The Evolution of the General Accounting Office: From Voucher Audits to Program Evaluations

Published by the Government Accountability Office on 1990-01-01.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

                 United   States   General   Accounting   Office

GAO         I    History Program

January   1990
                 The Evolution of the
                 General Accounting
                 Office: From Voucher
                 Audits to Program
                 Harry S. Havens
 Published by the United States General Accounting Office,
 Washington, D.C., 1990

ii                                                           GAO/OP-Z-HP   ;

          Harry S. Havens, Assistant Comptroller General of the TJnited States
          since 1980, spoke on November 3, 1989, at the Annual Conference of the
          Association for Public Policy Analysis and Management about the his-
          torical evolution of the General Accounting Office (GAO). In addition to
          providing an interesting perspective on the main eras in GAO'S develop-
          ment from 1921 to the present, Mr. Havens provides a perceptive analy-
          sis of change in GAO'S organization and work over the last 25 years,
          During this period, GAO shifted from an organization basically concerned
          with operational auditing to one also using policy analysis and program
          evaluation. In this way, GAO could provide decisionmakers with key
          information for use in dealing with major issues of the day, Mr. Havens
          also provides insight on such topics as GAO'S relations with the Congress,
          utilization of staff resources, professional training and development,
          and GAO'S institutional credibility.

          GAO is publishing Mr. Havens’s important speech to make it available to
          a wider audience. Copies of this publication are available to GAO officials
          and staff and other interested persons.

          Charles A. Bowsher
          Comptroller General
          of the United States

          111                                                             GAO/OP-Z-HP
Harry S. Havens

                  Harry S Havens

                  iv               GAO/OP-2-HP
Biographical Information

                Harry S. Havens was appointed Assistant Comptroller General in the
                IJnited States General Accounting Office on April 18, 1980. He has han-
                dled a variety of special assignments for the Comptroller General,
                including heading GAO’Swork under the Gramm-Rudman-Hollings legis-
                lation and GAO’S study of the 1987 stock market crash. Mr. Havens also
                serves on a number of internal management boards and committees.

                Prior to his present assignment, Mr. Havens served in the U.S. Navy
                from 1957 to 1964, with the Bureau of the Budget (now the Office of
                Management and Budget) from 1964 to 1974, and as Director of GAO'S
                Program Analysis Division from 1974 to 1980.

                Mr. Havens graduated magna cum laude from Duke University in 1957
                with a B.A. in economics. He attended University College, Oxford,
                England, from 1957 to 1959 as a Rhodes Scholar, receiving a B.A. in
                politics and economics in 1959 and an M.A. in 1963.

                Mr. Havens received t,he Comptroller General’s Award in 1978, 1981,
                and 1986 and the Comptroller General’s Certificate of Merit in 1981; he
                was awarded the rank of Meritorious Executive in 1983. He received the
                Roger W. Jones Award for executive leadership in 1980 and was elected
                a Fellow of the National Academy of Public Administration in 1981.

                Mr. Havens has published a number of articles on program evaluation,
                budgeting, and financial management.

vi   GAO/OP-2-HP
The Evolution of the General Accounting Office:
From Voucher Audits to F?rogrm Evaluations1
by Harry S. Havens, Assistant Comptroller
General of the United       States
                The United States General Accounting Office is a nonpartisan agency
                            in the legislative branch of government. It was created by the Budget
                            and Accounting Act of 192 lz with the statutory responsibility
                            (among other things) to “... investigate, at the seat of government or
                            elsewhere, all matters relating to the receipt, disbursement, and
                            application of public funds.. ..”

                            To understand the GAOof today, it is necessary to reflect on the origins
                            of the institution and its evolution over the seven decades of its exis-
                            tence. To recognize what GAO is today, and why, one must know from
                            whence GAO came.:’

                            The creation of GAO was an outgrowth of the same “good government”
0rig:ins and Early          era in the United States that gave rise to the executive budget at the
Years                       federal level and the creation of the U.S. Bureau of the Budget. Indeed,
                            they were established in the same legislative act. Moreover, it seems
                            clear that one of the primary motivations for including GAO in the pack-
                            age was the congressional view that if the President was to have the
                            enhanced power flowing from control over budget proposals and the
                            creation of “his” Budget Bureau, the Congress needed to have a counter-
                            weight in the form of enhanced ability to see how the money that it
                            appropriated was being spent. This counterweight was to be GAO.

                            The context in which GAO was created greatly influenced its early his-
                            tory and development,. The audit function had a long history in the exec-
                            utive branch, where the Comptrollers of the Treasury, GAO’S
                            predecessors, had existed since the earliest days of the Constitution. The
                            Congress, however, was suspicious of any arrangement leaving the audi-
                            tor subservient to the President. Accordingly, the 1921 act included
                            strong protections for the independence of GAO, of which the most
                            important are those relating to the tenure of its leader, the Comptroller
                            General of the United States.

                            ‘This paper was originally presented at the Annual Conference of the Association for I’ublic Policy
                            Analysis and Management. Arlington, Virginia, Nov. 3, 1989.

                            “I’ublic Law 67-13, approved .June 10, 1921

                            “For a more detailed examination of the history of GAO, see The GAO The Quest for Accountability
                            in American Government (Houlder, C,olorado: Westview Press, 1979) and A Tale of Two Agencies
                            (Hat.on Rouge, Louisiana: Ltrmsmma State IJniversity Press, 1984), both by Frederick C. Mosher.
                            Moshcr’s work is the primary basis for the historical discussion in this paper. Another valuable
                            source is May Hunter Wilbur’s An Early History of the General Accounting Office, 1921-1943, written
                            in 1943 but first published in Mar. 1988 (GAO/OP-l-HP).

                            Paye 1
The Evolution   of the General Accounting

The office of the Comptroller General is unique for the head of an
agency. The President appoints the Comptroller General, subject to Sen-
ate confirmation.3 Once confirmed, the Comptroller General serves a
fixed term of 15 years, from which he is removable only by impeach-
ment or by joint resolution. Neither removal procedure has ever been
attempted. Ilpon completion of his term, the Comptroller General may
not be reappointed and retires with full salary. The effect of these
arrangements is to insulate the Comptroller General and GAO to a very
substantial degree from improper external influence.                                                     1

At the beginning and for the next two decades, the Comptroller Gen-
eral’s independence from the President was a contentious issue. Wilson
vetoed the first version of the act in 1920, arguing that the Comptroller
General should be removable at the pleasure of the President. The Con-
gress disagreed and, after minor changes, resubmitted the bill to Har-
ding, who approved it.

Initially, GAO’S location in the structure of government was ambiguous.
Clearly, GAO was created at the instigation of the Congress, to satisfy its
wishes for an independent auditor to whom it could look for informa-
tion. At the same time, the presidential appointment of the Comptroller
General, together with GAO’S inheritance from Treasury of some argua-
bly executive functions, led some to conclude that GAO straddled the
executive/legislative boundary, having a foot in both camps. Only 65
years later was the issue settled once and for all.

GAO has gone through three distinct eras since its creation in 1921. The
first lasted until the end of World War II. In this period, GAO'S work con-
sisted largely of detailed auditing of individual vouchers. It was charac-
terized by a highly formal, legalistic review of each voucher, with
approval for payment and the settlement of the affected accounts being
dependent on the payment’s conforming to an elaborate set of rules gov-
erning the use of public funds.

This control-focused, quasi-judicial interpretation of the audit function
was typical in the 19th and early 20th centuries. In many nations, the
government audit agency is called the Court of Accounts, indicative of

“Under a new procedure. instituted by the General Accounting Office Act of 1980 (Public Law 96-226,
approved Apr. 3: 1980), thr President is expected to select from a list prepared by a special biparti-
san congressional commission. It would be constitutionally impermissible to require the President to
select from such a list. Thus, the statute provides that the President may ask the commission to
sugpest additional names and, at least in principle, the President could nominate someone who did not
appear on any list. Hnwever. in the one case in which this procedure was used (the appointment of
the present Comptroller General). President Reagan accepted the first list and selected from it.

