F/0 DEFENSE ADP UNCLASSIFIED ;momhmhhhmmu; N … · DAP - Data Automation Panel DAR - Defense...
232
F A-AIL3 755 BOOZ-ALLEN AND HAMILTON INC BETHESDA ND F/0 15/5 DEFENSE ADP ACQUISITION STUODY. (U) NOV 81 MOA93-3..C-0490 UNCLASSIFIED N ;momhmhhhmmu; hmhhhhhhmsl
Transcript of F/0 DEFENSE ADP UNCLASSIFIED ;momhmhhhmmu; N … · DAP - Data Automation Panel DAR - Defense...
F A-AIL3 755 BOOZ-ALLEN AND HAMILTON INC BETHESDA ND F/0 15/5DEFENSE ADP ACQUISITION STUODY. (U)NOV 81 MOA93-3..C-0490
UNCLASSIFIED N
;momhmhhhmmu;hmhhhhhhmsl
FINAL REPORT
DEFENSE ADP ACQUISITIONSTUDY
30 NOVEMBER 1981
PREPARED FOR:HEADQUARTERS, U.S. AIR FORCE.1; ADP ACQUISITION IMPROVEMENT GROUP
and the .
DEFENSE SYSTEMS MANAGEMENTCOLLEGE
FORT BELVOIR, VIRGINIA 22060
a-:
BOOZ*ALLEN & HAMILTON Inc. 0
* 82 04 23 030
UNCLASSIFIEDSECURITY CLASSIFICATION OF THIS PAGE (1hem Date Entered)
PRDDPAGE N STRUCTIOWREPORT DOCUMENTATION PBEFORE COMPLTI G FORM
,. REPORT NUMBER 2. GOVT ACCESSION NO. 3. RECIPIENT'S CATALOG NUMBER
S A
4. TITLE (and Subtitle) S. TYPE Of REPORT & PERIOD COVERED
DEFENSE ADP ACQUISITION STUDY - FINAL REPORT6. PERFORMING ORG. REPORT NUMBER
7. AUTHOR(s) S. CONTRACT OR GRANT NUMBER(e)
U.S. Air Force ADP Acquisition ImprovementGroup & Booz Allen & Hamilton Inc.
9- PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT, PROJECT, TASK
AREA & WORK UNIT NUMBERS
U.S. Air Force ADP Acquisition Improvement
Group
II. CONTROLLING OFFICE NAME AND ADDRESS 12. REPORT DATE
U.S. Air Force ADP Acquisition Improvement 30 November 1981Group 13. NUMBER OF PAGES
17814. MONITORING AGENCY NAME & ADDRESS(If dillerent from Controlling Office) 15. SECURITY CLASS. (of this report)
Defense Systems Management College UNCLASSIFIEDDepartment of Research & InformationFt. Belvoir, VA 22060 SCHEDULE
16. DISTRIBUTION STATEMENT (of thie Report)
Approved for public release
17. DISTRIBUTION STATEMENT (of the abstract entered In Block 20, if different from Report)
Approved for public release
IS. SUPPLEMENTARY NOTES
19. KEY WORDS (Continue on reveree side It neceeeary and Identify by block number)
Air Force Computer ResourcesADP ADP RequirementsAir Force ADPAutomatic Data Processing
20. ABSTRACT (Continue en revere ede if neceesar and identify by block number)
The purpose of this study was directed at determining methods of reducingthe time and costs involved in the Air Force acquisition of general purposeautomatic data processing (ADP) resources. The focus of the study was onmethods of improving the ADP acquisition process by (1) reducing ADPacquisition lead time, (2) minimizing ADP technology lag, and (3) improvingthe ADP requirements definition and approval process.
DOI JAN 7 1473 EatTiON OF I NOV 65 IS OBSOLETE
SECURITY CLASSIFICATION OF THIS PAGE (1l1con Data Reo)
311
SECURITY CLASSIFICATIO ON THI THISVb PAIF(Me Dae ih
FINAL REPORT
DEFENSE ADP ACQUISITIONSTUDY
30 NOVEMBER 1981
PREPARED FOR:HEADQUARTERS, U.S. AIR FORCE
ADP ACQUISITION IMPROVEMENT GROUPand the
DEFENSE SYSTEMS MANAGEMENTCOLLEGE
FORT BELVOIR, VIRGINIA 22060
DOOZ*ALLEN & HAMILTON Ino.
/ ,.. !_ ,,____ __ .
DEPARTMENT OF THE AIR FORCEHEADQUARTERS UNITED STATES AIR FORCE0 WASHINGTON, D.C.
, 2 4DEC
SUBJECT: Booz Allen and Hamilton Final Report on the DefenseADP Acquisition Study
1. The research and analytical work documented in this reportwas sponsored by the Defense Systems Management College undercontract No. MDA 903-80-C-0490, and monitored by the Air ForceADP Acquisition Improveme-nf- up, HQ USAF. The report is theproduct of a comprehensive examination of the Federal ADPacquisition environment and it addpresses a wide range oftechnical, statutory, regulatory, and management issues. Italso provides broad insight into the nature and causes of problemsin the ADP acquisition process and offers several strategiesfor improvement.
2. As a major source of information on ADP acquisition practicesand associated problems, the report has been extremely usefulin developing a plan to improve internal Air Force ADP acquisitionpractices. It does not, howeVer, necessarily reflect the opinionsor conclusions of its sponsors, nor does it represent an Air Forceposition on specific improvement approaches.
DONALD W. SAWYER, JR, Col, USAFDirector, ADP AcquisitionImprovement Project
I
'3 .... _ _
r
BOOZ ALLEN & HAMILTONInc
Madnagemnt Censulta tS
4330 EAST WEST HIGHWAY
BETHESDA, MARYLAND 20014
CHRISTIAN S. YOUNG 951-2537
SENIOR VICE PRESIDENT AREA CODE 30,
November 30, 1981Defense Systems Management CollegeDepartment of Research and InformationFort Belvoir, Virginia 22060
Attention: Mr. John McKeown
Subject: Letter of Transmittal for Delivery of Final Report,Contract No. MDA 903-80-C-0490/
Dear Mr. McKeown:
We are pleased to submit the attached Final Report of the DefenseADP Acquisition Study in accordance with Task 9 of the Statement ofWork. Written comments to the draft report were discussed with theAir Force Acquisition Improvement Group and have been incorporated inthis Final Report.
The major findings, conclusions, and recommendations are summarizedin an Executive Summary which precedes the main body of the report.
We have enjoyed our participation in this study of the DOD ADPacquisition procedures, and we would like to express our sincere appre-ciation to the Acquisition Improvement Group of ihe Air Force Directorof Computer Resources for their support in obtaining much of the in-formation used throughout the study. If you have any questions regard-ing the enclosed report, or if we can be of further assistance, pleasecall on us.
Very truly yours
BOOZ ALLEN &HAM LT( Inc.
Christian S. YoungSenior Vice President
enclosure 0
1 0
4 C:
a 1-4 lF, 1 0)u-Inc C r
TABLE OF CONTENTS
PageNumber
EXECUTIVE SUMMARY 1
I. INTRODUCTION I-I
1. Objective and Scope I-I2. Methodology I-2
II. THE ADP ACQUISITION PROCESS II-1
1. The Policy Environment II-i2. Process Description 11-27
III. EVALUATION OF FINDINGS III-1
1. Process Performance Evaluation III-i2. Analysis of Issues 111-22
IV. STUDY CONCLUSIONS IV-1
1. Air Force Implementation IV-12. Objective of the ADP Acquisition Process IV-43. ADP Acquisition Process Design IV-94. Process Organization and Management IV-135. External Air Force Constraints IV-27
V. RECOMMENDATIONS V-I
1. Objectives of the ADP Acquisition Process V-22. Revisions to the Process V-43. Organizational Roles and Responsibilities V-154. External Considerations V-30
VI. IMPLEMENTATION APPROACH VI-1
1. Steps to Implement the Revised Process VI-22. Realignment of Roles and Responsibilities VI-53. Activities of the Acquisition Improvement VI-7
Group
ii
REFERENCES
APPENDIX A - Literature Review
APPEN4DIX B - Government Survey
APPEN4DIX C - Industry Review
APPENDIX D - Performance Statistics
APPENDIX E - DSMC CaeStudies
APPENDIX F - Study Issues
APPENDIX G - Abstracts
I NDEX OF FIGURES
PageNumber
II-i Diachronic Chart 11-2
11-2 Application of Policy Guidance toto Management of ADP and ComputerResources 11-29
11-3 Air Force Program Organization 11-31
11-4 ADP Acquisition Process Work FlowDiagram (Defined by AFR 300-series) 11-38
11-5 ADP Acquisition Process FlowDiagram (observed) 11-43
III-1 Cost Distribution of AFCACProcurements 111-7
111-2 AFCAC Procurement Cycle Timevs. Cost 111-8
111-3 GSA Approval Time vs. Cost 111-9
111-4 Age of 978 Large- and Medium-ScaleFederal Computers Based on TheirAcquisition Dates and Technological Age 111-18
V-1 Proposed ADP Acquisition Process V-7
i. iiiI
INDEX OF TABLES
PageNumber
II-1 AF ADP Resource Approval
Thresholds 11-46
III-1 DAR Workload in Hqs., USAF 111-5
111-2 Trends in Types of Procurements 111-6
111-3 Processing Times for Test Cases 111-13
111-4 Sample Commercial ADP Acquisitions 111-16
111-5 Comparison of AF and Industry 111-16
111-6 Summary of Economical SystemReplacement Illustrations 111-21
111-7 Issue-Source Cross-Reference 111-24
111-8 Categorization of Problems 111-45
IV-l Issue-Conclusion Cross Reference IV-2
IV-2 Engineering Disciplines inSystems Acquisitions IV-18
IV-3 AFR 300 Series Skill Requirements IV-24
V-I Revised ADP Acquisition Process-Phase'I Description V-9
V-2 Revised ADP Acquisition Process-Phase II Description V-12
V-3 Revised ADP Acquisition Process
Phase III Description V-14
V-4 Proposed Maximum Thresholds V-17
V-5 Summary of Air Staff Roles andResponsibilities for the Manage-ment of ADP Resources V-26
iv
TABLE OF ABBREVIATIONS
A
ACD - AF Directorate of Computer Resources
ADAPSO - Association of Data Processing Service
OrganizationsADP - Automatic Data Processing
ADPE - Automatic Data Processing Equipment
ADPR - Automatic Data Processing Resources
ADPS - Automatic Data Processing System
ADS - Automated Data System
ADTS - Automated Data and Telecommunications
Service
AFADPP - Air Force ADP Plan
AFCAC - Air Force Computer Acquisition Center
AFCC - Air Force Communications Command
AFDSC - Air Force Data Services Center
AFDSDC - Air Force Data Systems Design Center
AFDSEC - Air Force Data Systems Evaluation Center
AFR - Air Force Regulations
AFLC - Air Force Logistics Command
AFSC - Air Force Syst js Command
AIG - AcquisitioroImfWovement Group
AIS - Automated Information Systems
AL - Acquisition and Logistics
ALS - Advanced Logistics System
AMP - ADPS Master Plan
ANSI - American National Standards Institute
APR - Agency Procurement Request
ASD(C) - Assistant Secretary of Defense
(Comptroller)
ATC - Air Training Command
V
C
CBD - Commerce Business Daily
CBEMA - Computer and Business Equipment
Manufacturers' Association
CCIA Computer and Communications Industry
Association
CCPC - Communications Computer Programming Center
CDI - Comprehensive Dissertation Index
CDR - Critical Design Review
CI - Configuration Item
CINC - Commander-In-Chief
CIS - Congressional Information Service
CMP - Configuration Management Plan
CPCI - Computer Program Configuration Item
CRISP - Computer Resources Integrated Support Plan
D
DAP - Data Automation Panel
DAR - Defense Acquisition Regulations
DAR - Data Automation Requirement
DDA - Directorate of Data Automation
DCS - Deputy Chief of Staff
DMP - Data Management Plan
DOC - Department of Commerce
DOD - Department of Defense
DODD - Department of Defense Directives
DODI - Department of Defense Instructions
DPA - Delegation of Procurement Authority
DPD - Data Project Directive
DPP - Data Project Plan
DS - Data Base Specification
DT - Development Test Plan
DTIS - Defense Technical Information Service
vi
E
EA - Economic Analysis
EO - Executive Order
F
FAR - (Proposed) Federal Acquisition Regulation
FCA - Functional Configuration Audit
FEDSIM - Federal Computer Performance Evaluation
and Simulation Center
FIPS - Federal Information Processing StandardsFM - Financial ManagementFMC - Federal Management Circular
FPMR - Federal Property Management Regulations
FPR - Federal Procurement Regulations
FPS - Federal Procurement System
FS - Feasibility Study
FSS - Federal Supply Schedule
G
GAO - General Accounting Office
GFE - Government Furnished Equipment
GP - General Purpose
GSA - General Services Administration
H
HAC - House Appropriations Committee
HASC - House Armed Services Committee
HGOC - House Government Operations Committee
HOI - Headquarters Operating Tnstructions
vii
HOL - Higher Order Language
HQ - Headquarters
HW - Hardware
I
IFB - Invitation for Bids
ILSP - Integrated Logistics Support Plan
IN - Intelligence
IOC - Initial Operational Capability
IPSC - Information Processing Standards for
ComputersIRM - Information Resources ManagementISO - International Standards Organization
L
LCC - Life Cycle Costs
LCM - Life Cycle Management
LE - Logistics and Engineering
LTOC - Lowest Total Overall Cost
M
MAJCOM - Major Command
MCAP - Major Command or Separate Operating Agency
ADP Plan
MIS - Management Information System
MM - Maintenance Manual
MP - Manpower and Personnel
MPC - Military Personnel Costs
viii
I
N
NBS - National Bureau of Standards
NTIS - National Technical Information Service
0
OEM - Original Equipment Manufacturer
OFMP - Office of Federal Management Policy
OFPP - Office of Federal Procurement Policy
0I - Operating Instruction
OIRA - Office of Information and Regulatory
Affairs
OJT - On-the-Job TrainingOM - Operator's Manual
OMB - Office of Management and Budget
OPR - Office of Primary Responsibility
OSD - Office of the Secretary of Defense
OSD(C) - Office of the Secretary of Defense,Comptroller
P
PA - Programs and Evaluation
PAR - Projected Automation Requirement
PCA - Proposed Configuration Audit
PDR - Preliminary Design Review
PL - Public Law
PM - Program Manager
PMD - Program Management Directive
PMO - Program Management Office
PMP - Program Management Plan
POM - Program Objectives Memorandum
PPBS - Planning, Programming, and Budgeting System
ix
PRP - President's Reorganization Project
PS - Program Specification
PT - Test Plan
PVR - Product Verification Review
QA Quality Assurance
R
RD - Data Requirements Document
RD - Research and Development
RFP - Request for Proposal
S
SAC - Strategic Air Command
SAC - Senate Appropriations Committee
SADSC - San Antonio Data Services Center
SASC - Senate Armed Services Committee
SAF - Secretary of the Air Force
SAF/FM - Secretary of the Air Force/Financial
Management
SDN - System Development Notification
S&I - Surveys and Investigations
SDR - System Design Review
SEMP - System Engineering Management Plan
SOA - Separate Operating Agency
SRR - System Requirements Review
SSA - Source Selection Authority
SS - System Specification
SVR - System Validation Review
SW - Software
x
T
T/A - Table of Allowances
TEMP - Test and Evaluation Master Plan
U
UM - Users Manual
UPS - Uniform Procurement System
USC - U.S. Code, Annotated
wWWMCCS Worldwide Military Command and Control
System
x
XO Operations, Plans, and Readiness
tx
xi
EXECUTIVE SUMMARY
EXECUTIVE SUMMARY
14 i In early 1980 the Assistant Secretary of the Air Forcefor Financial Management (SAF/FM) directed that a study bedone to determine methods of reducing the time and costsinvolved in the Air Force acquisition of general purposeADP resources. Specifically, the focus was on methods ofimproving the ADP acquisition process by:
. Reducing ADP acquisition lead time
* Minimizing ADP technology lag
* Improving the ADP requirements definition andapproval process.
A secondary objective of the study was to provide anADP resource library and case study materials to supportacquisition management instruction programs.
This report documents the results of the Defense ADPAcquisition Study, prepared for the Defense Systems Man-
agement College, Ft. Belvoir, Virginia, in conjunctionwith the Air Force Directorate of Computer Resources,
ter. Echchapter is summarized in this Executive
1. HE DPACQUISITION PROCESS
The ir FrceADP acquisition process can best bedesribd i temsof the policy environment and the spe-
cific activities comprising the process.
(1) The Policy Environment Is the Regulatory Struc-ture and Organizational Hierarchy Impacting theProcess
Federal ADP policy has its foundation in publiclaw and is' implemented throughout the Federal struc-ture by directives, circulars, regulations, pamphletsand standards from the numerous organizations in-volved. The policy environment within which the AFADP program functions includes the following:
Congress enacts laws, particularly PL 89-306(the "Brooks Act").
Executive Orders direct Federal organiza-tional roles and responsibilities.
OMB communicates policy and procedural guid-ance through issuance of circulars.
GSA issues both Federal Procurement Regula-tions (FPRs) and Federal Property Manage-ment Regulations (FPMRs) governing ADPacquisition.
DOD issues Defense Acquisition Regulations(DARs) as the principal DOD procurementregulation system.
OFPP issues pamphlets to provide procurementpolicy guidance to agencies in applying OMBcirculars to their ADP acquisitionactivities.
The National Bureau of Standards (NBS)issues Federal Information Processing Stan-dards (FIPS) setting forth Federal standardsand guidelines.
DOD Directives and Instructions implementFederal policy and provide direction to themilitary departments.
Secretary of the Air Force Orders delineateADP roles and reponsibilities within the AirForce.
In 1965 the Brooks Act, PL 89-306, amended theFederal Property and Administrative Services Act of1949 "to provide for the economic and efficient pur-chase, lease, maintenance, and utilization of auto-matic data processing equipment by Federal departmentsand agencies." The law established a structuralframework and centralized procurement authority tocontrol the acquisition of ADP resources. The basicauthorities of OMB, GSA, Commerce, and the user agen-cies relative to ADP are outlined in the law.
-2-
I. _ _ _ _
(2) The Process Description Is Presented in Terms ofthe Participants and Their Activities
The Air Force acquisition process for generalpurpose ADP resources can be described in terms of itsscope, the organizational participants, the processsteps, processing thresholds and management of the ADPprogram. The following comments touch upon thesesubjects:
* The AF 300 Series Regulations encompass theacquisition of all general purpose ADP re-sources except those acquired in combat
J weapons systems and specially designed
equipment.
* The senior ADP policy official for the AirForce is the Assistant Secretary of the Air.1 Force for Financial Management.
* The Director of Computer ResourcesK' (USAF/ACD) is the Air Force ADP programsingle manager and has been delegated man-agement authority for developing plans,policy guidance and procedures for planning,programming and budgeting for general pur-pose ADP; for approving data automationrequirements and directing and monitoringADP acquisitions; for performing liaisionwith GSA and other Federal organizations;and for providing technological expertise tothe Air Force.
* Major Commands (MAJCOM) and Separate Oper-ating Agencies (SOA) have their ADP ProgramSingle Managers who are responsible for pro-viding ADP support for the Command mission.
* An ADPS (automatic data processing system)is an aggregation of software and the re-sources required to support it (equipment,software, manpower and facilities). ADPSmanagers are appointed for command uniqueand UJSAF standard ADPS.
* The Deputy Commander for Data Automation ofAir Force Communications Command (AFCC) isresponsible for managing AFCC resources toprovide ADP support to Air Force activi-ties. The subordinate (ADP) elements ofAFCC include:
-3-
Air Force Computer Acquisition Center(AFCAC)
Air Force Data Systems Design Center
(AFDSDC)
- Air Force Data Services Center (AFDSC)
Federal Computer Performance Evaluationand Simulation Center (FEDSIM)
Communications Computer ProgrammingCenter (CCPC)
Air Force Data Systems EvaluationCenter (AFDSEC)
- San Antonio Data Services Center (SADSC)
- Phase IV Program Management Office.
The ADP Review Board reviews the Air ForceADP Program and advises the SAF/FM on majorADP policies and projects.
The Data Automation Panel (DAP) conducts theAir Staff Board Structure programming andbudgeting review of the Air Force ADPProgram.
The Air Force ADP acquisition process startswith the identification of a requirement forADP and proceeds from requirement definitionto requirement approval, to technical speci-fication development, to procurement, toinstallation and test.
Requirements definition and approval is arigorous series of steps described in greatdetail in the AFR 300 Series regulations.The requirement is documented in a DAR (DataAutomation Requirement).
Threshold values govern the level of ap-proval authority required for an ADP ac-quisition. The following individuals areauthorized to approve ADP acquisitions:
- The Air Force Senior Policy Official(SAF/FM)
-4-
- The Air Force ADP Program Single Manager
- Command ADP Program Single Managers
- USAF ADPS Managers
- Command ADPS Managers.
2. EVALUATION OF FINDINGS
The performance of the Air Force process for acquiringgeneral purpose ADP was evaluated by both a statisticaland a comparative analysis. The process was also examinedin relation to two fundamental questions.
* Is the Air Force incurring damage by the currentADP acquisition process?
* If so, what are the causes of the damage?(1) Process Performance Evaluation Revealed That the
Process Is Unnecessarily Lengthy
The performance of the Air Force process for ac-quiring general purpose ADP was evaluated along twoapproaches. First, a classical process evaluation wasperformed by statistically analyzing the acquisitionsthat go through the process. This approach focused onthe dwell times required to complete each step and thevariances in processing time, depending upon what isbeing processed. Voids in the available statisticaldata inhibited the statistical analysis. However, tothe extent that the statistical evaluaton could becarried out, it indicated that the variances in pro-cessing time are more related to unique aspects ofindividual acquisitions than to factors that could becaptured by aggregate statistical analysis. There wasonly a slight correlation between the size and type ofacquisition and the time required for the acquisitionprocess.
The second approach to process evaluation was tocompare the Air Force performance to that of privateindustry in acquiring similar ADP resources. Usingthe performance of industry as an outside standard, itwas determined that the Air Force takes an average ofthree times as long as industry to acquire ADP. Thisindicates that there is nothing inherent in the ac-quisition of general purpose ADP that generateslengthy acquisition times, but rather, the basis forthe delays is related to the Air Force acqusitionprocess.
-5-
(2) The Current Air Force Acquisition Process ImposesTwo Types of Damage on the Air Force: ExcessiveAcquisition Cost and Capability Loss
ADP acquisition time in the Air Force is exces-sive. A more significant cost imposed by the currentacquisition process is technology lag. Newer equip-ment not only offers greater capability but also re-sults in cost savings. The following changes occurwith newer technology equipment and result in reducedcosts:
* Less maintenance
* Fewer operators
* Less energy consumption
* Improved system reliability
* Less floor space
* Less personnel
.34 * Lower power for uninterruptible powersupplies.
Capability loss arises because of either a fail-ure to acquire needed capability or a failure toaccommodate growth. Failure to acquire needed capa-bility may result from a decision by a user thatsatisfying a requirement is not worth the effort ofgoing through the acquisition process.
(3) Analysis of Issues Reveals the Root Causes fortEhe DamageI
The collection of data from literature, govern-ment and industry sources resulted in a large volumeof information addressing a variety of perceived prob-lem areas which contributed to the damage cited in theprevious section. These data were sorted into 33 is-sues and analyzed to determine the recurring causesfor the poor process performance. They were furtheranalyzed to identify whether the causes were withinthe capability of the Air Force to correct or if theyrequired corrective action by organizations externalto the Air Force.
-6-
The analyses revealed the following root causeswhich have a direct beaAng on the current state ofthe ADP acquisition process:
* Hardware orientation and solutional focus ofthe ADP acquisition process
* Ineffective use of skills and personnelresources
* Inflexibility in the acquisition and manage-ment of-ADP
* Failure of the ADP acquisition process toadapt to a changing environment
* Lack of accountability and end-to-end man-agement in the process
* Lack of a mission orientation
. Lack of systems management and life cycleperspectives
* Lack of effective leadership and policydirection
* Inconsistency in policy, direction and man-agement of ADP.
3. STUDY CONCLUSIONS
The five major conclusions of the study follow. For
emphasis, those areas internal to the Air Force are pre-
(1) While External Influences May Be the Source ofMany Problems in the Air Force ADP Ac~uisitionProcess, Major Improvement Can Be Achieved byRedirections Within the Air Force
Many of the major constraints on ADP acquisitionsare self-imposed. There are a variety of causes forthese constraints. In some cases, policies which oncepromoted effectiveness and efficiency now constrainthe achievement of these goals because of a changingenvironment or differing needs. Some policies havealways been constraining to the process because theywere based on erroneous perceptions of the externalenvironment.
-7-
(2) The Current ADP Acquisition Process ObjectivesUnderemphasize Mission Essentiality in Favor of aPrimary Emphasis on Achieving the Goals of Econ-omy and Efficiency
The most frequently mentioned objectives citedfor the ADP acquisition process are achieving economyand efficiency and maximizing hardware competition.While these were legitimately given primary considera-tion in an era when it was appropriate to do so, theADP environment has changed considerably.
1. The Tenets of Efficiency and Economy AreBased on Assumptions Which Do Not ReflectCurrent Realities
Hardware is no longer the primary cost com-ponent. The cost of software, services andfacilities is increasing while hardware costsdecrease.
2. Emphasis in the ADP Acquisition Process MustBe Placed on Satisfying Operational Needs asWell as on Achieving Efficiency
Because the current process responds pri-marily to the goal of maximizing efficiency, costversus effectiveness tradeoffs are not generallyperformed.
3. The ADP Acquisition Process Lacks the ProperDegree of Planning to Effectively Focus onMission Needs
Current long-range strategy planning failsto provide the appropriate mission perspective.Curfent top-down strategic planning does not pro-vide the necessary guidance for the AF ADP pro-gram. Long-range planning at the MAJCOMs andSOAs does not provide a mission-oriented focus.The AFADP Plan is of little value in planningactivities at MAJCOM/SOAs.
4. The Current ADP Acquisition Process IsInflexible and Overly Restrictive
The AFR 300 Series regulations reflect thisrestrictiveness and inflexibility. The 300Series documentation requirements are duplicativeand solution-oriented. They result in delays and
-8-
less than optimal decisions. The great extent of
the documentation contributes significantly tothe acquisition lead time.
ization Contribute to the Confusion and Thus toDelays in Acquiring ADP
The Air Force has established a unique structurefor the management and acquisition of ADP; it is un-like the process used for acquiring other systems.There is confusion as to what acquisitions belong inthe 300 and the 800 acquisition process, and there islittle corporate-level review to specifically insurethat acquisitions are channeled correctly. There isalso a lack of end-to-end management accountability inthe current process design.
(4) Acquisition Under the Current Process DemandsSkills That Are Scarce, and This Inhibits theEfficiency of the Process
A variety of skills are required in the ADP ac-quisition process; not only are they lacking, buttraining to provide these skills is limited. Definingrequirements, preparing a DAR, and developing specifi-cations do not occur frequently in a career, andpeople are untrained for the tasks.
(5) The Confusion and Inefficiencies Introduced Intothe ADP Acquisition Process by Top-Level FederalPolicy Have Their Roots in Conflict Over the
Proper Direction of Acquisition Reform
There are conflicting philosophies at the con-gressional level between the House Government Opera-tions Committee (which stresses maximum competition)and the House Appropriations Committee (which stresseslowest total overall cost). The HGOC exerts greaterinfluence in the ADP acquisition process. The indis-tinct roles and responsibilities of 0MB and GSA hindereffective policy development and implementation.Faced with conflicting philosophical positions andcontradictory guidance, ASD(C) has responded inade-quately to the needs of the services' ADP acquisitioncommunity.
4. RECOMMENDATIONS
Correction of the deficiencies identified in the studyrequires an integrated set of recommendations which ad-dress fundamental changes in the acquisition process as
oo
-9-
well as specific changes in policy, procedures, roles andresponsibilities. The recommendations presented here aremore appropriately labeled strategies for change, ratherthan specific actions for improvement.
(1) There Must Be a Clear Recognition That the Pur-pose of the ADP Program Is to Fulfill MissionNeeds, and Responding to That Mission Must Be thePrimary Objective of the ADP Program
While mission primacy is not exclusive of other[acquisition objectives, satisfaction of operationalneeds cannot be sacrificed to achievement of theseother objectives. The process objective should be tosatisfy mission needs, within the required time, atthe lowest life cycle cost. Competition should be
used to the maximum extent practical to satisfy thisobjective. Strategic planning, as an essential ele-ment of mission primacy, should be instituted.
(2) In Light of the Mission Primacy Objective, theProcess Must Be Modified to Incorporate Flexi-bility to Accommodate Changing Needs and Adapt-
ability to a Changing Environment
Fundamental changes are required in the process
design. Five major changes are recommended.it There must be a clear, strong and obviouslinkage between Air Force mission planningand ADP acquisition program identification.
The decision to approve a need should bemade as part of funds approval in the Plan-ning, Programming, and Budgeting System(PPBS).
* The decision to approve the ADP solution and2 the acquisition strategy should be separated
from the need approval and be delegated to alower decision authority.
The process design should be based on provensystems management practices:
- Program Management System
- Life Cycle Systems Management
- Decentralized execution of theacquisition.
-10-
Acquisition process regulations and proce-dures should be simplified to enhance under-standing and encourage flexibility.
(3) Internal Roles, Responsibiliies and Organiza-tional Missions Must Be Modified to AccommodateChanges in the Process
This study identified four major deficiencies
(concerning people) which constrain the effective ac-quisition of ADP -- the mismatch of roles and skills,the inconsistencies in the management of and organiza-tion for ADP, the lack of end-to-end accountability,and the lack of effective leadership and guidance. Tocorrect these deficiencies, together with the recom-mended changes in the process, necessitates changes inthree areas.
i. The role of the single managers at theMAJCOM/SOA level should be strengthened by in-creasing their approval thresholds, by giving
them greater authority to reallocate funds, andby expanding their authority to initiate pro-grams. Also, skill levels in certain disciplinesmust be enhanced and made available to theMAJCOM/SOA program managers, e.g., systems engi-neering, program management and planning, logis-tics support, configuration management and opera-tional planning. For those MAJCOM/SOAs withrecurring acquisition programs requiring thosedisciplines, personnel with these skills shouldbe under the direction of the single manager.
2. The Deputy Commander for Data Automation inAFCC should have primary responsibility for USAFstandard ADPS, and support responsibilities tothe MAJCOM/SOA single managers. Within the AFCCCenters are areas of expertise which are limitedelsewhere in the Air Force. This expertiseshould be made available to support theMAJCOM/SOA non-standard systems on an as-neededbasis.
3. Adoption of the preceding recommendationsrequires a reallocation of roles and responsi-bilities on the Air Staff. Future considerationshould be given to consolidating the responsi-bilities for the management of telecommunica-tions, ADP, and computer-based office systems.Responsibility for the management of ADP shouldbe consistent with the organization of the AirStaff.
-11-
Contracting and acquisition responsi-.3 bilities should be assigned to AF/RDC.
} Responsibilities concerning the genera-tion, approval and management of re-quirements should be distributed within
* AF/RD.
* operational responsibilities for ADPsystems should reside within AF/XO.
Certain ADP support responsibilitiesshould be assigned to AF/LE.
(4) External to the Air Force, Previous Recommenda-tions Should Be Complemented by Changes in Policyand Modification of Organizational Roles andRelationships
Achieving maximum benefit from the recommendedAir Force process improvements requires some attentionto the external organizations which influence theprocess -- principally OSD. The Air Force should con-centrate on encouraging OSD to take a stronger, more
active and visible role in Federal ADP matters. The*ifAir Force should encourage OSD to:* Adopt consistent realignments of roles and
responsibilities within DOD similar to therealignments proposed for the Air Force.
* Simplify DOD directives and instructions and iensure consistency with Federal ADP policy
and acquisition regulations.
.4 . Regain credibility with Congress throughimproved relations and demonstrations ofcompetent ADP resource management.
* Assume a more dyamic role in relations withCongress, OMB, and GSA so as to assert DODinfluence on Federal ADP policy formulation,particularly regarding the implementation ofPL 96-511 and UPS.
5. IMPLEMENTATION APPROACH
The use of the Air Force Acquisition Improvement Group(AIG) to implement the recommendations is key, if notcritical. The implementation addressed here is in termsof the ADP acquisition process, the realignment of roles
and responsibilities, and the activities of the AIG.
-12-
(1) Implementation of the Revised Process Requires aSeries of Steps That Recognize That the ADP Ac-quisition Process Is Ongoing
Many ADP acquisition activities are keyed to thePPBS. At this time it is realistic to expect that allrecommendations could be implemented in time to affectthe FY85 budget call. The following specific stepsshould be taken:
* Brief external agencies on the impendingchanges
* Appoint program managers for all programsthat have survived the PPBS process
* HQ USAF begin developing planning guidancefor the FY85 budget call
* MAJCOM/SOAs implement new planning guidance
* Revise Air Force Regulations
. Initiate personnel and training actions toeliminate shortages of skills
* HQ USAF establish a program to monitor andevaluate the process performance.
(2) Realignment of Roles and Responsibilities ShouldBe Accomplished at HQ USAF and MAJCOM/SOAs
The following implementation actions are pre-sented in their recommended sequence:
* Strengthen the single manager's roles andresponsibilities
* Revise the role of AFCC in the acquisitionprocess
* Realign ADP responsibilities within the AirStaff
* Transition the ADP acquisition executiveresponsibilities from SAF/FM to SAF/AL.
-13-
(3) The Acquisition Improvement Group Should Under-take Certain Implementation Actions While FurtherAnalyzing Other Subjects
Implement:
* Brief external agencies
* Develop interim guidance for single managers
* Develop planning guidance
* Develop alternative ADP acquisitionapproaches
* Revise AFR 300 Series
* Support development of the plan to transferADP responsibilities within the Air Staff.
'I Study further:p
* Life cycle model for ADP
* Criteria for defining technological and eco-nomic obsolescence
* Alternative ADP acquisition approaches
* ADP Program Manager's Handbook
* Need for a HQ USAF office responsible fortelecommunications, ADP, and computer-basedoffice systems.
-14 -
I. INTRODUCTION
* I. INTRODUCTION
This report describes and documents the research, con-clusions and recommendations of the Defense ADP Acquisi-tion Study, completed for the Defense Systems ManagementCollege under the auspices of the Air Force Directorate ofComputer Resources, HQ rSAF/ACD.
J 1. OBJECTIVE AND SCOPE
This study investigated the existing Air Force ADP:1 acquisition process in order to formulate methods of re-ducing the time and costs involved in the acquisition ofgeneral purpose ADP resources. The study focus was onmethods of improving the ADP acquisition process by:
Reducing ADP acquisition lead time
* Minimizing ADP technology lag
* Improving the ADP requirements definition andapproval process.
A secondary objective of the study was to pro~vide anADP resource library and case study materials to supportacquisition management instruction programs.
Specific study activities included:
* Investigating the environment of the acquisitionprocess through a review of the management struc-ture that influences the process, i.e., the regu-latory and organizational hierarchy
* Reviewing previous reports from GovernmenLt andindustry sources concerninig problems in ADPacquisition
* Interviewing representatives from the Air Force,civil agencies, and private organizations
* Examining a number of cases of ADP acquisitionsby private commercial firms in order to compareGovernment and industry methods
* Identifying and analyzing issues to determine theroot causes for poor process performance
1-1
* Suggesting process improvements that address the ]underlying and pervasive problems inhibiting suc- icess of the acquisition process
* Preparing case studies of recent ADP acquisitionsto evaluate the effectiveness of the process.
While the purpose of the study was to identify im-provements to the ADP acquisition process for the entireDepartment of Defense, the study methodology was confinedto an analysis of the procedures implemented within theAir Force. The problems and constraints in acquiring ADPare not unique to the Air Force; the recommendations pro-posed in this study can be applied in concept to the otherservice branches.
2. METHODOLOGY
The methodology used to study the ADP acquisition pro-
divided into five phases:
. Problem identification And research planning,including determination of data sources and datacollection methods
* Data collection, including formatting and docu-mentation for subsequent use
* Data reduction, analysis, and integration
Formulation of recommendations
Documentation of research results.
(1) The ADP Acquisition Process Definition Providedthe Project Orientation
The initial study efforts focused on defining theAir Force ADP acquisition process and analyzing itfrom the point of view of the study objectives. Theapproach was to flowchart the ADP acquisition process,as it is defined in the AFR 300 series regulations,and while doing so, to identify problem areas in theacquisition of ADP as perceived by senior Air Forceofficials.
