Holding the Edge: Maintaining the Defense Technology Base

Holding the Edge: Maintaining the DefenseTechnology Base

April 1989

NTIS order #PB89-196604

Recommended Citation:U.S. Congress, Office of Technology Assessment, Holding the Edge: Maintaining theDefense Technology Base, OTA-ISC-420 (Washington, DC: U.S. Government PrintingOffice, April 1989).

Library of Congress Catalog Card Number 89-600711

For sale by the Superintendent of DocumentsU.S. Government Printing Office, Washington, DC 20402-9325

(order form can be found in the back of this repent)

Foreword

Technological superiority has been a cornerstone of United States security and industrysince World War II. That cornerstone is not crumbling, but over the past decade it hasweathered significantly. Foreign companies have made deep inroads into high-technologymarkets that had been more or less the exclusive domain of U.S. industry. In addition tocausing economic problems, this has fostered dependence on foreign sources for defenseequipment at a time when the technology in defense systems comes increasingly from thecivilian sector. At the same time, the Department of Defense reports that Soviet defensetechnology is catching up with ours, and sophisticated Western military equipment is routinelysold to third world nations.

These trends-and others-have prompted the Senate Committee on Armed Services toask what needs to be done to maintain the base of high technology on which U.S. nationalsecurity depends. This report, the second of OTA’s assessment “Maintaining the DefenseTechnology Base,” looks into that question in some depth. An earlier report, The DefenseTechnology Base: Introduction and Overview (OTA-ISC-374, March 1988), provided a broadview of the defense technology base and the concerns regarding its health.

This report develops some of the ideas introduced in the first report. It examines themanagement of DoD technology base programs and laboratories. It also analyzes the processthrough which technology is introduced into defense systems, in order to understand why ittakes so long and what might be done to speed the process up. Finally, this report examinesthe exploitation of civilian commercial sector technology for defense needs. It concentrateson the dual questions of expediting military access to civilian technology and keeping thenecessary base of technology alive and well in the United States. Volume 2 of this reportcontains extensive appendices and will be published in the summer of 1989.

The help and cooperation of the Army, Navy, Air Force, the Office of the Secretary ofDefense, the Department of Energy, NASA, and the National Institute of Standards andTechnology are gratefully acknowledged.

JOHNDirector

Defense Technology Base Advisory Panel

Michael R. BonsignorePresidentHoneywell International

William CareyConsultant to the PresidentCarnegie Corp. of New York

Thomas E. CooperVice PresidentAerospace TechnologyGeneral Electric

John DeutchProvostMassachusetts Institute of Technology

Walter B. Laberge, ChairVice President of Corporate Development

Lockheed Corp.

Robert FossumDeanSchool of Engineering and Applied SciencesSouthern Methodist University

Jacques GanslerSenior Vice PresidentThe Analytic Sciences Corp.

B.R. InmanAdmiral, USN (retired)Chairman and Chief Executive OfficerWestmark Systems, Inc.

Paul KaminskiPresidentH&Q Technology Partners, Inc.

Lawrence KorbDirectorThe Center for Public PolicyBrookings Institution

George KozmetskyExecutive Associate—Economic AffairsUniversity of Texas SystemUniversity of Texas, Austin

Ray L. LeadabrandPresidentLeadabrand & Assoc.

Jan LodalPresidentINTELUS

Edward C. MeyerGeneral, USA (retired)

Robert R. MonroeVice Admiral, USN (retired)Senior Vice President & Manager, Defense & SpaceBechtel National, Inc.

William J. Perry (ex officio)Managing PartnerH&Q Technology Partners, Inc.

Richard PewPrincipal ScientistBBN Laboratories, Inc.

Herman PostmaSenior Vice PresidentMartin Marietta Energy Systems, Inc.

Judith ReppyAssociate DirectorCornell Peace Studies Program

Richard SamuelsProfessorDepartment of Political ScienceMassachusetts Institute of Technology

John P. ShebellManager, RAMP EngineeringCustomer Service Systems EngineeringDigital Equipment Corp.

Michael ThompsonExecutive DirectorIntegrated Circuit Design DivisionAT&T Bell Laboratories

S.L. ZeibergVice President Technical OperationsMartin Marietta Electronics and Missiles Group

NOTE: OTA appreciates and is grateful for the valuable assistance and thoughtful critiques provided by the advisory panel members.The panel does not, however, necessarily approve, disapprove, or endorse this report. OTA assumes full responsibility for thereport and the accuracy of its contents.

iv

OTA Project Staff-Defense Technology Base

Lionel S. Johns, Assistant Director, OTAEnergy, Materials, and International Security Division

Peter Sharfman, International Security and Commerce Program Manager (through February 1989)

Alan Shaw, International Security and Commerce Program Manager (from March 1989)

Alan Shaw, Project Director

William W. Keller

Gerald L. Epstein

Laurie Evans Gavrinl

Christine Condon2

Congressional Research Service Contributor

Michael E. Davey

Administrative Staff

Jannie Home (through November 1988)

Cecile Parker

Jackie Robinson

Louise Staley

Contractors

P. Robert Calaway

Arnold Levine

MIT/Japan Science and Technology Program

1~ ~~~lgment from (YIA’s Energy and Materials ~of?am,2~ ~sslgment from the Department of Defense.

Workshop on the Relationship Between Military & Civilian Fiber OpticsJohn R. Whinnery, Chair

University Professor EmeritusDepartment of Electrical Engineering and Computer Sciences

University of California, Berkeley

James H. DavisDirector, Fiber Optics Program OfficeNaval Sea Systems Command

Brian HendricksonChiefElectro-Optics Technology BranchU.S. Air Force

Raymond E. JaegerPresident and CEOSpecTran Corp.

Donald B. KeckDirectorApplied Physics Research&Development LaboratoriesComing Glass Works

Tingye LiDepartment HeadLight Wave Systems Research DepartmentAT&T Bell Laboratories

John W. LyonsDirectorNational Engineering LaboratoryNational Institute of Standards and Technology

Alan McAdamsProfessorJohnson Graduate School of ManagementCornell University

William C. McCorkleTechnical DirectorU.S. Army Missile Command-Redstone Arsenal

Kenneth NillExecutive Vice PresidentLasertron

Paul PolishukPresident and ChairmanInformation Gatekeepers Group of Companies

Jan H. SuwinskiSenior Vice President and General ManagerTelecommunicationsCorning Glass Works

Robert W. TarwaterLight Guide Fiber & Cable ManagerAT&T Network Systems

Victor R. BasiliProfessorDepartment of Computer SciencesUniversity of Maryland, College Park

Elaine BondSenior Vice PresidentThe Chase Manhattan Bank

Mike DevlinExecutive Vice PresidentRational

Barry BoehmChief ScientistTRW. Inc.

Workshop on the Relationship Between Military & Civilian SoftwareLarry E. Druffel, Chair

DirectorSoftware Engineering Institute

Carnegie-Mellon University

Jeffrey M. HellerSenior Vice PresidentElectronic Data Systems

Dana P. LajoieTechnical DirectorGovernment Systems GroupDigital Equipment Corp.

John A. LytleDirector of Technical DevelopmentPlanning Research Corp.

Allan L. ScherrVice President, Development & IntegrationApplications Systems DivisionIBM

Mike Weidemer David M. WeissDeputy DirectorMission Critical Computer Engineering

Principal Member-Technical Staff

Air Force Systems CommandSoftware Productivity Consortium

Workshop on the Relationship Between Military & Civilian PMCsDick J. Wilkins, Chair

DirectorCenter for Composite MaterialsUniversity of Delaware, Newark

Ric AbbottPrincipal EngineerAdvanced Composites ProjectBeech Aircraft Corp.

James N. BurnsVice President of MarketingHercules, Inc.

Samuel J. DastinDirector, Advanced MaterialsGrumman Aircraft Systems

Bernard M. Halpin, Jr.Manager, Composites Development BranchMaterials Technology LaboratoryU.S. Army Laboratory Command

James J. KellyProgram Area ManagerMaterialsOffice of Naval Technology

Robert ManildiManager, Advanced CompositesHexcel Corp.

Michael J. MichnoDirector of TechnologyAdvanced CompositesAmocoPerformance Products

Alan G. MillerUnit Chief, Chemical TechnologyBoeing Commercial Aircraft

Thomas F. O’BrienSegment ManagerAdvanced Composites DivisionDupont Co.

Frances RensvoldPhysical Science AdministratorAero Mechanics Technology, Andrews Air Force Base

Douglas C. RuhmannChief Design EngineerManager, Materials & ProcessesMcDonnell Douglas Astronautics Co.

Nick SpenserSales ManagerComposites MaterialsCIBA-GEIGY

NOTE: OTA appreciates and is grateful for the valuable assistance and thoughtful critiques provided by the participants in theworkshops. The workshop participants do not, however, necessarily approve, disapprove, or endorse this report. OTA assumesfull responsibility for the report and the accuracy of its contents.

vii

ContentsChanter Page

l. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. Options for Congress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .. .. . $ . . . . . . . . . . . . . . . . . .3. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. Planning and Funding DoD Technology Base Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5. The Management of Defense Department Laboratories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6. Exploiting Other Management Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

39

19416389

7. Implications for the Defense Technology Base: Options for Congress . ...................1118. Lab to Field: Why Soling? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......1299. Civilian Technology and Military Security . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ......161

Acronyms

AAAS

AATR

AECAMCANSI

APLASD

AT&TATD

ATRATTD

BTICDPCERN

CEST

CINCs

COBOL

COCO

COECOTSCSIS

D-RAMDARPA

DCS(T&P)

DCSPER

—American Association for the Ad-vancement of Science

—Aided Automatic Target Recogni-tion

—Atomic Energy Commission—Army Materiel Command—American National Standards Insti-

tute—Applied Physics Laboratory—Aeronautical Systems Division (Air

Force Systems Command)—American Telephone and Telegraph—Advanced Technology Demonstra-

tion—Automatic Target Recognition—Advanced Technology Transition Dem-

onstration—Balanced Technology Initiative—Chief of Defense Procurement (U. K.)—Controller, Establishments, Research,

and Nuclear—Centre for the Exploitation of Sci-

ence and Technology—Commanders in Chief (of Unified

and Special Commands)—Common Business-oriented (program-

ming) Language-contractor-owned, contractor-oper-

ated—Corps of Engineers—Commercial Off-the-Shelf—Center for Strategic and Interna-

tional Studies-dynamic random access memory—Defense Advanced Research Proj-

ects Agency—Deputy Chief of Staff for Technol-

ogy and Plans—Deputy Chief of Staff for Personnel

DDDR&E(R&AT) —Deputy Director of Defense Re-search and Engineering for Researchand Advanced Technology

DDR&E —Director of Defense Research andEngineering

DFG —German Research SocietyDG —Defense GuidanceDGA —Delegation Generale pour l’Arme-

ment (France)DoD —Department of DefenseDOE —Department of EnergyDRB —Defense Resources BoardDSB —Defense Science BoardEC —European CommunityEN —Engineering Directorate (of ASD)

FARFOG-MFSXGAOGNPGOCOGSIDAIEPGILIR

IRSTISDNISO

ISTO

ITT

JCSJDAJRC

JSEPJSPDKDDLABCOMLAMPFLANLHXMADOMMIL-STDMITI

MMIC

MoDNASA

NDINIFNIST

NOSCNRLNSFNSIA

NIT

NWCO&SODDR&E

—Federal Acquisitions Regulations—Fiber Optic Guided Missile—Fighter Support Experimental—General Accounting Office—gross national product—government-owned,-contractor-operated—General Schedule—Institute for Defense Analyses—Independent European Program Group—In-house Independent Laboratory Re-

search—Infrared Search and Track—Integrated Services Digital Network—International Standards Organiza-

tion—Innovative Science and Technology

Office—International Telephone and Tele-

graph—Joint Chiefs of Staff—Japan Defense Agency—Joint Research Centers (European

Community)—Joint Services Electronics program—Joint Strategic Planning Document—Kokusai Denshin Denwa (Japan)—Laboratory Command—Los Alamos Meson Physics Facility—local area network—Light Helicopter Experimental—Magneto-acoustic Detection of Mines—Military Standard—Ministry of International Trade and

Industry (Japan)—Monolithic Microwave Integrated

Circuit—Ministry of Defense (U. K.)—National Aeronautics and Space Ad-

ministration—non-developmental items—Naval Industrial Fund—National Institute of Standards and

Technology—Naval Ocean Systems Center—Naval Research Laboratory—National Science Foundation—National Security Industrial Asso-

ciation—Nippon Telephone and Telegraph

(Japan)—Naval Weapons Center—Operations and Support—Office of the Director of Defense

Research and Engineering

OECD

OMBOPMOSDOSTP

OTAPCPEPEOPIPMCPPBS

R&ATR&DR&LM

RD&E

RDT&E

-Organization for Economic Coop-eration and Development

-Office of Management and Budget-Office of Personnel Management—Office of the Secretary of Defense-Office of Science and Technology

Policy—Office of Technology Assessment—personal computer—Program Element—Program Executive Officer—Principal Investigator—polymer matrix composite—Program Planning and Budgeting

System—Research and Advanced Technology—research and development—Research and Laboratory Manage-

ment—Research, Development, and Evalu-

ation—Research, Development, Testing, and

Evaluation

S&TSDISDIO

SESSTARSTARS

TCPTOATRDI

TSGUAVURCURIUSD(A)

USD(R&E)

—science and technology—Strategic Defense Initiative—Strategic Defense Initiative Organi-

zation—Senior Executive Service—Strategic Technologies for the Army—Software Technology for Adaptable

Reliable Systems—Technology Coordinating Panel—Total Obligational Authority—Technical Research and Develop-

ment Institute (Japan)—The Surgeon General—Unmanned Airborne Vehicle—University Research Centre (U. K.)—University Research Initiatives—Under Secretary of Defense for Ac-

quisition—Under Secretary of Defense for Re-

search and Engineering

x

Chapter 1

Introduction

Chapter 1

Introduction

Not long ago, the United States was theundisputed technological leader of the world.U.S. military equipment was meaningfully andundeniably more sophisticated than that of theSoviet Union, and our allies sought Americantechnology for their own defense efforts.American companies developed and sold high-technology products to a world that could notproduce them competitively. Defense-relateddevelopments led American technology andoften ‘‘spun-off” into the civilian sector, creat-ing products and whole industries. This rein-forced a U.S. defense posture based on usingtechnological superiority to offset whateveradvantages the Soviet Union and other potentialadversaries might have.

As we approach the 21st century, much haschanged. The model of U.S. technology leadingthe world, with defense technology leading theUnited States, still retains some validity. But itis a diminishingly accurate image of reality.Soviet defense technology increasingly approachesour own, and sophisticated weapons appear inthe hands of third world nations not long aftertheir introduction into Western and Sovietarsenals. At the same time, the U.S. military hasbeen plagued with complex systems that do notwork as expected, work only after expensivefixes, or simply do not work. Most are high-priced and take a long time to develop. Increas-ingly, leading edge technology comes from aninternationalized, civilian-oriented economy, whichputs a premium on exploiting technology as wellas developing it.

As a result, the Nation faces a complex set ofinterrelated problems that bear on its ability tocontinue to develop and manufacture in suffi-cient quantity the technologically advancedmateriel on which we base our national securityposture. There are specific concerns about:

1) the continued ability of the Department ofDefense (DoD) and its contractors to developthe technologies it needs; 2) the ability of DoDand the defense industries to turn these tech-nologies into useful, affordable products in atimely fashion; and 3) the ability of DoD toexploit the technology that is being developedworldwide in the private civil sector.

Concern over the availability of the latesttechnology for defense applications, and theability of U.S. industry to engineer and produceequipment based on that technology rapidly andaffordably, led the Senate Armed ServicesCommittee to request that OTA undertake anassessment of the defense technology base. Thisis the second report of that assessment. Theprevious report, The Defense Technology Base:Introduction and Overview,1 described what thedefense technology base is and presented themajor problems facing the Nation. This reportlooks in depth into some of the issues raised inthe previous report. It identifies strengths andweaknesses of the U.S. defense technology baseand analyzes options for enhancing the strengthsand remedying the weaknesses.

The summary of this report (ch. 3) is dividedinto three sections. The first addresses thestrategic management of DoD technology baseprograms. It examines the system by which thegoals of the technology base programs areidentified as well as the methods used to allocateresources in order to reach those goals. Theemphasis there is on the role played by theOffice of the Secretary of Defense (OSD) inguiding and coordinating the efforts of theArmy, Navy, Air Force, and other DoD ele-ments. It also addresses the management of thelaboratories run by the three Services. Theseissues are explored in greater detail in chapters4 through 7. The second section of the summary

Iu.s. co~~~~s, Office of Techn~lou A~%~~ent, T~ D~f~~~ T~chnolo~ Base: /ntroduc.tion ad overvie~ special Report, OTA-lSC-3TQ(Washington, DC: U.S. Government Printing Office, March 1988).

-3 -

4 ● Holding the Edge: Maintaining the Defense Technology Base

analyzes delays in getting technology into thefield (see ch. 8 for supporting details). The finalsection is concerned with ‘‘dual use” technol-ogy, i.e., technology used in both the civilianand defense sectors (see ch. 9). Volume 2 of thisreport contains detailed supporting material onselected topics for those wishing to explorethem in greater detail.

The remainder of this chapter provides a briefbackground on the topics of the report: man-agement of defense technology base programsand facilities; technology transition; and dual-use technology. Those familiar with these sub-jects may wish to skip directly to chapter 2,which presents issues and options for Congress.

A large part of the technology that ultimatelywinds up in weapons and other defense systemsis either developed or directly sponsored byDoD. This is particularly true of technology thatis altogether new, makes a major difference inthe performance of defense equipment, and is oflittle interest to commercial industry. How DoDruns its technology base programs is therefore ofmajor importance. In recent years DoD hasspent roughly $9 billion per year on its technol-ogy base programs: research (budget category6.1), exploratory development (6.2), and ad-vanced technology demonstration (6.3A).Roughly 40 percent of this is spent by the threeService departments (Army, Navy, and AirForce). Another 14 percent is controlled by theDefense Advanced Research Projects Agency(DARPA, formerly ARPA). Another 39 percentfinances the Strategic Defense Initiative Organi-zation (SDIO).2 Although all of SDIO’s fundsare allocated in the 6.3A budget category,according to SDIO only about 15 to 20 percentis actually spent on technology base activities.

The three Services run their technology baseprograms and their R&D institutions differ-

ently. 3 Some of this is the result of recentplanning, while much of it results from organ-izational ‘cultures” developed over many years.The Army’s effort emphasizes decentralization.The Army runs some relatively small researchlaboratories which focus on selected topics,while larger research, development, and en-gineering centers are closely tied to “buyingcommands.”4 The Navy stresses in-house re-search and development both in the NavalResearch Laboratory, a broad-based corporatelab that serves and underpins the Navy’s entiretechnology effort, and in full-spectrum researchand development (R&D) centers that nurtureideas from basic research through pre-production stages. These centers have tradi-tional ties to the equipment needs of variousfunctional parts of the operational Navy, but arenot formally tied to specific buying commands.The Air Force, which contracts out more of itsR&D effort than either of the other Services,centralizes its efforts within the Air ForceSystems Command. Its technology base pro-grams are seen as a link between buyingcommands (the divisions of Systems Com-mand) and the defense industry. The basictheme is to buy technology and make sure it getsto industry. The Air Force has recently adoptedthe position that technology base programsshould be a ‘‘corporate investment” funded atsome fixed fraction of the budget. The Air Forceputs a greater emphasis on R&D-related careerpaths than do the other Services.

With such diversity (including that added byDARPA, the other defense agencies and SDIO),if the program is to have overall planning andcoordination—and not everyone agrees that itshould—leadership almost has to come from theOffice of the Secretary of Defense.

z~er defense agencies account for approximately 7 percent of DoD technology base program funding. (See fcxxnote 1, P. 19 of tiis repofl.)

3A]] ~ree, ho~evcr, Orient their pro~ms heavi]y [oWard c~ent product areas. Nontra~itiona] ideas do not fit well into the SySteItL

4A buying c~mmand is one of a nm~r of Orgmizations wi~in the Armed Services responsible for developing and buying miiitary equipment andsystems.

Chapter l-introduction ● 5

Actual R&D is performed primarily byindustry, universities, and the laboratories runby the Services.5 In most cases laboratory is amisnomer, although a convenient shorthand.These latter institutions, as a group, performtechnology base work in addition to advancedand even full-scale development. They alsoprovide other functions to DoD. Their efforts aregenerally divided among performing in-housework, contracting out work (and monitoringcontractors’ efforts), and providing technicaladvice to program managers and buying com-mands (a function often referred to as being‘‘smart buyers”). It is very difficult to describea typical DoD lab because they differ in size, inthe mix of these functions, and in a number ofother basic elements. However, what they allhave in common is that they are owned and runby the government, staffed by governmentemployees, and subject to a large number oflaws and regulations. There has been a continu-ing and, in recent years, rising concern that theyare inefficient, ineffective, self-serving andduplicative of industry work, and increasinglyhampered in doing their jobs by the conditionsof being part of the government.

DoD has some important unique characteris-tics, but it is not the only large organization thatrelies heavily on new technology nor the onlyestablishment that runs R&D programs andfacilities. Large corporations and the govern-ments of other nations do the same. Theirspecific goals may not be the same: DoD buysdefense equipment to meet a threat, corporationsseek to develop and market products in acompetitive market, and other nations seek toenhance their economic positions as well as theirsecurity. But all share the general goal ofmarshaling technology assets to achieve somepurpose. To some extent, these other entitiesprovide some of the background against whichDoD must plan and execute its programs (cer-tainly the evolution of the threat is another). But

they also provide potential models of manage-ment techniques that might be useful to DoD insolving its management problems.

The technology base programs and laborato-ries produce technology, but that technology isof no use unless it makes its way into fieldedsystems that the military can use. There is greatconcern that it simply takes too long to get newtechnology into the field. Systems take upwardsof 10 years to develop and produce, and whenthey finally become operational, they oftenembody technology that is viewed as obsolete,either because better technology exists in thelabs or in industry, or because consumers canpurchase better technology at their neighbor-hood stores. In the previous report, OTA foundthat delays are not a technology base problem:they occur after the technology is developed.However, delays are a major obstacle to keepingour technological lead in fielded equipment.

While a majority of the most visible technol-ogy in defense systems comes from DoD andcompanies that contract with it, a significant partcomes from the “nondefense” sector. Mundanetechnology—like bolts—has often come fromindustries that sell to both military and civiliancustomers. And at the subcomponent level,much also comes from the civilian side. Increas-ingly, these “dual-use industries” are sourcesof advanced technology, sources from whichDoD should be able to draw (and in some casesmust draw, because the technology is ahead ofwhat the defense world is building). Increas-ingly, leading-edge technology is developed inthe civilian sector and then finds its way intodefense applications. But government rules thatmake doing business with the governmentdifferent from selling in the commercial sectorcreate significant barriers to companies movinginto government work. Some of these compa-nies are heavily involved in defense work, whileothers now do little or no business in the defense

5Work is a]w done by o~er government laboratories (e.g., the Department of Energy national labs and NASA labs) and VariOUS private Profit-makingand non-profit organizations.


sector. Moreover, dual-use industries are be-coming increasingly internationalized, raisingissues of the competitiveness of U.S. firms in theworld market and dependence on foreign suppli-ers in defense procurement.

DoD has become less able to drive thedirection of technology. While some areas arepursued primarily for defense applications, oth-ers are molded by the consumer market. Largecommercial markets generate enormous amountsof capital that fuel research and development.That R&D is primarily directed toward applica-tions and products with large potential commer-cial payoffs. The relatively small amount ofbusiness represented by sales for defense appli-cations is in many cases not significant enoughto swing the direction of development. There arestill many important areas of development thatare primarily, or exclusively, defense-oriented.But the pattern of technology originating in thedefense sector and “spinning off” into thecommercial sector is being replaced by paralleldevelopment and, to use the Japanese term,‘‘spin on” of commercial technology to militaryapplications. Faced with this situation, DoD canbuy cutting-edge technology developed in thecivilian sector, or it can spend large amounts ofmoney to keep a comparable leading edgeresident in-house or with defense contractors.

As a consequence, DoD finds itself (or itscontractors) having to buy from companies thatdo not need its business. Large aerospacecompanies have to play by DoD’s rules: defenseis their only business, or at least an over-whelming component of their business. But

small, leading-edge technology companies canmake much more money in the private sectorwithout the trouble of playing by governmentrules. They can opt out of doing defense work.

This report examines dual-use industriesthrough the mechanism of case studies, concen-trating on three industries: advanced compos-ites, fiber optics, and software. These presentdifferent perspectives. The advanced compositesindustry is heavily involved in defense business,but U.S. companies may see their commercialbase erode as international competition heats up.Moreover, many of the major companies areinternational or integrated with foreign firms.U.S. fiber optics producers now sell very littlefor defense applications. But DoD has importantuses for their products. Government buyingpractices form major barriers to these companiesdoing business in the defense market, and theyare beginning to face stiff competition in thecivilian market from foreign competitors. Fi-nally, the software industry is one that straddlesboth worlds, and moves very rapidly. Softwareis at the heart of most new defense systems,particularly command, control, communications,and battle management systems.

All of these topics have been the subject ofnumerous studies, which have produced con-flicting conclusions. This report pulls togethermuch of that work, along with original researchand analysis. Moreover, while DoD managementand industrial/trade issues have been the subjectof legislation and proposed legislation, theproblems are not yet solved. The next chapterdiscusses the major issues before Congress.

Chapter 2

Options for Congress

Chapter 2


The U.S. defense effort rests on a strong, broad,dynamic base of research and development. Govern-ment and private institutions, and civil and militaryestablishments all contribute. But this defense tech-nology base is also characterized by:

●

●

●

a heavy burden of government rules, regu-lations, safeguards, and procedures that stiflethe ability of the Department of Defense (DoD)to develop and exploit technology;the lack of an effective system for high-levelplanning and coordination; andthe lack of a clear government policy andcoherent strategy for dealing effectively withdynamic trends in the international high-technology economy.

To those who have followed defense industry,technology, and procurement, none of this will comeas a surprise. These problems—and more—havebeen noted and studied for at least three decades. Butdespite repeated attempts to fix it, the system hasremained resistant to major improvements. Indeed,the major problems have continued to worsen,although probably more slowly than if no measureshad been taken.

The U.S. is not faced with a defense technologybase that is in deep crisis. The Services and otherdefense activities fund a great diversity of researchand development, run a large number of laboratoriesthat do credible—and often outstanding-work, andsuccessfully exploit that technology and technologydeveloped elsewhere. But the process has a numberof serious shortcomings that may be amenable tosignificant improvement. Moreover, important re-cent trends threaten to intensify these shortcomingsand magnify their importance. U.S. leadership inhigh-technology industries that are vital to defenseis eroding in the face of strong international compe-tition. Budget restrictions predicted both by Con-gress and by the Administration will reduce fundingfor technology base activities at a time when thecosts of research and development are increasing.And DoD’s ability to compete successfully for keytechnical and managerial personnel is declining.

On top of all this, a heavy burden of rules andregulations impedes the development and exploitation

of technology and the successful transition ofdevelopments into fielded systems. The accumu-lated actions of past Congresses are a major con-tributor to the difficulties. Laws passed for a varietyof good reasons, taken together, bog the systemdown. Lack of clear policy on the part of bothCongress and the executive branch impedes thesolving of important problems.

Virtually all the easy solutions have been tried. Itis unlikely that any fruitful but painless approachesremain. Congress and the executive branch will haveto face some hard choices. These include alteringinstitutional arrangements that—despite their defi-ciencies—have become comfortable, and sacrificingexisting goals in order to achieve more efficientdevelopment and exploitation of technology.

Based on the analysis in this report, OTA hasidentified seven basic issues that profoundly affectthe welfare of the defense technology base. Theseare not specific action items, but rather broad agendaitems that warrant congressional attention. For eachof these there are many different choices as to whatindividual policy directions to take, and withinthose, a myriad of measures (and choices amongmeasures) for implementation. Implementation isclearly important, for without any sense of how toimplement a policy, it remains simply an abstrac-tion. There are options that can be implemented onlythrough legislation, because today the law forbidsthem or provides no way to make them happen. Andthere are options that can be implemented withoutchanging the law—through executive action orchanges in DoD’s internal regulations. Congress canhave a hand in effecting these sorts of changes bymaking its wishes known or by using its consider-able powers of persuasion.

Chapters 7, 8, and 9 discuss various specificpolicy options.

ISSUE 1: Reforming the Defense AcquisitionSystem

The defense acquisition system is a major contribu-tor to the long delays in getting new technology intothe field and erects formidable barriers to exploitingtechnology developed in the civilian sector. While


Congress did not intend the system to be slow,cumbersome, and inefficient, laws passed to fostergoals other than efficient procurement have made itso.

The system has weathered many attempts atreform because its problems are rooted in severalbasic causes. It is dictated in part by our basic systemof government which demands checks and balanceson the expenditures of large amounts of publicfunds, provides for a tug and pull between theinterests of the executive branch and those ofCongress, and permits both branches to reevaluateprograms yearly in light of changing factors andinterests. But much of the problem can be traced tolaws that Congress has enacted to curb abuses and tofoster goals other than efficient procurement ofdefense equipment. Laws and regulations have beenadded to ensure:

●

●

●

●

●

●

●

●

●

●

●

�❵

●

•

civilian control over military procurement,Administration control over Service activities,congressional control,protection of congressional constituent interests,environmental protection,fairness,competition,accountability,honesty,controllable business practices,minority interests,small business interests,protection against conflicts of interest, andprevention of large profits at taxpayer expense.

These many ends often conflict with each otherand with the objective of quick and efficientprocurement, which leads to compromises that cansatisfy few, if any, completely. Thus, the conse-quences of achieving these other objectives haveincluded high costs, long procurement times, ineffi-cient production, and restricted access to technol-ogy.

To promote these and other goals, the govern-ment has developed business practices and criteriathat differ markedly from those of the civilianmarket. Buyer and seller have an adversary relation-ship; accountability is stressed over efficiency andprice; and the government insists on visibility intohow its contractors conduct their business. Govern-ment imposes restrictions on profits, trade secrets,

and accounting procedures that are at variance withtypical commercial practices. This discourages manyinnovative companies from seeking defense busi-ness.

History provides little hope that a few clever,relatively painless moves will be sufficient to makethe system significantly more efficient while satisfy-ing other goals. If Congress is serious aboutmaking the system work better, it will have toface some hard choices. One choice is to giveefficient procurement greater emphasis over othergoals. This would most likely mean that the systemwould become less fair, less competitive, lessaccountable, less responsive to minority and smallbusiness interests, etc. Another option would be forCongress to give up some of the power it has overmajor defense programs, or to curtail sharply someof the many centers of power within the executivebranch. This would not necessarily make anyparticular program run better-two layers of manage-ment could be just as ineffective as 20-but it wouldremove major impediments. Instituting multi-yearbudgeting, which could also make programs runmore quickly and smoothly, would likewise requireboth Congress and the executive branch to give upsome power. Finally, Congress could loosen up therules under which DoD conducts business, allowingbusiness practices to move closer to those of theprivate sector. But inherent differences betweengovernment and private operations will alwaysremain. For example, the government is accountablefor the expenditure of public funds and is verysensitive to allegations of misuse. Where a businesswould be willing to absorb some pilfering if it wereexceeded by the cost of prevention, the governmentis usually willing to spend whatever is necessary toprevent fraud.

Few such moves would come for free. Forexample, relaxing accountability rules could make iteasier for companies to cheat the government. It maywell be that, weighing all these factors together,Congress will decide that the current balance amongall these interests is proper, and that inefficientdefense procurement is an acceptable cost. Whileconcerns for efficient procurement will push in thedirection of loosening up the system, a need torespond to a recent history of procurement scandals,failed programs, and high-cost low-quality equip-ment will likely push in the opposite direction.

Chapter 2-Options for Congress ● 11

ISSUE 2: Independent Research and Development(IR&D) Recovery

Current law permits companies having contractswith DoD to bill to the government, as a cost ofdoing business, part of the cost of their internallygenerated R&D program. Industry generally be-lieves that current rates of recovery are inadequate.Some think recovery rates are too high. DoD cannotseem to present a coherent position. IR&D recoveryis not treated in this assessment, but it is very likelyto be on the congressional agenda. Interested readersare referred to OTA’s previous report The DefenseTechnology Base: Introduction and Overview.1

ISSUE 3: Reforming the DoD Laboratory System

As a whole, the DoD laboratory system performsits function of supporting defense procurement. Asa group, laboratory managers are capable andexperienced and provide much of the corporatememory for technology base activities. But thesystem is vast, complicated, and uneven in perform-ance. The structure of the system as a whole—thenumber, types, sizes, orientations, and institutionalconnections of the labs—may be restricting theirutility and effectiveness. Moreover, the managementsystem under which these government owned andoperated facilities are run is rendering it increasinglydifficult for them to function effectively. A long listof rules impedes their daily operations and makesthem increasingly unable to compete for highlyqualified scientists and engineers. In general, Con-gress can choose to:

● reform the system itself,. order DoD to reform it according to congres-

sional guidelines, or. leave the job to DoD.

Whatever course Congress chooses, it is unlikelythat the correct approach will be either simple orobvious.