Page 2
The Evolution   of the General Accounting

the quasi-judicial origins, and some are still limited to reviewing the
legality of transactions.

For GAO, this concept of auditing included functions that remain part of
its mission. These include reviewing the validity of government con-
tracts (a responsibility now embodied in the Competition in Contracting
Act; > and providing advisory opinions to executive branch officials on
the legality of expenditures. Usually these activities receive little notice,
but occasionally they attract attention, as when the Navy seeks to pass
the buck to the Comptroller General in deciding if it is legal for Lieuten-
ant Colonel Oliver North, of Iran Contra fame, to continue receiving pay
as a retired officer after his conviction for improperly destroying gov-
ernment documents.‘1

Officially, GAO in the early years took no responsibility for judging the
wisdom of a particular expenditure, only its legality. However, the con-
servative leanings of the first Comptroller General (J. Raymond McCarl,
who served out his full 15-year term from 1921 to 1936) brought him
into frequent conflict with Roosevelt’s New Deal, which McCarl saw as
full of wasteful spending. In one now humorous incident, GAO questioned
how the Interior Department intended to use a camera it had purchased.
The penciled response to the auditors from Secretary Harold Ickes was
brief and to the point, “To take pictures, you damned fool.”

McCarl’s retirement in 1936 created a special opportunity to reassess
GAO’S  status. But this was not a new question. Harding, who had initially
signed the 1921 act, almost immediately changed his mind and wanted
to move GAO back to Treasury. In like vein, Hoover sought to strip GAO of
functions deemed executive in character. Roosevelt was of a similar
view. The landmark Brownlow Committee7 recommended a major
revamping of GAO that would have limited GAO exclusively to a postaudit
role. All these proposals were rejected by the Congress, which refused to
countenance any tinkering by the President with its GAO.

“Enacted as part of the Deficit lieduct.ion Act of 1984, Public Law 98-369, approved July 18, 1984.
The Competition in Contracting Act appears at 98 Stat. 1175.

“For technical reasons, GAO did not issue a formal opinion in this case. In a letter responding to the
Kavy, however, the GAO General Counsel pointed nut the long-standing interpretation of law that
reduced pay for retired officers (commonly called retired pay) is for contmued service as an officer
wth reduced responsibilities. North was convicted under a stat&- for which one of the prescribed
penalties is a prohibition on serving in any public office. Service as a tommissioned officer, even in
retired status, falls within 1his prohibition.

iKcport of the President’s Cbmmittre on Administrative Management (\f’zshington,      DC.: 1T.S.   Gnvrrn-
mcnt Printing Office, 1937).

Page 3
                 The Evolution   of the General Accounting

                 When Roosevelt finally accepted this outcome, he selected a Comptroller
                 General whom he expected to be more politically sympathetic. Thus,
                 Lindsay Warren, a Congressman from North Carolina, became, in effect,
                 the second Comptroller General. (Fred Brown, appointed in 1939, was
                 officially the second but served barely a year before resigning for health

                 The ideological conflict between GAO and the New Deal, which had
                 incurred Roosevelt’s wrath during the McCarl era, never surfaced with
                 Warren. With the advent of World War II and the crushing burden of
                 mobilization on all the institutions of government, ideological issues
                 evaporated. Instead, Warren, who served from 1940 to 1954, eventually
                 presided over the first great institutional transformation of GAO.

                 The voucher audit process, which had begun to creak badly under the
The Second GAO   weight of a growing government during the New Deal, utterly collapsed
                 in World War II. Even with a staff of over 14,000, of whom most were
                 auditing clerks, GAO could not begin to keep up with the waves of paper.
                 The GAO annual report for 1945, for example, reported a backlog of 35
                 million unaudited vouchers.

                 Comptroller General Warren recognized the futility of trying to return to
                 the old ways. Thus, in 1947, after the end of the war, he joined with
                 Treasury Secretary John Snyder and Budget Director James Webb to
                 design a new approach to financial management and auditing. The
                 departments and agencies would do their own voucher checking and
                 accounting; GAO would concentrate on prescribing accounting principles
                 and checking the adequacy of financial management procedures and
                 controls. This concept was subsequently endorsed by the First Hoover
                 Commission and embodied in the Accounting and Auditing Act of 1950
                 (title II of the Budget and Accounting Procedures Act of 195O).x

                 This first transformation of GAO was traumatic. From a total staff of
                 almost 15,000 at the end of the war, GAO shrank to only about 6,000 by
                 the end of the Warren era. The actual numbers who left GAO during this
                 period was even larger than this would imply. The new GAO had no use
                 for the army of auditing clerks who had populated the organization
                 before and during the war. Instead, GAO needed-and hired-the new-
                 st,yle auditors (trained in accounting) who were required for the new

                 ‘Public Law 81-874, apptwwl Sept. 12, 1960.

                 Page 4
The Evolution   of the General Accounting

Under the leadership of Warren and Joseph Campbell, who was
appointed by Eisenhower, served from 1954 to 1965, and was the first
certified public accountant (CPA) to hold the office of Comptroller Gen-
eral, the GAO of the 1950s and 1960s was modeled on the public account,-
ing firms of the time, which had, themselves, changed dramatically since
the Great Depression. It was increasingly staffed by professional
accountants, many being CPAS, and a significant number of GAO’S leaders
in this era were recruited directly from public accounting firms.

The shifting composition of the GAO staff under Campbell extended the
trends begun under Warren. The size of the staff continued to decline
and was barely 4,000 when Campbell left office. By then, well over half
the staff had college or university training in accountancy and very few
of the auditing clerks remained. (One vestige of the earlier GAO-the
transportation rate audits-remained      until the mid-1970s.)

Even in the Campbell era, with its emphasis on replicating the model of
the CPA firm, GAOnever confined itself exclusively to financial reporting
matters. A good deal of effort was still devoted to reviewing the legality
of agency activities. While this was not the old-style voucher auditing,
the underlying attitudes had much in common in that both involved
what some observers characterized as a nit-picking search for impropri-
eties. Therein lay the seeds of a problem for GAO.

In the early 196Os, GAOhad grown increasingly shrill in its criticism of
the defense contracting community, repeatedly issuing public reports
alleging overpayment and demanding that individually named contrac-
tors make voluntary refunds. This incurred the wrath of the defense
contractors. The controversy culminated in 1965 in what GAOers still
call the Holifield Hearings, named after the Chairman of the Military
Operations Subcommittee of the House Committee on Government Oper-
ations, a Congressman from southern California.

Chairman Chet Holifield’s district was not heavily defense industry ori-
ented. According to at least one observer,” Mr. Holifield was ambivalent
about the issue but was under pressure from other members of the Cali-
fornia delegation (urged on by the contractors and, some believe, by the

“Mr. Alex Silva, currently with GAO, was on the staff of Congressman Ronald Cameron of California
at the time of the Holifield hearings

Page 6
                 The Evolution   of the General Arrounting

                 Defense Department) to use his position to force GAO to end its aggres-
                 sive audits.IO In any event, Chairman Holifield severely castigated GAO
                 for its supposedly unfair treatment of the defense industry.

