oversight

Problems in Developing the Corps of Engineers' Automated Management Information System

Published by the Government Accountability Office on 1971-04-21.

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

e· I~+i-··~i-4




        .4UNI
         D           STA,                   TS




                  Problems In Developing
                  The Corps Of Engineers'
                  Automated Management
                  Information System
                 ACCOUNTING
                 GENERAL
                                                          OFFICE.


                  Department
                  Department of the Army          '   °




                 UNITED STATES
                 GENERAL ACCOUiNTING OFFICE


                                                  APRIL21. 1971
                              lit7~-~ii~F~i~j~-
                                    . _ .
                           UNITED STATES GENERAL ACCOUNTING OFFICE
                                    WASHINGTON, D.C. 20548

DINBASE:     DIVISION


            B-163074

            Dear Mr. Secretary:

                  This is our report on problems in developing the Corps of Engi-
            neers' automated management information system.

                  We believe that our review revealed the need for emphasis on
           factors governing the responsiveness of a proposed syrsetem to the needs
           of the organization for a reasonably long period of time. Such emphasis
           could be achieved by requiring that this matter be included in the formal
           studies prescribed for completion before substantial investments in
           equipment for the system .re authorized.

                  We recommend, therefore, that the existing directives be revised
           to include guidance to the proponents of major new automatic data proc-
           essing systems concerning the need for documented studies of the fac-
           tors affecting the long-range usefulness of the system.

                  We are encouraged by the attention being given by the Department
           to resolving the problems involved in giving appropriate consideration
           to the residual values of dissimilar equipment proposed for automatic
           data processing systems. We recommend that the Department, in con-
           nection with its study of the use of residual values--or a comparable
           alternative--obtain and consider the views of the computer industry
           with the objective of developing guidance in this complex area.

              Your attention is invited to section 236 of the Legislative Reorga-
       nization Act of lc?70 which requires that you submit written statements
       of the action taken with respect to the above recommendations. The
       statements are to be sent to the House and Senate Committees on Gov-
       ernment Operations not later than 60 days after the date of this report
       and to the House and Senate Committees on Appropriations in connec-
       tion with the first request for appropriations submitted by your agency
       more than 60 days after the date of this report.




                               50TH ANNIVERSARY 1921- 1971
  B-163074



        Copies of this report
 Management and Budget;       are being sent to the Director,
                           the Secretary of the Army;         Office of
 trator, General Services                               and the Adminis.
                           Administration.

                                        Sincerely yours,




                                       Director, Defense Division
The Honorable
The Secretary of Defense




                                 2
                        C on t e nt s
                                                          Page
DIGEST                                                         1
CHAPTER

     1     DELAY IN DEVELOPING PLANNED SYSTEM              4
               Functions and organization                  4
               Nature of system development problems       6
               Long-range plan for systems development     8
               Comparison of actual progress with fore-
                 cast                                      9
  2       SYSTEM DEVELOPMENT STARTED BEFORE RESEARCH
          COMPLETED                                       12
              Government policy requires studies          12
              Development started without adequate re-
                search                                    14
              Research disclosed numerous problems        16
              Conclusions                                 19
              Agency comments and action taken            20
              Recommendation                              21
  3       FACTORS NOT CONSIDERED IN ACQUIRING ADP
          EQUIPMENT                                       22
              Status of equipment contract                23
              Equipment contracted for before systems
                were developed                            25
                  Agency comments and our evaluation      25
              Prescribed formula not used in equip-
                ment selection                            28
                  Agency comments and action taken        29
              Residual value of ADP equipment not rec-
                ognized                                   30
                  Agency comments and our evaluation      32
                  Recommendation                          33
 4        SCOPE OF REVIEW
                                                          35
APPENDIX

       I   Letter dated June 10, 1970, from the Deputy
             Assistant Secretary of Defense (Comp-
             troller) to the General Accounting Office

      II   Letter dated May 28, 1970, from the Acting
             Director, Bureau of the Budget, to the
             General Accounting Office

  III      Letter dated June 9, 1970, from the Admin-
             istrator, General Services Administration,
             to the General Accounting Office

                         ABBREVIATIONS

ADP        automatic data processing

DOD        Department of Defense

EIDS       Engineer Information and Data System

GAO        General Accounting Office

P-BAMS     Programming, budgeting, accounting, and
             management system
GENERAL ACCOUNTING rFFICET                    PROBLEMS IN DEVELOPING THE CORPS
REPORfT TO I'HE SECRE'.AP.
                      Y OF DEI.'iSE           OF ENGINEERS' AUTOMATEC
                                              MANAGEMENT INFORMATION SYSTEM
                                              Department of the Army B-163074

DIGEST

WHY THE REVITW ,eA
                ·  ;D
      Prior work by the General Accounting Office (GAO) showed that the
      Corps of Engineers was planning to create a Corps-wide computerized
      data system covering technical and business applications. GAO wanted
      to know whether the methods used in developing the system and the
      guidance furnished by the Army and the Office of the Secretary of De-
      fense would result in a system that would be initially responsive to
      the needs of the organization and that would remain so for the fore-
      seeable future.

FINDINGS AN'D COX':CLUS~iO:S

      Before 1965, Corps divisions and districts could individually develop
      their own automatic data processing (ADP) systems and applications.
      Those individual efforts were then supplanted by a long-range plan
      covering the development of a Corps-wide system during the period No-
      vember 1965 through June 1970. During GAO's review the Corps esti-
      mated that work under the long-range plan would be substantially com-
      pleted by June 1974, about 4 years later than originally estimated.
      (See p. 10.)
     Tie Corps' efforts were started before it completed research into the
     factors affecting the long-range usefulness of the system. The Corps
     did not start the research studies until 1968, about 2 years after
     had started systenm development. The studies disclosed problems     it
                                                                      in
     organization, operatio.g procedures, and policies that could affect
     system needs. Directives by the Office of Management and Budget and
     by the Arny require evaluation of such factors; however, Department
     of Defense guidelines are silent on the matter.
     The Corps decided not to continue the systematic research envisioned
     by the studies. GAO found that the data needed to make informed Judg-
     ments about the specific problems or changes needed to overcome them
     was not obtained on a systematic basis. At the time of the GAO re-
     view, the system impact of the identified problems had not been is-
     sessed.
     GAO believed that the project needed to be reevaluated to reduce the
     risk of implementing a system that would have to be continually mod"-
     fled. (See pp. 12 to 22.)
     The Army purchased ADP equipment and services valued at about $3 mil-
     lion and in a later procurement obtained an option to buy additional,
     similar equipment costing about $7.6 million. The mathematical for-
     mula used by the Army in the later procurement to make a comparative
     rating of vendors' equipment had been superseded about 3 months earlier
     by a revised formula in the agency's standard operating procedures.
     Use of the new formula might have resulted in another vendor's receiv-
     ing a higher rating than that of the one selected. (See p. 28.)
     Further, GAO noted that existing Government regulations did not re-
     quire Government agencies. in determining which vendor had offered the
     best buy, to give weight to residual value--a quantitative measure of
     the continued usefulness of equipment for other applications when it
     is no longer needed for its original purpose. In this case some of
     the equipment offered by the low bidder had been previously used, had
     been out of production for several years, and currently is considered
     obsolete by some consultants. The losing bidders proposed to furnish
     new equipment. (See pp. 30 to 33.)

rEC2.9..,     7T-S   OR SUGGESTIONS

     Considering the lack of a demonstration that the system would serve
     its purpose adequately for an extended period and considering the
     failure to follow certain prescribed policies in acquiring equipment
     for a prototype installation, GAO suggested that the Department of
     Defense reappraise the project before further equipment was pur-
     chased.

AG;;Ž'Y ACTIO:.'S A.1 UI'.RESOLVED ISSUES

      In response the Department of Defense advised GAO that:
        --Its review of the long-range plan of the Corps of Engineers had
          confirmed the need for documented studies.
        --The Assistant Secretary of the Army (Financial Management) had
          been advised that the studies would have to be completed before
          extension of the prototype system.
        --A task group had been established to develop and recommend an eco-
          nomically sound and relatively simple technique for evaluating dif-
          ferences in residual values when dissimilar ADP equipment is of-
          fered.
        --The Army had amended its procedures to require that in the future
          all deviations from prescribed procedures would be documented, re-
          viewed by appropriate offices, and entered into the official rec-
            ord. (See app. I, p. 39.)



