You are not logged in. LOG IN NOW >

Four Years In, Code for America's Experiments In Disrupting Govt Still Just The Beginning

BY Sarah Lai Stirland | Monday, October 14 2013

One of Code for America's projects this year uses text messaging to notify food stamp recipients about their benefits

Code for America's projects may not end world hunger, overhaul a broken criminal justice system, or solve municipal budget crises in of themselves — but both citizens and government officials see promise in using the organization’s philosophy and techniques to work more closely together to incrementally solve these kinds of problems.

That’s becoming increasingly evident as the San Francisco non-profit enters its fourth year and convenes its annual summit this week downtown. Officials from 85 cities both in the United States and abroad are getting together to discuss everything from the merits and limits of acquiring talent and technology through the social coding site Github to implementing municipal entrepreneurship-in-residence programs. Last year, officials from 30 cities attended.

“We’re using this conference as an opportunity to push for more action,” said Abhi Nemani, Code for America’s interim co-exectuive director. “How can we leverage what happened over the past year with our 10 fellowship sites, and dozens of brigades and our peer network, and take it to the next level, with a broader audience of people interested in this work?”

It’s a question that city leaders and deep-pocketed funders and academics are interested in as well. The past year has seen the launch of the Data-Smart City Solutions program at Harvard’s Kennedy School of Government, the John S. And James L. Knight Foundation’s awarding of more than $3.2 million to several open government projects that appear to be gaining traction, and the launch of Bloomberg Philanthropies’ partnership with five American cities to develop an ‘innovation delivery model,’ i.e. programs to establish the conditions that encourage city officials to feel free to experiment in order to improve the way they deliver city services.

And no fewer than 13 American cities and counties now operate innovation offices of one form or another where a government worker and colleagues are tasked with working with local bureaucrats to brainstorm on how to address kinks in the system, or improve what they’re doing with their existing resources.

“We’re really seeing innovation institutionalize itself at City Hall,” Nemani says. “The notion of an innovation office, like Urban Mechanics, has really taken off.”

Some of the fruits of Code for America’s projects over the past four years seem to be pointing the way. The group’s fellowship program sends coders to work with local governments to come up with tech-based solutions to some of the problems they’re facing, but in the process, the CfA fellows also end up training key government staff to replicate the startup culture of Silicon Valley in their own work: That is, idea generation, building a minimum-viable-product to address those ideas, measuring its efficacy and figuring out if it’s working, or what needs to be improved and then iterating upon that.

For example, in San Francisco this year, four CfA fellows worked with the city’s Human Service Agency (HSA) and the Mayor’s Office of Civic Innovation to figure out how to make the interfaces that benefits recipients use more intuitive. The fellows, with a wide variety of of backgrounds from anthropology to Web development to policy research, built HSA a text-messaging application named Promptly, which alerts food stamp recipients to the fact that they’re about to lose their benefits because they didn’t fill out a form. But in the process, the fellows also conducted workshops to teach HSA staff about user-experience research and design and usability testing. The fellows also worked with staff to find tech-savvy employees who could maintain and carry forward their work after they left.

Shannon Spanhake San Francisco’s deputy chief innovation officer, will be sharing those experiences and lessons learned at this week's conference.

There are other examples of success that CfA’s conference will highlight as well. As disastrous as the rollout of was, it received praise for the way the federal Department of Health and Human Services used entrepreneur Eric Ries’ principles of agile development to develop the front-end. Code was shared on GitHub. HHS’ Chief Technology Officer Bryan Sivak will be on hand to discuss that project. GitHub’s Evangelist Scott Chacon will also discuss how governments have been using the service to develop open-source projects.

CfA’s San Francisco project was just one of 10 fellowship programs that the group ran this year. Its fellows fanned out to cities and counties both big and small to spread their gospel of agile development:

  • In New York City and Louisville, Kentucky, they worked with city officials to create a pre-trial criminal justice dashboard that gathers statistics on suspects and those incarcerated to enable key players in the criminal justice system to make better pre-trial detainment decisions;
  • In Kansas City they built a system to train business owners to gain Web skills to make the most of broadband in order to promote and manage their companies;
  • In Las Vegas, they helped the city use data to improve the way it zones businesses;
  • In Oakland, the fellows built a public records request tracker;
  • In San Mateo, the fellows built the Ohana API, which enabled them to build SMC-Connect, an online interface that helps citizens to more easily find the benefits for which they are eligible.
  • South Bend, Indiana got a prototype app called City Voice that seeks to garner community input via text message on what should happen to abandoned properties. The fellows also redeployed code from a 2012 CfA project named Local Data to help the city conduct its surveys of abandoned properties more efficiently so that it could more quickly achieve its demolition goals.
  • Meanwhile, Summit County in Ohio, had its park trail information system overhauled so that residents could more easily discover trail information.