                 These criticisms of GAO were not unanimously supported within Chair-
                 man Holifield’s own Subcommittee. There was some strong opposition,
                 most notably in the person of a future Chairman of the full Committee,
                 Congressman Jack Brooks. Nevertheless, in the Committee’s report,”
                 Chairman Holifield prevailed. Whether or not his criticism was justified,
                 the episode had two effects. First, it almost certainly hastened the
                 retirement of Campbell for reasons of health and, second, it probably
                 caused GAO to become much less aggressive in its audits of defense

                 Lyndon Johnson chose Elmer Staats as Campbell’s successor. Staats was
The Modern GAO   an economist and a career civil servant with almost 30 years’ experience
                 in the Bureau of the Budget, including service as Deputy Director under
                 four Presidents, This set the stage for the second great transformation
                 of GAO. While less traumatic in outward appearances than the shift from
                 voucher auditing to the CPA model of the late 1940s and early 195Os, the
                 transformation begun by Staats was no less substantial in its effects.

                 Staats, who served a full term as Comptroller General from 1966 to
                 1981, brought with him from the Budget Bureau the analytical
                 approaches that were gaining currency in the executive branch as part
                 of the planning, programming, budgeting system (PPBS), a highly struc-
                 tured, analytically based framework for decision-making. Staats
                 believed, and quickly convinced others, that the forms of analysis
                 employed in PPRS would be useful to the Congress in overseeing federal
                 programs and that providing them would be a logical extension of GAO’S
                 auditing function.

                 “‘Mr. Holifield, in an oral history interview two decades later, had little memory of the events that
                 loom  so large in GAO’s history. This would be consistent with the view that the impetus for the
                 hearings came from elsewhere. See Chet Ilolifield (GAO/OP-5-OH, Apr. 1988).
                 ’’17S Congress, House, Defense Ccmtract Audits, House Rept. 1132,NJth Gong., 2nd sess., Mar, 23,

                 ‘“Other factors undoubtedly contributed to GAO’s lessened interest in contract auditing, including
                 the creation of the Defense Contract Audit Agency (DCAA), an organization devoted exclusively to
                 that task.

                 Page 6
                        The Evolution   of the General Accounting

                        These new GAO responsibilities were soon given a statutory foundation
                        in the Legislative Reorganization Act of 19701:’and then reenacted in
                        expanded form in the Congressional Budget and Impoundment Control
                        Act of 1974.‘” Introducing these concepts of policy-oriented analysis into
                        GAO would ultimately involve changing the character of the

                        Initially, the Staats initiatives required grafting onto the existing CPA
                        and management audit functions the new roles of program evaluation
                        and policy analysis, together with a staff having the training and the
                        technical skills needed for this work. Over the longer run, the new func-
                        tions (and the people performing them) had to be integrated into the
                        mainstream operations of the institution, creating a truly interdiscipli-
                        nary organization capable of carrying out the full range of audits, man-
                        agement reviews, evaluations, and other studies of widely varying
                        complexity demanded of it by the modern Congress.

Introducing   Program   In 1967, only a year after Staats took office, GAO was asked to make its
Evaluation              first major program evaluation. Senator Winston Prouty sponsored a
                        provision of the Economic Opportunity Act Amendments requiring GAO
                        to review the effectiveness of the poverty programs. This work, per-
                        formed by GAO staff with the assistance of a number of consultants,
                        culminated in one overall report in early 1969’” and numerous separate
                        reports on issues and problems at the particular sites reviewed by GAO,

                        The success of GAO’S efforts under the Prouty Amendments demon-
                        strated that we could examine complex issues affecting politically sensi-
                        tive programs without great risk to the institution, so long as the work
                        was done carefully and professionally. These were standards with
                        which the GAO staff could be very comfortable, even though they were
                        still learning how to apply those standards to a new type of work.

                        Success in GAO breeds emulation, and as GAO moved into the 197Os, pro-
                        gram evaluations (usually called program results audits) became an
                        increasingly common activity for the organization. GAO reviewed the
                        effectiveness of the municipal waste water treatment construction grant
                        program and the New Jersey negative income tax experiment. By the

                        ‘,‘Public Law 91-510, approved Oct. 26, 1970.

                        “Public Law 93-344, approved July 12, 1974.

                        ‘“Review of Economic Opportunity Programs (B-130515, Mar. 18, 1969)

                        Page 7
                    The Evolution   of the General Accounting

                    mid-1970s, GAO'S growing capacity for this type of work was touching on
                    issues as diverse as income tax withholding for military personnel,
                    planned bed capacity of Defense Department hospitals, the cost-effec-
                    tiveness of automobile safety devices, and the interchange of data
                    among agencies administering welfare benefits.

The Vietnam War     Ventures into the field of program evaluation were not the only new
                    directions for GAO in this era. The troubles in Southeast Asia touched
                    GAO as they did the rest of the nation. As the tJ.S buildup of forces pro-
                    ceeded, the Congress began asking questions about how the money was
                    being spent; it became apparent that a GAO presence was needed.
                    Accordingly, for several years, GAO maintained an office in Saigon with
                    up to 30 permanent staff, often augmented by staff from other locations
                    on temporary duty.

                    GAO'S  work in Vietnam focused not on the actual conduct of the war but
                    on the closely related issues of supply and logistics. Questions involved
                    everything from how U.S. assistance for the Vietnamese government
                    was being spent, to the wisdom of the huge construction projects to sup-
                    port the tJ.S. involvement, the appropriateness of inventory levels, and
                    the efficiency with which material was shipped from the IJnited States
                    and other points of origin. An army’s ability to wage war depends on its
                    supply lines, so some of the issues were vital to the war effort. In other
                    cases, GAO was simply going about its traditional task of trying to save
                    money. One veteran of that effort, for example, recalled a report ques-
                    tioning why it was deemed necessary to ship table tennis balls from the
                    United States by air freight.“’ In addition to examining matters directly
                    related to the military effort, GAO also became involved in assessing the
                    problems of refugees, an area of continued GAO attention today.

The Energy Crisis   Trouble in Vietnam was followed by trouble at home, and the oil supply
                    disruption of 1973 was a major factor in GAO'S movement into policy
                    analysis. As the gas lines lengthened, the Congress turned to GAO for
                    reliable information on what was happening. In a relatively traditional

                    "'Defense-Helatcd Audits, 1937-1975 (GAO/OP-g-OH, Feb. 1989). p.58.

                    Page 8
The Evolution   of the General Accounting

role, GAO began assessing the quality of available data. This was formal-
ized in several pieces of legislation, culminating in the Energy Conserva-
tion and Production Act,17 which requires GAO to review, on a regular
cycle, the quality of data produced by the Energy Information Agency.

In addition to seeking reliable data, however, the Congress also sought
advice on how to deal with some of the critical issues of the day. This
reflected a growing distrust of the executive branch after the experience
of Vietnam and Watergate. GAO responded at several levels. One
involved looking at energy supply and demand, leading GAO to develop
forecasts of future conditions. Another involved looking at some of the
proposed “solutions” to the energy shortage. An example was the pro-
gram to develop breeder reactors, designed to produce more fuel (in the
form of plutonium) than they consumed (in the form of enriched

The breeder reactor concept emerged in the 1960s when nuclear power
was assumed to be a major future source of energy in the United States
and there was concern about the long-term adequacy of the uranium
fuel supply. As part of the program to develop the Liquid Metal Fast
Breeder Reactor (LMFBR), one of several possible breeder reactors, the
Nixon administration proposed building a demonstration plant, the
Clinch River Breeder Reactor (CRBR). The project was soon mired in

Opponents argued that breeder reactors were unneeded because of the
declining public acceptance and deteriorating economics of nuclear
energy. They also emphasized the risks of nuclear proliferation implicit
in the plutonium fuel cycle. Supporters pointed to the need to keep
available the nuclear power option, which would ultimately require
developing the breeder concept, The Carter administration tried to kill
the CRBR; the Congress kept it alive but without enough money to com-
plete the project.

Throughout the 1970s, GAO issued numerous reports on the LMFBR pro-
gram and the CRBR project, discussing problems and progress, or the lack
thereof. Finally, in 1980 and 1981, GAO advised the Congress to “fish or
cut bait.“18 GAO favored retaining the nuclear option and continued
research on the breeder concept, including completion of the CRRR

“Public Law 94-386, apprtwed Aug. 14, 1976.