                                        2
The General Services Administration (GSA) which has certain responsibil-
ities for the acquisition and management of computers in the Federal
Government advised GAO that it agreed that long-range system studies,
rather than short-range expediency, should form the basis for the acqui-
sition of ADP equipment. (See app. III, p. 44.)
The Office of Management and Budget which has responsibility for estab-
lishing Government-wide ADP oolicy stated that residual values should
be a factor in future management decisions relating to the acquisition
of ADP equipment and offered a number of reasons why current regula-
tions did not require consideration of residual values in the procure-
ment of dissimilar equipment. The Office further stated that it was
hopeful that the special studies being made on residual values would
prove to be useful. (See pp. 41 to 43.)
GAO is recommending that the Secretary of Defense
  --revise existing directives to include guidance to the proponents
    of major new ADP systems concerning the need for documented studies
    of the factors affecting the long-range usefulness of the system
   and
 --obtain and consider the views of the computer industry and coor-
   dinate the Department's efforts with the Office of Management and
   Budget before concluding the study of residual values or a compar-
   .ble alternative.




                             3
                              CHAPTER 1

                 DELAY IN DEVELOPING PLANNED
                                               SYSTEM
         The Committee on Appropriations,
                                             House of Representa-
   tives, in its report on
                              the Department of Defense
   propriation for 1969, expressed                        (DOD) ap-
   to spend substantial amounts        concern over the DOD plans
 v systems.
                                   on new  computerized management
              As part of
   ment of such systemsbyits continuing interest in the develop-
                            Federal agencies, the General
   counting Office (GAO)                                    Ac-
                           initiated a review of the
  of the Army Corps of Engineers                       activities
                                     in developing a Corps-wide
  integrated automatic data
                               processing (ADP) system.
  our review we were concerned                             During
  ment policies and procedures primarily with basic manage-
  to the development of           which are generally applicable
                          major ADP systems. The
  review is detailed on                             scope of our
                          page 35.
 FUNCTIONS AND ORGANIZATION

       The Corps, in discharging
  ties, spends about $2.3          its extensive responsibili-
                           billion annually, uses
 tary personnel, and employs                         1,500 mili-
                               46,500 civilians. Its
 are obtained from 22 different                          funds
 specific requirements            sources,   each
                        for the manner in which of which has
 should account for and                            the Corps
                         report on its expenditures.
 fice of the Chief of Engineers                           The Of-
direction of the Corps'           is  responsible  for overall
                          activities and is assisted
commissions, boards, advisory                            by 18
trative offices. The            groups,   and  various adminis-
                       Corps is organized into
torates, 13 divisions,                             four direc-
                        and 10 operating activities.
the Corps operates under                                   Since
                           decentral:zed management
the divisions and operating                            concepts,
                              activities have considerable
latitude in making decisions
                               in their area of responsibility.
     The four directorates
ning, and technical liaison are responsible for policy, plan-
                             activities, as follows:

    Civil Works Directorate--Carries
    authorized by the Congress       out projects directly
                               for rivers, harbors, canals,
    and waterways; coordinates
                               activities witvh State and
    local Governments.




                              4
      Military Construction Directorate-- Carries out projects
      for DOD requiring services in the areas of designing,
      contracting, supervising, and inspecting construction.
      These activities are conducted with funds provided by
      the appropriations of sponsoring agencies.

      Real Estate Directorate--Acquires, manages, and dis-
      poses of all Army real estate and provides similar
      services to the Air Force and other agencies.

      Military Engineering Directorate--Performs training,
      mobilization logistics, and operational planning and
      development for engineer elements of Army forces.

The Civil Works Directorate is responsible to the Congress
through the Secretary of the Army. The other directorates
are responsible to the Secretary of Defense through the
various Army command levels.

     The actual work of the Corps is performed by the 13
divisions and the 10 operating activities.  The efforts of
each of the 10 operating activities are basically oriented
toward a specific function, such as topographic services
and ballistic missile site construction; but within these
functional areas the activities have semiautonomous respon-
sibilities. The 13 divisions, including three which are out-
side the continental United States, are semiautonomous
supervisory offices that have broad responsibilities within
a specified geographical area. They direct the work of 40
subordinate district offices which are primarily respon-
sible for the design and construction of civil and military
facilities. 1

     Specifically, the districts prepare engineering studies;
develop the design for facilities; construct military-civil
works and other facilities; operate and maintain flood con-
trol, river and harbor facilities, and related installations;


1 Military
          construction is assigned to most, but not all,
 continental United States engineer divisions and dis-
 tricts.




                             S
  acquire, manage, and dispose of real estate; award
  minister contracts for construction activities;    and ad-
                                                  and perform
  other functions as assigned.

       The organizational pattern of the divisions or
                                                       dis-
  tricts provides groups of technical personnel
                                                 for civil
  works, military construction, real estate, or
                                                 military en-
  gineering, as appropriate, and an advisory and
                                                  administra-
  tive staff consisting of such support service
                                                 groups as
  legal counsel, personnel, data processing, accounting,
                                                          and
  others.

 NATURE OF SYSTEM DEVELOPMENT PROBLEMS

       Prior lo 1965 the divisions and districts developed
  their own ADP systems and acquired equipment when
                                                     they
  could attribute significant monetary savings to
                                                  its use.
 The equipment was used primarily by engineering
                                                  personnel
 to solve technical and scientific problems. The
                                                   use of ADP
 equipment became widespread in the engineering
                                                 area, but at
 that time the Corps did not centrally develop computer
 plications for Corps-wide use because of the wide        ap-
                                                    variation
 in equipment among the various districts and of
                                                  their dif-
 fering missions, sizes, workloads, and types of
                                                  projects
 being designed and constructed.

     Over the years the use of ADP for nontechnical
                                                     appli-
cations increased materially. The principal nontechnical
ADP applications are found in the areas of civil
                                                  works,
military construction, real estate, personnel,
                                                civilian
payroll, procurement, finance, cost accounting,
                                                 property
management, and planning and budgeting. The Corps
that technical applications now account for about estimates
                                                   half of
its total ADP requirements.

      By 1965 the divisions and districts had dissimilar
                                                          ex-
 periences in designing and installing ADP equipment
 tems. Some installations were preparing tc install  and  sys-
                                                     very
 sophisticated equipment and systems; some were
                                                oriented to-
ward solving technical problems and had not acquired
                                                      much
experience in nonengineering applications; and
                                                some had not
acquired any ADP equipment. All the early ADP
                                                applications
were developed independently to meet the requirements
the individual divisions and districts and were        of
                                                 not

                              6
standardized either Corps-wide or divisionwide. Conse-
quently, a conglomerate of methods or systems was created
or adopted, which varied in detail, structure, scope, de-
gree of computerization, and methods used to accumulate and
record data.

     Studies completed in 1965 indicated that the districts,
rather than the Office of the Chief of Engineers, had taken
the lead in planning, acquiring, and implementing systems
because the Corps did not have a long-range plan for sys-
tems development. The independent approach to systems de-
velopment fostered a number of problems, such as:

     -- Computers were used by divisions for individual
       business reports, and integrated applications had
       not been developed.

     --Project reports were furnished to all organizational
       levels and generally provided each level with the
       same amount of detail. The reports were nor de-
       signed for use on a management-by-exception basis.

     Further, the studies indicated that the design of the
systems was such that technical personnel at the district
level could not obtain adequate and timely data and there-
fore were maintaining their own cost records independent
of the formally authorized accounting records.

     The Corps' problems with its cost and financial systems
were described by one of its officials,as follows:

     "These cost and finance accounting systems gener-
     ate data for numerous recurring and one-time re-
     ports. The total volume of data contained in
     those reports each year is huge. In FY 64 a single
     Engineer District *** produced 3,774 pages of re-
     quired reports, and OCE LOffice of the Chief of Eh-
     gineers3 compiled from field feeder reports some
     7,264 pages of summary reports. Despite this
     large volume of data, much of which has not been
     useful to management, there 4s indication that
     management is not being furnished with t'e infor-
     mation it does need to insure proper direction and
     supervision of the several programs assigned.

                              7
       Therefore, there is a need at this time of utmost
       importance to revise these systems so that they
       will produce the data management needs and to elim-
       inate, insofar as it is within the authority of the
       Chief of Engineers to do so, all data that is not
       now, or likely to be in the future, of value to
       Management levels in the Corps."

     To solve these ADP problems, the Office of the Chief of
Engineers deci'    in late 1965 to develop an Engineer In-
formation and Data System (EIDS). This system was to en-
compass the data requirements of the Office of the Chief of
Engineers and of the divisions and districts for all techni-
cal, business, and management applications. This workload
was to be processed by nine regional data processing cen-
ters,l each consisting of a large central processing unit
located in the divir.on office and of smaller.computers
located in each of the district offices. The regional cen-
ters were to be interlin'ked to a central hub in the Wash-
ington, D.C., area which also would serve the needs of the
Office of the Chief of Engineers.

LOFG-RANIGE PLAN FOR SYSTEMS DEVELOPMENT

     During November 1965 the Office of the Chief of Engi-
neers prepared a long-range plan to serve as the basic
guide through June 1S70 for all actions necessary to de-
velop EIDS. The long-range plan was conceptual in nature
and did not include dates for all the detailed actions nec-
essary for EIDS development. The long-range plan consisted
of five parts which are summarized below.

