Should Local Government Be Run Like Silicon Valley?
Cities are contracting with Code for America -- what some call “the Peace Corps for geeks” -- in an effort to seed Silicon Valley virtues in local government.
Mark Bolton has a problem. To be more precise, he has 250 problems. Bolton runs Louisville, Ky.’s metro jail system. Its main facility -- two aging structures joined by a pedestrian sky bridge -- is designed to hold 1,353 inmates. Another facility across town has another 440 beds. But on this particular day in early February, Bolton has 2,043 people behind bars. For weeks, he’s been adding cots to the barracks-style rooms that house most inmates. But the jail is so overcrowded that he’s now considering opening another wing of the facility, even though it’s not up to code.
“It’s actually illegal for us to open it, but I have to consider it,” he says. “[W]hat’s the worse of two evils, just packing them in here like sardines, or putting them in a section of the jail that we don’t want to put them in?”
Jails don’t fill themselves. Police, prosecutors and judges do. “A jail is like a rain barrel,” says Kim Allen, executive director of Louisville’s Metro Criminal Justice Commission. “If you want to change the water level, you’ve either got to put less in or let more out.” But while everyone in Louisville knows the barrel is overflowing, they’ve had a hard time adjusting the taps.
One of the biggest problems is sharing information. Louisville has a long history of coordinating criminal justice initiatives, including what may be the oldest multiagency criminal justice commission in the country. Sharing information electronically, though, is a different story. The courts’ computer systems are administered by the state; the jail’s by the city; and then there are the systems used by Louisville’s various prosecutors, the public defender’s office and the police department. Although significant progress has been made in linking up the courts and law enforcement, the overall system is still far from optimal. Louisville’s 40 elected judges use their computer system in different ways to different degrees. Unnecessary duplication of forms is rife. Prisoners stay in jail for roughly 20 days on average before going to trial, even though many are there for nonviolent offenses. Some are there for failing to appear in court on bench warrants -- warrants they didn’t even know about. Louisville doesn’t have the technology to provide public access to its criminal justice system. As a result, it often communicates with offenders by arresting them.
Metro officials know these problems could be fixed with better systems and new technology. But like most city and county officials, they’re hard-pressed to find the money to make these sorts of investments. The state-operated court system has experienced years of deep budget cuts. Kentucky’s chief justice recently estimated that it would cost $28 million to begin updating the system. Software and support from vendors are expensive, and Louisville’s 70-person IT department includes just two programmers, which limits its ability to develop customized solutions.
All of which explains why Bolton and his top managers are willing to talk at length with a San Francisco-based team of computer programmers, designers and data experts. On a recent rainy day, a trio of them is gathered around the conference table off Bolton’s office on the third floor. They represent an unusual skill set for local government.
Marcin Wichary, 35, is a user interface designer and programmer. Originally from the Polish city of Szczecin, he’s worked for the past seven years at Google, primarily on its Chrome and search teams. Laura Meixell, 27, was previously a Presidential Management Fellow at the U.S. Department of Housing and Urban Development, where she worked on HUDStat, the department’s performance tracking system. Shaunak Kashyap, who turns 32 next month, grew up in Mumbai, India, before emigrating to the U.S. His first job as a programmer was writing code to control satellites. He then made his way to Yahoo and Netflix.
With their impressive credentials, attentive demeanors and buttoned-down business attire, the youthful group could easily be a team of high-powered consultants. In a sense, they are. But these are consultants with a twist. Louisville isn’t paying for their services. Instead, they are here as volunteers with a San Francisco-based nonprofit called Code for America (CfA). The idea is for the city to lay out its problems and for the team to explore ways software codes -- new apps -- could solve them. In particular, they are looking at the issues Louisville is having with its criminal justice computer system to see whether there might be a customized software fix.
CfA founder Jennifer Pahlka calls it “the Peace Corps for geeks.” Programmers and other data specialists apply for a one-year fellowship, for which they receive a stipend of $35,000. Cities apply too, presenting potential projects and also lining up funding for the fellows. (Some cities, including Louisville, rely on foundations; others appropriate funds to cover the $120,000 cost of a three-fellow team.) Last year, 29 cities and counties proposed projects. More than 550 people, many of them from top technology companies, applied for the 28 fellow positions. In a sea of 14 million local government employees, that’s not a large number. But the idea Code for America embodies is very big indeed. “The notion,” says Pahlka, “is that we should make government work like the Internet itself.”
Louisville and other cities, such as Boston and Philadelphia, have used their partnerships with CfA as part of a conscious effort to seed Silicon Valley virtues, such as creativity, speed and experimentation, in local government. That raises the question: What’s in Silicon Valley’s secret sauce, and why are Louisville -- and a growing number of other cities -- so eager to get it?
