James Webb Space Telescope: the $1.6bn telescope that cost $9.7bn and ran a decade late

The James Webb Space Telescope was sold as Hubble's successor: a segmented mirror three times Hubble's diameter and a tennis-court sunshield, orbiting a million miles from Earth. The early estimate was around $1.6bn, with construction starting in 2004 and launch then planned for 2016 — itself already years behind the projection GAO cited from 1999. By 2011 the cost had soared past $6.5bn, and Nature had nicknamed it 'the telescope that ate astronomy' because other research budgets were being cannibalized to feed it.

In July 2011 the House appropriations subcommittee for commerce, justice and science had enough and voted to strip $1.9bn from NASA's budget — cancelling the project with the words 'billions of dollars over budget and plagued by poor management'. The telescope survived: late in 2011 it was rebaselined at $8.8bn lifecycle cost with launch readiness in October 2018 — what GAO called almost nine times the cost and more than a decade later than originally projected in 1999.

The replan did not hold. In 2018 NASA's Independent Review Board reported remaining technical tasks at prime contractor Northrop Grumman, including work performance challenges on the sunshield and propulsion system, and reset launch to March 2021 at a lifecycle cost of $9.66bn. GAO's 2021 assessment put the cost at $9.7bn — nearly double the 2009 baseline — with launch delayed more than seven years and reserves burning faster than planned. Webb finally lifted off on Christmas Day 2021, five years after even the 2016 launch date.

Development went from a $1.6bn estimate to $9.7bn; launch slipped more than a decade; other NASA astronomy budgets were cut to absorb the overrun

The project promised never-flown technology — an unfolding segmented mirror and a five-layer sunshield — on a schedule set before any of it had been demonstrated.

NASA baselined to optimism: dates and dollars were set to win approval, and each replan ratcheted the numbers up instead of exposing the risk earlier.

The hardest phase, integration and test, is exactly where GAO said schedules slip; Northrop Grumman's sunshield tearing in test and a propulsion valve mistake confirmed it.

Absorbing the overrun meant cutting other astronomy programs, which built the institutional revolt that nearly killed the mission in 2011.

Baselining to optimism does not make optimism true: promising flagship results on immature technology turns every unknown into a late, unbudgeted surprise paid for by the whole portfolio.

Webb launched on 25 December 2021 from Kourou on an Ariane 5 and has since delivered landmark early-universe science — the engineering gamble eventually worked. But the financial record stands: the 2018 replan set a $9.66bn lifecycle cost, and GAO reported the total at $9.7bn — nearly double the 2009 baseline — with more than seven years of delay. The case became the standard citation for flagship 'conspiracy of optimism' budgeting, and for why oversight bodies demand mature technology before baselining.