Once the acquisition process was charted, andmajor problem areas identified, the next step was thedevelopment of a research plan. Hypotheses and key
1-2
questions were developed to guide the study efforts.At the same time, information sources were identified,and interview guides were designed for use in the datacollection phase.
(2) The Detailed Data Collection Was Conducted Uti-
lizing Three Major sources
The research plan required that data collectionefforts focus on three major sources:
* Literature* Government* Industry
These data sources are described below.
1. The First Data Source Was the Extensive Bodyof ADP Acquisition Literature
K A comprehensive literature review was con-ducted in three parallel efforts. The develop-ment of the ADP acquisition library and review ofincoming reference material provided the initialscreening for key resources. A parallel data
collection task involved an indepth review of theconclusions and recommendations of previous keystudies and research reports. The third datacollection task in the literature review con-sisted of an analysis of the policy environment,focusing on the legislative and regulatory hier-archy. The activities comprising the policy re-view included tracing legislative history andidentifying inconsistencies in law or policy.Details of the literature review are presented inAppendix A.
2. The Secon-' Source Was Data Drawn FromGovernment Participants in the Process
Government survey teams conducted interviewsand reviewed documentation at all levels of par-ticipation. With the exception of GSA in Denver,major Federal agencies involved in the ADP ac-quisition process were interviewed in theWashington, D. C. area. Documentation review andpersonal interviews were conducted throughout theAir Force. Interviews were conducted at Head-quarters Air Force, at the headquarters of MajorAir Commands (MAJCOMS), and at field activities
1-3
that play a key role in the Air Force ADP acqui-sition process. Those MAJCQMs not visited by thesurvey teams were surveyed by written question-naire. In addition to the general Governmentsurvey, personal interviews and case file reviewsat HQ Air Force and at some MAJCOMs (for the fiveacquisition cases) were conducted as part of thisstudy. Details of the government survey are con-tained in Appendix B.
3. The Third Source Was Data Collected FromPrivate Industry
The industry surveys were aimed at obtainingindustry's perspective of how the Air Force pro-cess is working, obtaining views on problems andsolutions, and finally, comparing industry prac-tice for acquiring ADP with those of the Air-1 Force. Information was gathered from hardwarevendors, system integrators, Federal ContractResearch Centers, and industry trade associa-tions. In addition to investigating ADP sup-pliers, the ADP acquisition practices used byseveral firms in procuring fairly large ADP sys-tems were investigated. Details of the industryreview are presented in Appendix C.
(3) The Data Synthesis and Analyses Led to the Devel-opment of Study Conclusions
The data analysis employed a structured integra-tion and interpretation of the qualitative and quanti-tative data from the three major data collectionactivities.
Quantitative data were subjected to statisticalanalyses, including correlation and regression anal-ysis, to explore the relationship of key factors invarious types of acquisitions. Critical Path Method(CPM) techniques were employed to analyze case studydata.
Qualitative data obtained from the literaturereview and Government and industry surveys providedkey insights into the acquisition process and the~critical interactions of individuals and organiza-tions. An understanding of the constraints imposed byexternal regulatory environment, gleaned from thepolicy review activity, was vital to the analysis offeasible alternatives for improvement.
1-4
Based on review and interpretation of the find-ings integrated from all sources, the perceived prob-lems were postulated as issues requiring further anal-ysis. To verify, amplify and document valid issues,the investigators thoroughly reviewed the source mate-rial from the Government, literature and industry sur-veys. Each significant reference to one or more ofthe postulated issues was summarized or recorded ver-batim on an individual file card. The citation cardswere then sorted by issue area. The end result was 33clearly delineated issues, supported by a number ofcards pertinent to the issue -- a clear audit trailback to original source data.
Detailed issue papers were prepared, based on thelarge volume of citations drawn from the data base ofinterviews and library resources. Analysis of theissues focused on synthesizing the numerous specificfindings into a subset of basic conclusions as to theunderlying causes of the pervasive problems in the AF
ADP acquisition process. Study efforts concentratedon those issues that represented deficiencies correct-able by the Air Force. Analyses revealed that theroot cause of each of the problems was the violationof one or more principles requisite to effective man-agement and efficient acquisition. Development ofconclusions was, therefore, directed toward describingthe implications of these violations for the processand its organizational participants. Exploring theseimplications prepared the study group for the formula-tion and analysis of alternative improvements thatwould incorporate the missing principles into theprocess.
(4) The Study Conclusions Led to the Development ofReeommendation Strategies
Following verification of the inadequacies of theADP acquisition process traceable to the violation ofmanagement principles, a series of analyses was ini-tiated. These analyses were designed to identifyalternatives and to develop significant, workablerecommendations. Recommendations were evaluated basedon their feasibility for implementation, projectedbenefits and costs, and contribution towards achieve-ment of the project objectives of reducing lead timeand technology lag and improving the ADP requirementsgeneration and approval process. The development ofthe recommendations emphasized the derivation of asimplified process and modification of policies, pro-cedures, and organizational roles and responsibilitiesrequisite to implementation of the revised process.
1-5
Al
(5) The Final Results of the Study Are Documented inThis Report
This report contains the study findings, conclu-
sions, recommendations, and implementation strategies.Chapter II deals with the regulatory environment and
presents a detailed description of the current pro-cess. Chapter III is an evaluation of the findings,
including a performance evaluation and analysis of the
issues upon which the conclusions rest. The studyconclusions are presented in Chapter IV, and the
recommendations in Chapter V. Chapter VI outlines the
approach recommended to implement the requisitechanges within the Air Force. The Appendices, A-G,
contain supporting data from the specific research
tasks, as well as the five acquisition case studies.
1-6
~1
j
I
I II. THE ADP ACQUISITION PROCESS
*.*~~-.' -
II. THE ADP ACQUISITION PROCESS
This chapter provides the foundation for presentationof conclusions by describing the ADP Acquisition Processin terms of the policy environment and the. specific ac-tivities comprising the process. The chapter is dividedA into the following two sections:
* The Policy Environment* The Process Description.
Each of these foundation elements is discussed in detailto provide a background for the remainder of the report.
1. THE POLICY ENVIRONMENT IS THE REGULATORY STRUCTURE ANDORGANIZATIONAL HIERARCHY IMPACTING THE PROCESS
The first part of this section discusses the regula-
tory structure, and the organizational roles are describedin the second part.
(1) Federal ADP Policy Has Its Foundation in PublicLaw; It Is Implem-ented Through Directives at theOSD Level and Regulations Within the Air ForceA Air Force ADP acquisition is governed by a multi-
tude of laws, executive orders, circulars, regula-tions, bulletins, and standards issued at the Federal.level, as well as implementing directives and regula-
nj tions developed by OSD and the Air Force.
The Federal procurement system has evolved into acomplex interdependent series of functions, circum-scribed by a proliferation of documents. In additionto two basic statutes, there are approximately 4,000other individual statutory provisions addressing pro-curement, with unique requirements governing ADP. Indeveloping the Federal Procurement System (formerlythe Uniform Procurement System), the Office of FederalProcurement Policy (OFPP) recognized the existence ofthis nonintegrated body of procurement law and non-uniform procurement procedure as a principal contribu-tor to the complexity of the procurement process.Figure II-1, a diachronic presentation of the policyenvironment, illustrates the sequence and relation-ships of the documents discussed here.
1950 1352 464 1356 11531143 1151 1353 1355 1067
FIGURE II-i 1TUM OW
DIACHRONIC CHARTHOOM _____________210 NOOWI
ccmmsmu A usa~
141 AMEb.AIrS
FEDERM R0f TYAND ADUVUTEY EIUDUAtI o1ff USERVcS ACTOF1141 PIKn.6 I ACTOF I
.1- / ....
04 1356 1953 19 1382 1364 1956 -168 197 172 1376 97 1373 1910 1332 19
1 1355 1957 1153 1961 1983 1315 1367 1363 1971 1373 1375 1377 1373 1331 1933 11
EISENHOWER KENNEDY JammSON NimO FORD- CARTER REAMDA
A-
AMENDMENTS
P.L 89 306. PL 9. 8 P . f 51BROOK ACT__ P. L. 34U. OF"P ACT OF" PAPEMWONU
- LUAU EDUCTIONTHIN EENIf NEONGS*ZATION N ACT
NO I ACT Of 19"
RERANIZATION S
HEARINGS
-71-71
a A-S4 SELECTIONI AND ACQUISITION
I A-Il
I GUIDELIES
I A-27AI INTRAARD INTERSAGENCY SAING
A-76 POLICIES F' C GCOMUa RL OR 0UUSTRIAPRODUCTS AND SERRIES
PAMPHLETN00..2
The policy environment within which the AF ADPFprogram functions includes the following:* Public Laws
* Executive Orders
* Office of Management and Budget Circulars
* Federal Procurement Regulations and FederalProperty Management Regulations
* Defense Acguisition Regulations
. Office of Federal Procurement PolicyPublications
* Federal Information Processing Standards
* Department of Defense Directives andInstructions
* Secretary of the Air Force Orders
* Air Force Regulations.
These documents are described in depth in thefollowing paragraphs.
1. There Are Two Basic Statutes Providing theFramework for Government Contracting andSeveral Key Laws Specifically Related to ADPAcquisition and Management
The two basic statutes governing contractingand providing authority to issue regulations arethe following:
Armed Services Procurement Act of 1947
. Federal Property and AdministrativeServices Act of 1949.
The Armed Services Procurement Act appliesto the procurement activities of the DOD, whilethe Federal Property and Administrative ServicesAct primarily governs the civil agencies. How-ever, the Brooks Bill was passed in 1965 asamendment to Title I of the Federal Property Act,adding a section on ADPE, and encompassing DODactivities. Each of the basic statutes has beenamended without regard to the other, resulting in
11-3
I7
~1 numerous inconsistencies, the effect of which hasbeen magnified through the issuance of implement-ing regulations and procedures.
Since 1965, the following laws have beenenacted related to some aspect of ADP acquisitionand management:
*PL 89-306:1 . PL 93-400
* PL 96-83* PL 96-511.
Public Law 89-306, commonly referred to asthe Brooks Bill, was enacted Oto provide for theeconomic and efficient purchase, lease, main-tenance, and utilization of automatic data pro-cessing equipment by Federal departments andagencies." The law established a structuralframework and centralized procurement authorityto ensure the wise expenditure of ADP funds inthe face of rapidly increasing ADP use. Thebasic authorities of 0MB, GSA, Commerce, and theIuser agencies relative to ADP are outlined.
The bill was enacted as an amendment to theFederal Property and Administrative Services Act
of 1949. Legislative hearings and reports indi-cate that the intent was neither a marked depar-ture from previous government policy, nor theadvancement of new policies. PL 89-306 simplycalled for efficient and economic acquisition ofADP, and defined organizational responsibilitiesrequired to achieve that goal. Unfortunately,basic assumptions of 1965 technology have madethe law's implementation somewhat less than
A: effective in the face of a changing technologicalenvironment. The thrust of implementation hasunduly emphasized hardware procurement and pricecompetition.
Public Law 93-400, the office of FederalProcurement Policy Act, passed in August 1974,directed the creation of the Office of FederalProcurement Policy (OFPP) within 0MB, to 'provideoverall direction of procurement policies, regu-lations, and forms." With the passage of thislaw, Congress declared its policy to be the pro-motion of economy, efficiency, and effectivenessthrough the establishment of policies, proce-dures, and practices for acquisition of propertyand services *of the requisite quality and withinthe time needed at the lowest reasonable cost,
11-4
utilizing competitive procurement methods to thei--ximum extent practicable.0 The law seemed totike a rational approach to the relative priorityof fulfilling a need versus achieving competi-tion, as goals of the procurement process.
Public Law 96-83, the Office of Federal Pro-curement Policy Amendments of 1979, stated thatOFPP should "provide overall leadership in thedevelopment and implementation of procurementpolicies,u and instructed the development of awUniform Procurement System." The law also ex-pressed a subtle shift in priorities from the1974 OFPP Act. This amendment declared thatCongress' policy was to, first, promote the useof full and open competition, and second, estab-lish policies and procedures to ensure the acqui-sition of requisite quality goods and services,within the time needed, at the lowest reasonablecost. The goals of procurement, as stated in theproposed Uniform Procurement System, reflect thispriority, referencing the mandate of P.L. 96-83.
Public Law 96-511, the recently enactedPaperwork Reduction Act of 1980, legislates theauthority of OMB to "develop and implement Fed-eral Information policies, principles, standards,and guidelines ... and oversee the acquisitionand use of ADP, telecomunications, and othertechnology for managing information resources.w'The law establishes the Office of Information andRegulatory Affairs to serve as a focal point forleadership and central direction of Federalinformation resource management. It incorporatesthe P.L. 89-306 objectives of economy and effi-ciency while consolidating OMB's ADP responsibil-ities under P.L. 89-306 with information manage-ment functions, and indicates an underlying shiftaway from equipment management towards informa-tion management.
Implementation-requires active agency par-ticipation and presents an opportunity for theAir Force to be instrumental in structuring pol-icy to improve information resource management.The Air Force is an agency under P.L. 96-511, andis provided a direct link for communication withOMB to resolve ADP acquisition problems, and to-revisit" the Brooks Bill in terms of restruc-tured policy. In defining ADP, Congress specifi-cally excluded that which involves intelligence,cryptologic or command and control activities,and weapons systems or other systems critical todirect fulfillment of the military mission.
11-5
2. Federal Organizational Roles and Responsi-bilities Have Been Directed Through Execu-
-ive Orders
In light of their impact on the ADP policy Ienvironment, the following two executive ordersare of primary interest:
. EO 11717* EO 11893.
Executive Order 11717 was issued on May 9,1973, transferring certain functions from theOffice of Management and Budget (OMB) to the Gen-eral Services Administration (GSA) and theDepartment of Commerce. Specifically, OMB'sfunctions with respect to policy control over ADP(except in relation to standards) were trans-ferred to GSA, and its responsibility for approv-ing ADP standards was transferred to the Depart-ment of Commerce. OMB maintained general man-agement oversight and fiscal authority and,according to the interpretation of the JusticeDepartment, responsibility for ADP Opolicy formu-lation.0 (see ref. 1) EO 11717 was partiallysuperseded by EO 11893 in 1975, by which certainpolicy functions were transferred from GSA backto OMB. While these included most of thosetransferred to GSA in 1973, the EO did nottransfer ADP policy control responsibilities backto OMB.
The effects of these two orders are dis-cussed in subsequent portions of the report.
3. OMB Communicates Policy and ProceduralGuidance Through the Issuance of Circulars
A number of OMB circulars have applied tocertain aspects of ADP management. These include:
* OMB A-54OMB A-71
* OMB A-76* OMB A-109.
On October 14, 1961, OMB (then Bureau of theBudget) issued A-54, which prescribed policiesfor making selections of equipment to be acquiredfor use in the ADP program of the executivebranch, including determination as to whether the
11-6
ADPE to be acquired would be leased, purchased,or leased with an option to purchase. This cir-cular was superseded by GSA Federal ManagementCircular 74-5 in 1974.
0MB Circular A-71, Responsibilities for theAdministration and Management of ADP Activities,was published on March 6, 1965. The circulardefines the responsibilities of executive agen-cies for the administration and management of ADPactivities in an attempt to ensure maximum coop-.1 eration and coordination. Transmittal MemorandumNo. 1, dated July 27, 1978, amplifies that por-tion of A-71 dealing with responsibilities andrequirements for a Federal Computer Security Pro-gram. Since 0MB had in fact been fulfilling such
:1 a role for ADP since 1959, A-71 recognized OMB'srole as one of overall leadership and coordina-tion. GSA was directed to assist in the achieve-I ment of increased cost effectiveness in theselection, acquisition, and utilization of ADPE,and the Department of Commerce (National Bureauof Standards) would provide technical advisoryassistance and support the development ofstandards.
0MB Circular A-76, published March 3, 1966(and Transmittal Memoranda No. 1, 2, and 3)
establishes the policies for acquiring commercialor industrial products (not exclusively ADP) forgovernment use. An August 30, 1967 revision(Transmittal Memorandum No. 1) clarified someprovisions of the original circular and lightenedthe Federal agency workload in implementing itsprovision~s. The basic policy set forth is reli-ance an the private enterprise system to supplythe Government's needs, except where it is not inthe Government's best interest. Specific circum-stances under which the Government may provideproducts and services directly win the nationalinterest" are described.
0MB Circular A-109, Major Systems Acquisi-tions, dated April 5, 1976, describes policiesand procedures to be followed in the acquisitionof major systems by Federal agencies. Systems,as it used here, includes data processing equip-ment. Problems have arisen in the implementationof A-109 due to the absence of GSA proceduralguidance, the difficulty in adapting the circu-lar's provisions to ADP acquisitions, and theapparent conflicts between the A-109 approach andPL 89-306. These issues are analyzed in subse-quent sections of the report.
11-7
4. GSA Issues Both Federal Procurement Regula-tions (FPR) and Federal Property ManagementRegulations (FPMR) Governing ADP Acquisition
The Federal Procurement Regulations areissued by GSA under the authority granted by theFederal Property and Administrative Services Actof 1949. The FPRs are the governing regulationsapplicable to the acquisition of items or ser-vices required by Federal agencies. The FPR isI the civilian agency equivalent of the DefenseAcquisition Regulations (DAR), and only certainspecialized sections, including those applicableto data processing and ADP acquisitions, are man-datory for use by DOD. The relevant material isPart 1-4, subpart 1- 4.11, which was added to theFPR in October, 1976. Recent changes to the FPRindicate a willingness on the part of GSA ADTS toaddress the criticism of counterproductive,
if' overly restrictive policy and procedure, and tomove towards increased user agency autonomy anddiminished hardware focus in ADP acquisitions.
Federal Property Management Regulations(FPMR) are also issued by GSA to provide specificguidance in the management and operation ofequipment and activities. The sections relevantto ADP and Telecommunications Management are101-35, 101-36, and 101-37, which are embodied inSubchapter F: ADP and Telecommunications.Responsibility and authority for the FPMR aredelineated in the Federal Property and Admin-*1 istrative Services Act of 1949. Substantivechanges to the FPMR, Subpart 101.35.2, were made
concurrent to the recent FPR revisions.
5. The Defense Acquisition Regulations (DAR)Are Issued by the Department of Defense asthe Principal DOD Procurement RegulationSystem
The DAR applies to all elements of theDefense Department and governs all DOD purchasesand contracts for products, systems, and ser-vices. It provides detailed regulatory guidanceon all aspects of acquisition, but is subordinateto the FPR in relation to ADP. There is somestandardization between the DAR anti the FPR, butthere are many inconsistencies in definitions,procedures and forms. The proposed FederalAcquisition Regulation (FAR) will replace the FPRand major sections of the DAR, offering the pros-pect of uniformity and clarity in acquisitionregulations.
11-8
6. OFPP Pamphlets Are Published to Assist Agen-cies in Understanding the Intent and Appli-cation of Policy
OFPP is the central source for procurementpolicy direction under the authority granted byP.L. 93-400 and P.L. 96-83. OFPP pamphlets offerguidance to agencies in applying OMB circulars totheir activities. They are not official or bind-ing implementing documents. OFPP Pamphlet No. 1of August 1976 was prepared by OMB and OFPP toenhance understanding of the intent and applica-
tion of the policies contained in OMB CircularA-109. OFPP Pamphlet No. 2 (Draft) of October1978 was prepared jointly by GSA, OFPP and OMB toassist Federal agencies in applying A-109 to theacquisition of major ADP/telecommunications sys-tems. No final version of this pamphlet has beenreleased. In its Uniform Procurement System Pro-posal, OFPP has recognized ADP acquisition pro-cedures to be sufficiently unique so as not to besubsumed by the UPS, apparently leaving revisionsin such procedures to GSA ADTS.
7. The Federal Information Processing Standards(FIPS) Are the Official Publications Relat-ing to Federal Standards and Guidelines
In order to carry out the Commerce Depart-ment's ADP responsibilities under PL 89-306, theNational Bureau of Standards (NBS) provides tech-
nical advice and coordination of standards devel-opment through its Institute of Computer Sciencesand Technology. The FIPS Publication Series pro-vides general guidelines for numerous specificfunctions related to ADP, such as benchmarking,management of multi-vendor plug-compatible sys-tems, standardization, and security.
8. DOD Directives and Instructions ImplementFederal Policy and Provide Direction to TheMilitary Departments
The DOD Directives and Instructions relevantto the acquisition and management of ADP aredescribed in the following paragraphs.
DOD Directive 5000.1 - Major SystemsAcquisitions, dated March 19, 1980.
This directive is the third issuance of DODD5000.1. The first issuance in 1971 was prior tothe publication of OMB Circular A-109. The
11-9
January 18, 1977, revision implemented the pro-visions of A-109 within DOD. The applicabilityof this directive was for "Major System Acquisi-tion Programs" designated by the Secretary of
Defense. For programs not designated as majorsystem acquisitions the directive was to be usedas a guide. In the March 18, 1980 revision, thisstatement was changed to read "the principles inthis directive should also be applied, whereappropriate, to the acquisition of systems notdesignated as major."
DOD Instruction 5000.2 - Major SystemsAcquisition Procedures, dated March 19, 1980.
This instruction provides the supplementaryprocedures for implementing DOD 5000.1.
DOD Directive 5000.29 - Management of Comp-
uter Resources in Major Defense Systems,dated April 26, 1976.
This directive establishes policy for man-agement and control of computer resources whichare embedded in major defense systems. It estab-lishes the Management Steering Committee forEmbedded Computer Resources at the OSD level tooversee the implementation of the regulation intothe Defense system acquisition process. Thisdirective explicitly excludes general purpose,commercially available ADP assets that are admin-istered under DODDs 4105.55 and 5100.40.
DOD Instruction 5010.27 - Management of 0Automated Data System Development, dated
November 9, 1971.
DODI 5010.27 establishes uniform guidelines
systems administered under DODD 5100.40.
DOD Directive 5100.40 - Responsibility forthe Administration of the DOD Automatic DataProcessing Program, dated August 19, 1975.
DOD Directive 5100.40 establishes the DODADP Program and assigns responsibilities pursuantto OMB Circular A-31. It designates the Assis-tant Secretary of Defense (Comptroller) as theSenior DOD ADP Policy official and directs theSecretaries of the Military Departments to 'des-ignate a senior ADP policy official to administerthe DOD ADP Program within the organizationalelements under their respective jurisdictions.'
11-10
DOD Directive 7920.1 - Life Cycle Managementof Automated Information Systems (AIS) datedOctober 17, 1978.
This directive establishes joint technicaland functional policy governing the life cyclemanagement and control of AIS, including fullconsideration of functional, ADP, and telecom-munications requirements. It applies the prin-ciples of DODD 5000.1 and OMB Circular A-109 toall major AIS, and designates ASD(C) responsiblefor integrating and unifying the AIS managementprocess within DOD. It further calls for thedevelopment and maintenance of functional, ADP,and telecommunications plans within each DOD com-ponent. The directive governs only those AISsutilizing ADPE encompassed by DOD Direc-tive 5100.40.
DOD Instruction 7920.2 - Major AutomatedInformation Systems Approval Process, datedOctober 20, 1978.
DODI 7920.2 establishes the review and deci-sion process and procedures for implementation ofa major AIS. (Note: The scope of applicabilityfor both DODD 7920.1 and 7920.2 has been in ques-tion at both OSD and Air Force levels.)
9. Secretary of the Air Force Orders DelineateADP Roles and Responsibilities Within theAir Force
The two principal orders pertaining to theADP program are the following:
SAFO 100.1SAFO 560.1.
Secretary of the Air Force Order (SAFO)100.1 dated 27 May 1977 delineates the authorityand the responsibilities of the Secretary, UnderSecretary, and the Assistant Secretaries of theAir Force. It designates the Assistant Secretaryfor Financial Management (SAF/FM) as the AirForce Senior ADP Policy Official. It makes himresponsible for administration of programs fordesign, improvement, and standardization of auto-mated data systems, and also for the selection,acquisition, management and use of Automatic DataProcessing Equipment and associated software(ADPE/S). It designates the Assistant Secretaryfor Acquisition and Logistics (SAF/AL) as the AirForce Acquisition Executive.
II-ll
Secretary of the Air Force Order (SAFO)560.1 dated 4 December 1978 outlines specificresponsibilities of the Assistant Secretary ofthe Air Force for Financial Management (SAF/FM)as the Air Force Senior ADP Policy official, pur-suant to SAFO 100.1 It authorizes the SAF/FM toconduct the Air Force ADP Program and prescribesthe delegations of management authorities, and
*1 responsibilities. The order also designates the.2 Hg USAF Director of Computer Resources as the Air
Force ADP Program Single Manager and delegates tohim authority to approve ADP projects and expen-diture of resources below specific cost thresh-olds.
10. Air Force Regulations Implement DODDirectives and Instructions
~1 Acquisition policy is promulgated in the AirForce by two major bodies of regulations. The AF300 series regulations, which are the principal
4 focus of this study, apply to general purpose,commercially available ADP resources, arid imple-ment the policies evolving from PL 89-306. TheAF 800 series regulations apply to major defensesystems and implement the provisions of
DODD 5000.1. The following is a summary of therelevant Air Force regulations:
AFR 300-2 - Data Automation: Managing theUSAF Automated Data Processing Program,dated April 24, 1980.
AFR 300-2 establishes the Air Force ADPprogram and prescribes policies and responsibili-ties for managing ADP resources. In addition itdefines the ADP approval threshold values and therelationships with other AF regulations.
AFR 300-6 - Automatic Data Processing Re-source (ADPR) Management, dated July 11,1980.
AFR 3D0-6 implements the provisions of DODDand DODM 4160.19 establishes requirements andcontains guidance for managing ADP resources.
11-12
AFR 300-7 - Automatic Data Processing (ADP)Planning, dated February 11, 1977.
AFR 300-7 provides guidance for preparingand submitting AF ADP plans, Projected AutomationRequirements (PARs), Major Command ADP Plans(MCAPs), ADPS Master Plans (AMPs) and the AirForce ADP Plan (AFADPP).
AFR 300-12 - Procedures for Managing Auto-matic Data Processing Systems; Volume I -Documentation, Development, Acquisition andImplementation, dated September 12, 1977,and Volume II - ADPS Management, datedSeptember 2, 1977.
AFR 300-12 provides the procedures for man-
aging ADPs. It defines the requirements fordocumentation, milestone reporting, and hardwareand software acquisition management. Volume I
applies to all ADPSs and ADPS elements within thepurview of AFR 300-2, and Volume II identifiesUSAF ADPS managers for standard ADPS and presentsspecific management functions and procedures forgoverning individual ADPSs.
AFR 300-15 - Automated Data System ProjectManagement, dated January 16, 1978.
AFR 300-15 defines and outlines the method-ology for managing automated data system (ADS)projects. It provides guidance on organizing,planning, developing and maintaining an ADS. Itdescribes the configuration management, reviewand reporting, quality assurance, and test man-agement procedures to be followed in an ADSacquisition.
AFR 800-2 - Acquisition ManagementAcquisition Program Management, datedNovember 14, 1977.
AFR 800-2 implements DODD 5000.1 and DODI5000.2 and states the policies for managing allAir Force acquisition and modification programsfunded either through procurement appropriations,through the Security Assistance Program, orthrough the RDT&E appropriation. It specifiesthe roles and responsibilities for managing majorsystem acquisitions within the Air Force.
11-13
APR 800-14 - Volume I: Management ofComputer Resources in Systems, datedSeptember 12, 1975, and Volume II:Acquisition and Suport Procedures forComputer Resources in Systems, datedSeptember 26, 1975.
APR 800-14 establishes policy for the acqui-sition and support of computer equipment and com-puter programs employed as dedicated elements,subsystems or components of systems defined underAPR 800-2. Though APR 800-14 was published priorto DODD 5000.29, it embodies the DOD policies setforth in the latter.
(2) The Federal ADP Management Structure Is Comprisedof a Number of Organizations Which Are Responsi-ble for Policy Development and Implementation,and ADP Oversight Management
In addition to the intricate network of laws,circulars, directives and regulations governing theprocess, there exists an equally elaborate organiza-tional structure. There are numerous organizations inthe ADP acquisition hierarchy that interact with oneanother in the development and implementation of pol-icy. Each organization wields formal influencethrough defined roles and prescribed authority, whilesome wield greater informal influence in the practicalfunctioning of the process. These organizations,which direct and control the user agencies' attemptsto fulfill their ADP requirements, are interdependentin carrying out the Federal ADP Program.
The organizations forming the regulatory environ-men~t include:
* Congressional Committees
* office of Management and Budget (0MB)
* General Services Administration (GSA)
* Department of Commerce, National Bureau ofStandards
* Assistant Secretary of Defense (Comptroller)(ASD(C))
* office of the Secretary of the Air Force.
Each organization is discussed in detail in thefollowing paragraphs.
11-14
1. Congressional Committees Influence ADPAcquisition Through Legislative Development,Interpretation, and Oversight
Congress has broad legislative and oversightauthority over the functions of the Federal agen-cies and departments. Specific committees areassigned responsibilities in monitoring selectedactivities. These committees exert their influ-ence not only through the passage of legislation,but also through subsequent interpretation andoversight management. In ADP acquisition, prob-lems have arisen due to varying interpretationsby the relevant committees and the resultantissuance of contradictory committee guidance.
* Implementing agencies have attempted to respondto the conflicting direction, but have not alwaysbeen able to act in accordance with the commit-tees' intent. Perceiving deficiencies in perfor-mance, some committees have intervened in ADP
* acquisition and management.
The key committees influencing the Air ForceADP acquisition process are:
* The House Government OperationsCommittee (HGOC)
* Senate Committee on GovernmentalAffairs (SCOGA)
1MHouse and Senate Appropriations
Committees (HAC and SAC)
House and Senate Armed ServicesCommittees (HASC and SASC).
OMGSA, OFPP, OSD and the militarydepartments must interface with these committeesin carrying out ADP programs. major policy and
Ph. guidance must be submitted by executive agenciesfor committee review; budgeted programs must besubmitted for authorizations and appropriations.ADP acquisitions are reviewed in connection withthe implementation of PL 89-306 and the oversightof OMB's and GSA's ADP activities. The commit-tees communicate their philosophies, recommenda-tions, and instructions in the course of theirpolicy making, budgeting, and review activitiesthrough committee reports.
Congress has become increasingly involved inADP acquisition and management since the passageof the Brooks Bill. The congressional interest
11-15
that spawned this legislation was largely a re-sult of numerous GAO reports documenting problemsin ADP and calling for centralized management.Since PL 89-306, GAO has not diminished itsefforts, but has continually reminded Congress ofthe ineffective implementation of the law throughthe issuance of at least 200 separate reports.Tensions between the legislative branch and theexecutive branch agencies charged with PL 89-306implementation have mounted to the point that theHGOC's involvement has been termed "interference*Iwhich impedes executive branch decision-making(see ref. 2 p. 3 and ref. 3 p. 51).
HGOC has become the de facto manager of ADPacquisition, exercising their influence throughthe GSA APR mechanism, often creating delays and Ifrustration by requesting holds on acquisitionsI late in the cycle. Congress believes its exten-sive scrutiny and oversight are warranted in theface of abdication of responsibility by 0MB, GSA,and the DOC, and the failure of agencies to ef-fectively manage the ADP resource.
The HGOC and the FIAC have disagreed on theoptimum method to achieve the P.L. 89-306 goalsof economy and efficiency in acquisition of ADP.This conflict has resulted in contradictorydirection to the user agencies, and subsequentconfusion. The HGOC has stressed the achievementof maximum hardware competition, the MAC hasfocused on the achievement of lowest total costover the system's life cycle. Recently, in theFY 1982 DOD Authorizations Report, the SASC ex-pressed its intention that the DOD should acquireADP, taking account of lowest total overall costto the maximum extent feasible.
There is no doubt that Congressional in-volvement will continue until the agencies havedemonstrated their management competence to thesatisfaction of Congress.
2. 0MB Has Broad ADP Policy and Fiscal Author-ity as Well as Oversight Responsibility
As a cabinet-level component of the execu-tive branch, OMB is responsible for fiscal policyand general administrative management. In theADP arena, 0MB has been assigned specific respon-sibilities under PL 89-306, and these haverecently been reaffirmed under PL 96-511, thePaperwork Reduction Act of 1980. However, over
11-16
the years, policy roles have shifted between 0MBand GSA, creating considerable confusion in theADP community and hampering both agencies in thefuliilment of their prescribed roles.
OMB's policy and fiscal authority has gen-erally been communicated through the issuance of0MB circulars. 0MB first instituted a programdesigned to achieve better management of ADPtechnology within the Federal Government throughcentralized leadership in 1959, and administeredthis program through issuance of circulars andbulletins until the passage of PL 89-306 inOctober 1965.
* By that time there had been dramatic in-7 creases in the use and cost of ADPE. These
factors, in conjunction with at least 100 GAOreports in the previous decade, aroused theI interest of Congress, specifically the Chairmanof the HGOC. The legislative result was PL89-306, which emanated from the Brooks Bill (HR4845) and passed Congress as an amendment to theFedekral Property and Administrative Services Act
* of 1949.
The Brooks Bill conferred Ofiscal and policycontrol" (PL 89-306) authority upon OMB (then theBureau of the Budget). While the law incorpor-ated most of the provisions of earlier OMB circu-lars, it failed to amplify the role of 0MB,focusing primarily on GSA's role. In contrast,Circular A-71 had Outlined a broad role for 0MBas the lead executive branch agency with respon-sibility to "provide overall leadership and co-ordination of executive branch-wide activitiespertaining to the management of automatic dataprocessing equipment and related resources...develop programs and issue instructions ... for theselection, acquisition and utilization' of ADP(BOB Circular A-71).
While PL 89-306 was intended to enable GSAto exercise government-wide authority in the areaof procurement policy, and develop regulations toassure government-wide compliance, its ambiguousdelineation of GSA's role vis-a-vis OMB's,indeveloping "go-vernment-wide policies for improv-ing the management" of procurement (1949 Act),has detracted from the effective performance ofboth organizations.
11-17
In 1973, Executive Order 11717 effectivelydivested OMB of its policy control functions withrespect to automatic data processing, leavingintact OMB's fiscal and oversight responsibili-ties, by shifting the policy functions to GSA.OMB functions related to the establishment ofstandards were assigned to the Department of Com-merce. The apparent intent of the order was toreinforce GSA's "overall leadership" role, asdefined in the 1949 Act and subsequent amend-ments, in developing government-wide policy forADP. While OMB was not relieved of its policyoversight and formulation responsibilities, theEO nonetheless appeared to expand GSA's role asan ADP regulatory authority in relation to OMB.
Meanwhile, OMB had acted in a way so as toenhance the perception of its diminishing author-ity in the ADP area. OMB ceased the issuance ofmajor ADP policy guidance subsequent to CircularA-71 in 1965. In response to the enactment of PL89-306, OMB in 1966 provided "guidance" to GSAand NBS with respect to the provisions of the Act.
In 1975, certain policy functions that hadbeen transferred to GSA under Executive Order11717 were transferred back to OMB under Execu-tive Order 11893. ADP policy authority, however,remained with GSA. Neither this 1975 EO nor its1973 predecessor clarified the extent of GSA'srole in "developing" policy based on OMB's "form-ulation," in the terminology of a Justice Depart-ment decision related to EO 11717.
Until 1975, ADP policy oversight responsi-bility within OMB resided in the Information Sys-tems Policy Division. After the creation of OFPPin 1975, the relative roles of OFPP and OMBInformation Systems Policy Division were un-clear. The Office of Federal Procurement Policy(OFPP) was established within OMB pursuant to PL93-400, which mandated that OFPP "provide overalldirection of procurement policies, regulations,procedures, and forms" (PL 93-400). OFPP wascreated in response to recommendations of theCommission on Government Procurement (COGP),which specifically and urgently suggested"immediate consideration" by the President, and
establishment by law, of a "central Office ofFederal Procurement Policy in the ExecutiveOffice of the President, preferably within theOffice of Management and Budget," to exercise"leadership toward effective management of theprocurement function" (Summary of COGP, p. 5-6).
11-18
The Commission noted that the Office 'should beindependent of any agency with procuring respon-sibility, operate on a plane above the procure-ment agencies, ... have directive rather thanmerely advisory authority" (COGP, p. 5) andshould have 'responsibility for policies' relatedto acquisition (COGP, p. 71). These descriptionsof OFPP's role in "exercising leadership" (COGP)and providing "overall direction to procurementpoliciesu (PL 93-400) are strikingly similar tothe role described for GSA in the 1949 Act andsubsequent amendments (including PL 89-306).
This has posed some questions as to OFPP'srole in relation to GSA/ADTS, concerning ADPacquisition. However, OFPP has conceded, in itsUPS proposal to Congress in October, 1980, thatADP acquisition procedures are Ounique." Whileit is not entirely clear, it would appear thatADP acquisition policy is the responsibility ofGSA/ADTS as per P.L. 89-306 and E.O. 11717.
