REPORT TO THE PRESIDENT ON THE MANAGEMENT OF AUTOMATIC DATA PROCESSING IN THE FEDERAL GOVERNMENT

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP67B00446R000600050004-7
• Release Decision: RIFPUB
• Original Classification: K
• Document Page Count: 103
• Document Creation Date: December 19, 2016
• Document Release Date: August 30, 2005
• Sequence Number: 4
• Publication Date: March 4, 1965
• Content Type: OPEN

Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
BUREAU OF THE BUDGET
AND SUBMITTED BY
JOHN L. McCLELLAN, Chairman
COMMITTEE ON GOVERNMENT OPERATIONS
UN*ED STATES SENATE
89TH CONGRESS
1st Session
f DOCUMENT
1 No. 15
REPORT TO THE PRESIDENT
ON
THE MANAGEMENT OF AUTOMATIC DATA
PROCESSING IN THE FEDERAL GOVERNMENT
PREPARED BY THE
MARCH 4, 1965.-Ordered to be printed.
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
COMMITTEE ON GOVERNMENT OPERATIONS
JOHN L. MCCLELLAN, Arkansan, Chairman
HENRY M. JACKSON. Washington KARL E. MUNDT, South Dakota
SAM J. ERVIN, Ja., North Carolina CARL T. CURTIS, Nebraska
ERNEST ORUENINO, Alaska JACOB K. DAVITS, New York
EDMUND S. MUSKIE, Maine uILWARI) L. SIMPSON, Wyoming
ABRAHAM RIBICOFF, Connecticut
FRED R. HARRIS, Oklahoma
ROBERT F. KENNEDY, New York
LEE METCALF, Montana
JOSEPH M. MONTOYA, New Mexico
WALTER L. ROYNOLDs, Chief Cicrk and Staff Director
ANN M. ORtcKia, Assistant Chief Clerk
OLaNN K. SHRIV'eR, Professional Staff Member
ELI E. NOBLEMAN, Professional Staff Member
W. E. O'BarsN, Professional Staff Member
JAMER R. CALI.~.wAY.. Professional Stag Member
ARTHUR A. SHARP, Staff Editor
EXCERPT F ROSI TISE CONGRESSIONAL RECORD OF MARCH 4, 1965
Mr. MCCLELLAN. Mr. President, the President of the I?nited States
has submitted to the Congress a "Report to the President on the
Management. of Automatic Data Processing in the Federal Govern-
ment,' prepared by the Bureau of the Budget, which has been
referred to the Committee on Government Operations for consider-
ation. The retort contains a number of recommendations for legis-
lative action, along the general lines of a bill, H.R. 5171, which was
approved by the House of Representatives and referred to the com-
mittee during the 88th Congress.
In view of the interest that has been manifested in this report and
in the proposed legislat ion, I ask unanimous consent that the report
be printed as a Senate document for use by the Committee on Govern-
ment. Operations, and other committees of the Congress which have
an interest in ADP proc?urenlent, operations, and management.
The Vics PRESIDENT. Without objection, it is so ordered.
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
LETTER OF TRANSMITTAL
THE WIIITE HOUSE,
Washington, March 2,1965.
7~
The Honorable the PRESIDENT OF TILE SENATE.
The Honorable the SPEAKER OF THE HOUSE OF REPRESENTATIVES.
SIRS : The use of automatic data processing equipment during the
past 10 years has contributed significantly to increased effectiveness
and rising productivity in governmental operations. The electronic
computer has enabled the Government to carry out .programs which
otherwise would have been impossible. Better and more economical
services to the public have been achieved through the use of this
equipment.
Government policies with respect to the acquisition and use of auto-
matic data processing equipmenthave been a matter of interest to a
number of congressional committees. In 1963, in response to a con-
gressional request, President Kennedy directed the Bureau of the
Budget to undertake a comprehensive review of this subject and to
prepare a report to the Congress. This study is now complete. The
suggestions for improvement outlined in the enclosed report have my
approval.
I have requested the Director of the Bureau of the Budget to work
with the interested committees of the Congress and with the executive
agencies concerned to assure the most economical and effective use of
this highly important area of management.
The Bureau of the Budget at an early date will set forth, in a cir-
cular, specific Government-wide responsibilities of the Bureau of the
Budget, General Services Administration, Department of Commerce,
and Civil Service Commission, to carry out the recommendations con-
tamed in the report.
Sincerely,
LYNDON B. JOh NEON.
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
LETTER OF SUBMITTAL
COMMITTEE ON GOVERNMENT OPERATIONS,
U.S. SENATE,
March 4, 1965.
The Honorable IIuB;ERT 111. II1MPIIREY,
Vice President of the United States.
DEAR MR. PRESIDENT : I submit herewith a "Report to the President
on the Management of Automatic Data Processing in the Federal
Government,' prepared by the Bureau of the Budget, which has
been referred to the Committee on Government.0perations for con-
sideration. The report contains a number of recommendations for
legislative action, along the general. lines of a bill, H.R. 5171, which
was approved by the House of Representatives and referred to the
committee during the 88th Congress.
In view of the interest that has been manifested in this report and
in the proposed legislation, the Senate authorized the report to. be
printed as a Senate document for use by the Committee on Govern-
ment Operations, and other committees of the Congress which have
an interest in ADP procurement ~iild operations.
A resum6 of previous actions taken by the Committee on Government
Operations in the 88th and preceding Congresses is set forth in the
daily Congressional Record on September 21, 1964, at pages 21677
through 21684.
Sincerely,
JOHN L. MCCLELLAN, Cha-irnnan.
V
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
CONTENTS
Introduction------------------------------------------------------
Chapters:
Page
1
1. Differences among computer installations_____________________ 9
2. Determining the best use of computer capabilities --------------- 15
3. Meeting requirements for computer capacity ------------------ 19
4. Selecting the proper equipment for use----------------------- 23
5. Purchase or rental of computers----------------------------- 27
6. Contracting for the procurement of equipment----------------- 41
7. Standardization of equipment and techniques------------------ 47
8. Government-sponsored research and development in computer
sciences------------------------------------------------ 53
9. Government-contractor relationships------------------------- 57
10. Information for managing automatic data processing activities-- _ - 63
11. Organization and legislation--------------------------------- 67
Exhibits:
A. Letter to the President from Chairman Tom Murray, Committee
on Post Office and Civil Service, House of Representatives,
dated August 1, 1963------------------------------------- 77
B. Letter to Chairman Murray from the President, dated September
19,1963 ------------------------------------------------- 79
C. Letter to Chairman Murray from the Director, Bureau of the .
Budget, dated September 25, 1963__________________________ 81
D. Selected actions, by the legislative branch related to the manage-
ment of automatic data processing-------------------------- 83
E. R6sum4 of management activities within the executive branch_ _ _ _ 85
F. The manpower and personnel impact-------------------------- 91
G. Proposed change to Armed Services Procurement Regulation
15-205.34----------------------------------------------- 97
H. Bureau of the Budget Bulletin No. 64-8: "Study of the Manage-
ment of Automatic Data Processing in the Federal Govern-
ment," dated December 26, 1963--------------------------- 101
1. Members of the project staff_________________________________ 103
J. Non-Government organizations consulted --------------------- 105
K. Letter to the heads of executive departments and agencies, dated
March 6, 1965, from the Director of the Bureau of the Budget,
enclosing Circular No. A-71, related to the administration and
management of ADP______________________________________ 107
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
INTRODUCTION
In the short span of a decade, the electronic computer has had an .
unprecedented effect upon the conduct of Government activities. Use
of this equipment has enabled Government to carry out programs
never before possible, and it has facilitated the provision of services
more effectively and economically, thereby contributing significantly
to the rising productivity in governmental operations.
However, the use of the electronic computer has created a variety
of problems. Some of these have been successfully resolved; others
are still matters of concern. In order to aid in the solution of cur-
rent problems and those that may well arise in the future, a compre-
hensive study has been made of the management of electronic data
pprocessing activities in the Federal Government,. This report is based
largely upon that study.
BROAD USE OF COMPUTERS
No single technological advance in recent. years has contributed
more to effectiveness and efficiency in Government operations than
the development of electronic data processing equipment. Most of
the important advances that, have been made in such diverse fields as
space exploration, research of all types, weather forecasting, and
atomic energy would not have been possible without the computer.
In the field of large-scale clerical operations, such as insurance proc-
essing, checkwriting, and the tax system, the computer has materially
assisted in producing major economies. Furthermore, the computer
is becoming increasingly useful to managers in solving complex prob-
lems involving interrelated types of information. The most notable
of these have been in military areas and in supply management, but the
use of computers to support advanced management techniques is be-
coming common in a broad range of governmental activity. Based on
results achieved to date, this latter type of use of computers holds a
potential, of outstanding importance in the public service.
In addition, there is the largely untapped area of integrating re-
lated information systems that cross organizational lines.
Accordingly, it seems reasonable to assume that the impressive
advantages to the Government already achieved through automatic
data processing (ADP) are but steppingstones to the future.
