With Big Data Comes Bigger Goals

Today’s performance management tools eliminate the old ways of thinking about what government can and can’t do.
November 2014
New York Police Commissioner Bill Bratton implemented CompStat in the 1990s. Flickr/ Policy Exchange
By Mark Funkhouser  |  Publisher
Former mayor of Kansas City, Mo.

I once had a police chief tell me that there was nothing that he and his department could do about overall crime or citizen safety. He wasn’t alone in his thinking. When I started working in performance auditing decades ago, getting government organizations to accept responsibility, even for outcomes directly related to their own missions, was difficult. It was seen as unfair, for example, to hold an employment training agency accountable for its graduates getting jobs because so many other social and economic factors contributed to that outcome.

In the last few years, however, things have changed. I now hear agency directors and elected executives talking about “moving the needle on population-level outcomes,” such as reducing the incidence of childhood obesity or, more broadly, increasing the overall health of a community.

It all started with the late Jack Maple. In a new book, The PerformanceStat Potential, Robert Behn describes how in the early 1990s Maple, then a lieutenant with the New York City transit police, scrawled out on a napkin the basic outlines of what became CompStat, a strategy of using computerized data to predict and prevent crime.

When Bill Bratton became New York City’s police commissioner in 1994, he promised that the police department would reduce crime by 40 percent in three years using the CompStat methods. People said Bratton was crazy, but when he, Maple and the NYPD succeeded, it opened the way for a powerful new way of thinking about government performance. Now county executives, mayors and governors everywhere are hiring people like Theresa Reno-Weber, chief of performance and technology for Louisville Mayor Greg Fischer. “If you’re not talking about population outcomes, what are you talking about?” says Reno-Weber.

This new way of looking at government performance focuses on the government executive’s role not only as a manager but also as a community leader. An example in Louisville is “55,000 Degrees,” an effort to increase the number of working-age adults with college degrees by that number by 2020 to support economic development. Fischer has this as one of his strategic goals, and he chairs the board of the community organization guiding the effort with support from Reno-Weber’s office.

As Louisville’s approach suggests, improving population outcomes requires ever-broader cooperation and sharing of data among organizations. The data sharing has to be very sophisticated, allowing agencies, nonprofits and individuals to see the information they need without running afoul of privacy rights. As Beth Simone Noveck, a professor at New York University and former U.S. chief deputy technology officer, has written, we need to be able to use technology to allow organizations to work with people rather than for people.

Increasingly, mayors and governors are on board with that, and the old ways of thinking about what a government agency can and can’t do don’t cut it anymore. Behn writes that Bratton used to tell mayors: “If you have a police chief who can’t get crime down, get yourself a new police chief.”