Office of Management and Budget Skip Main Navigation
About OMB
President's Budget
Management
Information &
Regulatory Affairs
Legislative Information
Agency Information

Senate Committee on Government Affairs GPRA Report

Contents

Return to OMB Title page of guidance to agencies

103d Congress 1st Session
SENATE
Report 103-58
Calendar No. 96
GOVERNMENT PERFORMANCE AND RESULTS ACT OF 1993
R E P O R T
of the
COMMITTEE ON GOVERNMENTAL AFFAIRS
UNITED STATES SENATE
to accompany
S. 20

TO PROVIDE FOR THE ESTABLISHMENT, TESTING, AND EVALUATION OF STRATEGIC PLANNING AND PERFORMANCE MEASUREMENT IN THE FEDERAL GOVERNMENT, AND FOR OTHER PURPOSES

June 16 (legislative day, June 15), 1993.-Ordered to be printed

COMMITTEE ON GOVERNMENTAL AFFAIRS

JOHN GLENN, Ohio, Chairman SAM NUNN, Georgia CARL LEVIN, Michigan JIM SASSER, Tennessee DAVID PRYOR, Arkansas JOSEPH I. LIEBERMAN, Connecticut DANIEL K. AKAKA, Hawaii BYRON L. DORGAN, North Dakota WILLIAM V. ROTH, Jr., Delaware TED STEVENS, Alaska WILLIAM S. COHEN, Maine THAD COCHRAN, Mississippi JOHN McCAIN, Arizona Leonard Weiss, Staff Director David F. Plocher, Counsel Franklin G. Polk, Minority Staff Director and Chief Counsel John E. Mercer, Minority Counsel Michal Sue Prosser, Chief Clerk

(ii)

Calendar No. 96

103d Congress, 1st Session

Report 103-58

GOVERNMENT PERFORMANCE AND RESULTS ACT OF 1993

June 16 (legislative day, June 15), 1993.-Ordered to be printed

Mr. Glenn, from the Committee on Governmental Affairs, submitted the following

CONFERENCE REPORT together with ADDITIONAL VIEWS [To accompany S. 20]

The Committee on Governmental Affairs, to which was referred the bill (S. 20) to provide for the establishment, testing, and evaluation of strategic planning and performance measurement in the Federal Government, and for other purposes, having considered the same, reports favorably thereon and recommends that the bill do pass.

I. Purpose

The purpose of S. 20, the Government Performance and Results Act of 1993, is to improve the efficiency and effectiveness of Federal programs by establishing a system to set goals for program performance and to measure results.

II. Summary

On March 24, 1993, the Committee on Governmental Affairs voted to report S. 20 as amended. The bill requires that, beginning in FY 1994, there shall be at least 10 three-year pilot projects in program performance goal setting, measurement, and reporting, and at least 5 two-year pilot projects in greater managerial flexibility in return for commitments to greater program performance. In 1997, OMB and GAO shall report on the results of the pilot projects. By FY 1998, the requirements of the Act for five-year strategic planning, annual program performance plans, and annual program performance reports shall come into force governmentwide. Also in 1997, OMB will begin at least 5 two-year pilot projects in performance budgeting.

III. Need for Legislation

OVERVIEW

Public confidence in the institutions of American government is suffering from a perception that those institutions are not working well. Recent public opinion polls indicate that this is particularly true with respect to the Federal Government, as both Congress and the Executive Branch are held in low esteem by the American people.

Much has been made of the seeming inconsistency between the public's desire for a wide range of government services, and that same public's disdain for government and objections to paying higher taxes. The Committee believes that part of the explanation for this apparent inconsistency can be seen in the results of a recent public opinion poll which shows that Americans, on average, believe that as much as 48 cents out of every Federal tax dollar is wasted. In other words, the public believes that it is not getting the level and quality of government service for which it is paying.

The Committee shares the public's frustration with waste, inefficiency, and ineffectiveness in Federal programs. As the general oversight committee of the Senate, it has a long history of examining and exposing those types of problems throughout government. The committee has also authored a series of legislative remedies which have been, or promise to be, very helpful in addressing these problems-including the Inspectors General Act and the Chief Financial Officers Act.

Following on these measures, the Committee believes that the regular and systematic measurement and reporting of program performance, compared to pre-established goals, would be a major addition, providing a valuable supplement to the Committee's previous work in the area of management improvement. Governmental waste and under-performance will likely persist until there is a change in the behavior of federal agencies. James Q. Wilson, a noted student of government, stated it well in his book "Bureaucracy: What Government Agencies Do and Why They Do It": "* * * government management tends to be driven by the constraints of the organization, not the tasks of the organization." A key step in changing government behavior is to create a focus on results.

At present, congressional policymaking, spending decisions, and oversight are all seriously handicapped by the lack both of sufficiently precise program goals and of adequate program performance information. Federal managers, too, are greatly disadvantaged in their own efforts to improve program efficiency and effectiveness by that same lack of clear goals and information on results. The goal-setting, performance measurement, and results reporting requirements of S. 20 are intended to address these needs of Congress and of federal program managers.

This reform has the potential to mark a significant change in the way that managers, policymakers, and the American people think about what services the government should provide, and how well it does at providing them. The legislation will provide the information necessary to strengthen program management, to make objective evaluations of program performance, and to set realistic, measurable goals for future performance-ensuring that the information is reliable will, of course, require attention by agencies, OMB, GAO, and Congress.

The legislation begins this effort with a series of pilot projects, recognizing that while necessary, this is a new and challenging direction for Federal agencies. However, the Committee believes that the eventual result will be not just better oversight and improved performance, but greater public confidence in the institutions of government.

ADMINISTRATION POSITION

The Administration expressly endorsed S. 20 at the March 11, 1993, hearing of the Committee on Governmental Affairs. At that hearing, OMB Director Leon Panetta stated that President Clinton, "has reviewed and discussed S. 20, the Government Performance and Results Act of 1993, and I am pleased to be able to advise the Committee today that this Administration strongly supports this bill." He described the reasons for the Administration's support as follows:

S. 20 is a major step toward making the Government accountable to the American people by making it clear what the taxpayers are getting for their money and removing some of the red tape that bedevils all of us. As every other enterprise has learned, government officials must manage for results, not just rules and regulations. This accountability both empowers and rewards those who improve performance. S. 20 provides us with a sound foundation as we go about the task of re-inventing our government, and we urge its swift passage.

GAO INTEREST

The General Accounting Office has had a long-standing interest in improving government management through the use of strategic planning and performance measurement. Since 1973, GAO has produced over 70 reports on performance measures and currently has nearly a dozen ongoing efforts to assess measurement in specific agencies. At the Committee's October 3, 1990 hearing on "OMB's Response to Government Management Failures", Comptroller General Charles A. Bowsher testified:

We very much support the development of performance measurements. It is something that we have advocated for some time, and one interesting note is that when I was dealing with some of the people from London during the last few years on the issue of financial management, the one thing that surprises them is that we are still talking about improving our accounting system. They are moving on to performance measurement. In other words, they have got the financial management and the accounting in place, are getting good cost information, and their big debate now is what are the performance measurements. And they have even got it down to the local government level covering fire departments, trash hauling, and similar services. It is impressive, and I think we could do it here at the Federal level.

At that hearing, Mr. Bowsher was asked if regular program performance reports identifying objective measurable accomplishments compared against original goals would be helpful in focusing Congressional oversight, and if it would help GAO in its program evaluation activities. He responded:

Yes to both. It would be very helpful I think to Congress to have this information, and that is why I think you need an annual report from all these departments, just like you have in the private sector, where you actually tell what was your performance goals, what were you trying to achieve with the program objectives, and how did you do. And it would help us immensely. We waste half our time in doing our program evaluation work and our financial audit work even in trying to figure out what is the goal that was trying to be achieved and where is the information. We are always over there trying to gather the data. It should be brought together in an organized fashion by the agencies themselves.

Donald H. Chapin, Assistant Comptroller General for Accounting and Financial Management, added:

We have been working recently with the VA and with the Agriculture Department to try to take some of the goals that are clear to us and that are reasonable and try to relate those to financial information, and it is amazing when you do that what you can see about an agency's operations. You can see what the trends are. You can understand what the numbers mean. And if we can get the agencies to supply us with measurable goals, we can relate those to financial results and then you can see what you are spending your money on and whether that money is well spent. And that is my fond hope, that we can get that into our system of Government and have it reported to the Committees of Congress as a regular matter so that they can see the effectiveness of the money that they appropriate. And you don't see this now, you really don't see that.