There are three basic approaches to reforming labmanagement. The least disruptive would be to alter,within the current civil service system, the rulesunder which they operate. This could include:

. extending the principal features of the NOSC/China Lake personnel experiment to other labs,

. permitting the labs expedited procurementprocedures for scientific equipment and serv-ices, and

. providing multi-year funding.

Alternatively, Congress could decide that R&D isinherently different from other government activi-ties, and that the labs should be allowed to operatedifferently from the rest of DoD. This might includepermitting salaries for scientists and engineers torise above current civil service ceilings and allowingthe labs to build and modernize facilities by goingoutside the military construction process. The mostradical approach would be to convert some or all ofthese facilities to government-owned, contractor-operated (GOCO) facilities, like the Energy Depart-ment National Laboratories. Conversion to GOCOcould solve some of these problems, but would be nopanacea.

Congress should also seriously consider alteringthe overall structure of the laboratory system. Thiscould include closing some labs, consolidatingothers, shifting the internal make-up and missions ofsome, and creating new ones. Corporate researchlabs, like the Naval Research Laboratory, might beestablished for all the Services; or the in-housecapabilities of many labs could be greatly improved.In the process, the system should get simpler, notmore complicated. Greater integration of DoD labswith other government labs—reform of the overallgovernment lab system—might also be considered.This could include forming research centers ‘tospearhead major thrusts into areas of particularsignificance for both defense and commercial needs.These would be drastic steps requiring careful,detailed study and assessment of the individual labsbefore implementation. If done correctly, they couldlead to greatly improved benefits from DoD R&Dexpenditures. If done carelessly, they could becounterproductive. At the heart of the process wouldbe devising a system for evaluating the performanceof the laboratories and their component parts. Thisought to include the quality of work as well as itsrelevance to both identified Service needs andpotential future advances.

Restructuring the lab system may be a neces-sary response to budget pressures that reduce funds

IRelea~d Nfmch ICJM, repo~ No. (JTA.ISC-3’74. .&ailable from the U.S. Government Printing Office, washin~on, Dc.


available to run them. Significant reductions couldbe accommodated by reducing all efforts proportion-ately, but this would reduce good work as well asbad. Other approaches are closing the least produc-tive and useful labs or effecting a more extremerestructuring of the entire system to maximizeperformance and utility at a lower overall level ofeffort.

ISSUE 4: Reforming Strategic Planning ofResearch and Development Programs

Unlike many governments and large corporations,the Department of Defense does not have a centralheadquarters-level system for planning and coordi-nating its technology base programs. Planning iscarried out by the Services, the defense agencies, andthe Strategic Defense Initiative Organization (SDIO);coordination among similar projects is done at thelaboratory level. This lack of central focus isrepeated both higher up the chain-at the overallnational level—and within the individual Services.2

This is not necessarily bad, If centralization stiflesunplanned innovation and healthy competition, failsto support Service needs, or results in decisionmak-ing by the uninformed, then it is counterproductive.However, lack of overall planning can lead towasteful duplication of efforts, lack of critical massto solve common problems, fractionated efforts, andinattention to areas that are on no componentorganization’s agenda. It also risks failing to identifyareas of common or overarching significance. Ifthere is to be strategic planning and central coordina-tion, it will have to be done by the Office of theSecretary of Defense (OSD). Congress should de-cide whether-m many DoD studies have advocated—OSD ought to be given greater power (or encouragedto exercise the power the law already gives it) toplan, coordinate, and oversee technology base pro-grams; or whether Service dominance should besupported and reinforced. More forcefully, Congresscould order OSD to develop a strategic planningprocess to lead to a coordinated, department-widetechnology base investment strategy.

As currently organized, OSD oversees Servicetechnology base programs at one organizationallevel, DARPA at a second, and SDIO only at thehighest level. This inhibits real coordination. More-

over, it leads to the lack of a high-level advocatewithin OSD exclusively for technology base pro-grams, lowering the status of technology baseprograms within both DoD and Congress.

Strategic planning and program coordination aredifferent from central management. The formerrefers to a strategic OSD planning function provid-ing the ability to orchestrate the entire program.OSD could perform this planning role from a broadperspective over all the technology base activitiesthat the individual Services do not have, but it wouldlack the detailed information and insight into theworkings of specific programs necessary to managethem effectively. Planning and coordinating pro-grams and then letting the extensive Service R&Dorganizations manage them is different from aggre-gating similar programs and managing them fromOSD.

Congress could also define more clearly what itsown role is. It seems unlikely that Congress canprovide direction to the thousands of individualprojects. Congress could actively involve itself inthe strategic planning process or confine its activi-ties to demanding that OSD produce and defend astrategic R&D plan.

ISSUE 5: Reforming Government PersonnelPractices

Recruiting and retaining qualified scientists andengineers is a major problem for DoD laboratories.In the current sellers’ market, government salariesand benefits for technically trained personnel are notgenerally competitive with either industry or univer-sities. Many DoD labs have given up trying to recruitthe best and the brightest. Loosening up the rigidcivil service salary structure is a principal compo-nent of ideas to reform lab management, and beingable to pay competitively-above civil serviceceilings—is a major incentive for converting labs toGOCO status. Federal pay raises, if they are enactedand applied in any significant way to scientists andengineers, could substantially help the situation;alternatively, Congress could consider a separatepay scale for scientists and engineers more in linewith industry and academia. This may not be apermanent problem, since the market for scientists

Z’rhe Services seem to exercise more influence over their components than OSD does over tie Services.


and engineers tends to be cyclic. But until such timeas it turns around, defense technology base effortsare being hurt by a system that cannot adjust to themarket. It is also possible that this time the marketwill not turn around, that the current expansion inhigh-technology industry-coupled with demographictrends-will keep the supply short for along time tocome. Congress may also want to consider efforts toincrease the number of students in technical disci-plines. Defense efforts are particularly hard hit byshortages because they mostly require U.S. citizensand can take little advantage of the large number offoreign graduate students in U.S. universities.3

Some observers see similar problems in attractinggood managers of acquisition and technology baseprograms. People with the requisite skills andknowledge can command greater salaries in indus-try, and are reluctant to work for DoD. “Revolvingdoor” rules are also a disincentive to governmentservice. Congress may wish to consider reviewingsalary levels. It may also be worthwhile for Congressto gain deeper insights into the inhibitory effects ofother employment restrictions and reconsider themin this light.

ISSUE 6: Fostering Greater CoordinationBetween Defense and Civil Research andDevelopment

National defense benefits from a vibrant civiliantechnology base. Civilian research provides anotherlarge source of technology that finds its way intodefense systems, and effective civilian R&D under-pins a strong economy that provides greater reve-nues for defense efforts. The ability of the militaryto achieve and maintain leading-edge technologywill, in many cases, depend on the health ofcorresponding civilian industries. In a very generalsense, economic security is a major component ofnational security; the ability of the United States tocompete economically is intertwined with its abilityto compete militarily.

The U.S. defense and civil sectors are not isolatedfrom each other, but they are far from closelycoupled. Two relatively separate sectors have evolved—

one military and the other commercial. The diffusionof civilian technology into defense systems ishampered, as is the availability for commercialpurposes of technology developed in the militarysector. Some of this is unavoidable: security oftendemands that some technology be kept under wraps.But much is the result of government business rulesthat erect barriers to commercial companies sellingto DoD and of a weak, high-level technology policyapparatus.

Other industrialized nations—particularly inWestern Europe and Japan-construct their tech-nology efforts with a greater emphasis on economicdevelopment over military development than doesthe United States. They are increasingly demandingthat military technology support commercial develop-ment whenever possible. In Japan, almost all tech-nology is developed for commercial purposes, andsome of it is then exploited for military uses. Whatis appropriate for these other nations is not neces-sarily good for the United States, since neither Japannor any Western European nation aspires to be asuperpower. However, these are the nations withwhich the United States is competing economically.We maybe able to benefit from making both militaryand civilian R&D do double duty.

There are several things Congress could do tofoster greater symbiosis of civil and military tech-nology. Steps could be taken to expand the availabil-ity for commercial exploitation of the vast amount ofR&D done in DoD laboratories and under contractto DoD. Tying the Defense laboratories more closelyto those of other agencies—for example by fosteringexchanges of personnel or forming major researchcenters for dual-use technology-could benefit bothmilitary and civilian developments. Both the devel-opment of technology and its transition into engi-neering could be helped by movement of technicalpersonnel between government and industry.

The acquisition system could be reformed tomake it easier for DoD to do business with innova-tive companies in the commercial high-technologyindustries. Government regulations on profits, data

3The ~ue~tion of ~tentl~ shofifalls in the fut~e supply of Scientists and engin~rs is addres~d in U.S. Congress, Office of Technology Assessment,

Educating Scientists and Engineers: Crude School b Crud School, OTA-SET-377 (Washington, DC: U.S. Government Printing Office, June 1988);and U.S. Congress, Office of Technology Assessment, Higher Educarion jbr Scientists and Engineer#ackground Puper, OTA-BP-SET-52(Washington, DC: U.S. Government Printing Office, March 1989).


rights, and accounting procedures all discouragethese companies from seeking defense business.

Congress may find it worthwhile to reconsidercurrent mechanisms for setting technology policiesat the highest levels of government. In particular, itmay wish to provide for a high-level organizationthat would oversee and coordinate major government-sponsored R&D programs.

ISSUE 7: Dealing With International Trends inHigh-Technology Industry

The United States is failing to maintain a competi-tive commercial base for some technologies that areimportant for defense procurement. Long standingindustrial and trade policies may have to be reformedif the United States is to maintain the industrialcapacity necessary to support essential dual-usetechnologies.

Both Congress and DoD have been concernedabout the movement of high-technology industriesoffshore. This has spawned several responses, in-cluding attempts to legislate that DoD buy almostexclusively from domestic suppliers. This approachwould probably minimize foreign content in U.S.defense systems, but it attacks the symptom ratherthan the cause. It would have little effect on theability of U.S. companies to compete effectively inthe international marketplace-a key to havinghealthy, leading-edge companies here for DoD tobuy from.

Having dual-use companies in the United Statesand available to DoD requires that they be suffi-ciently competitive on the world market to stay inbusiness. Defense business alone is not usually bigenough to keep them afloat. And creating captivecompanies that exist only on assured DoD businesswill almost certainly guarantee that technology fallsbehind the state of the art. Furthermore, cuttingourselves off from foreign technology will meandepriving our defense efforts of important technol-ogy that is not available here but possibly isavailable to the Soviets on the open market.

The United States will have to deal with twofundamental phenomena. First, high technology is aworldwide enterprise. The United States no longerhas a monopoly on it. We can change our positionrelative to the rest of the world, but it is extremely

unlikely that we will regain the dominant positionthe United States once enjoyed. Second, individualcompanies and entire industries are becoming inter-nationalized. It is becoming increasingly difficult (ifnot impossible) to define what an American com-pany is. Plants in the United States are owned byforeign nationals or foreign-based corporations. AndU.S. companies open plants in other nations. More-over, international partnerships lead to foreigninterests in U.S. ventures and partial U.S. ownershipof foreign factories. Protecting U.S. interests andensuring U.S. sources of supply are therefore notsimple matters. This is complicated by the measuresthat other nations take to protect their companies andtheir home markets.

The United States has yet to begin to formulatea policy to deal with this situation, both with regardto defense procurement and as it relates to the futureof the U.S. economy as a whole. Congress will befaced with decisions on how dependent on foreignsources DoD can be, which high-technology indus-tries must be kept viable in the United States, how tomaintain those industries, and how to protect U.S.defense needs as companies become internationalized.Congress will have to formulate policy with regardto foreign ownership of U.S. plants and foreignsiting of U.S.-owned facilities-or encourage theAdministration to do so.

The solution is almost certain to be found amongthe choices that lie between the two extremes ofbuying defense components only from U.S.-basedand owned suppliers, and buying solely on the basisof getting the best deal. The former is likely to beincompatible with staying on the leading edge oftechnology, and the latter may well reduce the U.S.base of technology and manufacturing to a level thatis insufficient in time of crisis if not in peacetime.These intermediate choices include buying from:

● U.S.-based foreign-owned companies,. U.S.-owned companies regardless of location,

and. nearby sources (i.e., Canada or Mexico) regard-

less of ownership.

In formulating policy, the Nation will have todecide how important foreign ownership is and towhat degree domestic siting of development andmanufacture is necessary. That policy will have totake into account factors such as: international


patterns of trade, manufacturing, and corporateownership; the costs and opportunities of maintain-ing domestic capabilities; existing relations withother nations; and the effects of policy choices onforeign relations. It is one thing to be interdependentwith an allied nation, and quite another, as the oilshocks of the 1970s demonstrated, to be dependenton just any nation. Every nation ultimately presentsa different case, but the spectrum ranges fromCanada-which is adjacent, a NATO ally, anddefined as part of the North American industrialbase—through our European NATO partners, Japan,other European trading partners, and ultimately tonations with which our ties are very uncertain.

The intricacies of formulating policy are illus-trated by the problems of trade in defense equipmentwith our NATO allies. The United States is pursuingmultinational cooperation and integration of defense-

related development programs through vehiclessuch as the Nunn Amendment, both for political-military reasons and to promote sales for U.S.defense firms. But these actions will also lead togreater competition from European defense com-panies in the United States and abroad. Access toEuropean technology will be offset by the diffusionof U.S. technology.

Policy decisions regarding foreign dependencefor defense needs fall into the jurisdictions of DoDand the Armed Services Committees. But thebroader issue of how the United States should dealwith the international economic situation in order toachieve these and other goals will involve a muchmore diverse cast of players. Congress will have todecide both how it will approach the problem in amanageable way, and what restructuring might benecessary within the executive branch.

Chapter 3

Summary

CONTENTS

MANAGEMENT OF PROGRAMS AND FACILITIES . . . . . . . . .Department of Defense Technology Base Programs . . . . . . . . . . .Department of Defense Laboratories . . . . . . . . . . . . . . . . . . . . . . . . .Other Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

GETTING TECHNOLOGY INTO THE FIELD . . . . . . . . . . . . . . . .

. . . . . . . . . * 1 9. . . . . . .. . . . .. . . . . . .. . . . .. . . . . . .. . . . . . .. . . . . . .. . . . . . .. . . . . . .. . . . . . .. . . . . . .. . . . . . .. . . . . . .

● ✎ ✎ ✎ ✎ ✎ ✎ ✎ ✎ ✎ ✎ ✎ ✎

1

. . . . . . . . . . . . . 2

. . . . . . . . . . . . . 26

. . . . . . . . . . . . . 2

...*,.. . . . . . . 2

. . . . . . . . . . . 3

. . . . . . . . . . . . . 3

. . . . . . . . . . . . . 33

. . . . . . . . . 33

Technology Insertion . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Acquisition . . . . . . . . . . . . . . . . . .+ . . . . . . . . . . . . . . . . . . .Affordability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . .,.,

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .DUAL-USE INDUSTRIES .+ . . . . . . . . . . . . . . . . . . . . . . .Barriers Between Civilian and Military Industry . . . .International Competitiveness and the Health ofU.S,Internationalization of Industries . . . . . . . . . . . . . . . . . .Formulating

. . . . . . . . . .Industries. . . . . . . . . .

Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure

of Oversight .... . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . . . . . . . . .

Chapter 3

Summary

MANAGEMENT OF PROGRAMSAND FACILITIES

The system used by the Department of Defense(DoD) to run its technology base programs isdominated by two major characteristics that arepractically unique among large technology-basedorganizations. First, the system is inherently decentral-ized, with planning and management dominated bya bottom-up approach. Second, it relies heavily(although not exclusively) on a large, diverse groupof government owned and operated laboratoriesdevoted to defense research.

Planning of technology base programs is doneprimarily by the Army, Navy, Air Force, DARPA(Defense Advanced Research Projects Agency) andthe other defense agencies, and SDIO (the StrategicDefense Initiative Organization).* The Office of theSecretary of Defense (OSD) primarily serves as amonitor and data collector, deferring to these com-ponent organizations on matters of program direc-tion. OSD collects budget requests and passes themto Congress; after the funding is approved, thecomponent organizations run their own programs.Within OSD there is a hierarchy of oversight thatinhibits rational integration of programs: the Serv-ices report atone level, DARPA reports one level up,and SDIO reports only to the Secretary of Defense.While not unique in running its programs this way,DoD follows a minority path. Most organizationsexert much more top-down coordination and controlover planning and management of technology pro-grams.

The labs owned and run by DoD have two generalshortcomings. First, most are not strictly laborato-ries and lack the multidisciplinary pool of talentnecessary to be effective in developing a broad rangeof modem technology. Although they interact, theyare generally independent of each other. Developingtechnology is not the only (or even the primary)mission of most of these labs, but access to that

capability underlies the ability to perform othermissions. Second, the government-owned, government-operated (GOGO) management arrangement hascreated many problems that impair the ability of thelabs to function effectively. Other organizationsstructure their lab systems and lab managementdifferently.

Worldwide, there are three major trends in theplanning, management, and performance of technol-ogy development: top-down planning; centralizedmanagement; and collaboration. Moreover, amongthe governments of other industrialized nations thereis a movement away from concentration on defenseresearch and toward emphasizing civilian researchthat can be exploited for both economic and defensegains, as well as a movement away from governmentownership of laboratories.

Department of DefenseTechnology Base Programs

The Department of Defense does not have acentralized system for strategic planning of technol-ogy base programs. It has a federated system inwhich the central authority-the Office of the UnderSecretary of Defense for Acquisition-plays anadvisory and coordination role, but either lacks orfails to exercise the power to make major decisions.Those decisions are made by the component organi-zations—the Services, DARPA, and SDIO. Theplanning process is both top-down and bottom-up,but it is clearly dominated by the bottom-upapproach: most real decisions are made within thecomponent organizations. OSD provides generalguidance and reviews Service programs, but doesnot exercise any strong role in molding them.Attempts by OSD to mold programs (usually to keepto budget ceilings) are often viewed by the Servicesas uninformed, capricious, and arbitrary. This ar-rangement generally results in OSD not being ableto guide or coordinate the technology base pro-grams. However, OSD has in the past provided

IIn fiscat yew 1989 fie three Services together will spend 40,2% of the technology base funding (6.1 PIUS 6,2 phIS 6.3A). SDIO will spend 39.370;DARPA will get 13.8%; and the remaining 6.7% will be spent by the other defense agencies—the Defense Nuclear Agency, the Defense CommunicationAgency, the Defense Mapping Agency, the Defense Intelligence Agency, tie Defense Lmgistics Agency, and the National Security Agency. Among theagencies, DARPA occupies a special place because of its role as a source of R&D to complement Service programs. Efforts of the other agencies tendto be more specialized.

–19–

20 . Holding the Edge: Maintaining the Defense Technology Base

leadership for some special cross-Service programs,such as VHSIC, MMIC, SEI, and STARS.*

This system is not necessarily bad, but it seemsto be ineffective in producing a coherent technologybase program. Those who believe OSD ought toprovide strong leadership find the current systemdisappointing. To those who believe that OSD oughtnot to be controlling technology planning, it is theproper approach, even if OSD occasionally weighsin too heavily and disrupts programs. They believethat the users of technology—the Services-oughtto plan and control its development, that giving toomuch power to OSD risks losing Service support fortechnology base programs, and that the Services arebetter able than OSD staff to preserve technologybase funding.

Central planning and central management are twoseparate but related issues. Without top-down plan-ning a program lacks, as DoD’s currently does, abroad consistency of purpose and coordination toensure that important areas are not left unaddressed,and that healthy competition among competingdevelopments does not become wasteful duplica-tion. Central management can help ensure that theresults of central planning are carried out, but it canalso result in control of programs by those least ableto understand them.

Organizationally, the problems arise from twosources. First, OSD lacks either the ability or the willto exercise power over the Services. And second,there is no one individual or office that serves as afocal point and coordination center for the technol-ogy base programs of all the component organiza-tions. This results in diffusing the power to plan andcoordinate, and precludes establishing a high-leveladvocate for technology base programs who is freeof competing interests. The Goldwater-Nichols reor-ganization changed the players and their titles, butdid not correct these basic problems.

Within OSD, all technology base programs withthe exception of SDI are the responsibility of the

Undersecretary of Defense for Acquisition. This isshown schematically in figure 1. But the technologybase is only one small part of what he is responsiblefor—he also oversees the rest of research, develop-ment, test, and evaluation (RDT&E) as well as all ofprocurement. DARPA reports directly to the Direc-tor of Defense Research and Engineering (DDR&E)for oversight, but oversight for the Army, Navy, andAir Force programs rests one level farther down thechain with the Deputy DDR&E for Research andAdvanced Technology. The DDDR&E(R&AT) isthe highest ranking official with responsibility onlyfor technology base programs, but he only hasresponsibility for less than half the technology base.Thus, the Service programs are coordinated at theDDDR&E(R&AT) level, but they are coordinatedwith DARPA’s program one level higher up thechain, and balanced with SDI only at the highestlevel. This produces a hierarchy of influence amongthese component organizations and a mismatch thatmakes it difficult to balance their demands.3 More-over, no one with the power to oversee the entiretechnology base program can be an advocate for itunencumbered by other, possibly conflicting, re-sponsibilities.

Overall goals for technology base programs aresupposed to be specified in the annual DefenseGuidance document. But in reality, the DefenseGuidance devotes little space to the technology base,providing only very general guidance that can beused to justify just about anything the Services,DARPA, and SDIO want to do. The result is thatthese component organizations plan more or lessindependently, based on internally generated crite-ria, and link their plans to the general language of theDefense Guidance. The OSD review of Serviceplans is predominantly a data-gathering exercisewith little real power exerted from OSD. And realcoordination is hampered because DARPA andSDIO programs (which together account for overhalf of the funding) are considered only at higherlevels. Thus the Services and agencies dominate theplanning process.4

2vew High Speed Integrated Circuits; Monoli~ic Microwave Integrated Circuits; Software Engineering Institute; Software Wchnology forAdaptable, Reliable Systems.

3M~ufacturing techn~]ogy Progms, vi~ to ensuring producibility of items, are accorded a generally lower level of oversight ~d advocacy ~anproduct technology programs.

4’Top level Plming is typic~ly not done Witiin the services either; ideas come up from lower levels. However, in recent years the Services have b~nconducting high level studies of their future technology needs: Air Force Forecast 11; Navy 21; and Strategic lkchnology for the Army. The Air Forcehad been planning some of its technology base program around the results of Forecast 11.

Chapter Summary ● 2 1

Figure 1—A Hierarchy of OversightOff Ice of Tech no I o g y

t h e B a s e

S e c r e t a r y o f D e f e n s e Pro grams

Secret a r y St r ate g i c

D e f e n s eD e f e n s e I n i t i a t i v e

O r g a n i z a t i o n

. . .

Director ,D e f e n s e

Defense ResearchA d v a n c e d R e s e a r c h

a n d E n g i n e e r i n gP r o j e c t s A g e n c y

D e p u t y D i r e c t o r ,

D e f e n s e R e s e a r c h

a n d E n g i n e e r i n gf o r

R e s e a r c h a n d

Advanced Tech no I o g y

SOURCE: Office of Technology Assessment, 1989.

A r m y N a v y Ai r Force

It is not the case that the Services do not talk toeach other or to DARPA or SDIO. There isconsiderable coordination among projects havingsimilar technical foci, but this occurs at the projectlevel and not at the overall program level. There ismuch technical interchange but little programmaticcoordination. OSD could exert strong influence atthis level through its technology reviews, but it onlyconducts a few such reviews each year.

Because no single individual or office hasresponsibility for all technology base activitiesand only for the technology base, it is difficult tohave a strong and consistent advocate for tech-nology base both within the DoD bureaucracyand in relations with Congress and the Office ofManagement and Budget (OMB). (This problem ismirrored within the Services with similar results.)Nevertheless, OSD personnel spend a large part oftheir time defending technology base programs oranswering congressional mail, leaving little timeavailable to evaluate technology base programs. It is

not surprising, therefore, that OSD and the Servicesdo not have a systematic DoD-wide approach toevaluating technology base activities. Evaluatinglast year’s programs is a key to planning next year’s.If OSD personnel do not have the time to evaluatelast year’s programs, they lack a solid basis on whichto judge Service plans for next year.

The structure of the bureaucracy is not the onlycontributor. The relationships among institutionswithin DoD also play a major role. The Services andDARPA have traditionally had the upper hand withOSD. SDIO was designed to be able to proceedwithout interference from OSD or the Services.Typically, this sort of “pecking order” will persistin the absence of positive actions to change it.

Personnel is another factor. Although OTA hasencountered OSD staff who are competent, dedi-cated, and overworked, there is a consensus amongexperts that, like the labs, OSD suffers fromrestrictions that limit its ability to get and keep thebest people. While experts are divided as to how to


solve the problem, most agree that paying more anddecreasing career restrictions would help. Somebelieve that the problems would be best solved byvesting power in a professional staff with longtenure, removing it from the hands of politicalappointees and other “short timers. ” Others thinkthat only a constant infusion of “new blood” willhelp: rearranging the system so that very capablemanagers could take such jobs for a fixed term (e.g.,4 years) and then return to industry.

Department of Defense Laboratories

Reports on the shortcomings of DoD laboratoriesgo back at least 30 years. The mind-numbing arrayof specific issues that these earlier reports haveraised can be captured by two fundamental ques-tions:

. Does the DoD have the type and quality oflaboratories it needs?

. Are the management arrangements under whichthese laboratories are run inhibiting their abilityto perform as needed?

Type and Quality of Laboratories

To be precise, DoD has no laboratories. TheArmy, Navy, and Air Force departments own andoperate a large number of research, development,and engineering (RD&E) centers, none of which arelaboratories in the pure sense, i.e., institutions solelyfor conducting research. These centers perform avariety of functions ranging from research throughfull-scale development to occasional limited-scalemanufacture of military equipment items. The mixof activities varies from center to center, withsome—such as the Naval Research Laboratory andthe Army’s Harry Diamond Laboratory-beingmore heavily oriented toward research than others.As a shorthand, the term “defense laboratories” isused to refer to these government owned andoperated RD&E centers.5

The structures of the defense laboratories-howbig they are, what kind of work they do, etc.—haveevolved historically, based in part on the differentprocurement systems of the three Services and theroles each has seen for its laboratories. These

structures are quite different among the labs. How-ever, the management arrangements and modes ofoperation—which are similar across all of them—are a consequence primarily of law and also of DoDand Service regulations.

Comparing the defense laboratories to othergovernment R&D institutions is difficult becauseDoD’s role as a large purchasing agency makes italmost unique within the government. NASA isperhaps the closest analog because it too purchasesproducts of technology, but it also builds things andconducts research and space exploration. The na-tional laboratories that support the Department ofEnergy (DOE) build nuclear weapons and pursue abroad base of research for the furtherance of science.Industry, which also runs laboratories, ultimatelybuilds things.

Comparing DoD labs among themselves is alsodifficult because no two are really alike. They differin three distinct dimensions: the subject areas theyfocus on; the mix among categories of work (6.1,6.2, etc.); and the weighting of their missions amonga number of basic tasks. In addition to conductingresearch and development, these tasks include:

●

●

●

●

●

●

buying R&D from contractors and monitoringthe contracts;advising program offices on responding toproposals from industry to do development andproduction work (i.e., acting as “smart buyers”of technology);providing a base of technical expertise andknow-how that can be drawn upon to solveproblems as they arise or to follow new areas oftechnology;training young officers in science and engineer-ing;solving technology-based problems (or equipment-based problems) encountered by field com-mands; anddesigning and producing very small numbers ofspecial purpose items ‘needed by field com-mands.

They also differ in size, source of funding, and theorientations and “cultures” of the organizationsthey primarily work for.

5DOD is ~so supported by con~actor owned and operated laboratories such as the John Hopkins University Applied Physics Lab d the MIT LincolnLab, and by national laboratories operated by contractors for DOE. For more information on the institutions that support the defense technology basesee: The Defense Technology Base: Inrmduction and Overview, OTA-ISC-374 (Washington, DC: U.S. Government Printing Office, March 1988).

Chapter 3-Summary ● 23

All of these differences make objective evalu-ations and comparative ratings of these institutionsvery difficult to perform. Most evaluations andcomparisons appear to be subjective ones, evenwhen performed by highly qualified individuals. Forexample, Service labs are frequently criticized fornot doing top-flight science, especially when com-pared to national laboratories or major universitylaboratories; but performing scientific research isnot the major mission of these facilities.

Nevertheless, there is a common thread among allthe tasks the labs perform: they all require thelaboratory to be a center of technical expertise. Mostdon’t require the staff to be conducting research andcontributing to the advancement of science ortechnology, but all benefit from a staff that hashands-on expertise: a staff member who is contribut-ing to the leading edge is closer to it than one whois simply reading about it, and is more likely to geta seat at the table when the real experts meet.

There are three basic approaches to providing theresearch core of an R&D facility. The first is to builda large, diverse, multiprogram laboratory with a staffthat does research in a broad range of disciplines.The DOE national laboratories fit this description, asdo the corporate research centers of several verylarge corporations such as IBM, AT&T, and GeneralMotors. These labs push forward the frontiers,provide a large pool of talent that can be directed andredirected to solve problems or follow new areas oftechnology, and provide a base of knowledge fromwhich other labs can draw for more narrow applica-tions. Staffs typically number well over 1,000 andare heavily weighted toward advanced degrees. TheNaval Research Lab is the only DoD lab that fits thismold.

A second approach is to build labs with staffs ofa few hundred that concentrate their efforts in one ora few areas. Several Army laboratories-NightVision, Harry Diamond, Electronic Technology &Devices, etc.—are structured this way. These facili-ties can have programs that are at least as good asthose of the multiprogram laboratories in a fewselected areas. However, this focus is bought at thecost of loss of breadth and flexibility to respond toa broad range of problems. Moreover, as modemtechnology becomes more complex, even a singlearea of concentration can rest on a broad base of

underlying disciplines. Size can constrain theselaboratories from effectively pursuing their fewareas of concentration and from shifting their focus.This problem of lack of critical mass is even morepronounced in the third type, the model followed bymost Defense labs: a medium to large RD&E centerwith small cells of expertise embedded in it. Theselabs do not have in-house research as a focus, but asa supporting function. Hence the cells of expertise,however skillful and productive, tend to be narrowand thin: in some cases the departure of one or twokey individuals could destroy that expertise.

In detail, the Army, Navy, and Air Force run theirRD&E centers differently. But in general they allfunction the same way: technology generated in-house, in other Service labs, externally undercontract, and any other place the staff has access to,is assimilated with the aim of transitioning it into theprocurement system. The accumulated base ofknowledge is used to advise the procurement offi-cers regarding the technical qualities of variousproposals to develop and build systems.

The central question is whether this systemhas been, and is really capable of, delivering thegoods. Does the technology transit into and out ofRD&E centers, and are the staffs up to the job? Thisis a very complex question requiring an intensiveinvestigation, but it is absolutely key.

If the answer is “yes,” Congress ought to stopworrying about the labs and let them get on withtheir work. Steps might be taken to make their jobseasier by easing management burdens. However,even if the labs are judged to be doing a good job,budget constraints may make it necessary to con-sider restructuring.

If the answer is “no,” there area number of stepsthat might be taken to fix things. These range fromtaking steps to ease management problems (whichwill be discussed below) to drastic reorganization ofthe entire system. Some involve centralizing, con-solidating, closing, and moving institutions. How-ever, such steps have far-reaching consequencesand can be nearly irreversible. They ought to betaken only after much deliberation. One approachwould have each RD&E center include or be closelyassociated with a large multipurpose laboratory, thesmall cells of expertise being replaced by a large,diverse pool of technical talent. Clearly, doing so for

95-677 - 89 - 2


each RD&E center would be prohibitively expen-sive. An alternative approach would be to provideeach Service, or DoD as a whole, with a centralcorporate lab and tie the RD&E centers closely to it.The Naval Research Lab might be a model. Smallerlabs of more limited scope are a second choice, butbecause they are inherently less flexible than multi-program labs, arrangements would have to be madeeither to shift their focus or close them down as theareas of technological interest shift. As an alterna-tive to building up the research bases within theServices, greater use might be made of DOE nationallabs as technology bases for the Services. Consolidat-ing facilities either within each Service or acrossService lines under OSD could offset the cost ofexpanding the underlying labs. But this runs the riskof cutting the links of the RD&E centers to theirparent buying commands and further restricting thetransition of technology into the procurement sys-tem. Unless handled carefully, it could also sever thevery important links of the labs to the field com-mands.

Management Structure of Laboratories

The problems that plague the Services’ government-owned, government-operated laboratories (GOGOs)and the causes thereof have been extensively docu-mented. They are inherent in the laws and regula-tions that govern the operations of these labs. Whilethese laws and regulations have not changed greatlyover decades, the trend within the last few years hasbeen for their application to become more onerous,making the government labs less attractive places towork at a time when the market for technical talenthas become much more competitive.

The difficulties fall into three related categories:problems in recruitment and retention; difficulties inconducting the day-to-day business transactionsnecessary to get the work done; and long delays inupdating buildings and major equipment. The lattertwo are problems in their own right as well ascontributors to personnel difficulties. Effective man-agement is also impeded by funding that is oftenunpredictable and fluctuates from year to year.