‘“[J.S. Fast Breeder Reactor Program Needs Direction (GAO/KvlD-80-81. Sept. 22, 1980) and letter to
various  members (R-199272. May 4,1981).

Page 9
                           The Evolution   of the General Accounting

                           project. But the current situation was a waste of hundreds of millions of
                           dollars each year without any visible progress toward completion of the
                           project. GAO recommended that the Congress either require the Depart-
                           ment of Energy to complete the CRBR or terminate it. Shortly thereafter,
                           the Congress terminated the CRBR.

Reemphasizing Defense in   The direction of GAO'S evolution as established under Staats, with its
                           growing emphasis on policy-oriented evaluation and analysis and a close
the ’80s                   working relationship with the Congress, was further intensified under
                           Charles Bowsher, the current Comptroller General, who was appointed
                           by President Reagan and assumed office in 1981. Bowsher was the sec-
                           ond CPA to hold the position.

                           This continuing evolution can be seen in GAO'S work relating to defense.
                           The Holifield hearings of 1965 had put a damper on that work, but only
                           temporarily and, even then, on only one component-the audits of
                           defense contractors. In the 197Os, GAO relied primarily on DCAAto do the
                           detailed contract audits. Meanwhile, GAO (in addition to doing its work in
                           Vietnam) emphasized broader issues concerning logistics; communica-
                           tions; personnel management; and -most prominently-the        cost, the
                           schedule, and the performance of major weapon systems in development
                           and production. The experience gained in this work provided a strong
                           base for expansion in the 1980s in response to the defense buildup
                           begun by President Carter and accelerated dramatically under President

                           In common with the rest of GAO, the defense work in the 1980s involved
                           a steadily widening mix of issues extending from the very detailed to
                           the very broad. Routine assignments ranged from reviewing problems in
                           the way the military orders spare parts to assessing the problems in the
                           B-l bomber avionics systems and the resulting implications for the B-l’s
                           ability to perform its assigned strategic mission. In GAO defense work,
                           however, the “widening” process was particularly dramatic in both
                           directions. For example, GAO returned to detailed auditing of individual
                           defense contracts, but with a very different purpose than was seen in
                           the 1950s and 1960s. The focus was on the systemic causes of contract
                           overpricing and cost overruns. Revelation of abuses was not an end in
                           itself, but a way of dramatizing the systemic problems.

                           While reimmersing itself in some of the details of contract auditing, GAO
                           broadened its range of defense work in the other direction as well. The
                           “big” issues became bigger and broader. GAO went from looking at the

                           Page 10
                           The Evolution   of the General Accounting

                           ability of the Aegis cruiser to perform its assigned mission to                 examining
                           the vaunted 600-ship Navy, assessing potential financial and                    operational
                           implications of the mismatch between the composition of the                     fleet and
                           the maritime strategy it was supposed to carry out in time of                   war.

Financial Management and   The expanding scope of GAO’S activities in the 1980s is also exemplified
                           by GAO’S increasingly outspoken concern about the financial condition of
the Budget                 the federal government and the fiscal policy it was pursuing and by the
                           role that the Congress sought to assign GAO under the Gramm-Rudman.
                           Hollings (GHH) deficit reduction mechanism.

                           In the Congressional Budget Act of 1974, the Congress assigned most of
                           the responsibilities for supporting the congressional budget process to
                           the newly created Congressional Budget Office. However, GAO was given
                           an oversight and assistance role with respect to budget systems and con-
                           cepts and the quality of budget information. In addition, GAO retained its
                           responsibilities for overseeing the government’s accounting systems. In
                           the 197Os, GAO pursued these responsibilities in a low-key manner. As
                           the 1980s unfolded, however, GAO became increasingly concerned about
                           the government’s finances.

                           While recognizing that the underlying problems of fiscal policy in the
                           1980s were grounded in unresolved political and ideological conflict, GAO
                           nevertheless concluded that inadequacies in budget presentation and
                           process were significant contributing factors. At the same time, archaic
                           accounting and management information systems often made it impossi-
                           ble to obtain reliable information as a basis for managerial decisions and
                           policy oversight, while weak internal controls created a breeding ground
                           for the type of fraud and mismanagement that. engulfed the Department
                           of Housing and Crban Development (HIJD).

                           These concerns culminated in the 1985 publication of the first in a series
                           of GAO “white papers” on the federal government’s financial manage-
                           ment system.l!’ GAO assessed the problems of the current structure and
                           suggested a conceptual framework for a new financial management sys-
                           tem embodying an int.egrated approach to the budget and accounting

                           “‘Managmg the Cost nf Government: Building an Effective Financial Management Structure (GAO/
                           AFMD-86-36 and 36A-two vols.-Feb. 1985).

                           Page 11
The Evolution of the   General   Accounting

Complementing this, GAO renewed its emphasis on financial auditing.
GAO urged the departments and agencies to produce meaningful annual
financial statements and to ensure the reliability of those statements
through independent audit. Those views gained increasing acceptance,
to the point where GAO, in 1989, audited the financial statements of the
U.S. Air Force, an effort whose scope exceeded that of any previous
financial audit in history.

Meanwhile, the continued inability of the Congress and the President to
agree on an acceptable budget policy set the stage for enactment of the
GRHlegislation. “(IFor reasons that remain obscure, the Congress chose
GAO as the agency to judge whether or not the deficit was going to
exceed the statutory thresholds and, if so, for allocating the required
spending reductions.‘l

The Supreme Court, however, ruled that GAO'S involvement in the GRH
arrangement was unconstitutional.?’ The Court, pointing to the joint res-
olution procedure for removing the Comptroller General, concluded that
GAO was a legislative branch agency. The functions involved in imple-
menting GRII were of an executive nature and therefore could not be
assigned to the Comptroller General.

Many observers question putting so much weight on a provision that has
never been used (or even threatened), as well as the very rigid view of
the separation of powers adopted by the Court. Nevertheless, the Court
was clearly correct in recognizing that, whatever the original intent,
GAO’S evolution over the past 40 years had moved the institution much
closer to the Congress; it is now overwhelmingly a legislative branch
agency. The Court finally resolved the issue that had troubled GAO from
its early years.

GAO was removed from the GRH operating mechanism after pulling the
trigger once and stepped back to its more accustomed role of monitoring
the operation and reporting on results. Even without an operational role

“‘The Balanced Budget and Emergency Deficit Control Act of 1985, Public Law 99-177, approved
Dec. 12, 1985. For a more detailed discussion of this episode, see this author’s “Gramm-Rudman-
IIollings: Origins and Implementation,” Public Budgeting and Finance, vol. 6, no. 3 (autumn 1986)

“The decision was made in a late night session of the conference committee, and there is no formal
record of the discussion. Most observers believe the principal consideration was a legacy of distrust of
the executive branch and especially nf what was seen as a politicized Office of Management and

“‘Bowsher v. Synar, 106 Sup. Ct. 3181(1986).

Page 12
                     The Evolution   of the General Accounting

                     under GRH, however, GAO remains an active participant in certain aspects
                     of the budget debate. In a series of reports prepared for the 1988-1989
                     presidential and congressional transition, for example, the Comptroller
                     General took a particularly strong position on the urgent need to deal
                     with the substance of the deficit,2:i accompanied by a renewed call to
                     reexamine the way the budget is presented.‘”

                     In 1989, these positions were further developed in a report on the impli-
                     cations of the accumulating Social Security Trust fund surpluses25 and in
                     another “white paper” on the need to overhaul the budget structure and
                     process.2”Involvement with the budget, as with the rest of the financial
                     management system, is assuredly part of GAO’S future.