       Part I--Crcate at each district the capability to de-
       velop computerized systems to replace aid streamline
       existing manual and semiautomated systems and to op-
       erate in-house computers where economically justified.


1Theregional centers will serve 10 divisions located
within the continental United States. One division, how-
ever, does not have any districts; therefore it will be
served by a central processor located in an adjacent di-
vision.
    Part II--Select a-pilot division where a standardized
    system for programming, budgeting, accounting, and
    management (P-BAMS) will be designed, programmed on a
    regionally centralized computer, and debugged by
    June 30, 1967. This will then be standardized Corps-
    wide and installed in all other divisions on an in-
    cremental basis. Concurrently with the design and in-
    stallation of P-BAMS, determine requirements for a
    personnel management data system and for the Engineer
    Command and Control System and develop a centrally de--
                      system
    signed standardized        for eech and implement them
    at the division.

    Part III--Standardize technical engineering applica-
    tions and program these applications into a centralized
    computer with on-line communications from each dis-
    trict. Standardize in a pilot division and upon com-
    pletion initiate in other divisions insofar as practi-
    cable. Develop on an individual-division basis addi-
    tional technical engineering and operations systems
    peculiar to the specialized workload of the divisions
    concerned.

    Part IV--Purchase computer components needed by each
    division. Utilize the pilot division's experience
    with the fully standardized system specifications
    which cal be made applicable Corps-widle. Upon com-
    pletion of part III in the pilot division, install
    identical equipment on an incremental basis in each
    division.

    Part V--Beginning in March 1968, release leased equip-
    ment at division and district level upon conclusion of
    systems testing and paralleled operation of the old
    and new system. This part also encompasses the com-
    munications tie-in between each division and the cen-
    tral hub of EIDS in the Greater Washington area.

COMPARISON OF ACTUAL PROGRESS WITH FORECAST

     Since the Corps original long-range plan did not list
or include completion dates for all the sLps necessary to
develop EIDS, we could not make a direct comparison between


                             9
                       and forecast progress  by individual
actual system progress                 is  available,  how-
development steps. Enough information             behind
ever, to show that the project is significantly
                                             long-range plan
schedule. For example, the Corps' original
                                      and debugged by
stated that P-BAMS would be installed       would be com-
June 30, 1967, and that the entire project the prototype
Rleted by June 1970. The Corps established  or about 3 years
regional computer center on March 1, 1970,     began opera-
later than originally estimated. The center
                                         Corps-wide stan-
tions with existing local systems since
dardized systems had not been developed.
                                                and their
     Some of the significant development steps
                                          of  our review
scheduled completion dates as of the time
follow.
            Development steps                Completion date
                                                Mar. 1970
Formulate system goals
Determine management information require-       Oct. 1970
  ments
Design functional systems using existing        Mar. 1971
  designs as bases                              May 1971
Develop data bases
Develop coding structures for inputting         Oct. 1971
   data                                          Apr. 1972
 Implement management information system
 Refine forecasting and simulation tech-
                                                 June 1973
   niques                                        June 1974
 Install all equipment
                                             be completed by
 The Corps estimated that the project would later than orig-
 the end of fiscal year 1974, about 4 years
 inally scheduled.
                                                       the
       Subsequent to the completion of our fieldwork,under-
                                                this
 Corps developed a revised plan for completingby the Army in
 taking. Under this plan, which was approved
                                           be developed
 October 1970, an information system will constituting the
 within each of the major functional areas areas covered by
 primary operations of the Corps. The six estate, civil
  this plan are comptroller, personnel, real       The Corps
 works, engineering, and military construction.

                               10
has completed development of a prototype finance and ac-
counting subsystem for the comptroller area and expects to
start implementing this module of the overall system by
February 1971. Modules for other areas will be developed,
tested, and implemented on a phased basis with the expecta-
tion that the overall system will be completed by December
1975.
        SYSTEM DEVELOPMENT STARTED BEFORE RESEARCH COMPLETED

          One of the basic factors gove'_ning whether systems re-
     main responsive to the needs of an organization for a rea-
     sonable period of time is the organization's stability in
     missions, goals, objectives, policies, organizational pat-
     terns, operating procedures, proolen m!.xes, and workloads.
     Because these elements determine the nature and timing of
     management's information needs, it is desirable for organi-
     zations to study these elements and to institute any de-
     sired changes in advance of development of large-scale in-
     formation systems. Defense agencies are required to do this
     by a number of Government regulations.

         We found that the Corps had not conducted such studies
!   before initiating EIDS development but instead had relied
    on studies made for other purposes; e.g., for obtaining ADP
    equipment for one of the divisions and for revising some of
    the Corps' accounting practices. After the EIDS project was
    begun, the Corps initiated the first phase of a brcad man-
    agement study and this disclosed widespread dissatisfaction
    with the existing mission staterments, goals, objectives,
    policies, organizational patterns, and operating procedures.
    The Corps has since decided not to continue this effort,
    but action may be taken from time to time on the individual
    problems identified in the study.

         We concluded that, if the Corps continued EIDS develop-
    ment in this environment, the Corps would risk either the
    continuation of these problems or the implementation of an
    information system that would have to be continually modi-
    fied to acccmmodate changes in the nature and timing of man-
    agement's information needs.

    GOVERNMENT POLICY REQUIRES STUDIES

         The Bureau of the Budget (now the Office of Management
    and Budget) issued instructions in 1960, 1961, and 1963, re-
    garding the nature and scope of studies to be performed by
    agencies before initiating system development projects.




                                  12
Guidelines issued on March 18, 1960, concerning the planning
for and conduct of studies provided that:

     1. Surveys be made from the point of view of determin-
        ing the best method to accomplish the mission and to
        expedite-the workload and not be mere computeriza-
        tions of manual systems.
     2. Surveys define the extent of the ADP problem, the
        expected economies and benefits, and the overall ef-
        fect on personnel, procedures, and organization.
     3. Surveys be authorized by officials with sufficient
        authority to effect organizational and procedural
        adjustments.
     In transmitting these guidelines to executive agencies,
the Director, Bureau of the Budget, stated that a thorough
analytical study should be conducted before decisions were
made on the economic and operational feasibility of any ADP
application.
     Bureau of the Budget Circular No. A-54, issued on Octo-
ber 14, 1961, states that decisions to use ADP equipment
must be preceded by and based upon the results of well-
documented studies which provide an adequate factual basis
for concluding that (1) functions or processes for which the
ADP equipment can be used are essential to perform and (2)
the systems, procedures, and methods to be employed in per-
forming these functions or processes have been designed to
achieve the highest practical degree of effectiveness with
optimum efficiency and operational economy.
     The Bureau of the Budget provided further guidance in
August 1963 when it issued Circular No. A-61 describing how
agencies could effectively review their ADP programs. This
circular stated that decisions to use ADP equipment-should
be based upon a documented system study in which functions
were critically examined to establish that they were respon-
sive to current or projected needs. The circular stated
also that the study should compare the benefits and costs
of the proposed system with the advantages and costs of the
existing system and should explain how the proposed system


                              13
would contribute to the more effective accomplishment of
program objectives.

      DOD Instruction 7041.3, "Economic Analysis of DOD In-
 vestments," covers the latter requirement, but DOD direc-
 tives do not specifically cover the requirement that sys-
 tems be responsive to projected needs. The Department of
 the Army, however, has issued regulations incorporating the
 essence of the guidelines issued by the Bureau of the Bud-
 get and prescribing the procedures to be followed in obtain-
 ing approval from Headquarters, Department of the Army, for
 developing automated information and data systems. These
 procedures provide that analyses be made of the missions of
 the activity. The Army regulation states that:

      "Analysis should be made from the point of view
           _iermining the best method to accomplish the
      inAssion and to expedite processing of the work-
      load, and not of merely substituting an elec-
      tronic computer for current methods. In examin-
      ing these methods, it must not be assumed that
      every operation being performed is essential to
      the overall mission of the organization, to a
      segment of the organization, or that the opera-
      tion is authorized. *** Conversely, consider-
      ation should also be given to the addition of a
      process not now being performed but which should
      be performed, and the establishment of essential
      records and reports not heretofore maintained."
 DEVELOPMENT STARTED WITHOUT
 ADEQUATE RESEARCH

      The developmentr plan for the EIDS project was based
 primarily on two studies completed by the Corps' contractors
 in 1965. The first study was performed by Arthur D. Little,
 Inc., for one of the Corps' divisions and was limited in
 scope to the activities of that division. A second study,
 performed by Arthur Young.& Company for the Office of the
 Chief of Engineers, was directed primarily toward an evalua-
 tion of the financial management system. These contractors
 recommended that the Corps establish regional computer cen-
 ters and an integrated P-BAMS.


