Hearing of the
Subcommittee on Rules and Organization of the House
Subcommittee hearing on "The Government Performance and Results Act and the Legislative Process of House Committees."
Executive Associate Director & Controller,
U.S. Office of Management & Budget
Mr. Chairman, I appreciate the opportunity to appear before this Subcommittee, and commend the Committee on its review of the Government Performance and Results Act.
Why GPRA Matters
As one who has worked both in business and government, I am respectful both of the similarities and the differences between the two. Unlike business, government is not measured by a profit margin. Whereas the health of any business will ultimately be judged by a comparison of revenues and costs, in government that comparison is simply not meaningful. In no case is a government agency or program established solely or even primarily with the goal of maximizing net receipts. The purposes of government are far more varied: to provide for our common defense, to help those who cannot afford to help themselves, to protect our environment, and dozens of other purposes.
That's why GPRA matters: it forces all of us -- from agency managers to OMB and Congress -- who are responsible for stewarding the public's resources to focus on our "bottom line", performance and results. It requires agencies to create goals and performance standards, and to establish measures so we know whether and how we met them. GPRA reinforces accountability: What results are we as managers and decisionmakers getting? How well are we doing in implementing programs? GPRA helps us tell the American public what we are getting from our government.
In one sense, what GPRA requires is not new. We have always asked for, and used, performance information. In our role as stewards of the efficient use of resources, both OMB and Congress have always sought to know what programs actually do. We have always asked "How many people are helped?", "How long will it take to complete a project?" and "What effects will this program actually have?' Nonetheless, GPRA does establish an objective, a process, and a language with which to communicate about performance and results. The objective is clarity about the actions and effects of government programs. The process is the creation and review of strategic plans, performance plans, and performance reports. The language is the vocabulary of inputs, outputs, and outcomes: by using these processes and this language, we constantly challenge ourselves to tie decisions about resources and program management to actual performance and results.
My testimony today will focus on OMB's and agencies' experience in developing and applying performance information to budget and program reviews, and the lessons we've learned from it. I discuss this experience hoping that the lessons we have learned will be helpful to the Congress. OMB's work at implementing GPRA have been intensive and longstanding. Each year, we have worked to improve and use agency performance information in our internal discussions on agency funding levels. At the same time, we've worked actively with agencies to improve their performance plans and to increase their own use of GPRA information. Within OMB, the same analysts who work on a particular agency's budget and management issues every day also review the agency's performance reports and work with the agency on GPRA implementation.
How Congress uses the information provided by agencies under GPRA is, obviously, a matter for Congress to decide. Nonetheless, OMB's own experience may be helpful. OMB in its oversight role must, like the Congress, review agency activities. Like the Congress, we do so with a relatively small staff. And, like the Congress, we are always looking for accurate information.
LESSONS LEARNED
1. Walk Before You Run
GPRA can improve accountability and program execution, but to do so it requires changes in resources, processes, and systems. These changes are neither quick nor easy. They take time. Other countries that have introduced institutional and management practices similar to GPRA have cautioned that it will require years to take hold. They are right.
The changes GPRA is bringing about are not confined to programs, their design and their budgets. Internal agency management must be changed, too. Some agencies are beginning to do so. In a number of cases, major program examinations/evaluations are being designed to help identify contemporary needs, better program delivery mechanisms, and outcome measures. Some agencies are beginning the further step of integrating their financial management and reporting systems with program performance management systems. We have made important strides in a relatively short time, but we all recognize that much more remains to be done.
These improvements are not easy. They are not always successful on the first try. But they always take time. Fortunately, the Congress in drafting the law provided for a phased and progressive introduction of its various parts. Even more important, both GAO and the Congress in reviewing our progress have generally recognized that these changes will take time. In this business, patience is more than just a virtue -- it's a necessity.
2. Performance Measures Evolve
Government, like everyone else, learns by doing. As we have worked to implement GPRA and use it in decisionmaking, we have recognized that planning and reporting is a dynamic process. Circumstances change, and programs change with them. Experience and management improvements will bring about change as well. This is why OMB has not treated performance measurement as a static process. An agency's performance goals must not be frozen in time. They must be refined, changed, dropped, and augmented. If we freeze requirements too quickly, we will get the wrong measures. For example:
- The Department of Transportation recently changed its goals for its maritime safety program. In FY 1999, the Coast Guard set a goal of reducing the fatality rate per 100,000 workers aboard commercial vessels. In doing so, it became clear that over the course of the past decade that the worker fatality rate had declined steadily. However, at the same time, the rate of passenger vessel casualties had increased. For FY 2000 and 2001, DOT moved to focus on this problem and set a goal for reducing it.
