Code for America at 10: Reflections on a Decade of Tech-Driven Engagement

On September 1, 2009, Jennifer Pahlka formally founded Code for America (CFA). Over the ensuing decade, the organization has grown to be the largest non-profit in the United States  devoted to improving how government uses technology. CFA currently claims 25,000 members in 85 Brigades across the United States, while internally, some 75 employees support the development of a core set of technologies for modernizing the delivery of government services. Over the years, CFA has been funded through a mix of private and foundation sources, ranging from Google to the Knight Foundation. Recently, as CFA’s tenth anniversary approached, Pahlka announced she would be stepping down as Executive Director, marking the end of a chapter of Code for America. Accordingly, it seems wise to reflect on how the organization has evolved, and where it might go next. 

Many developments over the last ten years have worked in favor of Code for America’s brand of tech-driven activism. Even more than when CFA launched, it now seems natural for a generation of young people who have grown up living with the Internet to be interested in using it for pro-civic purposes. The backlash against corporate technology companies, often referred to as the “techlash,” has reinforced the need for publicly-designed and -owned technologies. When CFA started, Barack Obama was in the White House. There was a lot of interest at the federal level for opening up and embracing the ways of the web. Over time, the Obama administration realized that it needed to do more to bring tech talent into government, starting initiatives like the US Digital Service and 18F. Pahlka, as a leader in this space, was tapped as Deputy Chief Technology Officer in the Obama administration, where she worked from 2013 to 2014. Now, with a more dysfunctional and anti-government administration in power in Washington, much of the grassroots interest in tech-improved governance is happening at the local level. These trends have broadly worked in Code for America’s organizational favor by bringing new people into the organization and forging collaborations in local government. 

But Code for America in 2019 is not the same organization as it was in 2009. It started with Pahlka’s activism and management experience combined with longtime tech guru Tim O’Reilly’s style of messaging and deep industry rolodex. Before there was Code for America, O’Reilly laid the groundwork with ideas like “government as a platform.” Pahlka worked with and supported O’Reilly at events like the 2008 Web 2.0 Expo (which she co-chaired) and the 2009 Gov 2.0 Summit, and they got married in 2015. In a recent interview with me, Pahlka said that O’Reilly’s style – ”bringing people from different disciplines together and connecting the dots” – was an inspiration. But it was Pahlka who came up with the idea to recruit geeks to devote a year of service to cities looking for help modernizing their technology infrastructure. Hence CFA’s name, which was inspired by Teach for America. 

Email from Jen Pahlka describing the brainstorm that led her to start Code for America

Riffing on ideas about “Web 2.0” and open data, Code for America’s first website talked about the importance of the open web, posing the question: “What if interacting with your local government was more like using Facebook or Yelp?” This question echoed beliefs widespread at the time about the emancipatory potential of networked, web-based technologies espoused by authors like Cass Sunstein and Clay Shirky. Over the next ten years, however, government did not turn into a five-star review website. But Code for America was never just about installing social media in government—it was about creating a “big tent” to bring new people into administrative work. Pahlka sees these “translators” who can communicate across professions, cultures, and communities as essential to changing the status quo. To achieve this goal, the organization needed to grow in unexpected ways. “We’ve evolved our strategy every few years to get closer and closer to sustainable systems change in government,” Pahlka said. In other words, when confronted with challenges, the organization did what tech companies often do: it pivoted. 

A Peace Corp for Geeks?

Code for America was first known for its Fellowships, which Pahlka initially described as a “Peace Corps for geeks.” Techies were brought into local government for a year to work (for modest living wages) on pressing local problems that could be alleviated with more up-to-date technology approaches. The financing came from a mix of private partnerships, local government, and the organization itself. Fellows were a compelling sell for cities reeling from budget cuts and hoping to “do more with less” with tech. Mayors—particularly those in “strong mayor” cities—could find desks for fellows more easily than they could outright hire new full-time IT staff. They also wanted to quickly show progress on issues like housing insecurity and education. 

As Fellows spread across the country, so did word about Code for America. The image of CFA Fellows wearing branded track jackets walking into City Hall was compelling to the media and funders. However, once inside government, CFA teams found that getting buy-in on projects was difficult. Fellows only spent part of their time on-site. Frequently shuttling back to San Francisco created challenges of building political capital. Their prototypes and powerpoints could not open all political doors. In later years, inspired by CFA as well as Bloomberg Philanthropies’ Innovation Teams, some cities also decided that they should invest in their own tech teams rather than bring in CFA Fellows. Although this showed that Fellowships helped city administrators take tech talent seriously, it also demonstrated that CFA wasn’t necessary to implement a team-based innovation model. Imitation might be a sincere form of flattery and demonstrate impact, but it also diluted the uniqueness of what CFA had to offer. 