Code for America began with the vision of computer publisher Tim O’Reilly. In 2005, O’Reilly coined the phrase “Web 2.0” to describe the shift away from the personal computer to the Internet itself as a computing platform. Four years later in 2009, O’Reilly Publishing and UBM TechWeb organized a conference, Gov 2.0, at which he challenged programmers to start working on society’s most urgent problems. His call resonated with Pahlka, a computer gaming event organizer, who helped put the conference together.
“One of the things I love about Silicon Valley is the experimentation and willingness to play around,” says Pahlka. “That is wonderful, but sometimes becomes trivialization. As a society we can’t afford to have some of our brightest minds working on trivial things like, you know, Facebook apps.”
Barack Obama’s election in 2008 raised hopes in Silicon Valley that Gov 2.0 for the federal government was at hand. And indeed people such as Todd Park, one of the founders of Athenahealth and the federal government’s current chief technology officer, moved to the nation’s capital and became effective proselytizers for open data and startup government. However, one of Pahlka’s childhood friends, Andrew Greenhill, who worked as the chief of staff to then-Tucson mayor Robert Walkup, argued that Silicon Valley was focused on the wrong level of government -- that Gov 2.0 could have a bigger impact on cities.
“Andy kept telling me that people understand their relationship with city government more than they do with the federal government,” recalls Pahlka. Cities such as Tucson needed apps too, apps that addressed the issues people cared about most: Where was the closest park? Was it clean? When would the trash get picked up?
Pahlka was dubious. “That’s not going to happen,” she told Greenhill. People in Silicon Valley, she argued, were interested in two things. One was making lots of money. The other, which she considered even more important, was that “they value creativity, agility, speed, autonomy and independence. That is not what government offers them.”
For nearly a year, Pahlka and Greenhill went back and forth on the issue. Then Pahlka went to Flagstaff, Ariz., for a family reunion. There she met up with her friend. “One day, we were talking about his experiences with Teach for America, how it influenced him, how he went into public service.” For Pahlka, it was a eureka moment. Programmers would never see writing apps for local government as a way to get rich, but they might see it as a form of public service. That very night, says Pahlka, “I told my dad and stepmom that I was quitting my job and starting something called Code for America.” She did. A year later, in January 2011, CfA enrolled its first group of “fellows.”
Pahlka’s vision for the group was essentially that by taking start-up values to government, “this thing that works” (Silicon Valley) would “fix this thing that doesn’t” (government). She quickly concluded that idea was wrong.
“There is a lot that is broken,” Pahlka says of local government, “and it is a really complicated system.” But, she continues, “the biggest thing we have learned is that the people who work in government, by and large (and there are certainly exceptions), are actually amazing. They go into this field because they want to serve the public. The notion that we were going to fix them was terribly wrong.”
What was broken, though, was the relationship between government and the citizenry. “That’s a problem on both sides,” Pahlka says.
One way she saw to fix it was by encouraging citizens to see problem-solving as something they could do with government. CfA fellows would act as catalysts in that process. An early success story occurred in Boston. During snowstorms, many residents there dig out and assertively defend parking spaces. Meanwhile, the fire department struggles to clear hydrants. Fellows helped write an “adopt-a-hydrant” application that allows residents or businesses to take responsibility for shoveling out “their” fire hydrants. (Buenos Aires, Chicago and Honolulu have since adopted modified versions of the application for maintaining other shared infrastructure, such as sidewalks.)
The following year, fellows in New Orleans addressed a problem that frustrated local residents in neighborhoods affected by urban blight: They created a one-stop resource with extensive information on who owned what piece of property, as well as the restoration status of each lot. Called BlightStatus, the website allows citizens to enter an address and get a comprehensive update on remediation efforts. In Detroit, a team of fellows developed Textizen, an app that allows local officials to conduct quick surveys of city residents.
As Code for America has grown, Pahlka and her colleagues have sought to leverage its work in numerous ways. One has been to make the code for all of the applications developed by the fellows freely available online. Another has been to use local programming talent by inviting local computer programmers to become “captains” of civic hacking “brigades” that can assist local governments even in the absence of fellows. More recently, Code for America has created a business accelerator to encourage programmers to develop commercial applications for local government.
“Once you create the demand for a different way of doing things, you need to be able to fulfill it,” says Pahlka. “The typical way you fulfill it is through contracts. We need to be a much larger, richer ecosystem of companies that contract with local government.”