Even premier laboratories, like the Naval Re-search Lab, are having difficulty attracting the bestand the brightest. Many of the RD&E centers haveall but given up trying: they now recruit from a small

circle of mostly local schools and hope to “grow”their own in-house expertise. OTA’s observationssupport the points made in earlier studies:

. most of the labs have difficulty hiring andretaining highly qualified personnel;

. the government is at a major disadvantage incompeting with industry and academia; and

● the system makes it difficult to reward goodperformers, penalize the poor performers, or tiesalary closely to performance.

The “NOSC/China Lake Experiment,” in whichthe Navy loosened the salary structure for scientistsand engineers at the Naval Ocean Systems Centerand the Naval Weapons Center at China Lake,helped with recruiting and retention in the entry andmidlevels. Similar novel approaches including sal-ary structure and educational opportunities are underconsideration by the Services. But since these do notraise the ceiling on salaries; they do little to solve theproblem of attracting and retaining key seniorpeople. Losing senior researchers is a double liabil-ity: exceptional senior people do exceptional work,and they also attract younger people, many of whomwill accept otherwise less attractive work conditionsin order to work with someone special.

Interesting work helps to attract and retainpeople. Good people stay if the work is challenging,if discretionary funding is available to allow them to“follow their noses,” and if they have an opportu-nity to pursue a technical career without beingsidetracked into management. But increasingly,technical people in Service labs can only get aheadif they become managers, and in those managementjobs they spend an increasing amount of their timein administrative tasks and insulating their bench-level people from bureaucratic “paperwork” im-posed from above.

At most DoD labs the Tehnical Director haslittle or no control over the most important supportelements of his organization—the personnel office,the general counsel, the procurement office, etc., allof which report to parent commands. And construc-tion of new facilities is handled out of militaryconstruction (MILCON) accounts for which the labsusually fight a difficult, and often losing, competi-tion with a long list of other claimants. This resultsin obsolete facilities.

Chapter 3--Summary • 25

The Defense Science Board has recommendedchanging the laws and regulations that are causingthe problems, loosening up the system to enable theDefense labs to compete more effectively. Whilethis might be helpful, it would require a long list ofchanges in both legislation and government regula-tions. This involved agenda could be very difficultto complete. However, a congressional decision totreat the laboratories differently from other govern-ment offices might facilitate the changes.

An alternative would be to convert the labs togovernment-owned, contractor-operated (GOCO),or even to contractor owned and operated (COCO)facilities. This would seem an easy way out of themorass of government red tape. GOCOs do havegreater management flexibility in personnel man-agement, but the evidence for greater flexibility isambiguous in areas other than personnel. GOCOscan pay higher salaries, can hire and fire more easily,and have much greater flexibility in rewarding goodwork and shifting personnel. They also displaygreater flexibility in shifting the focus of their work,and have some advantages-although not so dramatic—over GOGOs in their ability to purchase equipmentand facilities.

DOE GOCOs appear to show a greater aggres-siveness in seeking out and developing technology.And, at least in the design and manufacture ofnuclear weapons, transition of technology intoapplications is more direct than it typically is inDoD. But this is not necessarily a consequence oftheir being GOCOs. Size, fill-spectrum stance, andresearch-oriented culture are all contributors. So isthe relationship that has evolved between DOE andits labs: the missions of the labs have been construedin a very broad way, facilitating changes in programdirections as technology evolves.

While there are some real advantages to convert-ing to contractor operation, there are some importantoffsetting factors. No government-funded institutioncan escape oversight merely by converting tocontractor operation. Funds derive from congres-sional appropriations, and Congress holds seniorofficials of sponsoring agencies accountable fortheir use. Thus, the tendency is for the government

to impose on its contractor laboratories many of thesame rules and regulations it lives under. Conse-quently, with time GOCO labs tend to become morelike government-operated laboratories. Governmentrules under which the sponsoring agencies operatetend to be passed down to the contractors, so theGOCOs are not free of the majority of governmentimpediments. Government policy appears to be thateven though government regulations do not apply toGOCOs, GOCO practices ought to be consistentwith them. OTA found that the perception of “redtape” and the burden of bureaucratic paperworkreaching down almost to the bench level was nodifferent at GOGOs, GOCOs, and COCOS.

Although there have been many studies ofgovernment labs since the 1962 Bell Report, nonehave questioned its finding that there are “certainfunctions which should under no circumstances becontracted out. The management and control of theFederal research and development effort must befirmly in the hands of full-time government officialsclearly responsible to the President and Congress."6

There are some functions that are inherently govern-mental: passing them off to contractors would raisemajor questions. For example, being a smart buyerand advising a program office on the technical meritsof proposals is probably not a responsibility thatought ultimately to be entrusted to a contractor,although today contractors are part of that process.

One advantage of government labs—and a majorfunction—is that they can respond immediately toproblems that emerge in the field. Staff can beordered to stop whatever they are doing and turntheir attention to the problem at hand. This would bemore difficult for contractors to do, unless thecontract had been carefully crafted to allow for thecontingency. 7 At several contractor operated facili-ties OTA was told that response times would have tobe measured in months, if not years.

While all DoD labs could benefit from fewerrestrictions, not all are equal candidates for conver-sion to GOCO status. Those that conduct in-houseR&D would be better candidates than those thatfunction primarily as “smart buyers.” Similarly,those that cannot solve their management problems

6Report t. the President on Govermnt co~ractingfor Research and Development, reprinted in W.R. Nelson (cd.). The politics ofsclence (New’York, NY: Oxford University Press, 1968), p. 200.

TFor ex~ple, level of effort support COntrXtS.


within the government system would be more likelycandidates for conversion than those whose manag-ers believe they can.

Other Approaches

After examining a number of approaches used byother organizations to manage technology programs,some basic themes emerge that may be applicable toDoD management of its technology base programsand laboratories. First, in most governments andcompanies, R&D policy is approved at the upperlevels of management and promulgated throughoutthe organization. Second, centralized control overresearch projects is the rule. It is supported byfrequent reviews and combined with a readiness tocut losses when projects do not pan out and to buytechnology outside the organization if that appearsto be a more economic approach. Third, both publicand private organizations are moving toward col-laboration as a means of affording research of themagnitude dictated by modem technology. Finally,on a broader note, the Europeans appear to bemoving toward the Japanese point of view thattechnology efforts ought to be focused on enhancingthe economy: a strong high-tech economy willproduce both more money available for defense and‘‘spin-on” of technology for defense purposes.

For at least two decades the Europeans have beenworried about their economic positions, particularlyrelative to the United States. But the emergence ofJapan and other Asian nations as economic powershas greatly intensified their concerns. This hasspurred efforts to integrate the European Commu-nity (EC), notably the movement to a “singleEurope” in 1992. Moreover, as their fears ofeconomic problems have increased, their anxietyover Soviet military power has receded. Hence themood is to reduce the drain on the economy ofdefense oriented R&D, while increasing substan-tially research oriented toward civilian products.The Europeans are looking for ways to make defenseR&D support civil production; defense labs areincreasingly viewed as national assets that can beused to help make civilian industries productive.The trend appears to be to do research and explora-tory development (the equivalent of 6.1 and 6.2)predominantly in civilian-oriented labs. Only in theadvanced development stage would the work take ona more military-oriented cast. The prevailing phi-

losophy appears to be that science and technologypolicy should be integrated whenever possible witheconomic and industrial policies. In this regard, theEuropeans are moving away from the U.S. modeland toward the Japanese model.

It is tempting to take the attitude that if our systemhas significant shortcomings we ought to adoptsomeone else’s. But this approach is fraught withperil. While there are important lessons to belearned-and these general themes appear to beworth considering-it is not necessarily true thatDoD can simply adopt some other system as its own.All organizations are different, and they do not allsee themselves as solving the same problems.Management approaches tend to be rooted in corpo-rate ‘‘culture” at least as much as they are the resultof dispassionate analysis. It is somewhat dangerousto adopt the attitude that what works for some otherorganization ought to work for DoD, For example,the sheer scope and size of DoD’s technology baseactivities dwarfs nearly every other organizationexamined, and might even rival the aggregate ofthem. Furthermore, it is not clear that other organiza-tions are significantly more successful than DoD isin developing and nurturing technology and using itto good effect. The success story everyone immed-iately turns to is Japan. But Americans are not, anddo not behave like, Japanese. And the Japanese seekto use technology somewhat differently than doesDoD.

Planning and Priorities

In contrast to DoD, in which a laissez-faireapproach and “bottom-up” planning predominates,most Western European governments set nationalcivil and military R&D objectives from the top.Working through central committees or advisorypanels, cabinet-level officials set priorities andensure that the goals are translated into specificprograms in either government or private laborato-ries. The technical experts are usually left free todetermine the composition of the specific programs,but they must be able to justify program relevance tohigher authorities. In addition, the European Com-munity is exerting an increasing top-down influenceon the member nations’ research programs. Exploit-ing allies’ work and avoiding duplication of effort isa growing theme. The Japanese approach is perhapsless formal, emphasizing government/industry con-


sensus building and the role of industry, but ulti-mately major decisions are made by a central body.

Industry generally follows a somewhat similarcentralized approach. Major corporations typicallyhave central procedures for establishing businessobjectives, including identifying the key technolo-gies that are expected to contribute. Once theseselections are made, the component companies arefree to decide how to pursue them. But corporateoversight typically remains continuous and close.

There has been strong criticism that U.S. defenseR&D focuses too much on the near term, both ingovernment and in the private sector. Europeancompanies are even more likely than U.S. companiesto spend their R&D money for near-term applica-tions. This trend has become more pronouncedrecently in both Europe and the United States astechnology base expenditures have declined as aproportion of defense spending. In contrast, how-ever, budgets for long-term research-particularlycivil research-are increasing for many Europeancountries and for the EC. This is tied to a perceivedlinkage between R&D, economic competitiveness,and prosperity. Governments are seeking to improvetheir industries’ competitive positions by makingcivil research the driver and blurring the distinctionbetween civil and military R&D. The Europeans’short-term focus and declining funding in defenseresearch appears to be offset by a longer term focusand more generous funding for civil research. InJapan, the government role is greatest in long-termdevelopments for which the risks are high and thepayoffs not evident.

Growing fear of Japanese and U.S. industrialcompetition has fostered European interest in large-scale, centrally directed technological initiatives.These have been largely multinational in nature,such as ESPRIT, EUREKA, RACE, and BRITE,8

although there have been single nation programssuch as the U.K. Alvey program. These are modeled,in part, after a succession of U.S. initiatives—beginning with the Manhattan project—that, whilenot always successful, propelled technology for-ward. Large collaborative efforts are also employedby the Japanese, but their efforts tend to have moreindustry funding and less government money.

A similar approach currently in favor in Europeand to an increasing extent within U.S. industry is toemploy special research teams, or “centers ofexcellence,” often in collaboration with universitiesor potential competitors. These groups concentrateon technologies where a large critical mass ofpersonnel and other resources, or interdisciplinaryresearch, is considered essential. U.S. examples areSEMATECH, the Electric Power Research Institute,Semiconductor Research Cooperative, and the Mi-croelectronics and Computer Technology Corp.

Management and Control

European governments not only plan their R&Dprograms centrally, they also manage the executionof those programs centrally. Large companies alsotend to keep tight central control. In both cases, thetrend is also toward centralized control of laborato-ries in an attempt to establish the optimum balancebetween generic research and product-oriented (ormission-oriented) research.

DoD’s laboratory system is basically mission-oriented, with most laboratories dedicated to spe-cific warfare specialties. Mission focus provides acloser link between technology and military applica-tions, but it also encourages duplication in facilities,resources, and projects. European labs and programsare increasingly organized along technology, notmission lines. In France, Germany, and the UnitedKingdom the defense research activites are planned,organized, and managed by central authorities inde-pendent of service requirements and developmentactivities. Centrally managed civil research pro-grams are generally oriented around generic tech-nologies. Similarly, EC programs are directed to-ward enabling technologies, with applications left toindustry.

DoD’s extensive network of government ownedand operated laboratories is unique among Westerndefense establishments. With the exception of theUnited Kingdom, European governments own few,if any defense labs, and the British are in the midstof drastically consolidating their laboratory system.However, there are many more European government-owned and government-sponsored laboratories doingcivil research.

8EWoWm Strategic ~o~m for Re~mch in Information Technology (ESPRIT); European Research Coordinating Agency (EUREKA); Researchand Development in Advanced Communications for Europe (RACE); Basic Research into Industry Tmhnology for Europe (BRITE).


Industry is generally moving in two directions.Most R&D is being moved out to the componentcompanies. Some corporate research centers arebeing pruned back or even closed. As money getstight, it is easy to view corporate research centers asexpensive luxuries-’’money Sinks ’’-racier thanas investments. But at the same time corporations areestablishing corporate level centers of excellence inkey technologies (or forming collaborative efforts inthem). Technology is transferred to the productdivisions, at least in part, by assigning personnelfrom the product divisions to temporary jobs in thecentral facilities and then moving them back to useand disseminate the technology they studied andhelped develop. Industry is also moving in thedirection of collaborative research, sharing theescalating costs of modem technology. This re-search is of necessity technology oriented, notmission oriented.

Collaboration, Coordination, andTechnology Transfer

Collaboration in research is now a way of life.High costs and worldwide competitive pressures areforcing governments and industries to pool theirresources. Collaborative projects play a central rolein Japanese R&D. European governments and in-dustries explored cooperative research in the 1970sand early 1980s, but in the mid 1980s growingconcern that they were falling behind the UnitedStates and Japan led to a series of serious collabora-tive measures. Moreover, the European members ofNATO, after more than 20 years of ad hoc collabora-tion on defense and other aerospace projects, arenow working on establishing a coherent, systematicprogram of collaboration, Breaking down the long-standing barriers that have isolated European com-panies from each other and fragmented markets is anexplicit objective of recent high-technology collabo-rative initiatives. In addition, European companiessee that they each have to draw on a broader base oftechnology than was necessary in the past. Recogni-tion that Germany’s strong position in world trade isdue, at least in part, to a collegial, collaborativerelationship between industry, academia, and gov-ernment also helped spur interest in collaboration.

U.S. companies are not only engaging in collabo-rative programs at home, they are also joining with

European (and Japanese) companies in variousventures.

Applications: Transitioning TechnologyFrom Lab to Products

DoD has been criticized both for leaving technol-ogy in the lab too long, resulting in obsoleteweapons, and for rushing it prematurely into produc-tion—which creates unreliable products. Neitherallegation is without foundation. Technology transi-tion is one of the most difficult problems ofdevelopment. European governments and industriesappear to be no better at technology transition thanDoD is. Japan appears to have a unique success attransitioning technology rapidly and effectivelyfrom the lab into production. The Europeans appearto be studying and beginning to apply the Japaneseexperience. Teams of researchers, designers, engi-neers, manufacturing specialists, and even marketersare being brought together early in the life of aproduct in order to perform in parallel what usuallygets done sequentially. The parallel development ofprocess (manufacturing) technology and producttechnology is considered a particularly importantfactor.

Examples of the close relationship that is essentialbetween research staff and those who developspecifications exist in all successful companies; butin large and diverse government organizations theliaison and communication that is required may bejeopardized by interdepartmental rivalries and paro-chialism which only strong management and direc-tion can dispel. In DoD, requirements for newsystems are set by the Service buying commands,and development is done by industry. These areobliged by law to stay at arms length; the govern-ment labs provide the primary link between them—and the labs are not always successful.

GETTING TECHNOLOGYINTO THE FIELD

Government officials and others have expressedconcern and frustration over the age of technology infielded U.S. systems, particularly those just begin-ning to roll off assembly lines. Comparisons usuallytake two forms. First, government and industryresearchers have laboratory developments that areclearly superior to what is going into the field.

Chapter .3-Summary ● 29

Second, dual-use technology in defense systemsoften lags significantly behind what is available inthe consumer markets, and by the time a system hasbeen in the field for 5 to 10 years it can seemoutdated compared to what Ford or Radio Shack isselling.

Technology in production will always lag behindtechnology in the lab. Taking developments off thebench, engineering them into real systems, andgetting those systems into production is a time-consuming process for military and civilian manu-facturers, as well as for movie producers, thinktanks, book publishers, and many other enterprises.Indeed, very little legislation moves instantaneouslyfrom brain storm to law. Major military systems aregenerally much more complicated than civilianproducts, and hence the product cycles are muchlonger.9 In addition, the process of getting approvalto begin a military project is generally considerablylonger than the equivalent process in the consumersector. Furthermore, military systems have longlives, and dealing with frequent updates is a logisti-cal nightmare, so it is not surprising that changesoccur much less frequently than the typical yearlychanges in consumer products. It appears to makesense to change the current model Toyota because ofa relatively small change in engine technology.(Indeed, it helps sales to tell consumers that thisyear’s model is “all new” and “innovative,” andtechnology is often changed just to enhance market-ing.) But it makes absolutely no sense to rebuild theentire fleet of tanks every year to take similarchanges into account. The problems of maintainingdifferent equipment types in the field mean thatdecisions to update part of the total inventory, whileoften made, are not taken lightly. Finally, DoD is notin the business of developing and fielding technol-ogy for technology’s sake; its job is to get bettercapabilities into the field in a reasonable time at areasonable cost. Up to a point, it is not unreasonableto argue that new technology ought to buy its wayonto a system.

Military-specific technology is usually the pacingtechnology for entire systems, determining theschedule for getting the system into the field andcontrolling the rate at which the dual-use technolo-gies in the system get fielded. The entire acquisition

system is geared to the pace set by these militarytechnologies. It is often the case that after a systemdesign is frozen, the commercial counterparts oftechnology embedded in it continue to move for-ward, sometimes dramatically, resulting in severalgenerations of products before the military system isproduced. This produces military systems that arenot as advanced as some commercial products; butif responding to rapid changes in dual-use technol-ogy were to prevent freezing the design of a systemlong enough to get it into production, none of thetechnology would ever get produced.

Thus, while it can be misleading to comparefielded military technology to laboratory technologyor selected consumer technology, it is important toask whether new technology can get more quicklyand more effectively into the field. (It is alsolegitimate to ask why the military cannot have thesame products—like radios, CRT displays, trucks,and clothing—that consumers can go out and buy.)

The problem of getting new technology into thefield is not that the United States is unable to developnew technologies with military relevance. It is rathera problem of the transition of that technology intoengineering, the time needed to begin manufacture,and the rate at which new systems are built. It can beimproved in three general areas: improving theinsertion of new technology into acquisition pro-grams (i.e., the transition from technology base toengineering and production); improving the acqui-sition process that engineers and produces systems;and improving the affordability of systems so thatthey can be bought more rapidly.

Technology Insertion

The technology development and system acquisi-tion processes are largely (but certainly not com-pletely) separate. Technology base work takes placein a variety of institutions, including some compa-nies that ultimately build systems. Engineering andproduction are done in private companies (notalways the same ones that did the technology basework) under the supervision of DoD programmanagers. This causes a major bottleneck at thepoint at which technology moves from technologybase to acquisition. Several mechanisms exist tobridge this gap: general technical interchanges

gIn commerci~ products, complexity is usually the enemy, something to be managed caefully..


between Service lab people and industry; IR&D andcontract research that involve some companies in adevelopment; involvement of lab people with theprogram offices (part of the “smart buyer role”);and formalized Service transition programs.

Many studies of the transition issue seem to agreethat nonsystem-specific prototyping, pursued with6.3A funds, presents the greatest opportunity toimprove technology insertion. It has the potential tosolve two problems. By demonstrating feasibility,these advanced technology demonstrations helpreduce the high risk carried by some technicaldevelopments. And they help correct overoptimismby demonstrating the limitations in the current stateof the art. Overoptimism leads to promising toomuch, which in turn leads to disappointing systemsand to lengthy and costly redesign efforts. The newemphasis within DARPA on prototyping is appar-ently an attempt to ease the transition into systemdesign of technology developed under DARPAprograms. DARPA has always been the focus oftechnology that does not fit neatly into what theServices want to do. However, if the Services do nottake DARPA seriously, it is not at all clear thatDARPA’s prototyping effort will have any use.

AcquisitionIn searching for the causes of delays, the acquisi-

tion process has been the primary candidate, Evenwhen the system is working smoothly it seems totake a long time to move programs through; but itusually is not working smoothly. And when it bogsdown, delays lead to further delays through escalat-ing costs, compensatory stretch outs, and time-consuming attempts to fix any particular program’sspecific problems. While the consensus is that thesystem is in trouble, it has weathered study afterstudy without apparent improvement,

Several studies have found that acquisition (ad-vanced development, full-scale development, andproduction) takes longer than it used to. But the dataare not all that clear: there is certainly no obvioustrend toward rapidly increasing times. It does takelonger in the 1980s than it did in the 1950s or 1960s,but there is not enough data to discern clear trendsover the past decade. Studies of fighter aircraftprocurement, the most-studied system type, con-clude that whatever increases have occurred are in

the front-end decision process and in production, notin full-scale engineering development. Data on othersystems are less conclusive.

It is generally held that commercial industrycompletes programs more rapidly than does DoD,but there are significant differences between govern-ment and industry that make it possible for industryto avoid many of the basic problems that plagueDoD acquisition. These basic problems are “builtinto the system,” they are consequences of thecharacteristics of U.S. Government. For example,canceling a program that has grown too much in costor schedule to be profitable is easier than cancelingone that, despite schedule slippages and cost over-runs, is judged essential for national security.

But enhancing national security is not theNation’s only goal. Goals like fairness, environ-mental protection, equal opportunity, jobs, andcompetition all figure into how both Congress andthe Administration judge defense procurement pro-grams. DoD itself has goals it must pursue inaddition to managing programs efficiently: main-taining the defense industrial base, ensuring that themost efficient producer does not drive the others outof business (contrary to what industry would do),etc. Government is not solely concerned that aprogram provide the best capability at the lowestcost most quickly. Moreover, the political process inboth branches of government-the tug and pull overresources and goals-introduces uncertainty intoprograms, even when Congress and high-levelexecutive offices do not micromanage programs.

The structure of the DoD acquisition system ismuch more cumbersome than that of private sectorcompanies. That structure is, in part, determined bygovernment’s size and unique role. DoD programmanagers are accountable to five or six layers ofbureaucracy up to the Secretary of Defense. Theselayers typically have extensive horizontal structure,so the program manager (PM) has to satisfy a largenumber of people, many of whom have power overhis or her program but no responsibility for it. Tocomplicate matters further, the PM reports up onechain for oversight of the program, and up anotherfor the planning, programming, and budgetingsystem which is responsible for determining thefunding for the program. But this involvement of theOSD bureaucracy, as well as that of OMB and

Chapter .%-Summary . 31

Congress, is part of the checks and balances on theexpenditures of billions of dollars.

While industry shares many of DoD’s problems,it has a very strong incentive to manage success-fully: failure could mean bankruptcy. In manyinstances industry works under a simpler systeminvolving a direct link between the program managerand a high company official having the authority tomake decisions, settle disputes, and insulate the PMfrom external pressures. The PM has responsibilityfor the program: if it fails it is his fault and his jobmay be at stake. The DoD PM typically has to obtainseveral levels of approval for any action; there aremany people who, in trying to ensure that the PMdoes not fail spectacularly, will also prevent him orher from succeeding spectacularly.

Several factors are major contributors to delays inprograms: the sequential processes of requirementsgeneration, resource allocation, and system selec-tion; program variability (or instability) caused bymany players making changes; bureaucratic paraly-sis; inappropriate organization for defense procure-ment; and the quality and incentive structure forprocurement personnel. Underlying these are thebasic structure of the government, the nature of thebureaucracy, the organization of the DoD procure-ment system, and the conservative risk-averse natureof government organizations.

Requirements generation and resource allocationinvolve the Services, OSD, OMB, and ultimatelyCongress. They are highly political, which oftenleads to overpromising in order to get programapproval. Overpromising leads to cost growth andschedule slippage. But the system makes it easier toreadjust the program to these realities rather than togo back and question the requirements that producedthem in the first place.

Constant changes in defense acquisition pro-grams are commonplace, leading to cost increasesand schedule slippages. Variability results from therequirements process, the risks inherent in newtechnology, the political/budgetary process, andpersonnel turnover. While the disruptions caused bythese factors can be somewhat controlled, theunderlying causes cannot be eliminated.

Baselining-a form of contract between programmanagers and their Services-was developed to

limit changes in programs. But making baseliningwork requires giving program managers more author-ity over their programs than they now have. Neitherprogram managers nor Services can control budgetsor other changes and conditions imposed by OSD,OMB, and Congress. Moreover, external factors thataffect a program-like threat, doctrine, and resources—will cause changes in the program no matter howwell it is managed.

However, Congress, OSD, and OMB can decideto limit their direct involvement in a program (orCongress can decide for the others). But, at least inthe case of Congress, this would involve giving uppower which it jealously guards. Congress hasalready agreed, in principle, to relax oversight for afew major acquisition programs, which would re-quire reauthorization only at significant milestonesrather than annually. As yet, none of these milestoneauthorizations have been submitted to, or approvedby, Congress. Not all members are likely to agreethat efficient functioning of defense acquisitionprograms is more important than other issues theyare concerned with, including the (possibly shifting)interests of their constituents. The budget processspecified by the 1974 Budget and ImpoundmentControl Act and public Law 99-177 (Gramm-Rudman-Hollings) increases Congress’ incentives to keepcontrol of as many budget items as possible so thatit can engineer the budget levels it agrees to.

Perhaps the most discussed problem is the bureau-cratic burden individuals and companies must strug-gle through in order to do their jobs. A 1977 DefenseScience Board (DSB) panel concluded that increasesin acquisition times are all bureaucratic: ‘it does nottake any longer to do something, it just takes longerto obtain the necessary approvals and acquirefunding . . . .“ The program manager’s job hasbecome increasingly complicated, accompanied bylengthening time to complete contracting actionsand increased regulation, oversight, and auditing ofcontractors. The overall perception is that of increas-ing regulatory and bureaucratic burden, but studieshave found the picture to be unclear. While someindicators of burden have been clearly increasing,others have remained the same or declined. More-over, measuring the effects of regulatory and bureau-cratic activity is even more difficult than measuringthe activity itself. For example, estimates of the


added costs due to regulations and bureaucracyrange from 5 to 200 percent!

This “red tape” is unambiguously greater ingovernment than it is in private industry. What inindustry can be a straightforward, one-step, projectinitiation process involving the manager and a highcorporate officer is in DoD a two-step processinvolving the PM, a committee within DoD,10 andCongress. Both the DSB and the Packard Commis-sion recommended bringing the system closer to anindustrial model in this regard, and the Goldwater-Nichols Act tried to implement that.

Since the bureaucratic burden arises in part fromgovernment attempts to have programs satisfy goalsother than getting the job done most efficiently,solutions can be of two types: those that try tostreamline the system without changing its mix ofgoals; and those that seek to change the balanceamong goals, particularly the balance betweenhaving an efficient and successful program andsatisfying all the other government goals. Onesuggested solution is to review all the regulations todetermine whether each is still necessary andwhether the aggregate could be streamlined some-how, a daunting task in its own right. Anothersuggestion is to shift the burden of proof from thePM to those who would slow down the project,making the PM innocent until proven guilty. Forexample, a competition advocate would have toshow that the program was insufficiently competi-tive or that taking measures to enhance competitionwas important enough under the circumstances towarrant tampering with the program. But somehigher authority would have to be responsible forbalancing these claims against the interests of thePM who would always be served by ignoring them.

Some DoD programs do better than most:“black” programs (so it is said), and other specialhigh-priority programs. This success is due in part totheir high-priority which affords them high-levelattention. Clearly, all DoD programs could not betreated that way or the system would overload.These programs also get special exemptions from

various regulations, Granting similar exemptions toall programs would nullify the regulations, defeatingthe purposes for which they exist.

There has been widespread concern about theprocess that produces PMs and their chief assistants.These people are either military officers or civilservants. In 1986 the average tenure of PMs wasabout 2 years. This makes it difficult to give themreal power over programs that run many times thatlong, and creates incentives for them to sacrificelong-term performance in order to look good on theirwatch. The military personnel usually, but notalways, rotate rapidly in and out of the jobs in 2 to3 years. They do not always have prior experience orrelevant training. Many of the civil servants do notrotate, and ‘‘remain for so long that they resistinnovation and change. ” 11

Affordability

One of the major contributors to delays in gettingnew technology into the field is the cost of modemdevelopment and procurement programs and theresultant program stretchouts and low buy rates.Almost all important systems cost enough to getclose scrutiny by OSD and Congress. The battles arefought each year. The result is often that the fundingrequested by the program is reduced (in some casesdramatically), which slows the development paceand slips the date at which production is initiated.

Once the program is in production, DoD’s ten-dency is to reduce the funding below what had beenprojected, in order to keep as many programs aliveas possible. This leads to buying fewer of anyparticular item per year, which has two majorconsequences. First, obviously the slower the rate atwhich a system is bought the longer it will take to getthe capability into the field. It may not delay InitialOperating Capability, but it will certainly delay thedate at which a significant capability is fielded.Second, providing insufficient funds to procure atplanned rates raises the unit costs, which furtherdecreases the number that can be bought per year.

l~he Defense Acquisition Board, and perhaps ofiers.t ]J. Ron~d Fox ad J~e~ L, Field, The Defe~e~a~ge~ntc~//en8e: weap~nsAcquisition (Boston, MA: H~~d Business School PRSS, 1988),

p. 312.


DUAL-USE INDUSTRIES

Most of the technology that is engineered intodefense systems is still developed in the “defenseworld” of DoD’s laboratories and contractors. Thisis particularly true of the exotic technologies that arethe centerpieces of advanced designs. But increas-ingly, building those systems depends on develop-ments that take place in the civilian sector, a civiliansector that is driven by the international market-place. This was dramatically illustrated by eventsduring the first week of November 1988. A companycalled Avtex, which manufactured rayon fibers forthe apparel industry, announced that it was shuttingits doors in response to foreign competition in theclothing business. This sent shock-waves throughDoD and NASA when it was discovered that Avtexwas the only producer of fibers that were critical tothe production of missiles and rockets. While othersources could be qualified, and other fibers might befound to substitute for the rayon that Avtex made,that process would take longer than the period oftime the available supply of rayon would supportproduction. Negotiations were soon completed tokeep Avtex open.

High-technology industries are becoming in-creasingly internationalized: foreign companies andmultinationals are technology drivers. Large inter-national markets generate huge amounts of capitalthat fuel research and development into new prod-ucts and underlying technologies. The defensecomponents of these markets are often small, givingDoD little or no leverage over the directionsdevelopments will take. DoD has to choose betweenplaying a follower role, or spending large amountsof money to keep a competitive leading edgecapability in defense laboratories and industries. Butbecause of the cost of developing modem technol-ogy, it seems unlikely that DoD can afford todevelop all the technology it needs in parallel withthe civilian sector. Dependence on the private sectoris not all bad: commercial development of technol-ogy is a basic strength of the industrialized, non-communist world. Failure to exploit developmentsin the civilian sector would be throwing away amajor advantage over the Soviets. But relying on theprivate sector means that defense development andproduction will depend increasingly on the health ofthe civilian sector and on the ability of DoD and its

contractors to gain access to the products of thecivilian sector. Thus DoD faces two challenges:maintaining access to the technology developed inthe commercial sector, and coping with the interna-tional nature of that sector.

DoD and Congress face three generic problems.The first is keeping dual-use companies interested indoing defense work. Some are leaving the defensebusiness. Others have technology that DoD coulduse, but are reluctant to get into the defense business.These attitudes are based primarily on perceptions ofthe difficulties of doing business with the govern-ment, and the problems of doing business in bothsectors simultaneously. Second, high-technologyindustries are moving offshore due to foreigncompetition. Some have almost vanished, others areon the way. Furthermore, it seems likely that in thefuture some new technology-based industries willdevelop in other nations and never take root here.Careful balance will be necessary to nurture U.S.industries while maintaining access to foreign tech-nology. Congress will have to consider U.S. tradeand industry policy carefully. Third, entire indus-tries, individual companies, and the many-steppedtrails that lead from raw materials to finishedcomponents cross many national borders. In manycases, it is nearly impossible to determine what aU.S. company is, while in others it is difficult toseparate U.S. companies from their foreign partners.Congress will have to come to grips with themeaning of foreign ownership and foreign siting forthe availability of technology, as well as with howdependent the United States can afford to be onforeign sources. These international relationshipswill complicate attempts to protect U.S. supplysources.

Barriers Between Civilian andMilitary Industry

Since World War II, the U.S. economy hasevolved relatively separate military and commercialsectors. They have different business practices, onedictated by government regulations and procure-ment practices and the other flowing from themarketplace. In recent years the international mar-ket has had a considerable effect on shaping thelatter.


Government practices have made it increasinglydifficult for DoD to obtain state-of-the-art technol-ogy in areas where civilian industries are leading,making defense business unattractive to innovativecompanies and contributing to traditional suppliersleaving the defense business. Many firms that are notheavily involved in defense business are reluctant todeal with the government because they consider it tobe a bad customer. Moreover, many do not needDoD’s business and can simply opt out. The barriersare not technological, but legal, institutional, andadministrative. Some are the direct result of legisla-tion, others flow from DoD regulations, includingoverly cautious interpretations of laws. Some com-mercial firms cite excessive regulation, burdensomeauditing and reporting requirements, compromise oftrade secrets, and loss of data rights. Large defensecompanies have similar complaints, but have ad-justed to working under these conditions. But forsmaller companies, getting into the defense businessmeans heavy investment and reorientation of busi-ness practices.