Relations With the   GAO’S  recent history shows a clear pattern of continuing evolution in
                     response to the evident needs of the Congress. New responsibilities are
Congress             assigned, sometimes by statute, often through a pattern of committee
                     requests and occasionally on GAO’S own initiative. Once the responsibili-
                     ties are accepted and the capacity to meet them is built, they rarely dis-
                     appear. Rather, they become further accretions to GAO’S accumulating

                     As the mission of the agency evolved, it was inevitable-indeed,      it was
                     intended-that    GAO’S relationship with the Congress would also be
                     transformed. GAO sought to make its work more immediately relevant to
                     the legislative policy process and, as a result, the Congress and its com-
                     mittees became more directly interested in the scope, the nature, and the
                     timing of that work.

                     2”The Budget Deficit (GAO/OCG-8%lTR,        h3v. 1988).

                     %nancial    Management Issues (GAO/OCG-8%7TR, Nov. 1988).

                     25Social security: The Trust Fund Reserve Accumulation, the Economy, and the Federal Budget
                     (GAO/HRD-89-44, Jan. 1989).

                     ““Managing the Cost of Government: Proposals for Reforming Federal Budgeting Practices (GAO/
                     AFMD-90-1, Oct. 1989).

                     Page 13
                  The Evolution   of the General Accounting

                  From its beginnings under the 1921 act, GAO was required to perform
                  investigations and other studies requested by the Congress and its com-
                  mittees.“’ For the first half century, however, that mandate was of rela-
                  tively little consequence. IJntil the 197Os, virtually all of GAO’S work
                  responded to GAO’S own sense of need and priorities. As late as 1969, for
                  example, it is estimated that no more than 10 percent of GAO’S reports
                  were in response to congressional requests, and GAO officials (almost
                  exclusively the Comptroller General, himself) testified only 24 times.

                  By fiscal year 1977, congressional requests and statutorily mandated
                  audits and reviews were absorbing 35 percent of GAO staff resources and
                  GAO officials (including the division directors as well as the Comptroller
                  General and the Deputy Comptroller General) testified 111 times. The
                  trend, once established, was irreversible. In fiscal year 1988, fully 80 to
                  100 percent (depending on the subject area) of GAO’S resources were
                  involved in responding to specific congressional requests and GAO offi-
                  cials testified at committee hearings well over 200 times. To handle the
                  work load of testifying, GAO now relies heavily on the third major man-
                  agement echelon, the issue area directors (formerly titled associate

Staff Resources   The effects on GAO and its staff of the second great transformation have
                  been pervasive. These changes, however, while cumulatively dramatic,
                  did not usually seem so at the time. For the most part, new ideas, new
                  responsibilities, and new ways of doing business were introduced incre-
                  mentally. Because the mission changed through gradual accretion,
                  rather than sudden changes in direction, the shifts were much less
                  threatening to the staff. IJnlike the Warren-era changes, the Staats/
                  Bowsher transformation did not require the wholesale dismissal and
                  replacement of staff that was probably inescapable in the earlier period.

                  Nevertheless, while less traumatic for the organization than the virtual
                  revolution of the Warren/Campbell era, the changes introduced by
                  Staats and extended by Bowsher clearly have affected the staff in sev-
                  eral respects. The first evident change was to reverse the decline in total
                  numbers. From a low point of about 4,100 in the mid-1960s, GAO grew to
                  “iSection 3 10(b) of tht‘ 1921 art provides “[The Comptroller General] shall make such investigations
                  and reports as shall be ordered by tither House of the Congress, or by any committee of either House
                  having jurisdiction over revmuc. appropriations, or expenditures.” The language has been modified
                  slightly over the years, but t,hr basic point has not been changed. Interestingly, the act also requires
                  GAO to respond to requests from t,hc I’rcsidcnt. No one can reczallreceiving such a request in recent
                  times, but data for earlier periods is ambiguous. Some reports were submitted to the President, but
                  whcthcr or not he was the source of the Inquiries is unclear.

                  Page 14
The Evolution   of the General Accounting

a peak of about 5,300 in the late 1970s then settled back to the 5,000-
5,100 range, where it has been for the decade of the 1980s. But the com-
position of the staff has continued to evolve, as can be seen in the
changing nature and background of the people being recruited.

Among Staats’ first initiatives were efforts to expand the base of
recruiting and to introduce outsiders into the GAO leadership structure.
On recruiting, Staats first expanded the base to include business admin-
istration graduates without substantial training in accounting. He soon
broadened the effort to include such disciplines as public administra-
tion, operations research, engineering, statistics, and economics. Bow-
sher continued this trend and, by the mid-1980s to late 1980s there was
no longer any “standard” GAO recruit. Today, the largest single group
comes with a background in public administration or public policy and a
strong base of quantitative skills, but virtually every academic disci-
pline is now represented in the GAO staff. In addition, the staff includes
people with advanced academic training and line experience in virtually
every substantive field touched by government, from defense to health
care, energy, and tax law.

GAO still recruits accountants in significant numbers, but most of them
are hired specifically to do our accounting and financial auditing work.
Similarly, GA0 recruits economists to perform our economic analysis
work, computer scientists to review computer systems, and trained
criminal investigators to pursue allegations of corruption.

In practical terms, GAO'S recruiting philosophy has shifted dramatically
over the past 25 years. Once GAO recruited people for the professional
staff who came from a uniform background and then trained them to do
the variety of assignments that faced GAO. That strategy was doomed to
failure as the diversity of GAO'S work grew and the complexity of that
work increased exponentially. Today, GAO starts with a systematic
assessment of the sort of work it expects to be doing over the next dec-
ade or so and then recruits the people whose skills it thinks will best
match that work.

This strategy is also reflected in some relaxation of the “promote from
within” tradition established in the Campbell era. A large majority of
the GAO management structure is composed of people who have made
their careers in the organization, and that pattern, characteristic of most
large professional organizations, seems likely to continue. Significantly,
however, almost a quarter of GAO'S Senior Executive Service (SES) incum-
bents had substantial professional experience with other organizations

Page 15
                           The Evolution   of the General Accounting

                           before joining GAO. And that ratio is even higher at the top. Of the 13
                           mOSt senior positions in GAO, 2Hmore than half have prior experience in
                           other government agencies or in the private sector. And of these posi-
                           tions, only one-that of Comptroller General-is filled by what some
                           would consider a political appointment. Every other position in GAO is
                           filled through a merit-based competitive selection process, from new
                           entry-level recruits to the top of the SES2”

                           Staffing the modern GAO, however, has meant much more than changing
                           the mix of skills. In the mid-1960s the professional staff was uniformly
                           composed of white males. In common with most public agencies of the
                           day, GAO was slow to recognize that this was a serious problem and early
                           efforts to deal with it were not very successful. Society as a whole was
                           struggling with the question of how to close the gap between rhetoric
                           and reality, and GAO experienced that struggle in microcosm. As the
                           commitment of top leadership became increasingly evident, however,
                           GAO began to learn what works, Over time, the successes have been
                           incorporated into an aggressive equal opportunity and affirmative
                           action program affecting all aspects of human resource management,
                           from entry-level recruitment to promotion into the SES.

                           One of the important advantages of the 15-year term of the Comptroller
                           General is the ability to establish long-term agency policy directions and
                           to sustain those directions for an extended period of time. The effects
                           can be seen in this area, one of the most difficult internal management
                           problems facing any agency over the last three decades. The GAO work
                           force is not yet as balanced as it would like at the more senior levels, but
                           GAO’S goals are ambitious and it is moving firmly toward them.

