process and inputs and focus on measures of the outcomes and impacts of its work. GAO's performance indicators should include not only estimates of net savings but also associated improvements in the efficiency and effectiveness of programs. GAO should also increase its internal resources and capacity to analyze the costs and budget impacts of its recommendations, as well as to assess agency cost accounting and performance measurement systems.

External Peer Review

GAO should establish and budget for external peer review of its completed work on a continuing basis. The volume and diversity of GAO work are not compatible with formal review before jobs start or as part of the report review process. GAO, however, should work with professional associations and other experts to organize a process for having specialized panels of objective peers from outside GAO assess and report on samples of completed GAO work, covering several selected categories each year.

GAO FUNCTIONS AND WORK PRODUCTS

The nature of GAO work has evolved over the past 30 years, resulting in less emphasis on audits, particularly financial audits, and increased resources directed to program- and policy-related work. The quality, usefulness, and importance of GAO's work products are uneven. The panel also found lack of clarity in the objectives of individual GAO reports and the standards or criteria for making findings and recommendations. Overemphasis on procedural controls still prevails in many studies. Categories of GAO Work

GAO currently undertakes six principal categories of work: (1) financial audits; (2) economy and efficiency audits (including non-financial compliance audits); (3) program evaluation; (4) policy analysis and policy development; (5) management studies; and (6) special investigations. These categories are not sharply defined and elements of several categories are frequently mixed in a single GAO study. Partly as a consequence, a job may lack clear objectives at the outset in conflict with the "yellow book" standards that GAO prepares as a guide for federal, State and local government auditors.

The panel recommends that GAO develop clearer definitions, objectives, and standards for all the categories of work it performs; strengthen results-oriented assessment of government activities (particularly in the category of program evaluation); continue to strengthen its capacity to audit and advise agencies on systems and standards for financial and information management; include substantially greater cost analysis in its work in audit, evaluation, and policy analysis; base its policy analysis and policy development on fact-based audit and evaluation; and restrict recommendations for added oversight and control that are not based on a demonstration of expected net benefits.

GAO's special investigations work (which seeks to identify criminal behavior in government) should be kept separate and distinct from audit, evaluation, and management analysis done by issue area staffs. The panel recommends that congressional oversight committees work with GAO to develop clear standards for GAŎ investigative work and guidelines for referring requests for investigations to federal law enforcement agencies.

General Management Reviews and Related Work

GAO is at a crossroads in its management work. It has developed a cooperative, consultative approach to conducting general management reviews and providing technical assistance to agency leadership on systems improvements, which many agencies have found constructive. The panel finds inconsistencies between the attitudes and approaches involved in process-oriented, adversarial audits and the skills and relationships needed for GAO to play a potentially growing role in providing general management analysis and technical assistance to executive agencies, with the support of Congress.

The research done for this study shows that adversarial relationships between auditor/evaluator and agency or program management seldom lead to productive general management improvements. Aggressive oversight can identify, expose, and punish mismanagement that violates law-a legitimate function, within limits-but it historically does not improve management in the long term. Providing effective review and technical assistance to agency management usually involves a cooperative relationship with executives responsible for management change.

The panel recommends that GAO launch a major internal training process to convey the lessons learned from its general management studies throughout its own issue area staffs to bring those lessons to bear on all audit and evaluation activities.

It should continue this area of work and build the lessons into the whole range of GAO work. To fulfill its statutory role under the Government Performance and Results Act of 1993, GAO should work cooperatively with OMB, executive agencies, and oversight committees, not to audit but to record the experiences in pilot projects under the act, in a non-adversarial context, in order to make a methodologically sound evaluation of the implementation of the law as required in 1997.

CONGRESS AND GAO

While the panel found no evidence of deliberate partisan bias in GAO work, Congress has increasingly embroiled GAO in political and policy controversy, threatening its impartial role and institutional standing. Increasingly adversarial relationships between congressional and executive entities in-the last 20 years and intraCongress rivalries make it hard for GAO to produce objective work and to be perceived as doing so. Any use of GAO that erodes perceptions of its fairness and objectivity reduces its usefulness.