The issue of performance measurement was a major focus of a February 1985 GAO report entitled, "Managing the Cost of Government-Building An Effective Financial Management Structure" (GAO/AFMD-85-35). That report emphasized that reform of federal financial management should stress (as one of four key areas) the "systematic measurement of performance", arguing that "effective management of resources requires examining the results of government activities as well as their costs." The report stated that:

Whether the goal is defending the nation or immunizing children against disease, government officials and the public need to know how well government is accomplishing its intended objectives. Assessing government accomplishments requires measuring employee and program performance. Though the size and complexity of the government make it difficult, developing effective performance measurement systems is clearly possible.

The work of nearly two-thirds of government employees, for example, can be measured by means of formal productivity measurement systems. For the remaining one-third, substantial time and effort may be required to develop reliable measures of performance. Indeed, there may be some government activities (such as basic research) for which quantitative measures are not feasible. Even in these cases, however, qualitative measures can usually be developed and used.

A well-developed financial management structure should include performance information that can be used both for day-to- day management and policy and budgeting decisions.

The report points out that several state and local governments have already done this to a far greater extent than has the federal Government, and concludes that "We are convinced that such a structure can be built for the federal government."

At the Committee's May 5, 1992, hearing on S. 20, the Comptroller General reiterated his position that:

[A]gencies need to clearly articulate their missions in the context of statutory objectives and, with regard to services, citizen expectations. These objectives need to be written in terms that can be used to judge progress toward achieving them. It is essential that agreement be reached between Congress, the Office of Management and Budget, and the executive agencies on realistic, outcome-oriented goals if they are to use the data to assess progress.

At the March 11, 1993, hearing of the Committee on S. 20, he urged that "action on the bill should not be delayed."

CHIEF FINANCIAL OFFICERS ACT

The Chief Financial Officers Act of 1990 (P.L. 101-576) (CFOs Act) acknowledges the need for more attention to this issue. Though the primary purpose of the legislation is the improvement of financial management activities within the Federal Government, the Committee did include in that law the beginnings of a greater focus on program results.

The CFOs Act provides that the newly created position of Deputy Director for Management at the Office of Management and Budget shall "Perform all functions of the Director * * * relating to-(A) managerial systems, including the systematic measurement of performance". Also, the Act creates at 23 large agencies the position of Chief Financial Officer, and requires that each CFO shall "develop and maintain an integrated agency accounting and financial management system, including financial reporting and internal controls, which- * * * provides for- * * * the systematic measurement of performance".

However, neither the Act itself, nor its legislative history, provides elaboration on the phrase "systematic measurement of performance". The February 1985 GAO report mentioned above does use that term, and discusses it in a context where clearly it is meant to refer to the measurement of program performance-that is, program outcomes and results, and not financial performance. The CFOs Act does expressly require that agency financial statements reflect "results of operations of those revolving funds, trust funds, offices, bureaus, and activities", but a related provision makes apparent that this presently refers only to an "office, bureau, and activity of the agency which performed substantial commercial functions during the preceding fiscal year." Because revolving funds, trust funds, and commercial activities are only part-often a small part-of an agency's operations, the annual financial statement will provide a limited basis for addressing program performance.

Thus, within the context of the CFOs Act, the legislative mandate for the "systematic measurement of performance" of federal programs is presented in a most minimal fashion. OMB has now begun to build upon that simply stated mandate. In September 1991, and again in February 1992, it issued guidance to agencies on preparation of annual financial statements. OMB specified that the overview and supplemental financial management information sections should present information and data on program performance, and described different types of performance measures that could be used and defined how the information should be presented.

The Committee commends OMB for its efforts to use the CFOs Act as a means to begin moving federal agencies toward systematic measurement and reporting of program results. However, the Committee is concerned that by itself the performance measurement mandate in the CFOs Act provides insufficient emphasis for extending performance measurement across the full range of agency program activities. Simply put, the broad program performance and management improvement goals of S. 20's performance measurement provisions are more inclusive than those of the CFOs Act's financial management reforms. Institutionally, OMB's implementation of S. 20 must reflect that broader perspective.

Regular, systematic, and comprehensive program goal-setting, performance measurement, and results reporting is not an easy undertaking. Nor is it generally a welcomed task by bureaucratic organizations. There is often much institutional, and individual, resistance. The mandate for its implementation must be unambiguous. The specific requirements must be clear. And the effort must be sustained.

The Committee believes that ultimate success in bringing about this new focus on agency performance requires a firm statutory commitment. S. 20 provides this commitment, underscoring a strong desire by Congress for reforming this area, and extends and expands the performances measurement begun with the CFOs Act. As the legislation makes clear, this is more than just an Administration directive that can be rescinded or whose implementation can be stalled while waiting for a new administration with a different managerial agenda. The legislation signals an ongoing effort at implementation of performance measurement and reporting, as fully as is feasible within as many programs as is practical.

LESSONS FROM HUD PROBLEMS

Senator Bob Graham (D-FL) was the Committee's first witness at its May 23, 1991 hearing on S. 20. Senator Graham had himself presided a year earlier at hearings of the HUD/Mod Rehab Investigation Subcommittee of the Committee on Banking, Housing, and Urban Affairs. He testified that his own subcommittee's investigation into the "HUD scandal" showed that many of the problems uncovered at the Department of Housing and Urban Development could have been avoided with better congressional oversight, which he felt S. 20 would encourage:

In terms of Congressional oversight, I am afraid the Congress has to bear a serious part of the responsibility for what happened in HUD. There were almost no hearings held during the relevant periods on what was happening in the agency. Those hearings that were held were largely episodic and reactive rather than focusing on the programs and how well they were being administered. The recommendations made in Senator Roth's bill would certainly go a long way towards creating the opportunity for effective Congressional oversight. I strongly agree with the direction of the bill as introduced by Senator Roth. I think its emphasis on setting goals and performance standards, both within the Congress in the shaping of legislation, and within the Executive Branch, are exactly on point.

At Senator Graham's May 8, 1990 hearing on Congressional oversight of HUD, Richard L. Fogel, Assistant Comptroller General, GAO, had testified:

It would help if congressional committees could press some of the agencies to develop performance measures and output measures of what the agencies think are good measures of whether they are accomplishing program objectives * * *, That gives the Congress a basis for then going in and tracking and saying how well things are happening, what type of problems there may be, and if there are problems it gives you a basis for focusing the oversight. Unfortunately, most agencies do not want to develop those types of performance measures because it is easier not to be held accountable if you do not have them.

Also at that same hearing on the HUD scandal, Richard A. Wegman, Chairman of the Congressional Oversight Panel of the National Academy of Public Administration, had testified:

It's essential that there be clear and explicit performance goals for executive branch programs. This would enable agencies to provide a better match between these goals and the resources available to carry them out. Clarity is extremely important; otherwise, it's very difficult for Congress to make a judgment about whether or not the agency is doing its job effectively.

GOVERNMENT MANAGERS

The Committee took particular note of the fact that organizations of current and former government managers strongly support the goals of S. 20, and that they made many useful suggestions in the development of the Committee's substitute amendment. These organizations stated that the systematic collection of program performance information, related to outcome-oriented goals, is vital to helping managers do their jobs well.

The National Academy of Public Administration rarely takes policy positions, so its adoption of a supporting resolution at its November 1991 annual meeting is especially significant. The NAPA resolution pointed out that few government agencies provide timely information on the quality and outcomes of their major programs, but that public officials and citizens need this type of information-"not only information on program costs and the amount of work completed-but also information on the quality of service delivery and on the outcomes achieved through the use of tax dollars and other public resources."

In calling for agreement between the policymaking and operating levels of government on appropriate indicators of program cost, quality and quantity of services, and important program outcomes, the NAPA resolution stated that:

The National Academy of Public Administration strongly recommends that units of government at all levels make a concerted effort to encourage agency heads and program managers to monitor program quality and outcomes as part of an overall system aimed at improving the performance and credibility of major public programs. Performance monitoring should be an essential part of program administration and the budget process * * * chief executives, agency heads, and program managers should propose realistic performance targets in terms of program goals and agreed-on program quality and outcome indicators-and subsequently monitor and report on progress in achieving those performance targets.

Similarly, the American Society for Public Administration also adopted a resolution of support at its April 1992 annual meeting. Noting the fact that program managers' technical data are seldom translated into performance information that policy executives and elected officials can understand, ASPA stated, "Regular measurement and reporting of program effectiveness (including quality, timeliness, and outcomes) and efficiency can provide a new performance-based language for improved communication among program managers, central policy executives, and elected officials." ASPA felt this would help in the development of mutual expectations of program performance at specific funding levels, and make for more informed policy and operational decisions. The resolution explained that:

There is a history of successful uses of performance measurement and reporting by government organizations at all levels, providing that it can improve decision making, accountability and responsiveness to citizens, and program performance. However, use of performance measurement is still the exception rather than the norm in American government organizations. Most reports on government operations focus on expenditures and activity counts or numbers served. Few provide timely information on program effectiveness and efficiency. Thus there is great potential to improve performance, accountability, and responsiveness by implementing systematic performance measurement, monitoring, and reporting, and by integrating performance information into regular policy and management processes.