Professional Development   The increasing diversity of GAO’S work and of the staff doing that work
and Training               has lent added importance in recent years to GAO’S training and staff
                           development functions. There are two main threads running through
                           these functions. One is the traditional training function found in most
                           large professional organizations, that of assuring that the staff acquires
                           the knowledge, the skills, and the abilities needed to perform the jobs to

                           “‘The Comptroller General. the Special Assistant to the Comptroller General, the General Counsel,
                           and 10 Assistant Comptrollers General.

                           “‘The position of Deputy Comptroller General is filled by presidential appointment, in a process simi-
                           lar to that for the Comptroller kneral. The Deputy poskion is currently vacant., but Mr. Milton Soco-
                           lar, a career civil servant and former General Counsel of GAO. has served as Special AssIstant to the
                           Comptroller General and dc facto Deputy Comptroller General since 1981.

                           Page 16
The Evolution   of the General Accounting

which they are to be assigned; the other function is aimed at accultur-
ating the individual to the agency’s core values and mission. The mul-
tidisciplinary nature of the work and the team-style operating
environment of the modern GAO puts a special burden on both elements
of training.

It is obviously impractical to try to give everyone the equivalent of a
terminal degree in every discipline that may be relevant to the work.
What GAO aspires to do, rather, is to give everyone sufficient exposure
to other disciplines to understand the contribution those disciplines can
make to the work and to ensure that the staffer can interact effectively
with experts in those disciplines. In addition, the continually changing
nature of GAO'S work means that GAO must provide supplemental train-
ing for experienced staff so they can direct and manage new kinds of
work, involving people with new kinds of expertise.

This, however, is only part of the training regimen. As GAO staff develop
along their career paths, they must also have or acquire sound commu-
nication skills, involving both written and oral presentation. These skills
are important for the production of reports and other written products
and are vital for briefing congressional staff and providing formal testi-
mony, an increasingly important mode of communication for GAO.
Broader management skills also take on greater importance, both in
terms of managing and motivating staff and in terms of directing
projects to meet tight congressional deadlines.

The other main thread of the training function centers on “socializa-
tion.” GAO is an institution, not just a collection of individuals. Its prod-
ucts carry the organization’s imprimatur and are vouchsafed by its
institutional stature. Yet, those products are developed by several thou-
sand individuals in a geographically and organizationally decentralized
environment,, often working as temporarily assembled audit teams.

For such an organization to succeed, the staff must share a common
body of institutional norms relating to such matters as standards of evi-
dence, operating procedures, and rules of ethical behavior. Thus, the
training and staff development function must inculcate in the staff the
GAO ethic of accuracy, independence, and objectivity; an understanding
of what those concepts mean in the GAO environment; and a practical
recognition of how to apply them in the day-to-day work of the office.

One measure of the importance GAO attaches to training and the contin-
ued development of its staff is the investment it makes in this area. In

Page 17
                      The Evolution   of the General Accounting

                      fiscal year 1988, for example, GAO invested almost $2 million in formal
                      training. The pace of this investment seems certain to rise. One stimulus
                      is the establishment of the GAO Training Institute, an organization dedi-
                      cated to meeting GAO'S training needs. The Institute manages the training
                      function and delivers most of the formal classroom training, The other
                      major impetus for increased training is the requirement that all profes-
                      sional staff obtain at least 80 hours of continuing professional education
                      in each 2-year period or risk losing the qualification needed to perform
                      the general work of the Office and to advance in their careers.

                      It is difficult to characterize in simple terms all the changes that have
                      constituted the second great transformation of GAO. Indeed, the changes
                      themselves are continuing today, as can be seen from the evolving
                      recruitment patterns and training strategies. Other changes are too
                      numerous and diverse to catalog and often too subtle to capture in a few
                      words. Their cumulative effect, however, has been to create an organiza-
                      tion in which diversity and the capacity for further change and develop-
                      ment are themselves the norms-diversity        and change in discipline,
                      subject matter, analytical approach, data-gathering technique, and
                      reporting method.

                      This diversity and capacity for adaptation are the keys to GAO'S ability
                      to respond to congressional requests for audits, evaluations, and other
                      studies over an enormous range of issues. But this diversity operates
                      within a set of institutional values emphasizing accuracy, independence,
                      and objectivity that have changed little, if at all, over the past two
                      decades and that provide a unifying force across the entire organization.

What Makes GAO Work   GAO is not the only organization looking at important questions of public
Special?              policy and management. Within the legislative branch, the Congres-
                      sional Research Service, the Congressional Budget Office, and the Office
                      of Technology Assessment also examine such issues. What is it that
                      makes GAO'S reports different, beyond the fact that they are published
                      in blue covers?

                      There are several special characteristics of GAO work and of the reports
                      it produces. One relates to GAO'S problem-solving approach to issues.
                      When GAO identifies a problem in the design or the implementation of a
                      program, there is a strong institutional bias in favor of finding and pro-
                      posing a solution to the problem. Thus, a significant part of most assign-
                      ments involves developing solutions to the problems that are revealed
                      and GAO reports routinely contain recommendations addressed to the

                      Page 18
The Evolution   of the General Accounting

Congress or to the administering agency. The recommendations range
from modest adjustments in operating procedures to major shifts in
national policies.

Moreover, GAO'S recommendations, and the analysis supporting them,
become public knowledge. All GAO reports, unless classified for national
security reasons, are available to the press and the public immediately
upon issuance or within a few weeks.

Other special characteristics of GAO center on its field structure and its
access to data. In addition to having its headquarters organization in
Washington, GAO has 14 regional offices in the continental United States
and 2 overseas offices. If GAO is asked to examine a matter, it is not
limited to using information filtered by others. When necessary, GAO can
place its staff on location, permitting direct observation of events and
collection of any raw data that may be required to answer the question
posed by the Congress.

To carry out its fieldwork, GAO has a statutory right of access to virtu-
ally any records or other information in the possession of government
agencies, In addition, GAO'S reputation for objectivity and for the care
with which it handles data means that the staff can often gain volun-
tary access to data outside government that may be unavailable to other

The access to information and the field staff that can examine matters
at the site give GAO reports a special quality. For example, when ques-
tions arose about the condition of the nation’s nuclear weapons complex
or about progress in cleaning up toxic waste dumps, GAO could back up
its analysis of the problems with dramatic pictures and other hard evi-
dence, coupled with testimony based on fir&and observation.

The emphasis on “hard” evidence is characteristic of GAO. Very little of
GAO'S work involves speculation about the unknown, and GAO rarely
relies on expert opinion as the principal support for conclusions and rec-
ommendations, The usual GAO report relies heaviIy on data that GAO has
either gathered itself or tested to assure its reliability.

This approach reflects the fact that GAO'S culture is grounded in the
auditing profession. One of the central themes of that profession-
indeed, its very reason for existence -is that the reliability of data must
be assured before it is used for decision-making. Thus, GAO is quite com-
fortable using existing data in its work so long as the quality of that

Page 19
The Evolution   of the General Accounting

data can be established with reasonable confidence. In its program
audits and evaluations, for example, GAO is a frequent user of the large-
scale data bases maintained by the departments and agencies and, when
needed, those maintained outside government. When necessary, how-
ever, GAO will invest substantial time and effort in testing the data
before using it or in collecting original data when the existing data is

In one recent case, GAO was asked to assess the effect on air fares of
increasing concentration in the airline industry. In particular, GAO was
asked to examine the consequences of one or two airlines’ achieving
dominance in serving a city. GAO compared the fare per passenger mile
for those enplaning in 15 airports dominated by 1 or 2 airlines with 38
airports that were not so dominated.

To make the comparison, GAO used fare data collected routinely by the
Department of Transportation. In checking the reliability of that data,
however, GAO discovered not only the usual random data entry errors
that are found in almost any large data base but. also some outdated edit
controls that were systematically screening out valid fares. Once these
problems were corrected, GAO was able to proceed with the analysis.“”
Predictably, yields were strongly and positively related to concentra-
tion. Even more revealing is the fact that yields typically rose signifi-
cantly after an airline established a dominant position.

