Interview: Mikey Dickerson, US Digital Service

How can government learn from the private sector?

Mikey Dickerson deals in specifics. “Personally, I try very, very hard to stay far away from trying to transform everything,” says the administrator of the US Digital Service. “If you set yourself the task of boiling the ocean then pretty much by definition you won’t be able to tell whether you’ve made any difference in a year or two – and I really want to be able to tell whether we’ve made a difference in a year or two.” Sometimes he doesn’t need that long. Dickerson’s first experience of government was serving as a key member of an eclectic team of engineers, programmers and designers that in just six weeks fixed the website. Turning a disastrous launch into a system that has enabled millions of Americans to sign up for health insurance prompted an offer to lead the newly-created USDS, and this was enough to persuade him to swap Google - where he had worked as a site reliability engineer for eight years - for an opportunity at the White House. Chasing the pack It’s not only his aversion to a suit and tie that has followed Dickerson from west coast to east. So, too, has his determination to utilise his wider experiences at Google to dive deep into the mechanics of the federal digital operation to drive up performance. There is much to do. “First of all, government still calls it ‘IT’ and ‘cyber’ which the tech industry does not, and that’s a clue right there,” he says. “It’s just different jargon and a symptom of how isolated the two cultures have become from each other.” [blockquote] "Government still calls it 'IT' and 'cyber' which the tech industry does not" [/blockquote] He goes on to say that the differences run far deeper than mere vocabulary, extending across general approach, risk tolerance and speed of innovation. “This issue has become particularly acute and visible to the public in a really painful way,” he says. “Ten years ago the iPhone didn’t exist and now innovations like smart phones, GPS and Uber are deeply intertwined across people’s everyday lives – with government looking flat-footed by comparison.” Dickerson, though, wants to reverse these trends. Working with his small team at USDS, he is leading efforts to improve government computer systems and websites, as well as seeking to anticipate future problems and improving how citizens interact online with government more broadly. Doing this, he explains, starts with ensuring that the right people are in position to make a difference. “There’s no way to make something like social security become something customers like using – like Priceline for example – without the people who possess the experience of building a product,” he points out. “The tech industry didn’t magically wake up one day and learn how to do this either. There have been tons of failed products and all this learning is baked into the people now at Google, Microsoft, Amazon, Twitter, and Facebook. You need access to the same talent pool. If it was easier for people to circulate between both sectors then government would benefit from that trade.” Winning hearts and minds Of course, there are many legitimate differences between government and the private sector – and Dickerson is quick to say that his is not an attempt to bridge every aspect of this divide. Take tolerance for failure, for example. “There is a big difference between a food delivery start-up in San Francisco failing to deliver a profit and the failure of a government programme to modernise social security disability payments or veteran’s healthcare,” he says. “Government can’t and should not gamble to the same extent.” However, he goes on to say that while a certain amount of risk aversion is understandable, it shouldn’t be allowed to permeate through government as a whole. “Fear of failure is a root cause of why these technology projects fail a lot of the time,” he says. “People in every agency that I work with are terrified that if a project does something different and doesn’t work out then they’ll be blamed for it. They’ll then have to go through congressional hearings and consider the possibility that they could lose their jobs.” Dickerson believes that that this approach to projects – and their potential failure – requires a radical shift. “We need to re-interpret the incentives model that people respond to,” he says. “What drives me crazy is we have collapsed into a state where technology projects fail all the time in very predictable ways. We hire a contractor for hundreds of millions, sometimes billions of dollars, put them on a seven to ten year timeframe and we repeat this pattern again and again and it almost never works. The only reason I say ‘almost’ is because somebody might pop up and surprise me with some success story but I am actually aware of zero success stories of this ever working the way anybody intended. But we don’t count these as failure because they don’t tend to cause hearings and rarely make the news pages.” USDS has been created to help address these issues by helping agencies apply technology in smarter, more effective ways that improve the delivery of federal services, information, and benefits – always with a focus on the customer experience in mind. Given the scale and complexity of federal operations, it would be easy to think that this is something of a mission impossible – but Dickerson uses the analogy of the iPhone to illustrate how numerous moving parts can lead to a product that is in sync with customer demands. “The iPhone was a success not because Apple thought of every single feature and integrated it all together into one model,” he points out. “The iPhone was a success because they invented the App store and now millions of people are writing millions of apps to do small defined things. I have dozen of apps and it is a fact that they do not provide one unified, seamless, integrated customer experience. The look and feel is different from app to app but it turns out that I, as a consumer, and every other consumer that I know, are sophisticated enough in interacting with technology to assemble the set of features that I want in my phone and then in the process allow each one of those pieces to iterate much faster.” With this in mind, USDS doesn’t seek to transform vast swathes of government operations in one go – instead it opts to focus on smaller individual projects, partly a reflection of their own limited resources. “We take on projects and get very involved in their management and operation but we can only touch a fairly small number because we’re not that big an organisation,” admits Dickerson. “In any case, the size of some systems – like Medicare processing or tax returns – means we can’t solve a whole problem at once. Instead, we seek to solve one piece of the problem at a time which is less risky but can give the mistaken impression that it’s not important and what we’re doing is not very exciting. This is a perception problem that holds us back but breaking the problem into smaller pieces, where each piece is small enough that it can fail without it being a catastrophe, is another thing that government has to learn how to do.” Onward and upward Such projects have helped ensure that Dickerson’s DC days are anything but dull. From dawn to dusk, his is a schedule packed with meetings, calls and projects – all designed to help the digital agenda forward. It certainly seems a long time since his first exposure to Washington’s corridors of power during the project. Memories of those long, coffee-fuelled days still hold strong, however. “There wasn’t really an understanding that you needed to have any of the nerds, any of the technology people, any of the engineers, any of the people with computer science or engineering degrees,” he recalls. “I went to tons of meetings where people weren’t necessarily hostile, but confused about why a computer programming person was in the room when they were talking about designing policy. I’m not saying that a government agency is going to become a tech company, but there is no policy we implement that doesn’t have a major implementation and technology component to it.” And that’s the key point. Technology isn’t going away – anything but. Instead its march into every aspect of government continues unabated and rightly so as such advances offer policymakers a hugely powerful opportunity to transform services for the better. “We’re looking for tangible results,” concludes Dickerson. “With we were able to put into place a system where 15 million people were able get health insurance and this is something that can never be taken away from us. I’m going to measure success by stacking up these real results against the energy, blood, sweat and tears that goes into doing this work. We’re just getting started.” This interview was written by Danny Werfel, Director of the Boston Consulting Group in Washington DC for the Centre for Public Impact.