A company can organize to do business in eithersector, but can rarely do both under one administra-tive roof. Companies that do business in both sectorstypically have separate divisions that are organizeddifferently and almost never share staff, productionand research facilities, data, and accounting pro-cedures. These differences are profound. In largeaerospace companies the commercial side respondsto market conditions, whereas the military sideresponds to Service programs, government regula-tions, and the Federal budget. Their planning is“slaved” to the Federal planning and budgetingcycle. Corporate structures and rules tend to mirrorthose of DoD and tend to pass government encum-brances down to lower level suppliers. Companiesdoing government contract work have to keep theirbooks in formats that are compatible with govern-ment auditing rules and procedures.

Following these and other government rules addsto the costs of doing business, costs that canlegitimately be passed on to the government cus-tomer. Tighter control of the defense businessultimately translates into higher costs to DoD. TheUnited States is apparently willing to bear thisincreased cost as the price of other benefits-forexample, knowledge that the government is trying tokeep the process honest. However, imposing the

same rules on dual-use industries has other, farther-reaching effects. It makes them reluctant to dobusiness with DoD and encumbers their productswith additional costs that may adversely affect theirinternational competitive positions. When dealing inboth sectors, companies can accept either the highercost of following government business procedures,or the higher costs of maintaining two separatebusiness practices-one for government businessand one for other business. With some exceptions,DoD product specifications are also seen as encum-brances; characteristics that are of no value in thecommercial marketplace are engineered into theproducts for sale to DoD.

Government contracts regulate profits, creating abusiness environment very different from that inwhich most high-technology companies deal. Thesecompanies are used to investing heavily in R&D,recovering their investments through large profits,and then reinvesting in the next generation ofproduct. Moreover, their customers see only theproduct, whereas DoD insists on knowing how theproduct was made. Defense contractors get by onsmall profits, in part because much of their R&Dcosts are covered either by contract or IR&Drecovery. But dual-use companies qualify for little ifany IR&D recovery and are reluctant to do contractR&D. The government owns the rights to datagenerated by contract R&D so that it can keep thesubsequent phases of a project competitive bymaking a data package available to all bidders. Butcompanies that live by their innovation in thecommercial market see this process as offering theirtrade secrets to the competition. DoD proceduresprovide the winner of a development contract poorprofit margins, no guarantee of a continuing rela-tionship with DoD, and little incentive to innovateand provide a superior product.

Some industries, like advanced composites, arecurrently so closely tied to the defense business thatthey are apparently willing to live with theseproblems. But they worry that their competitiveposition may be damaged as the commercial marketdevelops. At the other extreme, the companies thatproduce fiber optics are reluctant to get involved ina defense market they see as always being a smallpart of their business: they do not necessarily see thepotential payoff as worth the aggravation.

Chapter 3-Summary . 35

While the small amount of military fiber opticsbusiness might be seen as evidence that the industryis not really important to defense, some within theDoD see it as a critical new technology for futuresystems, one in which defense could gain tremen-dously just by exploiting what has been and is beingdeveloped in the commercial sector. But DoD hasbeen generally slow in adopting fiber optic technol-ogy. Program managers have much to lose byinserting risky new technologies which may delayschedules and increase costs, but little to gainbecause the advantages of the substitution willusually become apparent only on someone else’swatch.

In the software industry, the divergence betweengovernment and commercial practices has beenenough to produce separate defense and commercialbusinesses that often do not share technology. Theprocedures, policies, and management of large-scalesystems in the military and civilian sectors divergestarting with requirements definition, continuing inthe development or acquisition of software, andthroughout the entire life cycle of the software. Thisrestricts the flow of leading-edge technology fromdefense into the commercial sector and reduces DoDaccess to readily available commercial products.Most of the differences can be attributed to thepolicies, regulations, standards, and directives man-dated by DoD. DoD software requirements are morerigid than their commercial counterparts. Defensesystems tend to be overwhelmingly custom built,while commercial systems will use as much off-the-shelf technology as possible. Software companiesare particularly concerned about data rights, whichthey see as critical to competitiveness. Companiesare reluctant to deal under DoD restrictions; in theireyes the government would be taking and possiblygiving to their competitors the very basis of theirbusiness.

International Competitiveness and theHealth of U.S. Industries

The Department of Defense has been concernedfor some time about the implications for defense ofdeteriorating competitive positions of U.S. manu-facturing companies in the international market.12

The government is also concerned from a widerperspective that this trend is weakening and under-mining the U.S. economy. DoD shares the concernthat a weakening economy and a drain of resourcesinto purchases of foreign goods will reduce moneyavailable to produce defense equipment, but itsprimary concern is the continuing availability ofnecessary items and technology.

The government does not as yet have a policyregarding dependence on foreign sources for defensematerial and technology, let alone a game plan forimplementing such a policy. The Undersecretary ofDefense for Acquisition has recommended a plan tobolster defense-related manufacturing in the UnitedStates. 13 The report detailing that plan does not makea statement on how much foreign dependence istolerable, although it does imply that some isunavoidable.

The complexity of the problem is illustrated bythe issue of cooperative development and produc-tion of defense equipment with the European NATOAllies. It has been long-standing U.S. policy toencourage multinational procurement of similardefense equipment to foster commonality, to get thebest equipment into the forces of all the Allies, tosave money, and recently, to exploit a broadmultinational technology base. In recent years theDefense Department has made great progress ingenerating international memoranda of understand-ing for joint development, with the help of initiativeslike the Nunn Amendment. But as the Europeanshave become more interested in cooperative devel-opments, they have also sought a greater share ingenerating the technology and a larger market sharefor their defense industries. Interest by U.S. compa-

12For ~xmplcs, we Defense Science Botid) “Report of the Defense Science Board Task Force on Defense Semiconductor Dependency,” prepmedfor the Office of the Under Secretary of Defense for Acquisition, February 1987; Report to the Secretary of Defense by the Under Secretary forAcquisition, “Bolstering Defense Industrial Competitiveness: Preserving Our Heritage, the Industrial Base, Securing Our Future” (Washington, DC:Department of Defense, 1988); and Martin Libicki, Industrial Strength Defense: A Disquisition on Manufacturing, Surge, and War (Washington, DC:National Defense University, 1986). See also, U.S. Congress, Office of Technology Assessment, Paying the Bill: Manufacturing and Americans TradeDeficit, OTA-ITE-390 (Sprin#ield, VA: National Technical Information Service, June 1988).

13&x “Bo]st@ng Defense Industrial Competitiveness, ” Op. cit., fOOtnOtC 12.


nies in joint ventures with Europeans has beenspurred, in part, by fears that several trends inEuropean thinking could sharply curtail their sales inEurope. Thus, the cooperative programs are atwo-edged sword helping U.S. sales in Europe whilestimulating European sales to the United States; andhelping U.S. defense policy in general, while bothhelping and hindering the maintenance of the U.S.defense industrial technology base. Crafting a work-able policy will be a tricky job.

There are three basic policy choices:

●

●

●

demand that anything that goes into defenseequipment be built in the U.S. from U. S.-sourced components, taking whatever meas-ures are necessary to ensure that all thenecessary industries are alive and well in theUnited States;let the market dictate which industries will behealthy in the United States and look only forthe best deals wherever they can be foundworldwide; orchoose some industries that have to be locatedin the United States, take appropriate measuresto ensure that, and let the rest go with themarket.

The first and third require some sort of inter-vention in the international economy, either support-ing the international competitiveness of U.S. compa-nies or protecting, supporting, and subsidizing U.S.companies that cannot otherwise survive. Anotherapproach is to design nothing into U.S. defensesystems that cannot be domestically sourced. Butthis cuts off a great deal of modern technology, aWestern strength. In making these choices, theUnited States will have to decide how dependent wecan afford to be, and how much independence we arewilling to pay for. If the United States demandsself-sufficiency without taking measures to keepU.S. companies alive and competitive, the list oftechnologies available for defense systems is likelyto decrease as time goes on.

It will be necessary to decide how to treatdependence on various nations. There are significantdifferences in being dependent on Canada (alreadydefined as part of the North American industrialbase), Britain, our other NATO allies, Mexico,Japan, Korea, etc. U.S. and Canadian companies are

closely intertwined. Despite the recent controversyover the trade agreement and other arguments, weare each other’s largest trading partners. Canada isalso a NATO ally with a common security interest.The chances of being cut off from Canadian sourceseither by policy or by hostile act are minimal. We arealso close to our European Allies; much of ourdefense equipment is bought to defend them. But weare separated from Europe by an ocean, and theyhave not always supported U.S. military actions.Other nations are much less tightly tied to the UnitedStates.

The high-technology economy is an internationalone and responds to international market forces.These forces are likely to continue to move indus-tries offshore despite U.S. efforts to will (or legis-late) them to stay. In the vast majority of cases,defense business is far too small to provide thenecessary clout, particularly when faced with othernations that manipulate their civilian markets tokeep their companies healthy. Competition comesfrom Japan, the smaller Asian nations—Korea,Taiwan, Singapore, etc—and Western Europe. TheEuropeans are taking dramatic steps to improve theirinternational competitive position, particularly inhigh technology industries. These include the eco-nomic integration of the EC in 1992, and the fundingand encouragement of large cooperative R&D proj-ects.

Although all industries are different, the plight ofthe fiber optics industry is illustrative. While healthyin the United States, it faces increasingly stiffcompetition at home and continuing difficultiesabroad stemming from limited access to foreignmarkets. Both the Europeans and the Japanese aremaking major pushes in fiber optics and photonicsin general. U.S. technology and production costs areat least competitive. But while U.S. producers havebeen largely excluded from some important foreignmarkets, the U.S. market remains open to foreignvendors. Japanese companies can sell in foreignmarkets at low prices because their government hasdiscouraged foreign competition in Japan whereprices are kept artificially high. The closed domesticmarket supports overseas competitiveness.

The U.S. software industry faces a different sort ofchallenge. It is currently strong and competitive, butthe rapid growth in worldwide demand for software

Chapter 3-Summary .37

threatens to outstrip the capacity of U.S. firms tomeet it, leaving a large opening for foreign firms topenetrate the market. Japan, France, the UnitedKingdom, Korea, Singapore, Taiwan, and India havethe capacity to penetrate the global market. Andmany of these nations have trade policies that eitherdiscourage sales by U.S. companies or fail to protectthe intellectual property rights of those companies:“pirated” software is becoming a major problem.Moreover, the Japanese are making rapid strides inturning software design from art to manufacture,building software factories to increase productivitydramatically.

Internationalization of Industries

Efforts to protect and nurture U.S. companies willbe complicated by trends toward internationaliza-tion in high-technology industries. Examples arefound in the advanced composites industry in whichmany of the firms that appear to be American—because they have American names or U.S. facilities—are actually owned by foreign companies and in thefiber optics business where international joint ven-tures are used to get into otherwise closed markets.International ownership, vertical and horizontalintegration, and international siting make it difficultto define in any convincing way what an Americancompany is. Moreover, the sequence of steps thatleads to a final product often crosses internationalboundaries many times and shifts as prices andavailability of components shifts. Is a Pontiac builtin Korea any more or less an American product thana Honda made in the United States or a Chevrolet/Toyota assembled in California from U.S. andJapanese parts?

Difficulties in identifying U.S. companies willproduce difficulties in writing legislation to protectthem or establishing DoD policy to encourage thegrowth of important domestic industries. Foreignplants owned by U.S companies, U.S. plants ownedby foreign companies, joint ownership, and jointventures all offer different sets of problems.

Formulating Policy

These trends toward internationalization willcomplicate difficult issues that Congress and theAdministration are already facing. Paramount amongthese is to decide whether the U.S. Government willplay a major role in encouraging and supporting U.S.

commercial business and industry, or whether—almost unique among the governments of majornations—it will continue to remain more or lessaloof, confining its activities to a few internationaltrade negotiations. Other governments encouragethe development of commercial technology andassociated industry, help to foster a domestic situ-ation conducive to growth, and support aggressiveoverseas marketing.

Having decided government’s role, the next issuewould be to define goals. These might include:

●

●

●

●

keeping key nondefense manufacture and de-velopment in the United States,

keeping manufacture and development in thehands of U.S.-based (or U.S.-owned) compa-nies;

preserving some portion of the U.S. market forU.S.-based (or U.S.-owned) companies; and

gaining access to foreign markets for U.S.firms.

Defining such goals will entail arriving at a workingdefinition of a U.S. company, or at least of howlocation and ownership affect U.S. national securityinterests.

It would be necessary to decide how large a roledefense needs would play in deciding which indus-tries are in need of government attention. Thisdecision would have to balance the problems offoreign dependence against the risk of diminishedaccess to foreign technology and manufacture. Itwould also have to consider how much the UnitedStates is willing to pay to buy domestically thatwhich may be available at a lower price elsewhere.The lessons of “low-priced oil” from the PersianGulf are instructive here. Determining the accept-able degree of offshore dependence for defenseequipment will necessitate deciding the level ofcomponentry which DoD would have to specify ascoming from domestic sources. For example, is itsufficient to require that systems or subsystems bedomestically sourced, or does DoD have to assurethat some or all of the components are made in theU. S. A.? This decision would dictate the level atwhich DoD would need visibility into the manufac-turing process and have to keep a data base onsuppliers.


Whatever the goals are, Congress will have todecide what levers can be pulled to make thoseattainable. In most cases, simply controlling defenseprocurement will not be enough to influence theindustry: it may ultimately lead to an inefficient,backward, protected industry that is incapable ofcompeting on the world market. Such an industrymight only be capable of providing DoD withobsolete technology or overpriced products. Thegovernment has the option of getting more deeplyinvolved in stimulating the development of technol-ogy for commercial ends, including making govern-ment R&D facilities more available and providinggreater incentives for corporate investment. Yetanother option is to formulate a strategy—as Japan

and other nations have—for controlling access tocritical U.S. commercial markets in order to preserveand support domestic industrial capabilities. A thirdpolicy lever that can be manipulated, but not totallycontrolled, is the cost and availability of capital forconducting R&D. Major technological develop-ments are capital intensive, with costs measured inthe hundreds of millions to billions of dollars.European and Japanese companies pay less toborrow money than do U.S. companies—far less inthe case of the Japanese. This allows them to carryon more projects simultaneously, and to sell theresultant products at lower prices than those of theirU.S. competitors, putting U.S. companies at acompetitive disadvantage.

Chapter 4

Planning and Funding DoDTechnology Base Programs

CONTENTSPage

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . + . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41HOW THE DEFENSE DEPARTMENT MAKES AND IMPLEMENTS

TECHNOLOGY POLICY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41OSD’s TECHNOLOGY BASE OVERSIGHT ACTIVITIES . . . . . . . . . . . . . . . . . . . . . . . . . 44

Developing a Technology Base Investment Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . ..., 44Establishing Research Priorities and Direction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46Coordination of Technology Base Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47Acting as a Strong Advocate for DoD’s Science and Technology Programs . . . . . . . . . 48Promoting Cooperation Among the Services and DARPA . . . . . . . . . . . . . . . . . . . . . . . . . 49Evaluating the Goals of the Technology Base Programs ... . ..... .......... . 51Evaluating Research Activities . . . . . . . . . . . . . . . + . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

ORGANIZATION OF OSD FOR OVERSIGHT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51Managing Technology Base Activities at DARPA and SDIO . . . . . . . . +....... . . . . . . 52Recruiting and Retaining Scientific Personnel . .... ..+. . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

DoD SUPPORT OF ITS TECHNOLOGY BASE PROGRAMS . . . . . . . . . . . . . . . . . . . . . . 56CONCLUSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

FiguresFigure Page2. Management of the Department of Defense Technology Base Program . . . . . . . . . . . . . 433. Comparison of RDT&E and Technology Base . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

TablesTable Page1. Department of Defense Funding of Technology Base Programs, Fiscal Year 1989 . . . 422. DoD Technology Base Funding, Fiscal Years 1984 Through 1989 . . . . . . . . . . . . . . . . . . 493. Individual Service Funding for Research (6.1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . +. . . . . 494. DoD Technology Base Funding Trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

Chapter 4

Planning and Funding DoD Technology Base Programs

INTRODUCTIONIn today’s technology-intensive environment-in

both the military and commercial context—theability of an organization to compete and win ishighly dependent on its ability to discover, develop,and apply advances in science and technology to itssystems and products. Success in that endeavordepends, in turn, on the ability of the organization toplan its technology investment strategy, marshal theresources to support it, and build and sustain atechnology base vital enough to produce the neededadvances.

The development and management of the tech-nology base underlying defense systems is anexceedingly complex enterprise. It is as multifarious—and as important to national interests—as thecapabilities and performance of the defense systemsthemselves. This chapter examines the Departmentof Defense (DoD) system for managing its technol-ogy base programs. It reviews how the Office of theSecretary of Defense (OSD)-particularly the Of-fice of the Director of Defense Research andEngineering (ODDR&E)--carries out its technol-ogy base oversight responsibilities. The purpose ofthis chapter is to evaluate the ability of thepresent OSD technology base management sys-tem to do its job, and not to judge the perform-ance of any Administration or individual DoDofficers. It focuses on oversight activities such asstrategic planning and coordination of technol-ogy base programs-that is, on the role of OSD inplanning the programs of the Services, defenseagencies, and the Strategic Defense InitiativeOrganization (SDIO), and forming them into acoherent whole—and not on the management ofspecific technology base program elements (PEs).

These oversight responsibilities include: 1)developing an overall technology base investmentstrategy; 2) setting research priorities and directions;3) reviewing and evaluating the technology base

program goals; 4) coordinating the numerous re-search activities that make up DoD’s technologybase programs; 5) acting as an advocate for thetechnology base programs; and 6) evaluating theoutcomes and effectiveness of DoD-sponsored tech-nology base activities.

The next section of this chapter briefly describesthe activities that comprise DoD’s technology baseprograms and how OSD, the three Services, theDefense Advanced Research Projects Agency (DARPA),and the SDIO2 are organized to manage and imple-ment their technology base programs. The followingsection then reviews how OSD, the three Services,and DARPA fulfill their respective technology basemanagement responsibilities.

The second major portion of the chapter examinesissues associated with the way in which OSD isorganized to carry out its technology base oversightactivities. The chapter concludes with a discussionof the support within DoD for its technology baseprograms, including an analysis of past and currenttechnology base funding trends.

HOW THE DEFENSEDEPARTMENT MAKES AND

IMPLEMENTS TECHNOLOGYPOLICY

Although the Department of Defense will investless than 4 percent of its entire budget in technologybase activities in fiscal year 1989 (see table 1 ), manyobservers inside and outside the Pentagon considerDoD’s technology base programs to be a crucialinvestment in the Nation’s overall security. Themilitary’s technology base programs represent awide spectrum of ‘‘front-end” technology develop-ment, beginning with a broad base of basic researchsupport and extending through the demonstration oftechnology that might be applied in future defensesystems. The scope of DoD’s technology base

I For ~ more detailed discussion of how OSD and the Services organize their respective technology base programs, see the Mach 1988 OTA rePofientitled The Defense Technology Base. Introduction and Overview-+ .$pe~ial Report, OTA-ISC-374 (Washington, DC: U.S. Government Printingoffice).

zFor fuflher infomatlon on the SD] program, see Library of Congress, Congressional Rc*~ch se~i~e, ‘‘The Strategic Defense Initiative: ProgramDescription and Major Issues,” CRS Report No. 86-8 SPR, 1986.

- 4 1 –


Table l—Department of Defense Funding of Technology Base Programs, Fiscal Year 1989(in millions of dollars)a

CategoryArmy Navy Air Force DARPA

Research (6.1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $173 $355 $196 $88 $956Exploratory Development (6.2) . . . . . . . . . . . . . . . . . . . . . . . . . . $561 $431 $574 $624 $2,522Advanced Exploratory Development (6.3A) . . . . . . . . . . . . . . . . $422 $193 $764 $557 $2,099

Service or agency total . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $1,156 $979 $1,534 $1,269 $5,577Strategic Defense Initiative . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $3,606

Total DoD technology base programs . . . . . . . . . . . . . . . $9,183

appropriated,b~tegov totals al= i~lude ~nding for tie Ower defense agencies and UniWQIY Research Initiatives IW-Irm.

SOURCE: Office of the Secrwary of Defense.

programs includes such diverse concerns as meteor-ology technology and the technologies for autono-mous guided missiles capable of differentiatingamong various targets.

DoD organizes its technology base programs intothree budgetary categories: research (funded undercategory 6.1); exploratory development, the practi-cal application of that research (budget category6.2); and advanced exploratory development, whichprimarily consists of the building of prototypes todemonstrate the feasibility of applying a particulartechnology to a weapon system (budget category6.3A). Work funded under the remainder of theDepartment’s budget for research, development,test, and evaluation (RDT&E), representing about80 percent of the RDT&E budget, is not consideredto be part of the technology base.3

DoD’s complex technology base program isplanned, organized, and implemented by the threeServices (Army, Navy, and Air Force), DARPA, theother defense agencies, and the SDIO, with theoversight and guidance of the OSD. The largestportion of the technology base program is conductedoutside DoD by industry (50 percent) and universi-ties (20 percent), with DoD in-house laboratoriesconducting the remaining 30 percent.

In the last three years, each of the three Servicesand OSD have reorganized the management of theirtechnology base programs. As a result of theGoldwater-Nichols Act, the position of Under Sec-retary of Defense for Acquisition [USD(A)] was

established and given responsibilities for all RDT&Eactivities except for those of the SDIO. The Directorof SDIO reports directly to the Secretary of Defense(see figure 2). The Goldwater-Nichols Act alsoreestablished the DDR&E as the primary individualresponsible for DoD’s technology base activities.The DDR&E is responsible for assuring the appro-priate emphasis and balance for DoD’s entiretechnology base program, except for SDIO.

Once the Services-the Army, Navy, and AirForce—have formulated their technology base pro-grams, the Deputy DDR&E for Research andAdvanced Technology [DDDR&E(R&AT)] has thetask of ensuring that their proposals have respondedto OSD guidance. The Deputy for R&AT serves as‘‘the corporate guardian” of the technology baseprograms, ensuring that the Services’ programs arewell balanced, with little duplication of effort, whileattempting to meet the current and future scientificand technological needs of DoD. The Services’technology base programs are coordinated withDARPA’s programs at the next level, the DDR&E.Finally, conflicts among these four programs andSDIO are adjudicated only at the highest level, theSecretary of Defense.

Each of the Services conducts an extensiveannual top-down, bottom-up planning exercise.From the top, the Services receive OSD’s annualDefense Guidance document, which provides themwith guidance on developing their overall RDT&Eprograms. Planning begins with a review and

SDOD typlca]]y considers 6.1 and 6.2 as its “technology base PrWams~ “ with 6.1 through 6.3A normally referred to as its “science and technologyprograms. ” However, in recent years these distinctions have become blurred in everyday usage. This report uses both terms to refer to budget categories6.1, 6.2, and 6.3A.

Chapter 4-Planning and Funding DoD Technology Base Programs ● 43

Figure 2—Management of the Department of Defense Technology Base Program

Secretary of Defense

Director,Strategic Defense

I n I t i a t i veOrgan i zat i on

Under Secretaryof Defense,Ac q u is i t Ion

USD( A)

IDirector, Defense

Research and+Engineer i ng

[DDR&E]I I


evaluation of the previous year’s research activities.When this review is completed, the Services thendecide which activities to continue, which to transi-tion from 6.1 into 6.2 or from 6.2 into 6.3Aprograms, which to move beyond 6.3A, and whichactivities to end.4

Each of the three Services operates and managesits technology base activities differently. Comparedto the other two, the Army employs a less centralizedapproach, relying on major field commands-theArmy Materiel Command (AMC), the Corps of

Engineers (COE), and the Surgeon General (TSG)—as well as the Deputy Chief of Staff for Personnel(DCSPER), to help develop and implement itstechnology base investment strategy. Compared tothose of the Navy and Air Force, the Army’stechnology base headquarters staff is quite small.The Deputy for Technology and Assessment (DT&A)is considered to be the Army’s Program ExecutiveOfficer (PEO) for the technology base programs.The DT&A is responsible for coordinating thetechnology base programs of AMC, COE, TSG, andthe DCS for Personnel. The Army’s Laboratory

qTransitions may actu~ly OCCU at times from 6.1 or 6.2 to 6.3, 6.4 Or even directly to operational systems.


Command (LABCOM) is responsible for overseeingand managing 75 percent of the Army’s technologybase program, including its eight laboratories, sevenresearch, development, and evaluation (RD&E)centers, and the Army Research Office. The com-manding officer of LABCOM reports to the com-mander of AMC.

Of the three Services, the Navy has placed itstechnology base management institutions at thefarthest remove from its procurement institutions.But relevance to Navy needs still remains a powerfulfactor in selecting projects, especially in 6.2 and6.3A. Further, unlike the other Services, the Navyperforms the majority (60 percent) of its technologybase programs in-house. Many of the Navy laborato-ries are capable of performing the entire spectrum ofRDT&E activities. The Navy supports the oldest andlargest of the Services’ research programs, alongwith the smallest program in advanced technologydemonstration. The Navy claims to be rebuilding itsadvanced exploratory development program, which,unlike the other Services, it does not manage in thesame office as its 6.1 and 6.2 programs.

As of November 1, 1987, the Deputy Chief ofStaff for Technology and Plans (DCS T&P), AirForce Systems Command, was established to over-see the entire Air Force technology base program.The DCS T&P is also the PEO for, and the singlemanager of, the Air Force’s technology base pro-gram. The Air Force Chief of Staff has designatedthe technology base program as a‘ ‘corporate invest-ment” to help raise its visibility and to provide along-term, stable funding base for the program. TheAir Force operates the largest extramural technologybase program of the three Services. Its technologybase activities are more centralized than are those ofthe other Services. The Air Force places specialemphasis on technology insertion: It has the largestadvanced exploratory development program; and itslaboratories are more closely linked to its five majorsystems divisions than those of the other Servicesare to theirs.

The role of DARPA appears to be changing withthe recent establishment of its Prototyping Office,

DARPA was originally established to support high-risk, long-range research. It does not operate labora-tories or conduct in-house research. Consequently,the majority of DARPA’s budget is contractedthrough the three Services to industry (75 to 80percent) and universities (20 percent), with only asmall fraction of DARPA’s technology base activi-ties actually conducted by the military. There issome concern that allowing DARPA to enter thedomain of hardware development and prototypetesting might compromise its support of long-range,high-risk research.

The SDIO program is centrally managed, with itsdirector reporting to the Secretary of Defense.Although the entire SDIO budget is funded under6.3A, DoD estimates that approximately 15 to 20percent of the SDI budget is spent on research andexploratory development. The majority of SDIprojects are executed through the Services withsome additional efforts through other executiveagencies, including DARPA, the Defense NuclearAgency, the Department of Energy, and the NationalAeronautics and Space Administration.5

OSD’S TECHNOLOGY BASEOVERSIGHT ACTIVITIES

The DoD technology base programs play a crucialrole in the military’s ability to develop technologyand apply it rapidly to meeting the Nation’s securityneeds. DoD reports that its science and technologycapabilities continue to improve, but the technologi-cal lead over potential adversaries is shrinking. Oneway for DoD to counter this adverse trend is to makesure that its technology base programs are planned,managed, and executed as effectively as possible.

Developing a Technology BaseInvestment Strategy

DoD does not have an overall, coordinatedtechnology base investment strategy or plan toestablish science and technology (S&T) priorities.According to a recent report by the Institute for

sThis C]laptcr does not exmlne [he rese~ch activities of several smaller agencws within DoD which account for less tb 2 percent of RDT&E. ~oseagencies include: the Defense Mapping Agency; National Sccuriiy Agency; Defense Nuclear Agency; Defense Support Project Office; DefenseComn]unications Agency; Defense Intelligence Agency; Defense Logistics Agency; and the Uniformed Services University.


Defense Analyses (IDA),6 a significant amount oflong-range planning is taking place in the Services,the research and development (R&D) centers, andDARPA, but “these efforts are, at the moment,pursued independently within each of the Servicesand, to some degree, independently at the R&Dcenter (laboratory) level.”7

Many observers within the military believe thattechnology base planning should remain decentral-ized.8 The Services assert that they have a muchbetter understanding of their respective combatneeds than does OSD. As a result, the Services-notOSD-possess the knowledge and technical skillsnecessary to establish a rational technology baseinvestment strategy to meet future combat needs.Many analysts believe that any attempt to centralizeplanning for the technology base programs withinUSD(A) would be unsuccessful. Representatives ofthe Services and DARPA argue that OSD’s primaryrole should be as advocate, reviewer, and coordina-tor for DoD’s technology base programs, In thisview, USD(A) should make sure that the technologybase is clearly understood within OSD, that Serviceand DARPA programs are reviewed for adequacy,and that unwarranted program duplication betweenthe Services is avoided. Advocates of this view alsohold that OSD is less able to defend the technologybase budget than are the Services, and greater OSDinvolvement would result in less Service support ofthe technology base.

On the other hand, the report of the IDA task forcedoes not endorse these beliefs. The task forcerecommended that OSD adopt a strategic planningprocess to “tie together the investment strategies asthey currently exist in the Services and Agencies. ”9

This strategic planning activity would involve OSD

working with the Services, the defense agencies, andSDIO in order to establish technology base goals andpriorities. It does not imply the creation of an OSD“planning czar” who establishes goals and objec-tives for all of DoD’s technology base programs.The IDA report notes that strategic planning, in anyorganization, will not succeed if it fails to involve anadequate number of the right people. In this ap-proach, once a coordinated technology base invest-ment strategy has been developed, the actual plan-ning and execution of the various programs couldcontinue in the current decentralized fashion, As inthe past, the Services and DARPA would beresponsible for organizing and executing their tech-nology base programs. However, the DDR&E wouldbe in a position to evaluate each agency’s program,based on how well it responded to the prioritiesestablished during the strategic planning process.

The Services and DARPA assert that theannual Defense Guidance document-supplementedwith additional Service requirements documentation-provides adequate planning guidance to developtheir respective technology base investment strate-gies. However, many other observers criticize theDefense Guidance on the ground that it is developedthrough a fragmented process which fails to producea coherent, well-coordinated U.S. defense posture.The document is prepared by the UnderSecretary ofDefense for Policy, based on Administration guid-ance and inputs from the Joint Chiefs of Staff, theUnified and Specified Commands (CINCs),10 theService Secretaries, other OSD Staff (including theDDR&E), and other relevant sources. Once theGuidance is approved and published, the Servicesuse it to build their respective programs, includingtheir science and technology programs. There are

6Task Force Report, ‘‘The Improved Coordination of DoD Science and Technology Programs” (Alexandria, VA: Institute for Defense Analyses, July1988). At the request of the DDR&E, IDA assembled a [ask force consisting of numerous S&T managers from the Services and OSD to examineapproaches for improving coordination of DOD’s technology base programs.

7The Semices have conducted impressive technological forecasting activities, including the Air Force’s Forecast 11 study, the Navy’s 21 study. andthe Army’s proposed Strategic Technologies for the Army (STAR) study. Such studies have been used to establish S&T priorities in the Services.However, as the IDA Task Force indicated, these activities arc primarily pursued independently within each Service.

glnstltute for Defcn~ Analyses, op. cit., footnote 6 p. ~.9Summq Repon ~d Recommendations of the DA Task F~rcc on ]mproved ~’oordinatioll of the DOD Science and ‘kchnoIogy %OgTafnS

(Alexandria, VA: Institute for Defense Analyses, July 1988), p. 11-2.10A ~ified ~ommmd, Composed of Slwlficm[ forces from two or more sc~i~cs, has a broad, continuing mission (usually geographically based). A

specified command, composed primarily of forces from one Service, has a functional mission. The eight unified commands are Europe, Pacific, Atlantic,Southern, Central, Special Opertitions, Transportation, and Space. The spcclficd commands arc the Strategic Air Command, Acrospacc DefenseCommand, and Military Aidi ft Command. The names of the commands designate their primary geographic or functional area of” responsibility. CentralCommand, created in 1983, is conccmed with the Persian Gulf’ region.


also various inter-Service requirements documentsthat help to gear technology to future defense needs.

Nevertheless, the current 120-page documentprovides only one page of guidance for theDoD-wide technology base programs. More couldhelp generate a stronger technology base program.Such broad guidance allows the Services andDARPA to justify technology base programs thatthey view as being in their individual best interests,but which may or may not meet the overall futurescience and technology needs of the Department ofDefense as a whole.

In the absence of a centralized S&T investmentstrategy, it is extremely difficult for the DDR&E toassess the technology base programs of the Servicesand DARPA, other than for technical merit. The1983 Grace report indicated that planning whichpermits the bottom-up approach to predominate—the current situation-often results in duplication ofeffort, and ineffective coordination of science andtechnology programs.

11 OSD's technology baseinvestment review is primarily an information gath-ering function. When the Services present theirannual technology base investment strategy to theDDDR&E(R&AT), no formal written feedback isprovided, although there are usually verbal com-ments. Until this year, each of the Service’s pro-grams was reviewed separately, making cross-Service comparisons difficult. The Defense agencies(primarily DARPA) are not required to participate inthis process, although they usually do to a limitedextent.

Under current conditions, OSD cannot ensurethat DoD’s technology base programs are wellbalanced, properly coordinated, and capable ofmeeting the current and future science and technol-ogy needs of DoD. On the other hand, it is clear thatOSD would be unable to conduct an effectivetechnology base investment strategy without theclose cooperation and goodwill of the Services andDARPA. Because the Services currently dominatethe planning process, and act independently of oneanother, any effort to consolidate this function in

OSD could cause dislocation and disruption ofexisting technology base program management.