                               14
     The scope of the work perfolmed by these contractors
under the Corps' study contrac
                            d    did not include an overall.
reassessment of the Corps' missions, goals, objectives, poli-
cies, or organizational patterns.. Furthermore, the studies
did not cover other aspects of EIDS, including the person-
nel management system and the Corps' command and control
system.
      The Corps did not supplement these studies by research-
ing other specific planning elements. For example, the
Corps' plan was assembled without prior determinations of
 (1) whether management information requirements were con-
sistent with future organizational objectives, (2) how
quickly rarious personnel needed to be furnished with vari-
ous kinds of information, or (3) which organizations should
be responsible for developing each major subsystem (P-BAMS,
command and control, engineering.and technical, and person-
nel). In addition, the Corps did not identify and describe
all tale major development tasks or their sequence, timing,
or interrelationships.




                            Is
RESEARCH DISCLOSED NUMEROU     ROBLEMS

     Our-review showed that in April 1968 the Chief of En-
gineers authorized the Engineer Strategic Studies Group to
perform a management analysis of the Corps of Engineers.
The purpose of this undertaking, known as the MACE study,
was to identify, ar.2.,yze, and solve problems in current
operations and management systems in the Office of the
Chief of Engineers and in the division and district offices.
Project justification documents state that previous Corps
studies have been limited in breadth either to single major
organizational areas, such as civil works, or to specific
resources and functions that affect all organizational ele-
ments and that the current project would be the first com-
prehensive analysis of the Corps on a total-entity basis.

     The first phase of this study, completed in November
1968, was an overview of operations throughout the Corps to
identify, for further investigation, specific problems that
were impeding resource use and mission accomplishment. In
the course of this work, 110 separate dissatisfactions and
problems were compiled.

     The following examples show dissatisfactions which
could have a significant impact on the EIDS project.

    1. Civil works

        a. The basic civil works and field organization
           structures need evaluation in terms of (1) today's
           communications and transportation, (2) necessity
           for present levels of organization and assign-
           ments of-functions, and (3) delegation of author-
           ity.

       b. The research and development program in civil
          works lacks the necessary cohesive, coordinated,
          and responsive structure to relate the program
          to the Corps' needs.

    2. Military construction

       a. The Corps lacks a formal Corps-wide policy to en-
          able all its relevant experience to be brought to


                               16
            bear in improving the design process. This lack
            causes individual districts to experience every
            pitfall for themselves.

         b. A new architect-engineer design.system should
                                                          be
            developed because the present rigid method of
            design of standard construction makes no use of
            such creative concepts as functional multiuse
            buildings, design for maintainability, and reli-
            ability, or future technology.

         c. Existing reporting systems do not permit manage-
            ment to forecast technological needs, to estab-
            lish research strategy, or to evaluate research
            productivity.
     We were told that the Corps had considered performing
a second phase of this study either by Corps personnel
                                                        or by
contracting with management consulting firms. We
                                                  were in-
formed that the Corps' discussions with these firms
                                                    centered
around questions of the appropriateness of the Corps'
nizational structure, procedures, and systems as       orga-
                                                 they re-
lated to its anticipated future environment.   We were advised
that the Corps had decided not to continue the studies
detail, not to make decisions on proposed changes,       in
                                                    and not
to prepare implementation plans because the cost
                                                  and  person-
nel requirements were higher than the Corps was
                                                 willing to
devote to the project. Action may be taken from
                                                  time to
time, however, on t:le individual problems identified.
                                                         At
the time of our review, the Corps had not formally
                                                    assessed
the individual impact of these problems on EIDS.

     Our review also revealed that the Chief
later restated t'le goals of the organization of Engineers
                                              and directed,
in April 1969, that plans be developed for attaining
                                                      these
new goals. The following examples illustrate
                                               some of the
new goals established for Corps organizations.

     1. Engineer Comptroller

       Improving the evaluation of mission performance--
       Develop improved methods for informing the Chief
       Engineers of the adequacy of the Corps' mission of
       performance, including the selection of key


                               17
  indicators and control factors for major activities
  that influence the effective accomplishment of the
  Corps' mission.

2. Military Engineering

  Establish system for.monitoring engineer require-
  ments--Establish and operate a "closed loop" system
  for obtaining.engineer requirements through techni-
  cal channels from the field, for evaluating and
  forwarding tiese to the appropriate agency for ac-
  tion, for monitoring progress, and for reporting
  back to the field. This system would include engi-
  neer staff sections.
    CONCLUSIONS

         The success of a long-term
                                     undertaking--such as
    EIDS project--depends upon                             the
                                the willingness and ability
   an agency to apply,                                       of
                        to project development,
   criteria which are compatible                 management
                                   with the long-term
   of the project. Thus                               character
   for all aspects of a long-range solutions should be
                         project and the quick            sought
   by expediency should                         solution offered
                         be avoided.
        We believe that the
                               Corps and the Army
   greater emphasis, at                             should have put
                          the start of this effort,
   range solutions for                                 on long-
                        the EIDS
   perform sufficient research, project.      The Corps did not
   plan development of             in our opinion, to
                        EIDS. Subsequent research adequately
   that a number of aspects                            indicates
                              of the Corps' activities
  need of internal readjustment.                           may be in
  whether it should define            The Corps must now
                              and incorporate needed        decide
  ments into its overall                                readjust-
                           management system before
  further or should risk
                           continuation of potential proceeding
  for an indefinite future                               problems
                              period. The latter
  tious condition that                              is the vexa-
                         system research seeks
                                                  to avoid.
       In our opinion a function
 permit managers to make            of system research
                                                          is to
 what changes should        timely  and appropriate decisions
                       be incorporated into                       on
 der development and                           the new system un-
                      on what changes can
 ferred to a later date.                      reasonably be de-
 investment                  The  alternative  seems to be the
 procedures of  substantial
             to serve goals funds    to mechanize functions
                              and missions whose effectiveness and
 has not been currently
                          assessed.
       Thus without
 assurance that the systems  research the agency
                                                   has little
                     system under development
 the needs of the organization                  corresponds to
                                 or that it will be
 in the foreseeable                                   acceptable
                     future. In addition,
 the development effort                      large portions of
project may be delayed may be wasted and completion of the
                         significantly. In
lack of a demonstration                       this case the
                          of system durability,
usefulness, and the                               or long-range
                      agency's failure to
policies in acquiring                      follow prescribed
                        the equipment (see
reasonable doubt about                      ch. 3) raise a
                         expanding the EIDS prototype
other installations.                                    to




                              19
      We believe that this case demonstrates a need for
 guidance concerning the durability of ADP systems.
                                                      The de-
 velopment of new systems is too important and expensive
 leave unanswered the question of whether the system      to
 nent has conducted sufficient research irto the      prop)-
                                                  factors af-
 fecting the long-range usefulness cf ,.. system.
                                                    Existing
 DOD guidelines are silent on this aspect of research
                                                       prece-
 dent to ADP systems design.

      The Office of the Assistant Secretary
 troller) has responsibility for the review of Defense (Comp-
                                            and approval of
 major new systems. It would seem appropriate
                                               for
 fice.to obtain from the system proponent, before this Of-
                                                   authoriz-
 ing a development project, a statement describing
 and nature of the research performed, the problems the scope
                                                     existing,
 and the probable effect of these problems on the
                                                   long-range
 usefulness of the system.

 AGENCY COMMENTS AND ACTION TAKEN

      A draft of this report was furnished to DOD. Drafts
were also. furnished to the Bureau of the Budget
                                                   and the Gen-
 era! Services Administration because these agencies
                                                        are p-i-
marily responsible for ADP policy and procurement
                                                     matters.
The General Services Administration advised us
                                                 that it
curred with our views that long-range system studies, con-
rather than short-range expediency, should form
                                                  the basis
for the acquisition of ADP equipment (see app. III).
were advised informally that the Office of Management We
                                                          and
Budget did not specifically comment on this matter
                                                      because
we were citing an illustration of noncompliance
                                                  with an
existing policy and were not recommending a change
                                                      in pol-
icy.

      The Office of the Assistant Secretary of Defense
                                                        (Comp-
troller) advised us that, as a result of a number
                                                   of
changes that had occurred, it planned to review
                                                 the Corps'
entire long-range plan and that, at the conclusion
                                                    of
review, DOD would be in a better position to comment this
                                                       on the
Corps' future plans (see app. I).