The Committee realizes that, as valuable as the goal-setting and performance reporting under S. 20 will be to Congress in its policymakiung and oversight roles, its greatest value will likely be to the program managers themselves. First of all, they will have a much better sense of what is expected of them and their programs. Presently, many mangers are seriously handicapped by a lack of clear programmatic direction. And second, they will get regular consistent feedback on measurable progress toward meeting those expectations. Managers will use this information throughout the year to plan their activities and guide their subordinates. The Committee believes that S. 20 has the potential to be a powerful tool for strengthening governmental management.

IV. State, Local, and Foreign Experiences

The Committee found that the use of program goal-setting and performance measurement is a growing trend in State and local governments, and also at the national government level in several foreign countries. The May 12, 1992, edition of "Financial World" magazine labeled performance measurement and program evaluation "perhaps the most important trend in state government in the 1990s". At the Committee's first hearing on S. 20, on May 23, 1991, testimony was received on the experiences of the State of Florida's Children, Youth and Family Services, and from the City of Sunnyvale, California.

The State of Florida enacted legislation in 1986, requiring its Children, Youth and Family Services office (part of its Department of Health and Rehabilitative Services) to monitor and report client-outcomes for children serviced by 34 State program components covering more than 275 separate programs. As mandated, the CYF Outcome Evaluation focuses on the results of services, rather than simply on outputs such as numbers of clients served and the amount of service provided. Program goals are defined in terms of expected results, and there is an annual Outcome Evaluation Report to the secretary, the governor, and the legislature.

In response to the Committee's question about the effect outcome measurement and reporting has had on program management, the Department submitted testimony stating that it has had "very useful and far-reaching effects." Among the major impacts cited in the testimony were:

Outcome information is used increasingly in our budget and appropriations processes. * * * In the 1991 legislative session, for instance, our report was very much in evidence at appropriation committee meetings-on legislators desks, and in the hands of legislative and agency staff, lobbyists, and advocates * * *.

We have identified programs for closer scrutiny, that is, our report provides a useful tool for targeting our own monitoring efforts. Program improvements have been made as a result * * *.

Perhaps most important, our work has sensitized us and others (the courts, advocates, contract service providers, and even to some extent the media) to the importance of thinking and acting with respect to results or outcomes for those whom we serve.

The testimony also pointed out that the success of Florida's Children, Youth and Family Services outcome evaluation system "is evident in legislation passed earlier this year, to extend in law the requirement for outcome evaluation to all programs under the department's jurisdiction." Under this extension-to the entire $8-billion a year Department of Health and Rehabilitative Services-the Secretary is authorized to set aside up to one-half of one percent of all program budgets for outcome evaluation.

The Committee also heard testimony from the city manager of Sunnyvale, California. That city began to develop a very sophisticated system of program performance measurement, incorporated directly into its budget system, in 1973-as part of a pilot project developed with the U.S. General Accounting Office. OMB testified that Sunnyvale's system "stands out as the single best example of a comprehensive approach to performance measurement that we have found in the United States. * * * One underlying reason for the success achieved in Sunnyvale is the fact that every program manager uses the system to plan, manage, and assess progress on a day-to-day basis."

Sunnyvale began the implementation of its system on a pilot project basis, starting first with its Public Safety Department, and gradually adding sophistication and capabilities as it spread government-wide. The city now tracks hundreds of workload outputs (including unit-costs), and service outcomes. Examples of program performance objectives include: achieve fewer than 10 complaints a year for poor information services by the City Hall lobby reception and telephone information service; repair damage from vandalism within 3 days; successfully respond to library user information requests at least 95 percent of the time; achieve a positive rating from at least 90 percent of recreation class participants; respond to emergency police calls within 5.6 minutes or less 90 percent of the time; maintain crime rate within the lowest 25 percent of cities of comparable size; maintain files of all documents so that they can be retrieved within 5 minutes 85 percent of the time.

Each year the city manager submits a detailed Annual Performance Report to the mayor and city council, indicating how well the performance objectives have been achieved. All of those objectives, in turn, are tied into twenty-year strategic plans covering 28 areas of city service, showing long-term goals for the City.

At the Committee hearing, Sunnyvale's city manager was asked whether it was difficult to achieve bipartisan agreement on expected program results. He responded:

[W]hat can become possible is through articulating outcomes this way, things have a potential of becoming less political as opposed to more political. There tends to be a focus on what everyone wants, and all of a sudden people on either side of the spectrum find out they want the same thing, and so they are articulating it in outcome terms instead of political terms. It has reached the point in local political campaigns * * * that even newcomers on the scene have now begun to articulate their platforms in terms of outcomes, even before they get into the political process. So at our level it has sort of been institutionalized, and certainly I think, if any thing, depoliticized the appropriation process.

The effectiveness of Sunnyvale's system of performance measurement is cited in the book "Reinventing Government," by David Osborne and Ted Gaebler. The authors point out that:

Between 1985 and 1990, the city's average cost per unit of service went down 20 percent, after factoring out inflation. (In other words, its productivity increased by roughly 4 percent a year.) In 1990, when it compared its own costs to those of similar size and type cities, Sunnyvale found that it used 35 percent to 45 percent fewer people to deliver most services. Its employees were paid more, but its operating budget was still near the low end of comparable cities, and its per capita taxes were lower than those of any comparable city in its sample.

In its most recent citizen survey, over 90 percent indicated satisfaction with the quality of city services.

With regard to foreign countries' experiences with performance measures, work by the Paris-based Organization for Economic Cooperation and Development suggests that several countries may be 5 to 10 years ahead of the U.S. in this effort, and that their performance measurement efforts are a key part of broader efforts to better manage for results. The key themes in their efforts have been, first, to better clarify agency and managerial accountability for results by defining goals clearly, developing measures, and reporting on progress. And second, to give managers the flexibility to manage for results by providing them the tools and incentives to act.

At the second hearing on S. 20, on May 5, 1992, OMB's Deputy Director for Management, Frank Hodsoll, summarized the experiences of several foreign governments. He noted that Australia is "at the leading edge in performance measurement." That country has a well-developed system involving strategic planning, annual work plans, performance reporting, and program evaluation. Program objectives and strategies for achieving them are well-defined. In addition, the government annually publishes a 19 volume set of detailed program performance statements relating expenditures to program performance. He added that:

It should be noted, however, that Australian program officials have much more latitude than we do in how they spend operating funds. For example, Australian officials can shift funds from staff to computers without Parliamentary concurrence. But Australian officials are more accountable for the performance of their programs; the annual work plans link program objectives and the performance agreements on which ministry officials are evaluated.

The experience of the United Kingdom was also described. That country's effort began in 1982 with its Financial Management Initiative (similar to our CFOs Act). In 1988 the U.K. launched its "Next Steps Initiative", which places special focus on the development of output and performance measurement for executive functions. Ministries are required to development systems through which managers at all levels in the national government will have a clear view of their objectives, and the means to measure and assess performance in relation to those objectives.

The U.K.'s National Audit Office reports that the ministries have made "worthwhile progress" and that the quantity and quality of performance information has significantly improved in recent years. OMB's assessment of the U.K.'s experience is that progress apparently depends "on the complexity and structure of the ministry, the nature of the ministry's work, and the available resources. There are also often inherent difficulties in devising performance indicators that measure the impacts of policies, allow for the effects of factors outside of Governmental control, assess the quality of services provided, and quantify the results of basic research."

The Committee agrees with OMB's conclusion in its hearing testimony that because of the uniqueness of the U.S. Federal Government's structure, "there are no domestic or international proto-types on which we might model a Federal system". Those national governments which have been much greater progress than our own in program performance measurement have a Parliamentary form of government, where heads of ministries are usually also members of Parliament. Nonetheless, as OMB points out, "Even though no specific prototype exists, our review of these systems has been helpful in determining what elements appear to be essential for a successful system. The review can also help us to learn from others' experiences and mistakes."

V. Cost of Performance Measurement

The Committee believes it is important that performance measurement not be a major additional cost item or paperwork burden imposed upon Federal programs. This issue was a subject of specific inquiry during the hearings on S. 20. Most of the witnesses indicated that program performance measurement need not be a significant cost or administrative burden. The Committee, however, has included within the pilot projects of S. 20 the requirement that this issue be examined and reported on to Congress.

In response to a question about the implementation cost of performance measurement throughout the Federal government, Frank Hodsoll, OMB Deputy Director for Management, testified that based on OMB's work with various agencies in this area of performance measurement, he thought the cost would be very small and absorbable within the current system. This view was supported by Charles Bowsher, U.S. Comptroller General, who responded that he did not think the cost would be great, and that the primary costs would be in additional training expenses.