-
The Occupational Safety and Health Administration in the Department of Labor originally measured its performance in terms of the number of inspections made and citations issued. Then OSHA realized that, since its real goal was to reduce workplace injuries, they would be more effective by focusing on that measure. They have since redesigned their programs to target resources on workplaces that have a greater risk of injury; as a result, OSHA is both more effective and imposes less burden on parts of industry that have safe workplaces.
-
HUD is another example. If you compare HUD's FY 1999 performance plan with the one two years later, the improvement is dramatic. Performance goals for HUD's "empowerment communities" objective in FY 1999 focused on many output measures, such as developing standardized assessments, developing tracking systems, and maintaining historic levels of households assisted. The FY 2001 plan adds outcome goals and measures: HUD proposes to track and hopes to decrease disparities in the ratio of city to suburban unemployment rates; and reducing the unemployment rate among young, entry-level job seekers in central cities.
As we have worked with the agencies to implement GPRA, we have also begun working to develop some measures that apply across agencies. Although much more needs to be done, we are pleased that even in these early stages we've been able to use this information to improve government management:
- The Government buys approximately $200 billion annually in products and services. Working through the interagency Procurement Executives Council, we recently adopted a set of performance measures for core acquisition systems. This "Governmentwide Acquisition Performance Program" will enable the Government to assess -- for the first time -- its ability to obtain the best value products and services on a timely basis, while maintaining the public's trust and fulfilling public policy objectives such as promoting small business participation.
-
Federal delinquent debt is a significant problem and the focus of considerable Executive Branch and Congressional attention. To improve management of Federal delinquent debt, we've been working with the agencies to continually track and report delinquent debt using a common measure of delinquency ("debt delinquent more than 180 days").
As we continue to work, we want to expand the range of goals and measures, from inputs to outputs and outcomes. Output goals still predominate in most agency plans, but the number of outcome goals is increasing. At the same time, we need to make additional progress in developing goals for cross-cutting programs. And we need to increase coordination with States and local governments, who directly deliver Federally funded services and benefits.
We at OMB are also trying to prove you "can teach an old dog new tricks". As we review and respond to the agencies, we have changed our requirements for performance plans, to help make sure the information in these plans is useful to the Administration and Congress. For example:
- We have started requiring information covering the performance over a period of years, rather than the single year that GPRA requires; this helps to highlight trends and changes that would otherwise be missed. The FY 2001 performance plans now include at least three years of performance information.
-
Second, OMB responded to Congressional and GAO interest by having the performance plan address major management issues more extensively, and include an expanded discussion of data quality issues.
5. Active engagement.
This last example makes our next point: that these process require active communication and engagement to work. When the Congress or OMB raises questions or concerns, agencies can and will make changes. Both OMB and Congress need to continue to engage with agencies about their plans and reports, and about their use of performance information in everyday activities.
About 100 departments and independent agencies prepare GPRA plans and reports. OMB has worked with virtually every one of these agencies over the past four years on their GPRA efforts. And we know that Congress, GAO, and groups like the National Academy for Public Administration (NAPA) have worked with many of these agencies as well. We now have another opportunity to do so as agencies are submitting their first set of annual performance reports at the end of the month. Providing feedback to agencies is vitally important to make sure agencies' plans and reports are useful to decisionmakers at all levels.
6. Be Flexible
GPRA is not a one-size-fits-all proposition. No single template would meet the needs of 100 different agencies, many of which possesses hundreds of disparate functions. Agencies' programs and goals vary, and their performance and result measures should, too. The performance measures for, say, processing disability insurance claims in Social Security are markedly different from the measures used to award grants at the National Institutes of Health.
Different levels of performance information apply to different audiences. We want performance information to be used by everyone from individual program managers to Cabinet secretaries, to those who provide oversight, OMB and the Congress. But the kind of information and the level of detail will vary for each. A plant manager at General Motors tracks many physical operations in detail; the reports he sees and uses are quite different from those presented to the company's CEO and stock holders. The performance information the company's Chairman and Board need to make decisions about overall corporate strategy, and that its shareholders need to evaluate the results of that strategy, differs considerably from the information a plant manager needs to improve the plant's operations.
