The trick is getting front-line managers to see value in fostering a learning-driven, problem-solving approach to improving performance.
It was an early morning meeting in the New York City Police Department’s cavernous meeting room in the mid-1990s, filled with police officers, but also with support staff, analysts, representatives from the city’s housing police, the transit police, the district attorney’s office, and more. Legendary Police Commissioner William Bratton presided at one end of the U-shaped table with a podium in the center and two screens behind. A precinct commander for one of the city’s 77 police precincts was at the podium.
Watching this scene was instructive. Bratton had a summary sheet of data on what was going on in terms of crime, arrests, and training of officers in each precinct and called precinct commanders to the podium on a regular basis to question them about progress and problems. While it looked intimidating, it was not just an accountability exercise, it was a problem-solving session. One commander was asked why his officers were not on bike patrols and he said they hadn’t been trained. Bratton turned to the department’s HR director and barked “train them by next week.”
Bratton wanted not only to hear from his officers, but also to ensure he could help them bust bureaucratic barriers within his department as well as across the tangled law enforcement landscape across the city. And his small team of analysts kept digging through the data. For example, they found petty crime peaked on the subway system between 3 and 5 p.m.—when teens were out of school but before their parents got home from work. So Bratton boosted patrols in the subway system during those hours and petty crimes went down.
Overall, his novel approach—called CompStat—reduced major crimes in the city by 39% within two years after it was introduced in 1994. And his approach quickly spread to other city departments, to other cities, to state governments, and ultimately to federal agencies.
The Approach Spreads
Baltimore mayor, and later Maryland governor, Martin O’Malley pioneered using this approach across entire governments, which he dubbed as CitiStat and StateStat. His use of this approach became part of his signature leadership style that was seen as results-oriented.
Harvard professor Bob Behn dubbed the broader trend in the use of this approach as PerformanceStat. In his observations, he felt there is too much of a focus on the visible features of PerformanceStat (data, projectors, meetings) versus sorting out the underlying cause-and-effect of how it works. He concluded that it is better seen as a leadership and learning strategy, not a standardized, mechanistic system.
He defined PerformanceStat as: A leadership strategy that is designed to achieve specific public purposes, where the leadership team persists in holding ongoing series of regular, frequent, integrated meetings.
At these meetings, the leadership team:
- Uses current data to analyze specific, previously defined aspects of recent performance.
- Provides feedback on performance versus targets.
- Follows up on previous decisions and commitments to produce results and learn from efforts to improve.
- Identifies and solves performance-deficit problems, and sets the next performance targets.
Adoption at the Federal Level
At the federal level, the PerformanceStat approach has been used effectively as well. For example, in 2010, the Department of Housing and Urban Development joined with the Department of Veterans Affairs with the audacious goal of eliminating homelessness among veterans by the end of 2015. It used the PerformanceStat approach to target efforts in 21 communities with the largest number of homeless veterans. Both departments sent in joint, top-level teams into each of the 21 communities, developed a 100-day plan of action, and worked across organizational boundaries to overcome barriers in order to meet their goal.
The deputy secretaries held joint PerformanceStat sessions to both understand and jointly solve problems. For example, HUD provided housing vouchers to veterans, but analysts found they were not being used. After some investigation, the analysts realized that a security deposit was required to rent an apartment, and the voucher program could not legally provide those monies. But VA had funds and authority to do so, and once this was figured out, the number of vouchers being redeemed went up—and homelessness went down. While HUD and VA didn’t meet the goal to eliminate homelessness among veterans, they substantially reduced the number—by 47% by 2016.
In recent years, the PerformanceStat approach has been used episodically across federal agencies as a leadership strategy, oftentimes by deputy secretaries in their role as their department’s chief operating officer. However, a variant of PerformanceStat as a learning and problem-solving approach has evolved across all agencies.
Agency Strategic Reviews
Since 2014, the top leadership of each federal department and major agency holds a strategic review each spring, meeting with OMB to discuss the progress of their strategic objectives—the core building blocks and primary units of analysis of their four-year strategic plans. Governmentwide, these strategic reviews cover over 300 strategic objectives. Agencies and OMB jointly assess progress in order to inform next steps in budgeting, changes in implementation strategies, or other operational and administrative actions to improve program and organizational performance.
The reviews are ways to learn what works and what changes may be needed to improve. Both OMB and agency leaders find the review process helpful. The reviews consist of both a core set of analyses and assessments of progress of agency strategic objectives that are stable from year to year, and special areas of focus that change from year to year. For example, OMB’s guidance for the strategic reviews in 2018 highlighted three areas of special attention: mission assessment, risk assessment, and management priorities. In 2020, the special attention will likely be on what agencies have developed for their new statutorily-required “learning agendas” and how these agendas can be used to inform the development of the next cycle of 4-year agency strategic plans.
While accountability-focused performance management is a powerful tool for driving change in certain settings, learning-focused performance management also can be a powerful tool because it concentrates on evidence-based, joint learning among participants about understanding what works, and why, and what actions are needed to solve performance problems.
The GPRA Modernization Act of 2010 institutionalizes a set of routines—quarterly reviews of agency and cross-agency priority goals, annual reviews, and assessments of strategic objectives—and designates responsible officials for overseeing them. These routines can be a platform for either accountability-based or learning-based approaches to performance management at the tops of agencies. But what works best at lower levels in large organizations?
Key Lesson and the Next Challenge
Top leaders in federal agencies have embraced the use of evidence-based learning routines reflecting the principles of PerformanceStat over the past decade. The key lesson has been that continuity in the use of a defined administrative routine makes a difference in achieving performance improvement.
The next challenge for agency leaders today is to find ways to cascade the use of these types of evidence-based learning routines down within their organizations to front-line managers.
The trick will be to get front-line managers to see value in fostering a learning-driven problem-solving approach to improving performance. But this has not been easy. For more than 20 years, the Government Accountability Office has periodically surveyed front line managers as to their use of performance information and it has typically been only about one-third who respond positively.
Fortunately, the recently passed Evidence Act and the increased availability of performance information that is more granular, real-time, predictive, and intuitive. This could make it more likely that line managers will use it as a learning tool and tie it to how they solve problems in their day-to-day work, such as making better decisions on administering benefits, improving traffic safety, or reducing crime.