In looking at specific examples from State and local government, the Committee found similar reactions. The City of Sunnyvale, California is generally recognized as having the most comprehensive system of performance measurement of any government in the country. Its city manager, Tom Lewcock, testified that Sunnyvale's system of program performance measurement and budgeting was implemented without the addition of any new personnel. He explained that this was because it replaced a more traditional system of accountability, rather than being simply added to it.

Florida's Children, Youth and Family Services is a state program nationally recognized for its sophisticated measurement of program outcomes and results. Its Chief of Research and Development, Dennis Affholter, testified that it generates information very similar to that mandated by S. 20, and that, "A small staff properly equipped can do this job * * *. It cost us about a quarter of a million dollars to support the staff to do this job out of a total budget that approaches $500 million * * *."

The Committee received other testimony that this type of measurement might potentially be costly, but the agencies should be able to measure their performance effectively with no more than one percent of their program funds (and much less for some of them). It was also pointed out that there is already a great deal of such data collection going on in federal programs and that this activity could be re-directed, coordinated, and the data better reported and used.

This latter point was supported by a GAO study released at the second hearing on S. 20.

VI. GAO Study of Agency Performance Measurement

At the request of the Committee, GAO surveyed the existing status of performance measurement in federal agencies. That study, entitled "Program Performance Measures-Federal Agency Collection and Use of Performance Data" (GAO/GGD-92-65), was released May 5, 1992.

GAO surveyed 103 federal agencies to determine the extent to which they had created strategic goals and collected measures of progress toward meeting those goals. About two-thirds said they had a single long-term strategic plan, and three-fourths said they collected a wide variety of data to assess program performance.

In describing the conclusions GAO reached from further examination of agency activities, Comptroller General Charles A. Bowsher testified:

However, when we visited a sample of these agencies, we found that most used the information at the program level. While this information was useful at the program level, it was fundamentally different from that needed to manage or make strategic policy decisions for the agency as a whole. Only 9 of the 103 agencies reported having an administrative infrastructure in place for developing and reporting results. By this we mean that there were few offices that routinely collected performance data and prepared regular reports on progress toward goals set in strategic plans.

He stated that agencies were using their performance measurement systems for a wide variety of purposes (e.g., organizational accountability, budget decisions, individual employee rewards), and that in recent months there has been some movement toward developing results-oriented performance measures. However, as the GAO report points out, the uses were almost exclusively to provide internal information relating to past activities or present operations, and not to assess progress towards goals in their strategic plans.

Despite these limited uses of performance measurement by federal agencies, the Committee is encouraged by the findings of the GAO survey. The report verifies a great deal of existing activity that provides a foundation for implementation of S. 20. The legislation can give these efforts better direction, structure, and coordination. It can expand their development and broaden the uses. And it can provide for much better public and congressional dissemination of the information.

VII. Elements of Performance Measurement System

The key elements of an effective program performance measurement system are strategic plans, annual performance plans, and annual performance reports. The Government Performance and Results Act establishes these elements. In addition, it provides for the possibility of waivers for increased managerial accountability and flexibility, and it tests the development of program performance budgets. Development of these plans and reports is defined as being an inherently governmental function, meaning that while assistance may be provided by other parties, actual formulation of the final products must be the responsibility of federal managers.

STRATEGIC PLANS

Strategic plans are the starting point and basic underpinning for a system of program goal-setting and performance measurement that will be established throughout the Federal Government. A multi-year strategic plan articulates the fundamental mission (or missions) of an organization, and lays out its long-term general goals for implementing that mission, including the resources needed to reach these goals.

The clearer and more precise these goals are, the better able the organization will be to maintain a consistent sense of direction, regardless of leadership changes at the top. This is particularly important in the Federal Government, where turn-over in top-level positions (such as Assistant and Deputy Assistant Secretary) occurs on a perhaps too frequent basis.

Even when a change in Administration brings about a shift in political philosophy, the program's missions and long-term general goals remain largely intact. The priorities and means of achieving those goals, and the approach to problem-solving, may vary significantly, but the long-term goals usually remain the same. Plans for how an agency will maintain its continuous operations and progress towards those long-term goals is vitally important.

As has previously been pointed out, many agencies already have what they call "strategic plans", but these are generally inadequate and poorly used. A major problem with many is that they have little direct linkage to the agency's daily operations, which greatly weakens their effectiveness.

PERFORMANCE PLANS

Annual program performance plans are what provide the direct linkage between an agency's longer-term goals (as defined in the strategic plan) and what its managers and staff are doing on a day-to-day basis. These plans are often hierarchical in form, showing what annual performance goals need to be accomplished at each level in order for the next higher level to meet its own goals.

Performance goals may relate to either "outputs" or "outcomes", the latter usually being the most important for policy purposes, but the former often being a useful management tool (especially when their per-unit costs are also tracked). A common weakness in program performance plans is an over-reliance on output measures, to the neglect of outcomes. Eligible clients completing a job training program are outputs; an increase in their rate of long-term employment would be an outcome. There could be similar outcome goals measuring the effectiveness of Federal community development block grants, such as percentage increases in property values and net new jobs created within the targeted areas. Even at the lowest operational level, there can be goals for processing time, error rates, customer/citizen- satisfaction levels, etc.

It is very important that annual performance plans include goals, not just for the quantity of effort, but also for the quality of that effort. These goals should be as specific as possible, they should drive much of the daily operations of the agency, and they should aim at achieving the long-term general goals of the agency's strategic plan.

It is also important that the resources needed to achieve the goals be indicated as part of the plan. The Committee is concerned about the "hollow government" phenomenon-where an agency has inadequate resources to meet its public missions. Whether the proper remedy is to increase the level of resources allocated, or to reduce the level of service to which the agency is committed, both should be brought into balance. The annual performance plan should show how program goals will be supported through sufficient management skills and human, budgetary, and physical resources.

Not all governmental programs lend themselves easily to measurable goals. For some it will be very difficult, and for a few, perhaps impractical altogether. Nonetheless, managers should resist the temptation to decide too quickly that a particular program is unsuitable for measurable goals. The fundamental question is, what is the difference between a successful program and a failure? Between a well-run operation and one that is mismanaged? How can we tell the difference, and how should that be defined? If the difference cannot be defined, then is that not just an invitation to waste, inefficiency, and ineffectiveness?

PERFORMANCE REPORTS

Annual program performance reports are the feedback to managers, policymakers, and the public as to what was actually accomplished for the resources expended-in other words, how well the original goals were met. This type of information is ideally available to program managers on a more regular basis throughout the year, but at a minimum there needs to be an annual compilation and reporting of results.

There may be more performance information tracked by the agency for management purposes than is summarized in the annual report, but there should be a match between the report and the goals of the previous performance plan. And while the nature of some of what is measured might change periodically, that should not be a frequent, widespread occurance (especially after the first few years' experience). Otherwise, it will be difficult to spot trends in program performance, which is often the most revealing type of information for managers and policymakers.

The Government Performance and Results Act also asks that the annual performance reports include explanatory information on goals not met. This includes plans for achieving the goals, or reasons why that is not possible and recommended action. The goal itself might be unreasonable, given the resources allocated. Or the goal might be reasonable, if the program is restructured. Or an unforeseen occurrence might have interfered with the goals attainability. Or the entire underlying premise of the program might be flawed. Or the program might simply have been mismanaged. Each of these and other possible explanations suggest different responses by top executives and the Congress.

Finally, the reports should also relate performance measurement information to program evaluation findings, in order to give a clear picture of the agency's performance and its efforts at improvement.

MANAGERIAL FLEXIBILITY WAIVERS

The Committee recognizes that Federal managers, as a general rule, are greatly limited in their ability to shift resources around within their programs, and to exercise other forms of managerial discretion. Rather than being held accountable for achieving results, they are generally held accountable for following detailed and specific procedures, within programs whose structures are rigidly mandated.

The Committee heard considerable testimony that governmental program results can often be improved if the balance between those two forms of accountability were shifted somewhat. This is, it has been the experience of other governments (local, state, and foreign) that managers can improve program performance if there is more specific agreement on what they are to accomplish, if they are given greater managerial flexibility, and then held accountable for the result. The British government, for example, signs agency heads to employment contracts with measurable goals for program performance, while giving them wider latitude in how they expend allocated resources to accomplish those goals.

At the March 11, 1993, hearing on S. 20, Comptroller General Charles A. Bowsher testified that,

[T]he experience of some states and other countries suggests that providing greater flexibility and incentives for managers to act is critical to fundamentally improving agencies' performance. These governments granted managers greater freedom by

- reforming their civil service systems to make it easier for agencies to hire and to provide different compensation, incentive, and promotion systems;

- recasting their budget execution systems to allow multiyear budget allocations, gainsharing, and a reduced number of line items in their appropriations;

- devolving more responsibility for control of operations away from central management agencies and creating an environment where managers are held more responsible for their actions; and

- streamlining acquisition processes and allowing choice between government and nongovernment service providers.