The way Congress uses GAO largely determines GAO's continued value as an institution. Congressional views of GAO's mission vary widely: some see it as an independent audit agency; others treat it as a servant of Congress and an infinitely flexible addition to congressional staff resources. The demand for GAO work always exceeds the capacity and resources of GAO, particularly if GAO is to produce timely, high quality results. The result is a considerable variation in the time GAO can devote to, and the quality of work GAO can produce for, different jobs. Congressional Requests

Requests from congressional committees and subcommittees and individual members for GAO jobs have increased dramatically in number and scope, while GAO budgeted staff remained stable for a decade and declined 10 percent from 1992 to 1994. In fiscal year 1993, 80 percent of GAO staff-years of work were devoted to fulfilling congressional requests or mandates and the other 20 percent was for selfinitiated work, which falls under the name "basic legislative responsibilities" at GAO. This represents a sizable decline in self-initiated work from a decade earlier, when self-initiated work took 53 percent of GAO staff time.

The degree of actual change is not clear, however, because a large proportion of congressional requests involve continuing streams of work in the GAO issue areas or reflect GAO priorities raised in its internal strategic planning process. Many committee staff members are satisfied with the way GAO projects are developed, although some are uneasy with having GAO spend so much of its time fulfilling congressional requests and others would prefer that GAO not shape so much of its own work.

As much as possible, congressional requesters of GAO work should create a nonpartisan environment in which GAO can function. They should not assign to GAO report topics and research questions that would place GAO in the midst of valuebased debates and political controversies. Also, congressional requesters should expand their focus from immediate shortcomings of programs and agencies (the "gotcha" approach) to efforts to develop systems that will generate better management, timely and accurate information, and comprehensive program evaluation on a sustainable basis.

Detailees

An issue of continuing congressional controversy is the assignment of GAO staff to committee staffs as detailees. Detailees have been heavily concentrated in some committees and subcommittees, mainly serving the chair and working with the majority staff. Sometimes, detail assignments have continued substantially longer than the official 1-year limit. In the last year, GAO has committed to following the 1year limit on detail assignments and has cut the number of detailees by half. The panel recommends that GAO strictly follow the 1-year limit and that GAO publish an annual report of the numbers, length and scope of detail assignments.

Oversight and Reauthorization

GAO's congressional oversight committees need to do more regular, continuing oversight of GAO. A requirement for periodic reauthorization of GAO, however, could threaten GAO's ability to perform its mission as objective audit and evaluation agency.

The panel believes strongly that the Senate Committee on Governmental Affairs and the House Committee on Government Operations should provide continuing, systematic oversight of GAO. Congress can change GAO's authorizing statute at any time and affect GAO activities through work requests and annual appropriations.

The December 1993 reports of the Senate and House Members of the Joint Committee on the Organization of Congress recommended replacing GAO's permanent authorization with a required reauthorization every 8 years. Most panel members are concerned that a requirement for periodic reauthorization would put political pressures on GAO during every reauthorization cycle that run counter to the interest of Congress in ensuring objectivity and credibility of the central audit and evaluation agency.

ATTACHMENT II

BACKGROUND ON NAPA PANEL AND PROJECT DIRECTOR

Alan K. Campbell, chair of the NAPA panel, was most recently chair of the legislative-presidential Commission on the Social Security "Notch” Issue, which reported to Congress and the President in December 1994. Mr. Campbell was the first director of the U.S. Office of Personnel Management and chairman of its predecessor organization, the U.S. Civil Service Commission. He is former dean and professor at the Maxwell School of Citizenship and Public Affairs of Syracuse University, and also served as Executive Vice President and Vice Chairman of ARAMARK, and visiting executive professor at the Wharton School of the University of Pennsylvania. He has led and participated in several notable research projects and other activities of the Academy.

Annmarie Walsh, project director for the NAPA study, is the Luther Gulick Distinguished Scholar in Residence at the Institute of Public Administration in New York City, and former President of IPA, as well as former director of the Center for Urban and Policy Studies of the City University of New York Graduate School. Ms. Walsh has also participated in several major research projects at the Academy and is currently leading a number of public administration research and advisory services in central and eastern Europe.

Vice chair of the panel was Peter Szanton, who is also a Fellow of the Academy and a member of the Academy Board of Trustees, former Associate Director of the U.S. Office of Management and Budget, former President of Rand Institute-New York City, and currently President of Szanton Associates, specializing in organizational strategic planning. Mr. Szanton is also a member of the NAPA Board of Trustees and chair of its Program Committee.

The panel had 11 other distinguished members:

Norman A. Bolz-Former Vice Chairman, International Operation, Coopers & Lybrand, with 40 years experience in audit, tax, management and administration; currently Director and Chairman of Audit Committee for Computervision, Inc., and Intersolv, Inc.; former Associate Commissioner for Policy and Management, Internal Revenue Service, U.S. Treasury Department; former Director of Finance and Administration. Special Olympics International, Inc.; and President American/Saudi Arabian Business Roundtable.