Most recently, PL 96-511, the PaperworkReduction Act of 1980, yet to be implemented,states that the Director of OMB 'shall developand implement Federal information policies, prin-ciples, standards, and guidelines, and shall pro-vide direction and oversee ... the acquisitionand use of automatic data processing, telecommu-nications, and other technology for managinginformation resources.* The ADP and telecommuni-cations functions of the Director shall alsoinclude "monitoring the effectiveness of, andcompliance with, directives," and providing1advice and guidance on the acquisition and use"of ADP and telecommunications equipment. Thisbroad role in AD? policy development, implemen-tation, and oversight is stated to be consistentwith the authority conferred by PL 89-306,neither increasing nor decreasing the authorityof GSA, OMB, or Commerce.
In hearings on the Paperwork Reduction Bill,the Senate expressed its intention not to expandGSA's ADP authority under P.L. 89-306, but ratherto consolidate all policy-setting, coordination,and oversight related to information managementin a clearly defined focal point, specifically,the Office of Information and Regulatory Affairs(OIRA) in OMB. Despite statutory provisions ofauthority to OMB, GSA has impacted Federal ADPacquisition policy in practice since the passageof the Brooks Bill.
11-19
3. GSA Has Played a Key Role and Exerted Con-siderable Influence on the Development andImplementation of Both Procurement and ADPPolicy
The General Services Administration (GSA)has broad policy and operational authority overthe acquisition and management of governmentI property and services. GSA's influence in theADP acquisition environment has been expandedthrough statute, executive and legislative branchintervention, and through the assumption ofresponsibilities seemingly abdicated by OMB since
1965.
GSA was established in 1949 largely onrecommendations of the Hoover Commission. TheFederal Property and Administrative Services Actof 1949 implemented the Hoover Commission'srecommendation for an independent general ser-vices agency, designating that GSA should "pro-
* vide for the Government an economical and effi-cient system for (a) the procurement and supplyof personal property and nonpersonal services
i..; (b) the utilization of available property;(c) the disposal of surplus property; and (d)records management." (Fed. Property Administra-
tion Services Act of 1949, Sec. 2.) The Adminis-trator was given the power to prescribe policies
which functions should be performed by GSA and
which should be delegated.
*1 BOB Circular A-71 defined GSA's responsibil-ity in the administration and management of ADPactivities as "aiding in the achievement ofincreased cost effectiveness in the selection,acquisition and utilization" of ADP. OMB'sintent was to charge GSA with "providing for con-trol of Federal equipment acquisition costs andequipment proliferation and duplication," but notto involve the Administration in *the effectiveutilJization of operating agency ADP resourcesonce they were acquired" (see ref. 3).
In 1965, PL 89-306 authorized the Adminis-trator to "coordinate and provide for the eco-nomic and efficient purchase, lease, and mainte-nance" of ADP, subject to the fiscal and policycontrol of 0MB. Congress appears to have envi-sioned the achievement of economy and efficiencythrough centralization of management within GSAand the provision by the agency of technicalexpertise lacking in the user agencies.
11-20
GSA's role under PL 96-511 is one of anoperational manager subject to OMB's direction.While a central management role in informationresources is stressed, the Senate language in thebill, and the specific wording in the law, dis-claim an expansion of GSA's authority, as definedin PL 89-306. There is some hesitancy (under-
* standable in view of past events) to accept thisdisclaimer for fear that an excessively broadinterpretation and implementation could effec-tively expand GSA's authority.
GSA communicates its policy primarilythrough two series of regulations. The FederalProperty Management Regulations (FPMR) apply toall government departments, including DOD. GSAADP policy was initially published in Part 101.32of the FPMR. In 1974, FMC 74-5 consolidated ADP*1 policy into one directive, and became the prin-cipal implementing document for PL 89-306. Thetenets of FMC 74-5 provided the basis for currentprocurement regulations and procedures.
Subsequently, a subpart added to thV FPR inOctober 1976 set forth policies and proceduresgoverning the procurement of all ADP, includingby DOD (GSA FPR Subpart 1-4.11). This resultedin the coexistence of two overlapping bodies ofFederal ADP issuances originating within GSA, inaddition to the DAR, which has created a gooddeal of confusion. A notable inconsistency isthe definition of competitive vs non-competitiveprocurements. The presence of adequate price'I competition qualifies as competitive with commod-ities other than ADP in the FPR, and in the DAR.However, the DAR must comply with the FPR provi-sions regarding ADP acquisition, and the FPRdefines a requirement for ADP as non-competitiveor solesource "not withstanding adequate pricecompetition," if it is based on-*specific make ormodel specifications.
Within GSA, the Automated Data and Telecom-munications Service (ADTS) is authorized to per-form GSA's ADP functions under the Birooks Bill,as amended. ADTS was established within GSA byAdministrative order on August 15, 1972, combin-ing the ADP functions of the Federal Supply Ser-vice with the telecommunications functions of theTransportation and Communications Service. Theorder tasked ADTS with "the direction and coordi-nation of a comprehensive government-wide program
11-21
for the management, procurement, and utilizationof automatic data processing and communicationsequipment and services" (see ref. 4 pp. 3-15).
GSA's ADP acquisition activities, asdirected by ADTS, have been closely monitored bythe House Government Operations Committee (HGOC).HGOC maintains that this extensive oversight iswarranted, whereas the dissenting view is thatthis intervention into ADTS daily decision-makingis "impeding Executive Branch management with fewcommensurate benefits" (GSA Eval.). In any case,it is clear that the HGOC has exercised consider-able influence in its oversight role, regardlessof its lack of statutory authority over the pro-curement process. This is evidenced by the factthat "there is no recollection of GSA proceedingwith a delegation or with one of its own procure-ments over the opposition of Government Opera-tions Committee" (see ref. 3 p. 50).
Additionally, ADTS has never exercised theauthoritative leadership role envisioned by theBrooks Bill, and ADTS relations with customeragencies have been fraught with conflict. Thecenter of ADP technical and procurement expertisehas not been developed. The length of timerequired to obtain a DPA from ADTS for equip-ment procurement has been cited as a cause oftechnological lag in the dynamic field of ADP(GSA Eval. p. 63).
Within ADTS, the management of ADP and tele-communications has been carried out separately,despite the increasing technological merger ofthese two fields. Fragmentation of organiza-tional responsibility among a variety of rela-tively independent elements has resulted in"unnecessary duplications, inefficiencies, and alack of coordinated approach" (GSA Eval. p. 61).ADP policy groups reside within the ADTS Officeof Agency Assistance, Planning, and Policy and ata branch level within the Procurement Division ofthe Office of Automated Data Management Services.
In 1973, when EO 11717 increased the scopeand complexity of GSA's management functions, theOffice of Federal Management Policy (OFMP) wasestablished to assume the policy functions trans-ferred from OMB. Five subordinate officesfocused on the functional areas of procurement,property, ADP, financial and management systems.While OFMP was disbanded in response to the EO11893 retraction of most of the functions that
11-22
had been shifted from OMB, a small ADP policy'1 group remained to exercise the ADP functions thatthe latter EO had not affected. The responsibil-ities of this ADP Management Policy and PlanningFunction were never clearly defined.
GSA has been heavily criticized for itsinability to cope with its dual procurement andADP responsibilities. Since the delineation ofauthority between OMB and GSA has been vague,there has been a lack of centralized policydirection and planning coordination. In addi-tion, GSA has interpreted FMC 74-5 very narrowlyconcerning software conversions, reflecting theinfluence of the HGOC. Although the circularstates that conversion costs are a consideration,GSA's interpretation has precluded the considera-tion of all of the costs that need to be includedin the analysis in order to achieve lowest totaloverall cost (LTOC) in acquisition.
The practical focus has been on maximizingcompetition in order to achieve the lowest hard-ware cost, an objective which is not alwayssynonymous with the achievement of the lowestcost over the system's life cycle. This positionhas created problems for user agencies who, as aresult, must orient their Agency ProcurementRequests (APR) so as to ensure maximum free andopen competition, whether or not this achievesLTOC. An opposing viewpoint is advocated by theHAC; they reject the argument that maximum hard-ware competition will ensure optimal economy andefficiency, and assert that the agencies shouldstrive for LTOC. The SASC has recently expressedits support for the HAC position in the FY82Authorizations Report.
Recently there has been progress in relationto DOD efforts to free the Department from GSAregulatory involvement in the fulfillment of itsmission. The SASC has recommended statutory lan-guage which insures that regulations governingDOD ADP acquisition for critical defense missionsare limited to those applying to other DOD pro-curement under Title 10 of the U.S. Code. Theamendment to Section 2315 USC Title 10 statesthat no other provisions of law shall be appli-cable to DOD ADP procurement if the function,operation, or use of such ADPE or services in-volves intelligence, cryptologic, or command andcontrol activities, is integral to a weapon sys-tem or is critical to direct fulfillment of
11-23
'~ .. ~
military or intelligence missions. This exclu-sion effectively removes these categories of ADPfrom the requirements of PL 89-306, and is con-sistent with the exclusions in PL 96-511.
GSA has made recent efforts toward improvingits performance and facilitating the smootherfunctioning of the ADP acquisition process.Changes have been made in the FPR's and FRMR's
*1 Subchapter F which 'represent a major step towardcomputer acquisitions at lowest total overallcost' (see ref. 5). Revisions include aclarification of lowest overall cost to includeconversion, a definition of competitive procure-ments that allows for 'maximum practicable' ver-sus the more restrictive 'fullw competition, andsubstantial increases in agency procurementauthority for ADPE under competition.
Additionally, there have been organizationalchanges within ADTS, including the creation of anoffice of Planning and Policy to manage both ADPand telecommunications. The organization isMcommitted to accenting its regulatory role andstrengthening its management areas,' and therehas been a significant effort to delegate pro-Axcurement authority in more instances (see ref. 3
4. The Secretary of Commerce Provides FederalAgencies With Scientific and TechnologicalADP Advisory Services
The Secretary of Commerce was assigned arole in the administration of the Federal ADPProgram by OMB Circular A-71 in 1965. A verysimilar role was legislated for Commerce underPL 89-306 to improve the utilization and effec-tiveness of computer systems. In 1973, EO 11717transferred to the Secretary all functions of OMBrelated to the establishment of government-wideADP standards, including their approval on behalfof the President.
The Department of Commerce, National Bureauof Standards (NBS), is responsible for the Fed-eral Information Processing Standards (FIPS) Pro-gram. This role for NBS involves the developmentand issuance of standards related to ADP equip-ment, techniques, and computer languages, amongother subjects, some of which have constrainedthe efficient acquisition of ADP. NBS, through
11-24
its Institute for Computer Sciences and Tech-nology, provides leadership, technical guidance,and coordination of government efforts in thedevelopment of technical guidelines and standardsfor computer systems.
5. ASD(C) Interprets and Implements Federal ADPPolicy for DOD
DOD Directive 5100.40 establishes the DODADP Program and assigns responsibility for theADP Program pursuant to 0MB Circular A-71. TheAssistant Secretary of Defense (Comptroller),(ASD(C)), was assigned responsibility for admin-istration, with delegation within ASD(C) to theDirectorate of Data Automation (DDA). DDA is anorganization within the Office of the DeputyAssistant Secretary (Management Systems) estab-lished in response to the PL 89-306 and OMB A-71requirement for an ADP policy and managementfocal point with each agency. DDA is not a majorinfluence in ADP acquisitions since its specifiedfunctions do not include a direct role in theacquisition and approval process.
DDA integrates and implements Federal ADP~1 policy through the issuance of policy and direc-
tives applicable to OSD, JCS, military depart-ments, the defense agencies, and the unified andspecified commands. DDA oversees the implementa-tion of policies and plans within the DOD ADPProgram and monitors major ADP system developmentprojects.
While it has minimal formal influence overmilitary departments' acquisitions, the Director-ate functions as a central point of contactwithin DOD, where problems among the departmentsor between DOD and GSA, 0MB, or congressionalcommittees can be resolved. This rol~e is imple-
4 mented by DOD Directive 4105.55. The functionsof DDA are performed by three independent teamswith total staffing of 12 individuals. DDA'sproblems in responding to conflicting executiveand legislative ADP policy direction are dis-cussed in the Conclusions, Chapter IV.
6. Assistant Secretaries of the Air Force Exer-cise the Statutory Authority Assigned andDelegated by the Secretary
The Secretary of the Air Force, pursuant to10 U.S.C. 8012, is responsible for and has theauthority necessary to conduct all affairs of the
11-25
Department of the Air Force. Subject to his con-trol, the Under Secretary and the Assistant Sec-
* retaries of the Air Force are authorized to actfor and with the authority of the Secretary ofthe Air Force in matters within the areasassigned. The specific functions for these indi-viduals are detailed in Secretary of the AirForce Order 100.1. The two Secretaries of theAir Force (SAF) discussed in connection with ADPare the following:
Assistant SAF for Acquisition andLogistics
Assistant SAF for Financial management.
The Assistant Secretary for Acquisition and*1 Logistics (SAF/AL) is the Air Force AcquisitionExecutive. He exercises that authority in theareas of research and development, weapons andweapons systems development, test, evaluation,production and contract management, procurementactivities, and supply matters. While managementauthority for the Air Force ADP Program has beendelegated by the SAF to SAF/FM (as discussed
~1 below) contracting authority for ADP remains withSAP/AL.
The Assistant Secretary of the Air Force forFinancial management (SAP/FM) is the Air ForceSenior ADP Policy official and is responsible forthe overall policy, management, and administra-tion of the Department's ADP program; this is inaddition to his overall responsibilities for AirForce budgeting and finance. His ADP responsi-bilities include the selection, acquisition, man-
-r agement, and use of automatic data processingequipment and associated software (ADPE/S). TheSAP/PM has been delegated the authority toapprove ADP projects, ADP specification, andexpenditure of ADP resources.* He also is thesource selection authority for ADP acquisitions.He -may, however, choose to delegate this author-ity. Authority to contract for ADP and to issuepolicy direction governing this authority is del-egated through SAP/AL to the Air Staff, and maybe further delegated from there.
i7 Major system acquisitions, as determined under DODDirective 5000.1, will be processed in accordance withthat directive.
11-26
The SAF/FM has been appointed Senior IRM
Official under the requirements of PL 96-511.
2. PROCESS DESCRIPTION IS PRESENTED IN TERMS OF THEPARTICIPANTS AND THEIR ACTIVITIES
This section of the report presents a description ofthe Air Force ADP Acquisition Process, and outlines it interms of the specific activities which must be accom-plished, the relationship and sequence of those activi-tiqs, and the participating organizations. The primarysource is the Air Force 300-Series Regulations including:
AFR 300-2 Managing the USAF Automated DataProcessing Program, dated 24 April 1980
AFR 300-7 Automatic Data Processing (ADP)Planning, dated 11 February 1977
AFR 300-12 Procedures for Managing Data Process-ing Systems (ADPS) Volume I Documentation, Devel-opment, Acquisition, and Implementation, dated12.September 1977; and Volume II ADPS Management,dated 24 April 1980
AFR 300-15 Automated Data System ProjectManagement, dated 16 January 1978.
The ADP Acquisition Process is described in terms of:
• Its Scope
* The organizational Participants• The Process Steps• Processing Thresholds
• Management of the ADP Program.
These elements are discussed in detail in the remainder ofthis chapter.
(1) The AF 300 Series Acquisition Process EncompassesAll General-Purpose ADP Resources Except ThoseAcquired in Combat Weapons Systems and SpeciallyDesigned Equipment
In order to understand the scope of the ADPAcquisition Process, it is necessary to define anumber of terms. The following definitions are takenfrom 300-2 (urless otherwise noted):
USAF ADP Program - The sum of all objec-tives, plans, policies, directives, proce-dures, and criteria establishing and per-taining to:
11-27
- The acquisition, use, and management of
ADP resources.
- The design, development, control,improvement, and standardization ofautomated data systems.
* ADP Resources - The totality of ADPE, soft-ware, computer programs, ADP contractualservices, ADP personnel, and supplies.
* Automatic Data Processing Equipment (ADPE)General purpose, commercially available ADPcomponents and systems created from themwhich were not specially designed (asopposed to configured) for any specific
application.
* Combat Weapons System - An instrument ofcombat, either offensive or defensive, usedto destroy, injure, or threaten the enemy.
. Commercially Available (see ref. 6) -Offered for sale to the general publicand/or industry at established catalog ormarket prices.
Specially Designed (see ref. 6) - Govern-ment specified and not commerciallyavailable. Excludes specially configured.
Acquisition Program (see ref. 6) - Adiretedeffort funded either through
procurement appropriations, through theSecurity Assistance Program, or through theResearch, Development, Test and Evaluationappropriation with the goal of providing newor improved capability in response to avalidated need. Excluded from thisdefinition are general purpose, commerciallyavailable ADP assets.
A major source of confusion is in the decision toapply AFR 300-2 or AFR 800-14 procedures to the man-agement of acquisition programs. The possibility forconfusion is depicted by the overlap illustrated inAFR 300-2, Attachment 4, shown in Figure 11-2. As anexample, AFR 300-2 does not apply to computer equip-ment which is "integral to or in direct support" ofcombat weapons systems; however, according toAFR 800-14, Volume II, paragraph 1-7, ADP resources insystems acquired under AFR 800-2 may be subject to
11-28
ccp
> L17
Z c IM
CA 0
0 - C0-0 c~ o
cl4 5 -
F,: 0 ~ O O - a.
<~C 00-n(n 0E~ B9 CE E c
o 4 0r .0 7C 6 0.-0 O i - 0 0' C CLo- : OC <
E z a -
0 4c 0 a .Lo~~ > a__ _ _ _ _ _ _ E
ZH E-1 c E ~? o 62
0 C3EW C)
U- C) CD
E mC C)
0 )C) C>
C)0 w04Ftt z co c 5 u44~~ 0 i- 0C0' 04
cc
11-2
APR 300-2 procedures to the extent specified in thePMD. The source of guidance for making the decision
:1 to be included in the PMD is unclear.
(2) The Senior ADP Policy Official for the Air ForceIs the Secretary of the Air Force for FinancialManagement
The Air Force ADP Management Structure, depictedin Figure 11-3, is under the policy direction of theSecretary of the Air Force for Financial Management.AFR 300-2 outlines the mission of the organizationswhich support the AF ADP Program. These organizationsare described in the paragraphs that follow.
'I1. AF ADP Program Single manager Is the Direc-
tor of Computer Resources (USAF/ACD)I Under the direction of the Comptrollerof the Air Force, the AF ADP Program SingleManager has been delegated managementauthority for the Air Force ADP Program. Assingle manager, the Director of Computer
Resources is responsible for:
* Developing, coordinating andissuing plans, policy guidance,and procedures governing the ADP
* Planning, programming andbudgeting of Air Force ADPresources
. Reviewing and approving Data Auto-mation Requirements, and directing
4 and monitoring ADP acquisitions
* Performing liaison for the AirForce with GSA and other FederalGovernment authorities relative toADP procurements
Providing ADP technologicalexpertise to the Air Force.
2. Hg USAF Staff Offices Have the FunctionalArea Responsibili-ties Under the Deputy andAssistant Chiefs of Staff
Staff offices are located in thefollowing organizations:
11-30)
>
0(A
q
(A2
Q 0 (
z -0L
4c
UU
E-I U-]'
USAF/RD - Research, Developmentand Acquisition
USAF/XO - Operations, Plans andReadiness
• USAF/IN - Intelligence
• USAF/MP - Manpower and Personnel
• USAF/PA - Programs and Evaluation
USAF/LE - Logistics andEngineering.
Functional staff offices have responsibilityfor developing, reviewing, staffing andvalidating functional requirements. Inaddition, some staff offices such as RD, LEand MP provide review, coordination andadvice on ADP programs in other functionalareas.
3. Major Commands (MAJCOM) and Separate Oper-ating Agencies (SOA) Conduct the Studies andAnalyses Which Must Accompany the DAR toObtain Approval
To conform with these requirements theMAJCOMs and SOAs must possess expertise intwo principal areas -- the functional areato which automation is being applied, andADP. For this reason two separatemanagement chains exist under theorganization commander:
ADP Program Single Manager pro-vides ADP support for the commandmission. The Office of the SingleManager provides support to thefunctional users in developing ADPrequirements and is responsiblefor planning, acquiring, andmaintaining ADP support.
Functional ADS Managers providethe functional area expertise forthe particular application(s)being automated. (An ADS (Auto-mated Data System) is defined asan assembly of procedures, pro-cesses, methods, routines, ortechniques specifically designedto make use of ADPE.)
11-32
4. USAF and Command ADPS Managers Are Appointedfor Command-Unique and Standard ADPS
An ADPS (Automated Data ProcessingSystem) is defined as an aggregation ofsoftware and the resources required tosupport it (ADPE, manpower, and facilities).ADPS managers are appointed by the appropri-ate organizational level depending uponwhether the system is command unique or USAFstandard ADPS.
5. Air Force Communication Command (AFCC)Provides Specialized ADP Support Services
The Deputy Commander for Data Automation isresponsible for managing AFCC resources to pro-vide ADP support to Air Force activities. Thesubordinate elements of AFCC include:
Air Force Computer AcquisitionCenter (AFCAC) is the centralacquisition and selection officefor the Air Force.
Air Force Data Systems DesignCenter (AFDSDC) analyzes, designs,develops, tests, implements, andmaintains ADPS's that are assignedto it by the Director of ComputerResources, HQ USAF. These include:
Base Level Data AutomationStandardization Program
- USAF MAJCOM ADP Program
- Base Level Management System
- Standard Base Supply System.
Air Force Data Services Center(AFDSC) provides ADP services tothe Air Staff, Office of the Sec-retary of the Air Force, Office ofthe Secretary of Defense, andother Federal Agencies.
11-33
I
Federal Computer PerformanceEvaluation and Simulation Center(FEDSIM) develops and maintainsprograms, models, and techniquesfor simulating and analyzing ADPsystems and equipment for all Fed-eral agencies. The Air Force isthe executive agent for GSA in 1operating FEDSIM.
Communications Computer Program-ming Center (CCPC) provides com-puter programming services for thecomputers of the AUTODIN communi-cations systems.
Air Force Data Systems EvaluationCenter (AFDSEC) provides indepen-dent quality assurance assessmentsof ADPS during the system lifecycle.
San Antonio Data Services Center(SADSC) provides ADP services forthe San Antonio region and for anyother government organization thatcan tie into the Center.
Phase IV Program Management Officehas program management responsi-bility for the ongoing acquisitionof the system to replace the Sec-retariat Standard Base Supply Sys-tem and the Base Level System. It
also has been assigned programmanagement responsibility for theInterservice Automatic MessageProcessing Equipment (IS/AMPE)program.
6. ADP Review Board Reviews the Air Force ADPProgram and Advises the SAF/FM on Major ADPPolicies and Projects
It is chaired by SAF/FM and composed ofthe other Assistant Secretaries, the General
Counsel, the Comptroller of the Air Force,and the Director of Contracting and Acquisi-tion Policy. The Director of Computer Re-sources is the Executive Secretary.
11-34
7. Data Automation Panel (DAP) -The DAP con-ducts the Air Staff Board Structure program-ming and budgeting review of the Air ForceADP Program.
(3) The ADP Acquisition Process Starts With the Iden-tification of a Requirement for ADP and Ends Withthe Implementation of an Operational Capability
For clarity, it is useful to separate the processinto the following phases:
* Requirement definition* Requirement approval. Technical specification development* Procurement
A Installation and test.
~1 These acquisition phases are discussed in thefollowing paragraphs. For each phase, the specificactivities are defined in terms of the information anddocumentation requirements, the decision processes,and the participating organizations. A flow chart ofthe generalized acquisition process, described in theAF 300-Series body of regulations, is depicted in
Figure 11-4 (see page 11-38).Requirement Definition Consists of the Identification,originating organization
Specific activities included in this phase are:
Defining the Functional RequirementDefining the ADP RequirementConducting studies and analyses
Preparing the Data Automation Requirement.
Requirements for ADP capabilities or services mayoriginate at any organizational level within the AirForce. In order to proceed with an ADP development oracquisition program the functional requirement must bevalid~ted by the functional area authority and the ADPsolution approved by the designated ADP approvalauthority. The vehicle for obtaining this approval isthe Data Automation Requirement (DAR).
Before a DAR is initiated, the functional areaspecialists must establish the economic justificationand feasibility of alternative solutions. If exploi-tation of automation is the selected alternative, theresults of these analyses must be documented in theDAR.
11-35
Defining the ADP acquisition process required theestablishment of a consistent starting point whichcould be traced from acquisition to acquisition. Thetime at which an ADP requirement is first definedwould seem to be a logical point. However, there aredifficulties in consistently identifying such an event.
Ideally, requirements for automation are definedas a normal part of the planning process and are docu-mented two to seven years in advance. Requirementsare documented in a Projected Automation Requirement(PAR), funds are approved, and work begins on estab-lishing a program to acquire the ADP capability. Inmany cases, however, unplanned needs arise or existingneeds change, causing the requirements generation andapproval process to be accelerated. In some in-.1 stances, ADP acquisitions are not traceable to a PAR,Aor if they are, the PAR is much broader, coveringagreater number of ADP capabilities.
The second starting point considered was theprepaiation of the DAR. A DAR Package consists of thefollowing documentation:
Executive Summary of the DARI * Data Automation Requirement (DAR)
* Statement of Requirements (if telecommuni-cations support is required)
* Feasibility Study
Economic Analysis
Executive Justification Summary
* Draft Data Project Directive
Draft Agency Procurement Request
Statement of work and GSA Form 2068
* ADP Telecommunications Requirements Checklist
ADPE List and Sole Source Justification (ifrequired)
* System Development Notift.cation (if changeto existing WWMCCS stand r d systems isrequired
Software Conversion Study 'Uf required).
11-36
The DAR documentation is initiated when the analysesof the originating organization clearly indicate thatthe requirement is mission essential, and that ADPresources are the most economical alternative solu-tion. Once these analyses are complete, the resultsmust be included in the DAR. Thus, it is difficult toascertain precisely when the studies and analysesactivity ends and when the DAR preparation activitybegins.
Due to these difficulties, for the purpose ofthis study (and consistent with previous ones) thecalculations of acquisition lead times started withthe receipt of the DAR package at the Air Staff forprocessing and approval.
Requirement Approval Starts With The Staffing of theDAR Package at the MAJCOM or SOA ADP Program SingleManager
If the final approval authority is above theMAJCOM/SOA level, the DAR is submitted to the AF ADP
Single Program Manager, USAF/ACD, for approval. Spe-cific activities included in this phase are:
* Approve DAR
Obtain Delegation of Procurement Authority(DPA)
* Issue Data Project Directive (DPD)
• Prepare/Coordinate Data Project Plan (DPP)
Upon receipt in USAF/ACD, the DAR package isassigned to an action officer for initial review. Theaction officer examines the package to ensure that itis properly formatted; that all reviews, certifica-tions and coordinations have been completed; and thatthe DAR is consistent with AF plans and policies. Ifthe DAR is deficient, either it will be returned tothe originating office for reaccomplishment, or theaction officer may request the originator to provideadditional information and rework the package at theAir Staff.
Once the initial review is completed, the actionofficer removes the draft Agency Procurement Request(APR) (or the GSA Form 2068) and sends the remainderof the package to the Air Staff functional OPR forreview, evaluation, and Air Staff coordination. Thefunctional OPR validates the functional requirementand obtains coordination from the appropriate officesincluding:
11-37
APPROVE COORDINATE APPROVE
SOFTWARE *APR/PREPARE APR/ISSUEGSA CONVERSION DRAFT DPA OPA
STUDY
tNO-
IAPPROVE DAR YES PREPARE/
I ARAPPROVED? ISSUE
* FESISIITY DAV ~ AANLYESOI APPROVEAP
IES ILT R-p A
I INO
PREPARE CONDUCT) ID
DRAFT FD SRRAPRVD
.OVE
IISSUE
IOPA
IF'UINICIT,
PREPARE/ REVIEW ISSUE DPA 1OISSUE APPROVE
PODPP
lB
IA
PROGRAM 'sMANAGER EPREPARE L
DPP I
,NIEII
II
NO I r----------
YES +EDPREPARE PREPARE PREPARE CONDUCT
FINAL FD DRAFT RD DRAFT SS DR
I/(I
_ __ _7
rn
10
ITIEIDIs
IE
1o
IN
IE
NO REPRE I PREPARE
NOPEAR RPAE -- ---- A OPEAR
CONDUCT 85 YES CONDUCT SPECSRSPRVESPC FINAL ADPE
PDR APPROVED? SPECS
PREPARE I PEPARE
FINAL RD DRAFT PNT
D
IEIt
/I
ABBREVIATIf
APR-AGENCY PROCUREMENT REQUEST
CDR-CRITICAL DESIGN REVIEW
DAR-DATA AUTOMATION REQUIREMENT
DPD-DATA PROJECTIVE DIRECTIVE
DPP- DATA PROJECT PLAN
DS -DATA BASE SPECIFICATIONDT -DEVELOPMENT TEST PLAN
FCA-FUNCTIONAL CONFIGURATION AUDIT
IOC-INITIAL OPERATIONAL CAPABIUTYMM-MAINTENANCE MANUAL
OM -OPERATIONS MANUAL
PREPARE PREPAREFINAL AOPIE
SPECS DRAFT PS
DRAFT DS
DRAFT AOMI DRAFT UM
DRAFT MM
It..
ABREIAIN EGN FIGURE
QUEST PCA-PIYSICAL CONFIGURATION AUDITPOR-PRELIMINARY DESIGN REVIEW
REMENT PS - PROGRAM SPECIFICATIONSIWEPT -TEST PLANI
PVR-PRODUCT VERIFICATION REVIEWIRD -DATA REQUIREMENT DOCUMENT IPRFP-REQUEST FOR PROPOSAL IR
INd AUDIT SAR-SYSTEM REQUIREMENT REVIEW tokIITY SS -SYSTEM SPECIFICATIONS ID
SVR-SYSTEM VALIDATION REVIEW 'UUM-USERS MANUAL 'c
IL
CONDUCT I A CPRELIMINARY IEFA
PCA IS
I CO
CONDUC IV I SSECONDUCT IN
NOELVERNARY IECATIO I
II
IFIGURE 11-4, ADP ACQUISITION PROCESS
I WORK FLOW DIAGRAM
'PIRloID'U
CONDONDCT
PRELIMINARY I IA CFCAA
CONDUCTUC
PRELIMINARYPCA IE
CONDUCT I.ENVIRONENTAL/PVRSYTMTT
CONDUCTIACCEPTANCE/IDEVELOPMENTI
TEST EII
No VERIFI- L ------- Cm--- -- -- CATION PEE
PREPARE o
FINAL UM11-38
Program Element Monitor(s)
USAF ADPS Manager(s)
* USAF/ACDX-ADP Policy and Security
• USAF/ACDM or ACDS-ADP Technical Evaluation
* USAF/MPME-Manpower (if manpower resources,manpower benefits or contractor services are
involved)
* USAF/ACMM-Management Analyses (for review ofthe Economic Analysis)
USAF/XOKC - Communications, Command and Con-trol (if there are associated communicationsrequirements)
USAF/XOKS - World wide Military Command and
Control System (if WWMCCS SDN's are involved)
• USAF/DAAD(s) - Privacy DocumentationA USAF/MPPT - Formal or Contractor Provided
Training
* USAF/LEXY - Logistics Policy and Procedures
* USAF/LEEP - Engineering Support
* USAF/IN - Intelligence Requirements
• USAF/RDCS - Contracting and ManufacturingPolicy
* AFDSDC - Air Force Data Systems Design Cen-ter (if base level and WWMCCS related re-quirements are included)
• HQ AFCC/CDO - Headquarters Air Force Commu-nications Command (DPD tasking to an AFCCADP Center, if appropriate).
After the package is fully coordinated, recom-mended changes are received by the AF/ACD actionofficer from the functional OPR and they are resolvedand incorporated into the final package by the actionofficer. The action officer then completes the DARapproval package which includes:
t
i II-39
Staff Summary Sheet - An executive summaryof the action recommended, requesting USAF/ACD signature on the DPD or on the letter toSAF/FM, depending upon the approvalauthority.
Executive Justification Summary - A synopsisof the DAR, usually six pages or fewer.
• Data Project Directive (DPD)
Letter to SAF/FM requesting approval if theDAR is above the USAF/ACD approval threshold.
The DAR approval package must again be sent to thefunctional OPR to be reviewed by the same Air Staffoffices that reviewed the DAR package. When com-pleted, the approval package is returned to the actionofficer to present for signature. The DAR is approvedeither by the SAF/FM signing the approval letter, orby the Director of Computer Resources signing theDPD. If the SAF/FM is the approval authority, a CaseSynopsis is prepared and presented in lieu of theStaff Summary Sheet.
DAR approval is signified by issuing a Data Proj-ect Directive (DPD) to the originating organizationfor the implementation of the required ADP resources.The DPD documents key decisions; assigns responsibili-ties for management, design, development and procure-ment; defines the project scope; and authorizes spe-cific actions to be taken in satisfying ADP require-ments approved in the DAR.
The first action required of the office assignedprogram management responsibility in the DPD is toprepare a Data Project Plan (DPP). The DPP describesthe actions that must be taken to complete the projectwithin the specified cost and schedule constraints andprovides planning documentation to assist the projectmanager in the development, acquisition and implemen-tation of ADP capabilities. Once prepared, the DPP issent to the USAF/ACD action officer where it receivesthe same coordination as did the DAR package. Whenthe Air Staff review and coordination is completed,the DPP is sent to the Director of Computer Resourcesfor approval.
All Air Force Agency Procurement Requests (APR's)are processed through USAF/ACD regardless of the DARapproval authority. For those cases which are abovethe MAJCOM/SOA approval threshold, the APRs are
11-40
detached from the DAR package and processed concur-rently with the DAR coordination to expedite therequirements approval process.
To process an APR, the USAF/ACD action officermust prepare a proposed synopsis of the APR and atransmittal letter for the Director's signature. ThisAPR package must be coordinated with the followingoffices:
• USAF/ACDX - Policy and Procedures Group
* USAF/ACDR - Resource Management Division
USAF/RDCS - Contracting and ManufacturingPolicy.
Once reviewed and signed, the APR is sent to GSAfor action. GSA, upon receipt, verifies that thepackage is complete and that all required studies,
analyses and justifications have been accomplished.GSA will provide the House Government OperationsCommittee (HGOC) with a synopsis of each APR that hasa high dollar value or which limits competition.Barring requests for additional information, GSA willissue a Delegation of Procurement Authority (DPA) tothe Air Force. For those high priority APR's, theAir Staff may invoke the '20 day clock" underFPR 1-4.1105(b), which states that the requesting
agency may assume that procurement authority has beendelegated if no formal written direction to the con-trary has been received within the 20-day period.
The DPA grants the Air Force the authority toacquire specific ADP resources. The USAF/ACD actionofficer reviews the DPA upon receipt from GSA. Theoriginal DPA remains part of the ACD case file and theaction officer prepares a separate letter, redelegat-ing the procurement authority to the requestingorganization.
Technical Specification Development Defines theRequirement in Either Equipment or Fhtnctional Terms
AFR 300-12, Volume I, Chapter 4, and AFR 300-15establish the procedures for managing ADPS/ADS devel-opment and modifications. The formal specificationdevelopment process, shown in Figure 11-4, is charac-terized by developing a series of different levels ofspecifications, conducting design reviews to approveeach specification, and conducting configurationaudits and verification reviews to test and accept thesystem.
11-41
Most ADP resource acquisitions do not undergosuch a rigorous documentation process. Even at theAFDSDC, the process described by AFR 300-15 for ADSacquisition management is tailored to the specifictasks. Figure 11-5 represents a more realistic viewof the ADP acquisition process, as witnessed duringthe case study and Government survey tasks of theproject. While the requirements definition and2'approval process remain the same as prescribed in AFR300-2, the technical specification development processis simplified considerably.
In reality, the specification development processis driven by the DAR preparation. In order to func-tionally validate the requirements, the user mustfunctionally describe those requirements prior toinitiating the DAR. As the studies and analyses whichmust be documented in the DAR are developed, the useris forced from functionally specifying requirements todefining specific ADP solutions. Thus, by the timethe DPD is issued, the user will often have prepareddraft specifications which are oriented more towardsequipment characteristics than towards functionalrequirements.
The next step is to convert these draft specifi-cations into ones suitable for a Request for Proposal(RFP).