Along with this increased flexibility, the governments also increased accountability-but for results rather than processes.

The freedom to be innovative and creative, and to marshal resources as seen appropriate, is also one way to improve the morale and self-esteem of program staff.

For the reasons, the Act includes a provision for the granting of managerial accountability and flexibility waivers; that is, the opportunity to be exempt from certain specific types of non-statutory administrative procedural requirements, in return for achieving greater program results than would otherwise occur. OMB would have to approve such waivers-the agreed conditions and promised benefits of which would be specifically spelled out. The requirements eligible for waiver are exclusively those regarding the internal allocation and use of resources. They do not include any requirements that directly affect persons or activities outside the agency. The Act does not give agencies the authority to waive statutory provisions or regulations promulgated under the Administrative Procedure Act.

The Committee recognizes that there will always be a need for certain procedural controls on management discretion. This is one reason the additional flexibility granted under the Act is fairly limited. Also, such flexibility is not intended as a way around existing labor agreements, Civil Service laws, or to permit inappropriate favoritism. Nor should it undermine organizational morale. However, given the need for government programs to find innovative ways to "do more with less", the Committee believes that the Act provides an important first step in a direction that may pay significant dividends. Experimentation in this area would be worthwhile.

PERFORMANCE BUDGETING

Traditional line-item budgets sent annually to the Congress are rather imprecise policymaking documents, and are rarely effective as management tools. Line-item budgets show how much the President proposes to spend on each program, and how that money should be allocated among various accounts. That format, however, provides a fairly weak linkage to anticipated program results. In other words, it shows how the money should be spent, but not what should be accomplished.

Particularly during this time of very tight budget constraints, it is important that Congress develop a clear understanding of what it is getting in the way of results from each dollar spent, and how those results would change with an increase or decrease in funding. In all likelihood, Congress will face difficult, wrenching budget decisions for years to come. But even if the budget were balanced, and revenues strong, this information would be important in the making of wise spending decisions.

Therefore, it would be most useful if Congress received a budget showing a direct relationship between proposed spending and expected results, along with the anticipated effects of higher or lower amounts. To use a hypothetical example, a survey of National Parks visitors might show that they give their experience an average rating of 3.7 on a 5-point scale. After examining the specifics of the survey results (i.e., what were the problem areas, in which parks), the Park Service might indicate that for an additional 5 percent in funding, it expects to be able to raise the average score to a 4.0. On the other hand, a 5 percent cut might result in a drop to 3.5. Likewise, in this example, the Park Service could show how rising costs or needed capital improvements require increased spending to maintain current services, and to what extent those services might decline if current spending is maintained or decreased. Congressional committees, of course, would examine the rationale underlying those assumptions, but they would have more concrete information on which to base their decisions.

The Government Performance and Results Act addresses the need for this type of information in two ways. First, it requires the President, beginning with FY 1999, to submit an overall Federal Government performance plan along with the budget, derived from the agency performance plans. While this is not a performance budget as such, it would be a very helpful first step. It begins to explicitly link expected results with expenditures in the budget. And second, the Act requires that there be two-year pilot projects in performance budgeting (linking anticipated results to alternative spending levels) in at least five Federal programs, beginning in FY 1998. It then calls for a report from OMB on these tests, along with recommendations on whether the entire Budget ought to be cast in those terms. New legislation would be required for full implementation.

The Committee believes that this pilot project approach is best because, while performance budgeting promises to link program performance information with specific budget requests, it is unclear how best to present that information and what the results will be. For example, GAO, in "Performance Budgeting: State Experiences and Implications for the Federal Government" (GAO/AFMD-93-41, February 1993), reports:

Despite long-standing efforts in states regarded as leaders in performance budgeting, performance measures have not attained sufficient credibility to influence resource allocation decisions. Instead, according to most of the state legislative and executive branch officials we interviewed, resource allocations continue to be driven, for the most part, by traditional budgeting practices. Reasons for this condition include difficulties in achieving consensus on meaningful performance measures, dissimilarities in program and fund reporting structures, and limitations of current accounting systems.

Accordingly, pilot projects will allow OMB to test possible approaches and develop capabilities towards realizing the potential of performance budgeting. At the March 11, 1993, hearing of the Committee, OMB Director Panetta emphasized the Administration's commitment to this endeavor:

With this bill, we will immediately undertake a more limited-but very useful-form of performance budgeting, in which the performance goals that are annually set will conform with the level of resources requested in the budget. Starting next year with the pilot phase of S. 20, we will begin building a system that comprehensively sets out to correlate performance, particularly results-oriented performance, with budgeted amounts.

VIII. Implementation of S. 20

The Committee recognizes that the reforms of S. 20 are a major undertaking. Comprehensive program goal-setting, and performance measurement and reporting, on a government-wide basis will not be accomplished easily. Many Federal agencies will have to think about their programs in ways they are not now accustomed-with a focus on results. Determining what to measure and how to measure it, and then collecting information that is both accurate and meaningful, will be challenging for many organizations. It may be several years before a truly effective performance measurement system is operating.

Past efforts at comprehensive management reform, such as the Planning-Programming-Budgeting System (PPBS), and Zero-Based Budgeting (ZBB), though equally well-intended, were less than satisfactory. New information technologies, unavailable in past decades, should now be a great advantage in bringing about successful program performance measurement. However, this effort will require careful planning and thoughtful execution, because the ultimate objective is to change agency and managerial behavior-not to create another bureaucratic system.

PILOT PROJECTS

One of the lessons learned from the experience of other governments studied by the Committee, OMB, and GAO is that it is best to begin with several pilot projects. Focusing on doing it right in a handful of programs-often learning on a trial-and- error basis-maximizes the likelihood of ultimate, government-wide success.

Because the Committee feels it is important not to try to do too much, too soon, S. 20 mandates that the requirements of the Act first be tested on a pilot project basis for three years (FY 1994, 1995, and 1996). This will give OMB the opportunity to study those examples and to develop useful guidance for more full-scale implementation. Congress too will have the opportunity to make changes to the underlying statute, if the pilot project experiences suggest needed modification.

The legislation allows OMB to designate the pilot project programs, in consultation with agency heads, but specifies that there shall be at least ten such pilots, and that they shall "reflect a representative range of Government functions and capabilities in measuring and reporting program performance." In other words, there shall be pilot projects in defense programs as well as social programs, and in difficult to measure areas as well as presumably easier areas.

Another lesson from other countries is the need to create incentives for managers to want to use performance measures. Having failed at effective performance measurement in the past, countries such as Australia and the United Kingdom have more recently found that providing agencies greater flexibility to manage seems to increase the chances of success in getting performance measures used. This was done by reducing central agency constraints on actions in personnel, budget, and procurement.

Based on this experience the legislation mandates that, from among the pilot projects, at least five also test "managerial accountability and flexibility" to see if the influence of incentives will, as in the other countries, increase the chances for successfully implementing better accountability systems. This additional flexibility is defined as the granting of exemption from certain specified types of internal, administrative requirements (dealing primarily with the shifting of funds between certain internal accounts), in return for agreeing to achieve even greater program results. Statutory requirements could not be waived.

After the three-year pilot projects in program performance measurement, OMB and GAO will each issue reports to Congress on the results of those tests, in mid-1997. Governmentwide implementation of the Act's requirements will begin in FY 1998.

Many Federal programs assisting or affecting the public are administered by States and local governments. Their role in delivering services directly to the public is often greater than that of the Federal agency funding the program. A number of States and local governments have, or are developing, a capability to set strategic goals and extensively measure program performance. For these reasons, the Committee encourages OMB and the agencies to work with States and local governments during this three-year pilot project phase to examine ways for reflecting the role of third parties in agency performance plans and reports.

During this pilot phase, the Committee also encourages that studies be done on the use of waivers. These waivers would give State and local officials greater flexibility, in exchange for their sharing with Federal officials an increased accountability for program results and improved performance. The waivers should look at the application of existing statutory demonstration-type authority or waiver authority currently contained in agency rules. These studies of State and local waivers does not, of course, give agencies any new authority to waive statutory or non-statutory requirements, though recommendations in that regard could be contained in the OMB report on the pilot projects.

TIMELINE FOR IMPLEMENTATION

Just as important as beginning cautiously in implementing these performance measurement requirements, is the necessity of a clear, long-term commitment to the reforms. The Committee felt it important to outline a comprehensive plan for phase-in of the Act's requirements, from pilot projects through governmentwide program performance reporting. This will put all Federal agencies on notice, even those not participating in the pilot projects, that they should begin now preparing for a new focus on reporting the results of their programs. The general timeline for implementation of S. 20 is as follows:

October 1, 1993.....10 pilot projects in annual performance plans and reports (FY 1994, 1995, 1996).