SIZE OF ADP OPERATIONS
The Federal Government is currently making substantial expendi-
tures for the acquisition and utilization of electronic computers. In-
ventory statistics 1 show that in the fiscal year 1964 nearly $1 billion
1 See 1904 Inventory of Automatic Data Processing Equipment in the Federal Govern-
ment (July 1964). Superintendent of Documents.
Approved For Release 2005/11/21 : CIA-RDP67B00446R000600050004-7
 Ap$roved For Relea 6D 1/24rAQ1 R BOO446R000600050004-7
was expended for this purpose, and some 54,000 maii-years were uti-
lized in the operation of 1,767 computers. This number of computers
is estimated to represent about 10 percent of all the Computers in the
Nation. In addition, however, estimates indicate there are at least
an equal number of computers used in the. Government that are ex-
cluded from the inventory because they are employed primarily in
unique military applications. Further, (here is another group of
equal size operated by contractors who perform work for the Gov-
ernment, on a cost-reimbursement basis. The magnitude of the ex-
penditures itself provides sufficient reason for assuring that proper
attention is given not only to the effective use of computers, but. to
economy and efficiency as well.
The extremely rapid exploitation of the computer which has been
evidenced in the Federal Government has not been without problems
involving almost every aspect of acquisition and utilization. The fol-
lowing examples illustrate (lie problems :
The diversity of ADP equipment and its use under varying
circumstances has raised questions as to the appropriateness of
general policies and guidelines applied uniformly to all ADP
activities.
The great range of possible computer applications-that- is, the
data processing problems to which the computer may be applied-
makes it necessary to develop means for selecting those applica-
tions which offer (lie greatest return. Some applications produce
distinct advantages while others are marginal at best.
The tremendous effect of system design on the efficiency and
effectiveness of computer applications makes it desirable that
means be developed for assuring that techniques of Iiigh c ualit.y
system design are utilized. The problem is compounded by the
fact that designiing systems involves an evolving technology that
calls for a high degree of individual competence applied to unique
situations.
The selection of equipment requires extensive knowledge of the
use for which the equipment- is intended and of the performance
of available machines. In making the selection, procedures are
needed that are not overly expensive, lead to the right choice, and
encourage competit ion.
The high cost of com niters requires that, wherever feasible,
machines already available within the Government be utilized
in lieu of acquiring additional capacity. There are, however,
many obstacles some technical and some administrative---to the
optimum utilization of machine time.
Contracting for computers is made difficult. by the lack of stand-
ard characteristics of the equipment, the more or less intanible
nature of supporting services that are required, and the neegd for
timeliness in the issuance of contracts.
Decisions on rental versus purchase involve problems of pre-
dicting the economically useful life of equipment under condi-
tions of changin req~,uirements and technology as a basis for
deterpiiinirig whetter the costs associated with leasing will exceed
the costs associated with purchase.
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For Release A$QO 1A 1DA 'r1Ap6570446R000600050904-7
The disposal of excess and surplus machines creates problems
that will grow as the Government increases its equipment pur-
chases. The problems involve questions of timing, responsibility
for. choosing the best means by which agencies may accomplish
their missions, and economic obsolescence.
Maintenance of equipment is a relatively new problem. Rented
equipment is ordinarily maintained by the manufacturers. For
purchased equipment, policies must be developed for choosing
between commercial maintenance procedures and maintenance, by
Government employees.
The differences among electronic data processing equipment
make the transfer of data among machines and systems difficult
and expensive. As in most expanding technologies, there are
problems of providing the resources for, and of achieving, a
satisfactory degree of universally accepted standardization with-
out, inhibiting advances in the state of the art.
There is a need for coordinating research and disseminating
findings. This need steins from the fact that much of the growth
in the technology of equipment and of data systems for using it
has been brought about by the research and development work
carried out on. many fronts and by many different age=ncies that
are eager to use the computer as a. means toward accomplishing
their own missions. The problem of coordination, however, is
basically the same as that currently experienced in other areas of
research and development..
The way in which contractors, performing work for the Gov-
ernment, acquire and use computers is important because; there are
indications that. the amount of work done for the Government on
contractors' computers is large. Problems in assuring efficient
and economical use, particularly by contractors who are working
on a c=ost, reimbursement basis, Have arisen because of the need
for avoiding the usurpation by the Government of the manage-
ment responsibilities and authorities of the contractor and also for
avoiding procedures that would result in the Government's acquir-
ing expensive equipment that would rapidly become obsolete.
The effect of the computer on Federal employment has been felt
in two separate directions. On the one hand, there has arisen a
continuous and urgent need for personnel in all phases of com-
puter use., ranging from competent key punchers, programers, and
designers of systems to managers who understand the potentiali-
ties and limitations of the equipment. On the other hand, in
many applications, the computers have made large numbers of
skills obsolete; this has brought a concurrent need for enlight-
ened personnel policies and advance planning to alleviate unneces-
sary hardship without depriving the Government of attainable
economies. The Civil Service Commission has made a separate
study of this subject.2 Based on the findings of the study, the
Commission has embarked upon additional studies and action pro-
grans which will concentrate upon four in "'or problem areas:
(1) manning for computer operations, including consideration of
4 "A Study of the Impact of Automation on Federal Employees," a committee print
prepared by the U.S. Civil Service Commission and referred to the Subcommittee on Census
and Government Statistics of the Committee on Post Office and Civil Service,. dated
August 1904.
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Ad-proved For Releet ZOW UK: Q f,%7B00446R000600050004-7
skill shortages and technical training, (2) upgrading the compe-
tence of computer-related personnel, (3) the impact. of computer
automation on the job security of Government employees, andthe
need to forecast- and plan appropriately for this impact, and (4)
the need for keeping abreast of impending advances in the state
of the art of computer automation, in order to be able to forecast
the impact of these advances on personnel requirements in general
and computer-related technicians in particular.'
The assignment of appropriate roles to the different echelons of
management in the F+eceral Government is of great importance.
Some computer applications, particularly those involved in ad-
ministrative functions, have a great (lent in common and conceiv-
ably could he subject to greater centralization. On the other
hand, the more significant computer applications are integral
parts of agency programs; accordingly, each is a unique applica-
tion and its management is it responsibility of those officials
charged with mission accomplishment. The problem then be-
comes one of improving (lie effectiveness and the economy of com-
puter titilizat ion, both within an executive agency and in the Gov-
ernment as it whole, without derogating the proper authorities
and responsibilities of managers in the line.
Experience in the Government and in industry has demonstrated the
urgent need for management to concern itself intimately with ADP
activities. Because ADP is based on a new technology whose most
visible manifestat ion is equipment of an esoteric naature, there has been
a tendency to regard it its the particular responsibility of the tech-
nician. APP applications, from the simplest to the most complex,
raise problems and require decisions that are the direct responsibility
of managers themselves. APP may have a pronounced effect upon or-
ganizational arrangements, the kind of work employees perform, the
conditions under which they work, or the type of skills required.
Furthermore, APP can have have an effect- not only upon the means
and cost of providing service to the public, but also upon the kind of
service provided. APP applications that are primarily designed to
provide information can have it significant effect. upon the way a man-
ager makes decisions and upon the control lie exercises over his or-
ganization.
Accordingly, it is necessary that managers concern themselves with
all aspects of APP projects including the determination of objectives
sought, the decision to proceed, the design, installation and operation
of the system, and the utilization of the end product. This requires
that managers obtain it broad understanding of the potentialities and
limitations of ADP, and of the work of the skilled technicians on
whom they must rely, and take steps to assure that decisions that are
properly within the sphere of management are made by managers.
The earliest efforts to utilize the computer resulted largely from
local initiative and ingenuity. Significant. contributions to under-
staanding the potentials and problems in the use of this new technology
^ Bee exhibit F' for a more detatird discussion of the current and planned ADP-related
nctlvities of the Civil Service Commission.
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For Release /i 1,2tA, 1ApfQp %0446R00060005W04-7
were made by these pioneer installations. As the use of computers
began to increase, various organizations in the Federal Government
began developing policies governing the management of automatic
data processing in order to facilitate proper use of the equipment and
i:o assist in resolving the problems that arose.4
The Bureau of the Budget has developed., or sponsored the deyel-
opmeni, of, guidelines and policies relating ta-
(1) Studies that should be made In advance of the acquisition
of equipment;
(2) Selection and acquisition of equipment, with particular
reference to purchase-or-rental decisions;
(3) Agency practices in respect of ADP management; and
(4) Sharing of equipment.
In addition, the Bureau (1) publishes an annual inventory of equip-
ment in the Federal Government; (2) has initiated and led a project
to test the feasibility of sharing exchanges; (3) has published a glos-
sary to provide a set of terms that will be commonly understood, and
a directory of training opportunities; (4) has established a special
panel on standardization (consisting of representatives of the General
Services Administration, Bureau of Standards, and the Department of
Defense) to strengthen the Government's participation in the pro-
gram of the American Standards Association;. and (5) has created
the large Interagency Committee on Automatic Data Processing and
a small council to foster the exchange of experience and to obtain the
advice of experienced agencies. By the budget review process and by
providing individual consultation, the Bureau has exerted an influ-
ence over data processing activities of the Federal Government.