Procurement Begins the Process of Fulfilling the NeedWith the Required ADP Capability
The Data Project Directive (DPD) establishes theoffice responsible for conducting the actual procure-ment of ADP resources. Large competitive procurementsare usually assigned to AFCAC; however, procurementsmay be assigned to local contracting squadrons if theyare determined to be capable of handling the par-ticular procurement. The activities associated withthe procurement phase are:
Develop final specifications* Develop RFP* Issue RFP and receive proposals* Evaluate proposals
Negotiate and award contract.
11-42
UIH
0
CL. tn
HU 0
HO
0
w
11-4
In cases involving AFCAC, the first interchangebetween the user and procuring organization is usuallynot made until the DPD is issued. By this time, theuser has developed draft specifications which areusually too restrictive and confining for the purposesof an RFP. Meetings are held between the user and theprocurement organization to exchange technical infor-mation and to determine the appropriate level ofdetail to be specified in the RFP. At this point, ajoint specification team is formed to prepare thefinal specification.
Besides the specification, additional materialsare prepared in conjunction with the RFP; theseinclude:
* Cost tables and cost evaluation models* Acceptance test criteriaI * Special and general provisions* Data requirements* Instructions to offerors* Conditions and basis of award.
During this time letters of interest and notices forthe Commerce Business Daily will be prepared tosolicit competition.
The final step in the RFP development is to sub-ject the RFP to internal reviews and independent pro-curement and legal reviews. once completed, it isissued. Proposals are received and evaluated againstthe criteria established.
A competitive procurement is a lengthy process.The goal reported by AFCAC for completing a competi-tive procurement totals 15 months, which is brokendown as follows:
Specification Development 4.5 monthsRFP Development 3.0 monthsProposal Development 3.0 monthsProposal Evaluation 2.0 monthsNegotiation and Award 2.5 months.
Installation and Test Activities Conclude theAcquisition Process
As in the case of the Technical specificationsDevelopment Phase, AFR 300-12 and 300-15 prescribe anumber of reviews and audits to be conducted duringthe Installation and Test Phase. What is requiredduring this phase of the acquisition process variesconsiderably depending on the resources being acquiredand the complexity of the procurement. The level of
I1I-44
effort required is specified as part of the contract.General purpose equipment acquisitions usually requirean installation and set-up period followed by a 30-dayacceptance test with documentation requirements lim-ited to what is commercially available from the suc-cessful vendor.
(4) Threshold Values Govern the Level of ApprovalAuthority Required
The following individuals are authorized toapprove various types of acquisition and procurementunder the Air Force ADP program:
The Air Force Senior ADP Policy Official
(SAF/FM)
• The Air Force ADP Program Single Manager
Major Command/Separate Operating Agency ADPProgram Single Managers
• USAF ADPS Managers
Command ADPS Managers, when the MajorCommand ADP program single manager delegatesauthority to them.
The Director of Computer Resources, AF/ACD, hasbeen authorized by the Secretary of the Air Force toredelegate the approval authority granted to AF/ACDand those authorities have recently been redelegatedby AF/ACD to the MAJCOM/SOA ADP program single mana-gers and USAF ADPS managers. Table II-1 summarizesthe approval authority thresholds for various types ofacquisitions under the Air Force ADP program.
(5) The ADP Acquisition Process Is Part of OverallADP Management Control, Which Includes thePlanning, Programming and Budgeting System
Planning for ADP resources is the primary respon-sibility of the Resource Management Division of theDirectorate of Computer Resources (AF/ACDR). ADPplanning, as defined in AFR 300-7, Automatic Data Pro-cessing Planning, is accomplished through planningdocuments which are intended to integrate ADP planningwith both the Air Force Planning, Programming, andBudgeting System (PPBS) as well as the ADP acquisitionprocess. The ADP planning documents are addressed inthe paragraphs that follow. Each is described interms of its purpose and its relationship to the plan-ning and acquisition process.
11-45
TABLE II-I
Air Force ADP Resource Approval Thresholds*
MAJCOM/SOA, USAFADPS MGR Approval
ADP Resource Thresholds ($000)
ADPE (Government or Contractor Acquired)
Sole Source - Lease (annual) $ 200Sole Source - Purchase $ 500Competitive - Lease (annual) $3,000Competitive - Purchase $5,000
Commercial Software Packages
Sole Source - Lease (annual) $ 200Sole Source - Purchase $ 500Competitive - Lease (annual) $3,000Competitive - Purchase $5,000
Contractual Services
Sole Source or Competitive $3,000
Reutilization of ADPE
Government owned $5,000Continued Leased (annual) $ 200Lease to Purchase $ 500
Organic Systems Development (annual) $3,000
*AFR 300-2
11-46
The Air Force ADP Plan (AFADPP) - The ADPPlan is the annual, Hq USAF-prepared plan-ning document which identifies the Air Forceand DOD ADP goals and objectives for thecurrent year and six subsequent years. Itspurpose is to provide planning guidance anddirection to Air Force organizations interms of how the Air Force will employ ADPtechnology to support the AF mission. It isthe initial planning document from which allothers are developed.
ADPS Master Plans (AMP) - ADPS Master Plansare prepared annually by each of the USAFADPS managers. The AMP documents the ADPobjectives for each USAF standard ADPS withguidance from the AFADPP. Projected Automa-tion Requirements (PAR) included in the AMPdocument the detailed programmatic planninginformation. ADP resource requirements inthe AMP are identified by command so thatthe single managers can project requirementsin their MAJCOM ADP Plans.
MAJCOM and SOA ADP Plans (MCAP) - MCAP's areprepared annually by the MAJCOM and SOASingle Program Manager with guidance fromthe AFADPP. MCAPs document the command'sADP objectives and resource requirements andinclude the detailed programmatic planninginformation through incorporation of thePARs.
Projected Automation Requirement (PAR) - APAR identifies an ADP resource and fundingfequirement two to seven years in/advance ofthe need. PARs are submitted to Hq USAF aspart of the AMPs and MCAPs to obtain AirStaff validation and to adjust fundinglevels through the Program Objectives Memo-randum (POM) Process.
This chapter described the ADP acquisition process andthe environment in which it must function. The next chap-ter describes the results of the evaluation of the processand its environment.
11-47
III. EVALUATION OF FINDINGS
K III. EVALUATION OF FINDINGS
This chapter evaluates the performance of the AirV Force ADP acquisition process described in Chapter II.The evaluation results presented herein are subdividedinto two major sections, answering two fundamental
questions:
Is the Air Force incurring damage by the currentADP acquisition process?
If so, what are the causes of this damage?
The following two major sections of this chapter address
these two questions.4 1. A PROCESS PERFORMANCE EVAL~UATION REVEALED THAT THEPROCESS IS UNNECESSARILY LENGTHY AND OFTEN INEFFECTIVE
Two approaches were taken in evaluating the perfor-* mance of the process. The first, a classical process
evaluation, consisted of a statistical analysis of the
acquisitions that go through the process. The statisticalanalysis focused on the dwell times required to completeeach step and the variances in processing time dependingupon what is being processed. This evaluation approachwas restricted to examining the process internally; itsintent was to determine where the process experiences bot-tlenecks. Corrective measures can then focus on theseproblem areas. The quality of such an evaluation isdependent on the availability of aggregate performancestatistics. There were significant voids in the availabledata and consequently, only a partial evaluation was pos-sible using this approach. However, to the extent thatthis evaluation could be carried out, it indicated thatthe variances in processing time are more related to uni-que aspects of the individual acquisitions than to factorsthat can be captured by aggregate statistical analysis.
The second approach to process evaluation began withsupplementing the statistical data with informationobtained from case studies, from the Government and indus-try surveys, and from the literature review. The AF per-formance was then compared to that of private industry inacquiring similar resources. The performance of privateindustry was examined as part of the industry survey. Inthis approach, the process performance was judged against
111-1
an outside level of performance. Whereas the first evalu-ation approach addressed whether some classes of acquisi-tions are processed more efficiently than others, thesecond approach addressed whether the processing of eventhe acquisitions is satisfactory when compared to a stan-dard of performance. The comparison indicated that the A?takes an average of three times as long as private indus-try to acquire ADP. It~was therefore important to deter-
- I mine whether the excessive time expended in the AF acqui-sition process yields commensurate results in the form ofmore nearly optimal acquisitions. This was not found tobe the case. Rather, it was found that the AF incurs aserious technology lag in ADP, that the process oftenfails to satisfy AF functional requirements for ADP, andthat the AF pays an unnecessarily high dollar price forthe ADP resources that it does acquire.
(1) A Limited Classical Process Evaluation RevealsOnly a Slight Correlation Between Process Perfor-mance and the Size or Type of Acquisition
The acquisition process can be partitioned intothe following four phases:
* Phase I - Requirements Generation. Thisphase consists of the identification of arequirement and the preparation of a DAR.Requirements generation primarily occurs ata Major Command (MAJCOM) or at the baselevel.
* Phase II - Requirements Approval. Thisphase consists of DAR approval at the AirStaff and the granting (or refusal) of a DPAby GSA and issuing of a DPD by the HQ USAF.
Phase III - Procurement Cycle. This phaseconsists of the development of specifica-tions, source selection, and contractaward. In most (but not all) cases, pro-curement occurs either at a local contract-ing squadron or by AFCAC.
Phase TV - Installation and Test. Thisphase is measured by the period of perfor-mance of the contract.
Having partitioned the acquisition process into theabove four phases, the evaluation then examined thedwell times required to complete each of these phasesand how the variances in processing time depend uponwhat is being processed.
111-2
--A
AD7A113 ?S55 OOZ ALLEN AND HAMILTON INC BETHESDA MD F/4 15/5WE;;N ADP ACOUISITION STUDY. (U)NOV 81 M0A903-80-C-0A90
UN4CLASSIFIED N
In order to characterize which acquisitionsencounter greater difficulty in the process, a set ofproperties of an acquisition were defined. The pro-perties that characterize an acquisition for the pur-pose of this evaluation are the following.
* Cost - The dollar value of the contract
* Purpose - Indicates if the acquisition is anew system, system replacement, systemupgrade, interim upgrade, modification, or aservice
* Type of Resource - Indicates if the resourceis hardware, software, or services (e.g.,maintenance, time-sharing, etc.), andwhether the resource is purchased or leased
* Type of Procurement - Indicates if the pro-curement is fully competitive, limited com-petition (i.e., brand name or equivalent),or sole source
Procurement Authority - Indicates if theprocurement authority is AFCAC, base levelprocurement, or other (GSA, Systems Command,etc.)
Approval Level - Indicates if DAR approvalis by SAF/FM, USAF/ACD, or by the MAJCOM.
The intent of the statistical analysis is to determinewhich of these properties is highly correlated withunusual delays in one or more phases of the process.Hligh correlations would indicate that the processoperates less well for acquisitions having those pro-perties. In order to perform such an analysis, it isnecessary to collect aggregate statistics that includea representative sample of the acquisitions that wentthrough any part of the process over a representativetime frame.
Much ofl the data proved to be unavailable forthis investigation. Aggregate data were generallyunavailable for the requirements generation andinstallation and test phases of the process. Datawere obtained for the amount of time required for GSkLto issue a DPA, and the time required for AFCAC toconduct the procurement. For the GSA data a breakdowninto cost and procurement type was available. TheAFCAC data included only cost and time. Since the
111-3
samples for GSA and AFCAC represented different setsof acquisitions, it was not possible to track indivi-dual acquisitions through both the approval and pro-curement phases.
Limited available data indicated that the 300-series process is being used for a wide range ofacquisitions. Table III-1 indicates the total numberand average monthly number of DARs received at ACDfrom 1971 through September 1980. This table indi-cates the numbers of ADP acquisitions that require ACDapproval. The percentage breakdown of acquisitionsfrom 1978 to 1980 into procurement types is shown inTable 111-2. This table is based on DPA processingstatistics maintained by AF/ACD. Figure III-i indi-cates the distribution by costs of a sample of 37 pro-curements for which the procurement authority wasassigned to AFCAC. The data on which Figure III-i isbased are derived from a tally of 73 AFCAC procure-ments conducted since 1970.
A statistical analysis indicates that there isonly a slight correlation between the size or type ofacquisition and the time required for the acquisitionprocess. This conclusion is based on an analysis ofthe data from AFCAC, AF/ACDA and GSA (referenceAppendix D).
Figure 111-2 graphs the cost of an acquisitionagainst the time required for AFCAC to process theprocurement. This time represents the number ofmonths from issuance of a DPD to contract award. Thepoints plot a sample of 34 procurements.
The presence of a large number of outliers indi-cated that the model is mispecified. A detailedanalysis of residuals, presented in Appendix D, sup-ports the notion that a cost vs. procurement timemodel is mispecified. That is, cost is not the pri-mary determinant of procurement time.
Figure 111-3 graphs the cost of an acquisitionagainst the time required for GSA to process theacquisition. This time represents the time betweenthe sending of an APR to GSA to the issuance of aDPA. The points plot a sample of 27 procurements.The 13 points labeled Os" were sole-source procure-ments and the 14 points labeled "c" were competitive.Two points, which are properly considered outliers onthis graph, represent an acquisition costing $8 mil-lion and one requiring 175 days. The correlation be-tween cost and time is 0.02, or essentially zero.
111-4
-n R0 n R 11% w0 - 1 0 1ad cai cm C" C4j 4 Wd wd- - 1
-j cc> dcUi
-W C""- N % 10 10 . gq - 1 10 1LL
0 0 - 1 0 N ~ 0 1 0 1 0 P
- P c
C4~~~ ~ ~ ~ ~ ~ an: c c2 L -L r r 4 U
-n qt m ( 4U0, c - L n c r
IL.
1% Nc - 10 I
cc cr. A1 Sr1 0 0 N 1 P 0 1
CD " amT- c
-L CZ!P.
_j co 1 1
be - 10.. 10cc0IS 1 S 0 1
T a,-, -
1US
IS 10C4 NM NrO -" NC4 -T*P1- N- -0 Nm C
r-2 cc10~~~~~~C N 1 -N 10 1 0 1 11 0 1
- N N P N N N - -IL
10 N
10 -A - - -
7z CC -cr >- cc 0
4x >. cc LA uj A .'
RU V)J >0 1 a
10 z .= i C.) 0 WL 0RU -, . U c z 0 -1-.-U, a
111-5
TABLE 111-2
TRENDS IN TYPES OF PROCUREMENTS1978-1980
LIMITED SOLECOMPETITIVE COMPETITIVE SOURCE
% A % A % A
CY 78 58 13 - 29
CY 79 41 -19 25 +12 34 +5
CYU8 48 + 9 29 + 4 29 - 5
3-YEARAVER49% 21% 30%AVERAGE %
CHANGE DOWN UP NO1978-1980 10% 18% CHANGE
SOURCE OF DATA: - OPA STATISTICS 2 JUNE 1980- DATA FOR 1980 EXTRAPOLATED FROM
1 JUNE TO 31 DECEMBER
111-6
III-5
FIGURE 111-1COST DISTRIBUTION OF AFCAC PROCUREMENTS
LESS THAN
400%
GREATER THAN $5M20%
$300K TO $I M
111-7
A
00
C14
0
CD0
2
111-
LU.
wow
I-z
0
0 uju
rn 0CL .0
SHVI0 40 0N11WN 3Nd3s~
4111-9
~INN&.
This result is independent of whether or not the out-liers are included in the calculation. Based on thissample there is a difference in average time dependingon whether the acquisition is sole-source or competi-tive. The average time for all acquisitions is 63days. If the acquisition requiring 175 days isdeleted from the sample the overall average is 58days. The average for competitive acquisitions is 51days. with the one solesource outlier deleted, theaverage for sole-source acquisitions is 67 days.
An additional sample of GSA turnaround times,consisting of 60 acquisitions, was available. Costand procurement types were not available for this sam-ple. Based on this sample the average GSA processingtime is 65 days. This result is consistent with the63 day estimate from the sample described above.
The following conclusions emerge from this sta-tistical analysis:
The average processing time at ARCAC is 16months.
There is a positive correlation between costand processing time at AFCAC. However, costis not the primary determinant in AFCAC pro-cessing time.
GSA typically takes about two months toissue a DPA.
There is no correlation between cost and GSAprocessing time.
Sole-source procurements take approximately30% longer (about two weeks longer) thancompetitive procurements to obtain a DPA.
The data and detailed statistical analysis supportingthese conclusions are presented in Appendix D.
(2) A Comparative Evaluation Reveals That Air ForceADP Acquisitions May Take Three Times as Long asThose in Private Industry
The second approach to the evaluation of processperformance compared the AF acquisition performancewith the performance of private industry. Case study,industry survey and literature review data supplement-ing the statistical analysis revealed that the AF maytake approximately three times as long as industry toacquire comparable resources.
111-10
It is important to note that industry performanceis only a basis for comparison and is not necessarilya goal. There is no assumption that it is possible toachieve industry performance levels or that the stepsthat might achieve this performance level are desir-able when taxpayers' money is being spent. The con-clusion of this section is that, in a different insti-tutional environment, better performance is possible.Put another way, there is nothing intrinsic about ADPthat causes the acquisition process to take as long asit does in the AF. Having established this point, itis then appropriate to examine how the Government'sinstitutional environment and the acquisition processthat occurs in that environment can be changed toachieve improved performance.
This process evaluation is restricted to describ-ing the difference in performance between the AF andindustry. The consequences of this difference in per-formance, described later in this chapter, includedollars unnecessarily spent and needed capabilitiesnot achieved when needed. The reasons fo'r the per-formance differential are described in Chapter IV inthe discussion of the study conclusions. It is onlythrough understanding these reasons that correctivemeasures can be proposed and evaluated.
Performing a comparative evaluation required thecollection of data from the following additionalsources:
Commercial Cases - In order to compare theAF ADP acquisition process with commercialpractice, the industry survey includedgathering data on commercial users of ADP.The data included documentation of a sampleof commercial ADP acquisitions plus aggre-gate performance statistics for two verylarge commercial users of ADP.
* Literature - The literature review includedseveral reports that evaluate aspects of AFADP acquisition (described in Appendix A).
* Industry Interviews - The industry surveyincluded an extensive set of interviews withI, representatives of firms in various aspectsof the ADP industry who have been involvedin 300-series acquisitions. These inter-views resulted in some information relevantto evaluating the 300-series process. Theindustry survey is discussed in Appendix C.
Case Studies - A set of five case studieswere prepared as part of this investiga-tion. Each case study tracked a major ADPacquisition through all phases of the pro-cess that were completed (two of them arestill in the acquisition process as thisreport is being prepared). The case studiesare documented in separately bound volumescomprising in Appendix E.
1. Air Force Acquisition Process Was AnalyzedWith Data From Several Sources
Based on available data, it appears that thetime required to complete an Air Force acquisi-tion varies widely. The Preliminary ADP Acquisi-tion Study, conducted by AF/ACD during the summerof 1979 (see ref. 7), examined the average timeto process AF ADP acquisitions. The reportexamined the process from the time that a DAR isreceived at HQ USAF to IOC (Phases II-IV); therequirements generation phase (Phase I) was notincluded. The results of this study included thefollowing:
Based on a sample of nine acquisitions,the average time to complete Phases IIthrough IV is 24.5 months. The follow-ing is a breakdown of the 24.5 monthsinto the three phases:
- Phase II: Requirements Approval -
7 months
- Phase III: Procurement Cycle -
13.5 months
- Phase IV: Installation and Test -
4 months.
Based on a sample of nine differentprocurements conducted by AFCAC, theaverage procurement cycle time wasfound to be 23 months. This sampleincluded four procurements that AFCACconsidered complex and consequently the23-month average should probably beregarded as an overestimate of the trueaverage.
111-12
I.
As part of the Government survey, represen-tatives at MAJCOMs were queried about the timedevoted to requirements identification and DARpreparation (Phase I). A figure of 4-6 monthswas found to be representative. Combining thisnumber with the results of the Preliminary ADPStudy, an average acquisition process time of29.5 months is obtained. On the other hand, ifthe AFCAC survey is used as a source of data forPhase III, the estimated overall acquisition timeis 39 months. Using the statistics plotted inFigure 111-2, an average procurement cycle timeis 16 months. The sample in Figure 111-2 in-cludes some complex acquisitions, but since it isa large sample, the results are not as severelyskewed as with the AFCAC sample of nine acquisi-tions discussed above.
As part of this study, five test cases ofADP acquisitions were documented (Appendix E).Table 111-3 shows the processing times for fourof the cases for each phase of the acquisitionprocess. These test cases were chosen to berepresentative of the types of problems inherentin AF 300-scries ADP acquisitions. While theyare not cases with extreme problems (such asPhase IV or ALS) they were chosen to illustratesome of the specific problems that can occur inthe process.
TABLE 111-3: PROCESSING TIMES FOR TEST CASES
SAC Global MASIIS/Hybrid Terminals Weather TDSC Average
Phase I 24 25 33 15 24
Phase 11 8 12 33 18 18
Phase III 8 7 3 16* 8
Phase IV 13* 6 3 ii* 8
TOTAL 53 50 72 60 59
* PROJECTED
111-13
I
If complex or difficult cases are dis-regarded, then the average processing time forthe four phases of the AF ADP acquisition processis approximately 31-33 months. If the complexand difficult cases (the AFCAC survey and thefour test cases) are included, then the averageis 55 months. The data indicate that an acqui-sition time of 33 months is a conservative esti-mate for straightforward acquisitions thatexperience little or no problems, while diffi-cult and complex acquisitions can result inacquisition times of 59 months and more tocomplete.
2. Data for Comparing the Commercial Acquisi-tion Process to the Air Force Process WasGathered During the Industry Survey
For a comparable ADP acquisition in the pri-vate sector, the average duration was 8-10months. In order to determine the ADP acquisi-tion performance of private industry, a set offirms that were known to have recently madeacquisitions were contacted.
This survey identified eight specific acqui-sitions by commercial firms that are comparableto acquisitions that the AF would make under the300-series process. In addition, two firms whomake frequent major ADP acquisitions providedaggregate performance data. Data on three com-mercial ADP acquisitions were obtained but notincluded in the sample because they were deemedto be comparable to AF 800-series acquisitions.*
Each of these was a new system for new applications
(i.e., they were not replacements or upgrades). Eachinvolved distributed processing and state-of-the-artdata communications applications. Consequently, theacquisition process involved extensive systems engi-neering work. Despite a 6-8 month time to developspecifications and, in two cases, competitive procure-ment, the average and median acquisition time for thethree cases was 2 years.
111-14
Table 111-4 indicates the type of firm, thedollar value, and acquisition process time (fromrequirement identification to IOC) for the eightcases. In each of the eight cases, data wereavailable on the breakdown of the acquisitioninto separate steps and time spent on each step.Because of inconsistencies in how firms partitionthe process, the requirements identification andapproval phases were aggregated into a singlephase for purposes of comparison. Table 111-5shows the comparison between commercial acquisi-tion and AF acquisition. These data indicatethat commercial firms take an average of 10months to perform a general purpose ADP acquisi-tion. If the Federation of Savings and LoansSocieties is disregarded as an outlying datapoint the average (with a sample of 7) isslightly less than 8 months.
Two firms provided aggregate ADP acquisitionperformance data rather than specific cases. Theinformation provided by the two firms can be sum-marized as follows:
A major telecommunication companyaveraged ADP acquisition performanceover a large number of acquisitions andseveral years. The majority of pro-curements are over $1 million. Theacquisition time varies from 6 to 8months.
A large manufacturing company annuallyprocures approximately $60 millionworth of ADPE. The average acquisitiontime is 8.5 months.
This data supports the ronclusion that industryperformance, in general, results in 8-10 monthacquisition times. Hardware vendor representa-tives who were familiar with industry practicesindicated that the industry average is approxi-mately 8-9 months.
(3) The Current AF Acquisition Process Imposes TwoTypes of Damage on the Air Force: ExcessiveAcquisition Cost and Capability Loss
This section examines two types of damage attrib-utable to the AF ADP acquisition process performance.
111-15
Table 111-4sample Commercial ADP Acquisitions
Dollar Value of AcquisitionAcquisition Time
Type of Firm (millions) (months)
Financial Institution $ 1.0 11.0
Major Bank 11.0 13.0
Energy Exploration .5 3.5
Financial Data Services .5 4.0
Floor Products Wholesaler .8 14.5
Manufacturer .23 6.0
Time Sharing 6.4 2.0
Federation of Savingsand Loan'Societies 25.0 26.0
Average 4.25 10
Table 111-5Comparison of AF and Industry
Phase Industry Average* AF Average
Requirements Identi-fication to Approval 3 months 13.5 months**
Approval to ContractAward 3 months 16 months***
Contract Awardto IOC 3.5 months 4 months**
*Based on data pertaining to the sample in Table 111-4"*Based on Preliminary ADP study
***Based on the data plotted in Figure 111-2
111-16
1. Excessive Acquisition Cost Is Incurred
In the process performance evaluation, itwas concluded that the ADP acquisition lead timein the Air Force is excessive. The longer thannecessary acquisition time imposes additionalcosts on the Air Force, including the significantcost of technology lag.
The average age of computer processors incommercial firms is 2-3 years. In general,industry replaces ADPE after 2-5 years. In theGovernment, the average age of processors is 9-10years. Figure 111-4 from the GAO report on obso-lescence (see ref. 8) indicates that the majorityof medium to large-scale Federal computers wereacquired over 7 years ago. The charts depictedin Figure 111-4 show the age of a sample of 978medium and large-scale computers based on theiracquisition dates (top chart) and their techno-logical age (bottom chart). Based on these fig-ures, the GAO concluded that the Federal inven-tory is old and growing increasingly outmoded.
In addition to decisions to retain oldequipment, the process sometimes results in deci-sions by the Government to buy equipment thatindustry will no longer purchase. That is, theGovernment buys equipment which is obsolete whenit is installed. Figure 111-4 indicates thatmost Federal computers are pre-1965 technology.The GAO reported (see ref. 8) that only 2 percentof the medium and large-scale computers in theFederal Government are using 1975 or later tech-nology. One hardware vendor representativestated, "We can sell things to the Governmentthat no one else will buy." Examples of obsoleteequipment which the Government still buys arecard punches, paper tapes, and IBM 360/65s (suchas those replacing the IBM 7080s at AFLC).
The Air Force is incurring a greater expen-diture of dollars for their ADP capability,because of their use of obsolete equipment, thanthey would incur if the same capability wereacquired using newer equipment. In investigatingthe effects of technology lag, it was found thatnewer equipment not only offered greater capabi-lity, but also resulted in cost savings. These
111-17
FIGURE III - 4
AGE OF 978 LARGE-AND MEDIUM-SCALE FEDERAL COMPUTERS BASED ON THEIRACQUISITION DATES AND TECHNOLOGICAL AGE
500
400 -Qo (332)
S300t.- (253)= (245)
2000
100 10 (34) 12
(2)
PRE-1963 1963-66 1967-70 1971-74 1975-78 JAPRIL1979
ACQUISITION YEARS
500 1478)
400
=€ (308)
300A.
E 2000U (108)
100 (39) (45)
0PRE-1963 1963-66 1967-70 T 1971-74 1975-78 1979
YEAR MODEL WAS FIRST AVAILABLE
TAKEN FROM GAO's "CONTINUED USE OF COSTLY, OUTMODED COMPUTERS IN FEDERALAGENCIES CAN BE AVOIDED"
III-is
cost savings occur not only in the cost ofhardware itself but in many auxiliary costsassociated with the operation of a computer.facility. These auxiliary cost savings are suf-ficiently large that even if functional require-ments remain quantitatively and qualitativelystatic, equipment that is more than 5-7 years oldshould usually be replaced because the replace-ment will, in most cases, result in lower totaloverall dollar costs for the capability.
During the industry survey, and in the GAO'sreport on obsolescence (see ref. 8) many reasonsfor higher cost were cited. The following threecost components were among these:
Maintenance Costs - According to vendorsources for commercial experiences, themaintenance cost of a computer in-creases by about 10 percent per yearfor about the first seven years.Thereafter, costs rise at a signifi-cantly higher rate.
Cost of Operators - A decline in thecost of operators occurs when a smallernumber of computers, and thereforefewer operators, are required. Forexample, at Tinker AFB the work cur-rently being performed by three IBM360/65s can be performed by a singleIBM 3033 or equivalent. In addition,the single computer will accommodategrowth while the three computers aresaturated.
Energy Consumption - Newer computersand peripherals operate on less powerand have lower air conditioning re-quirements. It has been estimated thatcurrent processors can operate on aslittle as 20 percent of the electricityrequired by mid-sixties generationunits with the same relative computerpower (see ref. 8). Approximately thesame savings can be realized on powerrequirements for air conditioning. Fordisc memory units, the per-unit powerconsumption has not changed but thecapacity of units has increased by afactor of 20 in the last 15 years.
111-19
In addition to the three factors cited above, thefollowing factors would result in savings whichcannot be estimated as easily:
* system Reliability - Since newer sys-tems operate more reliably, there arefewer costs due to non-availability ofthe systems.
* Floor space - older systems includemore and larger machines and thereforerequire more floor space.
* Uninterruptable Power Supply (UPS) - IfUPS is needed, then the higher powerrequirements cited above imply greatercost for UPS.
* Personnel - Both the operation andmaintenance of older systems is morelabor-intensive than for newer systemsand consequently more expensive. Inaddition it is becoming increasinglydifficult to find and retain operatorsfor these older systems.
The seven factors cited above indicate that theAF incurs dollar costs from the use of obsoleteequipment. Table 111-6, which is excerpted fromthe GAO report, Continued Use of Costly, OutmodedComputers in Federal Agencies Can Be Avoided,(see ref. 8) illustrates potential savings forfour cases examined by GAO.
An additional factor which was cited by vir-tually all industry and several Government repre-sentatives is the quality of the programmingstaff. In general, functional requirements arenot static. Therefore, software maintenance and(when capacity permits) new software developmentis constantly needed. Quality programmers do notwant to work on obsolete systems. Although it isdifficult to assign a dollar value to the moraleproblems of programmers, the problem of attract-ing and holding competent computer professionalsin the Government is widely recognized.
111-20
0 0~
0~~ a 0 >..
4. 0
fl0 00 .. ~
I- 0
IV2 u~i4 0 0
CL1 - 0.6omo 0
u 0
El) 0 00"fS0 WO0 In 3 0 c.j -J )
OU m 00 o .. .
0~ r II 0 0 Wa 'j 0-.- 4 U 3 n q4 * 0.'. . . 4
I 0 0 . 1-
0 orI -
0~0. 0 -c
0 0)44 000 r' U C M >
o a 0 c0) C' OW2 0
A 0 C C II4
0.0 0
a 0 as u. c 0"
C' 1 Vcc0 .n . 4
a 0 > 0 0. .. . 0. 0 0 -- 071 ~ a- .a6 - 1 6- c
- 0 .22 0. u it U2 c c z o U4 (1;2e -0U '0 a .4t
CL - - !. 0 U0 o0a 01 0 0. Ur0a w 0
C 0 l~ 01 0 X X0 at~0 EI US<.' 0
oc C6 u e0 '600 u- I.- a
111 W*-'a.l-216 .~
2. Capability Loss Is Also a Consequence of theProcess
It is difficult to document a failure tosatisfy functional requirements; generally, thesystems that are acquired do satisfy the initialrequirements. Failure to satisfy functionalrequirements arises as either a failure to accom-modate growth or a failure to acquire a neededcapability. Failure to acquire a needed capabil-ity may result from a decision by a user thatsatisfying a requirement is not worth the effortof going through the acquisition process.
Virtually all vendor industry representa-tives indicated that they were aware of manysmall applications programs that were not imple-mented because existing systems do not adequatelyaccommodate growth. Throughout the Governmentand industry surveys, there was no support forthe premise that functional requirements arestatic even in the general purpose MIS world,while there is ample evidence that the acquisi-tion process has failed to accommodate changingneeds. Even such standard ADP applications aspersonnel and financial record keeping are sub-ject to change. Uniform cost accounting, affir-mative action and all-volunteer forces were citpdas typical examples of factors that cause func-tional requirements to change in these areas.Even in the absence of changing functional re-quirements, there are quantitative changes inrequirements that the acquisition process hasfailed to accommodate. As an example of thisfailure, considerable doubt has been expressedthat AFLC can perform essential wholesale logis-tics functions in a military surge, and thisperceived deficiency is a direct result of ADPacquisition decisions.*
2. THE ANALYSIS OF ISSUES REVEALS THE ROOT CAUSES FORTHIS DAMAGE
The collection of data from literature, Government,and industry sources resulted in a large volume of infor-mation addressing a variety of perceived problem areas
* Since many AFLC acquisitions became linked during thevarious approval processes, they are all mentioned inthe separate volume on the MASIIS/TDSC case study inAppendix E.
111-22
which contributed to the damage cited in the previous sec-tion of this chapter. For the purpose of analysis, thesedata were sorted into thirty-three issues and analyzed todetermine the recurring causes for the poor process per-formance. This section summarizes these issues andanalyses.
(1) The Investigation of Deficiencies and Problems inthe ADP Acquisition Process Resulted in the Iden-tification of 33 Issues
The issues were documented in 33 Issue Papers Ibased on collated data from all data sources. Thesource data (interview notes and published documents)were reviewed and pertinent facts and statements weresynthesized or extracted verbatim and catalogued byissue subject. The total collection of collated datawas then used to develop the issue papers. Theresulting issue papers are attached as Appendix F.
Table 111-7 depicts the primary sources of infor-mation for the development of each issue. Whilenearly every issue had contributing data from allthree sources, only the primary sources are indi-cated. The positions taken by the various sources arerigorously documented in the collated data used todevelop each issue paper.
Each issue was then analyzed to determine theroot causes for each documented problem area. Theanalysis addressed three questions:
Does the issue document a problem in the ADPacquisition process?
Is it an internal Air Force problem (i.e.can corrective measures be taken within theAir Force) or does it require significantaction external to the Air Force?
What is the root cause for the problem?
The following paragraphs highlight the results of thisanalysis for each issue.
ISSUE #1: Standards - Two elements were addressed:
Are standards appropriate in the ADP indus-try, which is characterized by rapid techno-logical change?
111-23
TABLE 111-7Issue-Source Cross-Reference
Primary Data SourcesGovernment Industry Literature
Issue Survey Survey Review
1. Standards X X2. Requirements
Definition X3. Technology Lag X X4. Testing X X5. Training X6. Planning X X7. Requirements Approval X8. Program Management X9. Solicitations/Proposal
Evaluation X10. Industry Response to
Solicitations X1i. Contract Management
and Procurement X12. Specification
Development & Approval X13. Control of ADP
Requirements Changes X14. Types of
Specifications X X15. AFCAC X16. Types of Solicitations
and Contracts X17. Legislative Role X18. Congressional
Conflicts X19. Executive Branch
Conflicts X20. OMB's Role X X21. GSA's Role X X X22. OSD's Role X X X23. Headquarters AF Roles X X24. Functional User/ADP
SpecialistInteraction X
25. ADPS Standard SystemsManager's Role X
26. AFR 300/800 SeriesConflicts X
111-24
TABLE 111-7 (Continued)Issue-Source Cross-Reference
Primary Data Sources
Government Industry LiteratureIssue Survey Survey Review
27. AFR 300 SeriesProcess X
28. The Brooks Act X X29. Competition,
Conversion and LTOC x30. OMB & GSA Policy
Performance x31. A-109 in ADP
Acquisition X32. Life Cycle Management X33. Commercial Practices X
111-25
Is the Government the appropriate organiza-tion to take the lead in developingstandards?
Research revealed marked disagreement as towhether or not NBS Standards (FIPS) result in greatercompetition for Federal business, with subsequentlong-term benefits outweighing the short-term imple-mentation costs, or actually cost the Government moneyby locking in technology and limiting competition.Standardization is critical to Government Operationsand has implications both in terms of external influ-ences imposed on the Air Force, as well as the abilityto set standards internally.
The ability of the Government to force industrycompliance with Federal standards is based on theextent to which they can influence the market place.As the Government share of the ADP market declines,the industry will be less responsive to Federal Infor-mation Processing Standards than to those set in in-dustry organizations, such as ANSI or ISO. Attemptsto standardize on such a scale may severely limit com-petition and thus drive up the cost of doing businesswithin the Federal Government.
Concerning DOD in general and the Air Force inparticular, leadership appears to be lacking in thearea of standards policy. Little active participationis evidenced in terms of both responding to externalstandards and establishing internal standards. Theseobservations were most evidenced in the lack of effec-tive long range standards planning and policy.
ISSUE #2: Requirements Definition - Several areas areaddressed in the requirements definition phase of theacquisition process:
* Should functional requirements be used?
* Is the Data Automation Requirement documen-tation excessive?
What studies and analyses should beperformed?
Can users adequately articulate requirements?
* What is and should be the role of industry?
111-26
The requirements definition phase was frequentlycited as a major problem area in Air Force acquisitionin that inadequate requirements definition usuallyresults in delays in the process. Among the problemscited were:
* There is an inability on the part of usersto articulate and communicate requirementsproperly.