Margaret ("Peggy") Crenshaw-Special assistant to chairman, U.S. Postal Rate Commission; former counsel and staff director for the Democratic minority. Senate Governmental Affairs Committee; publisher of Employee Benefits Review.

William Frenzel-Former Republican Congressman from Minnesota, now Guest Scholar in Governmental Studies, The Brookings Institution.

Edie N. Goldenberg-Dean, College of Literature, Science and the Arts, University of Michigan; Professor of Political Science and Public Policy and former Director of the Institute of Public Policy Studies, University of Michigan; former Chief, Civil Service Reform Evaluation Management Division, U.S. Office of Personnel Management. NAPA Fellow.

Richard L. Lawson-President, National Coal Association; retired four-star general of the U.S. Air Force, including service as deputy commander-in-chief of U.S. European Command, director of plans and policy for the Joint Chiefs of Staff and military assistant to the President.

Thomas E. Mann-Director of Governmental Studies and W. Averill Harriman Senior Fellow in American Governance, The Brookings Institution; former Executive Director, American Political Science Association. NAPA Fellow.

Charles McC. Mathias, Jr.-Former Republican Senator from Maryland; currently partner in Jones, Day, Reavis and Pogue in Washington, DC.

C. Thomas McMillen-Former Democratic Congressman from Maryland; currently Director of CliniCorp, Inc., West Palm Beach, Florida.

Chester A. Newland-Duggan Distinguished Professor of Public Administration, University of Southern California, Sacramento and Washington, D.C.; past editorin-chief, Public Administration Review; former Director, Federal Executive Institute. NAPA Fellow.

Constance Berry Newman-Under Secretary, Smithsonian Institution; former Director, U.S. Office of Personnel Management; former President, Institute for American Business. NAPA Fellow.

Roger B. Porter-Professor of Government and Business, Kennedy School of Government, Harvard University; formerly Assistant to the President for Economic and Domestic Policy, Executive Secretary, Economic Policy Council and Cabinet Council on Economic Affairs. NAPA Fellow.

PREPARED STATEMENT OF CHARLES A. BOWSHER

We are pleased to be here today to discuss the future direction of the General Accounting Office (GAO). We believe that periodic congressional oversight is critical to the effective operation of any federal agency, including GAO.

We are proud of GAO's long tradition of service to the Congress, the contribution we have made to improving government operations, and the billions of dollars in financial benefits achieved by Congress and the executive branch as a result of our work. Our mission is to achieve honest, efficient management and full accountability throughout government. We serve the public interest by providing Members of Congress and others who make policy with accurate information, unbiased analysis, and objective recommendations on how to use public resources in support of the security and well-being of the American people.

While we take pride in our accomplishments, we also recognize that our history is one of continual change, improvement and adaptation to the emerging needs of the Congress In that regard, GAO has a number of changes underway, many of which are consistent with the recommendations made in the report to this Committee by the National Academy of Public Administration (NAPA). We are pleased that NAPA recognized that GAO's statutory foundations are sound and that it found no evidence that GAO has been steering its research toward satisfying particular policy or partisan interests. Achieving and maintaining objectivity has always been a foundation for GAO work.

Both GAO and the Bureau of the Budget (now the Office of Management and Budget) were created in 1921, the outgrowth of efforts at that time to modernize federal budgeting and accounting practices after World War I. In return for creation of the Budget Bureau, which centralized budget authority of the executive branch in the Office of the President, Congress insisted upon creation of GAO within the legislative branch to monitor and audit federal expenditures.

Central to the ability of GAO to effectively carry out its mission over the past 74 years has been the nonpartisan nature of the organization-reflected most importantly in the 15-year term of the Comptroller General and the career civil service status afforded its staff. To effectively be the nation's auditor, GAO staff must have the independence necessary to report the facts, without fear of political retribution. GAO's ability to do this consistently over the past decades has served the Congress and the taxpayers well.

In fact, insuring GAO's independence and credibility were primary considerations in GAO's creation by the Congress. The Budget and Accounting Act of 1921 was drafted to severely limit the extent to which GAO could be subject to partisan political pressure. This is apparent not only in the Act's provisions regarding the Office of the Comptroller General, but also in the debate prior to the Act's passage. Representative Good (R-Iowa), a principal sponsor of GAO's original authorizing legislation, articulated this theme during floor debate.