The General Services Administration has negotiated general sched-
ules with equipment suppliers for the acquisition of equipment and
has published regulations governing the disposition of excess and sur-
plus equipment. The General Services Administration currently is
extending the equipment-sharing concept across the country.
The Civil Service Commission has (1) provided classification and
qualification standards for positions related to the operation of com-
puters, (2) provided assistance to agencies in developing aptitude
tests, (3) given assistance to agencies faced with employee displace-
mnent problems, (4) sponsored and provided extensive training courses,
and (5) made special studies of the effect of the computer on Federal
employees.
The Bureau of Standards in the Department, of Commerce has con-
tributed by conducting experimental work on the design and operation
of computers, by providing assistance in designing computer systems
and selecting equipment, and by operating a Computer Service Cen-
ter/Sharing Exchange for the Washingtonarea.
Agencies themselves are developing staff resources at intermediate
and headquarters levels to manage computer activities and to provide
guidance in specific policies. A significant aspect of efforts by indi-
vidual organizations has been the technological advancement of system
design and the conduct or support of research.
These activities of the various agencies of Government have materi-
ally improved both effectiveness and efficiency in the use of computers.
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 A- proved For ReleW1 ;/?t1 :I~7B00446R000600050004-7
An increasing amount of information is being assembled and used;
average hours of machine use are increasing; more advantageous con-
tract terms are being negotiated; the proportion of purchased machines
has increased sharply, with commensurate savings in rental costs; a
significant beginning has been made on the problems of standardiza
tion ; and more of the scarce skills are being provided.
Nevertheless, the ever-expanding involvement. of the Federal Gov-
ornment with the computer-expressed in increasing expenditures,
numbers of computers, and personnel-and the effects of certain of the
practices followed by various agencies have caused concern in the Con-
gress, in the General Accounting Office, and within the executive
branch itself. The General Accounting Office published a number of
reports which focused attention on ADP policies and management
practices. The House Govermuent Operations Committee sponsored
legislation on the management of ADP. The House Committee on
Post Office and Civil Service held extensive hearings and, after pub-
lishingg a report that raised numerous questions. recommended to the
President that a study be made of ADP management. The President,
aware of the increasing significance of the computer and the problems
raised, directed that the Director of the Bureau of the Budget under-
take a com yreltensive stud of the management of ADP activities of the
executive branch and make recommendations for such administrative
or legislative actions as may be appropriate.?
On December 26, 1961, the Director of the Bureau of the Budget.
announced the beginning of the study and the creation of an Advisory
Committee composed of former Congressman Robert Ranispeek as
Chairman; :fir. Bernard L. Boutin, Administrator of General Serv-
ices; Mr. Manuel R. Cueto, vice president. in charge of electronic plan-
ning and development, New York Life. Insurance Co.; Mr. Walter F.
Frese, professor of business administration, Harvard University; Mr.
Martin Gainsbrugh, vice president, National Industrial Conference
Board; Mr. J. Herbert. lIollomon, Assistant Secretary of Commence
for Science and Technology ; Mr. Dwight A. Ink; Assistant. General
Manager, ktouiie Energy Commission; Mr. Frederick I. Lawton '
former Director of the Bureau of the Budget and former Civil Service
Commissioner; Mr. John W. Macy, Jr., Chairman of the Civil Service
Commission; Mr. Thomas D. Morris, Assistance Secretary of Defense
(Installations and Logistics) ; Mr. Martin Shubik, professor of
economics, Yale University; and Mr. David Z. Robinson, Office of
Science and Technology. Staff work for the study was under the
direction of Mr. Carl W. Clewlow. Assisting him were persons
selected from Government agencies. During the study, Members of
Congre and representatives of Federal agencies and of industry,
labor, and professional groups overec_onsulted "
This report is based upon the results of the study. The recommend-
ations made outline actions that should be undertaken to bring about
the successful resolution of current and future problems occasioned
by the use of the computer in the Federal Government. These actions
are summarized below:
6 See exb1htts A. B. and C for an exchange of correspondence which led to this study.
See also exhibit n for a list of actions taken by the leglsintive branch.
c See exhibit 11 for the announcement of the study ; exhibit I for the protect Staff mem-
hers; exhibit J for the non-government organizations consulted during the study.
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For Release p 122liAIA p 10446R00060005QO04-7
1. Modify existing Government-wide policies so that their precise
application in different kinds of operating situations is more closely
defined.
2. Develop and furnish criteria to assist agencies in evaluating
whether computers are being used effectively.
3. Develop and furnish cost princi les to be applied uniformly by
agencies when computers and related services are shared with others
on ft reimbursable basis.
4. Expand existing policies for the selection of equipment to pro-
vide additional guidelines on (a) the preparation of systems specifica-
tions which are transmitted to suppliers when inviting proposals to
furnish equipment, and (b) methods for evaluating suppliers' pro-
posals.
5. Continue present, policies governing the purchase or rental of
computers, except (a) to include the cost: of money as a factor in com-
paring alternative costs, and (b) provide for a general suspension of
purchase activity if a review of computer technology indicates that
superior equipment will soon be available, or if prospective excesses
of Government-owned equipment indicate that additional purchases
should not be made. As a consequence of increased purchasing in re-
cent years, policies governing the replacement of equipment to avoid
unwarranted long-term use, and the use of alternative ways for main-
taining owned equipment will be formulated.
6. Establish a firm time schedule for the negotiation of annual con-
tracts with equipment suppliers, and seek improved contract terms.
7. Strengthen Government support of programs initiated by the
American Standards Association to achieve needed compatibility
among automatic data processing equipment and systems.
8. Give increased attention. to the coordination and evaluation of re-
search and development programs in the field of computer sciences.
Expand the resources of the National Bureau of Standards to advance
the development of computer technology and systems oriented pri-
rnarily toward Government needs.
9. Extend Government policies on the purchase or rental of equip-
ment and on the use of excess equipment to contractors who perform
work for the Government (primarily Defense, Atomic Energy Com-
mission, and the National Aeronautics and Space Administration) on
ya cost-reimbursement basis. Include contractor-operated equipment
in intra-agency sharing arrangements.
10. Develop and prescribe a Government, wide information system
to provide selected managerial levels with information needed to man.-
age computer resources more effectively.
11. Continue present organizational arrangements and general as-
signments of responsibility among central and line agencies, but
kstrengthen and augment the resources devoted to the management of
automatic data processing activities.
12. Propose the enactment of legislation by the Congress which
would (a) constitute an expression of congressional policy and interest
with respect to effective and economical use of automatic data process-
ing equrpment, and (b) strengthen the authorities for the development,
testing, and implementation of standards; the performance of research
in commuter sciences and the provision of advisory services by the Na-
t tional.l3nreau of Standards; and the establishment of a revolving fund
o o finance arrangements for the joint utilization of computer facilities.
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
MANAGEMENT OF AUTOMATIC DATA PROCESSING IN THE
FEDERAL GOVERNMENT
CHAPTER 1
DIFFERENCES AMONG COMPUTER INSTALLATIONS
Failure to give adequate recognition to the differences that exist
among computer installations has complicated the problems of man-
agement and contributed to misunderstanding. These differences are
due to variations in the missions of agencies using the installations, in
operating objectives, and in operating requirements surrounding the
use of computers. Also, installations vary in respect of the qualifica-
tions required of their staffs and the kinds of equipment that are used.
The purposes for which computers are used cover the breadth of
governmental activity. They are used to-
(1) Carry out missions, such as the administration of veterans'
benefits and social security benefits, management of natural re-
sources, regulatory functions, weather forecasting, control of air
traffic, and tracking of missiles.
(2) Perform such control functions as scheduling of produc-
tion, and review and evaluation of agency missions.
(3) Conduct research and development activities in all fields
of science, and perform special operations.
(4) Perform common administrative-type functions such as
those related to personnel, payroll, and accounting.
The kinds of computers used for these purposes vary in design.
They range from small desk-type computers costing in the neighbor-
hood of $50,000 to large-scale computers costing $3 to $6 million.
The availability of the computer to the user also varies. Some are
placed on-site, to facilitate their use by scientific or engineering per-
sonnel; others are operated as service centers, to meet the needs of
many users on a scheduled basis; still others are placed organization-
ally, to service a specific major program function.
Furthermore, the speed with which the computer must respond to
the data-processing requirement varies widely. In missiles tracking
and air traffic control, the response must be instantaneous (often
referred to as "real time") ; for other requirements, the response can be
controlled on a scheduled basis according to the priority associated
with the requirement.
Policies and guidelines on the use of computers are at present writ-
ten in very broad terms so that they may be applied to a wide variety
of situations. While this has allowed differing agency requirements
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For Rele 1 5 j1 t; j : R*-7B00446R000600050004-7
to be accommodated within the policies and guidelines, it lilts led to
misunderstanding and undue variations in agency practices. For ex-
ample, misunderstanding arises with respect to the policy on sharingg,
which is intended to increase the use of existing computers. Although
this objective is desirable, the policy does not give sufficient. recogni-
tion to the fact that certain types of computer installations (for ex-
ample, those which must be available for nis(auitaneous but. intermit-
tent response) may not be susceptible to sharing, even though they
may Rave lowv utilization. Failure to give ade.quate recognition to this
fact tends to brand such installations, by implication, as inefficiently
utilized, and unnecessarily involves them in the sharing program.