* DAR documentation is excessive and solutionoriented, particularly the feasibility studyand econcrnic analysis.
* The process results in premature commitmentsto solutions, particularly hardwaresolutions.
The analysis of this issue revealed several fun-damental problems in the acquisition process. Thefirst is that when requirements are poorly identifiedit is usually because the appropriate skills are lack-ing in the originating organization. Second, the pro-cess itself demands a level of documentation that isredundant, hardware oriented, and forces early commit-ments to solutions. Third, poor requirements defini-tion can usually be traced to a lack of sufficientlong range planning which has a total systems focusand life cycle management orientation.
ISSUE #3: Technology Lag - This issue focused on thecauses, consequences, and costs of technology lag.
Considerable evidence supported the assertionthat the Government incurs a significant penalty forusing outdated technology because of technology lagcreated by a len~gthy and excessively constrainedacquisition process.
Technology lag reflects the damage caused by alengthy and cumbersome ADP acquisition process. As aresult, the Federal Government is incurring unneces-sary costs by keeping old systems in operation ratherthan upgrading to more current technology. Previoussections of this chapter detail this viewpoint. Aprimary cause of technology lag is the long acquisi-tion lead time and the failure of the process to adaptto a rapidly changing technological environment; how-ever, more fundamental causes are a lack of a systemsmanagement approach and a life cycle orientation.These latter causes can be traced directly to Federaland Air Force policies, guidelines and practices.
ITI-27
ISSUE #4: Testing - This issue revolved around threequestions concerning the use of benchmarking and livetest demonstrations (LTD):
Are LTDs cost effective?
When are LTDs appropriate as a criterion forsource selection?
In which situations are pre-award versuspost-award LTDs more appropriate?
Research indicated that although LTDs are costlyand time consuming, they can be valuable in sourceselection; however, the following criticisms werenoted:
Expertise for managing effective LTDs islimited to AFCAC.
Costs are driving vendors, particularlysmall vendors, out of the competition.
LTDs often do not serve as an effectivediscriminator in the selection process andthus serve only to delay the acquisition.
LTDs impose additional burden on the usersto create representative benchmarks.
This is an area that is marked by much confusionwith little specific guidance. The use of LTDs is notrationally applied to acquisitions on a case-by-casebasis, but rather routinely applied to most acquisi-tions. Further, there is little guidance available onhow or when to conduct LTDs, or in what situations itis cost effective to do so.
ISSUE #5: Training - The issue analysis addressedthree areas concerning the skills and training ofthose who must function within the current acquisitionprocess.
Are specialized skills required to supportthe AF ADP acquisition program? If so, whatis the availability of the personnel withsuch expertise?
Does the lack of experienced, trained per-sonnel delay the process?
111-28
If training is inadequate, what are alterna-tives to improve skill levels?
The inadequacy of skills and training, and theineffective use of personnel are serious detriments inconducting AF ADP acquisition and are a major sourceof delays in the process. Among the problems identi-fied are the following more frequently mentioned areasof concern:
* There are serious deficiencies in the avail-ability of qualified and experienced peopleto properly formulate requirements, developjustifications, and translate requirementsinto technical specifications and procure-ment strategies.
* There is a lack of career opportunity andtraining in computer resource management.
The lack of job flexibility and technologi-cal challenges (primarily due to an obsolez-cent computer inventory) inhibit retentionof quality computer specialists.
Skilled personnel are not utilizedeffectively.
The shortage of skilled personnel in key discip-lines is a serious problem that pervades the entiremilitary. However, this problem is further compoundedwhen personnel resources are not effectively util-ized. Both training and the allocation of resourcesmust be considered in improving the present situation.
ISSUE #6: Planning - This issue dealt with theeffectiveness of long range ADP planning in the AirForce.
Research revealed the Air Force long range ADPplanning is programmatic and effective in obtainingthe necessary funds and resources to acquire ADP; how-ever, strategic planning is virtually non-existent.There is a failure to effectively link mission plann-ing to the identification of ADP programs.
The lack of a mission orientation in the acquisi-tion process is most evident in the planning process.While fiscal planning is effective and important, thelack of strategic planning has limited the effective-ness of the acquisition process. The consequences ofthis are exemplified by the failure to keep pace withtechnology and the lack of a life cycle focus.
111-29
ISSUE #7: Requirements Approval - This issue con-sidered the delays in the ADP acquisition processtraceable to the requirements approval process,specifically:
What approval methods have been employed,and how can they be improved?
To what degree should approval bedecentralized?
The multiplicity of reviews in the approval pro-cess is a significant delay factor. Raising theapproval thresholds so that MAJCOM Single Managers canapprove many requirements that previously requiredUSAF/ACD approval has been successful; however, othercirticisms remain:
There has been a great deal of confusion asto what is being approved and by whom.
Recent changes to AF HOI 300-4 put the bur-den of Air Staff coordination with the func-tional OPR. This has resulted in furtherdelays in processing DAR's.
The requirements approval process is whatdrives the solution orientation and thehardware focus of the DAR.
Decentralizing the requirements approval processis a step in the right direction. More fundamentalchanges such as separating needs approval from solu-tion approval, tying needs approval to funds approval,and making more effective use of personnel and skillsneed to be effected to make the approval process moremeaningful and less time-consuming.
ISSUE #8: Program Management - This issue analysisexamined the Program Management role and authority inthe ADP acquisition process.
The major problem is that responsibility transi-tions among users, ADP specialists and procurementduring the course of an ADP acquisition making it dif-ficult to establish accountability for a particularprogram. This lack of single accountability is evi-dent in the following areas.
Program Managers are not appointed earlyenough in the process to influence acquisi-tion strategy.
111-30
Program Managers lose control of theiracquisitions when they are turned over toAFCAC to conduct a procurement.
Program Managers lack the end-to-end controlof an ADP acquisition.
The analysis of this issue reveals certain fund-amental problem areas in acquisition management forADP. First is the lack of accountability previouslymentioned. Second is the ineffective use of personneland skills caused by a transfer of responsibility anda lack of early involvement by key personnel. Thirdis the lack of flexibility in the current role of theProgram Manager to influence the acquisition strategyand to carry out the acquisition strategy during theproblem phase of the process.
ISSUE #9: Solicitation/Proposal Evaluation - Severalareas are addressed in the solicitation/source selec-tion phase regarding whether the Air Force proceduresand methodologies provide for:
• Reasonable cost estimating• Encouragement of competitive procurements• Appropriate proposal evaluations
Proper level of review.
The solicitation/source selection phase was oftencited as a problem area because:
Cost estimates are often questionable due toa lack of understanding of cost factors andlack of a standard life cycle model.
There appears to be an over-emphasis onhardware competition without regard to totaloverall system cost.
The number of procurement reviews is dis-proportionate to the benefits derived.
Analysis of this issue indicated several basiccauses. There is a general lack of people skilled inADP costing except at AFCAC and there is no standardlife cycle cost model. Emphasis on competitive hard-ware acquisition has resulted in procurements thatmaximize hardware competition to the detriment of low-est total overall cost. The required procurementreviews tend to lengthen the procurement process andadd little overall value to procurement itself.
111-31
ISSUE #10: Industry Response to Solicitations - Thisissue considered factors that adversely affect indus-try's willingness to respond to Air Force solicita-tions for ADP resources.
The following factors were most often cited byindustry as problem areas:
* Transferring more risk to the winning vendor
* High cost of bidding, e.g., benchmarking
* Terms and conditions not normal to comnmer-cial practice
* Detailed cost data required of single bidders
* Short time allowed for bid responses.
The underlying issue is how much risk the AirForce should be willing to assume versus passing it onto industry. Currently, source selection procedureshave been inflexible in that much of the risk ispassed to the vendor. The result is the driving up ofcosts and the limiting of competition.
ISSUE #11: Contract Management and Procurement - Thisissue focused on the pros and cons of centralized ver-sus decentralized procurement support in two areas:
The responsibility for source selection andaward
The responsibility for post-award contractadministration.
The centralized procurement support offered byAFCAC was cited as being highly desirable for largerADP acquisitions (eg., greater than $3M). AFCAC ishighly skilled in ADP procurements; however, theirresources are limited. On the other hand, with fewexceptions, base level procurement support, skilled inADP acquisition, is virtually non-existent.
This issue points to two problems inherent in theADP acquisition process--ineffective use of skilledhuman resources and a lack of flexibility. There isan ineffective use of the highly skilled AFCACresources when they must support certain smaller ADPacquisitions, yet there is no alternative when thattype of support is not available locally.
111-32
ISSUE #12: Specification Development and Approval-This issue dealt with three areas:
* To what extent does the specification devel-opment process contribute to acquisitionlead time?
* How and when are specifications prepared?
* Who should prepare them?
It was found that the specification developmentprocess is a delaying factor in the acquisition leadtime for a variety of reasons:
* The user has difficulty articulatingrequi remients.
* Users in general do not know how to developfunctional specification and tend to "overspecify" a hardware solution.
* Little skilled support is available to theusers during the initial stages of specifi-cation development.
* Specifications must usually be rewrittenwhen procurement support (e.g., AFCAC)finally becomes involved.
Good specifications are a time driver in the pro-cess but they are essential to a successful ADP acqui-sition. Two causes are raised by this issue. First,the process drives users to over specify their needs,by being too hardware and solution oriented ratherthan functionally oriented. Second, the processimplies an ineffective use of resources when it asksuntrained users to prepare initial specifications.
ISSUE #13: Control of ADP Requirements Changes - Thisissue dealt with changing ADP requirements during theacquisition process and how to control these changes.
A number of reasons were cited for requirementschanges during the acquisition process:
Needs are not thoroughly analyzed prior todefining requirements.
* Requirements are not properly articulated.
111-33
External environmental factors were notconsidered.
There is an inadequate exchange of informa-tion between the user and the procuringactivity early in the process.
Requirements changes cannot be avoided in the acquisi-tion of ADP, particularly whent the acquisition processis lengthy. Control of these changes is further com-plicated when requirements are not appropriatelystated, solutions are constrained too early in theprocess, and the accountability for requirementstransitions among different organizations.
ISSUE #14: Types of Specifications - There is con-siderable opinion on the use of functional versustechnical specifications for procurement.
Congress and GSA favor functional specificationsin order to enhance competition. Industry, in gener-al, also prefers functional specifications becausethey enhance competition. Certain Air Force organi-zations feel that functional specifications presentdifficulties in ensuring compliance during proposalevaluation, and feel that the Air Force is avoidingits responsibilities by not specifying precisely whatis needed in ADP terms.
For differing situations there are clear advan-tages and disadvantages for using one type of specifi-cation over another. The current hardware and solu-tion orientation of the process tends to force thedevelopment of detailed technical specifications. Theflexibility needs to exist within the process to usedifferent levels of specificity as the situation dic-tates. Further, the expertise must be available earlyin the process to take advantage of one type of speci-fication over another.
ISSUE #15: Air Force Computer Acquisition Center(AFCAC) - AFCAC was a controversial issue with a largesegment of the Air Force ADP community. At issue isthe length of time AFCAC takes to conduct an acquisi-tion, AFCAC's role in the acquisition process, andwhen AFCAC should become involved with the user.
The duration of an AFCAC-conducted acquisition istoo long based on even AFCAC's own internally estab- 3lished standards. Some of the reasons cited are:
111-34
* Using the same acquisition proceduresregardless of program scope, size, and risk
* Over-zealous efforts to achieve competitionto the detriment of schedule
* Delays in resolving disagreements with theuser--a question of who is in charge of theprogram
* The lack of an early involvement in theprocess.
These criticisms concerning AFCAC's involvementin "-he process point to three deficiencies in the pro-cess. First is the apparent inflexibility with whichAFCAC conducts of procurement, not tailoring thestrategy to fit the size and scope of the program.Second is the lack of end-to-end accountability ex-emplified by the transition of responsibility from theuser to AFCAC late in the process. Third is the in-effective use of skilled resources characterized byAFCAC's involvement in programs without regard tosize, scope and complexity.
ISSUE #16: Types of Solicitations and Contracts -This issue dealt with the types of solicitations andcontracts used in ADP acquisitions and their effect onthe acquisition process. The analysis also addressedthe Air Force means for communicating with industryregarding ADP acquisitions.
Requirements contracts are favored by a number oforganizations, but there is dissatisfaction with GSA'shandling of requirements contracts. GSA schedule con-tracts are also a source of dissatisfaction with someusers. They cite the one-year duration, the non-competitive nature, inflexibility, the need to testthe market and GSA's frequent failure to promptlynegotiate contracts at the start of each fiscal year.
Many organizations cite the use of IFBs as aproblem. They find it difficult to develop statementsof work and specifications that do not generate a num-ber of questions. Most Air Force ADP contracts arefirm fixed price regardless of risk, uncertainty oranything else that would dictate using alternateapproaches. There is also a strong desire in the AirForce for increased systems life, i.e., number ofyears on the contract, and the ability to upgrade inthe vendor's line.
111-35
The Air Force, particularly AFCAC, was criticizedfor not being more open and flexible in its dealingswith industry. Draft RFPs, vendor conferences,libraries and advance notifications of procurementswere suggested as means of communicating the AirForce's ADP procurement plans. A further inflexibil-ity is characterized by the restrictions imposed onthe alternative acquisition strategies cited above.
ISSUE #17: Legislative Role - This issue analysisexplored the extent of, and rationale for, congres-sional committee intervention in the AF ADP acquisi-tion process. While various committees display inter-est in, or maintain oversight of, certain areas of theFederal agencies' acquisition processes, the HouseGovernment Operations Committee (HGOC) is particularlyinfluential.
HGOC's involvement stresses the limits of over-sight and contributes to delays. In its efforts toensure maximum free and open competition, and to forcebetter agency management of ADP resources, the HGOCintervenes and requests that GSA place holds onacquisitions.
Congress has become increasingly interested inagency ADP resource management, primarily due to abelief that management responsibility has been abdi-cated by the executive branch agencies. However,extensive oversight and scrutiny might be reducedthrough the demonstration by the AF and DOD that theyhave wearned autonomy" through effective ADP manage-ment, to the satisfaction of GSA and Congress.
ISSUE #18: Congressional Conflicts - This issuefocuses on the philosophical conflicts between theHGOC and the HAC. The House Government OperationsCommittee (HGOC) and the House Appropriations Commit-tee (HAC) are two of the most influential committeesin the acquisition and management of ADP resources.
Their disagreements as to how best toachieve the PL 89-1.06 goals of economy andefficiency have contributed to user agencyproblems in interpreting and implementingcongressional guidance, as instructions fromthe committees are often contradictory.
111-36
These committees have developed adversaryrelations with OMB and GSA, resulting in amutual lack of understanding of objectivesand basic mistrust of intentions, and anabsence of consistent leadership and guid-ance to the user agencies.
The DOD and AF have been forced to expend addi-tional resources to satisfy demands imposed by con-gressional oversight, and have been challenged to com-ply with conflicting congressional instructions.
ISSUE #19: Executive Branch Conflicts - This issuedealt with the confusion over the roles and responsi-bilities of executive branch agencies in the FederalADP Program. The continuing lack of a central coor-dinator, planner and policy responsible agency has ledto confusion, complexity and delays in user agencydealings with OMB and GSA and has fostered antagonis-tic relations, a prime contributor to ADP acquisitionproblems. Despite recognition of need for policy andprocedural changes, neither OMB nor GSA have acteddecisively due to unresolved authority and mutual"finger pointing."
Fragmentation of responsibility and resultantfailure to reach timely decisions have created apolicy void, which has led to excessive delays, con-fusion, and questionable actions in acquisition.Acquisition has become a focal point for politicalmanipulation. Agency actions have sometimes beencounter to intended goals of the system and haveevoked congressional intervention, GAO audits, andcharges of mismanagement.
ISSUE #20: OMB's Role - This issue focused on thefailure of OMB to provide managerial and policy lea-dership to the extent anticipated by the ADP acquisi-tion community.
The analysis revealed that the prime contributorto OMB's failure has been pervasive confusion and lackof executive level coordination and cooperation inadministering the Federal ADP program. While the per-ception that OMB has been the senior ADP policy offi-cial may have been ill-founded based on PL 89-306 andEO's 11717 and 11893, PL 96-511 clarifies OMB's broadrole in ADP policy.
111-37
The lack of clearly defined hierarchy of author-ity and responsibility has created an environment ofantagonism and policy confusion throughout the FederalADP Program.
ISSUE #21: GSA's Role - This issue analysis addressedGSA's performance as the central ADP procurementauthority under PL 89-306.
The criticisms of GSA focus on the organization'sfailure to adjust to its dual procurement and policyroles. GSA has attempted to control the use of ADPresources through restrictive controls over the pro-curement process. It is not apparent that GSA'sinvolvement with agency ADP management through the DPAprocedure contributes to achievement of PL 89-306objectives, provides additional technical expertise,averts wasteful expenditures, or ensures competition.
While the trend is toward significantly higherthresholds, and more flexible and consistent procure-ment procedures, the responsibility still remains forthe AF (and other agencies) to be assertive, cooperatewith the legislative and executive branches in for-mulating and implementing policy under PL 96-511 andthe FPS, demonstrate competent ADP resource manage-ment, and earn the autonomy commensurate with theirtechnical and managerial expertise.
ISSUE #22: OSD's Role - This issue focused on OSD'smanagement of the DOD ADP Program, and concluded thatOSD's provisions of ADP policy guidance and support tothe services has been inadequate.
There is a perceived lack of a strong, centrallymanaged, and visible ADP organization. Problems con-tributing to DOD's difficulty in managing informationtechnology resources include fragmentation of respon-sibility, limited size and resources, low visibilityin liaison with external organizations, and mutuallydistrustful relations with Congress.
While deficiencies in DOD management of informa-tion technology resources are impeding their effectiveand efficient use, improvements are possible withinthe scope of OSD's authority by:
*Alleviating fragmentation of responsibility
111-38
* Increasing visibility and credibilitythrough a more dynamic posture in communi-cating and cooperating with OMB, GSA, andCongress
* Addressing the criticisms of weak and inef-fective policy.
ISSUE #23: Headquarters USAF Roles - This issue dealtwith the roles, missions and performance of the AirStaff organizations and the office of SAP/FM.
ACD and other Air Staff offices were repeatedlycriticized for the excessive amount of time requiredto staff requirements with little value added to theprocess, and for failure to develop, promulgate andissue proper policy guidance. The general feeling isthat there are many reviews at the Air Staff, any oneof which can say no and none of which can say yes. Inaddition, there was considerable controversy concern-ing the placement of the Senior ADP Policy Official inthe SAF/FM.
The analysis of this issue points to two problemareas. The first is the ineffective use of personneland skills on the Air Staff in that many of the func-tions performed by ACD in the management of ADP areperformed by other Air Staff organizations for othertypes of support. Second, the lack of consistency inthe ADP management and acquisition process with otherprocesses (e.g., embedded systems, and telecommunica-tions) is the source of much confusion.
ISSUE #24: Functional User/ADP SpecialistInteraction - This issue covered the responsibilitiesand interactions of. the functional users and the ADPspecialists in the acquisition process.
The functional users tend to regard the ADPacquisition process (and personnel) as an obstacle toobtaining resources to satisfy mission requirements.The functional user frequently needs assistance inexpressing his needs so that they will "sell" in theADP community. He needs the assistance of the ADPspecialist. The user also has a problem in managingchanges in his requirements, particularly in largesystems.
There has to be a strong interface between thefunctional user and the ADP specialists, right fromthe onset of establishing requirements. Articulating
111-39
functional requirements and translating them into ADPtechnical solutions and procurement packages requireexperience and skill. This skill is in short supplyin the Air Force. Given the complexity of the ADPacquisition process, there is a need for specialistsknowledgeable in the technology, current ADP acquisi-tion policy, regulations, and contracting procedures,as well as the particular requirements of the mission.
Current procedures do not make effective use ofthe individual skills of three groups of participantsin the acquisition process--users, ADP specialists,and procurement specialists. In addition, responsibi-lities transition among the three to the point thatthere is a lack of end-to-end accountability by anysingle group which often results in adversaryrelationships.
ISSUE #25: ADPS Standard System Manager's Role - Thisissue covered the role of the Air Force System Mana-gers for the standard ADPS.
The role of the AF standard system manager variesgreatly depending upon the specific system.
Responsibilities are well defined for somesystems, but not for others.
For some standard systems, organizationsother than the program manager are involved,leading to a situation in which the programmanager has responsibilities without com-mensurate authority and resources.
Using the AFDSDC as the standard system programmanager has worked well for standard base level appli-cations, but has worked less well for ADPS 10 (MAJCOMSupport) and for ADPS 80 (WWMCCS). When the applica-tions are non-standard it may be well to examinewhether standard system management is feasible oradvantageous.
ISSUE #26: AFR 300/800 Series Conflicts - In com-paring the AFR 300 series approach with that of AFR800 series, two related findings emerged:
There is a great deal of inconsistency inthe management and execution of the twoprocesses.
111-40
There is much confusion and little cen-tralized guidance concerning the scope ofapplication for the two approaches.
The major inconsistency is that the AFR 800series process is designed for larger, more complexacquisitions than the AFR 300 series process, yet itpossesses greater flexibility and more decentralizeddecision making. The second inconsistency, and thesource of much confusion, is that ADP can be acquiredunder both processes, yet there is no single manage-ment oversight. Decisions to use one over the otherare seemingly based on convenience and personal pref-erence rather than established management criteria.
ISSUE #27: AFR 300 Series Process - This issue analy-sis examined the AF 300 Series Regulations and theirimpact on the ADP acquisition process performance.
The major areas of perceived deficiencies in theregulations include the following:
They define a process which is hardware andsolution-oriented.
They require an excessive level of documen-tation and number of reviews.
They lack many of the systems disciplinesnormally required for systems acquisitions.
They specify an inordinate level of cen-tralized management control.
The regulations are not clear andunderstandable.
While much of the inflexibility and restrictive-ness of the current acquisition process has beenattributed to the regulations themselves, there is apoint of view that contends that the regulations dopermit the needed flexibility. A close examination ofthe regulations revealed a certain degree of flexibi-lity designed into the process; however, this nominalflexibilty has been negated by the restrictiveness ofactual application.
ISSUE #28: The Brooks Act - This issue analysissought to discover:
What are the stated and perceived objectivesof PL 89-306?
111-41
* Has the implementation been effective?
* Is the legislation fundamentally sound?What principles are in contention?
The stated purpose of the Brook's Bill was toprovide for economy and efficiency; however, the per-ceived purpose varies from an attempt by Congress tocombat perceived abuses in Federal agency ADP manage-ment and procurement to a mechanism to force competi-tion and dissolve a monopolistic market environment.The law itself is usually deemed innocuous, but itsimplementation has not achieved the goals of economyand efficiency in acquisition. There is some contro-versy surrounding the soundness of the law itself.Many term the law as "harmless" and "ineffective" andstate that its goals are laudable but its implementa-tion has failed. Those criticizing the Act do sobecause it provides ambiguous and conflicting defini-tion of roles, restricts agency authority to carry outtheir mission, and is overly concerned with hardwaremanagement.
ISSUE #29: Competition, Conversion, and LTOC - Thisissue addressed Federal policies on ADP acquisition inthe areas of competitive procurements, software con-version and lowest total overall cost (LT0C) as cri-teria for selecting an acquisition strategy.
The analysis found that these policies have beenunrealistic, by reflecting an overemphasis on hardwarecosts and cost avoidance through maximum compet-tion,and neglecting the impacts of software convers4 .on andthe failure to use LT0C as an evaluation criteria.
ISSUE #30: 0MB and GSA Policy Performance - Thisissue focused on the quality and effectiveness of ADPacquisition policies promulgated by 0MB and GSA.
Federal ADP policy is complex, conflicting, over-lapping, and long overdue for consolidation, clarifi-cation and codification. Analysis revealed that therehas been a significant void of consistent leadershipin ADP policy. There have been strong and often con-flicting influences on policy, imposed by industry andCongress. Resulting policies have been ineffective inprompting timely acquisition of ADP, cost effectiveprocurements, and maintenance of technological cur-rency. Recent policy initiatives, includingPL 96-511, the proposed FPS, and FPR and FPMR revi-sions are steps in the right direction.
111-42
ISSUE #31: A-109 in ADP Acquisition - This issue con-cerns the applicability of OMB Circular A-109 to ADPacquisitions.
The purpose of analyzing an A-109 approach is todetermine its applicability to ADP as well as comparethat process to current ADP acquisition methods.While A-109 is not the panacea for ADP acquisition, itdoes provide some useful lessons. Certain key ingre-dients of an A-109 approach were found to be lackingin the current ADP process, such as flexibility inacquisition strategies and approaches, an emphasis onmission orientation, the infusion of technologicalopportunities and the early involvement of industry inthe process.
ISSUE #32: Life Cycle Management - This issue analy-sis addressed two questions:
Is life cycle management of Automated Infor-mation Systems (AIS) being properly employed?
Is lowest total overall cost (LTOC) givenproper consideration in ADP acquisitions?
The consensus of comments, largely from theDefense Audit Service Report on the Review of theImplementation of DOD LCM Policies for AIS (seeref. 9), is that ADP Life Cycle Management (LCM) isnot well implemented in the services for the followingreasons:
Insufficient OSD guidance on policy, proce-
dures and terminology
Lack of manpower resources in OSD
Inadequate cost estimating and trackingpractices
Lack of a standard life cycle model foranalysis.
Further comments indicated that ADP is being procuredwith too much emphasis on hardware costs, as opposedto LTOC, and that unnecessary costs are being incurredwithout the proper attention given to total life cyclecosting. Specific recommendations cited from the datasources included:
111-43
* Change Federal ADP procurement policy to
avoid unwarranted software conversion.
* Implement LTOC guidelines within DOD.
* Standardize LTOC and life cycle cost models
for ADP acquisitions.
ISSUE #33: Commercial Practices - In this issueanalysis, commercial practices were compared withthose of the Federal Government to determine whetheror not industry performs better than the FederalGovernment in acquiring ADP. The quantitative resultsof this analysis were presented earlier in this chap-ter. other specifics were observed in this comparison:
* Industry uses functional specification moreoften.
* Industry uses past performance, responsive-ness and other subjective factors as evalua-tion criteria for the award.
Industry will stay with a vendor as long asthey remain satisfied.
(2) Summarizing the Analysis of Issues Provides theFramework for the Study Conclusions
Following the separate analysis of each issue, acollective evaluation of all the issues was made todetermine the underlying causes for the poor processperformance. This analysis was made in two parts:
The first determined the degree to which theAir Force can influence the process throughits own redirections.
The second identified the recurring causesfor the problems evidenced in the analysisof each of the issues.
Table- 111-8 presents the results of the firstdetermination. It shows that twenty-four (24) of thethirty-three (33) issues address problems within thejurisdiction of the Air Force. While this figure maybe somewhat arbitrary in that issues were identifiedlargely from sources internal to or close to the AirForce acquisition process, it is indicative of thedegree of opportunity available to the Air Force.
111-44
TABLE 111-8Categorization of Problems
External Air ForceIssue Problem Problem
1. Standards X X2. Requirements
Definition X3. Technology Lag X4. Testing X5. Training X6. Planning X7. Requirements Approval X8. Program Management X9. Solicitations/Proposal
Evaluation X10. Industry Response to
Solicitations X X11. Contract Management
and Procurement X12. Specification
Development & Approval X13. Control of ADP
Requirements Changes X14. Types of
Specifications X15. AFCAC X16. Types of
Solicitations and Contracts X17. Legislative Role X18. Congressional
Conflicts X19. Executive Branch
Conflicts X20. OMB's Role X21. GSA's Role X22. OSD's Role X23. Headquarters AF Roles I X24. Functional User/ADP
SpecialistInteraction X
25. ADPS Standard SystemsManager's Role X
26. AFR 300/800 Series Conflicts X27. AFR 300 Series
Process X
111-45
TABLE 111-8 (Continued)Categorization of Problems
External Air Force
Issue Problem Problem
28. The Brooks Act X
29. Competition, Conversion,and LTOC X
30. OMB & GSA Policy
Performance X31. A-109 in ADP
Acquisition X X
32. Life Cycle Management X X
33. Commercial Practices X X
111-46
___ ___ ___ ___ __a ,_ - r.
In reviewing the individual issue analyses, itwas observed that for all the problems identified,there are a number of underlying causes which have adirect bearing on the current state of the ADP acqui-sition process. These are the root causes which wererepeatedly mentioned during the issue analyses and arelisted as follows:
1. Lack of effective leadership and policydirection
2. Hardware orientation and solutional focus ofthe ADP acquisition process
3. Ineffective use of skills and personnelresources
4. Inflexibility in the acquisition and manage-ment of ADP
5. Failure of the ADP acquisition process toadapt to a changing environment
6. Lack of accountability and end-to-endmanagement in the process
7. Lack of a mission orientation
8. Lack of systems management and life cycleperspectives
9. Inconsistency in policy, direction andmanagement of ADP.
Further evaluation of these deficiencies pointedtoward three broad areas in the acquisition andmanagement of ADP which require primary consideration.
The first deals with the overall objective of theADP acquisition process. The hardware and solutionalfocus and the lack of a mission orientation indicatethat the objective of the acquisition process iseither poorly defined or misdirected. Second, thelack of flexibility and adaptability of the process,coupled with the lack of a systems and life cycleorientation, indicates problems in the current processdesign. Third, the lack of effective leadership, con-sistent direction accountability, and effective use ofpersonnel skills point towards a mismatch of roles,responsibilities and skills.
111-47
The preceding analysis reveals that the problems anddeficiencies in the AF ADP Acquisition Process can betraced to one or more of four categories, three of whichwere traced to areas within the Air Force's control:
The objective of the Air Force ADP acquisitionprocess is improperly focused.
The Air Force acquisition process design cannoteffectively accommodate all the situations forwhich it is being used.
There is a mismatch of roles, responsibilitiesand skills within the Air Force.
External ADP acquisition policy provides inappro-priate and/or inadequate direction for the AF ADPprogram.
Further, major improvements can be achieved by addressingthe problem areas internal to the Air Force. The nextchapter presents the study conclusions in this framework.For emphasis, those areas internal to the Air Force arepresented first.
111-48
IV. STUDY CONCLUSIONS
IV. STUDY CONCLUSIONS
This chapter presents the major conclusions of thestudy. in the previous chapter, a broad spectrum ofissues was discussed and evaluated. Each issue was eval-uated in terms of its causes and whether or not it couldbe corrected by focusing efforts internal to the AirForce. The conclusions, presented herein, identify themajor causes and effects for the poor process performancecited in Chapter III. The following are the five majorconclusions:
* Substantial improvements in Air Force ADPacquisitions can be realized by focusing effortsinternal to the Air Force
* Current ADP acquisition process objectives under-emphasize mission essentiality
* The current process design cannot adapt effec-tively to the varying Air Force ADP needs
* The Air Force ADP management structure contri-butes to the delays in acquiring ADP
* Confusion and inefficiencies have been introducedby top level Federal policy external to the AirForce.
Table IV-l relates the issues discussed in Chapter III toone or more of the five major conclusions which theysupport.
1. WHILE EXTERNAL INFLUENCES ARE THE SOURCE OF MANYPROBLEMS IN THE AIR FORCE ADP ACQUISITION PROCESS,MAJOR IMPROVEMENTS CAN BE ACHIEVED BY REDIRECTIONSWITHIN THE AIR FORCE
A basic conclusion of this investigation is thatalthough there are external constraints imposed on the AirForce by higher authorities, many of the major constraintson ADP acquisitions are self-imposed. These constraintswould appear to have a variety of causes. In some cases,policies which once promoted effectiveness and efficiencynow constrain the achievement of these goals because of achanging environment or differing needs. Some policieshave always been constraining to the process because they
IV-l
TABLE IV-lIssue-Conclusion Cross-Reference
Conclusion No.
Issue 1 2 3 4 5
Ix1. Standards X iX X
2. RequirementsDefinition X X
3. Technology Lag X4. Testing X X5. Training X X6. Planning X X X7. Requirements Approval X X X8. Program Management X X9. Solicitations/Proposal
Evaluation X X10. Industry Response to
Solicitations X X X11. Contract Management
and Procurement X X X12. Specification
Development & Approval X X X13. Control of ADP
Requirements Changes X X X14. Types of
Specifications X X X15. AFCAC X X16. Types of Solicitations
and Contracts X X17. Legislative Role X18. Congressional
Conflicts X19. Executive Branch
Conflicts X20. OMB's Role X21. GSA's Role X22. OSD's Role X23. Headquarters AF Roles X X24. Functional User/ADP X
SpecialistInteraction X
25. ADPS Standard SystemsManager's Role X
26. AFR 300/800 SeriesConflicts X X X X
27. AFR 300 SeriesProcess X X
IV-2
TABLE IV-lIssue-Conclusion Cross-Reference
Conclusion No.-
Issue 1 2 3 4 5
28. The Brooks Act x29. Competition, Conversion
and LTOC X30. 0MB & GSA Policy
Per formance X31. A-109 in ADP
Acquisition X X X32. Life Cycle Management x X33. Commercial Practices X X X
IV- 3
.-- - l - --P- ,- ____________________ -
were based on erroneous perceptions of the external envi-ronment. Regardless of their cause, these constraints areaddressed in this section because they provide oppor-tunities for improvement that are within the jurisdictionof the Air Force.
This conclusion is supported by the issue analysis,which revealed that 24 of the 33 issues represent problemswhich are primarily internal to the Air Force. While allissues impact Air Force ADP acquisitions and thus presentproblems for Air Force considerations, these 24 can beaddressed within the Air Force itself. Table 111-8 pre-sented these specific issues.
The next three major conclusions (conclusion numbers2, 3, and 4) address those problem areas within the AirForce.
2. THE CURRENT ADP ACQUISITION PROCESS OBJECTIVES UNDER-EMPHASIZE MISSION ESSENTIALITY, IN FAVOR OF A PRIMARYEMPHASIS ON ACHIEVING THE GOALS OF ECONOMY ANDEFFICIENCY
A basic deficiency in the ADP acquisition process isthe lack of well-defined objectives from which policy,procedures and organizational roles and responsibilitiescan be developed. Many objectives have been cited for theADP acquisition process; the most frequently mentioned isachieving economy and efficiency through maximizinghardware competition. While these goals were legitimatelygiven primary consideration during an era in which it wasappropriate to do so, factors have changed the environmentin which ADP acquisitions must be accomplished. First,technological advances have resulted in categorizing agreater number of computer resources into the class of"General Purpose ADPE." Second, use of ADP has becomewidespread to the point that it is essential to the AirForce mission. These factors are evident in an analysisof the assumptions on which the current process is based.
(1) The Tenets of Efficiency and Economy Are Based onAssumptions Which Do Not Reflect Current Realities
PL 89-306 became law in 1965. Consequently thelaw, the policy framework that derived from it, andthe current Air Force regulations are based on assump-tions of that era. These assumptions include:
Hardware is the primary cost component ofADP.
IV-4
* Maximizing hardware competition will resultin efficient acquisition and maximum economy.
* There is high degree of commonality in ADPapplications.
Throughout the Federal ADP acquisition process,activities are oriented towards minimizing hardwarecosts and maximizing competition. Contracts aregenerally awarded based on the lowest hardware bid,without evaluation of relevant auxiliary costs. FMC74-5 proscribed consideration of most software conver-sion costs in the calculation of overall costs for theselection of equipment through competitive acquisi-tion. Although recent revisions of the FPMR havesuperseded this circular, the policy was designed toensure that "care be taken to avoid undue biases orpredispositions which are prejudicial to free and opencompetition." (FMC 74-5) A primary objective of theFederal ADP Standards Program is to reduce costs andimprove productivity through fully competitive pro-curement. Regarding the assumption of commonality inapplications, government policy encouraging reutiliza-tion of surplus computers is based on the premise ofrelative interchangeability of computers across avariety of applications.
These assumptions have been invalidated bychanges in both the technological environment and theuse of ADP within the Air Force. one of the majortrends in the ADP industry is that hardware costs havebeen decreasing while costs of software (includingsoftware conversion) , personnel, and facilities havebeen increasing. The industry survey revealed thattechnological advances have resulted in hardware costsdecreasing significantly each year. This trend isexpected to continue for the forseeable future. Sincesoftware development is still relatively labor inten-sive, its costs have been rising. one result of theseprice trends is that ignoring software conversioncosts often results in a greater expense in the longrun than that incurred from competition based on low-est total overall cost.
