Agencies improve management through analytics, report says

Federal agency managers are demonstrating measurable improvements in programs by integrating data-based analytics in decision making, according to a study released Wednesday by the nonprofit Partnership for Public Service and the IBM Center for the Business of Government.

The study, "From Data to Decisions: The Power of Analytics," reviews managers' handling and results in seven federal programs spanning eight agencies, in which participants knew "they needed to change agency culture to take full advantage of an analytics mind-set." The research comes as agencies are gearing up to implement the 2010 Government Performance and Results Modernization Act.

"In the midst of the tremendous fiscal uncertainty the nation now faces, and with public attitudes toward government at an all-time low, it is more critical than ever that federal leaders base their decisions on accurate data and not on anecdotes, incomplete information or the belief that things will work out for the best -- particularly when those decisions have huge consequences on tax dollars and society," the report said.

The Partnership and IBM defined analytics as "the extensive and systematic use of data, statistical and quantitative analysis and explanatory and predictive models to drive fact-based actions for effective management."

The report, aided by research from IBM's Public Sector Business Analytics and Optimization practice, was welcomed during a Wednesday panel discussion by Shelley Metzenbaum, associate director for performance and personnel management at the Office of Management and Budget. She echoed the Partnership's hope that such reports mark "the beginnings of a movement." She distinguished between the business world's analytics and "mission analytics" needed by agencies. The government "has long excelled at gathering large quantities of data, but it has not been good at analyzing and using them," Metzenbaum said.

The programs examined in depth in the report are the Veterans Affairs Supportive Housing initiative, a joint effort by the Housing and Urban Development and Veterans Affairs departments; the Safety Management System in the Federal Aviation Administration; and the Medicare nursing homes and transplant programs, run by the Health and Human Services Department's Centers for Medicare and Medicaid Services.

Profiled more briefly are the U.S. Coast Guard's Business Intelligence Tool; the National Highway Traffic Safety Administration's Click It or Ticket seat-belt campaign; the Navy's Naval Aviation Enterprise; and the Social Security Administration's use of mission analytics in customer service.

"Regardless of the level of sophistication in their analytics programs, data analysis helped provide insights into how to improve programs," the report said. "We learned that data is only the starting point. The data needed to be analyzed, turned into information, and made accessible to staff and executives, and the data also needed to meet varying needs and be understandable to different audiences."

How well an agency exploits such quantitative tools depends on its "analytics maturity," the report said. "Common elements for assessing maturity include the quality and accessibility of data, organizationwide focus, leadership involvement, tools and technology and staff talent."

As agencies developed performance measures to assess progress, researchers "found that those measures changed over time and it was important that they stay meaningful and reliable and are tied to results." It was also vital that staff "had a clear line of sight from where they stood to the desired goals and outcomes" and that agencies "cultivated and leveraged partnerships across the agency and with partners who deliver services."

In the Medicare nursing program, for example, the increased use of data and transparency led to a reduction in the frequency of patient bed sores.

In the veterans housing program, progress now is measured not in the number of housing units funded but by the number of veterans who are placed in housing, considered a more accurate gauge. In FAA's air safety program, standardization of data definitions industrywide and protection of sensitive data on aircraft mishaps and near-misses allowed analysts to more rapidly identify trends and develop safety benchmarks.

In the Coast Guard's intelligence program, inspection staff using data from multiple sources were better able to allocate resources to port, waterway and coastal security without neglecting conventional marine safety.

And the Social Security Administration showed success using analytics to fine-tune the satisfaction surveys it regularly conducts to gather feedback from customers.

The highway safety agency used analytics to help it arrive at the proper balance of education and enforcement in persuading states to up the rate of seat-belt use, which has risen from 40 percent two decades ago to 85 percent today.

"An analytics program does not have to start at the top," the authors wrote. "Performance improvement efforts can start small and grow via a pilot program approach, but it's important that efforts are tailored to fit the agency's culture and that of its constituents, as well as meet mission requirements."