As another example, there is a general policy requiring that de-
tailed systems specifications be developed as a basis for selecting proper
equipment ; drat is, the. data-processing requirements must be set. forth
in precise terms in order to deterrniue the kind of equipment. which is
best able to meet those requirements. This policy does not, however,
recognize the fact that the extent to which system specifications can
be set forth in advance de ends largely a roil the character of the oper-
ations to be performed. In research and development types of oper-
ations or in service-center operations, it is often exceedingly difficult
to state the requirements with exactness.
The extensive use of computers for a wide variety of purposes sug-
gests the need for some classification of these computers in order to
provide a reasonable basis for making appropriate distinctions in ap-
plying policies. This classification should reflect the different pur-
poses for which computers are used, the different kinds of computers
used, and the operating requirements surrounding their use. In at-
tempting to devise some basis for classification, we have considered
the following: (1) types of equipment, (2) costs of equipment, and
(3) funct ions and environment.
Types of equipment.-Ilist orically, there has been a tendency to
distinguish between computers used 'for processing business data and
those used for scientific purposes. While this distinction was reason-
ably clear cut in (lie early days of computer use, it no longer has any
real validity, because changes in computer technology have, enabled the
same computer to serve both par apses with erluaf facility. Conse-
quently, this distinction is having less and less significance. It is not,
therefore, a distinction that should be given serious consideration in
the determination of policies; nor are there other characteristics of
design which scene to be ap)prolpriate for this purJpose.
('oxt of equipment.-Thus distinct ion was quickly ruled out as a clas-
sification device to govern the application of policies and guidelines,
except to the extent ilia( consideration of cost ranges would be useful
in establishing management review criteria. As a general rule, for
example, (lye process of review and approval for computers costing
$1 million nay well be =arried to higher organizational levels than
is the process for computers cost ing 71.i,O00.
I**'rrnetiottxand epp.rrtotiiiieiit.-t?urstudy indicates that themost logi-
cal distinct ion that should he made is one (lint recoguiizes the differences
in the environment in which (lie computer is opci.ut ing, including the
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For Release gj /A1 /o2'DiQIfi~-PV 0446R00060005PP04-7
response which the computer is required to make. Because the com-
pater is so closely linked to the performance of an activity, these en-
vironmental differences largely control the extent to which manage-
ment actions and criteria can be applied uniformly. in achieving maxi-
mum practical use of the equipment.
The preliminary analysis made during our study suggests that a clas-
sification along the lines depicted in illustration No. 1 could serve
as the basis for the distinctions that are necessary. The major head-
ings across the top of the illustration represent categories of environ-
ments; the stub column at the left of the chart indicates broad classes
of time response that must be met by the computer system. An ex-
planation of these environment categories and response classes follows.
General examples are identified within the matrix on the chart.
ERVlro mlent.
Category
I
II
III
IV
V
VI
Professional
Central
Integrated
Real Time
Research and
Special
Support
Computing
Operations
Operations
Development
Operations
Response
Services
Class
-,-
---- -
A. Priority
Remote console
? High Response
. Command
time shared
Inventory
& Control
systems
Control
Communications
Include:;
High Response
. Network
Time Sharing
. Military
Informations
Control
Computations
Systems
Retrieval
Airline
Engineering
Involving:
Reservations
Design
? Guidance i
--
---- -
-
M
difi
ti
computers
B. Time
-? -
Experimental
---
Engineering
Off-line
Dynamic
on
o
ca
Interfacing
Range
Finding
Critical
k
& Scientific
Inventory
Simulation
for;
Co
ut
tions
Development
Data
Control &
mp
a
Projects
Processing
Information
Missile
Communications
&
Digital
Retrieval
Checkout
Display
other
'Single
Simulation
Weather
Control
Purpose"
Predictions
Data
Uses with
Data
Reduction
Equipment
C. Scheduled
Engineering
Reduction
Management
Range
In
sed
e
Designed &
&
Scientific
&
Safety
cr
a
built to
Design
Computations
Business
Process
processing
meet
Computations
Business &
Systems
Control
capacity
special
Management
capability
conditions
Data
Processin
1. Environment categories
(a) Professional support.-In this environment, small desk types
of computers are used, generally on-site, for experimental development
and scientific computations to provide support to professional engi-
neering or research groups. Utilization of this type of equipment is
closely associated with the working hours of the professional staff.
(b) Central computing services.-In this environment, a computer
installation provides support for many users. Applications are gen-
erally diverse and often independent of one another. The workload
tends toward mathematical or statistical. processing, although these
may be intermixed with administrative or management applications.
The installation is normally operated by a staff of specialists, but
the computer programing is frequently done by personnel assigned to
organizations other than the central computer installation. However,
the customer is usually not present while the work is being processed.
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 ARproved For Rele A
~Y p~ 1djj 9 -& 9 B00446R000600050004-7
The equipment used can be in the small, medium, or lag a-scale class.
Medium- or large-scale equipment is often supplemented by small-scale
peripheral computers.
(c) Integrated operations.-In this environment also, a computer
installation provides support fora number of users; but generally one
application, which consists of a series of interde(iendent processes,
constitutes the largest pint of the workload, In this case, a variety
of inputs, usually from different organizations or geographic sources,
are processed against it system of group of data files to provide infor
mation to support it nnniiier of group of data files to provide informa-
tion to support it number of related functions. Applications are cyclic
and recurring, with large volumes of data being processed. They tend
more toward administrative or management functions, although
matlieniaticat statistical, Or operating programs may be involved or
intermixed. The computer installation is operated by a staff of spe-
cialists, with the systems development and programing usually clone
by a central computer staff working closely with customer organiza-
tions. The computin equipment is generally of (lie medium- or large-
scale class, but it is often supplemented by smaller computers. Equi -
nient to permit direct inquiries from off-site locations may also be
associated wit It these integrated operat ions.
(d) Real-t ne operations.-In this enviromment, the computer is a
part of a large equipment complex which requires continuous "imme-
diate" response. Since the real-time environment generally requires
uninterrupted support-, the equipment is normally "backed up" with
another computer, or sufficient redundancy is included to insure con-
tinuous operation. ('omniunications and display equipment are often
incorporated in the equipment complex. Computer equipment usually
is large-scale, frequently supported by additional small- or mcdiuni-
scale peripheral computers.
(e) Research- and development.--This environment usually involves
the adaptation of conmercial off-the-shelf or special-purpose equip-
ment, in order to experinienti develop, or test new or improved ways in
which computers can be utilized for particular applications. During
the process, the equipment is often extensively modified and cluinged,
and the development of new computer programing techniques or con-
cepts is required.
(f) Special operations.-In this environment, general-purpose
equipment may be used for unique purposes, or specialty constructed
equipment or niodifted commercial equipment may be used to meet un-
usual and specific requirements. In the latter instance, the equipment
must, frequently meet, extremely rigid specifications with respect to
size, reliability, or conditions within which it operates.
2. Time-re8po ,se classes
Within each of the environment categories, three general time-
response classes can be dist inguislied.
(a) Priority.--This type of response involves the use of communi-
cations capability; that. is, the computer equipment must be in a con-
tinuous state of readiness to receive, process, and distribute informa-
tion, or be capable of interrupting the ongoing process to respond to
external demands.
(b) Time critical.-In some cases, the project or program being sup-
ported by a computer requires a very rapid response to provide the in-
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For Release I$IglyAg A ? 446R000600054904-7
formation before it loses its value or before the next course of action
can take place. While these computers do-not operate in real time,
the time-response requirements nevertheless place constraints on the
manner in which the equipment is utilized.
(c) Scheduled.-A large part of the computer workload is usually
processed on the basis of scheduled start of completion times. While
scheduled applications may have some degree op flexibility, and may
be postponed depending on external factors, the schedule becomes
fixed at some point and must be met.
Although a classification system of this kind will be helpful in deter-
mining policies, it must be noted that not all computer installations
will fit neatly into one of the classifications. There will be many cases
where a single computer installation functions within several of the
environment categories and is required to meet, varying time-response
demands. In these instances, subjective judgments based on those
factors which seem most predominant and overriding will often be
required in the application of policies and guidelines.
In the development and application of policies, guidelines, and
criteria, the Bureau of the Budget will use a classification system
which recognizes the essential differences among computer installa-
tions. The pattern of classification suggested by the analysis made
during this study will serve as the basis for developing this system.
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
DETERMINING THE BEST USE OF COMPUTER
CAPABILITIES
The uses now made of computers are in most cases justifiable; but
marginal and, at times, uneconomical uses are sufficiently prevalent
to cause concern. The computer-based. systems that are most efficient
and most responsive to the information requirements of those who
use their output are the systems that take optimum advantage of
advanced techniques in planning the procedures and resources needed
for complete operations; that is, the most advanced "system design"
techniques. In many of the systems now in operation, however,
greater advancement in design is needed. Attention to the two prob-
lems-marginal uses and advanced system. design techniques-should
be given high priority.
