Barack Obama came to power on a promise to deliver “a smarter, more effective government,” fit for the 21st century. It was a mission with a distinctly digital sheen. The Obama campaign famously used the internet in a manner more sophisticated than anything tried before in politics. Now the same team would transform the institutions of public service. Obama supporters talked about an “iPod government”.
Last month’s launch of HealthCare.gov was the moment this dream definitively turned to ash. The website is meant to be the portal through which Americans can shop for health insurance. It launched with a presidential speech in which Obama heralded it as the start of a new era.
It has proved to be a digital marshland into which putative healthcare consumers have been hurling themselves, only to return, broken-spirited, telling tales of frustration, confusion and despair. The site cost half a billion dollars but for much of its existence has been barely able to manage a thousand enrollments a day. Obama has been forced to apologize, repeatedly and humiliatingly.
The website may yet be fixed. But at the very least, this failure to deliver a successful launch of the signature accomplishment of a supposedly tech-savvy administration demonstrates what a chasm still separates the world of government from the world of technology.
This isn’t just a problem for the Obama administration. In the UK too, large-scale IT projects seem doomed to go badly, as with Iain Duncan-Smith’s Universal Credit scheme.
The fundamental reason for this is that they involve a clash of two cultures entirely alien to each other. Politicians and geeks come from different planets.
Geeks are from Beta. Among technologists, “beta” refers to software that is being trialed so that bugs can ironed out. In their hearts, however, geeks are always in beta. They recoil from plans, resist deadlines and are wary of things being “finished”: there is always more tinkering to do. Geeks don’t care much for hierarchy or lines of command: people are judged on the stuff they can do, not on their title.
Politicians are from Alpha. They have a high regard for rank, especially their own. They rely on people doing what they’re told, and what they tell them to do is to implement grand visions and five-year-plans. Politicians hate, above all, to be seen as uncertain: they aim to exude a sense of utter confidence about the future at all times.
These are two different species. When they collaborate – if that is the word – it’s like a horse and a cow attempting to procreate. This matters, because in the future our cratered public finances and rising welfare bills mean that government will be forced to deliver services using fewer people and fewer buildings; more and more public services will rely on the internet. Geeks and politicians will be forced together more frequently.
What divides the two species, above all, is their attitude to mistakes. When a relatively junior member of the Obama administration warned his superiors of problems with the website back in March, he was told that “failure is not an option”. As Clay Shirky points out, the phrase “failure is not option” is fatuous. Failure is always an option, and it is only the politician’s terror of being found out that forces the pretence that it isn’t.
On Planet Beta, mistakes are not just regarded as inevitable, but welcomed as a necessary part of the process of innovation. Mistakes are how you learn. On Planet Alpha, mistakes are feared and loathed and to be avoided at any cost.
Companies like Google and Amazon are constantly running controlled experiments with their services, applying massive computational power to the testing of different scenarios, gaining feedback on what works on what doesn’t work and adjusting accordingly. They don’t work to fixed timelines and eschew grand launches, preferring to nudge, fix and tweak as they go.
But the politician’s fear of mistakes rules out experiment, and when you rule out experiment, you store up failure. The Obamacare website would have benefited from early testing. But its overseers were determined to deliver on a grand vision that had been sketched out at the beginning and had to be “delivered”, no matter what, on the appointed date.
For all the opprobrium heaped on the universal credit, one thing that Duncan-Smith got right is the use of pilot schemes. The failures have led to bad publicity but they are nothing compared to what would have befallen the government had it attempted a grand national launch in the manner of Obamacare.
The government might still abandon the scheme, to avoid the drip of bad news as an election nears. It’s easy to berate politicians for being incompetent managers of IT and online projects. But the reason they’re terrified of mistakes is that voters are impatient and unforgiving of mistakes too.
How many of us are ready to live on Planet Beta?