Unlike some, GAO did not view these results as warranting a call for
reregulation of fares. Rather, GAO recommended (and performed) fur-
ther work, centering on the barriers to entry (e.g., restricted access to
limited gate capacity) and other factors that may be impeding

In many cases, however, existing data is inadequate to answer the ques-
tions posed by the Congress. In 1981, for example, the Congress ena.cted
significant restrictions on eligibility and benefits in the program of Aid
to Families with Dependent Children (AFDC). It then asked GAO to assess
the impact of these amendments in terms of (1) AFDC caseload and cost,

‘j”Air Fares and Service and Concentrated Airports,statement by Kenneth M. Mead, GAO, before the
Subcommittee on Aviation, Senate Committee on Commerce, Science, and Transportation (GAO/T-
RCED-89-37. ,June7, 1989).

“‘Barriers to Competition in the Airline Industry, statement by Kenneth M. Mead, GAO, before the
Subcommittee on Aviation, Senate Committee on Commerce, Science, and Transportation (GAO/T-
RCED-89-65, Sept. 20, 1989).

Page 20
    The Evolution     of the Geneml   Accounting

    (2) work effort of affected families, and (3) economic well-being of
    affected families.

    Initially, GAO examined the available national and state-level data,
    including data bases maintained by academic and other nongovernment
    researchers, in hopes that these would permit GAO to answer the ques-
    tions relatively inexpensively. After careful consideration, GAO con-
    cluded that while the existing data provided useful insights, they were
    inadequate to establish the cause-and-effect relationships being sought.
    GAO also concluded t,hat it would be worth the time and effort to collect
    original data for that purpose. Accordingly, GAO decided to supplement
    the available data with the results of detailed evaluations at each of five
    sites, chosen to reflect some of the great diversity in program character-
    istics, economic context, etc. At these sites, G,40 collected data from
    thousands of case records and hundreds of interviews with people who
    had lost AFDC. benefits because of the 198 1 amendments.

    The results of the analysis were communicated in two reports? In brief,
    the major conclusions were as follows:

. There was a significant, nontransient decline in AFDC caseload and cost.
. There was no indication that people quit working to regain lost AFDC
l Affected families suffered a significant loss of income.
. Some families suffered special hardship from losing Medicaid coverage.

    In another recent case, the Congress asked GAO to assess the effects of
    the economic sanctions imposed on South Africa. Of particular interest
    was the effect of prohibitions on importing certain products originating
    in South Africa.

    South Africa no longer supplies trade data in sufficient detail to be ana-
    lytically useful, so GAO worked with a IJnited Iiations (UK) data base on
    world trading patterns. This contained data supplied by all of South
    Africa’s principal trading partners but had never before been used for
    this purpose.

    In assessing the reliability of the IJN data, GAO was especially concerned
    about reports that traders were employing various methods (transship-
    ment through third countries, etc.) to disguise the source of products

    ‘“An Evaluation  of the 1981AFDC Changes: Initial Analysis (CLkO/PEMD-84-6,Apr. 1984)and An
    Evaluation of the 1981 AFDC Changes: Final Report (GAO/l%3lD-85-4. .Julg 1985).

    Page 2 1
                         The Evolution   of the General Accounting

                         originating in South Africa. GAO used several techniques to establish that
                         this problem, while real, was not of sufficient magnitude to threaten the
                         credibility of the analysis. GAO then concluded that the sanctions had a
                         clearly detectable effect, reducing by $469 million South Africa’s
                         exports of sanctioned commodities to its principal trading partners.:‘:’

                         These three examples illustrate GAO’S approach to data when dealing
                         with relatively complex analytical issues. GAO uses existing data, if its
                         reliability can be assured. But that reliability must be tested; it cannot
                         be assumed. If existing data cannot meet the need and GAO judges the
                         issue worthy of the investment, it collects the original data needed to
                         answer the question.

Assuring Institutional   GAO’S reliance on hard data and its insistence on establishing the reliabil-
Credibility              ity of data before using it is a visible indicator- of the institution’s con-
                         cern for its credibility. That credibility is built on GAO’S cultural values
                         of independence, objectivity, and accuracy, and it is on the line each
                         time a report is issued or testimony is presented. But the close working
                         relationship that has emerged between GAO and the congressional com-
                         mittees during the decades of the 1970s and 1980s could pose significant
                         risks to credibility.

                         Independence and objectivity could be severely threatened by a commit-
                         tee that is convinced of the rightness of a particular policy position and
                         wants a GAO report, not to illuminate the issue, but to support that posi-
                         tion. Accuracy and precision could be threatened by the temptation to
                         take short cuts in order to satisfy a committee’s demand that GAO meet
                         an unrealistic deadline. Recognizing these risks, GAO goes to great
                         lengths to protect against them. Elaborate (and sometimes very time-
                         consuming) safeguards are built into every stage of the performance of
                         an assignment.

                         These safeguards start with the GAO strategic planning process, in which
                         GAC) defines the issues it judges to be of central importance, identifies the
                         questions it believes should be pursued, and allocates the available staff
                         resources among the ma,jor issues. This planning is not done in an ivory
                         tower. The judgments about priorities within each major issue area are
                         formed from extensive discussions both with committee staff concerning
                         the anticipated legislative calendar and with subject matter experts and

                         ‘%mth Africa: Summary Reporton Trade. Lcnding,investment.and StrategicMinerals (GAO/
                         NSIAD-88-228,Sept. 1988).

                         Page 22
The Evolution   of the General Accounting

agency officials [often in roundtable or panel format) about emerging
trends, problems, and policy initiatives.

The dialogue, particularly with the committee staffs, is mutually benefi-
cial. GAO learns what concerns are uppermost in the minds of committee
members and staff. At the same time, GAO observations about emerging
problems often lead a committee to adjust its agenda to take account of
an issue of which the committee was not previously aware.:j” This may
result in the committee’s requesting an assignment that was already
high on GAO’S priority list-one GAO would have wanted to do, even in
the absence of a request.

Not all requests, of course, fit so neatly into GAO’s internal sense of rela-
tive priorities. And priorities change as external events occur and new
problems emerge. In any year, GAO initiates well over a thousand new
assignments. The first step, in the case of a congressional request, is for
staff to meet with the committee’s or member’s staff to define the
request. Experienced congressional staff understand this process and
usually call an informal meeting before sending the official request.

In these discussions, GAOdepends on its staff with experience in the sub-
ject matter to negotiate agreement on a study that can be performed in a
balanced fashion (no hatchet jobs, please), is technically feasible (please
do not ask GAO to count the number of illegal drug abusers), and involves
 a reasonable claim on GAO'S very limited staff resources (GAO would
rather not duplicate the efforts of the Inspectors General or DCAA,)IJsu-
ally these discussions at the staff level are quite successful. In the event
of problems, however, GAO is never reluctant to elevate the issue to a
discussion between senior officials of GAO and the Congressman or the
Senator initiating the request.

Once the request has been negotiated, the assignment goes through GAO'S
internal review processes before any substantial work can begin.‘L5Each
assignment (unless terminated at an early stage) is reviewed at least

“‘Needless to say. these efforts do not always have the expected effect. For a number of years, GAO
and the Inspector General of the Department of IIousing and ITrban Development had been warning
about weak mana#ment and Iah. controls in IKTD. For various reasnns, the committees chose not to
follow up un these matters lmtil the Hl!D scandal broke in the spring of 1989.

“51n some circumstances, GAO assigns staff to work directly for a cnmmittec. In those casts, the GAO
staff become staff of the commit,tct, which decides the scope and thr direction of the work and
assumes responsibility for any product that may result, whirh wtmld be identified as a committee
product. nt)t a GAO rqu-t..