     That Office reviewed the Corps' long-range plan in
June 1970 in accordance with the guidelines in
                                               its


                             20
  instructions recently adoptedL
                                  to provide for better control
  over authorizations for.new
                                system development projects.l
  Under these guidelines, system
                                   proponents are required to
  demonstrate tePhnical, operational,
  at early star                         and economic feasibility
                   'f the project. The review
 the Corps net     to improve and complete the disclosed that
 of the EIDS poject in several                   documentation
 economic analysis, justificationsignificant areas, including
 for completion of all system        of requirements, a plan
                                modules, a study depicting
 tem durability, and the approach                            sys-
 engineering applications           being  used to standardize
                            between and among engineer
 sions.                                                 divi-
       The Assistant Secretary of
 for-med the Assistant Secretary Defense (Comptroller) in-
                                  of the Army (Financial Man-
 agement) that :hese actions
                              were to be completed prior
 any request for the extension                            to
                                of the EIDS prototype.
      Although t' s action seems
the Corps' system will have       to provide assurance that
                             to
futiure needs before additional  be  responsive to the agency's
                                 equipment is acquired,  we
believe that consistent guidance
DOD agencies to prevent similar     should be furnished to all
                                  occurrences.
RECOMMENDATION

     We recommend
existing directivesthat the Secretary of
                     to include guidance Defense  revise
of major new ADP systems                  to the proponents
                          concerning trie need for documented
studies of the factors affecting
of the system.                    the long-range usefulness


1DOD Instruction 5010.27,
                          "Management of Automated
 Systems Development,,'dated                       Data
                             July 8, 1970.




                            2!
                             CHAPTER 3

        FACTORS NOT CONSIDERED IN
                                  ACQUIRING ADP EQUIPMENT
        The long-ierm cost of owning
                                     general-purpose ADP equip-
  ment is influenced by the
                             design of the system for which
  the equipment is acquired
                             and
  ment to execute the functions by the capacity of the equip-
  sign. Both should be consideredprescribed by the system de-
                                    carefully before decisions
  are mide on which equipment
                               should be acquired to perform
  particular task.                                           a
       Defense policy since 1966 has
  placement computers be obtained    required that new or re-
                                    only after systems have
   been redesigned to make full
                                 use of the improved capabili-
   ties of later model equipment.
  a contract for EIDS equipment     The Corps, however, awarded
  the process of being designed   while the EIDS system was in
  was purchased under workload and developed. The equipment
  obsolete local systems rather specifications prepared from
                                  than under workload specificaL-
  tions fo.- the standardized
                              EIDS which wa.s then under
  opment. (See p. 25.)                                   devel-

       In selecting the successful
                                    vendor, the Army used a
  mathematical formula for evaluating
  proposals, such as price and         elements of the various
 was in the Army's standard    performance.   The formula used
 early procurement stages butoperating  procedures in the
                               had since been replaced  by an
 improved formula. Had the
                             Army used the improved formula,
 a different vendor might have
 overall score, which might     been assigned the highest
                            have resulted in award of
 contract to a different vendor.                        the
                                   (See p. 28.)
      Government regulations
 procurement to the present in effect from the time of this
                            have not required, in the
 tion process, that consideration                       selec-
                                   be given to differences
 in residual values--usefulness
 longer needed for its original for other purposes when no
 equipment when determining     purpose--of dissimilar
                            which vendor has offered theADP
best buy. The need for a Government
values for use in future procurements policy on residual
cated by factors in this case;          of computers is indi-
ment offered by the low bidder  i.e.,  that some of the equip-
                                had been previously sed,


                            22
 that it had been out.of production for several
                                                 years, that
 the equipment selected has since been classified
 by consultants to the Blue Ribbon Defense         as obsolete
                                           Panel, and that
 there is a probability that large quantities
 ment will be offered to the Government when of used equip-
                                              the new equip-
 ment series rerently announced by equipment
                                              manufacturers
 becomes generally available. (See pp. 30
                                           to 33.)
 STATUS OF EQUIPMENT CONTRACT

      In June 1968 the Corps awarded
 eral Electric Company to purchase 11 aGE
                                        contract :o the Gen-
                                          225 computers that
 the Corps had been leasing and to provide
 services having a total value of about $3 for other ADP
                                            million. Also,
in May 1969 the General Electric Company was
tract to provide the ADP equipment for EIDS. awarded a con-
to be furnished included GE 425 computers        The equipment
                                            for use by eight
Corps divisions and GE 225 computers for use
tricts and one division. The equipment would by 33 dis-
                                                 be provided
during fiscal years 1971-75 under options
                                            of the contract
which, if exercised, permit the Corps to
                                           purchase or lease
the ADP equipment. .If the Corps elects to
                                             purchase all the
equipment, the cost to the Government will
                                             be
$7.6 million. Maintenance and communication about
                                                costs are es-
timated to total about $600,000 annually.

     We were informed by the Corps that the ADP
                                                 equipment
for EIDS was installed in March 1970 at a
                                           prototype instal-
lation and that it would be leased by the
                                           Corps for
1 year. Tne purpose of this prototype installation about
permit the Corps to test, study, and evaluate        is to
gional computer center concept, (2) P-BANS     (1) the re-
work is finished on this portion of EIDS,   when development
                                           and (3) standard-
ization of technical applications.  Upon successful comple-
tion of the prototype test, the Corps will
                                            extend'P-BANS to
other divisions and districts and provide
                                           each with the
same equipment.

     The contract for EIDS equipment was awarded
General Electric Company under negotiated        to the
cedures after proposals had been obtained contracting pro-
                                          from four poten-
tial suppliers. In evaluating the prices
                                          quoted by the



                            23
 four bidders, the Army calculated that the
Government to acquire the equipment from thecost to the
 tric Company was about $1.4 million, or about General Elec-
'less'than the cost to acquir, equipment offered15 percent
                                                  by any of
the other vendors. This w primarily because
                                                 some of the
ADP equipment offered by 2.e General Electric
been used previously whereas the other vendors Company had
and more sophisticated machines belonging to     offered new
                                              a
ment series. The GE 225 computer was manufacturedlater equip-
the period 1960-67. Under the request for proposalsduring
                                                       issued
by the Army, used ADP eTuipment was acceptable
the vendors could have offered such equipment. and any of




                           24
 EQUIPMENT CONTRACTED FOR BEFORE
 SYSTEMS WERE DEVELOPED

       On July 29, 1966, the Secretary of Defense issued a mem-
 orandum to the heads of all Defense components, stating
 that:

     1. Defense agencies must insist on systems which sat-
        isfy the total management and operating requirements
        and which exploit the unique capabilities of the
        computer.

     2. Defense agencies, prior to computer selection, must
        develop and issue system specifications which ade-
        quately describe the work to be performed and which
        will result in the selection of computers which can
        satisfy the requirements of that specification.

     3. Defense agencies should not be forced to acquire
        additional units at later dates or to replace com-
        puters prematurely because the equipment was se-
        lected on the basis of'inadequate system specifica-
        tions.

     4. Defense agencies must select and acquire new or re-
        placement computers only after systems have been re-
        designed to make full use of the improved capabili-
        ties of later model hardware and then only when
        there are proven cost benefits. In these cases,
        systems redesign and programming should be accom-
        plished prior to delivery of any equipment.

     The Corps developed its equipment specification and
contracted for the equipment on the basis of the engineering
and business workloads experienced by the installation se-
lected to serve as the prototype for the regional computer.
center and used the installation's local systems and non-
standardized procedures. The systems mentioned in the EIDS
long-range plan had not been developed at the conclusion of
our fieldwork, so their impact on the specification could
not reasonably be estimated. Further, the computer programs
for EIDS had not been prepared as of February 1970.




                             25
        Contracting for equipment
   seems to preclude an optimum    under these circumstances
                                 match .of equipment to systems
   and workload and, in our
                             opinion, is not consistent
   the above-w1intioned Defense                          with
                                policy.
   Aency comnents and our ejvaluation
         The Assistant Secretary of
                                     Defense (Comptroller) fur-
  nished us with the reasons
                               submitted
  ing from DOD's current prescribed         by the Army for deviat-
                                       procedures for selecting
  equipment. (See app. I.)
                               The Army stated, in essence,
                                                                 that:
        The design of P-BAMS was being
                                          hampered by the fact
        that the equipment configuration
        Therefore the selection of           could not be specified.
                                    hardware    was begun in Novem-
       ber 1967 on the basis of the
                                       old  system  modified by the
       P-BAMS concepts. This decision
       responsible parties since           was  concurred in
                                   the system described in by  the
                                                                  the
       hardware specification equaled
                                         the computer workload
       that would be required by the
                                        new system.
       We recognize that a necessary
 sign is the determination             condition to system de-
                             in generic terms of an appropriate
 equipment configuration including
 output methods and such devices      the selection of input and
 visual displays. After the        as  punched cards, tapes, or
                               system proponent has reached
 final decision on equipment                                    a
                               configuration, the design
 system can be completed.                                    of the
                            The fact that the agency was
quired to determine the equipment                             re-
                                     configuration for design
purposes does not, however,
agency to procure specific constitute justification for the
                             makes and models of equipment
fore it has complied with                                       be-
                            the policy guidance furnished
the Secretary of Defense.                                      by

