Why the Obamacare website was doomed

by John Naughton

The Observer

One of the most dispiriting spectacles of the last month has been the botched launch of HealthCare.gov, the website created to implement President Barack Obama’s landmark healthcare reforms. Obamacare had a desperately turbulent passage through Congress and survived various wrecking attempts by the Tea Party and their accomplices. Then the glorious day dawned and millions of U.S. citizens hit the URL, hoping that, finally, they would be able to find a health-insurance plan that they could afford.

Guess what happened? According to the New York Times, of the 20 million people who tried to access the site over its first three weeks, only 500,000 managed to complete applications for health cover and an even smaller percentage of them actually succeeded in obtaining insurance. In an unprecedented move, the president had to make a public apology for the shambles.

At this point, British readers will mutter: “Well, at least he had the grace to apologize.” Ministers in successive British governments of all stripes have, over the years, presided over some IT cockups that put the Obamacare one in the shade. My guess is that upward of £10 billion has been blown over the years in massive government IT projects that turned out to be death marches to cancelation. But usually the bad news was quietly buried and we heard about it only when the National Audit Office lifted the stone to see what lay beneath.

The Obamacare website fiasco has a quaintly antique ring to it. This is the kind of stuff that used to happen in the first Internet boom, when any clown with an MBA and an idea for a dotcom could attract investors. In those days, new websites were often overwhelmed on launch. The founders hadn’t bought enough servers to handle the surge. But this is 2013 and those kinds of capacity problems don’t exist anymore. You can rent as many servers as you need from Amazon’s cloud, add another 100 in an instant and pay for them on your credit card.

So why was the Obamacare site launch such a disaster? Writing in the New York Times, two politically experienced geeks argue that it’s mostly down to the way the government purchases IT services. “Much of the problem,” they write, “has to do with the way the government buys things. The government has to follow a code called the Federal Acquisition Regulation, which is more than 1,800 pages of legalese that all but ensure that the companies that win government contracts, like the ones put out to build HealthCare.gov, are those that can navigate the regulations best, but not necessarily do the best job.”

That strikes a chord in Britain. British civil servants have traditionally been technologically illiterate, so when ministers demand a new IT system to fix some failing that is annoying the Daily Mail, some archetypal Sir Humphrey breaks into a cold sweat. He knows nothing about this stuff, except that it costs a bomb and that it usually bombs. The specter of the National Audit Office looms over him. He does not want another IT disaster attached to his personnel file. So what does he do?

Simple: He calls up the big consultancy firms asking for tenders. These in turn call up their chums in brain-dead firms called “system integrators” who know only how to do one thing, namely to build massive integrated IT systems the way they were built in the 1960s. And thus begins another death march to oblivion; another project that is billions over budget and years behind schedule. But Sir Humphrey sleeps easy in his bed. After all, the shambles was approved and designed by the boffins who understand this stuff.

The truly amazing thing about this soap opera is that it was allowed to go on for so long. But eventually the penny dropped: The government simply had to smarten up. And smarten up it has. The Cabinet Office now has some geeks who can spot consultancy bullshit at 50 paces. The unit they belong to is called the Government Digital Service. Its brief is to build the right technology, using modern methods and approaches, or to employ agile computing firms that regard £100,000 as a lot of money and are accustomed to delivering on time.

There are not many good-news stories in British government at the moment, but GDS seems to be one. I first began to think that when I heard an anecdote from a friend who moves in these exalted circles. He reported an overheard conversation between a senior GDS officer and a system integrator who had feasted for years on government contracts. The GDS guy outlined the new thinking. “But,” expostulated the contractor, “this is a completely different way of doing things.” “You know,” said the civil servant, “I do believe you’re beginning to get it.” Hallelujah!