October 1, 1994.....5 pilot projects in managerial flexibility waivers (FY 1995, 1996).

May 1, 1997.........OMB reports on pilot projects.

June 1, 1997........GAO report on pilot projects.

September 30, 1997..All agencies submit 5-year strategic plans (and every 3 years thereafter), and annual performance plans (and each year thereafter).

..5 pilot projects in performance budgeting (FY 1998 and 1999).

January 1998 (approx.)..OMB submits Federal Government performance plan with FY 1999 budget (and each year thereafter).

..FY 1999 budget also shown in performance budget format for pilot projects in performance budgeting.

March 31, 2000.....All agencies submit annual performance reports for FY 1999 (and each March 31 thereafter).

March 31, 2001.....OMB report on performance budgeting pilot projects.

IX. Legislative History of S. 20

101st Congress

On October 3, 1990, Senator Roth introduced S. 3154, the "Federal Program Performance Standards and Goals Act of 1990", which was referred that same day to the committee on Governmental Affairs. No hearings were held on the bill, but it was discussed by Senator Roth and witnesses at the October 3 and 11 hearings on "OMB's Response to Government Management Failures" held by the Committee on Governmental Affairs.

102d Congress

On January 14, 1991, Senator Roth introduced S. 20, the "Federal Program Performance Standards and Goals Act of 1991", which was referred that day to the Committee on Governmental Affairs. The legislation differed from the version introduced in October 1990 only by the addition of a Findings and Purposes section.

The Committee on Governmental Affairs held hearings on S. 20 on May 23, 1991 and May 5, 1992.

May 23, 1991-hearing on S. 20

The Committee on Governmental Affairs heard testimony from the following witnesses:

Honorable Bob Graham, U.S. Senator from Florida;

Professor Steven Kelman, Kennedy School of Government, Harvard University, Cambridge, Massachusetts;

Professor Joseph S. Wholey, University of Southern California School of Public Administration, Washington Public Affairs Center, Washington, DC;

Thomas F. Lewcock, City Manager, Sunnyvale, California; and

Dennis Affholter, Chief, Research and Development, Children, Youth and Family Services Department of Health and Rehabilitative Services, State of Florida.

May 5, 1992-hearing on s. 20

The Committee on Governmental Affairs heard testimony from the following witnesses:

Honorable Charles A. Bowsher, Comptroller General, U.S. General Accounting Office;

Honorable Francis S. Hodsoll, Deputy Director for Management, Office of Management and Budget;

Russell Morris, Commissioner, Financial Management Service, U.S. Department of the Treasury;

Dr. J. Michael McGinnis, Deputy Assistant Secretary for Health, Public Health Service, U.S. Department of Health and Human Service; and

Raymond J. Uhalde, Administrator, Office of Strategic Planning and Policy Development, Employment and Training Administration, U.S. Department of Labor.

The Committee on Governmental Affairs considered S. 20 on August 5, 1992. It adopted by voice vote an amendment in nature of a substitute offered by Senators Glenn, Roth, and Cohen, retitling the bill the "Government Performance and Results Act of 1992", providing for initiation of implementation with a set of 3-year pilot projects before governmentwide application, and making other changes, and voted to report the bill favorably by voice vote.

The Committee on Governmental Affairs reported S. 20 (S. Rept. 102-429) as amended on September 29, 1992, and the Senate passed the bill on October 1, 1992, under unanimous consent.

103d Congress

On January 21, 1993, Senator Roth (with Senators Glenn, Graham, Metzenbaum, McCain, Akaka, Robb, and Lugar as cosponsors) reintroduced the bill as S. 20, the "Government Performance and Results Act of 1993", which was referred that same day to the Committee on Governmental Affairs. The legislation was very similar to the version passed by the Senate the preceding year.

The Committee on Governmental Affairs held a hearing on S. 20 and other legislation on March 11, 1993.

March 11, 1993-hearing on s. 20

The Committee on Governmental Affairs heard testimony from the following witnesses:

Honorable Harry Reid, U.S. Senator from Nevada;

Honorable Charles A. Bowsher, Comptroller General, U.S. General Accounting Office;

Honorable Leon Panetta, Director, Office of Management and Budget;

David Osborne, author of "Reinventing Government";

Martin Gross, author of "The Government Racket: Washington Waste A to Z"; and

Peri Arnold, Professor of Government, University of Notre Dame, South Bend, Indiana.

The Committee on Governmental Affairs considered S. 20 on March 24, 1993. It adopted by voice vote an amendment by Senator Glenn and an amendment by Senator Pryor, and voted to report the bill favorably by voice vote.

X. Section-by-Section Analysis

Section 1. Short title

The name of the legislation is the "Government Performance and Results Act of 1993".

Section 2. Findings and purposes

Congress finds that-(1) waste and inefficiency in Federal programs undermine the confidence of the American people in the Government and reduces the Federal Government's ability to adequately address vital public needs; (2) Federal managers are seriously disadvantaged in their efforts to improve program efficiency and effectiveness, because of insufficient articulation of program goals and inadequate information on program performance; and (3) congressional policymaking, spending decisions and program oversight are seriously handicapped by insufficient attention to program performance and results.

The purposes of this Act are to-(1) improve the confidence of the American people in the capability of the Federal Government, by systematically holding Federal agencies accountable for achieving program results; (2) initiate program performance reform with a series of pilot projects in setting program goals, measuring program performance against those goals, and reporting publicly on their progress; (3) improve Federal program effectiveness and public accountability by promoting a new focus on results, service quality, and customer satisfaction; (4) help Federal managers improve service delivery, by requiring that they plan for meeting program objectives and by providing them with information about program results and service quality; (5) improve congressional decisionmaking by providing more objective information on achieving statutory objectives, and on the relative effectiveness and efficiency of Federal programs and spending, and (6) improve internal management of the Federal Government.

Section 3. Strategic planning

This section requires that each agency prepare a strategic plan for program activities covering at least a five year period, and that such plans be updated at least every three years. The first of these plans shall be submitted to the Director of the Office of Management and Budget (OMB) and to Congress by September 30, 1997. The Committee expects that OMB may establish appropriate dates for submitting subsequent strategic plans, but once submitted by the agency head, a strategic plan is not subject to approval by OMB or any other government entity.

PLAN CONTENTS

The basic content of a strategic plan shall consist of a comprehensive mission statement, a set of general goals and objectives and the approach that will be used in achieving them, including the necessary resources, and a description of any key external factors that may significantly affect achievement of the goals and objectives.

The plans should be succinct and precise. Agencies may choose to develop separate strategic plans for major component organizations or functions, but these separate plans must be subsequently incorporated into a single, agency-wide document.

The agency mission statement should concisely summarize what the agency does, as required by law, presenting the main purposes for all its major functions and operations. The general goals should elaborate on that statement. These goals constitute a specific set of policy, programmatic, management objectives for the programs and operations covered in the strategic plan, and serve as a framework from which the annual performance goals are derived. The general goals should correspond to the purposes set forth in the mission statement, and develop with greater specificity how an agency will carry out its mission.

The general goals need not be in a quantitative or measurable form, but they must be expressed in a manner that allows a future assessment of whether a goal is being achieved.

The strategic plan is also to contain a description of how the agency intends to achieve the general goals. This description should cover the overall approach that will be taken over the time period covered by the plan, including a schedule for significant actions and the needed resources. It should indicate how the goals of the annual performance plans required by section 4 will be used to measure progress in achieving the general goals of the strategic plan, and the underlying basis for any assumptions or projections.

The key external factors that could significantly affect the achievement of the general goals and objectives, and which should be explained in the strategic plan, can include both governmental and non-governmental factors.

Because measurement of outcomes often relies on an analytic process known as program evaluation, the strategic plan is to contain a section explaining how completed evaluations were used to establish or revise general goals, and set out a schedule for periodic future program evaluations.

CONSIDERATION OF VIEWS

The strategic plan is intended to be the principal means for obtaining and reflecting, as appropriate, the views of Congress, and those governmental and non-governmental entities potentially affected by or interested in the agencies' activities. Although development of the plan is not subject to formal procedural requirements such as the Administrative Procedure Act, the agency shall solicit and consider views of interested members of the public in the process of preparing the plan. An agency's strategic plan shall be a matter of public record, and shall be disseminated as appropriate, with the public offered an opportunity to obtain copies of the completed plan. Copies shall also be sent to Congress, including the Senate Committee on Governmental Affairs, the House Committee on Government Operations, and the appropriate authorization and appropriations committees and subcommittees.

