of this writing, an open question.
The Ground Side: ELTs, the ngVLA, and the Rubin Ramp-Up
Space missions dominate the headlines, but the Decadal Survey also sets priorities for ground-based astronomy, where NSF is the primary funder. Astro2020 gave very high priority to both the U.S. Extremely Large Telescope Program and the next-generation Very Large Array (ngVLA), with the ELT program generally treated as the nearer-term first ground priority if staging was required. The ngVLA is a radio telescope array that would succeed the iconic VLA in New Mexico with a continental-scale network of 263 antennas spanning from Hawaii to the Virgin Islands. The estimated cost: $2.3 billion, with NSF expected to provide the majority.
The ngVLA had been in development at the National Radio Astronomy Observatory (NRAO) since roughly 2015, when then-NRAO Director Tony Beasley began assembling a science case. By the time Astro2020 published, the project had a detailed design and a community of supporters — but also a rival. The Square Kilometre Array (SKA), a multinational project headquartered in the UK with array sites in South Africa and Australia, was already under construction and promised many of the same science capabilities. Some astronomers argued that NSF should contribute to SKA rather than build a separate American array.
The survey came down clearly for ngVLA, arguing that its frequency coverage (1 GHz to 116 GHz) and its focus on the northern sky complemented rather than duplicated SKA. But the funding question remained unresolved. NSF’s annual budget for astronomy is roughly $250 million; a $2.3 billion construction project requires either a multi-year appropriation or a novel funding mechanism. As of 2024, NSF had not secured a construction start for ngVLA, and the project remained in design development.
Meanwhile, the Vera C. Rubin Observatory — formerly known as the Large Synoptic Survey Telescope (LSST), renamed in 2020 in honor of the dark matter pioneer — was completing construction on Cerro Pachón in Chile after being the top ground priority of the 2010 survey. Rubin released public first-look/first-light milestones in 2025, with survey operations following after schedule delays and cost growth from initial plans. Its science operations, funded jointly by NSF and the Department of Energy, represent exactly the kind of long-duration, high-data-volume astronomy that Astro2020 built its priorities around.
The NSF Tension: When the Community Wants More Than the Budget Allows
The structural tension in every Decadal Survey is simple: the community recommends; Congress appropriates; NASA and NSF execute. The survey has no legal force. It is, in the words of former NSF Director France Córdova, “the community’s best collective judgment about where the science leads.” But that judgment is rendered in a political vacuum, and the gap between the survey’s ambitions and actual appropriations has widened with every cycle.
In 2022, NASA’s Science Mission Directorate received $7.8 billion — a record, but one with significant portions already committed to JWST operations, the Mars Sample Return campaign (itself facing major cost growth, with projected lifecycle estimates eventually rising into the $8 billion to $11 billion range), and the ongoing costs of missions already in operation. The bandwidth for new starts was narrow. NSF’s astronomy budget, roughly flat in real terms since 2010, offered even less room.
The result is a familiar institutional pathology: the community produces a survey that is, in Harrison’s word, “executable” — but only under budget assumptions that Congress has never actually delivered. Projects are prioritized, technology development programs are funded at a fraction of the recommended level, and the flagship mission recommended in one decade gets built, if it gets built at all, in the next. JWST was the top priority of the 2001 survey and launched in 2021. If HWO follows a similar trajectory, it might launch in the 2040s — by which time a new Decadal Survey will have been written, possibly recommending something else entirely.
The Priority Dispute Nobody Talks About: Multi-Messenger Astronomy
One of the quieter fights in Astro2020 was over multi-messenger astronomy — the combination of gravitational wave, neutrino, and electromagnetic observations that produced the landmark GW170817 neutron star merger detection in August 2017. The LIGO-Virgo detection, announced that October with a simultaneous gamma-ray burst detection by the Fermi satellite, was widely called the dawn of a new era. The community’s question for Astro2020 was: what infrastructure do you build to capitalize on it?
The answer was complicated by the fact that LIGO is funded by NSF, Virgo is European, and the proposed next-generation detector — Cosmic Explorer, a 40-kilometer arm-length interferometer that would detect mergers across virtually the entire observable universe — would cost an estimated $1.5 billion to $2 billion and require a site that does not yet exist. Astro2020 endorsed Cosmic Explorer as a priority but did not rank it above HWO or ngVLA. The gravitational wave community, which had spent years building the science case through the LIGO Scientific Collaboration, was not entirely pleased.
“There’s a real tension between the traditional astronomy community and the gravitational wave community over how the survey weights different kinds of science,” said one member of the Astro2020 gravitational wave panel, speaking on background. “We’re not optical astronomers. Our instruments don’t look like telescopes. And the survey process was designed for telescopes.”
That institutional friction — between communities with different funding histories, different agency relationships, and different definitions of what counts as “astronomy” — is the subtext of every Decadal Survey. The survey’s job is to paper over those frictions with consensus language. It rarely succeeds entirely.
The Lesson the Episode Teaches
The Decadal Survey is often described as a model of community self-governance — and in some ways it is. No other scientific field in the United States has developed a comparably rigorous, publicly documented process for collective priority-setting. The survey’s recommendations have shaped the careers of thousands of researchers and the expenditure of tens of billions of dollars.
But the survey’s structural flaw is also its defining feature: it is a document produced by scientists and delivered to politicians, with no mechanism to close the gap between the two. The community recommends in good faith; the budget process responds in its own time and on its own terms. JWST’s overrun was not a failure of the 2001 survey — it was a failure of the institutions that were supposed to execute the survey’s vision. The Casani review said so explicitly, and the community spent a decade absorbing that lesson.
Astro2020’s innovations — the technology maturation program, the independent cost validation requirement, the explicit acknowledgment that flagship missions take longer and cost more than anyone wants to admit — represent genuine institutional learning. Whether they are enough to prevent the next JWST-scale overrun, or the next near-cancellation vote in a House subcommittee, is the question that will define the next decade of American astronomy.
The survey is not a wish list, Fiona Harrison said. But it is also not a budget. The distance between those two things is where the real story lives.
Leave a Reply