As significant as the reduction in hardware costsis the growth in the application of ADP. In the erafrom 1965 to the present, general purpose computershave gone from being used largely for accounting andpayroll applications to a variety of specializedapplications. Today, general purpose computers in the
'v-5
Air Force support special test equipment, flightsimulation, weapons and armament equipment, and sensorand surveillance systems, as well as the accountingand recora keeping functions. Data Base ManagementSystems, transaction processing and real-time pro-cessing are commonplace in today's general purposecomputers. As a result, the notions that ADP has acommon application and that maximum efficiency andeconomy can be achieved through centralization andsharing of hardware resources are no longer correct.Of even greater significance is the fact that, throughthis technological sophistication, the Air Force isbecoming increasingly dependent upon general purposeADP to accomplish its mission. Robert V. Head, in arecent article in Government Executive (February1981) , described the magnitude of this problem:
Information Systems Technology is so embedded inthe Federal structure that critical administra-tive, scientific and military functions are allbut totally dependent on the smooth functioningof computer hardware and software.
(2) Emphasis in the ADP Acquisition Process Must BePlaced on Satisfying Operational Needs as Well asAchieving Economy and Efficiency
Part of the Government and industry surveyaddressed the issue of which classes of acquisitionsproperly belong in a process whose only goal is effi-ciency of acquisition. The following distinct cate-gories of acquisitions emerged from these discussionsas a basis for comparison:
Acquisitions that affect the administrationof the AF as an institution
* Acquisitions that directly affect the abi-lity of the AF to perform its mission.
This distinction separates acquisitions into two cate-gories: those for which it is appropriate for effi-ciency of-the acquisition process to be the exclusiveconsideration, those for which the effectiveness ofthe acquired system also needs to be considered. Inlight of technological changes and increased depen-dence on general purpose ADP use, more acquisitionsare moving from the administration category to themission essentiality category. Recent legislativechanges establishing "exempt" and "non-exempt" cate-gories for ADP systems is a recognition of this factat the Federal level.
IV-6
Because the current process responds primarily tothe goal of maximizing efficiency, cost versus effec-tiveness tradeoffs are not generally performed. Thesource selection process at AFCAC consists of choosingamong qualified vendors on the basis of lowest cost.That is, if one vendor can provide much greater per-formance than another who still meets minimum speci-fications, this greater performance will not result inhis winning the award if his price is higher. Formany systems, the more effective they are, the betterthey respond to the AF mission. Yet the process isnot designed to examine the whole range of levels ofeffectiveness that can be achieved by performingtradeoffs between effectiveness and cost.
(3) The ADP Acquisition Process Lacks the ProperDegree of Planning to Effectively Focus onMission Needs
This statement does not imply that the Air Forcedoes not adequately plan and program funds to meet itsADP needs. On the contrary, managers feel that theirADP programs are well represented in the budget pro-cess and are usually successful in receiving requestedfunds. Instead, this conclusion asserts that currentlong range strategic planning fails to provide theappropriate mission perspective.
The lack of effective strategic planning has anadverse impact on the effective performance of the ADPacquisition process, the extent of which is difficultto determine. It was noted by many of those who re-sponded during the Government and industry interviewsthat current Air Force ADP planning is "programmatic"and "reactionary" in that it responds to existingneeds, but does not adequately plan for future re-quirements. Specifically, the lack of adequate stra-tegic planning results in a process which does notadequately respond to evolving needs, fails to exploittechnological trends and innovations, and fails toprovide a basis for rational oversight and control.
1. Current Top-Down Strategic Planning Does NotProvide the Necessary Guidance for the AFADP Program
Experienced ADP users in the private sectorperceived the function of ADP strategic planningas th~eefold. First, it identifies the evolvingfunctional needs beyond the immediate future
IV-7
chne) n hid tpoiessrtgero
(i.e., 2-7 years); second, it considers needs interms of changing environmental conditions (e.g.,technological, organizational and legislativechanges); and third, it provides strategies for
attaining goals and objectives. An analysis ofthe Hq USAF ADP Plan (AFADPP) was made in lightof these criteria. The following deficiencieswere noted:
The AFADPP is a compilation of ADPplans and programs based on individualfunctional area managers' views oftheir general purpose ADP needs anddoes not adequately address the imple-mentable top level strategic planningobjectives. Two areas were noted:
- Top-down ADP policy addressingstrategies for integrating theneeds of various functional userswas not apparent.
- Future impact of technologicalchange on the Air Force missionwas treated only generally.
The AFADPP is an after-the-fact docu-mentation of decisions on individualprograms which address present needs.It does not provide the necessary levelof technology planning to guide the ADPmanagers in developing subordinateplans for the programming and budgetingof future needs.
2. Long-Range Planning at the MAJCOMs and SOAsDoes Not Provide a Mission-Oriented Focus
The requirement for strategic planning can-not be confined to the Headquarters. MAJCOMs andSOAs are required to identify programs which arenot only cc5mpatible within the Air Force as awhole, but also consistent with internal goalsand objectives. A strategic planning functionwhich integrates mission objectives and environ-mental considerations into a realistic set of ADPgoals and objectives at this level was not ap-parent. The MCAP process was the only evidenceof long range planning. While it serves the PPBSprocess well, the MCAP process is largely pro-grammatic and is not effectively linked to a
IV-8
coordinated set of mission-oriented objectives.The following observations are further evidenceof the lack of strategic planning at the MAJCOMsand SOAs.
* The AF ADP plan was universally viewedto be of little value in the planningactivities at this level.
* The MCAP is a collection of ADPprograms (PARs) ranked according toneed. It does not provide a masterplan for the evolution of ADP withinthe MAJCOM/SOA.
* Initiation of ADP programs in the DARprocess appears to be independent ofthe planning process. The fact that aDAR must identify a source of funds isthe only linkage to the planningprocess.
This conclusion describes how the ADP acquisitionprocess underemphasizes mission essentiality by inappro-priately focusing on hardware savings through competitionand failing to provide effective linkage between missionneeds and ADP planning. The hardware and solutional focusof the process is further evidenced in a close examinationof the process itself presented in the next major conclu-s ion.
3. THE CURRENT ADP ACQUISITION PROCESS IS INFLEXIBLE ANDOVERLY RESTRICTIVE IN THAT IT IS NOT ADAPTABLE TO THEVARYING NEEDS FOR WHICH IT IS BEING USED
The current process applies the same documentationrequirements and processing procedures regardless of thescope or type of the acquisition. Tailoring of the pro-cess or the documentation to fit the acquisition seldomoccurs -- even in those instances where it is permitted.The process is inflexible both by design and by virtue ofthe implementation practices that have evolved. Thisinflexibility can be seen by examining current regulationsand the documentation requirements imposed by the process.
(1) The 300 Series Regulations Reflect ThisRestrictiveness and Inflexibility
The 300 series regulations are, in general,applied in the same manner to all types and sizes ofADP acquisitions. While dollar thresholds do control
IV-9
the level of approval, the same documentation andapproval procedures are required for equipment up-grades, new system acquisitions, system replacements,software acquisitions, maintenance services, and othercontract services.
AFR 300-2 and 300-12 impose documentationrequirements without regard to the type of acquisi-tion. AFR 300-12 states that "requirements fordeveloping, modifying or maintaining an ADPS or anADS, or for acquiring ADPS components, will be docu-mented using the DAR format." Under this direction,DAR documentation requirements are applied equally todifferent situations. For instance, DAR's are re-quired for maintenance of systems that are installedand operating. Also, feasibility studies are requiredfor system upgrades, despite the fact that feasibilityhas already been demonstrated since the process isalready automated. This inflexibility thus imposesinordinate costs and delays in preparing, reviewing,and approving unnecessary documents.
Approval thresholds based solely on cost areinflexible. Two acquisitions may be of the same dol-lar cost, yet the complexity of the procurement andimplementation may differ radically. Requirementsapproval fixed solely by the dollar thresholds mayforce decisions to higher levels than should be neces-sary. The workloads of higher level decision makersare thus increased, resulting in delays affecting allacquisitions and causing the risk of possible degrada-tion in the quality of oversight for those acquisi-tions that really merit high-level attention.
The procurement process exhibits similar inflexi-bility, in that the same degree of discipline isapplied to different types and sizes of acquisitions.AFCAC has a set of procedures which are uniformlyapplied to most procurements. The same procedures areapplied to a $20M new system acquisition as are ap-plied to a $300K( equipment upgrade. Statisticalanalysis of 17 AFCAC procurements (Appendix D) showedthat there was a wide variation in procurement time(mean = 14 months and standard deviation = 9.05months) . Cost acounted for less than half of thevariation in the mean time. Data was not available topermit positive determination of the other factorsassociated with procurement time.
IV-10
As indicated above, inflexibility results incosts to the Air Force and delays in the process.Statistically, it is highlighted by the low correla-tion between the size and cost of an acquisition, andthe acquisition lead time (Ref. Chapter III) . SomeMAJCOMs have initiated "mini-DAR" procedures foracquisitions under $100K( to alleviate some of thisinflexibility. There is, however, a need to incor-porate more of this type of flexibility throughout the300 series process.
(2) The 300 Series Documentation Requirements AreDuplicative and Solution Oriented, Resulting inDelays and Less Than Optimal Decisions
The need to develop and cost all the alternativesolutions in detail as part of requirements approvalforces the premature development of solutions.Development of functional requirements and acquisitionstrategies is best done as an iterative process,between the functional users, ADP personnel, and pro-curement specialists. This is particularly true ifthe scope of the functional requirement is large.Premature "freezing" of requirements results in deci-sions based on less than adequate information.
The documentation requirements specified in AFR300-12 contribute to the acquisition lead time. TheDAR package consists of the following documentation:
Executive Summary of the DAR
Data Automation Requirement (DAR)
Statement of Requirement (if telecommunica-tions support is required)
Feasibility Study
Economic Analysis
Executive Justification Summary
Draft Data Project Directive (DPD)
Draft Agency Procurement Request
Statement of Work and GSA Form 2068
ADP Telecommunications Requirements checklist
IV-11
ADPE List and Sole Source Justification (ifrequired)
System Development Notification (if changeto existing WWMCCS standard systems isrequired)
• Software Conversion Study (if required)
Case Synopsis (if SAF/FM is the approvalauthority).
Delays are caused by the duplication among these docu-ments, and the level of detail and review required foreach.
Although there is duplication throughout thepackage, the most duplicative documents are the DAR,the economic analyses, and the feasibility study.Each is a restatement of the ADP requirements origin-ally presented in a requirements document, andrepeated for different purposes in these three docu-ments. The repetition is compounded by the executivesummaries and the staff summaries which must be pre-pared at each level of review.
Not only are these documents duplicative, butthey also contain an inordinate amount of detail. Theoverall sense conveyed by AFR 300-12 and HQ AF 01300-2, Pir Staff Automated Data Processing ActionOfficc s Handbook, is that much detail is required.AF 1 300-2 describes the DAR, with its 15 sections,F the requirement to discuss and provide data on,iternatives, e.g., all equipment types, cost and.ource, and cost data for each alternative. It statesthat the feasibility study "should be as comprehensiveas resources permit .... " The economic analysis is aparticular case in point. AFR 178-1, Economic Anal-yses and Program Evaluation for Resource Management,requires that all alternatives be costed and ana-lyzed. AFR 300-12 states that "each economic analysisshould be only as detailed as the situation dictates,"but it goes on to require that the analyses be done inaccordance with the 11 steps prescribed in the AFR300-12 plus the requirements of AFR 178-1. The appli-cation of this rigorous development of costs so earlyin the process substantively adds to delays.
Each of these documents requires time for prep-aration, review, coordination and approval. The re-view, coordination and approval are repeated at all
IV-12
the levels as the DAR moves up the chain from theoriginator, through the MAJCOM and MAJCOM single man-ager, and up to Hq USAF (if required). Within thelarge ADP staffs of MAJCOMs like SAC, the number andlevels of review are considerable, each adding to thelead time and often contributing little value to thefinal outcome.
This conclusion describes the restrictiveness andduplicative documentation requirements of the current ADPacquisition process under the 300 series regulations, andhighlights the resultant costs and delays. Further,delays are introduced by anomalies in the AF ADP manage-ment structure, as discussed in the following conclusion.
4. THE AIR FORCE ADP MANAGEMENT STRUCTURE AND ORGANIZA-TION CONTRIBUTES TO THE CONFUSION AND SUBSEQUENTDELAYS IN ACQUIRING ADP
The following four specific problem areas relate tothe Air Force ADP management and organizational anomalieswhich impact the effective execution of ADP acquisitions.
* Inconsistency in the ADP management structure
* Confusion in the scope of application
* The lack of accountability and end-to-endmanagement
Ineffective use of human resources.
The process implications of these factors are discussed inthe following paragraphs.
(1) The Air Force Has Established a Unique Structurefor the Management and Acquisition of ADP, Unlikethe Process It Uses for Acquiring Other Systems
The Air Force ADP management structure at theHeadquarters USAF was cited by many as a major sourceof confusion in processing ADP requirements. Much ofthis confusion stems from the multiple and overlappingsets of policy which have been externally imposed onAir Force ADP management. much of it, however, is adirect result of the way in which the Air Force hasimplemented these policies. In response to DODDirectives 4105.55, Selection and Acquisition ofAutomatic Data Processing Resources dated May 19, 1972and 5100.40, Responaiix fo heAministration ofthe DOD Automatic Data Processing Program dated August
IV-13
1975 which implemented 0MB Circular A-71, the AirForce established a unique management structure forADP. Much of the conflict and confusion identifiedduring the study investigation is traceable tomanagement anomalies in the Air Force structure whichare neither directed nor implied in these higher leveldirectives.
In establishing an organization to manage thegeneral purpose ADP program, four areas of responsibi-lity were centralized in the predecessor organizationof the Directorate of Computer Resources (AF/ACD).This centralized assignment of responsibility for ADPis not consistent with the way in which the remainderof the Air Staff is organized.
In managing General Purpose ADP, AF/ACD hasresponsibilities which duplicate the responsibilitiesof three other organizations at the Air Staff.
* The DCS for Research, Development and Acqui-sition (AF/RD) is responsible for acqui-sition policy and acquisition direction.
* The DCS for Operations, Plans and Readiness(AF/XO) is responsible for analyzing andplanning for Air Force mission needs,including planning for high technology needssuch as command and control andtelecommunications.
* The DCS for Logistics and Engineering(AF/LE) has policy and management respon-sibilities for all "personal property"(i.e., property other than real property) inthe Air Force.
Interviews with these organizations highlighted fourareas of responsibility which demonstrate the anoma-lous aspects of Air Force ADP management.
Acquisition* Operations* Requirements* Support.
The responsibility for acquisition policy is car-ried out by AF/RD through the Defense AcquisitionRegulations (DAR), which dictate the acquisitionpolicy for DOD. The DAR, however, do not reflect thecurrent Federal Procurement Regulations (FPR) which
IV-14
r1
apply to general purpose ADP. Though changes are inprogress to correct this, FPR policy has been directedthrough AF/ACD in the 300 series regulations and theindividual procurement delegations.
The responsibility for operational policy (i.e.,setting Air Force goals and objectives for the use ofhigh technology) is also divided. AF/XO, as part ofmission planning, maintains technological expertise inareas such as command and control, and telecommunica-tions. Maintaining a similar, and somewhat redundant,level of expertise for general purpose ADP planning isthe responsibility of AF/ACD. This is particularlysignificant since telecommunications and ADPtechnologies and applications are rapidly merging.
The requirements validation and approval, andprogram direction for ADP is the third area ofirregular procedure. The functional validation ofnon-ADP requirements is analogous to ADP requirementsapproval. However, the following three points,offered for comparison, indicate that the treatment ofthese requirements is not analogous:
For non-ADP requirements, functional valida-tion is performed by the functional areaexpertise on the Air Staff such as AF/LE, orAF/XO.
For non-ADP requirements, functional valida-tion and approval is the basis for obtainingfunds.
Non-ADP requirements are approved by theRequirements Review Group.
By contrast, ADP requirements approval is the ultimateresponsibility of AF/ACD, by virtue of the signatureauthority for the ADP solution, and this occurs afterfunds have been identified. It was noted that theTable of Allowances (T/A) more closely resembles theDAR approval process in that a requirement is approvedindependent of the funding process, by the officeresponsible for management of the required commodity(i.e., AF/LE), and results in an authorization toacquire a particular item. The T/A, however, is notprimarily intended to be used for high technologyitems.
IV-15
There is a similar duplication of responsibilityin program direction. Both AF/RD and AF/ACD have theauthority to direct acquisition programs. AF/RD canissue Program Management Directives (PMD) under AFR800-14, and AF/ACD can issue Data Project Directives(DPD) under AFR 300-2.
The fourth problem area is the unclear delinea-tion of responsibility for support management of ADPresources in the Air Force. AF/LE claims responsibi-lity for managing all "personal property" in the AirForce including ADPE. AF/ACD, as the central managerof ADP, has responsibilities for sharing and reutili-zation programs, as well as overall management of theAir Force computer inventory.
These organizational anomalies at the Air Staffwere described in the field as being a major source ofconfusion in conducting ADP acquisitions. Management-related confusion is also evident in an analysis ofthe differences in the scope of application betweenAFR 300 and AFR 800 type acquisitions.
(2) In Setting Up a Separate Process for General Pur-pose ADP Acquisition, the Air Force Has Intro-duced Confusion as to Its Scope of Application
In establishing a separate process for acquiringgeneral purpose ADP, and thus a separate managementstructure, the Air Force has introduced confusionbetween the scope of application for AFR 300 seriesand AFR 800 series. The existence of this ambiguityis illustrated by the fact that a survey of AF offi-cials elicited varying responses as to what belongs ineach process approach. There is little corporatelevel review to insure that acquisitions are beingchanneled in the appropriate direction. Decisions touse a particular acquisition process are made largelyat the MAJCOM's based on personal preferences ratherthan official guidance. The details of this investi-gation follow.
1. The Lack of Precise Distinction Coupled Withthe Lack of Proper Direction Has CreatedConfusion in the Scope of the Air ForceRegulations
The scope of the 300 series process encom-passes General Purpose ADP resources includingADPE, software, computer programs, ADP contrac-tual services, ADP personnel and supplies. ADPEin this case is defined as:
IV-16
General Purpose, commercially available ADPcomponents and the equipment systems createdfrom them....which are designed to beapplied to the solution or processing of avariety of problems or applications, butwhich were not specifically designed (asopposed to configured) for any specificapplication.
This scope explicitly excludes embedded systems.An embedded system is ADPE that is "integral toand in direct support of a combat weapons sys-tems." A combat weapons system is defined in AFR300-2 as:
an instrument of combat, either offensive ordefensive, used to destroy, injure, orthreaten an enemy.
Embedded systems are acquired as part of theoverall weapons system acquisition process whichis governed by the 800 series regulations. Thedifficulty arises, in part, because there aremany systems for which there is ambiguity as towhether or not they should be regarded as inte-gral to and in direct support of a combat weaponssystems.
The consequences of the lack of precisedistinction in the scope of application of thesetwo types of acquisitions have been confusion andan expansion in the scope of use of the 300series approach. There was much criticism leviedat the 800 series process by the MAJCOMs. Themajor complaints were that the process took toolong, the MAJCOMs lost control over their ownacquisitions, and the delivered capabilitysometimes did not meet the requirement. Thesecriticisms, in conjunction with the absence ofprecise definition and corporate level guidance,have contributed to the heavier reliance on the300 series approach. Unfortunately, this processdoes not explicitly provide for the appropriatedegree of engineering discipline required formany acquisitions.
IV-17
2. The AF-300 Process Does Not Provide for ManySystems Disciplines Which Are Needed forMany of the Acquisitions That Fall Withinthe Scope of the Process
The AF-300 series was designed to acquireresources requiring little or no special designand development. Therefore it does not explicit-ly provide for many of the systems disciplines,noted in Table IV-2, that are normally associatedwith acquisition of complex systems. Many ofthese disciplines are provided for in the AF-800series regulations. Although provision for thesesystems disciplines does not guarantee success,they have been found to substantially improvechances of successful acquisition of complexsystems.
Table P1-2: Engineering Disciplines inSystems Acquisitions
Planning DocumentSystem Discipline (in AFR-800 Series)
Systems Engineering Systems EngineeringManagement Plan (SEMP)
Logistics Support Integrated LogisticsSupport Plan (ILSP)
Computer Resource Computer ResourcesIntegration Integrated Support
Plan (CRISP)
Configuration Management Configuration
Management Plan
Operational Planning Operations Concept
Test and Evaluation Test and EvaluationMaster Plan
Data Management Data Management Plan
Quality Assurance Quality Control Manual
IV-18
There are many acquisitions that fall withinthe scope of the 300 series process which appro-priately belong in a process that explicitly pro-vides for these disciplines. The Missile WarningBypass acquisition encountered delays in theapproval process due to poor understanding of thesystem at AF/ACD and GSA. Specifically, theEconomic Analysis was questioned and the need fora software conversion was incorrectly raised.ADCOF representatives indicated that a moredisciplined approach during requirements identi-fication might have avoided these problems. Thissystem could have been acquired under the 800series had the using MAJCOM decided to do so.
The Advanced Logistics System (ALS), how-ever, could not have been processed under the 800series, given current definitions. Some of theserious problems encountered by ALS are of thetypes that these systems disciplines are intendedto prevent. These problems included the choiceof a computer that lacked a needed parity check-ing feature; the fact that 90 percent of computercapacity was used by overhead software (i.e.,only 10 percent of capacity was available for theapplications programs); and the specification of90 percent reliability for system components whenthe entire system had a requirement for 99 per-cent reliability.
In a more recent acquisition, the Phase IVprogram also encountered problems that might havebeen avoided by a more disciplined acquisitionprocess. Although a program management officewas established, the office lacked personnel withadequate training and experience in the systemsdisciplines. For example, there were norepresentatives from ATC or AFCC despite thetraining and communications impacts of theprogram. Further, delays resulted becausefacility requirements were not recognized in timefor AF users to make necessary modifications.
The lack of explicit provisions for theseengineering disciplines in the 300 series processis not the problem. There are many acquisitionsfor which they are not required. However, thedeficiency is that acquisitions requiring a highdegree of engineering discipline can and arebeing processed using the APR 300 series approachand that sufficient corporate level screening,guidance, and direction are lacking.
IV-19
(3) There Is a Lack of End-to-End Management andAccountability in the Current Process Design ThatInhibits Operational Efficiency in ADP Acquisition
In discussing the anomalies in the Air Forcemanagement of ADP earlier in this section, it wasstated that the resulting confusion affects the per-formance of an ADP acquisition program. Responsibili-ties transition among functional, ADP, and contractingspecialists at the MAJCOMs/SOAs and at the Air Staff.Consequently, there is an apparent lack of account-ability and end-to-end control for an acquisition.
1. No One Organization Is Responsible and ThusAccountable for the Life Cycle of an ADPAcquisition
One element of the detailed evaluation ofthe ADP acquisition process was an examination ofthe organizational participants who must interactin the process. This investigation revealed thata lack of clearly defined roles in the acquisi-tion process results in confusing and delayingtransitions of responsibility. The followingthree concerns pertain to these transitions:
* Transitions of responsibility must be accom-panied by information exchanges which aredelaying and contribute to the confusion.
* Major decisions affecting the acquisitionprogram are ultimately made by differentorganizations with different interests. Theresult is a lack of accountability by anyone organization for an entire acquisition.
Transitions of primary responsibility withinuser, ADP and contracting organizationsencourage an adversary relationship amongthe acquisition process participants.
A description of the transitions of controlduring each phase of an acquisition illustrates
these three points.During each phase of the acquisition
process, there is at least one instance whereprimary responsibility transitions from oneorganization to another. The requirementsdefinition phase begins with the identificationof a need by an originating organization. During
IV-20
this phase, primary responsibility for defining,analyzing and justifying the requirement isassigned within the organization, normally to afunctional area manager. The ADP personnelsupport the functional area manager, and togetherplay a strong role in the critical decisionsaffecting acquisition strategy and programimplementation that are normally made at thistime. The data automation support continues tobe provided to the user until the DAR is com-pleted and ready for staffing. At this point inthe process, primary responsibility transitionsto an ADP manager, and with it the responsibilityfor the DAR and all supporting programdocumentation.
During the requirements approval phase, pri-mary responsibility remains within the data auto-mation chain of command. A MAJCOM or SOA singlemanager or an ADPS manager must review and ap-prove the requirements package and staff itthrough higher level authorities as required.When a delegation of procurement authority isrequired AF/ACD must then assume primary respon-sibility for the program in communicating withGSA, and the HGOC if necessary. This again in-volves a significant transfer of information,over and above what is, or should be, requiredfor requirements approval.
Program management authority is assignedthrough the Data Project Directive (DPD), markingwhat was observed as the most confusing transi-tion in responsibility. Normally, the DPD willassign program management responsibility to aMAJCOM/SOA ADP manager or ADPS manager, andprocurement responsibility to either AFCAC orbase level procurement. Thus, a working rela-tionship among the user, ADP and contractingstaffs must be established to effect the smoothfunctioning of an ADP acquisition. However, thedelineation of roles and responsibilities is notalways clear. The user, in parallel with re-quirements approval, normally prepares draftspecifications to be incorporated in the RFP.Since the DPD marks the initial involvement ofthe contracting staff (especially if it isAFCAC), there must be a "meeting of the minds"among the three parties. When primary and sup-port responsibilities are confused, this exchangeof information becomes more difficult andtime-consuming.
IV-21
Many of the difficulties discussed above areevidenced by the actions of the single managers.For example, to avoid delays, the managers fre-quently use base level procurement support in-Istcad of AFCAC, despite their admission thatAF ' uoe < a better job in the procurement.Th ir u criticism is that in using AFCAC theylos, Lrol of their own acquisitions. Thiscriticism is justified. AFCAC is responsible forall procurement decisions, ranging from decidingon the type of specification that will be releas-ed in the RFP, up to and including the sourceselection. Once the contract is awarded, primaryresponsibility is returned to the MAJCOM/SOAprogram manager to monitor the performance of thecontractor. After contract award, AFCAC'sinvolvement is significantly reduced.
2. There Is No Management Control System WhichMonitors the Effectiveness of the EntireProcess
A second indication of the lack of end-to-end management was the difficulty experienced incollecting aggregate statistics on process per-formance. Attempts to assess the costs incurreddue to extended acquisition lead time or lostcapability were hampered by the lack of an effec-tive management control system to monitor thesecosts. Omissions in the management control sys-tem include the following:
The lack of a data base of performancemeasurement indicators
The absence of a mechanism for auditingthe results of an acquisition.
The lack of aggregate performance indicatorsis discussed in Chapter III. Just as significantas performance measurement in a management con-trol system is the ability to audit the resultsof an acquisition to determine if the requiredlevel of capability was acquired and if thepredicted benefits were achieved. Acceptancetesting is the only audit of capability currentlyperformed. While the test verifies that theacquired capability meets a minimum set of speci-fied requirements, it does not insure that therequirements were properly specified, or that thebenefits documented in the economic analysis havebeen achieved.
IV-22
(4) Acquisition Under the Current Process DemandsSkills That Are Scarce Within the Responsibleorganizations, and This Inhibits the Efficiencyof the Process
One of the most frequently encountered difficul-ties in the management of ADP resources in the AirForce was the inability of the user and contractingpersonnel to respond to the requirements of thecurrent acquisition process. The two principal pro-blems are the limited experience and the lack of suf-ficient training of the staff. Air Force representa-tives at all levels from base up to Hq USAF have notedthese problems. They were cited as a major source ofdelay in preparing adequate documentation, particu-
larly at the beginning of a program.
The following findings support the conclusionthat personnel resource limitations are a major sourceof delay in the process:
Each phase of the process involvesactivities that require special skills.
There is a limited pool of personnel withadequate training and experience in therequired skill areas.
The remainder of this conclusion details these twofindings.
1. En Order to Conduct an Acquisition, SpecialSkills Are Required in Each Phase of theProcess
In each phase of the process, special skillsare required to prepare, review and approve thedocumentation required by AFR 300-12. Theseskills and their related activities are presentedin Table IV-3.
Requirements definition is the responsibil-ity of the functional personnel assisted by theADP personnel as necessary. The preparation of aDAR and the supporting documentation requires acollection of diverse skills. The requirementsmust be spelled out in functional terms thatclearly articulate what is needed and what bene-fits will be achieved. The process calls forrequirements to be defined in sufficient detailto permit the early development and costing of
IV-23
.-$ C -4
4-1 >1 (a 4-J E 4)(a.- 1 C- c -U) ) H - mQ -4 w.(a r 1,a, c a a'C r- 0 ( a o a).. UflC 4-'C 0 c. CO 1--4 C - C V
U) u) C ---- _)U <4- C) H) _'- .) -4- aU ((D a)t) .1 L U) 4) W)> A.) a > w a) L C C0>, 4-J M0) a)' M ~~ 4)' M ~ (1 ) a) O ) -O -0
(n U)-qU) ).-4CU U) (a 0) u U) Q a) 4) 0) - 4-JM0 4 H c C'- 4J C c -4-J-4-1 U)
-4 .- C3'C U C'4 '-4 V r-' '4 *44)C(a O cm 0) '-q M a m ~ Z m C w a') 'ma O~ )- u Oc C CNC - E CC) u .C L cr _ -Q=C IC ) c C4- E UZ N-414 0 U) - : W U 0 -- a) w E 0 -Wi'4 0~ a)O w a) C (F= *'4 E a E fu 0 - C ( ..IJ -J F- a CZ41 (U 0 r.. U~C 4-J0 O-C -; 0) V) U) -
() 4- >C)Q a) ~ c 0,a) uOa) -O rOL a)C( CW 4 CC: 0 C'"-a >C-) 0-.a >C ) .- c~)-. C. > u 0 0 'C),
4-,0 U)
4--)
CU)W
I- ra
4-)U)-m 3 m>
OH - C E -
c- 0 - ac CC) :3 a) 14 u U Z-&C) 4) a4<)
a) a) Q)Q . .
U- ) 4- U-)$ 0U4 U 0 4-4 (0- 4J
a a) ma a)>E ) a)' >- ca) 4
'4-)
0~ -- E4
-4-2
all alternative solutions. These solutions, andtheir costs, must include all aspects of theacquisition, namely hardware, software, mainte-nance, documentation, training, facilities, man-power and telecommunications.
Requirements approval implies the approvalof both the functional requirement and the ADPsolution. The two are tied together in the DAR.The approval of ADP solutions requires a broadknowledge of ADP hardware and software technol-ogy, and the relationship between this technol-ogy and a spectrum of functional problems. Thereview and approval process starts at the baselevel, includes a number of levels of authority,and requires a number of functional specialtiesin addition to the primary "user." In many casesthe other functional specialists are inexperi-enced in ADP. Nevertheless, they are required toreview and approve documentation that combinesthe functional requirements and the ADP solutions.
Specification development is done by the ADPpersonnel working in conjunction with the func-tional users. It represents the translation of afunctional requirement to a technical requirementfor hardware, software, services or combinationsof all these. The specification must also ad-dress facility modifications and telecommunica-tions. In many cases, it is a requirement todevelop a technical specification suitable forcompetitive procurement. Procurement personnelcited the problem of non-competitive specifica-tions as a major delaying factor in dealing withthe users.
Procurement of ADP resources is a special-ized branch of procurement, externally con-strained by the requirements of FPRs and FPMRs.This type of procurement is also complicated bylease vs. purchase decisions (including varyingoptions to purchase) and long term supportoptions. The ADP industry has developed a uniqueset of proposal/costing methods. These uniquemethods make ADP unlike anything else the AirForce commonly procures. Sophisticated ADP pro-curing agencies such as AFCAC have developedcomputer software models to evaluate costs ofcomplex acquisitions. Local procurementpersonnel need to be knowledgeable in the use ofthese models.
IV-25
2. The Skills Required by the Current ProcessAre Lacking and Training Is Limited
The previous section established that spe-cial skills are needed to execute the 300 seriesprocess. However, most AF organizations lack anadequate pool of personnel with the requisiteexperience and training. This lack of trainedand experienced personnel results from the infre-quency with which functional personnel are calledupon to articulate functional requirements.
For many personnel, DAR preparation occursonce in a career. Only in very few cases arefunctional personnel frequently called on to pre-pare DARs. Consequently, formal training in DARpreparation is not given. None of the personsinterviewed who had previously prepared DARs hadever received training for it.
ADP specification development and procure-ment are closely related. Like the functionalspecialists, the ADP personnel seldom have todevelop specifications. Technical specificationwriting of any kind is usually not done anywherein the Air Force except in AFLC, AFCC, AFSC andin AFCAC. The specification preparation guide ofAFR 300-12 is not a replacement for training orexperience. Training in ADP specification writ-ing is not generally available, and the abilityto produce truly competitive specificationsrequires the close attention of ADP technicianswho have extensive procurement experience. Thereis an inadequate pool of these experienced ADPtechnicians to support local contractingsquadrons.
The above-mentioned shortcomings in skillsand training are further compounded by the over-all shortage of middle level managers in theofficer corps. For example, the Air Force isusing Lieutenants to fill Captain and Major posi-tions in the technical career fields. one seniorADP offical commented that not only is there alarge number of junior officers, but there isalso a shortage of experienced personnel to givethem on-the-job-training.
This conclusion delineated four specific problem areaswithin the Air Force ADP management and organizationstructure that contribute to confusion and delays and
IV-26
inhibit the efficient and effective functioning of the ADPacquisition process. To this point, the conclusions havedealt with problems areas within the Air Force. The re-maining conclusion addresses the constraints imposed bythe external environment.
5. THE CONFUSION AND INEFFICIENCIES INTRODUCED INTO THEADP ACQUISITION PROCESS BY TOP-LEVEL FEDERAL POLICYHAVE THEIR ROOTS IN CONFLICT OVER THE PROPER DIRECTIONOF ACQUISITION REFORM
The Air Force ADP program functions within the Federallegislative and executive processes which comprise anexternal environment over which the AF has little con-trol. An intricate policy framework has evolved fromPublic Law 89-306. This act establishes the basic struc-ture and concepts for managing the acquisition of generalpurpose ADP within the Federal Government. Chapter IIdescribes the organizational structure and summarizes themultitude of policies, regulations, and directives subse-quently issued to implement the Government-wide policydirection at all tiers of the organizational hierarchy.This section addresses the constraints within the externalorganizational framework which adversely impact the func-tioning of the Air Force ADP program. In particular thefollowing three areas of concern are addressed:
* Conflicting philosophies at the congressionallevel
Indistinct roles and responsibilities of OMB andGSA as ADP policy authorities
* The role of ASD(C) in responding to contradictoryguidance and implementing policy.
The implications of these constraints are explored in thefollowing paragraphs.
(1) Philosophical Differences Between the Key Commit-tees Influencing ADP Acquisitions Have CreatedConfusion Within the ADP Community as Organiza-ti'ons Attempt to Respond to the Conflicting Com-mittee-Guidance
Chapter II describes the roles of the HouseAppropriations Committee (HAC) and the House Govern-ment Operations Committee (HGOC) , as two of the keycongressional committees impacting the acquisitionprocess. The dichotomous philosophies espoused bythese Committees in fulfilling their responsibilities
IV-27
in policy making, budgeting, and acquisition revieware the source of much of the confusion which cur-rently exists in the Air Force. Manifestations ofphilosophical conflicts at the committee level includeinconsistent congressional direction, micro-managementby committee oversight, and frequent intervention inpolicy development and implementation. The result isconfusion within the ADP acquisition community as im-plementing departments, agencies, and organizationsare forced to interpret and function under unclear andoften contradictory policy guidance.
The influence of organizational elements whichexercise oversight management forces the process tofunction in an environment of externally imposed,incompatible constraints. Due to this lack of clearpolicy and direction, a great deal of interpretationon the part of OSD and the AF is required in order toimplement ADP policies and procedures. In tracing theorganizational conflicts of the congressional level,two points must be addressed:
* There is a philosophical conflict over theoptimum method for achieving economy andefficiency in acquisition.
* There is a tendency for agencies to respondto the HGOC position due to its oversightresponsibility.
The following sub-paragraphs deal with the implications ofcongressional interest in the process.
1. The HGOC Stresses Maximum Competition toAchieve Efficiency, While the HAC Focuses onLowest Total Overall Cost
The HGOC and the HAC have a basic disagree-ment concerning the optimum method to achieve thegoals of economy and efficiency in ADP acquisi-tion. Their philisophical differences are em-bodied in their interpretation of economy andefficiency, and thus their method of achievingit. on the one hand, the HGOC mandates theachievement of short-term economy and efficiencythrough lowesi- cost procurement by maximum hard-ware competition, while the HAC espouses lowesttotal overall cost (LTOC) in achieving the lowestcost to the Government in the long-term. Thefollowing points of contention highlight theirdifferences:
IV-28
• Interpretation of PL 89-306
. The role of competition
Factors to be considered in lowesttotal overall cost evaluations.