"In creating the general accounting office and providing for the comptroller general and the assistant comptroller general, the Committee was guided by a single thought, and that was that these two officers should be placed upon a plane somewhat comparable to the position occupied by Federal judges. The positions are semijudicial, and it was the opinion of the Committee that we should remove them as far as possible from political considerations." 59 Cong. Rec. 8610(1920)

The Committee intended, "that the Comptroller General should be something more than a bookkeeper or accountant; that he should be a real critic." (61 Cong. Rec. 1090 (1921). To ensure that result, the Comptroller General was given a 15-year term, during which he is removable only by impeachment or, for cause, by joint resolution.

GAO'S HISTORY IS ONE OF CONSTANT CHANGE AND IMPROVEMENT

In the 1920's and 1930's, in an era of limited government, GAO served the Congress primarily by auditing vouchers of the government to make certain that federal payments were correct. But just as Congress adapted to an increasingly complex world as the decades passed, so GAO changed to help Congress grapple with more difficult and complex issues.

The modern era for GAO began after World War II. Key to changing the GAO was passage of the Budget and Accounting Procedures Act of 1950, which authorized GAO to review federal agency accounting procedures and controls in lieu of checking thousands of ledger accounts. Out of this change grew the development of the comprehensive audit approach-a way to determine how well each federal agencarried out its financial responsibilities.

су

In 1966, Elmer Staats, who had served in the administration of four presidents as Deputy Director of the Bureau of the Budget, became the Comptroller General. Under his leadership, GAO began to move into the area of program evaluation to determine whether government funds were spent efficiently and whether federal programs were effective.

To deal with these complex issues, GAO today is staffed with well trained specialists in many disciplines-accountants, computer scientists, public administrators, business school graduates, engineers, statisticians, lawyers, economists, and others. At this year's authorized staff level of 4,700, GAO is about the same size as it was in 1963, when the federal budget was just approaching $100 billion for the first time. The federal budget today is more than $1.5 trillion, a 15-fold increase, and many agencies audited by GAO have grown several times over since the early 1960's. GAO has worked hard to keep pace with the increasing demands placed upon it in those intervening years. By continually working to increase productivity, to take advantage of modern technology, and to improve the skills and expertise of its staff, GAO has been able to keep pace with the added requirements and responsibilities that it now faces as a result of the growth in size and complexity of the federal government. GAO's productivity doubled between 1983 and 1992.

And we are proud of the results of our efforts. Although total audit coverage has never been possible, we have contributed significantly over the years to saving billions of dollars for the American taxpayer. Since 1981, GAO's work has resulted in more than $200 billion in measurable financial benefits.

In 1994 alone, GAO contributed to legislative and executive branch decisions and actions that resulted in more than $19 billion in measurable financial benefits. GAO thus generated $43 dollars in financial benefits for every dollar appropriated.

Our goal is to continue that trend. GAO's efforts are designed to help the Congress produce a smaller, more effective and efficient government, in both the audit work it completes and in the way it manages its own operations. The plan I have submitted to the Congress to reduce our staffing level to 3,975 by 1997 (a 25 percent reduction over 5 years) would result in recurring payroll savings of $130 million each year while still enabling GAO to effectively fulfill its mission. GAO will thus be at the lowest staff level since before World War II. The House Budget Committee suggestion of a 15 percent reduction in our budget next year is based on our plan to downsize GAO by 25 percent. But, this plan would increase the pace of our downsizing effort and we are examining how we would absorb a reduction of that size.

Critical to our ability to successfully accomplish this reduction is our continuous improvement effort that has enabled us to focus on reengineering our job management processes and use enhanced technology to improve to the way we do the work. We are tracking these efforts and implementing ways to measure progress in terms of time and cost reductions and productivity improvements.

While GAO is willing to become smaller, the pace at which that happens must not cripple GAO's audit and investigatory responsibilities and capabilities. A precipitous cut would jeopardize billions of dollars of financial benefits that flow from GAO's work each year. Significant staff disruptions would occur, resulting in loss of expertise and disruptions to ongoing work that could last up to 12 years. No government organization I am familiar with has been able to absorb a 25 percent cut in its operating budget in 1 year, especially when between 75 and 80 percent of its budget goes for salaries and expenses as does GAO's, and still maintain its capacity to deliver quality products. GAO is not an exception.

Our assistance to the Committees of the 104th Congress reflects how we continually seek to fulfill our mission of supporting the oversight and legislative needs of the Congress. Since January 1995, about 90 percent of all new request work we have started has been for the majority. This work includes helping target budget reductions, proposals for civil service reform and downsizing government, privatiza