Ordinarily, the greatest advantage for the Government is derived
when a computer is devoted to accomplishing the missions of an
agency, as opposed to routine administrative tasks. Examples are
numerous : they include the use of computers in administering vet-
erans' benefits, social security benefits, and the tax system; in tracking
satellites; in performing theoretical investigations for scientific pro-
grams; in predicting demands for electric power; in weather fore-
casting; and in controlling air traffic. There are advantages in using
computers for administrative tasks such as payroll and personnel
accounting. However, since administrative functions normally are
a relatively small part of total operations, the potential benefits to
be derived from these applications are limited.
Comparisons in respect of costs and benefits of proposed new
computer-based systems and the existing systems are fundamental
requirements in determining the most effective use of computers.
Manifestly, benefits must outweigh costs. Benefits may be expressed
in tangible terms, such as reduced operating costs, or in intangible
.terms, such as improved service to the public, the accomplishment of
missions not otherwise feasible, or better management practices.
Unless benefits such as these are assured, the use of a computer would
be classified as marginal.
The availability of unused capacity is an invitation to find addi-
tional uses for the computer. This search is motivated by the tendency
to regard unused time as evidence of inefficient management. While
undoubtedly the intention is to improve utilization, the effect often is
merely to increase the running time of equipment, with uneconomical
applications tending to lessen the true effectiveness of the use of com-
puters. After such marginal applications have found their way onto
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 P Q roved For ReleegmRD 1 tQA: SJA 'B00446R000600050004-7
a computer, it becomes difficult to eliminate them, even though their
presence may later require the procurement of additional equipment.
Management officials in agencies should make decisions on the uses
to be made of computers, rather than leave these determinations to
specialists in equipment and in system design. As proponents of the
use of computers j specialists can be most effective; their technical advice
is an essential factor in decision making. However, those who are
responsible for mission accomplishment should be irepared to make
the final decisions. Involving line management officials in this way
will aid in insuring that objectives for the use. of electronic data
processing equipment in mission accomplishment are carefully' estab-
lished and cleary understood; that adequate resources are made avail-
able to insure that the objectives will be achieved; that proper recog-
nition is accorded to the magnitude and complexity of the task; that
full cooperation and support of the total organization is obtained; and
that marginal uses are minimized.
Fittally, it is important that the use of existing computers be
evaluated With two purposes in view. (1) to determine whether the
benefits anticipated in the cost/benefit analysis made prior to acquisi-
tion are being realized, and (2) to redetermine the relative priority
of each computer application to insure that, with the passage of time
and in view of new potential applic'aatlolls , any application gait is con
sidered marginal is removed.
General guidelines stressing the value and purpose of such evalua-
tions are given in Bureau of the Budget. Circular A-61: "Guidelines
for Appraising Agency Practices in the Management of Automatic
Data Processing I~quipnnent in Federal Agencies," issued in August
1963. Tine essential quest ions to be aiiswcred inn titccourse of theevalua-
tions are as follows:
(1) '\Vliat advantages have resulted from the computer system?
(2) What contribution have these advantages made to the ef-
fectiveness with which the arency`s function is being performed?
This question is crucial in testing the ultimate worth of the ad-
vantages of the system, part icularly in those cases where increased
costs of data processing ittaay' be, involved.
(3) What are the resulting costs of the system? Are they
justified in terms of the benefits being obtained?
(4) How efficiently is the system being o aerated? The fact
that substantial benefits are being obtained cues not necessarily
mean that they are being obtained in the most economical
manner.
Although formal programs of evaluation have been undertaken by
some agencies (notably by Department of Defense), it is essential
that renewed and continuing emphasis be given to this ]chase of man-
agement, review activity. There is, however, at general lack of criteria
which would be useful in measuring performance in such areas as
operation and maintenance of equipment, efficiency of programing,
and costs for common applications. The availability of basic data
used daring the evaaluat ion should, to the extent feasible, become a part
of the ADP management inforaitaation system, which is discussed in
chapter 10 of this report.
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For Release 2005/11/21 : CI R 00446R0006000 004-7
AUTOMATIC DATA rtog
PROBLEM OF ADVANCED SYSTEM DESIGN
The development of computer-based systems can often be facilitated
.if there is an adequate frame of reference, so that any given system or
subsystem can be viewed in terms of its relationship to the total struc-
ture of systems in an agency. This is particularly true in the business
type or program type of functional areas; it may be less feasible if
novel, experimental, scientific research, or pioneering applications are
involved. Therefore, a master systems plan for an agency, at the
highest feasible level, is desirable. All efforts to develop systems can
then be undertaken in relation to the plan, in order to achieve an order-
ly and coordinated program. Such a plan, of course, needs to be re-
viewed periodically for adjustment. Leadership and coordination
of a program of this kind requires the provision of adequate staff
resources at agency and intermediate levels.
System design, and its importance as a primary consideration in
effective utilization of equipment, has been given considerable emphasis
by the Bureau of the Budget in policies and. guidelines published since
1960.1 There are three basic reasons for this :
(1) System design serves as a basis for determining whether or
not computer capability is, in fact, necessary. Cases have been
reported where a detailed analysis of an existing system resulted
in improvements to the point where the computer could make no
further significant contribution.
(2) It governs the proper development of specifications for sys-
tems. They, in turn, are highly significant factors in the selection
of the most appropriate equipment for the task at hand.
(3) It aids in insuring that the most efficient and economical
means are employed in achieving the objectives, of the system.
Several concepts and techniques (made possible by the computer it-
self) can often make a significant contribution to the advancement
of system design. Included are the integration into a single master
system of several functions using common basic data; the use of mathe-
matical techniques; the use of "management by exception" techniques
in which only usual circumstances are selected out of the routine process
for human consideration; and the coordination of separately operated
computer systems to assure machine-to-machine communication where
an exchange of data. is involved.
There is an awareness among agencies of the importance of system
design in assuring the best use of computers. It is no idle claim to say
that some of the Government computer systems are outstanding
examples of efficiency and of the employment of advanced techniques.
Efforts to make such examples more widespread are being deterred in
two ways : one of these deterrents stems from the lack of adequately
trained personnel. While this is a. general problem, its effects are felt
most keenly in the smaller agencies, which are only now beginning to
consider the use of computers in their operations and which do not
have a nucleus of trained personnel to draw upon. Also, because these
agencies are not certain that computers will prove advantageous in
Bureau of the Budget Bulletin 60-6: "Studies Preceding the Acquisition of Automatic
Data Processing Equipment" (Mar. 18, 1960) ; Circular A-54: "Policies on the Selection
and Acquisition of Automatic Data Processing Equipment in the. Executive Branch"
(Oct. 14, 1900) ; Circular A-61: "Guidelines for Apppraising Agency Practices in the
Management of Automatic Data Processing Equipment in Federal Agencies" (Aug. 3, 1063).
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
AUTOMATIC DATA PROCESSING
their operations, the investment necessary to build a staff simply for
the purpose of finding out may be prohibitive.
This problem has been met in part. by the National Bureau of
Standards, Department of Commerce, which has upon request pro-
vided assistance to agencies, to the extent possible. The demands for
assistance have far exceeded resources to service them. Other agen-
cies have at tunes cooperated by lending t rained personnel-often at
a, sacrifice. to their own programs. In other cases, commercial Sources
have been used to carry out studies.
There is a. need for a pool of trained personnel who can be called
upon, as needed, to help agencies-particularly smaller ones-in solv-
ing problems related to computer systems.
The other deterrent is the general lack of criteria for use by agencies
in evaluating systems design. The absence of such criteria creates a
void in the process by which management assures itself that appro-
priate methods are being employed and that a quality product is being
obtained within a reasonable period of time at it reasonable cost. Cri-
teria to provide assistance to management in this respect should be
developed and furnished.
Greater emphasis is needed on agency evaluations of their computer-
based systems. Criteria need to be developed to assist in measuring
performance in the operation and maintenance of equipment. 'leas-_
ores of efficiency in computer programing are also needed. Develop-
ment of these criteria calls for the assembling of information not cur-
rently available.
The needs of agencies for assistance in the analysis and design of
computer-based systems must be met. Because the Nat ional Bureau of
Standards has performed this role in the past, to the extent its resources
permnified, and because of its established competence in this field, it
seems logical that its resources should he expanded to meet more fully
the needs of agencies for assistance in systems work.
/1. The Bureau of the Budget will develop a broadly based program
of cont inunns evaluat ion of com intt er s} scents, to provide an assessment
of accomplishments and to serve as a recurring source of information
for the development or revision of policies and guidelines. The re-
ponsibil ity for conduct lag evaluations and preparing appropriate re-
ports will rest -,with the azency leads, in accordance with their normal
management responsibilities.
2. The Bureau of the Budget will develop criteria to assist in evalu-
ating both systems desitrn and various aspects of system performance.
3. Agencies should develop master data-processing plans at appro-
priate levels, to serve as -irides in the orderly development of systems
and to assure the most effective use of staff resources available for that
development.