      We dO not agree that the system
 curement specification was             described in the pro-
                            an accurate projection of
 workload requirements. Prior                           future
ment specification, the Corps to the issuance of the procure-
research and resolved basic     (1) had not completed system
                             questions concerning its future
goals, policies, and procedures,
                                   (2) had not standardized
and reevaluated engineering
stated requirements for otherrequirements, and (3) had not
                               portions of EIDS, including
command, control, and personnel.
                                    There is a strong


                              26
probability that equipment acquired under such conditions
will not provide an optimum match between the equipment and
the workload.
    PRESCRIBED FORMULA
                         NOT USED
    IN EQUIPMENT SELE.CTION

         The ADP equipment
                            specification indicated
   Army's Computer Systems                           that the
                             Support and Evaluation
   develop an overall scorp                          Command would
   configurations and         for each of the proposed
                       would use this score             equipment
   equipment. Personnel                      in selecting the
                          of the command have
   the vendor with the highest                 advised us that
                                 overall score ordinarily
   lected.                                                 is se-

        The scoring
  ferent weights forprocedure gave consideration to the
                      such factors as cost,                   dif-
  other considerations.                         efficiency, and
  culating the number        The mathematical
                         of points given to formula used in cal-
  ciency factor differed                      vendors for
  command's standard        from the formula contained the effi-
                       operating procedures               in the
  that time, according                         manual in effect
                          to responsible personnel.               at
  that we examined did                                  The records
                          not reveal why the
 had not been used.                            prescribed formula
                        We believe that, when
 from prescribed practices,                      agencies deviate
 plaining why the               a record should be
                    prescribed procedure             prepared ex-
 indicating that the                       was not followed and
 and approval of the deviation occurred with the knowledge
                       managers of the agency.

       In this case the formula
 about 3 months earlier           actually used had been
                                                           replaced
 old formula a vendor      by a new improved
                        could propose a longerformula.   Under the
 ing time then the                               workload  process-
 number of points ormaximum allowed and still receive
                       could take less time               a larger
 but not receive any                         than the maximum
                      additional points
processing time.                          for proposing faster
                   Thus the formula used
ficiency in terms                           did not measure ef-
                   of a relationship
minimum times r'quired                 between
                          by various vendors'    the maximum
dividually process                             equipment to and
                                                              in-
                     the workload described
tion.                                         in the specifica-

      Our calculations showed
                               that, if the newly
formula had been                                  prescribed
                  used, one or more
might have received                  of the other vendors
                     a sufficiently higher
efficiency factor                           rating for the
                   to have given it a
than that of the General               higher
                          Electric Company. overall score




                              28
 Agency comments and action taken

      The Assistant Secretary of Defense (Comptroller) ad-
 vised us that the Army had amended the standard operating
 procedures used by its computer selection office to require
 that all future deviations from prescribed procedures
 documented, reviewed by appropriate functional offices be
                                                         prior
 to submission for approval, and entered into the official
 record of the solicitation and award.

     The Assistant Secretary of Defense (Comptroller) ad-
vised us also that the Army had furnished the following
                                                        ex-
planation of why the prescribed formula had not been used
in this case.

          "The currently prescribed formula for scor-
     ing ADPE efficiency was not used at the time of
     selection because this scoring procedure was de-
     veloped after the Request For Proposal (RFP) in
     question had been issued to industry. This pew
     procedure established a requirement fcr includ-
     ing certain parameters in RFPs for proper appli-
     cation of the scoring formula. The formula that
     was used in the evaluation of the Corps project
     was in accordance with prescribed procedures
     used at the time the RFP was issued to industry."
     (See app. I.)

     According to Army records, the proposal request was
issued to industry in January 1968 and the four competing
vendors were resolicited by the Army on December 20,
                                                      1968,
or 3 months after the date that the new scoring procedure
was established. Thus, although the Assistant Secretary's
explanation is te':hnically correct, we believe that there
was nothing to preclude use of the new procedure.




                            29
   RESIDUAL VALUE OF ADP EQUIPiNT
   NOT REqCOGNIZED

       We noted during our
  the respective equipment review  that the residual
                           was not recognized or usedvalue of
  the selection process. 'Residual                     during
                                    value, as used herein, is
  a quantitative measure of the
                                continued usefulness of equip-
  ment for other applications when
  its original purpose.             it is no longer needed for

        As in the case of other durable
  generally has a useful life that        goods, ADP equipment
                                     extends over a number of
  years. Frequently, this useful
  riod that the equipment is neededlife is longer than the pe-
                                      for the purpose for which
  initially acquired. At the time
  equipment may be transferred      of its replacement, the
                                to another use within the Gov-
  ernment, traded in for new equipment,
                                           or sold.
       Since the residual value may
                                      be quite large, it seems
 reasonable that it should be given
 time that the equipment is purchased. some recognition at the
 relevant to take residual value            It seems particularly
                                   into account when the cost
 of competing equipment is being
 of the equipment offered has a evaluated and when certain
                                 much greater usefulness for
 other applications--residual value--than
 equipment.                                   that of competing

       This circumstance applied
  acquired for EIDS. As mentionedin the case of the equipment
                                    above, the General Electric
 Company offered used equipment
                                 whereas the other bidders of-
 fered new and more sophisticated
                                    machines belonging to a
 later series. Government regulations
                                         do not require that
 consideration be given to residual
 competing bids. Thus, in making      value in evaluating the
                                    the award to the General
 Electric Company without considering
 Army was fol-owing its normal          residual value, the
                                practice and existing policy
 directives of DOD and the Bureau
                                   of the Budget.
      Subsequent to the completion of
July 1, 1970, the Blue Ribbon           our fieldwork, on
                               Defense Panel issued its re-
port to the President and the
                               Secretary of Defense. During
its consideration of ADP within
Panel received an independent     DOD, the Blue Ribbon Defense
                               analysis prepared by



                              30
consultants and published portions of the consultant's re-
port as appendix I to the Panel's report. According to
this document, the consultants found that DOD had in its ADP
inventory several machines which would be considered obso-
lete by most computer personnel. The General Electric Com-
pany GE 225 computer that the Corps contracted for was in-
cluded in this category. In the judgment of the consultants,
these machines were no longer capable of performing work ef-
ficiently. It should be noted that, in publishing the find-
ings of these consultants, the Panel stated that it did not
necessarily endorse each of the consultant's findings and
recommendations.

     Our studies indicate that ADP equipment does have a
substantial residual value after 5 years and, in some cases,
after 10 years or more. Further, the probable residual
values differ greatly by type of machine, degree of sophis-
tication, and manufacturer.

     Although to date DOD generally has not sold equipment
no longer needed for its original application, residual
value should not necessarily be ignored indefinitely. When
a DOD agency decides that it no longer needs a particular
piece of ADP equipment, alternative uses can be sought if
the equipment is not obsolete. The equipment may be put to
another use within DOD or transferred through the General
Services Administration to another agency. In the latter
case the General. Services Administration may lease the equip-
ment, the lease charges being based on the fair market value
of the equipment at the time that the lease is negotiated.
This seems to confirm the existance of a residual value and
its relevancy to ADP equipment selection. Further, even
when alternative uses for equipment are found within DOD,
it seems prudent that, in selecting the best alternative ap-
plication, cost-versus-benefit studies should be made. If
such studies are to be realistic, a residual value should be
assigned to the equipment.

     Also, it should be noted that DOD and Office of Manage-
ment and Budget policy directives provide for considering
residual value in cost-benefit studies to be used in deter-
mining the relative economy of purchasing or leasing ADP
equipment.



                             31
         Thus existing policy provides
   of ADP equipment in deciding         for considering the valt
                                  whether
   equipment and in the interagency        to buy or lease the
   Computer manufacturers have        transfer  of ADP equipment.
                                recently announced new equipmen
   series, and, as this equipment
   substantial quantities of        becomes generally available,
                              used  equipment will be available
   for reuse and resale. Consideration
  future procurements could               of residual values in
                             provide the Government with
  sound basis for distinguishing                            a
  buy and the low-price obsolete   between  the  economically best
  degree of assurance              machine.   In
                       of continued spare parts addition, the
  support for the used equipment                   and software
  in this respect.                 merits special consideration

       It therefore would seem
                                desirable to consider residual
 value--or a comparably weighted
 from the equipment offered        alternative--in selecting
                             by
 would be the most advantageous  competing suppliers
                                  to the Government. that which
 Agency comments and ourtevaluation

      The Assistant Secretary
 vised us that DOD had          of Defense (Comptroller)
                       established a task group           ad-
jective of developing and                         with the ob-
                            recommending an economically
 and relatively simple technique                           sound
                                   for future use in evaluating
differences in residual
                         value
is offered by manufacturers. when dissimilar ADP equipment
that the final report of         The Assistant Secretary
                                                         stated
pleted but that we would  the  task  group had not been com-
                          be able to obtain a copy
port upon its completion.                            of the re-
                             (See app. I.)
      The Office of Management
 sidual values should be         and Budget advised
                         a factor to cons.der in us that re-
quisition of ADP equipment.                         future ac-
arranged to obtain data        The Office stated that it
                         from special studies and          had
sources and was hopeful                              other
                         that
It offered a number of reasonsthis would prove to be useful.
not require the consideration why current regulations did
stated, in essence that:        of residual values. The
                                                           Office

     -- Residual values were
                             subject to change because
        merous factors, such as                         of nu-
                                age, popularity, market
        ditions, and suppliers'                          con-
                                status.



