The costs of science December 4, 2008Posted by Sarah in science.
Tags: astronomy, me, money, nasa, rant
On the back of today’s news conference at NASA, Bad Astronomer Phil Plait wrote a pretty scathing post on NASA record of cost overruns. I didn’t see this bit of the conference, but apparently NASA big boss Mike Griffin tried to convince the gathered press, and unsuccessfully I gather, that the James Webb Space Telescope project is *not* suffering overruns at a current cost of $4.5 billion. Its original estimated cost was $900 million. Interesting…
I do agree with Phil that it’s very irritating, and not at all constructive, of Griffin to try spinning this into “not a cost overrun”. It’s pretty obvious that it is just that. Griffin may as well be straight about that. And in his defense, when I tuned in later to the conference, he did speak very frankly about the difficulties of estimating the cost of a mission during the earliest design stages – and I think I can relate to that. It sounds like even back in the earliest days of JWST, NASA administrators knew the $900 million estimate was far too low.
The problem runs deeper: cost overruns seem to be pretty engrained in science. The cycle goes like this. Scientists write proposals with unrealistic estimates of cost because they (often rightly!) presume that a lower cost will increase the chance of their government or funding agency shelling out the cash. Once they then get the money for the instrument, they have no reserves, no overheads, and therefore quickly run into delays and money shortages. So they go back to their funding agency asking for more, and because the project is already in a more advanced stage, the funders are quite likely to provide a top up. Eventually the thing will get built, late and over-expensive. And so it goes on. Proposals for projects that do contain realistic estimates will automatically look overly expensive compared with all the others.
So yes, scientists involved in building instruments get good at spinning problems. And it’s irritating, and they shouldn’t, and we should all provide realistic cost estimates in proposals. But the way to avoid this happening is for governments to provide better long-term support for R&D in instrumentation for science so that there are people working on these very expensive and important projects who have more experience and better support. That’s why last week’s news about the Dutch funding for E-ELT research was so well received – the government is showing some real foresight and has committed to a long-term investment in astronomy to keep the country as excellent in astronomy as it currently is.