4. The Department of C omnierce, through the National Bureau of
Standards, should expand the advisory services currently being pro.
vided t o agencies in the analysis and design of computer-based systems.
Its resources allocated for this inurpnse should be increased to the extent
required to meet such needs as hilly as possible.
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
CHAPTER 3
MEETING REQUIREMENTS FOR COMPUTER CAPACITY
PUTTING UNUSED COMPUTER CAPACITY TO WORK
It is the Government's policy 1 that agencies will seek to satisfy
computer needs by utilizing unused capacities, and that agencies having
unused capacities will offer them to those who need it.
There are distinct economic advantages to be gained by utilizing
unused capacity on computers already. installed by the Government.
If the capacity is available on purchased machines, it can be used
with very little additional out-of-pocket cost. If it is available on
leased machines, it often can be used at the lower rates charged for
extra use; and if the available capacity is within the total number of
hours a month for which a minimum lease charge is paid, the cost for
additional use of the equipment is negligible.
The potentials inherent in the practice of one computer installation
sharing its unused capacity are evident from a review of Government-
wide statistics. The latest annual reports submitted by agencies to the
Bureau of the Budget indicate that during a given 3-month period
about 500,000 hours of computer capacity were not being used among
1,400 computers.
Realistically, however, only a fraction of this capacity can be con-
sidered for use by others who may have computer needs. Over 150,-
000 hours, or almost one-third of the unused hours, are reported as
not available for sharing because of security restrictions, unusual
workload contingencies, or special equipment modifications. Of the
remaining 350,000 unused hours, 140,000 are on relatively small com-
puters, usually located on-site to be available to scientists and engi-
neers for immediate access when needed for computational problems.
Another 40,000 hours are associated with systems requiring instant and
uninterrupted response and are not suitable for extensive sharing.
Thus only about 170,000 hours, or about one-third of the total, can
reasonably be considered available for sharing purposes. While this
still represents a large potential, there are certain factors including the
following that will prevent full realization of this potential: (1)
virtually all unused capacity is available only on weekends or on
third and second shifts, and this creates administrative difficulties;
(2) the computer may not be available at the time the requirement has
to be met; and (3) geographical distances or locations may preclude
satisfactory arrangements.
Nevertheless, it has been demonstrated that sharing can be facili-
tated by an arrangement whereby computing needs and availabili-
ties are recorded and matched at a central point. Last year, the
i Bureau of the Budget Circular A-27: "Policies and Responsibilities on the Sharing of
Electronic Computer Time and Services in the Executive Branch" (June 15h 1864).
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Amroved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
AUTOMATIC DATA PROCESSING
Bureau of the Budget, with cooperation of other agencies, completed
a computer sharing exchange experiment in the Philadelphia area.
Arrangements made through this exchange resulted in a fivefold in-
crease in sharing. This plan is now being extended to other metro-
politan areas by the General Services Administration in accordance
With responsibilities outlined in Bureau of the Budget Circular A-27.
As noted in chapter 1(1 of this report, the development of an ADP
management information system will give specific attention to the in-
formation needs of the sharing program to assure that it can be
effective in satisfying the needs for computer time. It is essential
that this response also he made with respect to supporting services
(e.g., systems design, key punching, programing).
The Bureau of the Budget has also broadened the sharing concept
by arranging for the National Bureau of Standards to serve as a
computer service center for the Washington area-also on an experi-
mental basis to test its value. Under this arrangement, agencies
needing computer services can go directly to the service center to
have them performed. .1 sharing exchange, operated with the service
center, assists in arranging for these services to be performed else-
where if the service center, for some reason, cannot meet the require
ment. The continuation or extension of the service center conce pt. will
depend upon the results of the experiment, which is scheduled for
completion in 1965.
A problem involved in sharing arrangements is the matter of prices
charged for services. There are at present no guidelines to govern this
matter. As a consequence, there may be considerable variation in
the prices charged by installations using the same kind of equiiilnlent-
caused primarily by the inclusion of different cost factors in the price.
The result, of course, is to overload (lie installation having the cheaper
price, while the other is unused.
To facilitate an equitable distribution of the sharing workload and
to provide the customer some means for better estimating what his
costs will be, early attention should be given to establishing cost prin-
ciples which would be applied uniformly by agencies when develop-
ingschedules of rates and charges for the use of ADP equipment and
services.
According to the last annual report submitted by agencies to the
Bureau of the Budget-, agencies spent 55,800,000 in the fiscal year
1964 to acquire automatic data processing services from contractor
organzations. Of this nmount, $37,500,000 was spent- for systems
development activities, and $18,300,000 was spent for machine time
and related services, despite the fact that considerable unused capa-
city existed in Government installations. This latter fact reinforces
the need for an effective sharing program. This is not to say, how-
ever, that the use of commercia organizations was not fully Justified
under the circustances. The existence of this practice simply adds
another dimension to the potentials of the sharing concept.
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For ReleasBL&ljd2dAI1N00446R000600091004-7
Contractor organizations are used to provide a wide variety of serv-
ices. They include :
1. Providing computer time, usually with operator service.
2. Preparing and converting data for machine processing.
3. Designing a system..
4. Programing a, system.
5. Evaluating equipment and recommending selections.
6. Operating a computer installation in a Government facility.
7. Designing and developing special equipment.
8. Maintaining equipment.
9. Training.
Types.of organizations that provide these services include nonprofit
organizations, research institutions, educational institutions, con-
sulting firms, service organizations specializing in equipment opera-
tion or in systems design and programing, and equipment manu-
facturers who, in addition, perform these related services.
The use of contractor organizations usually occurs when peak work-
loads arise and it is to the agencies' advantage to secure outside assist-
ance instead of augmenting its own work force for a, short period
of time; or when the talents needed for a task are not. available within
the Government either as to kind, quantity, or time needed (this
occurs particularly with respect to systems development and program-
ing). However, several problems have arisen in the use of contractor
organizations:
1. Techniques, processes, or products produced by a contractor for
one agency might often meet the needs of another agency that is con-
tracting for the same thing. Some effective means for exchange of
information is needed. This problem also extends to activities within
Government.
2. Potential conflicts are present when manufacturers of equipment
contract to develop systems and then enter the bidding to furnish the
equipment. The establishment of criteria by which the objectivity
of performance could be evaluated would remove doubts in this area.
USE OF EXCESS AND SURPLUS COMPUTER CAPACITY
When an agency has determined that its continuing requirements for
computer capacity cannot be met by its own computer or by sharing
computers operated by other agencies, Government policy 2 requires
that agencies first consider the use of excess or surplus equipment
already available within the Government before considering the ac-
quisition of new equipment. This policy requires that consideration
be given not only to exc
Government ess or surplus equipment that is owned by the
, but also to that which is leased and is about to be returned
to the manufacturer. In this latter case, agencies that can benefit by
the use of this equipment can ordinarily purchase it at substantial
reductions from the original list prices.
a General Services Administration Personal Property Management Regulation No. ae:
"Utilization Screening of Government-owned and Leased Electronic Data Processing
Equipment" (April 1064).
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 ApZoved For Relea, ,ggAqq1/,&1rAC - 00446R000600050004-7
CURRENT E31PIIASIS ON MORE EFFECTIVE USE OF EXISTING COMPUTER
CAPACrry
The Bureau of the Budget has taken steps designed to emphasize
current policies on the use of computer capacity and to encourage the
fullest. measure of implenientat ion. In July 1964, the Bureau requested
the heads of all agencies that were scheduled to increase their computer
capacity through new acquisitions during the fiscal year 1965 to re-
tudy their needs, to consider again the possibilities of substituting
unused capacity of excess equipment for new additional Capacity, and
to report their findings to the Bureau for consideration in making
apportionments of funds. Similar reports have now been incorporated
in our plans for budge( hearings in the future.
1. The Bureau of the Budget will establish an interagency group
to study and develop cost principles to be applied uniformly by agen-
cies in establishing prices for shared computer time and services.
2. The Bureau of the Budget ~v111 continue its evaluation of the
service center concept to determine a proper course of action to be
taken.
3. The Bureau of the Budget will, with the assistance of the major
agencies concerned, undertake a. study of the problems associated with
the use of contractor organizations for providing services related to
electronic data processing activities, with a view toward developin
policies, guidelines, or actions that the study may indicate are needed.
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
SELECTING THE PROPER EQUIPMENT FOR USE
The selection of equipment which meets adequately the current and
projected data processing requirements of an agency is a complex
task, owing largely to the diversity of equipment available. Specify-
ing as fully as necessary the requirements for data processing, as a
precedent to decisions on selection, is a. painstaking and time-consum-
ing task.
CURRENT POLICIES ON SELECTION OF EQUIPMENT
Policies and guidelines governing the selection of equipment to. be
acquired from manufacturers are set forth by the Bureau of the
Budget as follows: I
1. Decisions will be based on system specifications.
2. The selection process must accord equal opportunity and appro-
priate consideration to all manufacturers who offer equipment capable
of meeting the system specifications.