                              32
           -- Residual values
              into the future could not be projected
                               with any degree           4 to
              there was no broad                 of confidence6 years
                                   base of industry
             in assessing                             experience since
                                                                  to use
                           these factors.
          -- Procurement
                          decisions based
             the probable                   on subjective
                           residual                         judgment of
            ment could result         value of each
                                 in protests by      offeror's equip-
                                                 the losing vendor.
         -- Residual values
                              may tend to establish
            petitive advantage                         a built-in com-
            facturer (see         in favor of the
                           app. II).               most popular
                                                                  manu,
   The General Services
   these Views.            Administration generally
                                                        concurred with
                    (See app.
                               III.)
         The concerns
                       expressed by
    two things. First,                these agencies
                        there  is                      center around
   atively sound set of            a difficulty
                          values corresponding in obtaining a
   be realized at                                 to the amount rel-
                                                                   to
                   some distant time
   used ADP equipment.                  from the sale
   resold by the          Since the equipment           of a piece of
                 Government,   we                is
   secondary to                    believe          not  generally
                                            that this problem
                estimating the                                   is
  ond user at the                 value  of the equipment
                   end of the period                        to a sec-
  the function                           or at the conclusion
                for which the
  We have been                  equipment Was                   of
                advised that                    originally acquired.
  DOD tack group is in the     the  approach
                             direction of a being
                                               sound followed
  satisfy the latter                                  techniqueby to
                                                                   the
                      purpose.
       Second, equipment
 protest if the             manufacturers
                 residual-value            could be expected
 Government provided                technique implemented       to
 ular equipment        a built-in    unfair advantage       by the
                 vendor. We recognize                  to a  partic-
 exists and believe                        that this possibility
                      that
signing and implementing due care should be
                              any technique      exercised in
We believe that,                             fir                 de-
                   prior   to  the implementationGovernment   use.
developed by
              the                                  of any technique
quested to presentGovernment, industry members
                     their views on                 should be re-
sideratioa.                            the technique
                                                      under con-
Re--coniendat ion
      We recommend
                  that DOD obtain
                                  and consider
the computer                                   the views of
             industry and
                          coordinate the
                                         Department's


                               33
efforts with the Office of Management and
concluding the study of residual values   Budget before
                                        or a comparable al-
ternative.




                          34
                          CHAPER 4

                       SCOPE OF REVIEW
     We reviewed efforts by the Office
                                       of the Chief of En-
gineers to develop EIDS in relation
                                    to system development
concepts, policies, principles,
                                and procedures contained in
DOD directives and memoranda, Department
lations, and Bureau of the Budget        of the Army regu-
                                  circulars and bulletins.
      Specifically, we reviewed study
                                       reports prepared by
Arthur D. Little, Inc., Arthur
                                Young & Company, and other
 study groups; the documents used
                                   in obtaining Department of
 the Army approval of the EIDS project;
                                         the Corps' long-
range plan; equipment specifications,
tracts; and related memoranda and       proposals, and con-
                                    correspondence. Also, we
discussed the development of EIDS
                                    with management officials
of DOD, the Department of the Army,
                                      the Office of the Chief
of Engineers, divisions of the
                                Corps, and various ADP equip-
ment suppliers. We did not make
                                  a detailed review of the
subsystems being designed for EIDS.

     During our review, which was performed
                                            during the pe-
riod May 1968 through February
                               1970, we visited the follow-
ing locations.

    Department of the Army:
        Management Information Systems Directorate,Wash.
           ington, D.C.
        U.S. Army Computer Systems Support
                                            and Evaluation
          Command,Washington, D.C.
    Corps of Engineers:
        Office of the Chief of Engineers,
                                          Washington, D.C.
        Division offices located in Dallas,
                                             Texas; New
          York, N.Y.; Cincinnati, Ohio; and
                                             San Francisco,
          California




                           35
APPENDIXES
                                                                                   APPENDIX
                                                                                       Page


         tj. r7N
              . *,
                 '~               ASISTA14T   FCREttP.Y or CrFIJSE
                                       WASIINGTON. b.c.   so351




                                                                  10 JUN 1970

     Vr. C. M. Bqilvy
     Director, Def.':ue D1vir.ion
     U. S. Gcneral Accounting Office

     Dear ir. Ei(..-:
           Thio Jett'e   re.oonds to you': relest for c:,.Sents
     repIrt review'nV. it.e efforts                                on the urcft
     an integrated rtorition a:nd of .it.a
                                         tuh..
                                             ArLy Corps of,ri:.-e
                                            syster ({.(D Cr.se 3101l). to   develop
                                                                         Ti:is
     office ~apprec.atj, Loth the e).Ltnsive
     the re'o..,enr-z,tiois :.ue as a           rev.    rd,: by your office and
                                      result of that review.
          Currently, . nuitbecr of ch.ncr.$ have
                                                   occurred v*hieh r:yv have altered
    the situation e'cribnnd within the report.
    ch'`,e.s, My offir.c, ILL.,S to revicw            As & result of th'.se
    XRanc Plar, in rid-j-.e. At the        the entire C^.ps of Enginueors Long
                                        conclusion of thi. review *' should
    in a position to com.m'ent r.r. fully
                                             on the status and future computing  Io
    pa.uas for the Cowps of Engineers.
            Tlhe enclosure nrovides the reasons
     viating f:'6m th ctrrcrtly p-,res;crtbcd       svb.nitted by the Ariny for de-
                                                  prcc.:,--e for st:tinE Uuto-
       .,ti.. DaCta i'r;c.-r. ;:uii.--nt. (ADPE).
     t'hat the sc;ect'on c.ffice cend:               S're Ar-.y 'Ls advased tilis office
     require that all fu:tur- dev.ttin;s tJ.S. ELar.:Lrd oprztinr.g    pro.cdurcal to
                                             from pre^cr:bei p1'ccd'res are d.cu-.
    Lented, revicv'c. hy the aL.prc-'ri..te
    sion for appro-val, ard enter.d ihjto fulctio:j,l off ice prior tj submiiu-
                                              the ffic'.Ul record.
            Thi.s of'ice. :.rees wvth your rccoz.-end:ttlr
    t1icn pro.edurt.. be exp':nded to iLclude                  that the /aPfrE sclec-
                                                  a, t:-.h i:.:e for ev':iiutin6
    feec       s in reidtl                                                        dif-
                          value hcre dilsimrie.r Il..    is offcre' by
    rrnuftacturers.To this -'ud, a tulsk grosU wi,:;
    1910, uder th:-o                                  e.-tblishod on Fetbruary 13,
                    :id-r.ce or the Depa.-t-ent
                   Cr                               of Defense Ah-' l'olicy
   Co-%Itthee. %:hI], tre final rrort
   plets.d, it cb.ctivcj :re to c'vei-op o.'  t!-e tiLsk .rotp has not beer. con-
                                             and
   is bulh econn:'^all; sourcd -ni re'atively;r    re_-=:.end e.tech:iqalc uhich
   yosr offitc h:Ce beetCcfntactc(-.               s:i.- to asply. Mcr.bher of
                                     to discu.s t.is Ssr.se. ShouJd you
   Circ, a co.' or thc fitr.l report                                          dc-
                                      can be radc available at 'ompletioon.


rEST
 DOCUMEFNT AVAILABLE                       Sincerelyp


    Lncloaure
                                                  Icr:',:                    e,.