Early on, Code for America was also built on the idea that evangelizing for open-source and open data would help the organization achieve its goals organically. If government projects could be converted into open-source code stored on Github, the thinking went, anyone could apply their technical skills to shared problems. Cities could also save money and time. For example, in 2010, CivicCommons was born, a collaboration between Code for America and NYC-based OpenPlans, one of the earliest local open data nonprofits. The hope was to support collaborations through data and code that could transcend the formal boundaries of government. An app that worked in Boston could be customized to work in Alaska, and also for local community non-profits, saving everyone time and money. But there were few applications like Adopt-a-Hydrant that were simple, politically-agnostic, and addressed a problem common to different cities. CivicCommons, like other such projects, struggled to find sufficient support. 

Unfortunately, CFA didn’t have the infrastructure needed to properly sustain this flowering of creative activity. High turnover rates in staff at the San Francisco mothership also made retaining organizational knowledge a challenge. In other words, Code for America didn’t even know everything that Code for America had built. In 2012, Code for America also started an accelerator and incubator to spin out companies, thinking that early successes in specific cities could lead to larger successes. Despite predictions, no such wave of “civic tech startups” developed. Examples like Textizen and SeeClickFix–public sector tech for-profit companies that are financially successful, or at least sustainable–remain rare. CFA’s accelerator was quietly retired in 2014. And while most of the Fellowship and Accelerator/Incubator projects were uploaded to Code for America’s GitHub so they could be replicated, the truth is few of those 692 repositories are being used. 

Code for America still develops products, but they are now more carefully selected for their potential to meet core government needs. This move towards “service design” is currently represented by projects that increase participation in criminal justice reform and safety net programs. For example, Clear My Record enables people convicted of low-level marijuana offenses to have their records expunged by working with local district attorneys to streamline the process. Last year, Code for America set a goal of expunging 250,000 records by 2019, and aligned the yearly “National Day of Civic Hacking” with National Expungement Week to look to the Brigades for novel ideas. By refining its offerings, CFA’s mothership can devote more time to developing and encouraging the adoption of a smaller set of technologies that can potentially improve the lives of historically marginalized communities. Meanwhile, on the broader front, instead of benefitting from accelerators or open code repositories, CFA found itself buoyed by a type of open organizing that was not originally in the cards: Brigades. 

The Rise of the Brigades

Since so many people applied to be fellows, the acceptance rate for the first few rounds of Fellowships was around 5%. Code for America soon realized there were hundreds more interested techies nationwide than there were Fellowship positions. This shows that Fellowships had generated enthusiasm among young techies about using their skills to help government. The Fellowship format may not have endured, but enthusiasm for a new model of participation was real. In 2010, Pahlka and Kevin Curry had founded CityCamp, loosely modeled off the Sunlight Foundation’s Transparency Camp, under the tagline “Gov 2.0 goes local.” So in 2012, when Code for America was writing a $1.5 million Google grant to test the viability of the Brigades, Pahlka called Curry to help implement them. As Curry told me recently, he imagined Brigades as the volunteer wing of Code for America—like “fans” of Code for America—symbolized in their logo by hats of volunteer firefighters. Google’s grant may have been for a website, but Pahlka and Curry recognized that people, mobilized through a lightweight organizational framework, would be an enduring asset. 

Some 85 different Brigades are now run locally, enabled by Code for America’s tools and resources, each retaining a certain local flavor. North Carolina has a cluster of highly collaborative brigades in smaller cities, while BetaNYC is a registered non-profit, led by Noel Hidalgo, that after years as a volunteer operation has landed its own grants (for more on NYC’s vibrant civic tech ecosystem, see the Civic Tech Field Guide’s new NYC portal). The Brigades have also been leveraged in various ways, as a source of projects and an audience for ideas. The Brigades’ success shows that “civic tech” benefits from a social infrastructure supporting technological design and implementation. By working to foster the Brigades network, Code for America correctly recognized that these hyperlocal organizations were more valuable than the technological outcomes they initially sought. The Brigades also gave CFA a nationwide presence and a claim to be promoting civic engagement. 


Stickers representing many of the CFA Brigades on hand at the 2019 CFA Summit

That said, Code for America’s growth has not always been smooth. Some efforts, like the venture-capital oriented accelerator and incubator, were quick to fade. Others, like Brigades, have endured, though that too has not been without challenges. The Brigades program ran into a tough spot a few years ago with a loss of leadership and member frustrations about funding. That, along with discontent among Brigades (described in a 2016 Civicist story by Jessica McKenzie), led CFA to organize a National Advisory Council (NAC). With their input, CFA re-thought the Brigades’ volunteer funding structure and brought on dedicated staff like Christopher Whittaker, a longtime civic tech organizer from Chicago, to manage them. In 2018 Code for America received $2 million from the Knight Foundation to support the Brigades for two years. This ensured the immediate survival of the Brigades, along with the yearly “Brigade Congress” that provides a venue for Brigade members to discuss how they overcame common problems. This weekend in Chicago, some 1200 Brigade members are gathering for the third annual Brigade Congress. Pahlka sees events like the Brigade Congress and CFA’s yearly Summit as essential to changing government culture to adopt best practices like user-centered design and iterative development. 