3. Two prime factors are to be considered: (a) the capability of
the equipment for fulfilling the system requirements and (b) the cost
of the equipment and costs associated with its installation and opera-
tion.
Agencies ?ener~a.lly have followed one of two methods in applying
current policies for equipment selection :
1. They conduct "in-house evaluations" of equipment based on an
examination of characteristics and specifications contained in litera-
ture offered. by various manufacturers. This method is not used ex-
tensively, however. Information regarding latest equipment m:a*
not always be available. Regardless of how objective an agency s
in-house evaluation may be, manufacturers generally feel uneasy about
whether their equipment was properly evaluated.
2. Usually, the agencies provide system specifications to each manu-
facturer, with a request that he submit a proposal outlining which
of his equipment can best do the job. In addition, the proposal should
include such information as costs, personnel required, and the avail-
ability of programing aids. These proposals are then evaluated, to
determine those which may best meet the agencies' requirements.
The present study involved extensive discussion of selection poli-
cies and practices with manufacturers as well as with the agencies.
These discussions disclosed that implementation of current policies
1 Bureau of the Budget Circular A-54 : "Policies on the Selection and Acquisition of
Automatic Data Processing Equipment in the Executive Branch" (,Oct. 14, 1961).
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 AQifroved For Rele p>PA9 5 1'l91A: P ,6jB00446R000600050004-7
creates problems for both the agencies and the manufacturers. From
the agency viewpoint, the need to solicit and evaluate proposals (at
various review levels) delays the delivery date of the computer, to the.
disadvantage of the Government. The agency is unable to take ad-
vantage quickly of new technological developments, with the result
that. higher costs and lower product ivity are continued for longer than
would otherwise be so.
From the viewpoint of the manufacturer, there were several dis-
advantages:
(1) The cost to him of preparing a detailed proposal is-great.
Costs could be reduced if (a) there were more uniformity in the
format and content of the requests for proposals and the related
s stem specifications, and (b) more effective use of representative
(`benclunark") problems were made in the specifications.
(2) The time allowed by agencies for responding to proposals
is often too short. to pacrmit- proper preparation. Thirty to sixty
clays was suggested as an appropriate t hue period.
(3) Selections are often made without notifying the competing
and losing manufacturersof the reason why they lost.
Desp)ite these dissat isfactions, the practice of selecting equipment on
the basis of the manufacturer responses to requests for proposals is
regarded as sound. It provides the best paeans for each manufacturer
to present his case on an equal basis with all other manufacturers: the
inclusion of system specifications in the request for a proposal pro p-
erly places (lie emphasis upon the system, rather than upon a srmpe
comparison of hardware: and it helps to prevent biased selections
which may be unfair to the manufacturers. While the- practice may
cause some delay in improving operations, any costs that might be in-
curred on this account are, considered to be outweighed by the checks
andbalances which this pract ice provides against what otherwise night
be hasty, ill-conceived actions. Furthermore, the delay may not be
occasioned so much by (lie proposal process as by the internal review
and approval process which, again, is necessary to assure proper co-
ordination of all data processing activities.
Nevertheless, the need to simplify the proposals required of manu-
facturers is recognized. The preparation of systems specifications on
a more uniform basis by Government agencies, and the more effective
use of benchmark problems which are representative of the data proc-
essing requirements, would benefit both the supplier in preparing a
proposal as well as the Government in evaluating the proposals.
Further, the development and eventual use of standard, ninchine-
independent program lap rages will greatly, facilitate the comparative
evaluation of proposals from the standpoint of time manufacturer's
total product, including both equipment performance and related pro-
graming support. This emphasizes the importance of adequate Fed-
eral support of the program of the American Standards Association,
as discussed in chapter 7 of this report, since the association is now
in the process of developing American standard languages for com-
puter programing.
High on the list of factors affecting the selection of equipment are
the differences among the equipment. offered by the various manufac-
turers. (See "Chapter 7: Standardization of Equipment and Tech-
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For Relea 8"1/2*TC B00446R000600 004-7
niques.") This becomes particularly significant when selection con-
cerns the replacement of equipment. The operation of a given com-
puter often requires a large investment in the development of computer
programs. Once developed, these programs often cannot be used on
equipment offered by another manufacturer except, perhaps, by means
of a costly and often unsatisfactory conversion process. Consequently,
the user is "locked in" on a situation which may require him to replace
his equipment with compatible equipment offered by the same manu-
facturer, unless he wishes to ignore the cost of developing new pro-
grams, which in some cases cannot be justified from an economic stand-
point. Efforts are underway to free the user from such situations by
the development of more effective "conversion" programs, and particu-
larly by the development of machine-independent programing lan-
guages which can be used irrespective of the equipment used.
SHARING INFORMATION ON MANUFACTURERSI PERFORMANCE
There have been instances in which manufacturers have failed to
perform satisfactorily. In some cases equipment was not delivered on
time, was not maintained satisfactorily, or did not operate properly.
In others, the programing aids (usually referred to as software) were
not delivered with the equipment or did not perform as expected.
In general, this information has not been exchanged among agencies
on any routine basis, and therefore could not be given appropriate
consideration during the selection process and in the negotiation of
contract awards by the General Services Administration.
PRACTICES IN EVALUATING EQUIPMENT
The absence of detailed guidelines on the techniques and considera-
tions to be employed by agencies in their evaluation of equipment dur-
ing the selection process has led to a wide range of practices in this
regard. Undoubtedly, some are better and more effective than others
and, if so, should be uniformly applied. In addition, the current
policies and guidelines do not differentiate sufficiently between the
acquisition of additional equipment and the replacement or modifica-
tion of existing equipment. More specific guidance in this respect
would help to promote uniformity of action.
Present policies and guidelines on the selection of equipment need
expansion and updating ee the problems and needs set forth in
this chapter. Specifically:
1. Current policies, which require that system specifications be de-
veloped and used as a basis for selection, need to be supported by a
technical issuance, for Government-wide use, which provides criteria
and procedures in the preparation of specifications of this kind.
2. Guidelines are needed on the form, scope, and handling of invi-
tations to suppliers to submit equipment proposals. These guidelines
should include reference to the use of benchmark problems which are
representative of the data-processing requirements.
3. As reference material for the 'Government as a whole, updated
and authoritative information on the capabilities and characteristics
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Ap roved For Rea
i izf1P1' '? CW-M 00446R000600050004-7
of equipment is essential and should be provided. This guide should
also contain comparative studies of selected characteristics of avail-
able makes and models.
4. The distinctions between selection decisions for equipment addi-
tions replacements, and modifications, need identification and
clarification.
5. Methods should be devised to enable agencies to share their ex-
periences in the selection of equipment.
6. Government-wide criteria for evaluating suppliers' equipment
proposals should be provided.
7. Information concerning past. performance of suppliers of elec-
tronic data-processing equipment. would be helpful in the process of
equipment selection and should be provided to all executive agencies.
8. The factor of compatibility of equipment of the same or different
suppliers is an important consideration in the selection process. How-
ever, this fact needs to be emphasized; and most agency officials need,
and should be furnished, governmental criteria and guidelines on how
to take the compatibility factor into account.
9. Since electronic data-processing equipment in the Government
inventory, whether leased or owned which exceeds the needs of the
current holder competes with suppliers' equipment. in the selection
process, cost/effectiveness guidelines are needed to assist in determin-
ing when excess equipment should be used in lieu of acquiring addi-
tional equipment.
1. The Bureau of the Budget will provide for the publication of
criteria, gidelinrs, or regulations covering the selection of electronic
data-processing equipment. It will do this through new issuances or
by expanding upon current issuances, covering the following subjects :
(a) The preparation of system specifications, including bench-
mark problems, to be furnished equipment suppliers in requests
for proposals.
b) Evaluation of suppliers' proposals.
cj Compatibility considerations.
Consideration of excess and surplus equipment.
e) Distinctions to be made between additions, replacements,
an modifications when selection policies and criteria are applied.
(f) Interagency sharing of experiences in the selection and
performance of equipment.
2. The General Services Administration should p 'main current
d to on the characteristics and performance capabilities o a items
ofercially available general -purpose electronic data-processing
equipment. that are (a) currently in place in the Government, (U)
available from suppliers, and (e scheduled to become available front
suppliers. Based on this data, GSA should provide comparative in-
formation to agencies on request.
3. The General Services Administration should gather and make
ay dabble to executive agencies on request information on the per-
formance of the firms that sup ly electronic data-processing equip-
ment and programing aids to the agencies.
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For Release 2005/11/21: CIA-RDP&ZB00446R000600050004-7
CHAPTER 5
PURCHASE OR RENTAL OF COMPUTERS
Public controversy over the management of electronic data process-.
ing in Government has centered primarily on the following question :
Should computers be purchased or should they be rented? The answer
to this question involves the basic problem of deciding how long a
o'iven computer can. be used advantageously in the Government. If
curing this period the costs associated with purchase are less than the
costs associated with rental, the computer should be purchased; if
not, the computer should be rented.