                                      tirt nri Stecrl     .y   'e.;,o    C




                                         39
APPENDIX I
    Page 2


                                                             Enclosure
          In an effort to save approximately 2 years' time and stay on sched-
    ule with the Corps of 2ngineers Long RanGe Plan, the Aruy selected
    equipment for a prototype installation prior to completion of redesign
    of the AD'P sf,.ttCa. A contract with Arthur Yourn and Company to rede-
    sign the ex.i:ting Corps of Engineers Accour.ting System into a Program-
    Budget Accountin5 lanacement System (P-L:1,5) was let in July 1966 and
    was expected to produce a completely redeci(ned system by Inarch 196B.
    By mid-1967 slipiagec in this contract effort rade it obvious that re-
    des:,n of the system would not be completed on schedule. At this time
    a situation existed whereby de.ign of P-BA.L¶ was being hampered by the
    fact that thz· ALX.? coni'lguration could not be specified, and computer
    procure!ment wus icbing delcycd by lack of a redesigned ADI' system. There-
    fore, selection of hardware v3s begun in ;;oveLber 1967 based on existing
    accounting systems modificd by the Fi4B/MS concepts which had been devel-
    oped up to that tme. The Dcpartment of Army Hanagument Information
    Systems Directorate, the Corps of Engineers and the syst'ms contractor
    concurred in the decision to proceed with hard.are selection on the basis
    that the system described in the ADPE specifications equaled the com-
    puter workload that would be required by the redesigned system.

          The currently prescribed formula for scoring ADPE efficiency was
   not used at the tin- of selection because this scoring procedure was
   developed after the Request For Proposal (RP.?) in question had been
   issued to indust37;. Thin new procedure established a requirement for
   including certain parencters in P1iPs for proper application of the scor-
   ir.nj for.la. 5Tin' fcrmula tnat was used in the evaluation of the Corps
   proJect .as in Lccordsncc with prescribed procedures used at the time
   the FF? was issued to industry.




                       BEST DOdI£ENT AVILABLE




                                       40
                                                                APPENDIX II-
                                                                     Page 1


                          EXECUTIVE OFFICE OF THE PRESIDENT
                                                        BUREAU OF Tn E rtUGET
                                                          WASHINGTON, D.C. ROM




 MHAY 28 1970

 1.r.   A.   T.   Samuelson                 BEST DOCUMENT AVAILABLE
 Director, Civil Division
 U.S. General Accounting Office
                                            BEST DOCUMENT AVAILABLE
 Washington, D.C. 20548

 Dear Mr. Samuelson:
On March 26 you invited comments on a draft report entitled
 Review of efforts by the Corps of Engineers to develop
integrated information and data system."                 an
                                          Included in the
draft report is a recon.nendation that policies on
                                                   the selcc-
tion of cc..-puters should be modified to provide for the
consideration of residual values.

The objective of your recomnr.Endation is to assure
                                                       the selec-
tion of the lo?;est cost alternative by considering
                                                        the
estimated value of the eauipment at the end of the
                                                        period of
use as well as the initicl procurement cost. This
                                                        i_ consis-
tent with the general policy in bureau of the Budget
No. A-54 which states that selections will be based Circular
                                                         on ,cquiD-
ment capability an;  3.owest overall cost. liow.Bvcoi, we have
not, in this policy, specifically identified residual
                                                            values
as one of the factors to be considered in determining
                                                            the
lowest overall cost.   We foresee a number of perplexing
problems if we were to do so.

As pointed out in your report, residual values can
                                                          differ
greatly r.monrg equipn.ent models.  They are affected by age,
popularity, actual ant potential technological obsolescence,
market conditions, and even by the comoetititive and
                                                          financial
status of the suuplier. For these reasons, we have
                                                            resisted
the use of an arbitrary    formula a:. a uniform r.. ch,-nism for
computing residual values on different models.        Further,
there is no broad base of indu.try expcrience or
                                                       generally
accepted guidelines which enable us to ; sscss all of
                                                             th..e
factors and project an estir&atud residual value for a given
model four to six years into the future with    any degree of
confidence that the estimate would hold u-.    1 ithout such
experience and industry guidelines to support our cstirat:es,
procurcm.ent decisions that rcfl.ected an indoc:lndent, r.u!jec-
tive jurgmeat on the probable residual valier of eaich offeror's
equipment could becomne the target of continuing protests
                                                             by
the losing vendors that would bn difficult to resolve.




                                       41
APPENDIX II
     Page 2



   Also, in forecasting such residual values, it is probable
   relatively higher values would be assigned to computers   that
   the greatest degree of popularity, in the
                                                           with
                                            same vein that the
  more popular automobiles usually command higher trade-in
  allowances. This would tend to establish a built-in
                                                        competi-
  tive advantage in favor of the most popular manufacturer
  each instance, and would perpetuate his predominance      in
                                                        to the
  detriment of others. Under existing circumstances
                                                     we do not
  believe this to be in the best lonaer-range interests
                                                         of the
  Government or the computer industry.
  Your report notes that the Government has already
                                                       recognized
  the relevancy of residual values to proper management
                                                           of
  computer equipment in (a) the General Service Administration's
  practice of placing a fair market value on excess
                                                      equipment
  that is transferred from one agency to another,
  versus lease decisions under Circular A-54. Withand (b) purchase
                                                      regard to
  fair market values, these are, of course, determined
  General Services Administration on the basis            by the
                                                 of current
  conditions and a }:no%.n market, and do not involve
                                                      the more
  difficult judgment of estimating future values.
  to purchase versus lease decisions, you are aware With regard
                                                      that, for
  the reasons stated above, we have not yet found it
  to establ,.ish definitive guidelines for estimating feasible
                                                      residual
  values and implementing the policy on a uniform and
                                                         consistent
  basis.

 We are hopeful that it will become possible to establish
  acceptable means for projecting residual values because,
 you point out,. they should be a factor in management      as
 The experiences of leasing companies, iwho have thus   decisions.
                                                       far
 generally limited their activities to relatf.vely
                                                   low-risk
 situations, the emergence of a used equipment marketing
 industry, the experiences in the excess redistribution
 and the results of special studies being made on         program,
                                                   this problem
 may prove useful in this regard. Further, current
                                                     efforts to
 move toward co!petitive. procurement based on a supplier's
 offer of a firm price over the anticipated life of
                                                     the system
 may provide the orportunity for including consideration
                                                           of
 the supplier's own assessment of a residual value
                                                    in that
 price.

 We appreciate th- opportunity to comment upon this
                                                    matter.
 Sincerely,




 GO-kcttln A!r
           PSr('atin,   --
                 ~ Dzt~CtQ;    EST           AV
                                     DOCUMEN'TAVAIrBtE



                               42
                                                        APPENDIX III
                                                              Page 1

                         UNITED STATES OF AMERICA
                    GENERAL SERVICES ADMINISTRATION
                             WASHINGTON. D.C.   20405




      JUN 9 1970

Honorable Elmer B. Staats
Comptroller General of the
United States
General Accounting Office
Washington, DC 20548

Dear Mr. Staats:

This is in reply to your letter of March 27, 1970, which asks us for
comments on your draft report relating to the efforts of the Army Corps
of Engineers to develop an integrated information and data system.

We havejread your draft report with great interest. We are, of course,
in no position to comment on the factual findings. However, we have set
forth in the enclosure to this letter our observations which relate to
the policies against which your factual findings have been measured.

We hope that our comments will be of assistance to you. If you have any
questions in respect to our comments, please do not hesitate to contact
US.

Sincerely,



   ,obe0rt ':_.-I
  Aj:Aztr,;cr                              Drn
Enclosure                                  BEST DOCUMENT AVAILABLE




                                     43
  A.PPENDIX III
         Page 2
 Comments Relating to the General
                                   Accountin Officers Draft Report on
 the Efforts of the Army Corns of Engineers
                                            to Develop an Integrated
 Information and Data System

 I. We concur that under normal conditions
 possible, long range systems study              a thorough and, to the extent
                                       rather than short range expediency
 should be the basis for the acquisition
                                           of automatic data processing
 equipment (ADPE). This policy guidance
 Budget Circular A-54 and Bureau of            has existed in Bureau of the
                                        the Budget Bulletin 60-6 for many
 years. In our opinion, this guidance
                                         remains valid.
 2. The determination and use of residual
                                              value in respect to ADPE,
 although also provided for in BOB
                                     Circular A-54, is quite complex.
 reasonably broad base of historical                                      No
                                       data is available at present which
 would permit us to draw valid conclusions.
                                                We hesitate, therefore, to
 concur with your recommendation that
                                          under all conditions "in situations
 where dissimilar ADP equipment is offered
 agency will evaluate the differences          by manufacturers, the procuring
                                       in residual values so as to assure that
the equipment selected is the most
                                      economical for long term use. " We
have been and are continuing to explore
                                           the many facets of this area with
the Bureau of the Budget.

 3. While evaluation and/or selection
                                       criteria may, and indeed should
 vary from time to time, the application
                                           of the basic principles expressed
 in Bureau of the Budget Circular A-54
                                         (i.
 systems specifications and least overall e., capability to meet the
                                            cost) should be the guidelines
followed in the selection of ADPE.
                                     We believe that this principle also
is well founded.




                                44