INHERENTLY GOVERNMENTAL FUNCTIONS

The preparation of either an agency's or the Postal Service's strategic plan, annual performance plan, and annual program performance report under this Act are declared to be inherently governmental functions. In defining these activities in this manner, the Committee was guided by the OMB policy letter of September 23, 1992, which established Executive Branch policy relating to service contracting and inherently governmental functions. This policy letter defined an "inherently governmental function" as a "function that is so intimately related to the public interest as to mandate performance by Government employees."

While this Act specifies that Government employees are solely to be responsible for the final plan or report, this does not limit agencies from being assisted by non-Federal parties, such as contractors or grantees, in the preparation of these plans and reports. This might be necessitated, for example, when there is a lack of in-house expertise within an agency. The assistance of non-Federal parties may include collection of information, the conduct of studies, analyses, or evaluations, or the providing of advice, opinions, or ideas to Federal officials, or to provide training of Federal employees. This assistance by non-Federal parties in the performance of inherently governmental functions is also consistent with the OMB policy letter.

The Committee also recognizes that many Federal programs are carried out by States, local governments, and contractors-not by the Federal Government directly. Federal agencies regularly rely on these parties for performance data, and the Committee neither intends nor expects existing systems, processes, and requirements for measuring current or past performance, or which propose or forecast future performance levels to be duplicated by new parallel efforts involving only Federal employees.

Finally, the Committee notes that it is the longstanding policy of the Federal Government that Federal officials should perform the decision and/or policymaking and managerial responsibilities of the government. The basic principle is that accountable federal employees should not only be responsible for the "products" produced by their agencies (whether contractors or federal employees produced the product) but also should be involved in a significant manner in the "process" of formulating the product. Thus, agencies are not fulfilling the intent of this legislation if the required plans and reports are largely the products of contractors. To further this need for accountability, agencies should include in their plans and reports an acknowledgment of the role and a description of a significant contribution made by a contractor or other non- federal entity to the plan or report.

AGENCIES COVERED BY THIS SECTION

The legislation includes all federal entities defined by 5 U.S.C. 105 as being an agency, with certain exceptions. The Panama Canal Commission was excluded because its substantive missions will end within a few years. The Central Intelligence Agency is excluded because the Budget of the United States Government displays a very limited number of program activities for this agency. The Committee also recognizes that plans and reports prepared by the Central Intelligence Agency would most likely be classified in their entirety. The Director of OMB should assess whether, at some future date, the CIA should be made subject to this Act, and include the results of that assessment in the report to the President and the Congress required by Section 1118(c). This assessment should also review how those defense agencies whose primary mission is intelligence should be subject to this Act. In preparing this assessment, the cognizant Congressional committees, the Secretary of Defense, and the heads of the agencies should be consulted. Also excluded are the Postal Rate Commission and the General Accounting Office. The Postal Service is covered by this Act, but under a special provision (Section 7) designed to recognize its special status of independence within the Federal Government.

Section 4. Annual performance plans and reports

This section requires the preparation of an agency's annual performance plan and the Federal Government performance plan for the overall budget. The latter plan is to be submitted to Congress as part of the Budget of the United States Government. Agency annual plans are submitted to OMB for use in preparing the overall government-wide performance plan.

FEDERAL GOVERNMENT PERFORMANCE PLAN

Beginning with the submission for FY 1999, the annual budget of the United States Government shall include a Federal Government performance plan. The Director of OMB has discretion in determining the best manner and useful form for submitting the performance plan. The plan could be integrated with the detailed budget estimates, appear as an individual part of the main budget document, or be submitted separately. The Committee intends that this plan be submitted coincident with the principal budget documents, so that Congress will have this plan available when reviewing the agency budget estimates.

The Act does not set dates for submission of the annual agency performance plans to OMB. The Committee expects OMB to establish agency submission dates commensurate with the time it requires for preparing the overall government-wide plan.

The Committee intends that the Federal Government plan present a single cohesive picture of the annual performance goals for the fiscal year. In doing so, the Director may summarize or abstract the material contained in an agency's annual performance plan, presenting at least the key measures of program performance. Waivers of requirements and controls as provided for by Section 5 of this Act must be included.

Several Federal agencies, such as the Tennessee Valley Authority, have statutorily been afforded a greater degree of autonomy in conducting and administering their business operations than have other agencies. The requirements of this Act that Executive agencies submit strategic plans and annual performance plans to OMB is not meant to diminish such autonomy. Rather, this is simply a reflection of the requirement placed on OMB to develop annually a Federal Government performance plan, based on the agencies' performance plans. In preparing the Federal Government performance plan, the Committee expects OMB to make allowances, as appropriate, for the greater degree of autonomy of TVA and similar agencies. This legislation is not intended to alter in any way the statutory authority of TVA or any other agency to plan or conduct its operations. Moreover, as the Committee has stressed, performance measurement under this Act should focus on those measures and that information useful to and used by program managers. Such an emphasis very much relies on a bottom-up approach, rather than top-down imposition. This premise is particularly valid for those agencies with significant autonomy for their business operations.

To obtain a comprehensive picture of the government's performance, the Committee believes it important to take tax expenditures in consideration. Tax expenditures are revenues foregone, and are specified in the tax code. They represent preferential exceptions to the baseline provisions of the tax structure and are created to provide benefits to qualifying individuals or entities, or to provide an incentive to encourage particular activities. Tax expenditures are similar to spending programs in their impact on the deficit; and like spending, are established to achieve specific national objectives. However, the effect of tax expenditures in achieving these goals is rarely studied.

To increase significantly the oversight and analysis of tax expenditures, the Committee believes that the annual overall Federal Government performance plans should include a schedule for periodically assessing the effects of specific tax expenditures in achieving performance goals. (This schedule would be in addition to the primary content of the overall plan-the program performance goals tied to the direct expenditure of funds.) The Committee expects that annual performance reports would subsequently be used to report on these tax expenditure assessments. These assessments should consider the relationship and interactions between spending programs and related tax expenditures. The Committee hopes that such reports will foster a greater sense of responsibility for tax expenditures with a direct bearing on substantial missions and goals.

The Committee also expects the Director of OMB to establish an appropriate framework for undertaking periodic analyses of the effects of tax expenditures in achieving performance goals, and to describe this framework in the May 1, 1997 report to the President and the Congress. In establishing the framework, the Committee encourages the Director to consult with the Secretary of the Treasury, the General Accounting Office, and the Joint Committee on Taxation.

AGENCY PERFORMANCE PLANS

The principal parts of agency annual performance plans shall include a set of performance goals for the agency's program activities along with a summary of the necessary resources, the performance indicators that will be used to measure performance, and an identification of how the measured values will be verified. Additionally, the annual agency plans shall contain any proposed waivers of administrative and procedural requirements as provided for by section 5.

Each plan as submitted to OMB should be consistent with the agency's budget request. The agency should then consult with OMB in revising that plan to be consistent with the President's proposed Budget of the United States Government.

After publication of the government-wide performance plan, agencies should provide copies of their final plans to their authorization and appropriation committees in the Congress, and make such plans available to the public.

The Committee recognizes that Congressional actions on agency funding requests could materially affect achievement of the goals established in an annual performance plan. The Committee prefers that agencies address in their annual performance reports any significant effects on the achievement of performance goals resulting from Congressional action. However, an agency may also elect to revise its annual performance plan to reflect Congressional budget action. Any revised plan should be completed prior to the start of the fiscal years for which it will be in effect.

A copy of the revised plan should be provided to OMB, the appropriate Congressional authorizing and appropriation committees, and made available to the public. No revisions should be made to the Federal Government performance plan for the overall budget required by section 1105(a)(29). If an agency prepares a revised annual performance plan, the program performance report should compare the actual performance achieved against both the performance goals in the revised plan and the performance goals as set forth in the Federal Government performance plan for the overall budget.

PERFORMANCE GOALS

Annual performance goals are the major means for gauging progress toward accomplishment of the longer-term general goals contained in the strategic plan. The Act specifies that most performance goals are to be expressed in an objective, quantifiable, and measurable form. It is important that a performance goal be defined with sufficient precision to permit ready assessment of progress in meeting that goal. There may be several performance goals for any general goal in a strategic plan.

The performance goals should conform with the level of resources requested for the relevant program activities in the Budget of the United States Government. Agencies are expected to make appropriate adjustments to the annual plans originally submitted to OMB in order to reflect the staffing and funding levels in the budget.

PERFORMANCE INDICATORS

For most performance goals, a number of performance indicators should be developed. Performance indicators are the reference markers used to measure whether a goal is being achieved. For example, a performance goal for the Indian Health Service might be to improve maternal and child health on tribal reservations. To measure this improvement, a number of indicators might be used, such as morbidity and mortality rates, median infant birth weights, immunization coverage percentages, frequency of pediatric checkups, etc. In some instances, the performance goal may be self-measuring and no separate indicators are needed. For example, a performance goal calling for the Federal Aviation Administration to staff 300 airport control towers on a 24 hour basis would be self-measuring.