Other organizational efforts were recycled into new programs. In 2018, the CFA Fellowship program was rebranded as a “Community Fellowship” where Brigade members locate projects in their community to be supported by a mix of private and public funding, similar to the earlier model. The Community Fellowship approach was intended to address problems with Fellowships through a more community-centric approach driven by Brigades. Hashim Mteuzi was brought on from FedEx as Senior Manager for Talent, and guided applicants through the new process. However, the 2018 pilot of Community Fellowships had a deadline just one month out from its announcement. The timing aligned better with CFA’s yearly Summit event than the funding deadlines for local government. Brigades are generally run by a handful of busy, dedicated organizers. As a result, it was difficult for Brigade leaders to coax reluctant government partners to sign on. This is just one way that “moving fast” in Code for America has sometimes come at the expense of a long-term strategy. 

Like many organizations, Code for America has also been forced to grapple with other echoes of its origins in insular tech cultures. In May 2018, Clay Johnson—an early advisor and mentor at Code for America—was publicly revealed to be a repeat abuser and harasser of women. Pahlka responded quickly to these assertions by reinforcing codes of conduct and publicly apologizing to the affected members. “Organizations must change if we are to rid our movement of the kinds of behaviors that push valued contributors away,” she wrote. “Ours is changing, and for the better.” Johnson’s downfall and Pahlka’s response show that Code for America has tried to evolve past its origins and counter “tech bro” stereotypes. In a picture of all the California Brigade representatives from the last Brigade Congress—an event for Brigade members to gain a sense of community and share insider tips—I am one of two white men in a baker’s dozen. CFA is now transparent about its progressive demographics, and has a board that reflects the diversity of the brigades. These efforts show that they are actively supporting a fuller sense of diversity to try to counteract the tendency in tech to benefit the already privileged. 

Code for America has also wisely evolved its storytelling to lean less heavily on a tech-first, politically agnostic model of social change. At the start, its message revolved around the idea that government needed technological expertise, and Silicon Valley had it. “Politics isn’t government,” as Pahlka has frequently said. But members, recognizing that technologies have implicit politics, were also more interested in designing and implementing technologies ethically, not just more efficiently. As a result, Code for America has latched on to ideas that both democratized technology design and were palatable to government employees. For example, at the 2014 CFA Summit Laurenellen McCann’s concept of “build with, not for” was given central billing because it humanized CFA’s brand of tech-centric public service. Their phrase also sat comfortably with its branding as “civic tech” rather than “gov tech.” Code for America now regularly uses social justice storytelling to bring together activists in Brigades and more politically agnostic government officials. Through stories like that of Jazmyn Latimer, who developed Clear My Record, activists can see this service as justice for imprisoned Black Americans, while more apolitical public officials can simply view it as fulfilling the letter of the law. In these ways, Code for America brings together activists and career bureaucrats, people with very different ideas about politics who might not otherwise meet. These types of meeting spaces for communication across institutional and cultural barriers are necessary for building coalitions of support, but remain scarce. 

Over ten years, Code for America has demonstrated both the potential and precariousness of organizing around technology to fix what ails government. Most importantly, at its core, Code for America has fostered a transition from informational to infrastructural politics among reform-minded techies. If the message about informational politics was “information wants to be free,” infrastructural politics tries to directly reconfigure massive systems—like government bureaucracy and cities—to be more inclusive and participatory. Despite CFA’s early penchant for “Government 2.0,” Jen Pahlka was correct when she described Code for America as not primarily about social media or even technology; it is about embedding people inside of systems with the intent to change them. Code for America reinforced this frame at its Summit this past May. Pahlka said that, after announcing a search for a new Executive Director, “we’re going to look at what we can do to build a more fair, just and equitable systems that govern us and to whom we can bring into the tent to work with us and grow the movement.” Simultaneously, she said she is interested in growing the organization to be “50 times faster and 100 times bigger.” But can Code for America combine a deep, infrastructural approach with exponential growth? 

The history of Code for America shows that democratizing infrastructural politics is a slow, intensive process. CFA has succeeded when it has built trust, emphasized equity, and consciously bridged institutional boundaries. It has stumbled when assuming simple solutions from private tech companies can immediately fix problems in the public sector–what Pahlka calls “Death Star Thinking” (i.e., “One incredibly well-placed shot into the thermal exhaust port and the entire apparatus of our oppression explodes spectacularly”). Public sector projects are slow to gain a foothold in government because the very systems that need to change tend to resist it, and they lack investment potential because their benefits naturally flow to the people. So expecting exponential growth may not be prudent. As Code for America employs vocabulary of social justice and movements, it also finds itself occupying an increasingly crowded space. The last few years have also seen a flourishing of tech organizing: labor movements within tech companies, academic-practitioners interested in situating design as a form of social justice, and community groups. Yet, these organizations lack Code for America’s track record and vital concern of improving government. As we have seen over the last three years of Trump, we still need government in a liberal democracy. At the Brigade Congress on October 18-20, Code for America might be wise to think less about sheer growth, and more about how to leverage its growing institutional savvy and organizational maturity, while maintaining its distinctiveness in an increasingly crowded space.