Figuring Out What Makes the Best-Managed Cities Tick

OAKLAND, Calif. — One of the common themes in the world of municipal best practices is how to take the pioneering models of what works and adapt them for use in other local governments. While every jurisdiction is different, there’s increasing interest in trying to pinpoint formulas, methods and applications that are common in the highest-functioning and best-managed cities.

“We’re trying to get a better picture of what good looks like,” Beth Blauer, executive director of the Johns Hopkins University Center for Government Excellence, said at last week’s 2015 Code for America Summit at the Oakland Convention Center.

Blauer is one of the leading forces behind What Works Cities, a $42 million initiative launched in April and spearheaded by the Bloomberg Philanthropies in conjunction with Results for America, the Johns Hopkins University Center for Government Excellence, the Government Performance Lab at Harvard University’s Kennedy School of Government, the Sunlight Foundation and the Behavioral Insights Team.

The non-profit and academic coalition is in the process of identifying 100 so-called “What Works Cities” around the United States and announced the first eight participating jurisdictions in August—Chattanooga, Tennessee; Jackson, Mississippi; Kansas City, Missouri; Louisville, Kentucky; Mesa, Arizona; New Orleans, Louisiana; Seattle, Washington; and Tulsa, Oklahoma.

The What Works Cities program connects the participating cities with technical assistance on how to better use their data to make more informed policy and operations decisions.  

“The idea is that we’re creating maturity models here,” Blauer said during a Code for America Summit workshop featuring officials from three of the previously announced What Works Cities, Louisville, New Orleans and Seattle. “The idea is to look at what those best practices are and bring them back to the cities,” she said.

Around 110 cities had applied to be part of the program and additional cities will be named on a rolling basis through 2017.

Blauer detailed some of the characteristics of the cities that they’re looking at closely for future inclusion, noting that “the cities already on this panel were already advanced” in their data practices, referring to the representatives from Louisville, New Orleans and Seattle at the workshop session.

While the What Works Cities program is “looking for champions below the mayoral level” interested in data-driven approaches to municipal management, she said, the program wants mayors to apply as a sign of “commitment,” in particular, mayors who have already embraced many of the program’s driving principles.

Additionally, the program is looking for “stability,” Blauer said, so if the end of one mayoral administration is in sight, What Works Cities will likely choose a city with executive leadership that will be sticking around for few years.

Data, performance and analytics are central to the What Works Cities program.

Jackson and Mesa will be implementing open data practices for the first time while the rest of the eight initial cities will be strengthening their existing open data practices. Jackson and Tulsa will also be implementing mayoral-led performance-management programs while Chattanooga, Kansas City and Mesa will build upon existing programs.

“I’m a firm believer in when you have facts in front of you, it is transformational,” said Seattle’s deputy mayor for operations, Kate Joncas, who was part of Thursday’s What Works Cities workshop panel, along with Michele Jolin, CEO of Results for America; Theresa Reno-Weber, chief of performance and technology in the Louisville Metro government; and Oliver Wise, director of performance and accountability in city of New Orleans.

Joncas said that Seattle Mayor Ed Murray has emphasized efficient and effective public services as part of his administration. A big problem Seattle City Hall struggles with, Joncas said, is homelessness.

That’s not for a lack of trying to address the pervasive problem. Seattle spends roughly $40 million on programs to assist the city’s homeless population—a January 2015 headcount of the homeless population in King County, which includes Seattle, found 3,772 men, women and children living without shelter.

City funding for homeless services is distributed through 400 separate contracts.

It’s easier to count the number of sandwiches that are handed out, Joncas said, than assess the effectiveness of programs in getting the city’s homeless people off the streets and into more stable housing situations. And the What Works Cities program will help Seattle ascertain the costs and benefits of its contracting strategy.

In a follow-up workshop session on municipal data analytics on Thursday, Blauer and Wise joined Mike Flowers, New York City’s first chief analytics officer under then-Mayor Michael Bloomberg, and Abhi Nemani, former chief data officer for the city of Los Angeles.

Wise, whose efforts leading analytics and performance management efforts in New Orleans—particularly BlightStat—have been nationally recognized, stressed that the tangible results of a municipal data program isn’t simply a “giant map” that displays municipal data.

As Wise described it, the data analytics push at New Orleans City Hall “was about a process and a shift of perspective in how that data could be used.”

He also said that with data analytics, there’s a critical distinction when looking at those practices when compared to traditional performance management methods.

“It’s a different mindset than performance management which is about ratcheting up tension,” Wise said. “Data analytics is about ratcheting down tension.”

In other words, data analytics dissects and assesses the methods and information needed to manage municipal programs and make decisions where performance management is often focused on the pure results.

“Analytics is the really next logical step for performance management shops” in city halls that have them, Wise said.

The first big steps in the creation of a city hall data analytics operation aren’t necessarily about going out to find an advanced IT analytics platform, Wise said.

It’s far more important getting the right people within a municipal government to the table from the start, and that includes the technologists, policy leaders and data practitioners.

“When you get the right people in the room, the friction melts away” when everyone sees the value in the transformational opportunities that come with better data leveraging, Blauer said during the first session.

“The conversation should come first, then the data,” Wise said, echoing Blauer’s statement.

As things go along, one important goal in creating a shared vision within a municipal government regarding data and analytics is to create “a loop,” Reno-Weber said, “where people put in data and then use their own data, then they audit it themselves.”

Keeping things simple is also important, panelists at both workshop sessions agreed. A data analytics push for municipal governments shouldn’t be about an expansion of data collection for departments. That can add to the workloads of already overburdened civil servants.

If you want a department to collect new sets of data, you need to “illustrate the problem that you’re going to solve,” Blauer said. And that shouldn’t come before harvesting the low-hanging fruit.

“Inventorying the data that exists will get you where you want to be,” Blauer said.

Flowers noted that there’s already data out there that isn’t being used, pointing to a predictive analytics project with roots in New Orleans that used existing census block data to figure out areas of the city that were less likely to have homes with smoke detectors.

The company Flowers now works for, New York City-based data analytics company Enigma, worked with New Orleans to turn that project into a free, open-source tool that any city government can use to better target efforts to encourage the use of home smoke detectors or distribute free smoke detectors.

It was “crappy federal data that we turned into an actionable tool,” Flowers said, describing the relatively low-tech data approach for creating the tool, called Smoke Signals.

“There is existing data with exceedingly high value if used in the right way,” he said.

RELATED on Route Fifty:

• “Stats Surge Continues in Cities and States”
• “New Analytics Tool Helps Predict Where There’s a Lack of Home Smoke Detectors”

Complete Coverage of the 2015 Code for America Summit.

• “At Code for America, Hope for Procurement Reform Amid Ongoing Struggles”
• “Sometimes You Don’t Need to Think Outside the Box”