The Committee believes agencies should develop a range of related performance indicators, such as quantity, quality, timeliness, cost, and outcome. A range is important because most program activities require managers to balance their priorities among several subgoals. For example, the farm loan program's goal may be to provide farmers with a source of credit if they cannot get it from the private sector (intended outcome). A manager of a farm loan program will want to make the most loans possible (quantity), ensure the loan will be repaid (quality), provide the loan when it is needed (timeliness), and at the least cost to the government (cost). Reliance on any single one of these measures could create a perverse incentive for managers to achieve one subgoal at the expense of the others. While the Committee believes that a range of measures are important for program management and should be included in agency performance plans, it also believes measures of program outcomes, not outputs, are the key set of measures that should be reported to OMB and Congress.

Wherever possible, agencies should include performance indicators that correlate the level of program activity with program costs, such as costs per unit of result, costs per unit of service, or costs per unit of output. The Committee acknowledges that such indicators require a financial system capable of assigning and aggregating costs by unit, and a clear and precise definition of the unit that will be measured. Many agencies are developing the capability to report on unit cost measures in the annual financial statements required by the Chief Financial Officers Act of 1990. It is important, however, that the indicators of unit cost be those most useful to agency staff in managing programs, rather than being simply those indicators primarily developed for financial management. The Committee expects agencies to assign a high priority to the development of these types of unit cost indicators.

VERIFICATION AND VALIDATION

In verifying and validating the measured values, an agency may use an audit or any other procedure that would support the general accuracy and reliability of information contained in the annual performance report. To the extent that the annual performance report contains audited performance information that is also included in the annual financial statements required by the CFOs Act, no further validation of such information is required. Agencies should note that the use of audits is not required for performance data contained in reports under the Government Performance and Results Act. But, again, the Committee emphasizes that as the success of the Act depends to a large degree on the reliability and utility of the information presented, special attention will be needed to ensure credibility. This will require efforts by all parties at all stages-agency data collection, OMB guidance and supervision, and Congressional and GAO oversight.

ALTERNATIVE FORMS OF MEASUREMENT

The Committee recognizes that not all of the many diverse program activities conducted by the Federal Government are susceptible to objective and quantifiable measurement. To be useful, performance measurement must be relevant, and reflective of the primary focus of the agency. Agencies should not trivialize measurement by seeking to measure performance in a forced or artificial way, simply to present a quantifiable measure.

In areas where meaningful objective measurement is difficult, an alternative form of measurement may be authorized by OMB. Preferably, the alternative will be in the form of two, somewhat subjective performance definitions-one of a minimally or marginally effective program, and one of a fully successful program. Recognizing that in some cases an agency may be unable to define a goal using these two descriptive statements, OMB may in those instances authorize the agency to define and use another alternative form. All forms must be in terms that would permit an independent determination of whether the program's eventual performance corresponded to the performance statement.

The Committee also recognizes that for a very small number of program activities, it may be infeasible or impractical to set performance goals in any form. For these program activities, OMB may authorize a brief explanation in the annual performance plan of why this is the case.

COVERAGE OF PROGRAM ACTIVITIES

The activity structure in the Budget of the United States Government is not consistent across various programs, being tailored to individual accounts. This lack of uniformity results in-for purposes of this Act-too many program listings for some accounts, and an overly broad aggregation for others. A single account may contain ten or more projects or activities. The Committee intends that the annual plans not be voluminous presentations describing performance at every level for every program activity. The annual plans and reports are to inform, not overwhelm the reader. The agencies are expected to use good judgment in determining the array of program activities presented in the plan, and to consolidate, aggregate, or disaggregate the lists of program activities appearing in the budget accounts.

Though the annual report will be the most public and visible presentation of performance data, the Committee expects that it will not be the only such presentation that an agency will make. Performance information will be useful at many different levels in the Federal government, from line managers to policymakers, each of which will need information presented at different levels of detail and frequency. Agencies should account for these varying uses of and needs for information in designing their performance measurement systems.

CLASSIFIED PLANS

The Committee recognizes that for certain government functions and operations, many key performance goals and indicators should not be publicly revealed. Publishing them could, for example, compromise national defense or undercut negotiating strategy with other countries. The Act allows an agency to prepare a classified or non-public annex to its annual performance plan covering program activities that relate to national security, the conduct of foreign affairs, criminal prosecution, or revenue collection. The parameters set out in the Freedom of Information Act may be useful to agencies in determining whether material belongs in such an annex. An agency preparing a classified non-public annex is expected to minimize the number of performance goals and indicators contained therein.

The Act does not provide for a similar classified or non-public annex to the strategic plan. The Committee believes that general goals can be defined in a way that avoids the need for classified goals. Should this presumption prove incorrect, or if those general goals become so imprecise as to be virtually useless, this should be addressed in the reports required by sections 6(a) and 8(b) of this Act, and in the strategic plan(s) for the agency. The Committee believes, in the interests of public disclosure and accountability, that an agency should indicate in its strategic plan whether it will have a classified annex to the annual performance plan.

DEFINITIONS

An "outcome measure" assesses the actual results, effects, or impact of a program activity compared to its intended purpose. Outcome measurement cannot be done until a program or project reaches either a point of maturity (usually at least several years of full operation for programs continuing indefinitely) or at completion. Another prerequisite for measuring outcomes is the existence at the outset (in statute, directive, or other document) of a clear definition of what results are expected from the program or project. While recognizing that outcome measurement is often difficult, and is infeasible for some program activities, the Committee views outcome measures as the most important and desirable measures, because they gauge the ultimate success of government activities.

An "output measure" records the actual level of activity or effort that was realized, and can be expressed in a quantitative or qualitative manner. Output measures are often intermediate, in that they assess how well a program or operation is being carried out during a particular time period, such as a quarter or a year. The number of schools and students participating in a national test of reading skills, and the percentage of eligible students receiving additional reading instruction, might be output measures, while improved national reading scores might be an outcome. Output measures in the annual performance plans should emphasize those used by agency officials in day-to-day operations and program management.

A "performance goal" is the target level of performance (either output or outcome) expressed as a tangible, measurable objective, against which actual achievement will be compared. An example of a performance goal for a student reading program would be to have 2.3 million students receive an average of three additional hours of reading instruction per week during the 1990 school year.

A "performance indicator" is a specific value or characteristic used to measure output or outcome. In other words, it is what will be measured. Quantitative indicators are used in measuring work-load, production, transactions, records, and various rates, such as utilization, consumption, and frequency. Qualitative indicators are used to measure timeliness, stoppage or out-of-service conditions, and various rates such as error or defect rates, inventory fill, and maintenance or repair intervals. Quality of service indicators include measures of complaints, customer satisfaction levels, and responsiveness rates. Efficiency indicators measure relative transaction or production costs. Financial indicators are numerous and can include receipt, collection, and credit obligation rates. Other examples of indicators include milestone and activity schedules, design specifications (such as hardware performance levels), operating parameters (such as mean failure rates), status of conditions (such as highway miles in good repair), and percentage coverage (such as eligible population).

The term "program activity" refers to the listings of projects and activities in the appendix portion of the Budget of the United States Government. That appendix contains one or more program and financing schedules for each agency, one part of which is the "Program by activities" section.

"Program evaluation" is an objective and formal assessment of the results, impact, or effects of a program or policy. While most often aimed at assessing the degree to which a program's stated objectives are being or have been realized, program evaluations are also frequently used for measurement of "unintended" results, good or bad, that were not explicitly included in the original statement of objectives or foreseen in the implementation design. Thus, they can serve to validate or find error in the basic purposes and premises that underlay a program or policy. Finally, this definition should be read as including evaluations of program implementation process and operating policies and practices when the primary concern is about these issues rather than program outcome. However, the definition is not intended to include program monitoring activities that are (or should be) a routine component of good program management.

PROGRAM PERFORMANCE REPORTS

Beginning with FY 1999, the head of each agency shall prepare and submit to the President and Congress a report on program performance. The first of these annual reports is to be submitted no later than March 31, 2000. These reports will contain two main parts: a report on the actual performance achieved compared to the performance goals expressed in the performance goals plan; and of the steps to be taken to achieve those goals that were not met. If a performance goal becomes impractical or infeasible to achieve, the agency should explain why that is the case and what legislative, regulatory, or other actions are needed to accomplish the goal, or whether the goal ought to be modified.

The agency's performance report must be submitted to the appropriate authorization and appropriations committees of the Congress, and copies provided other committees and to the public upon request. The agency shall also provide to any committee of Congress, upon request, more specific information on the actual performance for any performance indicator established in its annual performance plan. Agencies are required to begin reporting performance trends, on a phased-in basis, so that for FY 2002 and thereafter, performance data will be shown for each of the most recent four years.

A performance report shall also describe the use and assess the effectiveness of