Code For America’s CTO Michal Migurski is helping cities use data to unlock some of the biggest questions. A U.C. Berkeley grad, majoring in cognitive science with some computer programming thrown in, he got his start running code for Stamen Design back in 2002. Along with founder Eric Rodenbeck, who sports a trademark trilby, he zeroed-in on bringing data visualizations to the next level in remarkable ways. A map of San Francisco’s taxi cabs, for instance, made it all the way to the Museum of Modern Art. Who knew data could be beautiful?
As Migurski tells it, data projects like the ones he executed while at Stamen are no small feat. Take “The City from the Valley”: Stamen noticed how San Francisco tech workers were being pumped to and from Silicon Valley and decided to investigate. Migurski and his team started counting and clocking the buses: trips, stops, companies, and frequency. Sometimes they turned to bloodhounding for numbers, hiring bike messengers to tail buses wherever they went. The results turned raw data into bubbly arteries of traffic, simpler to the eye, like a subway map. Where’s the pulse of the city? Look no further.
But given all the city’s problems, even the ones that produce crunchable numbers, what good’s a map?
“Around the office, we used to call it ‘An object to think with,'” Migurski says. It’s a tool to ask new question with, he says, and interrogate a larger space. In general, it can be an easy and often devastating way to show imbalance or patterns. When solving a problem, sometimes it’s best to work in from the edges.
In 2007, the Stamen team took on a starker data viz challenge: crime in Oakland. Using the city’s crime data API, they built Crimespotting, an interactive map to show crimes, when and where they happen, across the city, pins standing in for robberies or disturbances of peace. It effectively replaced the city’s slow and dismal Crimewatch site and used the datasets more effectively.
Crimespotting did get the Stamen into trouble though: they pinged the Oakland Police Department data site so frequently, the city shut them down until they could understand where all the data requests were actually coming from and why.
With civic tools and data, Migurski says, the key is set up a call and response between a user and the information. “If you have a question,” he says of citizens, “it’s important you feel you are being listened to and that [your interests] will be acted on.”
In 2013, Migurski left Stamen Design and joined Code for America. He now finds himself working head-to-head with agencies who access massive amounts of data, pulling it into giant repositories. “It’s like touching the source,” he says of his current role.
The biggest slog, he says, is getting local governments to understand how they can put two-and-two together, one data-set with another, to make connections and epiphanies. But it’s also difficult, Migurski adds, to get developers, “edgy,” on the cusp types, to slow down and let cities catch up. Sometimes they’re dealing with systems 10 to 15 years behind. Everyone suffers, he says.
“It’s a buying vs. building problem,” Migurki says.
He explains that governments frequently buy new technologies they don’t know how to work or why they’re using it in the first place.
But slowly, cities like Oakland are seeing solutions in open data, braving it themselves or hiring IT teams to build better tools from scratch. Migurski says Code for America looks for this brand of maturity and resourcefulness, cities ready to take off the training wheels. (Over the years, Oakland’s Crimewatch map has been overhauled and given more attention and now outpaces Crimespotting.)
Migurski explains the process behind some of Code for America’s civic data projects. It usually begins with the problem and then works backwards with data to solve it, not the other way around, he says.
The Oakland Brigade, for instance, launched RecordTrac last year: a way for the track down and request public records online, from storm drain damage reports to documents of police brutality.
It used to be that offices would “shove [record requests] on each other just to waste each other’s time,” Migurski says. Now, projects like RecordTrac cut down the frustration and drama, involve fewer moving parts, and use this data as a different kind of weapon.
Migurski argues that with projects like this, Oakland is showing it’s not just a “little brother city.” The norm of data literacy, he says, is also more leading to sharper journalists and citizens, who are asking more and getting used to the recent explosion of data tools and ecosystems.
Migurski imagines open data tools will someday soon morph into the ease of Google search engines. “Something a lot more human,” he says. (He doesn’t however put a lot of stock in the whole “quantified self” trend, and personally finds it exhausting. “It always turns into a ‘what color is your wristband?’ conversation,” he says.)
Communities, even if they start small, he says, are seeing you don’t have to be a “superhacker” to bend data and code, to make it do what you want. He points out that all he started with, fresh out of school, was a web browser and an Excel spreadsheet. Migurski looks back at the “scrapping problem” of projects like Crimespotting, pulling down data without talking to anyone. “It was like Robin Hood,” he laughs.
There is always the threat of getting lost in the weeds, he says, but open data doesn’t have to mean an “info dump.” Open data, in fact, can be about narrative, what Migurski describes as “story vs. dots” on a map.
Oakland is just beginning to write its own story, its own code, and Migurski’s work is an important part of the equation.