As interesting as Code for America’s plans for local government are, even more intriguing are the ways in which mayors and county leaders are using those ideas to implement Silicon Valley virtues in local government. For Mayor Greg Fischer, bringing CfA to Louisville isn’t simply a way of addressing jail overcrowding. It also offers a chance to promote innovation in government more broadly.
Prior to winning election as mayor in 2010, Fischer was a successful entrepreneur, with a passion for quality improvement. By all accounts, he’s brought that same passion to metro government.
“He lives and breathes continuous improvement,” says Margaret Handmaker, a longtime consultant for Booz Allen Hamilton and a former state secretary of commerce who came out of retirement to serve as Fischer’s interim director of economic improvement. “It’s what he is about.”
One of Fischer’s first acts was to create a planning department, whose primary function was putting W. Edwards Deming’s Plan-Do-Check-Act (PDCA) model for continuous quality improvement in place. (“There were a lot of cities and agencies collecting data,” says Fischer, “but when I asked them, ‘How are you then solving problems?’ most people looked at me with a what-does-the-question-mean look.”) He also embraced Baltimore’s CitiStat, which gathers data on government agency performance in a single computer database, allowing local officials to evaluate everyday performance and progress toward civic goals. “LouieStat,” now in its second year of operation, began with nine key agencies. Together with the PDCA process, it aims to address what Fischer calls “the daily work and continuous improvement” part of government. But there is also what Fischer calls “the breakthrough part.” In 2011, Louisville was one of five cities awarded a $4.8 million Bloomberg innovation award. With it, Louisville intends to make striking improvements in five areas: recycling, animal control services, vacant and abandoned properties, rezoning, and encouraging exports.
For Fischer, successful governance involves all three of these functions -- the daily work, continuous improvements, and innovation and breakthroughs. He sees the city’s partnership with Code for America (the cost of which has been paid for largely by the Houston-based Arnold Foundation) as a way of improving on all three of these functions.
“Number one,” says Fischer, “we have a real problem we can improve on” -- jail overcrowding. This is not a small problem. Nationwide, some 800,000 people are locked up in local or county jails at any given time. Many are nonviolent offenders awaiting a court date. Yet only about a tenth of the nation’s local jurisdictions use a risk-assessment protocol to decide who should stay behind bars and who can be safely released pending their court appearance. The cost of building new facilities instead can be considerable. Denver recently spent $159 million on a new 1,500-bed detention facility. While it has many features that should make it a more efficient and humane institution than Louisville’s facilities, it’s an expenditure Fischer would prefer to avoid.
The second reason to use Code for America is the role it can play as a dedicated innovation resource. “It’s hard in any business, or government for that matter, to pull out dedicated resources just to focus on innovation,” says Fischer, “so this is an intentional way to do that.”
The third attraction of the program is that it can give Louisville a global best-practices view. “When you’re doing culture change in an organization, there have to be signals from the outside that your organization is in fact changing and being valued on the national or international level,” says Fischer. “That’s why we compete for things.”
There’s another benefit too -- fresh eyes to assess how and what you’re doing.
“The last person you want to ask is the person who works in the government,” says Ted Smith, who heads Louisville’s economic growth and innovation initiative. “They’re the fish in the fish tank. They don’t see the water. The hallmark of all Code for America projects is, what is the user experience? How does it work? And is it inefficient or ineffective in some way?”
As for the CfA fellows, after their first month in Louisville, they’ve begun to circle around a handful of potential fixes to the criminal justice system. Resolving all the issues of inoperable computer systems may be beyond the capacity of three fellows.
“Every group of users has different systems and different levels of access,” notes Wichary. Simply trying to map them all -- a task the fellows have undertaken -- “is an interesting challenge.” They have also begun looking at other issues. One has to do with how Louisville interacts with the people it imprisons pending trial.
Outstanding warrants is on this list. Louisville -- a city of 600,000 residents -- has 100,000 of them. “We have heard that sometimes it is hard for people to re-docket,” says Wichary. Some citizens who have a warrant pending can walk into court, pay a fine or reschedule (re-docket) a court date, and walk out. But first they need to know there is a bench warrant. They also need to understand where they need to go to re-docket their case, and what words like “re-docket” mean.
Right now, there is no single source of information on outstanding warrants, no easy way to pay a fine and no way to reschedule. “It’s just one piece of the puzzle,” says Wichary, who also cautions that the fellows have not yet begun to discuss specific solutions in detail, but “it’s a general area we are trying to explore.”
That’s just fine with Criminal Justice Commission head Kim Allen. “There are so many opportunities for either applications or process changes or interfaces that can make a real difference in how effective and inefficient the system is,” Allen says.
“[H]aving them here helps us re-energize and focus our efforts. I think you can see the excitement from the players in the system. It’s a great opportunity.”