Establishing Research Prioritiesand Direction

The dominant position of the Services in deter-mining technology base initiatives arises from, andcontributes to, the lack of an overall technology baseinvestment strategy. The Services have filled apower vacuum and now protect their power. Insteadof working with OSD to establish priorities based onoverall defense needs, the Services allocate re-sources based on their own views of their individualneeds. A 1981 Defense Science Board study recom-mended that a DoD R&D investment strategy linkedto future combat needs be utilized in technologybase planning “. . . so that technologies fundedthrough the allocation processes would be moreexplicitly and consistently related to future opera-tional needs. ” *2

In the absence of broad strategic guidance,individual Service goals tend to supplant moregeneral strategic ones. As the primary civiliancomponent within DoD, the Office of the Secretaryof Defense is supposed to act as a counterbalancingforce to the Services, working objectively with theServices and DARPA to develop an overall technol-ogy base strategy in the best interest of DoD as awhole. In principle, once the strategy is articulated,the Services and DARPA develop science andtechnology goals to achieve that strategy.

The implications of inadequate OSD guidancecan be significant with regard to the types oftechnological priorities the Services are willing tosupport. For example, according to Samuel P.Huntington,13 the Services are extremely reluctant to

support “orphan” functions that are not central to aService’s own definition of its mission or fightingdoctrine. This can present great difficulties forsetting well-balanced science and technology priori-ties, since modern technology has provided capabili-ties that may not coincide with traditional ap-

11‘~esjdent’s Private Sector Survey on cost Control, ” report of the Task Force cm Research and Development, Executive Office of the President,Dec. 8, 1983, p. 30.

l’lJ.s. Dep~ment of Defense, office of tie DDR&E, “Report of tie Defense Science Board, 1981 Summer Study panel on tie Technology B~”(Springfield, VA: National Technical Information Service, November 1981), p. i-2.

13 SmucI P. Huntington, ‘‘Organization and Strategy, “ in Reorganizing America’s Defense (Washington, DC: Pergamon Press, 1985), p. 236.


preaches to mission accomplishment or to theaccepted division of mission responsibility.

The Services are often reluctant to supporttechnology base initiatives that challenge theircurrent mission or fighting doctrine, and becausethey dominate the technology base planningprocess, they are in a position to discourage suchinitiatives.

Coordination of Technology Base Programs

DoD lacks a strong and focused coordinatingcapability for its science and technology programs.Although DoD has over 200 tri-Service and inter-agency coordinating groups, in general they have notbeen effective at providing high level coordinationacross the DoD-wide technology base programs.(However, some, such as the Joint Service Electron-ics Program [JSEP], have produced impressiveresults.) Coordination efforts are further hamperedbecause a significant portion of DoD’s science andtechnology programs are not under the direct pur-view of the DDDR&E(R&AT).

In its task force report on improved coordinationof DoD S&T programs, the Institute for DefenseAnalyses concluded that it is necessary to differenti-ate between “technical interchange” and “pro-grammatic coordination, ” The IDA study con-cluded that currently there is much technical inter-change among the Services, but very little program-matic coordination is aimed at identifying scientificor technological gaps and overlaps. The IDA studystates that, without proper coordination, it is difficultto ensure that the total DoD S&T program isproperly addressing the overall science and technol-ogy needs of DoD.14

OSD uses annual science and technology reviewsto help evaluate and improve the coordination of itstechnology base programs. These reviews are de-signed to examine a particular technology baseprogram element (PE)15 (e.g., avionics) and theprojects in that element. However, such reviews arenot always effective. Due to manpower constraints,OSD can only conduct a limited number of S&Treviews each year. Further, since there is no uniform

OSD-wide format for conducting the reviews, themethodology and thoroughness vary greatly. Withover 200 coordinating groups producing a hodge-podge of different reports, it is very difficult for OSDto determine whether the resources of its technologybase program are being allocated wisely.

The IDA task force made three major recommen-dations for strengthening science and technologycoordination. The first is to establish a DoD-wideS&T Coordinating Group responsible for establish-ing 17 Technology Coordinating Panels (TCPs) forthe entire S&T program. Membership of the TCPswould consist of senior R&D managers from theServices, the agencies including DARPA, and SDIO.

The TCP panel members would be kept up to dateon the status of a particular technology, the justifica-tion for specific programs in which a technology isused, and why the users’ needs necessitate thepursuit of that technology. The purpose of the TCPpanels would be to reduce unwarranted technologyduplication, ensure that resources in a particulartechnology area are well balanced, identify potentialtechnology gaps, and identify critical long-lead-timetechnologies in a series of annual reports.

The second major recommendation is that OSD,the Services, the agencies, and SDIO develop aDoD-wide format for the annual TCP reports. DoDcurrently has no formal S&T reporting process for its200 coordinating groups. If these 17 TCP groups areto be effective, IDA believes, they should produceconsistent reports that outline important technologyactivities across all of DoD’s S&T activities,

The third, and final, recommendation is to absorband disband those existing coordinating groups thatare not needed to support the work of the TCPs. IDApoints out that each of the 17 TCPs would haveunder it a number of (existing) technology coordi-nating subgroups. For example, the TCP for Shipsand Submarines would have three subgroups: Hulls,Hydrodynamics, and Machining. Each of the sub-groups would be required to contribute to the TCP’sannual report. Those not needed for this processwould be disbanded.

l’llnslitute for Defense Analyses, op. cit., footnote 6, P. 3.

ISThc pE is the basic bui]ding block in DCID’S program planning and budgeting systcm (PPBS). There arc approximately 180 PEs in DoD’s entiretechnology base program, with each PE consisting of all costs associated with a research activity or weapon system.


Although these recommendations might help toimprove technology base coordination, getting themaccepted and implemented within DoD may prove tobe difficult. Each of the recommendations wouldhave to be approved by the DDR&E as well as theUSD(A) and then implemented by the Services andthe defense agencies. This is a process which in thepast has proven to be very difficult. For example,although DARPA was invited to participate in theIDA study, no DARPA representatives attended themeetings of the task force or participated in writingthe final report.

Distrust and misunderstandings among OSD, theServices, and DARPA are a major impediment toimproving S&T coordination. Efforts by OSD toimprove cooperation or coordination can be inter-preted as an attempt to tell the Services and DARPAhow to manage their science and technology pro-grams. Some OSD representatives believe that theServices will pursue an independent path whenpossible. Accordingly, improved coordination amongOSD, the Services, DARPA, and SDIO will bedifficult to achieve.

Acting as a Strong Advocate for DoD’sScience and Technology Programs

OSD currently lacks a strong defender of itstechnology base programs. A strong advocate wouldhave two primary responsibilities: 1) presenting acomprehensive review and defense of DoD-widetechnology base programs to Congress; and 2)acting as a strong proponent for the S&T programswithin the DoD.

The IDA task force concluded that there is nosingle individual within OSD who is responsible forpresenting and defending technology base programsbefore Congress or within DoD. The task forceindicated that the USD(A) should provide high-visibility advocacy for the S&T programs anddevelop a coherent DoD-wide position statement onthe technology base programs.16

The lack of an effective S&T advocate withinOSD has contributed to the erroneous perception, inCongress and even within DoD, that the technologybase programs have shared in the rapid growth of theRDT&E account. Between fiscal years 1984 and

1989, funding for DoD’s RDT&E programs in-creased 20 percent in constant dollars (see table 2).During the same period, however, research (6.1 ) andexploratory development (6.2) funding declined 3percent and 6 percent, respectively, in constantdollars. Between fiscal years 1984 and 1989, almostall of the growth in DoD’s S&T programs can beattributed to SDI. When the SDI figures are includedin DoD’s S&T activities, they present a distortedimpression of budgetary growth in the S&T pro-grams. The DDR&E testified before Congress thatthe rapid growth of the SDI budget has strengthenedthe technology base programs of DoD. By contrast,IDA task force members expressed the belief thatmost SDI efforts have been of little use to the rest ofDoD’s S&T programs.

In recent years, OSD has been unable to presenta comprehensive review of its technology baseprograms to Congress in a compelling way. Forexample, in DoD’s fiscal year 1987 RDT&E reportto Congress (the last year DoD produced such areport), all the major RDT&E goals were focused onshort-term objectives. The report did not makeadequate distinctions between technology base ac-tivities on one hand and development, testing, andevaluation activities on the other. OSD failed toconnect technology base advancements with thedevelopment of current and future weapon systems.Finally, the report provided no information on howfunding trends for technology base programs com-pared with the overall growth in RDT&E funding.

In some cases, OSD officials have not been ableto prevent the Services from shifting funds awayfrom their own S&T programs in order to supportmore immediate concerns such as procurement, or toprevent OSD from cutting technology base pro-grams. This is clear in budget reviews, and wasdemonstrated recently when the Army cut fundingfor its research program by almost one-third andcancelled its In-House Laboratory Independent Re-search (ILIR) program.

As illustrated by table 3, the Army-like the otherServices-supported consistent increases in its re-search program beginning in fiscal year 1980.However, in fiscal year 1987, when DoD facedbudget constraints, the Army cut its research pro-

lbInstitute for Defense Analyses, op. cit., footnote 6, P. ~.


Table 2—DoD Technology Base Funding, Fiscal Years 1984 Through 1989 (millions of 1982$)

% Change(constant $)

1984 1985 1986 1987 1988 1989 1984-1989

Basic Research (6.1 )a . . . . . . . . . . . . . . . . . . . . . . . . 778 760 831 756 740 755 - 3Exploratory Development (6.2) . . . . . . . . . . . . . . . . 2,051 2,032 1,984 1,985 1,924 1,928 - 6Advanced Technology Development Without SDI . 1,261 1,175 1,223 1,433 1,438 1,408b 12SDI office . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1,109C 1,243 2,318 3,156 2,957 2,849 157d

Total Technology Base With SDI . . . . . . . . . . . . . . . 5,199 5,210 6,356 7,330 7,059 7,191 38Total Technology Base Without SDI . . . . . . . . . . . . 4,090 3,967 4,038 4,174 4,102 4,342 6Total RDT&E . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24,829 27,371 29,322 30,464 30,568 29,663 19‘This category does not mcludehmding Iortha University Research lnihatives(URl) Program.~Thls figure does not include the transfer of $250 million ofOSD-managed projects to13ARfW.Ckcordmg to DoD, altfwugh SDIO was allocated $49 million to begin Its research actwitles, the three Services plus DARPA were akeadv conducting about $1.2 billion in SD1-related

research-in fiscal year l&34 (in 1984$).dReflectS $1 ,25cJ million of SD1-related work in fiscal War 1984.

SOURCE: office of the Secretary of Defense.

Table 3-individual Service Funding forResearch (6.1) (in millions of dollars)

AirYear Army Navy Force

1980 . . . . . . . . . . . . . . . . . . . . . . . . 130.7 214.9 119.21981 . . . . . . . . . . . . . . . . . . . . . . . . 144.4 241.4 126.61982 . . . . . . . . . . . . . . . . . . . . . . . . 179.2 276.5 147.41983 . . . . . . . . . . . . . . . . . . . . . . . . 206.2 307.6 166.41984 . . . . . . . . . . . . . . . . . . . . . . . . 216.5 320.6 191.41985 . . . . . . . . . . . . . . . . . . . . . . . . 231.5 341.2 201.31986 . . . . . . . . . . . . . . . . . . . . . . . . 250.3 342.3 210.21987 . . . . . . . . . . . . . . . . . . . . . . . . 219.5 354.3 223.31988 . . . . . . . . . . . . . . . . . . . . . . . . 168.9 342.1 197.71989 . . . . . . . . . . . . . . . . . . . . . . . . 172.7 355.3 196.4SOURCE: office of the Secretary of Defense.

gram by 12 percent and cancelled its ILIR programfor fiscal year 1988. In fiscal year 1987, ILIRcomprised about 7 percent of the Army’s researchbudget, or $16 million, spread among its 31 labora-tories on a competitive basis.

The ILIR programs serve a number of importantpurposes for Service laboratories. Because they area principal main source of discretionary researchfunds, the Service ILIR programs help the labsmaintain an atmosphere of creativity and researchexcellence, enhance their S&T base, provide seedmoney that can lead to new research efforts, andassist the laboratory directors in hiring new Ph.D.s.

In its 1987 summer study on technology basemanagement the DSB stated that ‘‘a successful

laboratory requires discretionary basic researchfunding for its long term vitality. ” The DSB went onto recommend that “at least 5 percent, and up to 10percent, of the annual funding of Federal laborato-ries” should consist of ILIR funds.17

In meetings with OSD the Army reassured theDDDR&E(R&AT) that funding for the researchprogram would be restored as soon as possible.However, in fiscal year 1988, the Army cuts itsresearch program an additional 23 percent, andcanceled the ILIR program. After several meetingswith top Army RDT&E officials, the DDDR&Edecided not to raise the 6.1 funding issue to theDDR&E level.

Promoting Cooperation Amongthe Services and DARPA

There is a long history of inter-Service rivalryand difficulty in cooperation between the Serv-ices and OSD. Further, cooperation between theServices and DARPA is hindered because DARPAreports to the DDR&E while the Service S&Trepresentatives report to the DDDR&E(R&AT).Starting with the National Security Act of 1947,Congress has taken a number of steps to strengthenOSD as a centralizing and coordinating body. Manyanalysts believe that these efforts have generally

ITU.S. ~p~ent of Defense, Office of tic DDR&E, Report of the Defense Science Board, 1987 Summer Study on Twhnolo~y Base Management”(Springfield, VA: National ‘lk.chnical Information Service, December 1987), p. 15.


been unsuccessful.18 It is probably too early to tell ifthe recent Goldwater-Nichols Act will be successfulin improving OSD centralization and coordinationcapacity.

Inter-Service rivalry is not necessarily all bad. Inhis book Bureaucracy and Representative Govern-ment, William Niskanen states that ‘‘competitionamong bureaus promotes efficiency by reducing thecost of certain services. ” Niskanen points out thatredundancy can guard against catastrophic failure ofone or more programs. 19

Certainly there are advantages in using competi-tive approaches. However, there are limits to theextent to which competition can contribute to thesuccess of the DoD S&T program. In an environ-ment where the rapid development and deploymentof a technology often is important, excessive compe-tition usually results in poor coordination, whichslows the introduction of new S&T capabilities.Representatives from OSD and the Services havestated to OTA staff that inter-Service rivalry oftenhas played a major role in delaying the developmentof important technologies (such as remotely pilotedvehicles).

There have been instances in which OSD, theServices, and DARPA have been able to overcomesome of these difficulties. One example is the recentestablishment of the Balanced Technology Initiative(BTI). Established by Congress in fiscal year 1987,the BTI is intended to develop new technologies to‘‘substantially advance our conventional defensecapabilities. ” The National Defense AuthorizationAct for fiscal year 1987 indicated that BTI fundswere to be used to ‘‘expand research on innovativeconcepts and methods of enhancing conventionaldefense capabilities,” and to promote “restorationof the conventional defense technology base. ”20

Responsibility for planning, development, and over-

sight of the BTI program was assigned to theDDR&E. The makeup of the BTI planning team,chaired by a representative from the office of theDDDR&E(R&AT), was unusual because it includedthe Services, DARPA, SDIO, and four other OSDorganizations: Tactical Warfare Programs, Strategicand Theater Nuclear Forces, International Programsand Technology, and the Undersecretary of Defensefor Policy. Funding for the program was appropri-ated by Congress to OSD to be divided among theServices, SDIO and DARPA.

Despite initial skepticism by the Services andDARPA, the BTI program appears to enjoy strong,although not universal, DoD and congressionalsupport. One reason might be that OSD did notdevelop program guidelines on its own and then askthe Services and DARPA to forward proposals basedon those guidelines. Rather, OSD made a deliberatedecision to include all of the interested parties in theprocess of developing the BTI guidelines. All of theparticipants played a role in the development of itsoverall goals, and knew that project selection wastied to the ultimate goals of the program rather thanjust technological competence.21 OSD officials triedto take advantage of European technological knowl-edge and capabilities during the initial planningstages. The BTI planning team was also successfulin designing broad project implementation andevaluation procedures. Finally, the BTI report toCongress tied each of the programs to a crucialcomponent of the air-land and maritime strategy, notto a program funding element; essentially, the BTIplanning team tied each project to a component ofthe conventional warfare doctrine.

It is still too early to evaluate the success of theBTI. Like other congressionally mandated pro-grams, the BTI was greeted with skepticism in theServices and OSD because congressional interestand funding support for such special initiatives often

‘8 Daniel J. Kaufman, “National Security: Organizing the Armed Forces, ” Armed Forces & Society, Mar. 16, 1988, p. 15.

lgwilli~ Niskanen, Bureaucracy and Representative Government (Chicago, IL: Aldinet Athefiom 1971).2ORo&fi C. ~ncm, DDR&E, Department Of Defense ,Ttatement on the Balanced Technology Initiative, presented before tie committee on ‘reed

Services, U. S. Senate, Apr. 11, 1988, p. 2.zlGuldellnes for selection included the followlng: 1) proJec[s had 10 bc consistent with the stated intent of Confless. 2) Emphasis had to be on

technology aeas that address recognized conventional force needs (e.g., chemical, biological defense, and nuclear programs were generally excluded).3) Projects should offer both short- and long-term potential for enhancing conventional force needs. Preference would be given to ongoing work thatoffered a high payoff in military effectiveness, with Iimitcd additional funding. 4) SDIO suggestions should be presented as technological spinoffopportunities with relevance to conventional defense needs (e.g., hypervelccity guns and projectiles, high-power microwaves, and advanced seekers andsensors). 5) A certain number of projects had to involve joint programs, such as Services/DARPA, multi-Service, or international cooperation.

Chapter 4-Planning and Funding DoD Technology Base Programs .51

have been fleeting.22 Nevertheless, rather than devel-

oping program guidelines and objectives independ-ently, OSD appears to have tempered this skepticismby creating an environment in which all interestedparties are willing to cooperate in the developmentand implementation of the BTI program.

Evaluating the Goals of theTechnology Base Programs

OSD and the Services have not developed asystematic, DoD-wide approach for determiningthe extent to which technology base programsactually satisfy goals set by OSD. OSD officialsand Service representatives typically describe twogoals of the S&T programs: maintaining technologi-cal superiority over the Soviet Union, and being asmart buyer of technology and technological exper-tise. Other technology base goals, such as reducingcomplexity and cost, improving productivity of theindustrial base, sponsoring the highest-quality S&Twork, and enhancing return on investment, receivecomparatively little emphasis.

Moreover, although there is seemingly a generalconsensus on what it means to keep ahead of theSoviet Union, there appears to be much less agree-ment regarding what it means to be a‘ ‘smart buyer. ”OSD and the Services appear to have no systematicway of determining whether they are smarter buyersof technology and weapon systems today than theywere, say, 10 years ago. It appears that OSD and theServices take it “on faith” that a sustained effort invarious S&T activities provides them with theability to make intelligent investments in S&T andweapon systems development.

Evaluating Research Activities

When OTA asked OSD representatives how theyspent their time, the responses focused on threethings. First, most of the respondents said they spendtoo much time on the long process of reviewing anddefending their programmatic budgets. Second,OSD personnel spend time responding to DoDinternal requests. These requests include technicalquestions, providing information for the DoD In-

spector General, responding to General AccountingOffice audits, or trying to prevent one of the Servicesfrom shifting funds away from an S&T program.Third, the respondents indicated that they spendmore and more of their time trying to satisfycongressional requests. According to a recent article,in 1970 Congress requested 31 reports or studiesfrom OSD. By 1985 that number had climbed to 458.Concomitantly, legal provisions detailing how DoDis to carry out certain aspects of its responsibilitieshave increased from 64 to 213, while annualcongressionally mandated actions requiring specificDoD compliance increased from 18 to 202.23 OSDrepresentatives gave the impression that they weredrowning in a sea of internal and external account-ability and bureaucratic red tape.

Many respondents indicated that they spent onlya small portion of their time performing duties thatrequire science and technology skills. It appears thattoo many OSD-as well as Service-R&D manag-ers are required to spend an inordinate amount oftime defending their budgets, responding to DoDbureaucratic red tape, or answering an ever-growingnumber of congressional inquiries, leaving littletime to evaluate R&D activities.

ORGANIZATION OF OSDFOR OVERSIGHT

DoD’s current organizational arrangement formanaging S&T activities contributes to the difficul-ties OSD encounters in shaping a coherent technol-ogy base strategy, and to the problems describedabove.

As a result of the Goldwater-Nichols Act, DoDhas reorganized the management of its RDT&Eactivities. The Act abolished the office of UnderSecretary of Defense for Research and Engineeringand replaced it with the USD(A). The legislationalso re-created the Office of the DDR&E, whichreports to the USD(A). (See figure 2.)

The USD(A) has oversight responsibility for allof DoD’s technology base programs, except those ofSDIO, That oversight responsibility is delegated to

ZzThe SeWices we generally satisfi~ with the conventional programs the BTI is currently supporting. However, Congress did nol Provide anYadditional funding for the BTI in fiscal year 1990. Consequently, OSD will fund the BTI program by taxing other conventional technology base effortsof the Services. The Services argue that OSD’S action in this instance has greatly compromised the original intent of the BTI program,

zsKaufman, op. cit., footnote 18, P. 5.


the DDR&E, who in turn delegates oversight of theServices’ programs to the DDDRE(R&AT). Thedirector of DARPA is supposed to work closely withthe DDDR&E(R&AT), but reports directly to theDDR&E.

The current DoD RDT&E organizational struc-ture raises a number of concerns. The first centers onthe primary responsibilities of the USD(A). ThePackard Commission stated that it was crucial forthe new USD(A) to have full-time responsibility formanaging the defense acquisition system, settingR&D policy, and supervising the performance of theentire process including procurement, logistics, andtesting. The Under Secretary is also responsible fordeveloping contract audit policy, supervising theoversight of defense contractors, and preparingannual reports to Congress on major issues ofacquisition policy and program implementation.With the procurement budget many times larger thanthe tech base budget, members of the defense R&Dcommunity are afraid that their concerns will take aback seat to the USD(A)’s broad menu of acquisitionresponsibilities.

Some OSD and Service representatives believethat it is too early to tell whether the technology baseprograms will suffer as a result of the reorganization.However, other DoD officials contend that S&Tprograms have already experienced some setbacks.They note, for example, that the USD(A) recentlyremoved the office responsible for internationalR&D cooperative programs from the DDR&E’soffice. The newly created Deputy Under Secretary ofDefense for International Programs and Technologyis still responsible for cooperative foreign R&Dprograms but now reports directly to the USD(A).However, according to the 1986 Nunn Amendment,by 1994, 10 percent of the RDT&E programs are tohave foreign involvement. Representatives of theoffice of the DDR&E believe that they should haveoversight responsibilities for those programs. Cur-rently there are about 20 foreign S&T cooperativeprojects.

A second organizational problem concerns thereestablishment of the DDR&E. The DDR&E wasoriginally established as part of the 1958 Depart-ment of Defense reorganization Act. The DDR&Ewas given greater responsibilities in 1977, andelevated to the Under Secretary of Defense for

Research and Engineering [USD(R&E)] as part ofSecretary of Defense Harold Brown’s managementreforms.

Some DoD officials contend that the reestablish-ment of the DDR&E as subordinate to the USD(A)might be interpreted as a lowering of status for R&D,since the DDR&E no longer has direct access to theSecretary of Defense. Various OSD representativeshave argued that if science and technology are thecornerstone of the military’s defense capabilities,then S&T programs should have direct access to theSecretary’s office. They fear that, without suchdirect access, important S&T issues such as coopera-tive foreign R&D programs will get lost in theacquisition shuffle. Others, however, assert that onlytechnology that gets fielded in military systems hasany value for defense, and that the DDR&E isappropriately placed under the USD(A). They claimthat the reestablishment of the DDR&E will notpresent a problem if the USD(A) strongly supportsthe S&T programs.

Managing Technology Base Activitiesat DARPA and SDIO

The Role of DARPA

A third concern is the role that DARPA and SDIOplay in supporting DoD’s technology base pro-grams. DARPA was established in 1958, partly as aresult of the launching of the initial Sputnik satel-lites. The President and Congress also recognizedthat DoD needed an organization that could “takethe long view” regarding the development ofhigh-risk technology. DARPA was thus setup to beDoD’s ‘‘corporate” research organization, reportingto the highest level, and capable of working at thecutting edge of technology. DARPA’s organizationallows it to explore innovative applications of newtechnologies where the risk and potential payoff areboth high, and where success might provide newmilitary options or applications-or revise tradi-tional roles and missions. In theory, since DARPAhas no operational military missions, it should beable to maintain objectivity in pursuit of researchideas that hold promise for important technologyadvancement for all of the Services,

DARPA executes its programs mainly throughcontracts with industry, universities, nonprofit or-ganizations, and Government laboratories. DARPA


now has a limited in-house contracting capability,but most of its contracts are still managed by theServices.

Although DARPA was originally established asa small agency to promote the rapid diffusion of newand creative ideas, in the past two years DARPA’sbudget has ballooned to over $1 billion, sponsoringalmost 25 percent of the military’s S&T work.24

Further, in 1986 DoD announced that DARPA’s roleas a developer of technology would include proto-typing.

The recent rapid growth in DARPA’s budget andits additional prototyping responsibilities haveraised several concerns within the defense commu-nity. First, OSD and Service representatives havehad problems coordinating technology base activi-ties with DARPA. They contend that part of theproblem is DARPA’s independence and its separatereporting chain within OSD. All three Services havecomplained that DARPA seldom involves them inits initial planning activities for joint DARPA/Service projects. The Services note that DARPAoften chooses not to participate in important technol-ogy base activities. Many experts believe that effortsto improve DoD-wide technology base planning andcoordination would require full participation byDARPA.

A second concern revolves around the changingnature of DARPA’s technology base activities. Ofthe $1,270 million DARPA budget for fiscal year1989, only $88 million is for basic research. Thereis growing concern, inside and outside DoD, thatDARPA may be supporting too much appliedresearch and technology demonstration activitiesrather than longer-term, high-risk basic research. Intestimony before the House Science and TechnologyCommittee’s Task Force on Science Policy, NormanR. Augustine, a member of the DSB and then theexecutive vice president of Martin Marietta corpora-tion, stated that:

In decades past, the effort of the [Defense]Advanced Research Projects Agency was focusedupon advancing basic research and applied research.Over the years since its inception, however, thefunds allowed to DARPA have been, to an increasingdegree, used for prototype demonstrations-a veryworthwhile undertaking in its own right but never-theless still a major drain on the basic researchresources originally intended at the time DARPAwas established.25

The changing nature of DARPA’s technologybase work leads to a third organizational issue:DARPA’s alleged past difficulties in transferringtechnology to the Services. DARPA is not the onlyorganization, public or private, to struggle withtechnology transfer problems, and over its 30-yearhistory DARPA has had a very impressive record ofsuccessfully transferring such technologies asstealth, directed energy, and some types of lasers tothe Services. Nevertheless, many OSD and Servicerepresentatives strongly criticized DARPA’s currenttechnology transfer activities, particularly with re-gard to prototyping and technology demonstrationprograms. This has taken on particular importance inrecent years, as Congress has turned to DARPA toaddress Service-related advanced technological prob-lems in such areas as anti-submarine warfare,anti-armor applications, and lighter-than-air tech-nology.

Two recent studies seem to reinforce the OSD andService technology transfer concerns. In 1985, atDARPA’s request, both the National Security Indus-trial Association (NSIA) and the Technology Trans-fer Center at George Mason University conductedstudies of the particular technology transfer processassociated with DARPA’s large technology demon-stration programs. Both of these studies concludedthat DARPA’s technology transfer activities rely toomuch on individual initiatives, resulting in a veryweak and haphazard approach to the technologytransfer process. The NSIA study noted that “DARPAis often too insensitive or unaware regarding theneeds and problems of the Services. ” The NSIA

240sD e5timates~at for fISCaI year 1$)89 DARPA’s budget will be about $1,270 million. Of that amount, $250 million will consist of work transferredout of OSD to DARPA. These projects include the SEMATECH initiative, the Monolithic Microwave Integrated Circuit (MMIC) program, the SoftwareTechnology for Adaptable Reliable Systems (STARS), and several other programs. The OSD projects will be primarily managed by OSD personnel whohave been transferred to DARPA from the disbanded Computer and Electronics lkchnology Directorate in OSD. The remaining $1,000 million(approximately) consists of about $700 million requested by DARPA and a total of about $300 million added on by Congress,

25Nomm R. Au~stinc, M~in Marietta COrp., testimony at hearings ~efl>re tie HOUSC Committm on science and T&hnology, Task Force on SciencePolicy, Oct. 23, 1985, pp. 3-4.


panel indicated that an “increase in the awarenessand sensitivity to the Services needs need notcompromise DARPA’s essential free thinking. ”26

Both studies recommended that DARPA developa written Agency-wide technology transfer plan forall technology demonstration activities. Amongother things, this plan should require all programmanagers to work closely with the appropriateService when undertaking a technology demonstra-tion project. Both studies recommended that DARPAmake available to all Program Managers a centralhistorical database of successful and unsuccessfultechnology transfer strategies based on actual pro-gram experience. The studies pointed out that giventhe short tenure of DARPA professional staff (about3 to 4 years), a written plan and central databasewould help improve the long-term continuity ofDARPA’s technology demonstration programs.

Despite these strong recommendations, DARPAhas not yet developed a formal technology transferplan. OSD and the Service representatives assert thatif DARPA continues to pursue its technologytransfer activities as it has in the past, many goodtechnological opportunities could be wasted. How-ever, this problem is not unique to DARPA; OTAhas not found any formalized written technologytransfer procedures developed by OSD and theServices.

The Role of SDIO

The Strategic Defense Initiative Organization(SDIO) was established as a separate agency of theDepartment of Defense in 1984, with the directorreporting to the Secretary of Defense. The SDIO’smission is to ‘‘provide the technological basis for artinformed decision regarding feasibility of eliminat-ing the threat posed by ballistic missiles andincreasing the contribution of defensive systems toU.S. and allied security. ”27

As indicated in table 3, the SDIO budget hasgrown rapidly over the past 5 years. Although theentire SDIO budget is contained within the 6.3Abudget category, not all SDI activities are advanced

technology demonstration efforts. For example,SDIO’s Innovative Science and Technology Office(ISTO) has the mission of establishing the feasibilityof revolutionary concepts with the potential forapplication to specific SDI technological needs.Like DARPA, the ISTO executes its researchcontracts through the Services. The ISTO estimatesthat in fiscal year 1988 and fiscal year 1989 it willsupport $100 million in SDI-related research.

SDIO also supports exploratory development(otherwise funded under category 6.2). Becausethere is no separate office that manages such work,it is very difficult to determine how much explora-tory development SDIO is currently funding. OTAasked SDIO if it could determine the dollar amountof research and exploratory development projects itsupports on an annual basis. According to the SDIOcomptroller, SDIO does not fund any true researchactivities. 28

SDIO's research efforts do not matchDoD’s accepted definition of research. However, thedirector of ISTO indicated that 80 percent of theprojects his office supports do qualify as research.OSD and Service representatives agree that ISTOsponsors short-term research programs, with heavyemphasis on solving specific SDIO challenges.

The comptroller maintains that all of ISTO’sefforts really fall under the definition of exploratorydevelopment. The report concludes that approxi-mately 20 percent of the SDIO budget is devoted toexploratory development work.

SDIO is funding about $700 million of explora-tory development work with no formal coordi-nating ties to the three Services and DARPA.Presently, the only coordinating activities occurinformally, as the individual Services and DARPAmanage SDIO’s exploratory development contracts.OSD and Service representatives have stated thatSDIO should participate in OSD’s technology baseinvestment strategy activities. Taken together, thecurrent organizational arrangement and the missionof the SDIO program make such participationunlikely. According to the Air Force, however,SDIO projects conducted in Air Force laboratories

LbNational security Industrial Association, “DARPA’s Technology Transfer Policy,” December 1985, p. 7.zTGcra]d Yonas, Acting Deputy D1re~tor and Chief Sclen(lst of SDIO, “The Stra[egic Defense initiative Science in the Mission Agencies & Federal

Laboratories,” testimony at hearings before the House Committee on Science and Technology, Science Policy Task Force, Oct. 23, 1985, p. 543.18Thc Conlpuoller of SD1O provided OTA with a written estimate of how, milch technology base work SDI() is current]y supporting, by Ciltf2gO~.

Chapter 4--Planning and Funding DoD Technology Base Programs .55

differ from DARPA projects because they are wellintegrated into the laboratory program. Such coordi-nation and integration usually occurs at the labora-tory (or lab division) level.

Recruiting and RetainingScientific Personnel

According to OSD and Service representatives,DoD is often unable to recruit the very bestscientific, technical, and managerial talent. Be-cause of growing salary disparities between thegovernment and the private sector, OSD is losingmany top level S&T managers.

The late Philip Handler, former president of theNational Academy of Sciences, once pointed outthat in science the best is vastly more important thanthe next best. Both the 1983 Packard Report and the1987 DSB report concluded that OSD and theServices face serious disadvantages in hiring andretaining top S&T personnel for three primaryreasons: inadequate civil service compensation,‘‘revolving door” restrictions, and a lowering ofstatus associated with Federal employment.

