The latest round of cost growth on NASA’s James Webb Space Telescope (JWST) is disheartening for a number of reasons — prominent among them that it took an independent review to uncover it — but hardly surprising. The infrared observatory, billed as the successor to the flagship Hubble Space Telescope, has a history of steep cost growth: Its estimated price tag had steadily risen from $1.6 billion in 2003 to $5 billion when NASA unveiled its 2011 budget request in February, and there were warning signals that it would climb even higher. That’s what prompted U.S. Sen. Barbara Mikulski (D-Md.), who chairs the Senate Appropriations subcommittee that oversees NASA and is a JWST supporter, to order the review in the first place. Sure enough, the review team determined the telescope would not be ready for launch until September 2015, 15 months later than NASA’s latest target date, and as a result would cost $6.5 billion.

Space programs of the JWST’s scale and complexity are notorious for exceeding cost estimates by a wide margin, not just because unforeseen technical issues inevitably arise but also because of the way they are sold and managed. Government agencies and program managers have a well-established track record of winning approval for projects based on unrealistically low cost estimates — contractors do the same thing to win contracts — and of sweeping problems under the rug until they grow too big to ignore. By the time that happens, so much money has already been invested and so many jobs are at stake that the program is difficult if not impossible to kill.

This phenomenon is not unique to the United States. In Europe, for example, it has been publicly suggested that the Galileo satellite navigation system, whose latest cost projections far exceed what the European Union has said it is willing to pay, is simply too big to fail. In other words, whatever it ends up costing, the taxpayers are simply going to have to suck it up. Although seldom invoked in public, the “too big to fail” argument has kept numerous space programs alive well past the point where the sponsoring agency would have been wise to cut its losses.

Without benefit of hindsight, of course, it is nearly impossible to determine when a program has crossed that threshold. Canceling the JWST now would mean writing off an investment of $3 billion, which probably doesn’t make sense given the estimated $2.9 billion cost to finish and launch the observatory and the hundreds of millions of dollars NASA would have to pay in contract termination fees. What’s unknowable at this point is whether the cost growth will stop at $6.5 billion. If the final price tag winds up being $8 billion or $10 billion — and that’s a perfectly plausible scenario — walking away from the program now looks a lot more attractive.

NASA Administrator Charles Bolden has made management changes on the JWST and said he is not ready to concede as inevitable the latest round of cost growth, which would gut the agency’s budget for new astronomy programs over the next five years. This is probably wishful thinking; the real question is whether NASA can stop the bleeding at $6.5 billion. Replacing the program manager was a necessary first step given the review team’s conclusion that the cost growth was due primarily to management rather than technical issues. But the flawed management practices cited in the report — accepting unrealistic cost estimates, setting aside insufficient program reserves and deferring work to stay within budget in any given year — are well entrenched and, apparently, widely accepted at NASA and other government agencies.

Another big problem uncovered by the review is that senior NASA officials apparently were oblivious to the status of a program that in 2005 was singled out for special scrutiny by then-NASA Administrator Mike Griffin after he signed off on a $1 billion cost increase and 22-month launch delay. This is both incredible and inexcusable and raises serious doubts about whether NASA can be entrusted with big development initiatives.

Neither NASA nor the U.S. Congress should be satisfied with steps aimed simply at bringing the JWST program under control; they need to find out where the breakdown occurred that allowed this to fester in the dark and develop measures to prevent it from recurring. For too long government space programs have been run according to a philosophy that bears resemblance to the adage that it is easier to ask forgiveness than it is to get permission. Given the current state of the nation’s finances, NASA may well be approaching the limit of the public’s capacity to forgive.