The Federal government is no different. The level of information the head of OMB uses to, for example, evaluate the effectiveness of VA health care differs from what the Secretary, the head of VA Health, or the head of an individual hospital might use. Although individuals at all levels are concerned with all of VA's performance measures, individual hospital administrators primarily track their performance in screenings for colorectal, cervical and prostate cancer, mammography, and the like. The Veterans Health Administration headquarters focuses on a chronic disease care index, which measures adherence to industry practice guidelines for diseases affecting veterans. And the VA Secretary or OMB would evaluate overall improved health care to veterans, as measured by, for example, an index of how well VA is doing in preventing disease.
From an OMB standpoint, our guidance and instructions have given agencies the flexibility they need from the very start. Performance planning should be a bottom-up process: we did not develop a model plan for the agencies. We chose instead to encourage agencies to think on their own about what was best for them, and to work with them by providing feedback and guidance.
OMB also allows agencies substantial discretion to combine plans and reports. For example, agencies may combine their annual plans with their annual performance reports, or their annual performance reports with their accountability reports. Ultimately, our goal is to encourage agencies to integrate financial information and performance information into a single accountability report. This will present a complete picture of each agency's program and the resources used to achieve it.
7. Integrate performance measures into resource and management processes
Unless GPRA is tied to decisions by agency management, OMB and the Congress, it risks becoming a paper exercise. In our view, GPRA was intended to do much more than require the writing of reports; to be effective GPRA efforts must become tools for management and oversight. As OMB works with agencies to develop their performance plans and reports, our purpose is to make the both goals and the performance information useful to decisionmakers at all levels.
Each year, OMB works to better integrate agency performance information into its own budget development process. And, each year, we're making progress. For example, data from the Bureau of Prisons showed that the overcrowding rate in Federal prisons had increased from 22 percent in 1997 to 32 percent currently. Even though we have had a major program of new construction, the Federal prison population has greatly increased, and now numbers over 136,000 inmates, more than double the number of inmates in 1990. The President's Budget funds additional prison capacity to reduce overcrowding to 30 percent in 2001, and to 28 percent in 2006.
To give another example, the Indian Health Service developed a service plan aimed at using a cost-effective public health approach to reduce the health disparities in Native Americans. This performance plan played an important role in OMB's development of the FY 2001 budget: we aimed funding increases at improving health outcomes. Reviewing the performance measures in IHS' plan led us to target additional resources to increase glycemic control efforts as a means to reduce diabetic complications, to provide additional home-based well-child visits, and to improve water fluoridation compliance to reduce the prevalence of dental decay.
8. Don't shoot the messenger.
When an agency makes a good faith effort, but doesn't meet their goal or if the measures need to be revised, we need to be careful: If the response from OMB were public criticism, then agencies would soon decide that the moral of the story is "Do analysis and you'll be punished." If agencies come to that conclusion, GPRA will never work.
Therefore, we try to provide a combination of public encouragement and private criticism. Both are important; either without the other would quickly turn GPRA into an exercise in rhetoric and reports. We think this is the best approach to encourage agencies to develop and use performance measures in day-to-day management and oversight -- and that is our performance goal.
9. Be realistic.
The final lesson we at OMB have learned is to be realistic about what performance information can – and cannot – do. We recognize and accept that performance information will always be but one factor in policy, program, and resource decisions.
Theoreticians tend to look at performance information and budgeting from either of two abstract directions. At one end, there are those who posit that budgets are determined entirely by external political circumstance, and performance information is irrelevant. We have a word for such cynics: They are "wrong".
At the other end, there are those who believe that that performance information can create decisions that are entirely objective -- that some day the OMB, the President, and the Congress can all be replaced by an analyst armed only with a computer and megabytes of performance data. This is equally wrong. Performance information is and should be an input into decisions about program management and resource levels; but it is not the only input and, by itself, cannot make decisions. Even with the best information in the world, ultimately the President of the United States and the Congress must decide how much and what kind of programs we will have in national defense, in education, in environment, and the myriad of other concerns of the public.
In advising on and making these choices, we have always sought performance information to help. With GPRA, we have the force of law and an established process to help us obtain that information. Consequently, we are generally getting more and better performance information. But performance information, even perfect information, can't make policy decisions. It can only be a tool to inform them.
The standard should be to produce measures and information that help decisionmakers reach judgments, helps agency staff manage more effectively, and help the public be informed about the programs and services carried out in their name.
I hope it is clear that both that we have made very real progress and that we have much more yet to achieve. The vision created by GPRA is a challenging one, but it is important and, with effort, achievable. We look forward to continuing to work with the agencies and the Congress to achieve these goals, and deliver the kinds and quality of programs that people expect from their government.