A recent unpublished Navy study found that sincethe early 1980s, the disparity between Federalsalaries and salaries in industry and academia hasgreatly expanded. For example, the average com-pensation for S&T managers in the upper 10 percentof the private sector was $40,000 to $50,000 higherthan for their Government counterparts. Anotherinternal Navy survey of university principle investi-gators (PIs) found that, for the first time, the majorityof PIs’ salaries were higher than government sala-ries. Some 60 percent of university PIs are paidsalaries that exceed the Federal pay cap, withapproximately one-third of them exceeding $90,000.

This problem is likely to become more pressingin the future. Changing demographics will producea work force with greater ethnic diversity and morewomen. Minorities and women have not contributedin substantial numbers to science and engineering inthe past. The challenge will be to expand the

participation of women and minorities in scienceand engineering college programs and graduateschools and, ultimately, to offer them rewardingcareers working in defense technology base andrelated program areas.29

Salary disparity has also contributed to a highlevel of turnover among top-level OSD politicalappointees. For example, between 1981 and 1988there were three different USD(R&E) officials, andfive individuals have held the position that is nowDDDR&E(R&AT). An internal OSD study indi-cated that the overall quality of its S&T politicalappointees was very inconsistent.

Pre-employment and post-employment personnelrestrictions also mitigate against recruiting first-ratepolitical appointees. Such officials are required todivest themselves of any financial interest in anycompany conducting business with DoD. Thisrequirement can result in serious tax consequencesfor the political appointee. Further, many prospec-tive employees resent the prospect of filing anannual financial disclosure statement.

The main postemployment restriction concernsthe recently amended “revolving door” legislation.The revised law restricts the kinds of services formermilitary officers and DoD employees may performfor a future employer that does business with theDefense Department. Among other things, this lawimposes a‘‘2-yearban on certain former Departmentof Defense personnel receiving compensation ofmore than $250 from defense contractors (who havecontracts in excess of $10 million with the govern-ment) if the former officers or employees hadofficial procurement duties relating to that contrac-tor during the 2-year period prior to separation fromgovernment service. ”30

According to OSD and Service representatives,the revolving door legislation has significantlylimited DoD’s ability to hire top-level S&T manag-ers from the private sector who have had experienceworking in the defense arena. Compared with theirpredecessors, many top-level S&T managers now

Z9U.S. Congess, Ofiicc of Technology Assessment, Educating SLienti,st.s and Engineers: Grade School to Grad School (Washington, DC: U.S.Govemmcnt Printing Office, June 1988); and U.S. Congress, Officeof’Technology Assessment, Higher EducationforScienti.!ts and EnguzeerQackgroundPaper, OTA-BP-SET-52 (Washington, DC: U.S. Government Printing Office, March 1989).

~~Jack Maskel], Library of Congress, Congressional Research SCWiCC, ‘‘Post Employ rncnt ‘Revolving Door’ Restrictions on Department of DefensePersonnel,” July 5, 1988, p. 3.


come to DoD with little or no defense experience.This situation has contributed to the increasingperiod of time it takes new DoD S&T managers tounderstand the complexities of the overall defenseenvironment.

DoD SUPPORT OF ITSTECHNOLOGY BASE PROGRAMS

In 1953, President Eisenhower said that, despitethe establishment of the National Science Founda-tion, Federal agencies such as DoD would have tocontinue performing and supporting basic researchclosely related to their missions. Since then DoDofficials have asserted that basic research providesinformation on natural phenomena that DoD coulduse in the development of modem weapons.

In 1963 Harold Brown, then DDR&E andsubsequently Secretary of Defense, said that ‘‘as thelargest user of scientific and technical information inthe Federal Government, DoD had an obligation toreplenish this information. ” Brown went on to say

that DoD has to support a broad range of researchthat may or may not be directly related to itsmission. 31 DoD representatives contend that in anera of rapid technological change and growingSoviet S&T competence, DoD support for a strongand diverse technology base program is imperative.

Despite these strong statements of support,funding for DoD’s S&T programs has been inconsis-tent over the past 20 years. As table 4 indicates,beginning in 1970, funding for research (in constantdollars) began to decline and did not exceed its 1970level of support until 1986. Moreover, since the peakyear of 1986, funding for research has declined morethan 4 percent in constant dollars.

Similarly, support for exploratory developmentdeclined until the late 1970s. Then it rebounded,nearly returning to its 1970 level by 1983—inconstant dollars, Between 1984 and 1989, however,support for exploratory development again fell byalmost 10 percent.

By almost any measure—total constant dollars,fraction of DoD budget, fraction of RDT&E budget—

the level of DoD support for its research andexploratory development programs has decreasedover the past 20 years. In the mid-1960s, researchand exploratory development represented 25 percentof the total RDT&E budget. By 1989 it had shrunkto less than 9 percent. Between 1970 and 1988, 6.1and 6.2 funding declined as a percent of DoD’s totalobligational authority (TOA), from 1.79 percent to1.27 percent. Further, as table 4 shows, since 1983DoD has moved its resources from research andexploratory development programs to advancedtechnology development (ATD) programs.

Between 1984 and 1989, constant dollar fundingfor 6.1 and 6.2 programs declined 3 percent and 6percent, respectively (see table 2). During the sameperiod, funding for advanced technology develop-ment (ATD) exclusive of SDI increased 12 percent,while support for the SDI program rose 157 per-cent,32

The recent rapid growth of both ATD and SDIprograms has taken its toll on the basic research andexploratory development programs. OSD and Serv-ice representatives have indicated that DoD isputting greater emphasis on ATD activities toimprove the transfer of new technology to weaponsystems. For example, in 1984 DoD reduced fundingfor its exploratory development program by $300million while increasing ATD by $500 million.According to OSD personnel, the switch in fundingwas nothing more than an accounting change: areview of the programs supported under exploratorydevelopment revealed that a good portion of thework should have been classified as ATD.

In testimony before the House Armed ServicesCommittee, the DDR&E, Dr. Robert Duncan, saidthat the growth in the ATD program and the SDIprogram has offset the losses in research andexploratory development. However various OSDand Service representatives contest this statement,insisting that those technology base activities whichSDI currently supports are aimed exclusively atsolving SDI-related problems. Consequently, poten-tial benefits flowing from SDI into technology baseprograms will be long term, and probably more

~lRa]ph Smders (cd.), “Re~~ch: Meting of the Term,” in [)efemre Research and Development (Washington, DC: Industial college of tie A~cdForces, 1968), p. 73. ‘

s2This calculation is based on DoD info~ation that the three Scrviccs and DARPA were supporting about $1.2 billion in SDI-related resemch in 1984.


Table 4-DoD Technology Baae Funding Trends (millions of 1982$)

AdvancedExploratory Technology Total Total

Research Development Development without withYear (6.1) (6.2) (6.3A) SDIOb SDI SDI

1970 . . . . . . . . . . . .1971 . . . . . . . . . . . .1972 . . . . . . . . . . . .1973 . . . . . . . . . . . .1974 . . . . . . . . . . . .1975 . . . . . . . . . . . .1976 . . . . . . . . . . . .1977 . . . . . . . . . . . .1978 . . . . . . . . . . . .1979 . . . . . . . . . . . .1980 . . . . . . . . . . . .1981 . . . . . . . . . . . .1982 . . . . . . . . . . . .1983 . . . . . . . . . . . .1984 . . . . . . . . . . . .1985 . . . . . . . . . . . .1986 . . . . . . . . . . . .1987 . . . . . . . . . . . .1988 . . . . . . . . . . . .1989 . . . . . . . . . . . .

779728712629579530528556576608653660697754778760831756740755

2,4182,2382,4142,3062,1261,9231,9021,9471,9371,9722,0212,1342,2332,3572,0512,0321,9841,9851,9241,928

————567631677734697725676600738792

1,2611,1751,2231,4331,4381,658

——

—

—1,1091,2432,3183,1562,9572,849

3,1972,9663,1262,9353,2733,0843,1073,2373,2103,3063,3503,3933,6683,9034,0903,9674,0384,1744,1024,342

——————————————

5,1995,2106,3567,3307,0597,191

aThe6.3Acategory wasestablishedin 1974.bEstablishedm 1984.

SOURCE: Office oftheSecrWary ofoefense.

expensive than if they had been supported directlythrough S&T activities.

The IDA task force stated the consequencesbluntly:

If the decline in resources devoted to science andtechnology is not reversed, the impact on the relatedtechnological capabilities of U.S. weaponry andforces maybe compromised so much that we willneed to rethink our basic strategy of using qualita-tively superior weapons to offset numerical disad-vantages. 33

While DoD’s RDT&E program has experiencedsignificant growth in the 1980s (see figure 3) theS&T portion of the budget has not shared in thatexpansion. Between 1980 and 1989 the RDT&Ebudget increased 90 percent in constant dollars,while the S&T programs (excluding SDI) increasedonly 16 percent.

According to OSD and Service representatives,two primary reasons explain this relatively small

increase. First, technology base programs do notenjoy strong support at the highest levels within themilitary. Similar findings were reported by the IDAtask force, and the DSB in its 1987 summer study ofDoD S&T programs. In its report the DSB stated:

Where once OSD exerted a centralized point ofunified leadership and budgetary authority andcontrol for the 6.1 program, the Study Group isconcerned that this leadership is fragmented bydelegation to the Services and agencies; the 6.1program has, in effect, been relegated to a positionof second or third order of importance and lacks topmanagement attention. Stated bluntly, DoD “corpo-rate management” has essentially abrogated some ofits responsibility for long range vitality and competi-tiveness.34

OSD and Service representatives believe thatmilitary leaders do not appreciate the role that pastS&T accomplishments have played in providingtechnologically superior weapons. Top Pentagonleaders are often willing to sacrifice 6.1 and 6.2

qqlnsti[utc for Defense Analyses, op. cit., fOOtnOte 6, P. 4.

34U.S. Dep~ment of Defense, op. cit., fOOtIIOte 17, p. 13.


Figure 3-Comparison of RDT&E andTechnology Base

Billions 1982$353 0 -2 5 -2 0 :15-10-5 -0 ( I 1 1 1 1 1 I [ I I I I 1 1

1970 75 8 0 8 5Year

— Technology base — Total RDT&/3E— Tech. base & S D I

SOURCE: Data provided by the Office of the Secretary of Defense,

activities in order to protect budgets for immediateand visible needs, such as tanks and planes. Theconsequences of reducing the force structure, termi-nating weapon systems, or delaying procurement aremuch more visible than a particular research pro-gram which may not bear fruit for 10 to 15 years, ifever. Unfortunately, this attitude has not served thetechnology base programs well in times of tightbudgets. It has resulted in military leaders ‘raiding”S&T programs to help pay for downstream systemdevelopment programs. As was noted earlier, therecent growth in the ATD budget to improve DoD’snear-term technology transfer concerns has come atthe expense of the research and exploratory develop-ment,

Focusing on near-term defense needs in resolvingbudgetary conflicts tends to bias the subject matterof DoD research. There is considerable agreementwithin OSD and the Services that much of DoD’sresearch program is aimed at meeting the short-termneeds of the military, and that it is easier to obtaintop-level support for research activities that can berelated to specific military needs. This is a point ofcontention. Many of those actively involved in theS&T programs believe that this is a misuse ofresearch funds. They contend that it is unwise todirect the research program toward the solution of

near term problems because military utility cancome from all areas of science and engineering. Bythis logic, it is in DoD’s best interest to be involvedin a wide range of research problems, to follow theirprogress carefully, and to apply long term scientificreserch to present and future needs.35

A recent internal OSD evaluation of DoD’sresearch programs concluded that many of itsresearch projects in such fields as mathematics,chemistry, computer sciences, and physics were‘‘too well connected to current military needs. ” TheOSD review instructed Service 6.1 managers that“you should be reaping the fruits of seeds sown byyour predecessors, and you should be sowing theseeds which will bear fruit for your successorsseveral times removed. ”

This very same concern was discussed in theDSB’s 1987 summer study on the technology base.The DSB report concluded, “The need for short-term results and immediate ‘relevancy’ has becomethe governing criterion in framing a program. Wehave experienced a ‘research menu squeeze’ inwhich the most popular programs, justifiable interms of clearly perceived near-term military rele-vancy, survive the cut.”36 The DSB report urged theServices to pursue more research activities withlonger term objectives. On the other hand, someargue that basic research is funded within DoDprecisely because the Defense Department can giveit a focus that makes it relevant to military needs.

CONCLUSIONThere is a serious question as to whether OSD is

currently fulfilling its technology base oversightresponsibilities in a satisfactory way. There isgeneral agreement, inside and outside of the Penta-gon, that OSD has not developed an overall technol-ogy base investment strategy. Many within theServices contend that, for a number of reasons, OSDshould not attempt to develop a coordinated technol-ogy base investment strategy, and that the currentdecentralized system is probably better. But othersassert that such a large technology base program,with important national security implications, ought

35 George Gamota, ‘‘How Much Does the Defense Department Advance Science?” in proceedings of an American Association for the Advancementof Science (AAAS) symposium, Naval Research Center, Washington, DC, Jan. 8, 1980 (published Sept. 24, 1980), p. 4.

3SU.S. Dep~ment of Defense, op. cit., footnote 17, p. 12.

Chapter 4-Planning and Funding DoD Technology Base Programs . 59

to possess some overall central leadership andguidance.

Implementing an OSD-guided investment strat-egy would not be a panacea for all the challengesconfronting DoD’s technology base programs. Acoordinated investment strategy could: 1) helpcreate a process for making OSD-directed strategicdecisions, 2) allow OSD and the different agenciesto focus on the outputs of the S&T programs and notjust the inputs, and 3) enhance the understanding ofDoD technology base programs.

A coherent technology base investment strategywould assist Congress in its review of defense S&Tprograms. In the absence of a clearly articulatedtechnology base strategy, Congress is forced tofocus its review on numerous individual programelements. A technology base strategy that includeda rational list of priorities would enable Congress totake a broader view of the Pentagon’s S&T pro-grams. Congress might then focus its attention onthe extent to which DoD’s proposed technology baseprogram satisfies its overall strategy and statedpriorities.

Despite the Goldwater-Nichols Act, OSD’scurrent organizational arrangement presents prob-lems for coordinating the different technology baseprograms. Without the full participation of DARPAand SDIO, a coherent technology base program willbe very difficult to achieve.

Clearly there is no magic formula for DoD to usein determining the ‘‘right” level of support for itstechnology base programs. After numerous discus-sions with individuals outside and inside the defensecommunity, OTA has identified several criteria thatmight usefully be applied to evaluating the overallstrength of DoD’s science and technology programs.

First, it is essential for an organization tomaintain strong support for a broadly based scienceand technology program, Top corporate managers,responsible for maintaining the overall health oftheir science and technology programs, must have adeep understanding of how a strong technology baseprogram can help an organization attain both its

short and long term S&T goals. DoD’s technologybase programs do not enjoy consistent high-levelsupport within OSD and the Services. An organiza-tion’s research program should be strong and diverseenough to attack any problem related to the organi-zation’s mission. As the director of research for alarge industrial corporation told OTA, he wants hisS&T people to be “swimming in a sea” ofcompany-t-elated research problems.

Second, individuals responsible for managingS&T programs need a clear mission statement thatguides the overall makeup of the S&T programs.The mission should be developed by a criticalnumber of people throughout the organization andunderstood by all. DoD asserts that the primarymission of its S&T programs is to offset thenumerical advantages and growing technologicalsophistication of Soviet forces. But recent studiescriticize DoD for focusing too strongly on the SovietUnion, arguing that the military must be prepared toengage in a number of different combat arenas.37

There is little agreement within OSD and theServices on how the technology base programsshould be structured to meet the diverse securitychallenges that will confront DoD in the future.

Third, a strong S&T organization must be able torecruit, hire, and invest in the very best S&T talent.These new people should be exposed to a strongorientation program that helps them understand howtheir work will contribute to attaining the overallS&T mission. In order to conduct a vital S&Tprogram, DoD must achieve the ability to recruit andretain top flight scientists and engineers,

Fourth, many researchers, both inside and outsidethe Pentagon, contend that DoD needs to maintaingreater funding stability for its technology baseprograms. This is especially true for the early stagesof research activities.38 DoD's research and explora-tory development programs have suffered since theestablishment of SDI. Over the last six fiscal years(1984-89), DoD has been the only major FederalR&D sponsor to experience a funding decline, inconstant dollars, for basic research. A continuation

37scc, for Cxmple, Fred C, lkle and AlbtwI Wohlstetter, ‘‘Discriminate Deterrence,” Report of The Commission On Integrated brig-Term Strategy,Jan. 11, 1988.

SW.S. Dep~men[ of Defense, op. cit., footnote 17. P. 11.


of these trends could jeopardize a pillar of U.S.defense strategy.

The director of research at a Department ofEnergy (DOE) laboratory speaks of “recoveryresearch. ” When an organization fails to support abroadly based research program, it often experiencesdifficulty with new products as they move intodevelopment. Consequently, in order to correct suchproblems, the organization is forced to engage inrecovery research, which is costly and time-consuming. The DOE official stated that the more anorganization has to perform recovery research, thegreater the probability that its S&T programs are notreceiving enough support. An OSD official toldOTA that he believes that DoD has to support toomuch recovery research.

Finally, a strong S&T program must be closelycoupled to the developers and ultimate users oftechnology. This is an important avenue of commu-nication for managers to ensure that their S&Tprograms are solving the right problems, SomeService officials complain that technology basepeople are not always consulted when new weaponspecifications are developed. For example, ArmyS&T representatives told OTA that they were notconsulted when the Light Helicopter Experimentalprogram specified an automatic target recognitioncapability (ATR). According to these officials, theyknew that an ATR capability was (and still is) notfeasible. The Army now refers to this concept asaided automatic targeting recognition (AATR).

Chapter 5

The Management of DefenseDepartment Laboratories

CONTENTSPageINTRODUCTION ....... + . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

63GOVERNMENT-OWNED , GOVERNMENT-OPERATED LABORATORIES . . . . . . . . 64. . . . . . . . . . . . . .

Justifications for In-house Work . . . . . . . . . . . . . . . . . . . . . . . . . .. .. .., ., .. ... 64Personnel Management. ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67Laboratory Management Issues: Funding. .. .. ... ... 72

GOVERNMENT-OWNED, CIBTRACTOR-OPERATED FACILITIES:AN ALTERNATIVE MODEL . . . . . . . . .... . . . . . . . . . . . . . . . . . . . . . . . . . 74

.,...,. . . . . . . . . ,,. 74The GOCO Relationship at the Department of Energy: Contractual Arrangements . . . 75 . . . . . . . . . . . .

The GOCO Relationship at the Department of Energy: Complying WithFederal Norms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . 76. . . . . . . . . . . . . . . . . .

78S U M M A R Y A N D C O N C L U S I O N S .. .. .. .. .. .. .. .. .. ... ..... : 80

Chapter 5

The Management of Defense Department Laboratories

INTRODUCTION

Three problems—personnel, funding, and size—hinder virtually all of the laboratories operated byDepartment of Defense (DoD) employees in per-forming inherently governmental functions, actingas smart buyers, and incubating new concepts. Thegovernment’s personnel system is too rigid becauseit makes it difficult for laboratories to compete withthe private sector for professional staff, because payscales are inflexible, and because of the obstacles itsets to rewarding performance. The laboratories findit difficult to get funds as and when they need them:They must spend their funds within the 12-monthbudgetary cycle, and they have limited authority tomove money between accounts, approve start-ups,and target discretionary money to building theirtechnology base. Finally, most labs often cannotdeploy a critical mass of scientists and engineers intonew areas that may be vital to the lab’s mission.

Alternative models to the government-owned,government-operated laboratory exist, for examplethe facilities operated under contract to the Depart-ment of Energy (DOE). Such models have certaincommon features. They operate under the contrac-tor’s own management systems. Their personnelsystems enable them to compete on almost equalterms with universities and industry for scientistsand engineers. Their funds are often available untilspent. And they have the size and depth of expertiseto work in the various disciplines needed fortechnology development.

The question that DoD and its technical managersconfront is whether to continue the current system orto restructure the laboratories. As grave as the labs’problems are, conversion to government-owned,contractor-operated facilities (GOCO) may not bethe answer. A conversion to GOCO could improvethe laboratory’s operations in the short term, whileleaving its basic problems unchanged. No institu-tional approach can be divorced from the ends theinstitution is supposed to serve. The military Serv-ices must first decide what purposes their laborato-

ries serve before taking the next step of alteringlong-standing institutional arrangements.

This chapter describes, compares, and contraststhe basic management arrangements of DoD’sgovernment-owned and -operated laboratories withthose of comparable facilities, whether government-owned and contractor-operated or federally fundedresearch and development centers (FFRDCs). It alsoraises a fundamental question: Why does the govern-ment in general, and the DoD in particular, need todevelop technology through its own laboratories?

The problems of DoD’s in-house laboratories arewell documented. The next, and more difficult, taskis to take the argument a step further: to considersome alternative approaches to technology develop-ment. The Defense Science Board (DSB) did this inits 1987 summer study, recommending that undercarefully specified conditions some labs considerconverting to a GOCO model But the costs andbenefits must be carefully weighed. After ail, aGOCO military lab would still be dealing with thesame sponsor whose rigidity prompted the conver-sion in the first place. Additionally, no government-funded institution can escape oversight merely byconverting to contract. The reason is simple: Whe-ther government-operated or GOCO, operating fundsultimately derive from congressional appropria-tions, and Congress holds the senior officials of thesponsoring agencies accountable for their properuse.

Beginning with a look at the roles of in-housemilitary laboratories, this chapter explores the sys-temic problems they face in getting work done—problems of personnel management and develop-ment, starting and completing work, relations withthe sponsoring agency, and the like. The discussionthen turns to an alternative model, the multiprogramlaboratories operated under contract to the DOE.The DOE national laboratories merit close study,first, because the relations between what is nowDOE and its contractors have endured over fourdecades; second, because these labs seem to have the“critical mass” needed to bring very large technol-

IDefense Science Board, “Report of the Defense Science Board 1987 Summer Study on Technology Base Management, ” Washington, DC,December 1987.

-63-

64 • Holding the Edge: Maintaining the Defense Technology Base

ogy development programs to fruition; and third,because the labs and their sponsoring agency haveused the concept of ‘‘work for others” to redeployprofessional staff as projects wind down, and tomove into areas contiguous to their principal mis-sions.

After outlining both approaches to technologydevelopment, this chapter assesses both kinds ofinstitutions with respect to five topics: 1) manage-ment flexibility, 2) the extent to which GOCOinstitutions tend to become more like governmentlabs over time, 3) the ability of both kinds ofinstitutions to transfer the results of their 6.1-6.3programs to user organizations, 4) their mechanismsfor diversifying within their basic missions, and 5)the ability of government labs to assimilate the moresuccessful features of GOCO institutions.

GOVERNMENT-OWNED,GOVERNMENT-OPERATED

LABORATORIES

Justifications for In-house WorkGovernment-owned and operated facilities can be

justified for many reasons. First, apart from the issueof whether such labs serve as smart buyers, performinherently governmental functions, or provide tech-nical assistance for fielded systems, the relation oflaboratory to sponsor is more clear cut than it is forother arrangements. Government operation avoidsthe criticism sometimes made of GOCOs that thegovernment does not know whether a lab’s execu-tives identify with the government or the contractorwho pays their salaries. As one authority notes,‘‘in-house laboratories can be expected to share thesense of mission of their agency and to be responsiveto their needs. ” Such labs provide stability andcontinuity “by simply continuing arrangements thathave evolved historically. ”2

But such justifications skirt the important ques-tion: Why should technology for weapons systems,space exploration, or measurement protocols bedeveloped out of government-staffed labs at all?

There are five principal justifications for suchfacilities:

●

●

●

●

●

that certain functions are inherently govern-mental and may not be delegated to others,that the lab serves as a smart buyer, evaluatingits contractors and keeping them at arm’slength,that, through basic research, the lab can origi-nate new concepts that, with support from itssponsors, may develop into fielded systems;that the lab can do special-purpose work formilitary customers that either is of no interestto industry or is kept from industry for securityreasons, andthat the lab can provide support to users once aweapons system has been successfully fielded.

Moreover, in-house laboratories can react quickly tomilitary emergencies, as the Naval Research Labora-tory’s (NRL’s) recent support for the fleet in thePersian Gulf shows.

Inherently Governmental Functions

The concept of ‘inherently governmental” func-tions is perhaps the fundamental justification fortechnology development institutions run by civilservants. As Budget Director David Bell toldCongress in 1962, there are certain functions thatmay not be contracted out, functions that include:

the decisions on what work is to be done, whatobjectives are to be set for the work, what time periodand what costs are to be associated with the work . . .the evaluation and responsibilities for knowingwhether the work has gone as it was supposed to go,and if it has not, what went wrong, and how it can becorrected on subsequent occasions.3

This position has important implications for theconduct of research at military laboratories. In thisview, a laboratory or research and developmentcenter should have the capability to conceive ofweapons development projects, develop technicalspecifications for industrial contractors, and super-vise contractor efforts to ensure the reliability ofsystems and components in the early stages of

ZT.J. W;] btis, “Domes[ic Models for National Laboratory Ulili.mtion,” in Energy Research Advisory Board, Final Report of the MultiprogramLuborafory Panel, Vof. //: Support Studies (Oak Ridge, TN: Oak R]dgc National Laboratory, September 1982), p. 63.

‘3 David Bell, Bweau of the Budget, ~~ykrtem$ Deve/opmen/ u~ Ma~g~rnent, testimony at hc~ings before the HOU,SC Committee on GovernmentOperations, Military Operations Subcommittee, June 21, 1962, 87th Congress, 2nd Session, p. 44.

Chapter S--The Management of Defense Department Laboratories . 65

development, regardless of cost.4 Carried to itslogical conclusion, this view holds that governmentscientists and engineers are a national asset that,within broad limits, should be retained beyond theimmediate programs for which they were hired; thatthe lab is a going concern, not a job shop; and thatits professional staff must do basic and exploratoryresearch, simply to evaluate work done outside itswalls.

This basic philosophy is perfectly compatiblewith a number of arrangements. Government labora-tories exist in every phase of dependence on theirprime sponsor. A lab may: work exclusively for onesponsor or for several; perform reimbursable workfor other organizations, as many service labs aredoing for the Strategic Defense Initiative Organiza-tion (SDIO); serve as a corporate laboratory over andabove its responsibilities to its parent agency; or(what often comes to the same thing) do work that isonly loosely coupled to its sponsor’s missions.There are even cases, like the optical facility at theAir Force Weapons Laboratory (Kirtland Air ForceBase, New Mexico), of small GOCO units embed-ded in a government-run engineering center. Fur-ther, many military and civilian labs, particularly atthe National Aeronautics and Space Administration(NASA), have chosen to contract out virtually all oftheir support functions-functions ranging fromcarting trash to managing the cafeteria to program-ming and operating tracking stations—without com-promising their principal functions.

Smart Buyer

The laboratory may also function as a smart buyer,a role that complements its mission to plan develop-ment projects. The lab acts as smart buyer when itdevelops a particular technology-say, a new kindof integrated circuit, wideband recording device, orfault-tolerant avionics—that it can hand over to acontractor for further development and production,or when it evaluates private sector developments. Ineffect, the lab’s R&D work presupposes, and de-pends on, the existence of a strong private sectorR&D infrastructure. At this level, the justificationfor research is that a lab cannot assess technologywithout being thoroughly knowledgeable. As one

DoD engineer put it, with some exaggeration, ‘if wego to a symposium and see something new, we’renot doing our jobs. ”

In another sense, the in-house lab serves as asmart buyer when it proves a concept that may leadto new technology. The job of an engineering centerworking on, for example, very-high-speed-integrated circuit (VHSIC) technology is not tomake the systems work, but to show that they willwork. Actual operational success is in the hands ofthe buying commands, systems developers, andproduction people. A military service will insertVHSIC or other technologies only where the tech-nology can ‘‘buy its way” into a weapons system.

Where a lab really acts as smart buyer is inbringing its expertise to bear in deciding when workin a certain area has gone as far as it should. As onetechnical director put it, his job “is to kill offprojects that will not fly before costs get out ofhand. ”

Long-term Research

The laboratory also serves as an incubator of newconcepts. In fact, it is this role that serves as theprincipal justification for much 6.1 and 6.2 workcarried out by DoD institutions. It is more obviousat NRL than at the other Navy R&D Centers—butnowhere is it insignificant. If NRL has become acorporate lab for all of DoD, it is because of its workin technologies that had no immediate applicationbut that would ultimately define the technology fora new generation of weapons systems. NRL’s workin a variety of fields-designing x-ray astronomyexperiments, developing a unique class of electroac -tive polymers, perfecting ceramic-air composites forunderwater sensors—positions the Navy to moveinto the development phase, confident that thetechnology to make systems work is available.

The importance of such advanced work may begauged by the efforts of lab technical directors toincrease their pot of 6.1 and discretionary funds—though it should be noted that the two are by nomeans the same. Most of them would agree, in thewords of one of them, that the government ‘shouldsupport some tech base work that is independent of

@n his point, sw Feder~ Coordinating Council for Science, Engineering and Technology, ‘‘A Research and Development Management Approach:Report of the Committee on Application of OMB Circular A-76 (o R& D,” Washington, DC, Oct. 31, 1979, pp. VI-VII,


any particular program, because otherwise you mayshut out some technologies that could become veryimportant. ”

The sponsoring of basic research at militarylaboratories serves a number of ends. Basic researchcan be justified as a means of enhancing the lab’sreputation; indeed, most facilities try to hire the bestscience and engineering graduates by holding outthe possibility of their doing some basic research.Such research makes scientists available to engi-neers to work on serious technical problems, withoutgoing into development work full time. Finally,basic research allows an engineering-oriented facil-ity to develop a few special applications to militarytechnology.

Yet the laboratory executives are caught on thehorns of a dilemma. To try to justify basic researchon the ground that it will lead to some definite payoffis self-defeating, especially in an environment whereeverything militates against risk-taking. Basic re-search can be justified because it helps to define thetechnology out of which weapons systems maydevelop; new defense systems are made up offragments of new defense technologies coupled to anexisting base. According to this view, 6.1 work is the“push” that really changes the technology base,with 6.2 and beyond as the “pull.” Nor need basicresearch always precede the product developmentcycle; a military microelectronics facility producingcustomized chips for military customers, for exam-ple, may do fundamental research into the propertiesof matter as part of its design program.

Special-purpose Work

The in-house laboratory also exists to do work thatis not of interest to commercial industry but is ofinterest to the military. A case in point is the kind ofsmall-batch production of radiation-hardened chipsdone at a few labs. Such R&D serves two relatedpurposes: it produces highly specialized chips insmall runs for military customers and, more impor-tant, it leads to new technology for subsequentinsertion into existing systems. The drafting of newspecifications can, by itself, lead to new develop-ment projects. What is more, the effort to improveproduction cycles can itself lead to new technology:in areas like silicon-on-sapphire microelectronicstechnology, military labs are far ahead of industry.

Such special-purpose facilities help to tie indus-try to the work that is being carried on at militaryfacilities. By pushing the state of the art, thesefacilities force industry to focus on applications ofmilitary interest, such as the radiation hardening ofintegrated circuits and the applications of iridiumphosphide and gallium arsenide technologies. Bydoing very advanced research, these small, special-ized production facilities stimulate the right kind ofwork, so that it becomes available for industrialproduction.

User Support

For many labs, work does not end when they handover technology to the buying organization. In thiscontext, the term “laboratory “ is something of amisnomer, applying as it does to the NRL and a fewsmaller institutions. The preponderance of serviceR&D facilities are product development or engineer-ing centers, whose staff will often continue work tothe fielding of a new system and beyond.

Viewed in this light, much of the research at theNaval Weapons Center (NWC), the Naval OceanSystems Center (NOSC), and even NRL is done thebetter to support their principal customers or, asnecessary, to provide quality control support to thecontractor. Thus NWC was brought in by the NavalAir Systems Command to assist in redesigning theSparrow-a medium-range, air-to-air guided missilethat had run into serious problems when deployed inVietnam; NRL has consistently sent its scientistsand engineers in to support the fleet; and many AirForce R&D centers have sophisticated approaches toinserting new technologies in existing systems.Indeed, much of this technology insertion can occurindirectly: for example, a company may do develop-ment work under contract to an engineering center,adapt the new technology and sell it back to themilitary.

If stress has been laid on the role of Navy centersin supporting the fleet, it is because this is one of thefeatures that most distinguish them from the otherServices, especially the Air Force. Compared to theNavy, the Air Force uses its labs more exclusivelyfor technology exploration and component work,and uses industry for bringing technology to produc-tion. The Navy, on the other hand, uses its labs in‘‘full spectrum mode” for 6.1 through 6.4 (engineer-ing development) work, and acquisition support and

Chapter S---The Management of Defense Department Laboratories ● 67

fleet support thereafter. The reason, as one Navyofficial explained, is that the Navy’s mission ‘makescontinuous support for industry necessary, because[otherwise] a contractor might have to go out on acarrier for six months. ”

Still, it must be said that much of the justificationfor military work conducted out of military facilitiesis somewhat after the fact. The current institutionalarrangement of military R&D is more a matter ofhistory than of cold logic. In fact, the two alternativemodels for technology development within thegovernment were generated outside DoD, by NASAand the Atomic Energy Commission (AEC). Eachmodel embodied a philosophy of how differentresearch and engineering centers could be groupedin related fashion.