Under its oversight responsibilities withrespect to implementation of Public Law 89-306,the HGOC appears willing to incur significantdiseconomies in order to achieve more widespreadcompetition. This view is not shared by HAC,whose primary interest in ADP acquisition andmanagement is the achievement of LTOC over thesystem life cycle. While PL 89-306 provided forthe "economic and efficient" management of ADP,ADP technology has changed sufficiently thatoveremphasizing hardware costs hampers theachievement of these goals. In the interest ofeconomy and efficiency, agencies are required tooffer ADP hardware procurements to open competi-tion to obtain lowest prices. However, thistechnique does not give proper consideration toother cost factors. The result is that LTOC isnot achieved because significant cost factors areoverlooked and, as a result, agencies suffer fis-cal loss and even mission degredation at theexpense of maximum hardware competition.
PL 89-306 designated the General ServicesAdministration (GSA) as the responsible agent foracquisition, use, and maintenance of ADP, and theOffice of Management and Budget (OMB) was givenfiscal and policy control. GSA exercised itscentral procurement authority by issuing FederalManagement Circular (FMC) 74-5 as implementingpolicy guidance for ADP procurement. Until itsrecent revision and incorporation into the FPMR,this circular was the key policy directive imple-menting PL 89-306. Its practical interpretationby GSA emphasized maximum hardware competition.
While the HGOC supports the achievement ofmaximum hardware competition as the primarystrategy to maximize efficiency, the HAC believesthat a strategy which emphasizes LTOC over thesystem life is more realistic. In conductingacquisitions urder FMC 74-5, GSA has emphasizedthe hardware cost factors over other LTOC costfactors on the premise that a complete cost eval-uation, including conversion costs, inhibits com-petition by favoring the incumbent vendor. While
IV-29
competition does lower prices when hardware isthe primary cost factor, these savings may benegated when other factors that determine overallcost are included. The HAC, the SAC, and many ofthe agencies believe that all relevant life cyclecosts must be considered (whether or not theycontribute to competition) to achieve the mosteconomic and efficient solution. The HAC hasinstructed DOD to implement LTOC guidelines; how-ever, agencies such as DOD must respond to thedemands of the HGOC and GSA (which conflict withthe HAC direction) in order to obtain a DPA.
2. HGOC Exerts Greater Influence Than the HACDue to Its Direct Involvement in ReviewingAgency Procurement Requests
DOD and other agencies have attempted torespond to the conflicting congressional guid-ance, but a dominance by the HGOC and GSA in theacquisition process has directly influenced agen-cies' responses. Due to the HGOC and GSA manage-ment oversight responsibilities, the Air Forcehas been constrained by the philisophical pre-ference for maximizing hardware competition overLTOC. Agencies must follow the GSA procurementregulations and structure their APRs to ensuremaximum free and open competition, or run therisk of being unable to satisfy their ADP needs,despite the expressed intent of the HAC that theagencies strive for LTOC.
The influence of the HGOC on agency ADPacquisition is further evidenced in the risingattention being given to approaches aimed atincreasing hardware competition. Two suchapproaches are:
U tnbundling of hardware and softwareacquisitions to encourage hardwarecompetition
* Standardization of computer higherorder languages (HOL) to minimize theeffects of software conversion.
Both approaches are potentially costly to thefunctioning of the AF ADP acquisition process.Unbundling increases the number of acquisitionsrequired to support a total system and thusincreases the administrative burden and costs of
IV- 30
conducting the acquisition. The issue of stan-dardization is receiving more emphasis at alllevels of policy making. Attempts to standardizein the past have had adverse impact on acquisi-tion lead time and technology lag. Converting toHOL software for the purpose of broadening com-petition is itself costly and extends the acqui-sition lead time.
Both the "unbundling" and "standardization"issues have serious implications to the ADPacquisition process. Their costs and benefitsmust be carefully considered. Because of thepotential impact, these issues warrant specialconsiderations, over and above what could be ac-complished within the time frame of this study.
(2) A Further Impasse to Effective Policy Developmentand Implementation Results From Indistinct Rolesand Responsibilities of 0MB and GSA in the ADPPolicy Environment
The conflicts evidenced at the congressionallevel are not unique. Incompatible acquisition phi-losophies, obscure lines of authority, fragmentedresponsibility, and uncooperative interactions per-meate the Federal poli~cy-making environment. The lackof concise ADP acquisition guidance which has charac-terized the acquisition environment since the passageof the Brooks Bill has been widely attributed to 0MBand GSA. Factors associated with these two executiveagencies that have contributed to the policy implemen-tation confusion (as previously discussed in relationto the congressional level conflicts) include thefollowing:
* Indistinct role definition of apparentlyauthoritative policy organizations
* Conceptually paradoxical policy direction.
This discussion explores these characteristics of theexecutive br anch policy environment.
1. There Are Jurisdictional Conflicts Between0MB and GSA
0MB and GSA each have ADP acquisition roles,assigned by PL 89-306. GSA was made the soleprocurement authority, responsible for the acqui-sition, use and maintenance of ADP equipment; 0MB
IV-31
was assigned responsibility for policy and fiscalcontrol. The coexistence of these organizationshas led to diverse interpretations as to thefunction and responsibility of each. In addi-tion, uncooperative organizational interactionshave "yielded narrow, inconsistent, and incom-plete policy." (see ref. 1). Since the 1965Brooks Bill, GSA has acquired, through a seriesof executive orders, and assumed, due to OMB'sinaction, much of the ADP policy making respon-sibility, in addition to its procurement respon-sibilities. This shifting of roles has resultedin a procurement-oriented policy, with over-f lapping organizational authority. There havebeen shifts of responsibility and authoritybetween and within 0MB and GSA that have hamperedtheir effectiveness in issuing consistent policyguidance. Two sources of the jurisdictionalconfusion warrant amplification:
The shifting functions and authorityrevealed in historical perspective
The coexistence of policy organizationswithin each agency.
These two factors have heavily contributed to thejurisdictional conflicts and resultant policydeficiencies.
(1) The Organizational History of ADPPolicy as It Relates to 0MB and GSA IsCharacterized by Continuous Confusion
In 1959, 0MB instituted a program andorganized a staff toward better overallmanagement of ADP technology in recognitionof Executive Branch concern for centralleadership. From 1960 to 1965, 0MB exer-cised its policymaking responsibilitiesthrough issuance of bulletins and circu-lars. In 1965, PL 89-306 established GSA asthe central procurement authority.
Prior to the passage of PL 89-306, andconsistently since its passage, 0MB hasopposed the Act because they disagreed withthe need for legislation to establish themanagement functions that 0MB had performedsince 1959 (see ref. 1 p. 67). The BrooksBill exhaustively defines the role of GSA
IV-32
and fails to amplify the role of OMB, exceptto state that it should exercise "fiscal andpolicy control." While the law incorporatedmost of the features of OMB's previous ADPstudies and circulars, it failed to clearlyidentify OMB as the lead executive agencyfor the management of ADP, the roledescribed by OMB Circular A-71.
In 1973, Executive Order 11717 trans-ferred d considerable portion of OMB's re-sponsibilities for government-wide adminis-tration and financial management to GSA andthe Department of Commerce. GSA receivedADP policy control authority (except inrelation to standards, the establishment andapproval of which were transferred to Com-merce) and OMB maintained general oversightand fiscal control. While the order statesthat "no function vested by statute shall bedeemed to be affected by these provisions,"OMB was effectively divested of its policycontrol responsibilities with respect toADP. According to a recent report by aPanel of the National Academy of PublicAdministration, the intent of the Presidentwas for GSA to "assume a broader managementrole" through "overall leadership" indeveloping government-wide policy for ADP,and the other areas affected by the order(see ref. 3 p. 15-16).
Subsequently, a Department of Justicedecision held that EO 11717 did not relieveOMB of its policy oversight and policy for-mulation responsibilities. This ruling alsoheld that GSA was responsible for "develop-ing policy based on OMB's formulation" (seeref. 1 p. 9-10). This attempt at clarifi-cation served only to increase the existingambiguity by making an unclear distinctionbetween policy formulation and policydevelopment.
In 1975, Executive Order 11893 calledfor certain policy functions that had beentransferred to GSA by EO 11717 to be trans-ferred back to OMB. However, ADP policyauthority was not transferred. Neither GSAnor OMB has exercised a strong leadershiprole, and each has blamed the other for the
IV-33
failure to provide direction to the ADPacquisition community. The result has beena continuing lack of centralized policydirection and long range planning.
0MB has been perceived to be theresponsible official, based on PL 89-306,but has not issued any major ADP policyguidance since 0MB Circular A-71 in 1965.As a result, 0MB has been faulted for"abdicating its policy responsibilities andexpecting GSA to both formulate and developADP policy." GSA has charged that 0MB hasnot provided adequate policy "formulation"to enable them to fulfill their policydevelopment responsibilities under EQ 11717(see ref. 1 p. 10) . OMB has apparentlyperceived its role to be one of "encouraginguse of ADP technology" and "supporting thebudget examination process," as opposed toone of developing ADP policy (see ref. 5).
GSA stepped in to fill this void byincorporating ADP policy into FPR's andFPMR's. Later, in July 1974, FederalManagement Circular (FMC) 74-5 was issued,with the expressed intent to "update andconsolidate ADP policy in one directive."This circular has since been superseded byrevisions to the FPMR.
The recently passed Paperwork ReductionAct of 1980, PL 96-511, directs the estab-lishment of an Office of Information andRegulatory Affairs within 0MB, the head ofwhich shall report to the Director. TheDirector of 0MB "shall develop and implementFederal Information policies, principles,standards, and guidelines" (PL96-511) forautomatic data processing and telecommunica-tions functions and activities. OMB willalso maintain its oversight responsibili-ties. The law states that "nothing in thischapter shall be interpreted as increasingor decreasing the authority conferred by PL89-306 on GSA, Department of Commerce, or0MB." Since PL 96-511 casts 0MB in the kfyADP policy role, it would logically followthat 0MB was assigned to this role by PL89-306, and despite convoluted authoritytransfers and responsiblity adbicationssince 1965, the organization is still thesenior ADP policy official.
IV-34
The only clear aspect of this organiza-tional history is the obvious confusion ithas caused in the ADP acquisition community.This confusion is further evidenced in thesecond source of jurisdictional conflictbetween OMB and GSA -- the coexistence ofpolicy organizations within eachorganization.
(2) A Number of Policy OrganizationsCoexist and Others Have Been Createdand Abolished
The perception of overlapping authorityand fragmented responsiblity has been en-hanced by the presence of a number of or-ganizations that have shifted and evolvedsince the passage of the Brooks Bill. Thefollowing discussion focuses on two organi-zations:
Automated Data and TelecommunicationService (ADTS)
Office of Federal Procurement Policy(OFPP).
Within GSA, ADTS was given responsi-blity to execute GSA's ADP functions underPL 89-306. As previously discussed, theassignment of these responsibilities in GSAis open to varying interpretations. ADTS isauthorized and directed under an amendmentto the Federal Property Act of 1949 to"coordinate and provide for the economic andefficient purchase, lease, and maintenanceof ADP..." (see ref. 3 p. 50). User agen-cies have complained that ADTS "lacks thetechnical competence to be of genuineassistance" (see ref. 3), and takes an in-ordinate amount of time for approval. (Itshould be noted that GSA has made changestoward improving the problems. "ADTS iscommitted to accenting its regulatory roleand strengthening its management areas so itcan provide oversight," (see ref. 3) andthere has been a significant effort to dele-gate procurement authority. An Office ofPolicy and Planning has been created todevelop policy and orchestrate up-frontplanning for both ADP and telecommunica-tions.)
IV-35
Fragmentation of organizational respon-sibility within GSA contributed to the prob-lems in ADP policy development during the1970s. In the wake of GSA's expanded roleunder EQ 11717, the office of Federal Man-agement Policy (OFMP) was established toperform the ADP policy functions transferredfrom 0MB. This office was subsequentlyabolished in 1975 when EQ 11893 transferredmost of the functions assigned to GSA byEQ 11717 back to 0MB. Since ADP policyauthority was not transferred back to 0MBunder the second Executive Order, a smallADP Policy group, the ADP Management Policyand Planning function, remained intact; how-ever, confusion resulted because their re-sponsibilities were never clearly defined(see ref. 1). This group may be the nucleusof the current Office of Planning and Policy.
Two other ADP Policy groups have re-sided in ADTS: one within the ADTS Officeof Agency Assistance, Planning, and Policy,and a second located within the ProcurementDivision of the Office of Automated Manage-ment Services. The latter group has beencredited wih the development of "most of theADP policy now in effect" (see ref. 1).
Within OMB, ADP policy oversightresponsibility resided in the InformationSystems Policy Division until 1975, when PL93-400 (Office of Federal Procurement PolicyAct) created the Office of Federal Procure-ment Policy (OFPP) within 0MB, "to provideoverall direction of procurement policy pro-cedure, regulations and forms." The resultwas a lack of clear delineation of rolesbetween OFPP and the Information SystemPolicy Division regarding ADP policy de-velopment in 0MB.
The location of OFPP within 0MB, asopposed to GSA, has accentuated the juris-dictional disputes between these agencies.Lack of confidence in GSA has been noted asan overriding factor in this placement ofOFPP in 0MB (see ref. 3); however, theCommission on Government Procurement'srecommendation to create an Office of
iv-36
Federal Procurement Policy effectivelyexcluded GSA by stating that OFPP should notbe placed in an agency with a procurementfunction.
To further obscure the delineation ofFederal policy making responsibilities, theNational Telecommunication and InformationAgency was established in 1977. This of-fice, within the Department of Commerce,evolved from the Office of Telecommunica-tions Policy within the Executive Office ofthe President.
The coexistence of duplicative policyorganizations has created ambiguity as tothe roles and responsibilities of OMB andGSA. This has adversely impacted the func-tioning of these agencies and challenged theADP acquistion community in its efforts torespond to confusing direction.
2. There Are Seemingly Contradictory PoliciesBeing Implemented by OMB and GSA
The indistinct delineation of roles andresponsibilites is further evidenced in the con-flicting nature of the resulting policy direc-tion. OMB and GSA are mutually dependent incarrying out the Federal ADP Program, and yet theacquisition philosophy embodied in the guidanceissued by these agencies is conceptually para-doxical. The ADP community has encountered prob-lems in policy implementation due to an inabil-ity to reconcile the tenets of OMB A-109 withthose of PL 89-306, as implemented by GSA FMC74-5.
In response to the Commission on GovernmentProcurement's recommendations (see ref. 11), OMBissued Circular A-109 in 1976, establishingpolicies to be followed in the acquisition ofmajor systems. There have been roadblocks in theimplementation of A-109 in the present form, asthe requisite level of detail has proven prohib-itive for the majority of ADP acquisitions and"there appears little that applies specificallyto the ADP environment" (see ref. 1).
IV-37
It has been alleged that GSA is in opposi-tion to A-109, based on their failure to issueimplementing instructions. Until recent revi-sions of the FPMRs and FPRs, FMC 74-5 served asthe premier ADP policy document. This issuance,and its interpretation by GSA and the HGOC,favored the centralized procurement philosophiesof PL 89-306. GSA has conducted acquisitionswith hardware solution focus diametricallyopposed to the mission needs focus of A-109.
The coexistence of GSA procurement policywith A-109, which lacks implementation guidance,has caused anxiety in the ADP community, despitethe issuance of OFPP Draft Pamphlet #2, ADiscussion of the Application of A-109 to theAcquisition of ADP/Telecommunications Systems,"to assist Federal agencies in the understandingand application of OMB Circular A-109" (DraftOFPP Pamphlet #2). In 1978, GSA and the OFPPjointly prepared OFPP Pamphlet #2 expressing theintent that ADP acquisitions follow A-109, withthe coordination and approval of GSA and OMB. Inaddition to the problems of translating theactivities outlined in A-109 to apply to ADP,"there is little inclination to embrace a new setof rules promulgated by an acquiescent and here-tofore uninvolved OMB" (see ref. 1).
(3) Faced With Conflicting Philosphical Positions andContradictory Guidance, ASD(C) Has InadequatelyResponded to the Needs of the ADP AcquisitionCommunity
In response to PL 89-306 and OMB A-71, the Direc-torate of Data Automation (DDA) was established as thefocal point for ADP policy and management with theOffice of the Secretary of Detense (OSD). DDA is notspecifically assigned a role in ADP acquisition, andits present staff size would prohibit involvementother than on an exception basis. The Directorate haslittle formal influence over the military departments'computer acquisitions, as current directives do notinclude DDA in the approval process. To the extentthat DDA is involved, their policy issuances haveimplicitly reflected that they best serve as a facili-tator, rather than an instigator. Directives relatedto acquisition have acquiesced to the GSA hardwareproc-irement orientation legislated in PL 89-306,implemented by FMC 74-5, and interpreted by HGOC. Inrecognition of GSA and HGOC influence, there has been
IV-38
reluctance to issue implementing guidance on achievingLTOC in acquisitions. ASD(C) attempts to deal withthe chaos and conflicting philosophies in the FederalADP policy environment have highlighted two problemsfor DDA in administering the DOD ADP program:
Effective guidance on achieving LTOC has notbeen issued.
Lack of direct involvement in the function-ing of the ADP process has contributed tomarginally effective policy guidance.
The remainder of this section further defines thesecharacteristics of ASD(C) 's performance.
1. ASD(C) Has Failed to Issue Solid Directionto Achieve LTOC in Acquisition Despite theRepeated Instructions of the HAC
In 1978, the HAC instructed DOD to developand implement "lowest overall cost" criteria forits future procurements. In 1979, the committeereminded DOD of these instructions. DOD hasfailed to respond to these requests to the satis-faction of the Committee. While OSD officialshave expressed acceptance of LTOC concepts, guid-ance has lacked potency, and OSD appears to havemade a "conscious decision to defer availableconstructive action because of the application ofexisting policy" (see ref. 3). While DODDirective 7920.1 on Life Cycle Management ofAutomated Information Systems and DOD Instruction7920.2 on the Major Automated Information SystemsApproval Process address LTOC, their light treat-ment does not qualify as solid direction.
OSD's inaction is partially attributable tothe conflicting guidance of the HGOC and the HAC,and the overall confusion surrounding LTOC. OSDinformed the Surveys and Investigations Staff ofthe HAC that unilateral DOD guidance is useless,since GSA has ultimate procurement authority, andconsequently, any change in policy must come fromGSA.
IV-39
2. ASD(C) 's Minor Role in the Functioning ofthe ADP Acquisition Process Contributes toProblems in Policy Development andImplementation
The 1978 HAC Surveys and InvestigationsStaff Report (see ref. 1) concluded that as aresult of the DDA's minor role in the ADPacquisiton process, its contribution to assistingthe military departments has been marginal. Inthe absence of DOD guidance and direction inusing the available cost analysis tools, themilitary departments have been acquiring systemsunder their own guidelines, with varying degressof effectiveness. The charge of ineffectiveguidance in the area of LTOC can, therefore, betraced to insufficient involvement at thefunctional level, as well as to confusion in theFederal policy environment.
In summary, the problems found in acquiring ADP werecategorized into five major conclusions. The first con-clusion is that many problems can be corrected by focusingefforts internal to the Air Force. The second, third andfourth major conclusions then describe the problems withinthe Air Force in terms of misfocused process objectives,inappropriate process design, and a mismatch of roles,responsibilities, and skills. The final conclusion de-scribes the problems imposed on the Air Force by externalconstraints.
The next section presents the recommended strategiesfor correcting the observed deficiencies and accomplishingthe overall study objectives. In light of the first con-clusion, the major emphasis of the recommendations is onactions which are internal to the Air Force. The recom-mended strategies are presented in the same framework asthe study conclusions -- the four recommendations cor-responding to conclusions two through five respectively.
IV-40
V. RECOMMENDATIONS
V. RECOMMENDATIONS
Many of the problems identified in this study havebeen discussed in the preceding chapters. Thirty-threeseparate issues were highlighted in Chapter III and theirmajor causes concluded in Chapter IV. Correction of thedeficiences identified in the analysis of these issuesrequires an integrated set of recommendations addressingfundamental changes in the acquisition process as well asspecific changes in policy, procedures, roles, andresponsibilities.
The recommendations presented in this chapter providethis integrated approach to correct the basic deficienciesdelineated in Chapter IV. As such, they are more ap-propriately labeled strategies for changes rather thanspecific actions for improvement. These recommendedstrategies describe the changes which must be incorporatedinto the acquisition process and the ADP management struc-ture to achieve the objectives of this study. Specifica-tions for implementation are described in Chapter VI.
The recommendations of this study are categorized intofour broad areas as follows:
There must be a clear recognition that the pur-pose of the ADP program is to fulfill missionneeds, and thus responding to that mission mustbe the primary objective of the ADP acquisitionprocess.
In light of the mission primacy objective, theprocess must be modified to incorporate flexibil-ity to accommodate changing needs and adaptabil-ity to a changing environment.
Internal roles, responsibilities, and organiza-tional missions must be modified to accommodatechanges in the process.
External to the Air Force, implementing amission-oriented, flexible ADP acquisition pro-cess should be complimented by changes in policy
V-1
and modification of organizational roles andrelationships.
Specific recommendations in each of these areas are dis-cussed below.
1. THERE MUST BE A CLEAR RECOGNITION THAT THE PURPOSE OF
THE ADP PROGRAM IS TO FULFILL MISSION NEEDS, AND THUSRESPONDING TO THAT MISSION MUST BE THE PRIMARY OBJEC-TIVE OF THE ADP ACQUISITION PROCESS
Chapter IV concluded that the current acquisition pro-cess tends to place primary emphasis on economy and effi-ciency at the expense of fulfillment of operational mis-sion needs. The analysis revealed the lack of effectivepolicy direction and leadership, and the absence ofmission orientation to be root causes of many of theproblems in the ADP acquisition environment. The lack ofwell-defined objectives from which to develop policy andprocedures and define roles and responsibilities con-tributes to difficulties in fulfilling mission needs. Tocorrect this deficiency, the objectives of the processmust be clearly defined and communicated through policydirection which accentuates the primacy of satisfyingmission needs through acquisitions conducted in a mannerwhich addresses the need, at the lowest life cycle cost,using maximum practicable competition.
(1) While Mission Primacy Is Not Exclusive of OtherArcc-isition Objectives, Satisfaction of Opera-.onal Needs Cannot Be Sacrificed To Achievement
of These Other Objectives
Analysis revealed that the procedural focus onhardware acquisition as opposed to fulfillment of themission need is at the core of much of the damageincurred by the AF in ADP acquisition and management.Inflexibility, restrictiveness, and solution orien-tation create a lengthy and cumbersome process whichmay result in less than optimal decisions. Inordinatecosts and delays are incurred, and the mission needmay not be satisfied in the most efficient andeffective manner. To address this deficiency, it isrecommended- that the acquisition process objectives bereconsidered. In defining and communicating theseobjectives, the fundamental purpose of the ADP programmust be underscored. In light of the need to conduct
v-2
acquisitions consistent with that purpose, the processobjective should be to satisfy mission needs, withinthe required time, at the lowest life cycle cost.Completition should be used to the maximum extentpractical to satisfy this objective.
The recognition that satisfying the missionrequirements is the primary objective of the processis the keystone for all other recommendations of thestudy. This premise is the essential ingredient ofthe recommended changes in the process--changes thataddress the underlying causes of the pervasive prob-lems identified in the analysis of issues. This ismore than a philosophical change, for the implica-tions are concrete in terms of changes to the processand modifications of roles and responsibilities.While each specific recommendation will be addressedin detail in this chapter, a mission orientationimplies such pervasive changes as a life cycle andsystems management orientation, increased flex-ibility, and decentralization of responsibility andaccountability in defining and satisfying require-ments. In addition, to effect this shifting focus tomission primacy, a link must be formed between ehemission and its ADP requirements.
(2) Strategic Planning, as an Essential Element ofMission Primacy, Should Be Instituted
if satisfaction of mission needs is to be para-mount, then strategic planning becomes vital as themechanism for linking the recognition of need with theutility to fulfill that need. Since ADP represents autility to support the organizational mission, therequirement to satisfy mission needs cannot be separ-ated from the acquisition and management of ADP, andassignment of responsibility and authority must beconsistent with that premise. Strategic planning isthe link between the mission and the means to satisfyit, and is defined as accomplishing the following:
Translates mission goals into ADP goals andobjectives based upon changing technolo-gical, organizational, and politicalconsiderations
Institutes a formalized disciplined approachto identifying needs beyond the immediatefuture (i.e., 2-7 years)
V-3
Controls system growth to be compatible withmission growth
Provides the basis for long-range, program-specific planning.
Provides the basis for evaluating theeffectiveness of the AF ADP program
Formalizes communication of goals andobjectives to external agencies andthroughout the Air Force.
The next section addresses strategic planningmore specifically in terms of the proposed revisionsto the ADP acquisition process.
2. IN LIGHT OF THE MISSION PRIMACY OBJECTIVE, THE PROCESSMUST BE MODIFIED TO INCORPORATE FLEXIBILITY TO ACCOM-MODATE CHANGING NEEDS AND ADAPTABILITY TO A CHANGINGENVIRONMENT
Regarding the acquisition process design (i.e., thesteps which must be performed in order to acquire ADP),
the analysis identified three major deficiencies whichconstrained satisfactory performance -- process in-flexibility, the failure of the process to adapt to achanging environment, and the lack of systems and lifecycle man- agement perspective. Correcting these majordeficiencies will improve the process performance in terms Iof all three of the study objectives.
Making the process more flexible and incorporatingsystems and life cycle management approaches will reducethe acquisition lead time in two ways. First, varying thedegree of management and control for programs of differentcomplexity, as well as size, will lessen the review andapproval time for many programs and will concentrate theoversight on fewer individual programs. Second, adoptinga systems and life cycle perspective will reduce the totalnumber of acquisitions which must be processed, and extendthe useful life of the systems being acquired.
In terms of reducing technology lag, the increasedflexibility will encourage users to enter the processsooner and replace ineffective and obsolete equipment
V-4
earlier. By making the process more adaptable to a chang-ing environment, planners will take advantage of techno-logical opportunities, (i.e., pre-planned system improve-ments) rather than reacting to existing capability needs.
Improving the requirements generation and approvalprocess is an area where the process design recommenda-tions will have considerable impact. The major intent is
to reduce the time required and to improve the quality ofdecisions made during this phase of the process. Therecommendations that follow address these improvementsmore specifically.
(1) Fundamental Changes Are Required in the ProcessDesign
In order to correct the major deficiencies citedabove and to achieve the objective of mission primacy,certain fundamental changes must be adopted in theacquisition process design. Five major changes areproposed.
There must be a clear, strong and obviouslinkage between Air Force mission planningand ADP acquisition program identification.
The decision to approve a need should bemade as part of funds approval in the PPBS.
The decision to approve the ADP solution andthe acquisition strategy should be separatedfrom the need approval and delegated to alower decision authority.
The process design should be based on provensystems management practices:
- Program Management System
- Life Cycle Systems Management
- Decentralized execution of theacquisition.
Acquisition process regulations and proce-dures should be simplified to enhance under-standing and encourage flexibility.
V-5
-hese changes, which conform with the major decisionsannounced by Deputy Secretary of Defense, Frank C.Carlucci, in April 30, 1981, (see ref. 12) are thebasis for redefining the ADP acquisition process.This recommendation is presented in the followingsection.
(2) The Proposed Revision to the Process Demonstratesthe Effects of These Changes
Many acquisition process models and descriptionswere evaluated and compared with the process describedby the AFR 300 series. Among these are the SystemsAcquisition Process described by AFR B00 series, theDODD 5000.1 and DODI 5000.2 models as amended by theCarlucci decisions, and the A-109 approach. All aredescribed in terms of major decision points (mile-stones) which, if approved, authorize the acquisition
to proceed to the next major milestone.
Using this framework to describe an acquisitionprocess for ADP systems, it was determined that threemajor decisions are required in the process:
To initiate an ADP program based upon anapproved need
* To authorize implementation of an acquisition
* To accept the system or capability as satis-fying the original need.
Figure V-1 depicts the overview of the proposed ADPacquisition process. Three phases are identified,each of which results in one of these major deci-sions -- program milestones.
Phase I, Program Planning, defines the plan-ning activities which lead to the identifi-cation and approval of ADP programs and re-sult in Milestone I, Program Initiation.
Phase II, Alternatives Exploration, definesthe activities required to determine theappropriate solution and acquisition strat-egy of the approved program, and results inMilestone II, Implementation Authorization.
V-6
zz
00
z L
00
U.'.Fca QVx L c oCL C3 SCLolC3XC)c)CU.'LJCc:. L
C00La-
C.37
Phase III, Acquisition, includes the activi-ties required to obtain or procure theneeded capability or service. This phaseends in Milestone III, System Certification.
The following paragraphs describe each phase in moredetail. To clarify what is being recommended, eachwill be presented in terms of the typical activitiesand decisions by the key participants, the requireddocumentation, the phase objectives, and the resultingchanges and improvements to the process.
1. Phase I Is Described As Program Planning
Table V-1 presents the major activities ofthis phase in terms of the typical activities anddecisions by the key participants. It also pre-sents the required documentation corresponding toeach of the activities and decisions.
The overall objective of this phase is toconduct the requisite planning and related activ-ities to define and initiate an ADP program. Thespecific phase objectives are:
To define ADP goals and objectives forthe Air Force and for the operatinglevels within the Air Force
To identify plans and programs to ac-complish these ADP goals and objectives
To plan and allocate resources neededto support specific ADP programs
To initiate programs to satisfy ap-proved needs.
Phase I corresponds to the pre-DAR activi-ties of the current ADP acquisition process. Itparallels the budget process which commences withthe release of the Air Force ADP Plan in Marchthrough the submission of MCAPs in October. Themajor changes proposed in this phase are designedto provide a more mission-oriented focus to theprocess, to appoint a program manager earlier inthe process in order to influence key decisions,and to provide the Single Manager with greaterflexibility to initiate ADP programs. Thefollowing specific changes are recommended:
V-8
P ADA113 755 BOOZ-ALLEN AND HAMILTON INC BETHESDA NO F/6 15/SDEFENS ADP AC*JISITON STUDY.(U) ANOV 81 MDA
9 03-
8 0-Cm
049
UNCLASSIFIED
3.'IuuIuuhhffIhu
EEEIIIEEIIEIIEEIIIIEEEEEI
a .5.
-0 '-,'c- m !2L
-~~~ ~ ~ ~ B;I-- 9McU3 I- -
0!0 <W I. wM4UCa 4c . a aa a io.
Wa wA Co cc UI- U
10 -XWia~ 0~ 4 . -~~IA a-N4 U)4v z~;u~.4
~~~S4ww oll S. LU: =,z
~~~L 0. COW~ aU C-~ ~ >-' 02~~
W. M~ 9L ai .L A
>F X>UU 0 CC ,Z c e c cC7 LDII .C...L wC. C~. C
ca
z U
UAU
SM - C zL3a CA
4 UU
4 ' .2 m S M W >
a-- CD o Uw rnC.-M CD C4 w4CU~ M
Icc CA 4t 0at .. ~ z-~ z CmCC3 r C Ira a- ) 0 a- S CC- I0 5.j C
CD2 L40
C2
~j ).L) SM ~= ~42 ~ I~C l-~~ -
....o C. >Da U)U
;, C4 U 2U C S
W +Ia j .- 0 III a- C 41
a a a aaa a aaa4.
CZ a !ww wMam cw c . -
x 2 __ __r
CDt
CD m -
coa.a
z co-z
ZK C- _icoLUL1C
Greater emphasis on top-down ADP stra-tegic planning, particularly in formu-lating strategic ADP goals and objec-tives at the operating levels
The identification of needs in missionterms
The approval of those needs as part offunds approval in the PPBS
The initiation of an ADP program on thebasis of an approved mission need priorto the selection of a solution and ac-quisition strategy.
These changes are presented in the Phase Idescription contained in Table V-1; however, therecommended change concerning strategic planningwarrants further discussion.
The strategic planning recommendation re-quires changes at two organizational levels.First, the Air Staff must prepare ADP goals andobjectives which are realistic and implement-able. For example, the goal -- "to replace allobsolete equipment in the Air Force, within thenext five years, where it is cost effective to doso" -- is one that can be achieved throughout theAir Force. This structured evidence of planningprovides the basis for rational oversight manage-ment by presenting a clear message to externalagencies concerning the intended direction of theAir Force ADP program. Such clearly defined andrealistic goals provide the rationale for sup-porting or refuting proposed changes in Federalpolicies and ADP standards.
Secondly, at the MAJCOM and SOA level, theremust be more active involvement in the strategicplanning process. Implementation of this recom-mended change calls for the Single Managers toprepare operating level ADP strategic plans,which must be approved by the respective CINCsand Commanders. Effective strategic planningforces earlier identification of need and a bet-ter awareness of technological opportunities tosupport those needs. The long-term effect is to
V-10
reduce the number of unplanned requirements andimprove the quality of ADP support in the AirForce, in addition to presenting an image of com-petent resource management to external oversightagencies.
2. Phase II Focuses on Alternatives Exploration
This phase performs the required studies andanalyses necessary to determine the preferred ADPsolution and the acquisition strategy. The ac-tivities, described in Table V-2, partially par-allel the activities surrounding the preparationand approval of the DAR in the current ADP acqui-sition process. The changes and resulting bene-fits in this phase are described below:
A major change is that an ADP programhas been initiated and a program man-ager assigned prior to the evaluationof alternative solutions and strate-gies. The resultant benefit is thatthe program manager, who must eventu-ally implement the technical solutionand strategy decisions, can influencetheir outcome, thus establishing agreater degree of accountability in thedecision making process.
A second change, initiating a programprior to conducting the studies andanalyses, also adds needed flexibilityto the process. The directive docu-ment, either the program charter or thePMD, can tailor the requirement forstudies to the complexity and the sizeof the program, thus eliminatingredundant and unnecessary documentation.
The separation of the decisions made inthis phase from the need approvaldecisions made in Phase I is a thirdsignificant change. It allows thelater decisions (technical solution andacquisition strategy) to be delegatedto lower levels of authority, thusreducing the required approval time.
V-11
LU
aa
aE I- z C
mU i I C, r.
=5 .,Ugo C I-Luj -U . V Lau)Z.
.. m 41 . ,ot ~ ~ ~ ~ I. U, 0 -9 9
C')CC
0 LILWCDu
c U,CC- 0 )I 4WL
4U
Z- -CC
=> U .u
I-. CC l0 IIL ) co C' U= aL US V;
C2 -- UsE
'U P 9 Z .maOK-C g zO
a2cj an cw "
'm Ca en
101
'U Z0,itL
0 CC0 UAt: U CC I-'SILu
CC-j UCIL I l W M
mn CDC m 04
41-
V-CZ
3. Phase III Concludes the Acquisition
once implementation has been authorized, theobjective of this third phase is to take actionon the chosen technical solution in accordancewith the approved acquisition approach. TableV-3 describes the activities of this phase.
Depending upon the strategy selected, theacquisition phase can be particularly lengthy,especially for competitive approaches. Many ofthe steps which contribute to this time periodare dictated by Defense Acquisition Regulationsand cannot be significantly shortened. otherprocedures, such as benchmarking, are not dic-tated by the DAR, but their rigorous applicationalso contributes to the lead time. The intent ofthe recommendations for this phase of the processis therefore to:
* Inject greater flexibility in the useof acquisition strategies
* Promote rational application of theprocedures for a given acquisitionstrategy.
The most significant result is a carry-overfrom the changes in the previous two phases.That is, a single program manager who maintainsmanagerial, technical, and financial responsibil-ity for the ADP program is held accountable forthe acquisition end-to-end, and at completionmust certify the performance of the system to theuser. It is through the program manager's con-tinued involvement in the process that the acqui-sition cycle can be shortened in areas such aswriting appropriate specifications and solicita-tion documentation, selecting the appropriateacquisition strategy, and tailoring procedures toobtain a balance of cost versus risk.
Many of the changes proposed in the acquisitionprocess design require a reexamination of organizationalroles and missions at all levels within the Air Force.This reexamination is the subject of the next majorrecommendation.
V-13
I-i
E , ~
Wr Q 0
,- adz1-4~
a-p -C.-9V
"- Z- Uc m,. L, -5,
mcI C.C. a z -
C,, aUCC.
wxC.,a
CA -I-Wa
0J
wAJ WLICo >~~
W U 0 iiP ~ CA W a
4a >- L3~~JU
Z (4 0 a
t I. U i
-wc w~3W o CS
K ;E .- C IU I.
aL a ao0Z! O a a ao0 rnC
W 3ac - 3 >toF
I-, c
zj - zeL" .L
V-1
L" Is 0 1.1
3. INTERNAL ROLES, RESPONSIBILITIES, AND ORGANIZATIONALMISSIONS MUST BE MODIFIED TO ACCOMMODATE CHANGES INTHE PROCESS
Concerning the people who must function within theprocess, the analysis identified four major deficiencieswhich constrain the effective acquisition of ADP -- themismatch of roles and skills, the inconsistencies in themanagement and organization for ADP, the lack of end-to-end accountability, and the lack of effective leadershipand guidance. These deficiencies, in conjunction with therecommended changes to this process, affect the roles,responsibilities, and organizational missions at alllevels within the Air Force, and necessitate changes inthree areas:
* The role of the MAJCOM/SQA ADP single manager andassociated organizations
* The role of the Deputy Commander for Data Automa-tion (AFCC) and subordinate AFCC centers
* Roles and responsibilities for Headquarters USAForganizations.
