The Stat Approach

Baltimore's CitiStat holds agencies' feet to the fire.

In September, the Office of Management and Budget released its final evaluation of the performance of federal programs. This was the last act of OMB's five-year-old PART endeavor, the catchy acronym standing for a clunky title: Performance Assessment Rating Tool. PART found that 80 percent of 1,017 federal programs were performing acceptably. The other 20 percent either could not demonstrate results or were judged to be failing.

The program, an important element of a serious Bush administration effort to focus on management, made great strides toward systematic performance measurement, although it never was embraced by Congress and so had little influence on budgeting decisions.

PART won the Harvard Kennedy School Innovations in American Government Award in 2005. This signaled that accountability, once the sine qua non of public management, had given way to performance as the holy grail, though I suppose one could argue that the latter is a subset of the former. The ascendance of performance had been telegraphed when the innovations awards crowd crowned Baltimore's CitiStat as a winner one year earlier.

CitiStat's principles have a longer, and more closely examined, history than PART, its ancestry tracking back to the CompStat program former Mayor Rudolph Giuliani started in New York City to identify and address high-crime locations. So, when invited by Baltimore Mayor Sheila Dixon to a conference about CitiStat, I thought I might pick up some lessons of interest to the federal management cadre.

The conference started off with an actual CitiStat weekly meeting, this one to examine the progress of CleanStat, Dixon's effort to enlist several city departments in meeting her priority goal of a "cleaner, greener, healthier and safer" Baltimore. "Grime and crime go together," said CitiStat's leader, deputy mayor Christopher Thomaskutty.

On the hot seat was Valentina Ukwuoma, head of the Solid Waste Bureau, and her deputies. Why, inquired the inquisitors, was the Poplar Grove area of the city, though targeted for increased city services, suffering from 65 overdue high grass and weed requests, 35 overdue cleaning requests, and 58 overdue forestry requests?" And why was one of eight graffiti-removal crews closing only 1.81 service requests per hour, while every other was above 2.5 and the highest was at 3.4?

Such questions underline the emphasis on objective statistics in the "stat" approach. And this level of street detail is what makes CitiStat work in Baltimore's 80 square miles. But, one wondered, how would it translate one or two levels up in the federalist scheme?

The beginnings of an answer are coming out of Maryland's capital, where Gov. Martin O'Malley, who invented the CitiStat program while serving as mayor of Baltimore, is now striving to install State-Stat in Annapolis. At the conference, O'Malley and a top aide, Matt Gallagher, voiced frustration with slow progress at the state level. Departmental secretaries were not used to providing details about their agencies' performance, and as Gallagher observed, it's difficult to hold a single agency accountable for such goals as mitigating pollution in the Chesapeake Bay or ensuring a child's well-being. Still, as meetings with secretaries progress, the governor's expectations become clear: Programs need to measure and im-prove their performance, and their leaders will be held to strict account.

At the federal level, these challenges are compounded. Few agencies can claim they are solely responsible for social, environmental or other outcomes. But the insistence by O'Malley and Dixon that agencies do better, and the personal time they invest in this pursuit, offer examples that the next president and leaders of Congress could follow if they're seriously interested in pursuing the goal of a high-performing government.