In each case, the critical decisions on how theagencies would operate were taken right at thebeginning. NASA would operate through a networkof field centers staffed by government employeeswho would define the work to be done, select theprime contractors, evaluate the work done and, ifnecessary, be prepared to go into the contractor’splant and take over.5 By contrast, the AEC chose towork through a network of multiprogram laborato-ries operated under contract; in the case of theweapons laboratories, they would be part of avertically integrated system combining research,weapons design, and production.

The important point is this: the institutionalarrangements at NASA and AEC were matters ofdeliberate policy. By contrast, the military R&Destablishment has grown haphazardly without thekind of fundamental decisions that NASA or AECtook. Unlike those agencies, the Services and theOffice of the Secretary of Defense (OSD) havevacillated among a number of approaches: betweenbuilding up labs as full-spectrum organizations, orseparating generic technology base work fromengineering and development; or between doingresearch loosely coupled to service missions andpressuring the labs to work only in mission-relatedareas. While DoD’s stated policy is that its in-houselabs shall maintain “a level of technological leader-

ship that shall enable the United States to develop,acquire, and maintain military capabilities neededfor national security,” the actual policies of OSDand the services are somewhat less consistent.6

This inconsistency shows itself in organizationalarrangements that (so to speak) require one part ofthe organization to work around the rest. Despiteeverything that militates against it (see ch. 4), somegood work manages to get done. The problems thatafflict military tech base work are those that afflictall very large organizations. In general terms,bureaucracies are “tentacular”; that is, if you makea mistake, be very sure that every hole is plugged sothat it will never happen again. One official put itthis way, tongue firmly in cheek: “Central is better.If you want to buy furniture, have one guy in chargeof buying for the entire organization, even if you cango across the street and get the item at a much lowerprice. ”

In fact, that remark identifies one of the key flawsin the entire DoD technology development organiza-tion. At many labs, the technical director has nocontrol over the most important support elements ofhis or her organization—the personnel office, thegeneral counsel, the procurement people, possiblyeven computing services, all of whom report to thebuying commands or headquarters.

But this is to anticipate the ensuing analysis ofoperations at military R&D centers. While one cangeneralize about the problems of the military labs, abetter approach is to begin with specific issues and,in the light of those analyses, to derive some usefulconclusions.

Personnel Management

Virtually every study of military laboratories hasnoted critical deficiencies in the way they recruit,train, and manage their professional staffs:

. Most of the larger laboratories experiencedifficulty in hiring and retaining qualifiedscientific and engineering personnel, especiallyhighly qualified senior staff.

Son (he dwl~lo~ t. ~reatc ~enters staffed by government employees, we Amo]d S. ~vine, Ma~ging/VAsA in fh APO11O Era, SP-4102 (Washington,DC: NASA Scientific & Technical Information Branch, 1982).

6u.s. Dep~ent of ~fenw, Under SWretW of ~fen~e for Rese~ch and Engin~ring, Dep~ent of Wfenx Instruction 3201.3, MN. 31, 1981,p. 2. Cited in Library of Congress, Congressional Research Service, “Science Support by the Department of Defense,” December 1986, pp. 178-179.


●

●

●

●

The government is at a major disadvantage incompeting with industry and the universitiesfor the best technical and engineering gradu-ates.The job classification system requires elaborateposition descriptions that have little or nothingto do with the positions being filled.The system makes it difficult to reward thegood performers or remove the poor ones.Inflexibility in setting salaries means that pay isseldom commensurate with performance:

Lab directors do have some discretion to workwithin the system. Thus the Navy uses ‘‘managingto payroll” (MTP) as a discipline to keep from hiringtoo many people, while allowing trade-offs. MTPallows the naval centers to keep their dollarsconstant while changing the number of slots. Fromthe Navy’s perspective, the advantages of MTP are,first, that it gives technical directors flexibility indistributing work among different center employeesand contractors; and second, that it helps to“cleanse” the centers by shedding work that theyshould not have taken on in the first place. UnderMTP, centers can maintain a stable work force eitherby cutting back on contractors, or by carrying theirpeople on overhead.

Most military R&D facilities have tried to makesimilar, piecemeal improvements within the currentsystem. But two R&D centers have successfullyattempted a more comprehensive approach, withinthe terms blessed by Title VI of the 1978 CivilService Reform Act. Since 1980 the NWC (ChinaLake, CA) and the NOSC (San Diego, CA) haveparticipated in a personnel experiment, with twoother Navy centers as controls.

The China Lake experiment, as it is generallyknown, breaks with the standard Federal personnelsystem in four ways: 1) separate career paths, withdistinct paths for scientists and engineers on the onehand and technical or administrative specialists onthe other; 2) the consolidation of 15 General

Schedule (GS) grade classifications into no morethan 5 broad “pay bands” corresponding to careerpaths (professional, technical, administrative, tech-nical specialists, clerical/assistant); 3) abbreviatedposition descriptions and standards; and 4) a muchcloser linking of pay to performance. Although theOffice of Personnel Management (OPM), whichoversees the project, originally designed it to run 5years, Congress has extended it to 1990.7

Differences in the ways the two centers haveimplemented the China Lake experiment are minorcompared to the similarities. At both NOSC andChina Lake, pay is linked to the GS scheme ofclassification, There is a set formula for hiring juniorprofessionals determined by each center’s personneloffice; at higher levels, supervisors set salariesaccording to the pay bands, with salary tables basedon government-wide changes in GS pay scales. Eachcareer path is a separate competitive path: ifreductions in force occur at a center, they can occuronly within specified career paths.

China Lake has been one of the most closelyfollowed demonstration projects ever sponsored bythe Federal Government. OPM has monitored theproject since its inception, issuing annual progressreports and a comprehensive evaluation in 1986. Inthat report, OPM found that the project had largelysucceeded in doing what it was intended to do.Compared to the control sites, personnel at thedemonstration labs-employees and supervisors—perceived the system to be more flexible than theNavy’s conventional performance appraisal system.In reviewing compensation systems, OPM con-cluded that the positive results it found seemed tohave been “strongly influenced by the introductionof broad pay ranges corresponding to the newclassification levels . . . broader pay ranges providegreater latitude in making performance-based paydistinctions.”8

OPM identified other elements that, in theopinion of its staff, helped account for the project’s

vTherehave ken mmy dcscrip~i~ns Ofthe china L*C experlrncnt. This account draws on several, including: U.S. ~fice of personnei M~agement,Research and Demonstration Staff, Office of Performance Management, ‘‘Status of the Evaluation of the Navy PersonneI Management DemonstrationProject: Management Report I,” March 1984 and (s~e source) “A Summary Assessment of the Navy Demonstra[ionRoject: Management Report IX,”February 1986. The original opM propos~ is in Federal Register, VOI. 45, No. 77, Apr. 18, 1980, pp. 26504-26543. A good summary account may befound in Larry Wilson, ‘ ‘The Navy’s J3~rirnent with Pay, Perform~ce and Apprai@,” De~ense ManugernenrJournuf, Vol. 21, No. 3, 3rd quarter 1985,pp. 30-40.

W.!j. office of Personnel Management, “A Summary Assessment, ” op. cit., footnote 7, P. VI1.

Chapter 5—The Management of Defense Department Laboratories ● 69

relative success. One of these was the labs’ involve-ment in developing their system. Another was thatthe system covered employees at a wide range ofwork levels, and a third was the protection ofemployees from any initial adverse impact, by a‘‘buy-out” feature written into the project plan.

The China Lake experiment was not designed tobe cost neutral. That is, the Navy recognized thatthere would be certain start-up costs in moving to thenew system. Once the system was in place, averagesalary differences between demonstration and con-trol labs tended to flatten out, to the point that thedifference among scientists and engineers dimin-ished greatly or even disappeared after they had beenon board 3 or 4 years. 9 But the ways in which thesame pot of money was distributed were quitedifferent. As OPM put it, “the initial salary gap isgreat enough that in any year the remaining demon-stration v. control difference in the salaries of newand recent hires accounts for about 2 percent to 3percent in additional demonstration payroll costs.”

One of OPM’s most significant conclusionsabout China Lake was that although costs arecontrollable, “the decision to limit costs can sub-stantially alter the results achieved. Unless organiza-tions are willing to make some investment in the newsystems, employees are likely to perceive they willgain no benefit from the systems, or that they willactually be penalized under the systems.”11 In fact,OPM concluded, total salary costs had risen by 6.0percent (as of January 1986) over those of the controlsites as a direct result of the project. Costs rosebecause the demonstration labs were offering toscientists and engineers starting salaries that were17.5 percent higher than at the control labs, andbecause China Lake permitted greater salary in-

creases within pay bands than would have beenpossible under the General Schedule.12

A May 1988 report by the General AccountingOffice (GAO) confirmed many of these findings.GAO’s general conclusion was that the projectdemonstrated that a pay-for-performance systemcould be implemented to the general satisfaction ofmany supervisors and their employees. Despite thisqualified approval, GAO found that the OPMevaluation left many questions unanswered. InGAO’s view, “the overall weakness of the ChinaLake evaluation was that when all is said and done,the volume of data that were either missing ornon-comparable was quite large. ” Although GAOdid not know the reasons behind the data problems,it determined that ‘they were of such magnitude thatfirm conclusions about project effectiveness cannotbe drawn.”13

The problems GAO cited included the non-comparability of the test and control sites and a lackof information on how and to what extent the projectwas implemented at the test sites.14

But this begs the question. Underlying the GAOanalysis is the assumption that one is comparing atentative demonstration with a system that is inter-nally coherent and designed to address the sameissues (but in a different way) which the China Lakedemonstration was created to address. Leaving toone side the difficulty of evaluating so complex aprogram, the existing government personnel systemis even more vulnerable to criticism. Speaking onlyof current hiring procedures, former OPM DirectorConstance Homer said that the system “is slow; itis legally trammeled and intellectually confused,and it is impossible to explain to potential candi-dates.”’ China Lake and a comparable demonstra-

~Ibid., pp. 51-52.

ltllbid., p. 52.

11 Ibid., p. VIII.12[,, ~ “@ate “f its 1986 ~eP~, OpM concluded Mat tie average sa]~ries of dcm~n~tralion lab scientists ~d engineers continued tO gTOW re]atiVe

to those of their control lab counterparts. This difference could nor be cxpltiined by the effecxs of salary increases, since these were virtually identicalat both kinds of site. Higher starting salaries for scientists and engineers, which were 18.7 to 29.1 percent greater at the demonstration sites, secmcd toaccount for much of the differential. U.S. Office of Personnel Managcmcm, ‘ ‘Salary Costs and Pcrfcmnance-Based Pay Under the Navy PersonnelManagement Demonstration ProJcct: 1986 Update: Managcmcnt Report X,” December 1987, pp. 3-6.

IJu.s. General Amounting OfilCC, “Observations on the Navy’s Pcrsormcl Management Demonstration Project,” GGD-88-79, May 1988, p. 29.These remarks were included in a letter from GAO to Senator !hn Nunn, Chairman of the Senate Armed Services Committee.

141n light of its findings, 1( is wo~h noting that GA() has ins[ltu[cd its ow~ pay. fc>r-pcrformmce system, wifi tbec pay binds corresponding 10 Grades7 through 12, “leadership” positions at GS-13/14, and “managerial” posts at GS - 15.

l~Constance Homer, ‘‘Address to Career Entry Recruitment (’onfcrencc, ” Washington, DC, June 23, 1988, p. 4.


tion project at the National Institute of Standards andTechnology (NIST) (formerly the National Bureauof Standards) have a coherence and logic that thecurrent system altogether lacks.16

What has the China Lake experiment reallyachieved? In one sense, it “demonstrates” just howinadequate the current personnel system is. Note thatthe experiment is now into its ninth year, with noimmediate prospect of extending it to other govern-ment laboratories. A program that began underCarter and continued under Reagan awaits the BushAdministration for extension or termination. TheCivil Service Reform Act that authorized the projectprovided no mechanism for extending it beyond itstest sites. Assuming that the project is extendedbeyond 1990, the government may be faced with asuccessful experiment that will have no ramifica-tions, unless Congress enacts proposals to extend theproject to Federal laboratories generally .17

But if one concedes—as even GAO has done—that the China Lake experiment did what it wasmeant to do, its success is somewhat irrelevant to theproblems of DoD laboratories. One could conceiveof small-scale improvements to the current systemeven without instituting performance-based pay.After all, OPM has in place mechanisms that makeit easier for agencies to hire qualified professionalstaff. Thus, agencies can now apply to OPM forauthority to hire engineers directly, without an initialscreening by OPM. OPM has delegated to agenciesthe authority to negotiate starting salaries withtop-quality candidates for jobs at grades GS-11 andhigher. On a pilot basis, OPM is drafting simplerstandards that agencies can use to classify positions,including engineering positions, These new stan-dards, OPM says, ‘‘will give agencies more flexibil-ity to redesign work, to classify jobs and to writeagency specific guides if needed. ” *8

But the ultimate limitation of China Lake andsimilar proposals is that they simply divide up piecesof a pie whose overall size remains about the same.Salaries are adjusted within narrow parameters;despite increases at demonstration sites, OPM foundthat salaries for nearly all the occupations comparedgrew 5 to 19 percent less at those sites than didsalaries for the same occupations in the privatesector. 19

In this respect, the NIST project is superiorbecause it has the authority to adjust the salaries itpays its scientists and engineers to match those paidby the private sector for comparable work. Andwhile its demonstration project is supposed to becost neutral during the first of its 5 years, NIST canuse the surveys as a device to narrow (if not close)the salary gap between itself—NIST salaries arealready among the highest at Federal laboratories—and industry by work force attrition.

Even if the China Lake experiment were extendedgovernment-wide, it would take a long time to undothe damage wrought by the current system. Person-nel issues cannot be isolated from other issues—funding, research planning, the acquisition of majorsystems, and the like. Among the elements not yetmentioned that affect the labs’ ability to hire,promote, and retain are the periodic hiring freezesthat affect most government institutions; the newFederal Employees Retirement System, which makesit easier for government workers with portablebenefits to leave the government earlier; and cut-backs in travel budgets, which make it harder for labofficials to recruit. Compared to the larger DOElaboratories, which recruit from the top 10 percent ofgraduates from the major national technical schools,most DoD managers tend to hire locally-partlybecause of small travel budgets, partly because theyare resigned to the unavailability of top graduates.

lbThere me mtiJor differences in the design of the NIST and China Lake projects. Although it incorporates such concepts as pay bands and career paths,NIST has certain special features: direct-hire authority for all professional cmployees, an annual comparability survey of total compensation of NISTpositions to similar positions in the private sector, cost neutrality, and recruiting allowances for professionals that NIST particularly wants to hire.

ITThere have been a n~lber of legislative proposals to reform the Federal personnel system. OPM twice unsuccessfully introduced its Own ProPosaifor a “Simplified Personnel System, ” most recently in January 1987. A bill, S. 2530, introduced by Senator Jeff Bingaman (D-New Mexico) in June1988 would extend the China Lake experiment beyond the two naval centers currently involved. S. 2530 would authorize between six and ten personneldemonstrations, of which four would be instituted at DoD and onc at NASA. The biH would establish higher minimum rates of pay and an alternativecompensation system based on comparable rates for comparable private-sector work.

I~U.S. Office of Personnel Management, “Simplifying the Federal Manager’s Job” (n.d.), p. 3.19u,s, Office of permmcl ~m~gement, op, cit., footnote ~~, p. [], opM ~erive(f its sa]ary-comp~son fi~res from data provided by tie Bureau Of

Labor Statistics.


But personnel issues extend beyond competingwith the private sector for new hires. Retaining goodemployees in an environment where the work forceis overgraded but underpaid is just as formidable aproblem. Technical directors stressed the nature ofthe work itself as one of the strongest attractions fortheir best and brightest. Good people stay if the workis challenging, if they have the opportunity to dosome basic research and publish their findings, ifdiscretionary funding is available to start new work,and if the laboratory gives equal recognition toseparate career tracks for researchers and managers.

Although many lab officials spoke of having“two-track” systems, the evidence for such isambiguous. According to a survey by the ArmyLaboratory Command, lower-level engineers be-lieved “that a scientist had to become a manager inorder to get ahead in a government laboratory. ”20

From the lab director’s perspective, someone has totake on the responsibilities with which externalorganizations task the lab, as well as manage thelarger programs that constitute its mission. Thus,inevitably, many scientists and engineers comeunder pressure from their superiors to take on workoutside the disciplines in which they were trained.Sometimes, scientists and engineers make a success-ful transition into management. Other times, as oneformer government official said, “you take goodengineers and turn them into lousy managers. ”

There is also a more insidious threat to theintegrity of the professional work force. ThroughoutDoD laboratories, the increase in congressionaloversight has gradually transformed the role ofresearch executives, such as division heads andbranch chiefs. Rather than managing projects orensuring their technical quality, one of their princi-pal jobs is now to insulate their bench-level peoplefrom the requirements with which Congress tasksthe labs and their sponsors. In particular, the amountof oversight and paperwork appears to have in-creased the most at those laboratories where the bulkof the work is contracted out—thus forcing manag-ers and other senior professionals into contractadministration.

In sum, the personnel problems that afflict mostDoD laboratories are not personnel issues in thenarrow sense. They flow, rather, from the totalenvironment within which professional staff andmanagers try to get their jobs done. Even where acenter can attract the top graduates, it has to contendwith problems that are not ‘‘personnel” at all:uncertain budgets, long lead times in building newfacilities and procuring new equipment, and limitson the pot of discretionary funds available to startnew work. The incentives for a new hire to remainpermanently depend more on the total environmentof his or her institution than on personnel manage-ment practices in the narrow sense.

Two more points deserve emphasis. The first isthat the Federal personnel system creates someperverse incentives for retaining employees. Underthe Federal Employees Retirement System, thebetter employees can take their retirement benefitsand leave for industry and universities with many oftheir more productive years still ahead. At the sametime, mediocre performers remain; under the ChinaLake system or Managing to Payroll, they can expectno major salary increases, but they also stand littlerisk of being terminated. Thus, a low turnover rate ata laboratory can be a sign of health or a portent ofinstitutional decline.

The other point has to do with the optimal size oflaboratories, an issue discussed later in connectionwith GOCO facilities. It may well be that there aretoo many DoD laboratories, and that many of themare too small ever to achieve critical mass. If aninstitution is too small, there will be too littleflexibility for a few people to strike out into newterritory, or for new ideas to spill over into researchwork. At smaller facilities, there may not be enoughgroups of two or three or four people delving intoareas unconnected with their current missions butthat might lead to new missions. Governmentinstitutions seemingly must have more than about1,000 people before the kind of flexibility that makesfor their survival exists.21 Additionally, as weaponssystems grow ever more sophisticated, the numberof disciplines that a lab needs under one roof willincrease.

Z~<S. Army Laboratory Command, 4 ‘Innovative Personnel Practices, ” March 1988, p. 4.ZIHm~ Mwk ad AmOld ~vine, The Ma~8eme~ of Research ln~titutions. A ~ok atGover~entL&oratories, SP4181 (Washington, ~: NASA

Scientific & Technical Information Branch, 1984), p. 70.


The question of a lab’s optimal size bearsdirectly on the retention of quality staff. Unless a labis assigned a mission in one narrowly definedarea-the Army’s Night Vision Laboratory (Ft.Belvoir, VA) might be an example-it must havescientists and engineers drawn from a variety ofdisciplines. Even in that case, as a 1979 governmentreport noted, ‘‘the development and enhancement ofmodem technologies is an inherently multidisci-plined endeavor. The most narrowly focused ofresearch activities today involve several profes-sional disciplines as well as highly skilled technicalsupport personnel. “22

The DOE’s weapons laboratories have becomeadept at instituting a matrix structure, wherebymoney is pulled away from divisions and moved intoprograms. The result is that there is more mobilitywithin Energy labs like Sandia and Los Alamos thanat most DoD labs. With some exceptions, a newprofessional hired at a DoD lab is likely to spend hisor her career within the same research division ordirectorate. In contrast, professionals at DOE labshave more options: beginning their careers at (say)the lab’s research division, they may move intomission-related areas, return to research, or moveinto management.

In fact, the DOE labs have done more to controlpersonnel problems than virtually any DoD facility.For one thing, they have bypassed the entire issue ofsalaries comparable to those of the private sector. Asa rule, salaries and personnel systems correspondclosely to those of the contractor who operates thelab: the personnel system at Sandia National Labora-tories is modeled on AT&T’s Bell Labs, while thoseat Los Alamos and Lawrence Livermore NationalLaboratories are modeled on that of the Universityof California system. For another, the DOE laborato-ries tend not to hire for specific jobs. Their size andmultidisciplinary capabilities make it easier for labexecutives to move people to where they are neededand redeploy people as projects wind down.

Laboratory Management Issues: Funding

The ability of a government laboratory to accom-plish its mission depends on the ways it is funded.The amount of funding obviously matters, but so

does its predictability, flexibility, and the ability oflab managers to disburse funds once they becomeavailable. Laboratory executives say they preferfunding that is tight but predictable over larger butunpredictable funds.

At DoD labs, funding problems are at least asnumerous as the personnel problems they aggravate.Funding is unstable, making planning and staffcontinuity on projects difficult; it is inflexible, inthat most funds cannot easily be transferred to otheraccounts where they might be needed more; andmonies must be spent during the fiscal year forwhich Congress appropriated them, preventing thebuildup of contingency funds. This requirementaffects DoD’s ability to sustain long-term work.

Nevertheless, there are important differences inthe way the services do their getting and spending—with the Navy centers obtaining their funding fromthe Naval Industrial Fund (NIF) and the Army andAir Force receiving money through line-item appro-priations.

NIF is a shorthand way of saying that navalcenters must recover the full cost of their operations.Industrial funding provides working capital forindustrial-type activities, such as shipyards, theoverhaul of aircraft, or running a laboratory, Underthis approach, the activity pays all its expenses outof working capital and charges its customer the fullcost of its products and services. Each industrialfund activity group has a cost accounting systemspecifically designed for its operations, to identifyand accumulate the costs of their products orservices.

This approach has important implications for theconduct of naval research, development, test andevaluation. First, because the NIF is a revolvingfund, payments that naval centers receive from theircustomers should do no more than replenish theworking capital fund that finances operations untilpayments are received. Second, in relation to their“buying” commands, naval centers are contractorsde facto and de jure. A Naval center undertakes workfor (say) the Naval Sea Systems Command on thebasis of a contractual agreement that obligates bothparties until work is completed. A facility like theNWC at China Lake has virtually no line-item

zzFedera] Coordinating Council for Science, Engineering and Technology, op. cit., footnote 4, P. 35.

Chapter 5-The Management of Defense Department Laboratories ● 73

budget authority. Instead, it operates like a Battelleor SRI International, which would go out of businessif it had no customers.

A third, very important, feature of NIF is the assetcapitalization program (ACP).23 Effective fiscalyear 1983, the Deputy Secretary of Defense ap-proved asset capitalization as a way to fund themodernizing of industrial fund equipment. Underthe program, equipment costs are recovered over thelife of the asset by including depreciation costs in therates charged to customers, The availability of ACPmoney strengthens the cash position of industrialfund, and helps fund managers avoid shortages.

Thus, naval engineering centers obtain workquite differently than a NASA research center or anArmy laboratory-although Army research, devel-opment, test, and evaluation (RDT&E) was fundedindustrially at one time. At a facility like the NWC,the program offices serve as “shadow offices” totheir counterparts in the prime sponsoring organiza-tion, which in their case is the Space and NavalWarfare Systems Command (SPAWAR). Note thatthis is not the principal buying organization for thecenter; its principal customer, accounting for morethan half of its total obligational authority, is theNaval Air Systems Command. Although there aresomething like 3,000 customer orders in the systemat a given time, some two dozen cover most ofNWC’s work.

In the view of managers at the Navy centers,industrial funding is an effective way of gettingwork done. Among its advantages are that itprovides limited authority to start work on asponsor’s order prior to the receipt of funds, assistsmanagers to control their resources better, enablesthe facility to finance and carry inventories ofnon-standard materials, permits the use of workingcapital for initially charging all costs, including 6.1and 6.2 work, and serves to develop total costs foreach task, including overhead.

Most Navy lab managers consider a recent OSDproposal to terminate NIF over the next two fiscalyears potentially disastrous. DoD contends thatindustrial funds are more costly to operate than other

systems, that their advantages have not been demon-strated, that industrial fund clients are not bona fidecustomers who can take their business elsewhere,and that as structured, the system makes DoD andcongressional oversight difficult. The Navy, sup-ported by GAO, disputes these assertions, chimingthat NIF meets the criteria under which the Navy’sresearch and engineering activities are financed.24

If NIF were terminated, the Navy would haveseveral options. One would be to convert to aresource management system that combined customer-funded direct labor with Navy-funded overheadunder an appropriate budget line item. This could bedisastrous, in the view of some officials, becauseoverhead becomes very difficult to defend in acompetitive budget preparation environment. Alter-natively, the Navy could adopt a resource manage-ment system with ‘‘applied overhead,” which isidentical to NIF at the macro level, except that it hasno asset capitalization program. While the Navycould live with this arrangement, it would incursizable one-time costs to convert its financialsystems.

Compared to Army and Air Force labs, industrialfunding gives the Navy a certain flexibility instarting and accounting for work. But it is still firmlypart of the appropriations process, although at oneremove. The start of work at a Navy lab still dependson its customers having the necessary obligationalauthority-and, if that money comes in late in thefiscal year, that it remains available to complete thework it is funding. Further, there are important areasof naval lab operations not covered by industrialfunding, such as military salaries, non-appropriatedfunds, and military construction.

Military construction deserves special mention,since delays in new construction are one of the majorobstacles to lab performance. This is the case forseveral reasons: as with other functional areas, thoseresponsible for facilities management do not reportto the lab technical director; lab requests for newfacilities are thrown into one “pot” with otherconstruction requests, and new facilities for labsgenerally have rather low priority. At some DoD

zson the -t capi~i~ation program, see U.S. General Accounting Office, “Industrial Funds: DoD Should Improve Its Accounting for AssetCapitalization Program Funds,” NSIAD-86-1 12, May 1986.

ZQU.S. Gener~ Accounting Office, “Proposal to Change From Industrial Funding to Another Method, ” NSIAD-89-47, December 1988, pp. 1-3.


laboratories, many facilities are 40 years or evenolder.

This is doubly unfortunate, because good facili-ties not only drive a lab’s mission, but also attractgood people. In turn, an excellent staff will, to adegree, generate good facilities, The process isself-perpetuating; people tend to generate newprograms around the facilities, so that when anRDT&E organization matures, its roles and missionsdepend primarily on the facilities available: windtunnels, clean rooms, anechoic chambers, simula-tors, and the like. As other authors have noted,“facilities have a longer ‘half life’ than people. Afacility like the 40-by-80-foot wind tunnel at [NASA’SAmes Research Center] might be used for fortyyears, while an individual researcher will change hisinterests every three or four years and move on tosomething new. Thus a vigorous research anddevelopment program demands an efficient facilitiesdevelopment staff, more particularly where onefacility serves a number of projects. ”25 By thiscriterion, few DoD laboratories have the power todevelop facilities to keep pace with either theequipment that they will house or the missions theyare designed to support.

Thus, DoD laboratories are subject to all thedisadvantages and few of the advantages of facilitiesowned and operated by the government. But it isimportant to understand that these problems do notflow automatically from the status of these facilitiesas government-owned, government-operated insti-tutions. Both the NASA centers and NIST haveshown greater flexibility: NIST, because its role aslead agency in measurement science is highly valuedby its government customers; NASA, because of themuch stronger ties between the centers and theirprincipal buyer, the headquarters program offices,than in the DoD system. At NASA, the centerslargely define the programs that the agency funds. AtDoD, by contrast, the relation of the R&D infrastruc-ture to the buying commands is much less certain.

The next section describes alternative approachesto developing technology—those represented by

GOCO facilities of the DOE and the somewhatsimilar FFRDCs under DoD.

GOVERNMENT-OWNED,CONTRACTOR-OPERATED

FACILITIES: AN ALTERNATIVEMODEL

Introduction

The GOCO facilities are an unparalleled resourcefor the United States. In particular, the nine multi-program, or ‘‘national” laboratories represent one ofthe heaviest investments in basic and appliedresearch made by the United States or any othercountry. Besides conducting about 70 percent of theDOE’s weapons development and a quarter of itsenergy-related research, the national labs have otherroles. As systems engineers for DOE, as consultantsto State and local governments, and as stewards ofunique facilities, the labs contribute in many ways tothe Nation’s technology base.

From their inception, all the multiprogramlaboratories have been government-owned and contractor-operated. The Atomic Energy Commissioners chosethis course, although they were not barred fromoperating their own laboratories; indeed, the AtomicEnergy Act of 1946 authorized “a program offederally conducted research and development. ”According to Harold Orlans, the AEC contractedwith private organizations as the principal means of‘‘retaining a degree of normalcy and freedom in theevolving system of nuclear science and industry. ”26

By contracting with outside groups, AEC could keepthem informed about highly classified activities thatwould normally be confined to official circles, andbring to the government experience and advice notnormally available to it. Orlans concluded that thisdecision helped, as much as anything, “to keep theAEC more alive and alert, administratively andtechnically. ”27 The result was an arrangement thathas no counterpart in the Federal Government, savefor the contract between NASA and the CaliforniaInstitute of Technology to operate the Jet PropulsionLaboratory.

zshlwk and t.,evine, op. cit., footnote 21, pp. 83-84.‘2~H~O]d orIanS, contracting for Atoms (Washington, DC: Brookings institution, 19~7). P. ~.271 bid., p. 8.

Chapter 5-The Management of Defense Department Laboratories ● 75

Before turning to the applicability of theEnergy model to DoD institutions, somethingshould be added about FFRDCs like Lincoln Labo-ratory and university-affiliated research centers likethe Applied Physics Laboratory (APL) of The JohnsHopkins University.

28 DOD sponsored these centersfor much the same reasons that DOE chose tooperate its national labs through contractors: theServices sought independent outside expertise fromorganizations unfettered by many Federal regula-tions; they wanted to develop long-term relationswith such organizations; and they specificallywanted to deal with institutions tied to the universitycommunity. The 30-year collaboration between theNavy and the Applied Physics Laboratory on theFleet Ballistic Missile Program shows how effectivesuch a special relationship can be.

The main difference between contract centers likeAPL and a DOE laboratory is that the former areprivately owned organizations working for a pri-mary sponsor. At APL, for example, Johns Hopkinsowns the land and the buildings—although theNavy, APL’s prime sponsor, furnishes the equip-ment. There are other differences of degree ratherthan kind. Compared to the DOE weapons labs,which have their own audit organizations, FFRDCsare audited regularly by the Defense Contract AuditAgency (DCAA). Additionally, many centers haveto go through special procedures to avoid the fullweight of Federal regulations. For APL to avoid thebroad mandate of the Competition in ContractingAct for competitive procurement, the Navy’s SPA-WAR must draft a “justification and approval,”which the Assistant Secretary of the Navy forShipbuilding and Logistics ultimately signs,29

Although these contract research centers performsome of the functions of DOE’s national laborato-ries, there are significant differences. One differencehas to do with areas of emphasis: Compared to themany functions of the larger national laboratories,

the FFRDCs and university affiliates tend to concen-trate on systems integration and engineering serv-ices. They are more likely to work almost entirely forone sponsor, and to devote most of their resources toa few programs, than the national laboratories are.While the same could be said of DOE’s weaponslaboratories, their size, their diversity, and theircapacities for advanced research make them a moreappropriate model for DoD’s consideration.

The GOCO Relationship at the Departmentof Energy: Contractual Arrangements

The organization and operation of the multipro-gram DOE labs’ are in dramatic contrast to those oflabs operated by DoD employees. In the former, wefind the vertical integration of research, develop-ment, and operations; a long-term relationship withthe sponsoring agency; a critical mass of scientificand technical disciplines; and (compared to DoD)much greater flexibility in moving people betweendivisions and projects.

Although superficially complex, the administra-tive relations between the labs and DOE are actuallymuch simpler than those at DoD. Through its staffand program offices, DOE headquarters in Washing-ton sets broad policy and develops the overall budgetout of which funds to operate the labs will come.30

Eight field operations offices monitor the operatingcontracts, although their roles encompass muchmore. Finally, the labs carry out broad programs ofresearch and technology development within theguidelines approved by headquarters.

Arrangements between DOE and its contractorsvary within narrow limits. Management and Operat-ing contracts normally run for 5 years, with thecognizant operations office performing a‘ ‘compete-extend” analysis before the contract expires. Com-pared to standard commercial contracts between aFederal agency and vendors, the terms are moregeneral and until recently were based mainly on

28u.s. Gener~ Accounting Office, “Competition: Issues on Establishing and Using Federally Funded Research and Development Centers, ”NSIAD-88-22, March 1988.

2~Johns Hopkins University, Applied Physics Laboratow, “Report to The Johns Hopkins LJnivcrsity Trustees Committee on the Applied PhysicsLaboratory,” March 1988, p. 2.