The problem of deciding how long a computer can be used advanta-
geously is not an easy one. It is complicated by the need to consider
the effects of future changes in workload and system requirements,
the impact of improvements in technology, and the potential longer
term use of the computer elsewhere in Government after its original
purpose is served. All these considerations involve judgments. Either
decision-purchase or rental-resulting from these judgments can
be costly if it turns out to be incorrect. If a rented computer is used
beyond the break-even point, unnecessary rental costs will be incurred.
If a purchased computer cannot be used until the break-even point is
reached, a loss on disposal is likely to occur.
Policies and criteria governing these decisions must therefore be
developed cautiously, with a view toward minimizing losses. Further-
more, they must give attention to the longer range effects involved in
the. purchase of equipment. These include. the determination of the
point in time when purchased equipment should be replaced to avoid
uneconomical use; the management problems related: to the utilization
and disposal of Government-owned equipment when it becomes excess
to an agency's needs; and the provision of maintenance and operating
support for the equipment which, under a rental arrangement, is the
responsibility of the supplier.
As of June 30, 1964, 681 or 38.5 percent of the computers in use by
the Government had been purchased. Estimates ` indicated that by
June 1965 the percentage will have increased to 46.
These statistics reveal a significant change in attitude toward 'the
advantages of purchase. Until 1962, most computers in Government
(about 85 percent) were rented. This method was usually chosen
because computer technology was new and people were inexperienced
in its use. Rental offered a certain protection against obsolescence or
mistakes in selection because, presumably, the rented equipment could
be replaced more readily. In addition, the rental method had become
almost traditional because until 1956 the principal supplier did not of-
fer data processing equipment for sale.
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Ai-oved For Releaslf1AGpYk'800446R000600050004-7
In 1961, however, it. became apparent that the advantages of pur-
chase were being overlooked in many cases where long-term usage of
the computer was anticipated. As as consequence, the Bureau of the
Budget undertook the development of policies and criteria to govern
an agency's decision whether to buy or rent. A statement of policy
was issued in October 1961,' and the trend toward increased pur-
chasing was set in motion. Because of the need to make budget ad-
justments to finance the large capital investments, and because of a
general and understandable hesitancy to break with the existing wide-
spread rental practices, substantial increases in purchases were not
noted until the fiscal years 1963 and 1964.
The policy on which purchase or rental decisions are now based
is one which requires agencies to snake an objective evaluation of the
relative merits of both alternatives and to select. the one which, under
the particular circumstances, offers the greatest advantage. The
guidelines are relatively simple. They require the agency to-
(1) Calculate the point. in time (break-even point) when ac-
cumulated rental costs begin to exceed the cost. of purchase.
Normally this will range between 21/2 and 41/2 years. Because
rental costs accumulate according to the rate of usage, computers
with high utilization rates reach this point earlier.
(2) Decide whether the computer will satisfy the user for at
least that length of time. If so the equipment should be pur-
chased. If not, it should be rente .
This latter decision involves a judgment which the responsible
agency is best able to make. If an ,agency's programs are relatively
stable, future needs for equipment can be determined with relative
ease, and a decision to purchase a computer can be made with reason-
able certainty that the investment is adequately protected. This is
true of the large percent age of computers removed that have been pur-
chased. Statistics show that. purchased computers removed during
the rascal year 1964 had been installed an average of 51 months.
On the other hand, the decision concerning usage. is often very
difficult to make.. Many programs on which computers are used are
subject to rapid and unpredictable changes in workload. Others
may demand the latest, most powerful computer as it becomes avail-
able. Therefore, the length of time that the computer being ac-
quired for these programs will be capable of satisfying the need is
uncertain. Justifiably where such uncertainty exists, it would be
unwise to risk capital funds until it is reasonably clear that the com-
puter will be satisfactory for a period beyond the break-even point..
These cases account for the substantial number of computers still
being rented. Statistics show that leased computers removed during
the fiscal year 1964 had been installed only 34 months, on the average.
Government agencies have not been uniformly zealous, however, in
keeping rental contracts under constant review, in order that Prompt
action to purchase can be taken when it appears evident that the
computer will remain in use beyond the break-even point. This is
particularly true where the predicted rate of use is being greatly ex-
ceeded and the break-even point consequently will be reached earlier
iBureau of the Budget Circular A-54: "Policies on the Selection and Acquisition of
Automatic Data Processing Equipment in the Executive Branch" (Oct. 14. 1961).
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For Releasg9/X11'/2jATj~-Ik&P00446R0006000?0004-7
than had been anticipated in the original rental decision. In order
to avoid unnecessary rental costs, agencies must assure that the con-
ditions justifying rental are being continuously evaluated and that,
when rental is no longer justified, prompt purchase action is taken.
Current policies do not require that the cost of money be included
in determining whether to lease or purchase ADP equipment. It
would be desirable to require consideration of this additional factor,
although there will probably be few instances when such considera-
tion will be significant enough to alter the decision.
Government policy recognizes the effect of technological advances
by stipulating that if the break-even point for purchase is not ex-
pected to be achieved within a 6-year period, then (because replace
ment with superior equipment might be warranted by that time) the
computer should be rented. However, since the break even point is
usually reached between 21/2 and 41 years, the policy, in essence,
assumes that the rate of technological improvement will not be sub-
stantial enough to warrant replacement of equipment during this
relatively short period.
On the other hand, in their decisions to buy or rent, private industry .
gives substantial consideration to technological advancements. A
recent survey in a national magazine indicates that about 85 percent
of all computers in use by private industry are rented. Although
investment policy and tax considerations are also factors, the main
reason given for favoring rental is that this gives the user greater
flexibility in changing computers as significant technological advances
are made.
The effect of technological advances upon the production cycle of
selected computers is evident in illustration II. The length of time
in which computers remained in production before being superseded
range from 18 months to 79 months. Of those still in production
in June 1964, only two had been in production for as long as 61
months.
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Ag0roved For Releg[bq,/JW4A q"- P&7B00446R000600050004-7
ILLUSTRATION_ II
PRODUCTION CYCLE OF SELECTED COMPUTERS
MWOUt ZEO
ROOUCTIO14
MODEL NO.
A9RO%
MOS. IN
ROO'N
1950
8101
1911
RS2
1953.
1954
1755
Fp56
It7I
1931
1954
1%0
1%]
963
I%3
I964
1%5
UNIVAC I
69 ?
Y1111
UNIVAC 1101
41,
Y///
IBM 701
39
is
UNIVAC 1101/A
47 ?
Y//J
/1JJJ
IBM 707
II ?
%
////
kw&,ilr 209
77
/44
ISM 650
79
U C 0
Ill-
'I//
//1J
%4
IBM 704
40.
%//
/////
IM. 701
ill-
'
/////.
UNIVAC II
Si ?
i//1
V4%
4444
IVA 7011
37 ?
4bny+l1 1000
L
%//
//1J
IBM 709
2S.
1111
/1/
UIC 1
47 ?
V/1J
NCB 304
54 +
B 770
a
I VA 705 111
61.
////
/
UNNAC 1105
47 ?
IBM 7070
49 +
ACA 501
61 +
1/
1114 70}0
4? +
////
UNIV.MC SS 14t50
43?
%111
GE 210
44 +
IVA 1401
46 +
wak.d1 100
u +
CDC 1604
54 +
1MA 7010
34 +
NCI 790
3* +
?~~~
JNIVAC AG
31+
.INIVAC 111
23 +
JIIIVAC 1107
23 +
%//
MRmwlr 333
33 +
End
I llbd
uct{nn r
?Il?!i
?d,
ICA 301
41+
lim
E. d
?Iha
1M J.
- 64,
~oLl
//
f
will
1. poi
kn.
ICA 601
20 +
GE 775
L+
(
I
t
f
(
f
! 1
%1
Approved For Release 2005/11/21: CIA-RDP67B00446R000600050004-7
 Approved For Releasee21{31DA~JAP~g0446R000600059404-7
ILLUSTRATION II. (Cont'd)
PRODUCTION CYCLE OF SELECTED COMPUTERS
///////////// ANNOUNCED
PRODUCTION
MODEL NO.
APPROX.
1950
1951
1952
1953
1954
1955
1956
1937
1958
1959
1960
196,
1962
1963
1961
1965
MOS.IN
8
PROD'N
IAIOR
160
9
is,
NCR 310
39 +
IBM 1410
32 +
NCR 315
- 26 +
Hon?yw?II 40D
+
31
CDC 140A
36 +
BumoBh. 5000
+
17
UFC it
is
?
%////
BWTOR9lr 260/210
23 +
IBM 7040
15
IBM 7094-I
22 +
WI
UNIVAC 1004
+
Hat?ywdl 1800
+
CDC 921
20 ++
CDC 3600
13 +
1844 7010
6 +
IBM 1440
8 +
i
///4
IBM 1160
9 +
%//.1
GE 215
Homyo.II 1400
8+
UNIVAC 1050
8 +
GE 235
3 +
3
0
0
CDC 3200
0
GE 415/425
+
1
Hon?y.?II 200
0
? End of P,ndo?fion
wttm?td.
"1
I
CDC 3100
0
+ N,nb. of n,ont s
thou J.. 64,
.0&1
Non?y