Before discussing each of the above, two general com-ments must be made. The first concerns organizationalchange. A major consideration for this recommendation isto minimize the organizational impact. The changes pro-posed in this section are, for the most part, changes inroles and responsibilities rather than organizationalrealignments. The majority of changes recommended to thispoint can be instituted within the current organizationalframework. However, certain organizational realignments,particularly at the Air Staff, are inevitable in the longrun. Legislative changes, such as the passage of PublicLaw 96-511, the language in the proposed 1982 DOD Autho-rization Bill on "exempt vs non-exempt ADP,l technologicaltrends such as the merging of telecommunications and ADP,and other political trends are the catalysts for thecurrent rethinking of the organizational structure at thehighest levels within the Air Force.
The second general comment concerns the issues oftraining and the retention of qualified personnel. Bothare problems facing the military in many technologicalfields,' and ADP acquisition is no exception. Thelimitations of skilled, qualified personnel account formany of the delays in acquiring ADP. While simplifying
V- 15
the process and making more effective use of these lim-ited human resources will improve the situation to someextent, there can be no substitute for a firm commitmentto training in the areas of planning, acquisition, andmanagement of ADP resources.
(1) Adopting the Program Management System Requiresan Expansion of Roles at the MAJCOM/SOA Level
Fundamental changes in the process design includethe adoption of a program management system and a con-trolled decentralization of the execution of the ADPacquisition process. The program management systementails appointing a program manager earlier, increas-ing his authority, and making him accountable for theprogram throughout the acquisition life cycle. Toaccomplish this, the role of the MAJCOM and SQA ADPSingle Manager must be strengthened, and additionalsupport skills are required.
1. Strengthen the Role of the Single Managers
Strengthen role of the single managers inthree areas:
Increase the approval thresholds of thesingle managers
Give single managers greater authorityto reallocate funds
Expand their authority to initiateprograms.
The approval thresholds for the MAJCOM/SOASingle Managers and the Standard ADPS Managerswere recently increased to include the approvalauthority previously held by USAF/ACD. This nowgives single managers approval authority for pur-chases up to $5 million competitive and $500,000sole source (reference Table 11-1). Initiativesare underway to give the single managers moreauthority to reallocate funds.
Table V-4 describes the proposed maximumthresholds for program initiation. If imple-mented, these thresholds will further enhance thesingle manager's authority. Three changes arenoted in establishing the threshold criteria.
V-16
0
H 0 0
w w 0 0
00a4
W U) I)
HO
E-U)
>
0 pq a4r-, 4w a4 >4, 0z 04 < U)
"I U E-4 U -
U) z o CAm
H0 0 14P 0
U) U-
H0
00
a44 i
U0 0 0 i
E-4 nA Dni
CQL
H0)
v-ii7
I, *n -n
* First, threshold levels are based onsystem life cycle cost rather thanhardware cost to reflect a change inemphasis in the process.
* Second, since program approval is tiedto funds approval in the PPBS process,the thresholds change depending onwhether the requirement is processedwithin the PPBS cycle and those out ofcycle.
* Third, since program approval and ini-tiation occur prior to solution ap-proval and acquisition strategy formu-lation, competitive versus non-competiive acquisitions are not a cri-teria in the threshold determination.These and other decision and reviewlevels are determined by the level ofprogram milestone reporting specifiedas part of program direction.
These thresholds are intended as maximum guide-lines for program approval. They are below theDODI 7920.1 thresholds* for major systems, andare not intended to be applied equally to allMAJCOMs and SOAs. Thresholds for individualMAJCOMs/SOAs should vary. Considerationsshould include such items as the level of ADPsophistication and use, the size and mission ofthe organization, past performance in conduct-ing acquisitions, and the effectiveness ofstrategic planning. The actual secretariatapproval thresholds should be established at apoint that provides appropriate visibility toeach of the major commands and encompassesapproximately 5 percent of the total Air ForceADP Program.
Accompanying the increased single manager'sapproval threshold should be the increased flexi-bility to reallocate funds and initiate pro-grams. This, too, is an area where progress has
* DODI 7920.1 defines major systems as those greaterthan $100 million program acquisition life cycle orgreater than $25 million in any single year.
V-18
been made. Single managers have the authority toreallocate funds within Program Elements andcertain Element of Expense Investment Codes(EEIC's). This flexibility is paramount for thesingle manager to be able to reallocate funds toinitiate programs for higher priority needs with-in the budget and program years.
2. Expand Skills and Resources
To fully execute an acquisition program us-ing program management system procedures, addi-tional support is required at the MAJCOMs/SOAs.Skill levels in certain disciplines must be en-hanced and made available to the program man-ager. These areas were described in Chapter IVas those disciplines required for certain AFR 300series acquisitions, but which are not explicitlyprovided. They include:
Program management and planningSystems engineeringLogistics support
* Configuration management• Computer resource integration
Operational planningTest and evaluation.
For those MAJCOMs and SOAs with recurring acqui-sition programs requiring these types of disci-plines, personnel with these skills should beunder the direction of the single manager. Forothers, these skills should be made availablethrough AFCC centers or other centralized sources.
Similarly, procurement support should bemade available to the single managers. Contract-ing support capable of conducting competitive ADPacquisitions should be established at bases wherethe need is warranted. Where the volume is notlarge enough, procurement support should be madeavailable through AFCAC or other regional ADPprocurement offices to be established.
(2) The Deputy Commander for Data Automation (AFCC)Should Have Primary Responsibility for USAFStandard ADPS's, and Support Responsibilities tothe MAJCOM/SOA Single Managers
AFCC is the center of expertise for the acquisi-tion, management and operations of ADP resources.
V-19
Within the AFCC centers, under the direction of theDeputy Commuander for Data Automation, are areas ofexpertise which are limited elsewhere in the AirForce. Making effective use of this expertise iscritical to the success of the Air Force ADP Program.
The primary role for the AFCC centers, with theexception of AFCAC, is to support the USAF StandardSystems. This role should remain unchanged. However,to make more effective use of these talents, theyshould be made available to support the MAJCOM/SOAnon-standard systems on an as-needed basis. Such sup-port should be made available at the request of a pro-gram manager and should be in the form of a programmanagement review team or an acquisition consultantteam to assist with such tasks as:
* Conducting required studies and analyses,including requirements analyses, feasibilitystudies, and software conversion studies
* Formulating acquisition strategy
* Developing specifications and relatedtechnical documentation
* Preparing solicitation documentation
* Assisting in source selection
* Supporting in-house software developmentactivities
* Supporting live test demonstrations (LTDs)and other testing activities.
Concerning the AFCC ADP roles, an area deservingspecial consideration is the support provided by theAir Force Computer Acquisition Center (AFCAC) . AFCACserves as the centralized ADP selection office andprovides other program management support functionsduring the acquisition process. To make best use ofthis limited expertise, AFCAC's role should be reducedto specialized support functions of the acquisitionprocess in the following areas:
Provide contracting and source selectionsupport to USAF standard systems and toMAJCOM/SOA systems above a certain threshold(e.g., $25M) or at their request
V-20
Assume primary responsibility for technologyforecasting and ADP standards evaluation
Provide acquisition assistance to programmanagers at their request.
These modifications in AFCAC's role will make mosteffective use of their expertise.
(3) Adoption of the Preceding Recommendations Re-quires a Reallocation of Roles and Responsibili-ties on the Air Staff
A major conclusion of this study is that the AirForce ADP management structure and organization con-tributes to the confusion and thus the delays in ac-quiring ADP. Of particular concern is the apparentuniqueness in the management and acquisition of gen-eral purpose ADP compared to other systems and re-sources. There was near universal criticism that thisunique management stucture for ADP is responsible formuch of the confusion and results in little value ad-ded to the acquisition process. Opinions concerninghow to improve this situation are not as universal.Suggestions range from establishing a separate DCS forinformation systems to shifting certain areas of re-sponsibility within existing organizations.
Concerning ADP management, four areas of respon-sibility were described in Chapter IV as being anomal-ous to the rest of the Air Staff:
ADP Contracting & AcquisitionADP RequirementsADP OperationsADP Support.
Each area was examined in terms of three broad roleswhich reflect the types of functions performed by mostorganizations on the Air Staff:
Policy Formulation and PromulgationResource ManagementControl and Oversight.
The first step in the analysis of roles and responsi-bilities was to identify and define the ADP functionsin each area and determine where those functions are
V-21
• -- - -,., .... ,-. A
duplicated in other Air Staff organizations. The re-sult of this first step is presented in Chapter IV aspart of the study conclusions. The second step was todetermine where these functions are best performed ont.he Air Staff. To make this determination, the fol-lowing criteria were established:
Are the functions consistent with otherfunctions performed by that organization?
Does the organization possess the skills andresources necessary to perform thosefunctions?
Can the organization adapt to changingtechnological and political environments?
Is the impact on current organizationsminimized?
An additional technological issue was also con-sidered. Should responsibilities for ADP, telecom-munications and other information systems technologiesbe consolidated, or should they remain separate? Ig-noring the political ramifications of such a decisionfor the moment, there are compelling reasons why thesetechnologies can no longer be considered separately.
From a technological viewpoint, the technol-ogies are merging to the point that there islittle distinction among them. These dif-ferences are made more in terms of theirapplication than their capabilities. Fur-ther, interface points are becoming increas-ingly difficult to establish.
From a management viewpoint, the trends to-ward decentralization and mission orienta-tion support their consolidation.
- The decentralized execution of the ac-quisition process requires a less spe-cialized technological focus in a cen-tralized management organization (i.e.,the Air Staff).
- A mission orientation and a systemsapproach to the acquisition process
V-22
------- --------
requires that consideration be given toall the information technologies aspotential solutions. Early considera-tion of one aspect of technology overanother constrains the solution.
Recognition of the technological merger of ADPand other information technologies is widespread inthe technical and business literature, and has beenreaffirmed in the last decade by many studies and re-ports. In notable instances, study teams have takenthe additional step of recommending the functionalmerger of the information technologies for managerialand/or policy control, and policy and legislation havebeen supportive. The 1972 establishment of ADTS with-in GSA recognized the growing interface between ADPand telecommunications by consolidating direction andcoordination functions for ADP and communications. In1977, the Commission on Federal Paperwork (see ref.13) recommended that the President consolidate thedispersed and fragmented major paperwork, informationand communications-related policy oversight functionsand authority under central management direction andleadership. In that same year, the Privacy and Pro-tection Study Commission (see ref. 14) reported thatdata communication systems had become inextricablyinvolved with data processing systems. The 1978 Re-port of the Federal Data Processing ReorganizationProject of the President's Reorganization Project(PRP) brought to light numerous deficiencies inFederal ADP management and made a wealth of recommen-dations to alleviate the perceived problems. TheCentral Agencies Team of the PRP (see ref. 15) notedthe inattention to the confluence of computers andcommunications as a major deficiency and the GeneralGovernment Team Report (see ref. 16) found inadequatelinkage between information processing and telecom-munications in terms of long-range planning and sys-tems design and development. The National SecurityTeam (see ref. 17) recommended that informationtechnology, to include all computer/telecommunica-tions activity, be elevated to a recognized Air Stafflevel function, commensurate with its importance as aresource.
The December 1980 Evaluation of the GSA by aPanel of the National Academy of Public Administration(see ref. 3) faulted GSA ADTS for managing ADP and
V-23
telecommunications separately, despite thleir techno-logical merger. GSA, however, recently reorganized,creating an office of Planning and Policy to manageboth ADP and telecommunications. The Paperwork Reduc-tion Act, PL 96-511, consolidates authority for policyrelated to management of all information resources,including telecommunications, in 0MB. Additionally,the proposed Federal Procurement System (FPS) callsfor a comprehensive, uniform, integrated Government-wide system for procurement of goods and services toreduce fragmentation of policies, regulations, andprocedures.
As a result of this evaluation the followingrecommendations are made concerning the realignment ofroles and responsibilities for the management of ADPon the Air Staff.
1. Future Consideration Should Be Given to Con-solidating the Responsibilities for the Man-agement of Telecommunications, ADP, and Com-puter Ba sed Office Systems
For the reasons cited previously, it is im-portant to consider any change in roles and mis-sions in light of the inevitable merger of theinformation technologies. Since the focus ofthis study was limited to ADP resources, the pro-cedures for managing telecommunications (cur-rently under AF/XOK) and for office systems(currently under AF/DA) were not examined to thesame level of detail. Thus it is difficult todetermine the political impact of such an orga-nizational realignment. However, it is recom-mended that these three areas of informationtechnology be managed within the same structure,and the same body of regulations, consistent withthose proposed for the management and acquisitionof ADP in the following section.
2. Responsibility for the Management of ADPShould Be Consistent With the Organizationof the Air Staff
organizational consistency is a driving con-sideration in the assignment of roles at the AirStaff. The major changes proposed for the pro-cess design are not new concepts, but rather ones
V-24
which are currently being used successfully inother Air Force acquisition processes. Policiesand procedures currently exist to implement con-cepts such as the Program Management System,mission-oriented needs statements, life cycle andsystems management, and a decentralized executionof the acquisition process. These policies andprocedures should not be redeveloped, but ratheradopted as the procedures which govern the ADPacquisition process.
in addition to consistency, and in keepingwith the goals for the process recommendations,the following goals were considered in formulat-ing organizational recommendations:
* Maintain a simplified but disciplinedstructure for the management of exemptsystems.
* maximize the commonality in proceduresfor exempt and non-exempt systems.
* Capitalize on existing procedures.
* Encourage organizational decision-making which fosters technological fu-sion of ADP, telecommunications andoffice systems.
* Make most effective use of people andskills.
Table V-5 describes the proposed realloca-tion of ADP roles and responsibilities. Thetable identifies those responsibilities for whichAF/ACD currently has responsibility. Since thereare no changes proposed in the role of the func-tional OPR on the Air Staff, these organizationshave been omitted from the table. Recommenda-tions concerning the other ogranizations involvedin the management and acquisition of ADP are asfollows:
Contracting and acquisition responsi-bilities should be assigned to AF/RDC.
Responsibilities concerning the genera-tion, approval and management of re-quirements should be distributed withinAF/RD.
V-25
0 0 P4 0HH ~ I C0 H
H ~ > 0~~0H- 2:~ m>4 H z) E-4 L4 >1
0l N Z 2H z (D -I :-1 a4 w EWF:W< L 0 Q F
0 a. () Z Ix .- u I- UC) U 4
Cl) M ' 4 '4' ** * 4
0 p U-C) Cl
E-C Cl) p~ ~1:4~ - Cl)<w
u U) 0 >4 w ) , 4 m
0 " E-:F -4P >40 q>4 E, U) D E) 4-4UZ
H < 1 0 Ix04 -4 &AW <& ZE
H < wuzu 42: w 0 w .m f) w C < a4-4 12 < > xWU D 0U )> D> ,E
co 0 4' 4 4 ' '4
Cl)
E-0 0 0
Cl) U) tJ< E ,~ W * -u~~ ~z Cl) >4 2:2 u 7- 0 Z -q i - - >4H E-
P Cu) (n J w 0 w> P OLrl OCNx Q UC)F 0 w!D u
x 0 <2 >~H <~~~E~ 2: 4U ULr FEQ w~1 z a 0 wWQ0,E Zl Zu~C)E
> 01 WCJ u < 0 r) U c- 0< 124 z 4 > 0C4 U 0
0- 4-4 El m0 E - ) C L0 E
cr wf -K) * *
2: z ilFU4
0- C) 4- -)N)
r-4 z 00W-
o0 E. l)l -. 000W -
E- <' W E- Eww<E
z) u )w 47 lU) - , 0 W n o <&AW:
00
< E- 0
D, E cv ~s"U)~ E-4
0 :D 0- 00 00 4:A
a >1 . m-w6
* operational responsibilities for ADPsystems should reside within AF/XO.
* Certain ADP support responsibilitiesshould be assigned to AF/LE.
In terms of implementation, the ordering ofthe above recommendations is critical. Wherethere are compelling reasons for consolidatingADP contracting and acquisition responsibilitiesin AF/RDC, there is a less urgent need for theother proposed functional realignments; however,long range plans should consider these recommen-dations as the preferred alternatives. The fol-lowing paragraphs summarize the rationale sup-porting these proposed alternatives.
(1) ADP Contracting and Acquisition
There are compelling reasons for AF/RDCto assume contracting and acquisition re-Isponsibilities for ADP (i.e., promulgatingpolicy and guidance in the areas of sourceselection, contracting, and contract admin-istration). Plans are currently in progressto transition this responsibility. Thereshould be one, single source for contractingand acquisition policy and guidance -- TheDefense Acquisition Regulations. Attemptsto separate the special Federal require-ments imposed on ADP have had an adverseeffect on the process.
(2) ADP Requirements
There are three reasons why AF/RD isconsidered the best alternative for assumingthe responsibility for policy, direction andcontrol of ADP requirements (i.e., managingthe process for generating and approving ADPrequirements, issuing program direction, andsatisfying the requirements through systemengineering and acquisition procedures).included in this area of responsibility isthe interface with GSA required for non-exempt systems:
V-27
First, AF/RD currently is the OPRfor policy governing requirementsdefinition, approval and manage-ment. Many of the procedures be-ing proposed in the revised pro-cess, such as the program manage-ment, systems engineering, andtest and evaluation procedures,fall with the scope of control ofAF/RD.
* Second, the career force managerfor many of the disciplines re-quired in the revised process isAF/RD.
* Third, AF/RD is considered thetechnology manager responsible fortechnological planning and devel-opment. While these activitieshave been restricted to researchand development, advanced develop-ment, and production systems, RDis in the best position to assumea similar role for commercial ADP.
Alternatively, consideration was givento both AF/XO and AF/ACD for this area ofresponsibility. in both cases it would meanmaintaining separate structures for exemptand non-exempt systems, as well as separateprocedures. (At this writing, the responsi-bility for management of exempt systems hasnot been established. There are indicationsthat this responsibility will be establishedunder SAF/AL and that AF/RD is a likely can-didate to assume that role.)
(3) ADP Operations
Within this area of responsibility,there is little proposed change to existingroles and responsibilities. The two signi-ficant areas of change are:
AF/XO should be responsible forestablishing operational policyconcerning the employment of ADPsystems in the Air Force.
V-28
USAF/XO should act as the func-tional OPR for multi-user systems.
These are both areas in which they are nowresponsible for telecommunications and inwhich they are in a position to assume re-sponsibility for ADP with very little impact.
(4) ADP Support
The final area of responsibility con-cerns the items of supply and inventory man-agement responsibilities for ADP. Many ar-gue that AF/LE, as personal property man-agers for the Air Force, already possessthis responsibility for ADPE management.The specific recommendations in these areasare as follows:
Responsibility for the reutiliza-tion and sharing programs, cur-rently managed by AF/ACD shouldtransition to AF/LE.
Configuration management and lifecycle costing procedures, cur-rently AF/LE responsibilities,
should be consistently applied toADP.
Tables of allowance should be es-tablished for certain classes ofADPE and should be an alternativeto the requirements generation andapproval process. (Considerationis currently being given to thisapproach in USAF/LEY.)
The recommendations discussed to this point addressproblem areas within the Air Force's ability to change.While these areas are of major concern in this study, theycannot be addressed without considering the impact of theexternal environment. The next section addresses therecommendations concerning organizations external to theAir Force, and their relationships to the Air Force.
V-29
4. EXTERNAL TO THE AIR FORCE, PREVIOUS RECOMMENDATIONSSHOULD BE COMPLEMENTED BY CHANGES IN POLICY AND MODI-FICATION OF ORGANIZATIONAL ROLES AND RELATIONSHIPS
The recommendations to this point have focused onactions internal to the Air Force. This is consistentwith the study conclusion that although there are exter-nally imposed constraints, many of the obstacles to anefficient ADP acquisition process can be overcome withinthe scope of AF authority. However, achieving maximumbenefit from the process improvements requires some atten-tion to the external organizations which influence theprocess -- principally OSD.
The time is right for promotion of positive change inthe ADP acquisition process. Recent initiatives indicatethat the Federal regulatory environment is receptive toacquisition reform. The Deputy Secretary of Defense,Mr. Frank Carlucci, announced on April 30, 1981, (see ref.12) that DOD has initiated major changes both in acquisi-tion philosophy and in the acquisiti-n process itself.Executive agency reform proposals such as the Federal Pro-curement System (FPS), recent revisions in FPRs and FPMRs,and organizational changes at OMB and GSA reflect thewillingness to address the persistent problems in Federalacquisition and information resources management. Con-gressional awareness of the need to act is indicated bythe passage of the Paperwork Reduction Act, PL 96-511, andthe Senate's inclusion of legislative reform in the 1982DOD Appropriations Bill.
opportunities exist to correct many of the problemsthat have constrained the acquisition process for years.The preceding recommendations of this chapter describeopportunities to implement changes within the Air Force.To complement these changes, the Air Force should forwardfor review and adoption at OSD a set of recommendationswhich are conceptually similar to those proposed in thischapter.
It is felt that the Air Force can most successfullypromote change external to the Air Force to the extentthat it can influence OSD to enact reforms and representDOD interests at the Federal level. Toward this end, theAir Force should concentrate on encouraging OSD to take astronger, more active and visible role in Federal ADP mat-ters. Consistent with the strategy recommended in thischapter, the Air Force should encourage OSD to:
V- 30
Adopt consistent realignment of roles andresponsibilities within DOD similar to therealignments proposed for the Air Force.
Simplify DOD directives and instructions andensure consistency with Federal ADP policy andacquisition regulations.
Regain credibility with Congress through improvedrelations and demonstration of competent ADPresource management.
Assume a more dynamic role in relations with Con-gress, OMB, and GSA to assert DOD influence onFederal ADP policy formulation, particularlyrespecting the implementation of PL 96-511 andthe FPS.
These initiatives, coupled with the Air Force's ownefforts to improve the acquisition and management of ADP,will have the greatest impact on reducing the constraintsimposed by the external environment.
The above recommendations were designed such that,taken in aggregate, they represent the full set of inte-grated strategies needed to correct the deficiencies inthe ADP acquisition process and to achieve the studyobjectives. The specific implementing actions are delin-eated in the next chapter. As stated earlier, the orderof the recommendations is important -- the first being themost critical in terms of acceptance and implementation.The next chapter further defines this ordering through thesequence of actions needed to implement these recommendedstrategies.
V-31
VI. IMPLEMENTATION APPROACH
VI. IMPLEMENTATION APPROACH
The implementation approach is based on taking thewhole set of recommendations and implementing them in anorder that recognizes that the ADP process is an ongoingeffort, with many activities keyed to the PPBS. The totaleffect of the recommendations will be felt after the lastimplementation step has been completed. At this point itis realistic to anticipate that all recommendations couldbe implemented in time to affect the FY85 budget call. Onthe other hand, since the acquisition process is ongoing,various recommendations may be expected to impact theprocess as soon as they are implemented. Implementationof recommendations that eliminate whole steps in the pro-cess obviously shortens the acquisition process; raisingthe dollar approval threshold keeps more programs out ofthe HQ USAF staffing cycle. The benefits of planning andtraining are more long term.
Implementation of organizational changes is recognizedas a sensitive area. As mentioned in the preceding chap-ter, it is possible to implement the process changes with-out changing the existing organ'ization for ADP acquisi-tion. However, many factors previously discussed appearto make organizational change inevitable. Only slightlyless sensitive than organizational changes per se arechanges in roles and responsibilities. The implementationof such changes is, in some cases, closely linked to or-ganizational changes; however, in most cases it does notaffect the overall organizational mission.
The use of the Air Force Acquisition Improvement Group(AIG) to implement these recommendations is key, if notcritical. They have the advantage of the knowledge andthe insight gained from being able to step back and viewthe process with a critical yet constructive eye. Throughtheir close interaction with the multitude of participantsin the process, they have a firsthand knowledge of themany problems in processing ADP acquisitions.
This chapter is not an implementation plan. It is arecommended implementation approach; detailed implemen-tation plans would be required before proceeding with thevarious steps. Implementation will be addressed here interms of:
VI-l
. The ADP acquisition process• The realignment of roles and responsibilities• The activities of the AIG.
The implementation steps that follow are presented intheir recommended sequential order.
1. IMPLEMENTATION OF THE REVISED PROCESS REQUIRES ASERIES OF STEPS THAT RECOGNIZE THAT THE ADP ACQUISI-TION PROCESS IS ONGOING
The following are the steps required to implement theprocess recommendations proposed in Chapter V.
(1) The First Step Should Be To Brief ExternalAgencies
The Air Force ADP acquisition process interfaceswith a number of organizations at both the policy andoperating levels. It is important that the Air Forceacquaint these organizations of the impending changesfor two reasons:
At the policy level it is important thatASD(C) and OMB know what changes are anti-cipated, that the changes are being madewithin the framework of existing Federalpolicy, and that they are aimed at improvingthe Air Force ADP acquisition process. Thissame briefing could be used as an opportu-nity to float some "trial balloons" onpolicy changes that the Air Force would liketo see occur to further enhance the ADPacquisi- tion process. During the course ofthis study, the Air Force AcquisitionImprovement Group has briefed most of theFederal organ- izations involved in ADPacquisition on the purpose and status of thestudy and, without exception, all organiza-tions have expressed an interest in hearingthe results.
At the operational level, it is importantthat organizations like GSA, which haveregular interface with the Air Force onAPR/DPA matters, be made aware of proposedAir Force changes.
These external briefings could afford the Air Force anexcellent opportunity to foster the concept of "earnedautonomy." Even though a significant number of ADPacquisitions will be exempt from GSA and to some
VI-2
extent HGOC review and influence, there still remainthose non-exempt acquisitions that will requireexternal review.
(2) Program Managers Should Be Appointed For AllPrograms That Have Survived the PPBS Process
The early appointment of program managers isessential in establishing program accountability. Assuch, program managers should be appointed for pro-grams that have survived the PPBS as soon as one canbe identified. Three specific steps should be taken:
Provide interim guidance to the single mana-gers on their enhanced role and authority toappoint program managers.
• Strengthen the program manager's role by:
- Issuing HQ USAF/ACD policy statement tothe field
- Revising DPDs where appropriate
- Requiring MAJCOM/SOA single managers todevelop program manager charters
- Reemphasizing AFCAC'S role to supportthe program manager.
Institutionalize single managers proceduresfor reducing documentation for ADP programsthat are within single manager approvalauthority. Many single managers had pre-viously instituted "mini-DAR" procedureswithin their organizations for small pro-grams. These mini-DAR procedures should bereexamined by the single managers in thelight of all the pending changes to theacquisition process, and, with the appro-priate changes, should be institutionalized.
The implementation of the above changes will have themost immediate impact on improving and acceleratingthe ADP acquisition process as it can immediatelyaffect those proposed acquisitions that are already inprocess.
VT-3
(3) HQ USAF Should Now Start Developing PlanningGuidance for the FY85 Budget Call
FY85 is the earliest budget year that can betotally affected by the new procedures, particularlyconcerning the strategic planning done by theMAJCOMs. The FY85 guidance should be developed inthree areas:
HQ USAF should develop and publish the stra-tegic planning guidance for the MAJCOM/SOAs.
HQ USAF/ACD should prepare the Air Force ADPStrategic Plan.
HQ USAF/ACD should develop the format andpublish the instructions for preparingmission need statements for input to theLong Range Plan (current MCAP).
Until the above actions have been completed, theMAJCOM/SOAs will not be able to take the next step.
(4) The MAJCOM/SOAs Should Implement the New PlanningGuidance.
Based upon improved strategic planning and usingthe HQ USAF Strategic Plan, the MAJCOM/SOAs shouldprepare to respond to the FY85 budget call.
MAJCOM ADP strategic plans should beprepared.
Mission need statements should be preparedand submitted, using the existing MCAPprocess.
The Air Force panels that operate as part of the PPBSwill have to be instructed that if a mission require-ment survives the PPBS review it will not be formallypresented to the Air Staff again for requirementsreview (i.e., no DAR processing) unless it exceedsnewly established thresholds.
(5) Current Air Force Regulations Will Have To BeRevised
The revisions will be required in three generalareas, as follows:
VI-4
Regulations and guidance must be revised toprovide ADP strategic planning guidance.
The AFR 300 series regulations which definethe acquisition process must be revised andsimplified.
Guidance must be developed to define accept-able alternative acquisition approaches.
Interim guidance should be provided to the field, inthe areas deemed necessary, pending the completion ofdrafting and staffing of revisions to the regulations.
(6) Final Actions Are Necessary To Complete Implemen-tation of the Revised Acquisition Process
Final implementation actions are required inthree areas:
The MAJCOMs/SOAs must identify the addi-tional ADP and acquisition skills that arerequired to support the new acquisition con-cept. Reassignments must be made where pos-sible within existing manpower resources.Shortages must be addressed by manpower pro-curement and training, within the Air Forcetotal prioritization.
MAJCOMs/SOAs should identify their trainingneeds in the ADP and acquisition supportskills.
HQ USAF should establish a program to moni-tor and evaluate the process performance andto make changes as appropriate.
Taking these steps will enhance the successful implementa-tion of the strategy recommended in Chapter V.
2. REALIGNMENT OF ROLES AND RESPONSIBILITIES SHOULD BEACCOMPLISHED AT HQ USAF AND MAJCOMs/SOAs
As discusssed previously, some changes in roles andresponsibilities are mandatory in order to achieve thetotal benefits of the recommended changes in the process;other changes are highly desired. The following para-graphs address the sequence of actions required to effectchanges in organizational roles and responsibilities.
(1) Strengthen the Single Manager's Roles andResponsibilities
To strengthen the single manager's role, thefirst implementation step in revising roles andresponsibilities is to issue interim policy directionto: VI-5
* Authorize single managers to appoint programmanagers to programs for which funds havebeen approved
Increase single manager's approval authority
Raise the thresholds under which the singlemanagers are authorized to approve ADPsolutions and direct program implementations.
This interim direction should be followed by revisedregulations which implement the total processrevisions.
(2) The Role of AFCC in the ADP Acquisition ProcessMust Be Revised
The recommendations detailed a dual role for theAFCC centers--direct responsibility for multi-usersystems, and support responsibilities as an acquisi-tion consultant in support of MAJCOM and SOA programmanagers. The following implementation actions arerequired to clarify these roles:
Issue interim direction that redefines theroles and responsibility of the AFCC centers.
Direct AFCC to prepare a transition planwhich revises the role of AFCAC and consid-ers other organizational realignments.
Revise AFCC's charter to provide increasedacquisition support to MAJCOMs/SOAs.
(3) ADP Responsibilities Should Be Transferred FromAF/ACD Into AF/RD, XO and LE, as Appropriate
A phased implementation of these responsibilitiesis encouraged. AF/ACD should remain intact as long aspossible to effect a smooth transition. Transfer ofresponsibilities should be accompanied by an appro-priate transfer of resources.
Direct AF/ACD to take the lead in developinga transition plan with RD, XO and LE.
Direct the assumption of the responsibili-ties by RD, XO and LE upon approval of thetransition plan.
VI-6
This transfer of responsibilities should coincide withthe next step as much as possible, but neither stepshould be allowed to be a reason for delaying theother step.
(4) Transition the ADP Acquisition Executive Respon-sibilities From SAF/FM to SAF/AL
The completion of this last step will completethe changes at HQ USAF and will result in the respon-sibilities at the SAF and Air Staff being similarlyaligned for ADP acquisition.
3. SPECIFIC ACTIVITIES ARE REQUIRED BY THE ACQUISITIONIMPROVEMENT GROUP (AIG) TO IMPLEMENT THERECOMMENDATIONS
Because of their expertise acquired during this study,the AIG personnel are the best qualified to handle many ofthe implementation actions. They have a depth of back-ground information and are not encumbered by any assignedduties other than this study.
(1) The AIG Should Take Certain Implementing Actions
Specific implementing actions for the AIG are asfollows:
Brief selected external agencies on theresults of the study.
Brief MAJCOMs/SOAs and Centers on theresults of the study.
Develop the interim guidance for theMAJCOM/SOA Single Managers.
Develop planning guidance.
Develop alternative ADP acquisitionapproaches.
-Revise the AFR 300 series regulations.
- Incorporate the planning guidance.
- Define the revised ADP acquisitionprocess.
- Define alternative acquisition
approaches.
Support the development of the plan to tran-sition responsibilities from ACD to RD, XO,and LE.
VI-7
Support the development of directions forthe realignment of responsibilities of theAFCC ADP centers.
The preceding implementation activities are within theskills and capabilities of the AIG personnel. Duringthe study subjects were identified which require fur-ther study. These are described in the followingsection.
(2) The AIG Should Further Study and Analyze CertainSubjects
In addition to the implementation actions pre-viously discussed, certain actions require furtheranalysis to complete the implementation. Specificareas are as follows:
* Develop a life cycle model (LCM) for ADPsystems.
* Define criteria for determining technologi-cal and economic obsolescence.
* Develop, analyze and recommend alternativeADP acquisition approaches.
- Research and define alternatives.- Make recommendations.- Define applicability.
Develop an ADP Program Manager's Handbook.
Study and make recommendations on the needfor the establishment of an Office of Infor-mation Technology in HQ USAF which wouldhave responsibility for telecommunications,ADP and computer-based office systems.
The first three items above are needed to completeimplementation of the recommendations. The study ofthe need for a single office to be responsible fortelecommunications, ADP and computer-based office sys-tems is not essential for the implementation of theADP acquisition improvement study. However, with thepresent division of the three responsibilities withinthe Air Staff, together with the requirement to imple-ment the Paperwork Reduction Act, PL96-511, it isappropriate to undertake such a study now.
VI-8
Implementation actions can begin almost immediatelybut could extend over a period of several years beforefinal regulation changes and field level OIs are formallypublished and distributed. However, the near-term imple-mentations can be expected to produce near-term improve-ments in the ADP acquisition cycle, i.e., shorten thecycle. The implementation of the qualitative type ofimprovements in the acquisition process takes longer to befelt. The expertise of the Acquisition Improvement Groupcan be profitably utilized to make a smooth implementation.
VI-9
REFERENCES
REFERENCES
1. The Acquisition of Automatic-Data Processing SystemsWithin the Department of Defense and Other FederalAgencies, Report to the House Appropriations CommitteeSurveys and Investigation Staff. May 10, 1978
2. Acquisition Team Report of Federal Data ProcessingReorganization. President's Reorganization Project.office of Management and Budget. June 20, 1978
3. Evaluation of the General Services Administration:Report to the Administrator. National Academy of PublicAdministration Panel. December 31, 1980
4. Automatic Data Processing Strategy Study, Volume II:Technology, History, and Evaluation. Decisions andDesigns, Inc. March 1975
5. The Acquisition of Automatic Data Processing SystemsWithin the Department of Defense. Report to the HouseAppropriations Committee. Surveys and InvestigationStaff. March 1980
6. Report of Ad Hoc Working Group on Embedded ComputerResource Acquisition Policy. Joint Services Ad HocCommittee. April 22, 1980
7. Air Force Preliminary ADP Acquisition study. FrancisBurke, Gordon Ray, Kevin McManus. USAF/ACDX
8. Continued Use of Costly outmoded Computers in FederalAgencies Can Be Avoided. Government Accounting office1980
9. Review of the Implementation Of DOD Life CycleManagement Policies for Automated Information Systems:Draft of Proposed Report. Defense Audit Service.September 16, 1980
10. "Federal ADP Systems Atrophy in the Sinews of Govern-ment" Robert Head, Government Executive. February 1981
11. Report of the Committee on Government Procurement.Volume 2, Part C Acquisition of Major Systems.Committee on Government Procurement, December 1972
12. 'Improving the Acquisition Process" Memo, 30 April 1981,Frank Carlucci, Deputy Secretary of Defense, April 1981
13. Report on Information Systems Management. Commission onFederal Paperwork. 1977
14. The Report of the Privacy and Protection StudyCommission. Privacy and Protection Study Commission.July 1977
15. Report of the Central Agencies Team of the Federal DataProcessing Reorganization Project. President'sRe-organization Project. December 1978
16. Report of the General Government Team of the FederalData Processing Reorganization Project, President'sReorganization Project. June 1978
17. Report of the National Security Team of the Federal DataProcessing Reorganization Project. President'sReorganization Project. October 1978
:1