Page 23
The Evolution   of the General Accounting

twice by the leadership of the division in which the work will be per-
formed and (separately) by the Comptroller General and his senior man-
agement team.

These reviews center on certain key considerations. Is the assignment
appropriate for GAO? Questions about estimated budgetary costs, for
example, would be more properly directed to the Congressional Budget
Office. Is someone else already working on the issue? GAO staff are
responsible for ensuring the absence of unnecessary duplication, not
only between different parts of GAO but also between GAO and the other
congressional support agencies-the Congressional Research Service,
the Congressional Budget Office, and the Office of Technology Assess-
ment-and the Inspectors General and internal auditors in the executive
branch, as well.

Once the general appropriateness of the assignment has been estab-
lished, GAO turns to different sorts of questions. Is the issue worth the
resources needed to explore it‘? Is the right question being asked? Does
GAO have an efficient and a reliable way of developing the data needed
to answer the question? Is the proposed analytical approach likely to
yield a valid result‘? Are the right people working on the assignment?
Will the results be available in time to meet the anticipated need? These
questions are inherently subjective, but the process of asking them and
debating the answers is an essential safeguard against the waste of
scarce resources and, perhaps more importantly, against the risk of
starting a study that is biased or technically flawed from the outset.

The safeguards continue throughout the assignment. As the fieldwork
and the analysis are being performed, there are routine supervisory
reviews of the progress of the job involving senior staff from the divi-
sion responsible for the work and, if applicable, from the regional office
where the work is being performed. Usually, these reviews center on
assessing progress against the assignment plan and on solving the inevi-
table problems that arise in any study (e,g., the absence or the demon-
strable unreliability of key elements of data). In addition, on
congressional request assignments, there is frequent contact with the
requesters to ensure both that they are informed of progress and prob-
lems and that the requested work is still relevant to congressional needs.

As an assignment nears completion, the safeguards become intensified,
focusing on what will be said in the written product that is expected to
emerge (formal report, statement for hearings, etc.). The conceptual
framework for safeguards at this stage is multiple independent reviews

Page 24
The Evolution   of the General Accounting

before the product is permitted to be published. These reviews take sev-
eral forms, but the intended cumulative effect is (1) to provide the func-
tional equivalent of the professional peer review to which academic
research is commonly subjected and (2) to apply sufficient tests to the
content of the report t,o ensure that it meets GAO quality standards and
is worthy of receiving the institutional imprimatur.

First, each draft report is subject to supervisory and technical review at
several levels within GAO. The more important or complex the issue, the
more reviews to which the report will be subjected. The initial draft of
the report is prepared by (or under the supervision of) an evaluator-in-
charge (EIC), the functional title given to the day-to-day leader of the
project. If the EIC is located in a regional or an overseas office, the draft
is reviewed within that office. The draft is also reviewed in the head-
quarters division controlling the assignment. This review involves an
assistant director, the responsible issue area director, the division’s
director of planning and reporting, the Assistant Comptroller General
heading that division, and the division’s quality assurance team, as well
as its editorial staff.

Second, each report is “referenced.” To satisfy this requirement, a fully
qualified staffer must check every number and every statement of fact
in the report, identifying the source in the fieldwork or the analysis (the
“workpapers” in GAO parlance) and confirming that the report accu-
rately reflects the information in the workpapers. Any discrepancies
must be corrected before the report can be published.

Third, unless explicitly prohibited by the requester, written comments
on GAO'S findings, conclusions, and recommendations are routinely
requested by GAO from the agency on whose activities it is reporting.
When these official comments are received, they are published as part
of the report. (Even when a requester has directed GAO not t,o share the
draft report with the agency, GAO staff meet with agency officials in an
“exit conference” to ensure the accuracy of factual material.)
If an agency’s comments indicate disagreement with the contents of the
draft, the GAO staff must analyze those comments and respond to them
before the report can be published. This response is also published as
part of the report. If, as sometimes happens, the agency provides addi-
tional information that was previously not available to GAO, or points
out flaws in the analysis or logic supporting the report, GAO reexamines
its conclusions and recommendations. If GAO decides that the agency

Page 25
The Evolution   of the General Accounting

arguments are valid, it makes an appropriate change before publishing
the report.

A fourth type of independent review is used on a selective basis, with
respect to reports that involve complex forms of analysis or deal with
sensitive policy issues. For example, any report dealing with questions
of economics or relying on economic analysis to support the conclusions
must be reviewed by GAO'S Chief Economist, who is independent of the
operating divisions and reports directly to the Comptroller General. In
addition, the Comptroller General and members of his immediate staff
routinely review a substantial portion of the reports being published
and always review any report in which the issues arc tricky or politi-
cally sensitive. Finally, GAO often seeks review of a draft by outsiders
who are expert in the subject matter of the report or in the analytical
techniques being employed, to ensure that GAO has not overlooked an
important point.

To ensure that safeguards on quality remain strong, GAO has instituted a
mechanism for systematically reviewing its procedures for conducting
studies. The Post Assignment Quality Review System (PAQRS) involves
an independent staff that, each year, selects a sample of the reports that
were issued and examines in detail how the assignments were planned
and carried out. For each step of the assignment, actions by the GAO
staff are compared with the required procedures. Any deviations are
analyzed to determine why they occurred and what effect they may
have had. The results of the PAQRSreviews are reported to GAO senior
management both to reinforce the importance of the quality control pro-
cedures and to ident.ify procedures that may need modification.

The laborious, tedious, and often time-consuming lengths to which GAO
goes to ensure the quality of its products are a frequent source of frus-
tration both to the GAO staff and to congressional clients. They are one
of the reasons-along with the complexity of the work itself-that       the
average GAO assignment takes 9 to 12 months to complete. (Some espe-
cially difficult evaluations with extensive original data gathering may
take as much as 2 years or more.) But GAO'S investment in these safe-
guards is only a measure of the importance it attaches to accuracy,
integrity, and objectivity.

When GAO issues a report or when GAO officials present testimony before
a congressional committee, the Comptroller General and GAO as an insti-
tution stand behind that product. Conclusions and recommendations are

Page 26
             The Evolution   of the General Accountjng

             supported by GAO'S institutional stature, not just the professional repu-
             tation of the individuals who produced them. Indeed, it is a relatively
             recent innovation for GAO even to identify in a report the people who
             worked on it. This is now being done as a means of providing individual
             recognition, but there has been no diminution of institutional responsi-
             bility and accountability.

             In summary, today’s GAO is an organization of about 5,000 people who
Conclusion   take great pride in carrying out a modern version of the original man-
             date from the 1921 act: “ . ..to investigate, at the seat of government or
             elsewhere, all matters relating to the receipt, disbursement, and applica-
             tion of government funds....”

             As the world has changed around us, the work needed to satisfy that
             mandate has also changed, as has the institution of GAO. The work and
             the staff performing it have become more technically sophisticated,
             more policy oriented, and more closely attuned to the needs of the Con-
             gress. In the process, change itself has become a constant. But the pur-
             pose of the mandate, GAO'S central mission, has not changed. To GAO, it
             means always finding a way to learn the facts, no matter the issue about
             which the Congress has inquired, and putting those facts together in a
             story that is accurate and balanced, with recommendations that are
             appropriate for solving any identified problems, no matter whose ox
             may be gored.

             Fulfilling that mandate is a challenge that can take GAO people any-
             where in the world, from examining files and records in an agency’s
             Washington headquarters; to observing the tests of a new weapon sys-
             tem and analyzing the results; to following the cash of a foreign assis-
             tance grant in t,he third world; to assessing the role of computer failures
             in the 1987 stock market crash; and to evaluating the effects of new,
             federally mandated procedures in a local welfare office. Wherever GAO
             must go to learn the facts, there you will find the GAO staff.

             Page 27