Many of 2013’s fellowship projects sound more mission-critical than previous years. That, in part, may be because city officials are learning how to navigate their own bureaucracies in order to enable them. But there’s a sense that this is still only the tip of the iceberg. For every project that CfA embarks upon, there are dozens of other transformative ones that are rejected. For example, Californians still have to deal with a relatively opaque system of discovery if they want to know about potential conflicts of interests government officials as a result of their financial holdings. The state’s equivalent of the Federal Election Commission, the California Fair Political Practices Commission, requested fellows from CfA to come and help change that. Ultimately, CfA wasn’t able to commit because of the state’s onerous procurement contracts, said FEC Commissioner-elect and CFPPC Chair Ann Ravel in an interview.

“There were provisions in the contract that were not palatable to Code for America,” she said, such as liability issues that are standard, boilerplate contract terms for large IT vendors — a legacy from the numerous past failed giant tech projects in California.

“That was unfortunate for us because moving the ball forward in any way — even if the project wasn’t finished within the year — would have been beneficial,” Ravel said. The CFPPC could have finished what they started on their own, she explained.

The CFPPC chair chalked the experience down to her own newness in the position at the time. In retrospect, she said, she could have avoided that contract altogether and gone to the head of California’s Department of General Services to hammer out an arrangement that could have accommodated the fellows.

“I do think that it’s a mistake to think of [the CfA fellows program] as an IT project,” she said. “They really should be treated as fellows. In Sacramento, they hire young people who are called Capitol Fellows. They get the benefit of learning, they get a stipend, you ask them to do certain things that other people do for you, including special projects. You might at the end of the day have a Capitol Fellow that doesn’t do as much as you’d like. But that’s how you should think about it — you should be pleased with smaller successes.”

The Senate confirmed both Ravel and Lee Goodman, a Republican counterpart, to be FEC commissioners late last month. Ravel said that Goodman is also interested in increasing the FEC’s technological capabilities to boost transparency. She said that she intends to call CfA Founder Jen Pahlka (who is now White House Deputy CTO) to see how they can work together.

Ravel’s experience with procurement is indicative. Outdated and cumbersome government procurement policies have presented another big hurdle for civic hackers and their infant companies. Many of them simply do not meet the criteria mandated in such policies. So that subject is also going to be a big theme at the conference, Nemani says. Much of the third day of the conference will be devoted to brainstorming sessions on modernizing procurement — always a sticky and difficult subject given that city officials are always mindful that they’re experimenting with taxpayers’ money.

CfA, along with the Sunlight Foundation* and the Omidyar Network, conducted a local procurement survey among city officials earlier this year. Among other things, they found that many of those cities are trying to accommodate civic startups with some form of workaround to traditional IT procurement procedures. Kansas City, Missouri, Philadelphia and San Francisco appear to be the three cities who are actively experimenting with new approaches to procurement. Kansas City, for its part, passed a resolution that would allow the city to test out a startups’ products using city infrastructure, and San Francisco has created an exemption to its procurement policies for experimental projects that run on an evaluated basis for under two years.

In Philadelphia, Chief Data Officer Mark Headd experimented this spring with the city’s procurement policies by using GitHub to solicit proposals for a piece of software. The contract cap was set at $30,000. In a blog post, Headd said the goal was to contract with small, local technology companies. He characterized the experiment as a success because it attracted “9 quality responses from local vendors,” which was triple the number required by the city’s procurement rules. GitHub also provided a helpful channel through which to communicate with the vendors, he wrote.

It’s this willingness to experiment on a small scale and evaluate the results that’s crucial to changing the current model of governing, Nemani says.

“One thing about our model is that, unlike traditional vendors who have a very specific app that they’re going to build, our model focuses on emergent outcomes,” Nemani says. “We push for the cities to be comfortable with, frankly, some ambiguity in the process — but sometimes some of the most interesting things come out of that.”

*Disclosure: Personal Democracy Media co-founders Andrew Rasiej and Micah Sifry are senior advisors to the Sunlight Foundation.