OAKLAND, CALIF. — When the city of Los Angeles released data on trash collection, the L.A. Times followed it with a story about how the service was worse in low-income areas.
And that immediately made other departments wary of opening their own data.
It’s not just L.A. That’s been a problem in many places when it comes to opening up data and making it usable for government, residents, nonprofits and the private sector — some people are simply afraid that the data will make them look bad.
“There are people in the community who have been wanting to know certain pieces of data forever, and so I try to be the conduit to the department, like, ‘You really should publish water shutoffs because that’s a huge issue that people are concerned about,’” said Joel Howrani Heeres, director of open data and analysis for Detroit. “But that doesn’t necessarily resonate, and then there’s a fear of the bad story that prevents them from publishing that data.”
And indeed, many open data sets can paint an unflattering picture of government work — at one point during the 2016 Code for America Summit this week, one presenter showed a series of maps demonstrating how police in Oakland, Calif., tend to focus drug enforcement efforts in minority-heavy parts of the city.
Also at the summit, two people working with government on data projects talked about how to get away from that fear and show people the value of data.
A lot of the time, according to Vyki Englert of Hack for LA, it comes down to trust. Building a relationship with a department based on data can make people more comfortable with the idea of sharing. A good place to start, she suggested, is with simple data governance: Are the boundaries of a water district usable if they’re made available outside a department?
“Start somewhere simple, get the process lined up," Englert said, "and once people are starting to see and feel the value of something that’s not sensitive data and could never be sensitive data … because they’re comfortable with sharing already, [then] it’s not two things for them, which is, ‘Oh, you’re going to share and you’re going to share sensitive information.’”
A lot of that work can be done in small chunks. John Gravois, a developer advocate with Esri, said he likes data outreach events as a means for collecting feedback that can make data more useful and wedge open the door to sharing. People who have worked with a data set for a long time might have slathered it in acronyms that are meaningful only to a small group of people, for example — an audience not familiar with the data could point that out, or suggest other ways to put the data in context.
“If you respond to those sorts of requests for improvement, you can build a really solid relationship,” Gravois said. “And I think doing that around data sets that are more benign is the baseline standard for developing a more constructive relationship going forward and tackling some of the stickier problems.”
Even as “the bad story” from one department can inspire fear in another, Englert said that success stories can also help bring more people around.
“In L.A., in California, we’re lucky, because some of the folks leading this work are actually the Sheriff’s Department in Los Angeles County, the Department of Justice and the state Health Department — the last people you’d expect to jump on board with open data,” she said. “We have good examples so we can go, ‘If the sheriff’s department can do this, you can do this.’”
The successes don’t have to be big, either — it doesn’t need to be a huge, comprehensive data dump that revolutionizes the way a department works. Oftentimes what matters at the ground level is process improvement; things that save time and money or make somebody’s job a little easier.
“It’s not always about the big stuff," Englert said, "and the more people see those little stories, the more they’re like, ‘Oh yeah, well I have a little story too. Let me tell you about this little thing we did, and now … the clerk’s office is using a more transparent FOIA system and they’ve saved 10 work-hours a week because now they’re using open data.'"
NEW ON THE PODCAST