NASA's Constellation program: $9bn spent, canceled before anyone flew

In 2005, carrying out the Vision for Space Exploration, NASA set up Constellation: an Orion crew exploration vehicle plus the Ares I and Ares V rockets, meant to return astronauts to the Moon by 2020 and eventually go on toward Mars. In August 2006 NASA awarded Lockheed Martin the contract to build Orion. Constellation became NASA's flagship human spaceflight effort and its largest development program of the decade.

The program ran ahead of its own engineering. GAO spent years flagging that Constellation had moved into development without mature technologies, firm requirements or a realistic cost estimate, and in September 2009 reported that program cost and schedule would remain uncertain until a sound business case existed. The Augustine Commission, the expert panel the President convened that year, reached the same verdict: milestones were slipping and the goal of returning to the Moon would not be met in any reasonable time or at any affordable cost.

In February 2010 the FY2011 budget request canceled Constellation outright, offering $2.5 billion over two years for facility and close-out costs. Congress pushed back — an appropriations act barred NASA from spending FY2010 money on termination — but the NASA Authorization Act of October 2010 made the end formal. About $9 billion had been spent, and no one had flown on any of it. The program's technology and know-how were redirected into a new heavy-lift rocket.

About $9bn spent by the time the 2010 authorization act formally ended it; the FY2011 budget added $2.5bn over two years just for close-out

NASA began full-scale development before the business case existed — GAO's central finding was that implementation had outrun requirements, technology maturity and cost realism.

The schedule did not match the money; the Augustine Commission found far more funding would have been needed to make the Moon timeline real.

The program carried a 2020 Moon deadline that made early, optimistic baselining politically convenient and technically costly.

Constellation depended on continuous political support across administrations, and its justification eroded when the independent review withdrew confidence in cost and schedule.

Development that starts before requirements, technology and costs settle pays for discovery the expensive way. A sound business case comes before implementation — in any industry.

The fight over the wreckage lasted most of 2010: Congress first barred termination spending, then legislated the program's formal end while directing its hardware base into a successor heavy-lift rocket. Orion survived as a crew vehicle on the new rocket, and the same cost-estimation failures later resurfaced on the successor program. The archive's lesson record of that successor — awarded fees while schedules slipped — shows how little the institutional incentives had changed.