30FOT Pupxs of contract a~ifistration, the fie]d Operations offices located c]o~ to the labs oversee them. For purposes of program phrming andinstitutional management, the nine muhiprogram laboratories are ‘‘administmuvely assigned” to two cognizant program offices, The Assistant Secretaryfor Defense Programs oversees the Idaho National Engineering, Lawnmcc Livermore, Los Alamos, and Sandia Nationat Laboratories. The Director ofEnergy Research is the “cognizant secretarial officer” for the Argonne, Brookhaven, Lawrence Berkeley, Oak Ridge, and Pacific Northwest NationalLaboratories.


reimbursable costs. Thus, AT&T manages Sandia ona no-profit, no-loss basis; the contracts for LosAlamos and Lawrence Livermore National Labora-tories reimburse the University of California foroperating costs and award it a management fee.More recently, at Oak Ridge National Laboratory,DOE has instituted a cost-plus-award-fee arrange-ment, in which the contractor, Martin MariettaCorp., receives a special fee based on performance.

If one looks at these contracts after reviewing astandard contract between DoD and one of itscommercial suppliers, they seem extraordinarilybroad. Here is virtually the entire scope of work inthe contract for managing Los Alamos NationalLaboratory:

Work under this contract shall, in general, com-prise research, development and educational activi-ties related to the nuclear sciences and the use ofenergy in mutually selected military and peacefulapplications, engineering services, and such otheractivities as the parties may agree upon from time totime . . .

Due to the critical character of the work from thestandpoint of the national defense and security, it isunderstood . . . that very close collaboration will berequired between the University and DOE withrespect to direction, emphasis, trends and adequacyof the total program.

How can anything so vague serve as the basis foroperating a laboratory with an annual budget of $900million? There is no single answer; instead, there areseveral reasons that this contract is a successfulinstrument for managing a national laboratory. Oneis that there is much more to the contract than thestatement of work just cited; there are, in fact,numerous powers, especially the power of the purse,by which DOE fleshes out the very broad mandatejust cited. Another reason is that after 40 years’experience of working together, both parties under-stand the terms very well. By itself, the GOCOcontract does not lead to a long-term relationship: itpresupposes it.

A special feature of the contracts between theUniversity of California and DOE is the provisionthat work shall be set by mutual agreement. These

“mutuality clauses” are unique, although at onetime NASA had such a clause in its contract withCalTech to operate the Jet Propulsion Laboratory.(CalTech has an R&D contract with NASA, not anoperating contract for administrative services.) To adegree, the mutuality clause gives a false impres-sion, since it implies that DOE may unilaterally taskother laboratories that do not have such a clause.There is actually a very complex give-and-takebetween all of the multiprogram laboratories, theirsponsors, their clients (including DoD), and theuniversities. The mutuality clause simply affirms theunderstanding that runs through all of these con-tracts: that DOE is tapping the expertise of outsideorganizations to run the labs; that this expertisecannot be used effectively if DOE elects to micro-manage the contractor; and that the contractor musthave freedom to select the technical approach mosteffective in carrying out the lab’s mission.

In this system, the operations offices are muchmore than contract administrators. This is why DOErejected a 1981 recommendation by the GAO thatthe operations offices report directly to each lab’scognizant program office, rather than to the Depart-ment’s Under Secretary. DOE officials contendedthat such a proposal would not only require a hugeincrease in Federal staffing, but would lead to “thebalkanization of the field structure.”31 A morecompelling justification for leaving the field struc-ture intact—as DOE did—is that the structure of theoperations offices mirrors the vertical integration ofthe Department as a whole. For example, besidesoverseeing the Sandia and Los Alamos laboratories,the Albuquerque Operations Office administers 7widely scattered weapons production facilities andthe system for transporting all government-ownedspecial nuclear materials.

The GOCO Relationship at the Departmentof Energy: Complying With Federal Norms

How far are the GOCO laboratories bound byFederal policies? There is no simple answer, perhapsbecause neither DOE nor its contractors wish to belocked into anything too definitive. Yet there hasbeen a gradual shift over the past decade, with DOE

31u.s4 Gener~ Accounting Office, ‘‘A New Headquarters/Field Structure Could Provide a Better Framework for Improving Department of EnergyOperations,” EMD-81-97, Sept. 3, 1981, See especially the comments of Assistant Secretary for Management and Administration William Heffelfmgerat pp. 48-49.

Chapter 5—The Management of Defense Department Laboratories . 77

trying to get the labs to conform more closely tolegislation and regulations.

The official view reflected in DOE directives andin congressional legislation is that the labs’ statusdoes not exempt them from complying with thespirit of Federal policies. As stated in an opinion ofthe Deputy Comptroller General, the labs mustcomply with “the Federal norm”:

It is our view that while Federal statutes andregulations which apply to direct procurement byFederal agencies may not apply per se to procure-ment by prime operating contractors . . . the primecontractor’s procurements must be consistent withand achieve the same policy objectives as the Federalstatutes and regulations. This, we believe, is what ismeant by the “Federal norm. “32

While a laboratory like Sandia follows AT&Tprocurement and personnel management policies, itis also bound by a variety of regulatory constraints.These include DOE acquisition regulations anddirectives that apply the Federal Acquisition Regu-lation to departmental entities, the Buy AmericanAct, and prevailing-wage legislation on Federallysubsidized construction contracts. As Federal con-tractors, the national labs also come under thesupervision of the Labor Department’s Office ofFederal Contract Compliance Programs,33

On the other hand, the labs are exempt from anumber of requirements that bind Federal agencies,among them formal advertising, set-aside programs,and the Competition in Contracting Act. Althoughmajor purchasers of supercomputers, the nationallaboratories are also exempt from complying withthe Brooks Act, which governs the acquisition ofcomputers and telecommunications equipment, atleast as it applies to scientific computing. Thesituation is less clear for administrative and general-purpose computers; the consensus at the labs is thatthey must sponsor full and open competition forthese machines. Finally, because Sandia and otherfacilities have their own audit capabilities, they donot require the services of the Defense ContractAudit Agency in monitoring their own contracts.Lab officials believe that they can handle small andmedium-sized procurements much faster than DCAA

can, although there is evidence (see below) thatprocurement lead times have increased substantiallyat the national labs.

These exemptions affect the labs’ operations inmany ways. First, they enable the labs to buildlong-term relations with industry in a way notpossible for Federal agencies bound by the Competi-tion in Contracting Act. Second, the labs find itexpedient to comply with the spirit of the law, evenwhen they are not bound by the letter. Thus theweapons laboratories set aside a substantial numberof smaller procurements for minority-owned smallbusinesses. For example, under pressure from GAO,Oak Ridge National Laboratory dropped its percent-age of sole-source procurements from 50 to 20percent. Third, these exemptions make it possiblefor the labs to function; to impose the full weight ofFederal regulations would undermine the rationalefor having them run by contract.

One area where the labs are free to set their courseis in personnel management. The personnel systemat each laboratory corresponds to that of the primecontractor because, as one official explained, “in aGOCO you have not only the people, but also theorganization’s management system. ” There are noassigned slots at DOE labs, and the very besttechnical people can make as much as $95,000,although a lower figure is more usual. The mostsenior executives at the weapons labs earn between$100,000 and $150,000, roughly twice what theircounterparts at the military labs earn.

The principal constraint on the willingness of alaboratory’s prime contractor to set the highestsalaries is the DOE review triggered at the $60,000threshold; the local operations office has approvalauthority up to $70,000, and DOE’s Office ofAdministration up to $80,000, with higher salariesrequiring the Director of Administration’s approval.Additionally, at some laboratories, DOE approvesthe appointments of the most senior executives.

DOE also approves facility-wide salary in-creases, based on cost-of-living adjustments, recruit-ment and retention rates, and the like. The facilityproposes an increase to the cognizant operationsoffice, which forwards the proposal with its recom-

qzDecision of Bputy comp~ol]er General in protest of Piasecki Aircraft COW. (B-190178, JUIY 6, 1978), P. lo.

33 Alone ~ong DOE labs, Sandia has tO file its accounting system with the governmen[’s Cost Accounting St~dmds Board.


mendations to headquarters. For its part, the Officeof Administration sponsors generic surveys of scien-tific and technical salaries. DoE is now developingcriteria to remove individual salary reviews andconvert to more “systemic” approaches to deter-mining appropriate levels.

Although practice varies from lab to lab, there isa certain uniformity in their hiring and promotionpolicies. When officials say that “there are noassigned slots at DOE laboratories,” they do notmean that people move randomly from assignmentto assignment. What they mean is that laboratoriesdo not hire for specific jobs. Instead, they hire peoplewith the technical disciplines that fit the laboratory’smission, and who give promise of performing wellin a number of environments. Again, many labs liketo move their “high-potential performers” withinand between program divisions, especially thoseindividuals with management potential.

The laboratories can hire and move around thebest people because of their sheer size. Each of theweapons labs has about 8,000 employees and, whilethis creates problems of its own, the number anddiversity of projects does make it easier to attract andretain the top engineering and scientific graduates—some of whom the labs hire on the spot. In particular,lab officials note that facilities are a key selling pointin hiring and promotions. Although layoffs do occur,the labs can keep them fairly small, since they haveother options not available to DoD laboratories, suchas finding slots at production centers for labemployees no longer needed at the main facility.

The personnel practices that DOE ratifies havemade the national labs far more competitive thanmost of those staffed by government employees.Thus, salaries are far more in line with industry andthe universities; the surveys sponsored by DOE helpto keep salaries realistic. Moreover, the labs recruitaggressively. Most of the larger ones recruit nation-ally and hire directly—something government labsare only beginning to do-and can attract the top 10percent of graduates from the best engineering andtechnical schools.

Funding Arrangements and Work for Others

The three weapons labs receive level-of-effortfunding for Defense programs. DOE allocates fundsto each institution annually, based on its mission and

the size of its staff. When the money becomesavailable to the laboratory, each program or divisiondirector negotiates with the lab director for a portionof the funds. But unlike many government labs,DOE facilities do not obtain their funds in one lumpsum. Instead, they receive money from hundreds ofseparate contracts with other DOE components—theheadquarters program offices-each of which speci-fies the task covered by its agreement. In this respect,the closest government analog is the NIF describedearlier.

DOE-sponsored work is funded with “no-year”monies, available until spent. This does not meanthat the labs have complete discretion in schedulingoutlays. DOE provides budget outlay guidance onwhen money shall be spent during the fiscal year,and DOE weapons labs must obligate DOE funds towithin 1 percent of allocation. But as one DOElaboratory executive observed, “it is the technicaldiscretion of the lab management (not accountingdiscretion) which is crucial.”

Other funds are obligated on a project basis by theend of the fiscal year, like those for DoD non-nuclearprograms, although some DoD money for R&D is2-year funding.

Consider how this system works at one weaponsfacility, Sandia National Laboratories. Its principalmission is research, development and engineering ofthe components of nuclear weapons (other than thenuclear explosive). In light of this mission, Sandiaexecutives regard their technical programs as havingtwo components: a technology base (basic andapplied research, computing, analytic techniques,advanced components) and deliverables (materialsfabrication, system design, quality assurance, stock-pile surveillance, nuclear safety). It is this twofoldmission that drives the program and determines thekinds of work Sandia will take on, particularly fromnon-DOE organizations.

For its purposes, Sandia’s no-year budget author-ity has two advantages. First, it enables the lab to letcontracts beyond the current fiscal year; and second,it allows long-term planning, even though DOE willdirect the lab through budget outlay guidance onwhat may be spent in a given year. Beyond that,Sandia officials can view their funding in differentways. In terms of sponsorship, DOE defense-relatedfunding in fiscal year 1987 accounted for 60 percent

Chapter 5---The Management of Defense Department Laboratories ● 79

of total operating funds, with other energy-relatedwork accounting for between 9 and 10 percent. Theremainder of Sandia’s funds came from reimbur-sable work from outside organizations—the impor-tant category of “work for others.”

Sandia’s policies on work for others follow DOEguidelines. Briefly, Sandia will not undertake workif it interferes with DOE weapons programs. Even ifresources are available, Sandia will not commencework unless it meets several criteria: The work mustbe of national and technical importance, match thelab’s mission and capabilities, avoid competitionwith the private sector, and complement existingDOE programs with integrally related work. WhereSandia participates in reimbursable programs, itincorporates DOE’s policy of full cost recovery.Sandia will seek to recover all costs including labor,direct charges, overhead (the lab charges the samerates to DOE and outside organizations), and generalpurpose equipment.

Like the other two weapons laboratories, Sandiaalso applies a surcharge-a tech base ‘‘tax” on allwork for others—that it uses to fund new, long-rangeresearch. At Sandia the tech base tax currentlysupports 70 people, most of whose work runs for upto 3 years. Note that this taxis only a portion of whatSandia spends on tech base work. According to DoDfunding categories, approximately 8.4 percent ofSandia’s 1988 budget went for 6.1 work and another17.4 percent for 6.2, or exploratory development.34

Thus, Sandia is effectively spending just over aquarter of its $1.1 billion budget on tech base-a farhigher amount than any DoD laboratory or engineer-ing center, save the Naval Research Laboratory,spends.

The major difference between DOE and DoDpolicy on work for others is that the DOE multipro-gram laboratories consider it a normal and desirablepart of their missions, while the latter does not. ForDoD, work for others—primarily non-defense workfor civilian agencies—is a distraction from the labs’missions and to be confined within narrow limits.For many years, DoD has had a policy of limitingsuch work to 3 percent of professional staff-years atindividual laboratories. Since DoD labs are con-strained by total personnel ceilings and are not

allowed to keep revenues for work for others, anywork done for external users comes directly at theexpense of their DoD clients. At the DOE labs, bycontrast, work for external agencies is much moreopen-ended: up to 20 percent of operating budget forEnergy Research labs, and as much as 30 percent forthe weapons labs.

This raises a fundamental question about themissions of the multiprogram labs: Why are they soeager to diversify? The easy answer is that asself-consciously “national” facilities, the laborato-ries regard diversification as an essential part of theirmission. But there is more to it than that. Thesefacilities have the preconditions for successfuldiversification. The first is the presence of secondparties willing to sponsor a laboratory’s venture intonew fields, just as industry sponsored Sandia’s workin drilling technologies, or Du Pent worked withArgonne National Laboratory on neutron diffractionstudies of catalysts, or SDIO funded work at LosAlamos in directed-energy weapons.

Next, lab executives believe that while theirorganization’s mission remains relevant, currentprograms do not exhaust the organization’s capacityto carry it out. And not least, there are fewinstitutional barriers to prevent laboratories fromtaking abroad view of their missions. Here, DOE hasplayed an important part by its policy of permittingwork for others, bringing in outside scientists andengineers for advice and joint ventures, and improv-ing conditions for cooperative work. Indeed, theremoval of obstacles may accomplish more thanwell-intentioned, but largely fruitless, efforts tostimulate two-party ventures.

This philosophy has implications for the defensetech base. As funding for nuclear weapons shrinks,DOE laboratory executives want to involve theirorganizations more closely in nonnuclear defensework. Diversification protects existing jobs and theability to hire fresh graduates. Their laboratories, sotheir argument would run, are already working inthese areas and have the experience to move intorelated fields. DoD funding for nonnuclear work isactually growing much faster than DOE funding is;at Lawrence Livermore, DOE funding betweenfiscal years 1982 and 1986 increased by 34 percent—

sA~fomation supplied by Sandia budget and program officers.


a real annual growth rate of 1 percent—DoDfunding, by 256 percent.35

The labs can bring their enormous resources tobear on the most important technical problems; theirnondefense work often has defense applications; andmuch of the technology that the weapons labs havedeveloped for SDI can be transferred to tacticalbattlefield problems. Further, the enormous comput-ing power at the labs—Los Alamos alone hascomputer power equivalent to 60 Cray-ls—is aresource for expanding the defense tech base in amuch more sophisticated way.

In sum, the DOE’s multiprogram laboratoriesmay serve as one (not “the”) alternative model tofacilities owned by the government and operated byits own employees. They have avoided the rigidityof government personnel classifications and much(though not all) of its regulatory apparatus, and theyhave benefited from DOE’s level-of-effort funding.They have the critical mass to move on several frontssimultaneously—although their size, as will be seen,may be a double-edged sword. The final section ofthis chapter examines the relevance of the DOE’sGOCO facilities, and comparable federally fundedR&D centers, to the problems of military laborato-ries.

SUMMARY AND CONCLUSIONS

Growing dissatisfaction with the operations ofDoD labs has led to proposals that some of themconvert to a GOCO status. In substance, this is whatthe DSB tentatively proposed for some labs in its1987 summer study.36 And in one sense, certainparts of DoD might accept such a transition. TheServices have long relied on outside laboratories forsophisticated exploratory work. One thinks of theestablishment of the Aerospace Corporation andLincoln Laboratory as contract research centers forthe Air Force, and the reluctance of the Navy’sStrategic Systems Program Office to use navallaboratories in developing the Fleet Ballistic Missilein the 1950s and 1960s. Since the early 1970s, DoDin general and the Air Force in particular have

moved to reduce the proportion of basic andexploratory research carried out by governmentemployees, with the results noted in OTA’s earlierSpecial Report on the Defense Technology Base.37

To a degree, Service skepticism about the value oftheir own laboratories becomes a self-fulfillingprophecy.

The ultimate justification for converting a gov-ernment facility to contractor operation is that itmore effectively provides the government with aproduct or service, while ensuring that inherentlygovernmental functions are carried out by civilservants. The remaining sections of this chapterweigh the virtues and drawbacks of this approach, inlight of what is known about the operation of DOE’snational laboratories.

Do contractor-operated facilities have greatermanagement flexibility than in-house govern-ment facilities? What are the advantages anddisadvantages, to the government and its operat-ing contractors, of GOCO arrangements?

The evidence is unequivocal in personnel man-agement but ambiguous elsewhere. Clearly, theDOE laboratories have much greater freedom thanDoD facilities to hire directly from the universities,to pay salaries comparable to what industry and theuniversities pay for comparable positions, and tomove people through the organization with relativefreedom. Because the laboratories’ personnel sys-tems reflect those of their operating organizations,they tend to be less bureaucratic and more attuned tomarket conditions than the generality of governmentcenters.

Some of this flexibility carries over into budget-ing and program management. It should be notedthat a significant portion of the labs’ funding is fortech base work and that, within broad guidelines,much of their manpower is earmarked for work forothers. Much of this work is, in a sense, diversifica-tion within the laboratory’s primary mission, ratherthan outside it. Thus at Los Alamos a largeproportion of work for others is sponsored by DoD,although some of it, as in laser technology, may have

W-J.S. Gener~ Accounting Office, op. cit., footnote 28, P. 13.

s~Defense Science Board, op. cit., footnote 1.

STU.S, Congress, Office of TechnoloW As=sment, “Ch, 4—Managing Department of Defense lbchnology Base Programs,” in The DefenseTechnology Base: )ntroducrion and Overvie# Special Report, OTA- ISC-374 (Washington, DC: U.S. Government Printing Office, March 1988).


important commercial applications. There are alsoprograms, like Lawrence Livermore’s work on insitu coal gasification, which grew out of AECresearch into the peaceful uses of nuclear explosives.

Clearly, the labs benefit from a managementstructure that enables the government to achieve itsends through a quasi-industrial system. What theoperating contactors derive from this arrangementis less clear. At one extreme, AT&T, in runningSandia, and Du Pent, in operating the SavannahRiver Plant, are essentially working pro bono.38 Atthe other, Martin Marietta is operating Oak RidgeNational Laboratory for commercial reasons. Itwants the award fee, it wants access to Oak Ridgepersonnel, and it wants access to technology—although Martin Marietta gains access to technologydeveloped at Oak Ridge on terms no better thanother corporations receive.

In an intermediate category are the labs operatedby universities: Los Alamos, Lawrence Livermoreand Lawrence Berkeley by the University of Califor-nia; Argonne by the University of Chicago; andBrookhaven by Associated Universities, Inc. Al-though the University of California receives amanagement fee for operating its laboratories, this isnot the main reason for the long-term relationship ithas had with AEC and DOE. From the University’sperspective, the laboratories enable it to do one ofthe things it exists to do-research. The laboratoriesoffer matchless opportunities to do “big science, ”to use unique facilities, and to develop researchideas. At some university-operated laboratories, asizable number of professional staff hold jointappointments, while many graduate students takesummer jobs that ultimately lead to full-time posi-tions. In these and numerous other ways, theuniversities gain at least as much as they put intorunning the laboratories.

There are, however, three disadvantages to theGOCO arrangement as DOE has adopted it. Thefirst, the sheer size of the Energy weapons laborato-ries, was not inherent in the GOCO status. Rather, itstemmed from the Atomic Energy Commissioners’

decision to make the laboratories full-spectruminstitutions tied to the production facilities. Al-though this arrangement worked well for manyyears, it became more and more difficult formanagement to stay intellectually on top of institu-tions of the size of the weapons labs. All of themhave now placed their own ceilings on institutionalsize, although this owes as much to the constraintsof Gramm-Rudman-Hollings, and the likelihoodthat arms negotiations will lead to major changes inprograms, as it does to a belief that a givenlaboratory has reached its natural limit.

Another problem with GOCOs is a certain lack ofaccountability. True, the operations offices aresupposed to oversee the labs and production facili-ties, but evidence is mounting that the oversight hasnot gone far enough. Perhaps the evidence isstronger at production facilities, like the problemswith reactors and nuclear wastes at the SavannahRiver and Rocky Flats Plants, than at the laborato-ries themselves. Weapons labs like Lawrence Liver-more and Los Alamos oversee each other to someextent; this competition does not exist in theproduction sector. What seems to have developedover many years is a relationship between thegovernment and the operating contractor, withvirtually no continuing external oversight since thedemise of the congressional Joint Committee onAtomic Energy in the mid-1970s.

This leads to the third problem, the abdication oftechnical responsibility by the government. Becausethe AEC elected to contract out almost all of itstechnology development, it happened that virtuallyall of the scientific and engineering expertise residedin the laboratories, with the headquarters organiza-tion at a real disadvantage in evaluating the laborato-ries’ technical programs. This did not mean thatheadquarters or the operations offices could notoverrule something the labs wanted. They could—but for administrative, financial, and political rea-sons, not technical ones. Just as AEC turned overresearch and development to outside organizations,so it also turned over much of its evaluation to

s~[n ]i@t of Du ponl’s decision tO withdraw as operating contractor, the DOE has awarded a contract 10 Westinghouse to operate Savannah River whenDu Pent’s contract expires in 1989.


outside advisory panels. In this respect, DOE is alineal descendant of the AEC.39

The experience of the DOE weapons laboratoriesconfirms the thesis that “the technical capability todo something is often the trigger that causes theestablishment of a national policy based upon thatcapability. ”40 This is true of the DOE weapons labsin a way that it is not of any DoD lab, except for theNaval Research Laboratory and a few engineeringcenters. And yet, because neither AEC nor DOE hadany independent technical organization of their own,they had to defer to the labs on the technical meritsof strategic weapons. It may well be that thedevelopment of many weapons programs or thecreation of a civilian nuclear power industry wouldhave occurred very differently had AEC sponsoredan in-house organization to evaluate its contractors’proposals.

Do GOCOs tend to become more like govern-ment labs, since they face the same pressures toaccount for the use of public funds? To put itdifferently, do GOCOs develop analogs to Fed-eral policies in acquisition, information manage-ment, and personnel, thus losing the flexibilitythat contractual status confers?

There does indeed seem to be a rule that, withtime, contractor-operated and government-operatedlaboratories tend to become more like each other,because both are accountable for their use of publicfunds. In practice, no Federal agency has beenwilling or able to give its contractor-operatedfacilities complete independence to set policieswithin the framework of their missions, even whenthere were no specific regulations to prevent this.Nor does DOE’s delegation of “inherently govern-mental functions” to the national laboratories con-tradict this. An agency can delegate those functions,while micromanaging its facilities in every otherrespect.

As asserted by the Deputy Comptroller General(quoted earlier), the government’s position is that

even when a contractor-operated facility is exemptfrom certain regulations, it must still comply withthe “Federal norm. ” For instance, the national labsmay not be directly subject to the Federal Acquisi-tion Regulation; indirectly, they comply with itthrough regulations that DOE, their prime sponsor,imposes, Again, independent centers like APL needspecial waivers exempting them from full and opencompetition; must be prepared to respond to outsideaudits from different agencies; and must fine-tunetheir accounting systems to reflect the separate typesof appropriations from which their funding origi-nates. All of this adds to administrative overhead andto the demands on technical staff to shield bench-level workers from government paperwork.

GOCO facilities react to these demands in severalways. One is to comply with the spirit of governmentpolicy without being bound by its letter. This is whysome DOE labs voluntarily synopsize their procure-ments in the Commerce Business Daily, reserveprocurements for small businesses, and try to limitthe number of sole-source contract awards. Anotherapproach is to justify a deviation from Federal policyfor special reasons, as the Energy labs do when theyapply for authorization to purchase supercomputers.

For all that is known about the GOCO facilities,it is surprisingly difficult to acquire quantitativeinformation about their operations, partly becauseDOE laboratory contractors are reluctant to supplythe information, and partly because DOE tends totreat it as proprietary. The little that is knownsuggests that the advantages of a GOCO operationmay be overrated. True, at Los Alamos, according toa government source, the contract staff can annuallyhandle some 45,000 small purchases—those under$5000----over the phone. However, anecdotal evi-dence suggests that lead times at some of the Energylabs are at least as great as those at the larger militarylabs, and in some cases greater. To the extent thatthis is true, a GOCO institution is no guaranteeagainst micromanagement and the kind of inflexibil-ity found in government organizations.

WSc)mc ~b~ewers have noted fllal ~D suf{”ers from tie same prob]cm, in ~hat [he headquarters Org:lnixaliOnS that SpOnsOr tic work of in-houseLlaboratories may also lack the technical expertise to judge the pcrforrnancc of those labs. The difference is that DoD cxccutivcs did not dclibcralcly turnover almosl all of the military laboratories and related R&D opcrtilions [o outsi& organi~ations, as the Atomic Energy Commission did with its labs,which were contractor-operated from the start.

~{)Mark and ~>vine, op. cit., fo(lmotc 21, p. 221.

Chapter 5—The Management of Defense Deptzrtment Laboratories ● 83

How and in what ways do GOCOs differfrom government-operated laboratories in theirability to transfer the results of their 6.1,6.2, and6.3 programs to user agencies?

Again, the GOCO DOE laboratories show a muchgreater ability to move from basic research down thespectrum of technology development. DOE GOCOsare closer to the ultimate application than DoD labsare. Specifically, the weapons labs’ responsibilityextends from basic research to the retirement ofweapons in the national nuclear stockpile. Asmentioned above, the labs are only part of avertically integrated complex that extends frombasic research through the production of weapons-grade materials to the assembly of the weapon itself.In DoD, by contrast, the process by which technicalwork at the laboratories eventuates in operatingsystems is much harder to trace.

The very depth of expertise at the DOE weaponslaboratories has two notable effects. The first is theircommitment to a substantial amount of basic re-search and advanced exploratory work. Where aDoD facility might do only enough exploratoryresearch to keep abreast of technology—keeping, asit were, a window on the world—the nationallaboratories tend to be more aggressive. They canafford to be: Where expertise at a DoD lab might goto a depth of two or three persons, at the national labsit can encompass an entire branch working inleading-edge technologies. Second, weapons-related work and nondefense programs cannot besegregated in terms of research. There is a constantgive-and-take in these areas that leads to new ideasand new applications. Thus, the Los Alamos MesonPhysics Facility (LAMPF) was originally designedin 1967 for research into the structure of the atomicnucleus. That research led, a decade later, to thespinning off of a separate group, leading in turn toneutral particle beam work that became a coretechnology for SDI.

In sum, the national laboratories have theresources to develop aggressively their portion ofthe defense technology base. In their operatingagreements, DOE specifically recognizes basic re-search as a function that needs no extraneous

justification. Beyond that, the interplay of technolo-gies, the respect the university community has forthe laboratories, and the tech base tax that the labsplace on work for others, give them a markedadvantage over the DoD labs in technology develop-ment.

What mechanisms can both kinds of institu-tions use to diversify within their basic missions?

Because the DOE laboratories have construedtheir missions in the broadest way, they havemanaged to diversify within, rather than outside,those missions. A more fundamental differencebetween Energy and Defense laboratories is that theformer consider such diversification an integral partof their missions. The latter have diversified inresponse to directives from organizations external tothe laboratories-the Service commands or OSD.Thus, the Army Laboratory Command developed astrategy for investing in next-generation and “no-tional” systems; and the Air Force sponsoredProject Forecast 11 as part of its tech base strategy.Much of the military’s tech base work will occuroutside its own laboratories, while the reverse is truefor the DOE.

Given the capabilities of the DOE weapons labs,it would be surprising if their primary mission didnot spill over into related areas. Siegfried Hecker,Director of the Los Alamos National Laboratory, hasput it succinctly. At the weapons labs, “nondefensebasic and applied work is done in an environmentoriented toward national defense. Thus, multiplepayoffs are common and occur quite naturally.While contributions are made to the solution ofnondefense problems and significant additions aremade to the international scientific knowledge base,considerations of potential defense applications ofresearch results come as a natural by-product. ”41

This approach has led the DOE laboratories tomove increasingly into nonnuclear defense work.Although the DOE technology base developedseparately from that of DoD, the two are converging.Here, too, Los Alamos has moved aggressively:applying (as mentioned before) LAMPF to theneutral particle beam program for SDI; working withthe Army, the Marine Corps, and DARPA on the

qlslgfricd Hecker, Las Alamos National Laboratory, “Review of Managmncnt of the Nation’s Defense Twhnology Base, ” testimony at hearingsbefore the Subcommittee on Defense Industry and lkchnology, Senate Armed Services Committee, Mar. 18, 1988, p. 15.

&# ● Holding the Edge: Maintaining the Defense Technology Base

armor/anti-armor program; and investigating theuses of free-electron lasers for ground- and space-based weapons. Given the Energy labs’ computingpower, they can apply sophisticated modeling to theproblems they choose to attack.

If the experience of the DOE laboratoriesdemonstrates anything, it is that the greater the depthof expertise, the greater the ability to apply it toproblems pertinent to the organization’s mission.Most DoD laboratories lack this ability to movequickly into new areas, first, because most military“laboratories” are really engineering centers; sec-ond, because their charters restrict their freedom ofaction in any case; and third, because they lack staffand facilities comparable to those of the nationallaboratories. It will take fundamental changes beforeDoD laboratories can make greater contributions tothe defense technology base. Such changes couldinclude closing some facilities, consolidating othersinto weapons development centers, or convertingsome laboratories to GOCO facilities. This thirdoption is considered below.42

Can the relations of government laboratoriesto their sponsor agencies be placed on a quasi-contractual basis comparable to those of GOCOs?Are any government labs considering such arrange-ments?

It is possible to imagine an arrangement underwhich government laboratories could take on theflexibility of GOCOs while remaining government-operated. For the sake of argument, a laboratorycould combine the China Lake personnel systemwith the freedom NIST has to seek support fromFederal customers, an NIF-style funding scheme,and an approach like NASA’s effort to turn over itssupport functions to contractors-in fact, just whatthe DSB had in mind when it recommended thatDoD sponsor a laboratory management demonstra-tion. An organization run along those lines wouldhave a degree of freedom that few governmentfacilities now enjoy.

But such incremental improvements might not gonearly far enough, The China Lake system does notmake government salaries more competitive withthose of the private sector. Demonstration projectsrarely have unambiguous results; and as China Lake

shows, they tend to remain insulated from othergovernment establishments, perhaps out of fear thattheir example might metastasize through the system.

Similarly, NIST’s freedom to take on work forother agencies or the standards community resultsfrom its unique mission to support the U.S. technicalinfrastructure. And the contracting-out of supportservices raises legal and political issues, such as howone distinguishes between services like running acafeteria and providing scientific computing serv-ices, which is more mission-related. In other words,when do such services impinge on governmentalfunctions? How can one avoid the on-site supervi-sion by government employees of contractors,which is illegal? And how can an agency avoid theinefficiency of converting base operations, functionby function? In any case, such hybrid facilitieswould remain bound by government policies inprocurement and accounting that would attenuatethe freedom gained in other areas.

A single-step conversion to GOCO status couldavoid these problems while bringing others in theirtrain. However, all the GOCO institutions describedearlier have enjoyed that status since their inception.If a Defense laboratory were to convert to GOCO, itwould be the first instance of an existing facilitytaking that route. Because it would be unprece-dented, a changeover would be very complex. Therewould be numerous issues to be resolved along theway: changeovers in employee benefits, relationswith the Office of Federal Contract CompliancePrograms, the need to restructure its procurementauthorities, and the like. But the main issue confront-ing such an institution would be the nature of itsrelationship with its sponsoring agency. Under thenew arrangement, the laboratory might be operatedby an industrial contractor, a university, or even thelab’s senior managers acting as a corporate body.

Alternatively, a laboratory could elect a hybridstatus-contracting out all support functions, whileconducting R&D as a Federal entity. A few NASAfacilities, like the Johnson Space Center, haveadopted such a mixed system, with all supportfunctions turned over to a prime base supportcontractor. This has several advantages: for exam-ple, the contractor has direct-hire authority for

Q-@n the firs( two options, see ch. 7 below.

Chapter .5--The Management of Defense Department Laboratories ● 85

professional staff if their work falls within thecontract’s scope, and it can pay them market rates.But the legal questions—the demarcation betweencommercial and inherently governmental functions—remain exceptionally complex.

What kind of DoD laboratory is the bestcandidate for conversion to GOCO? While anyfederally operated laboratory would benefit fromfewer restrictions on operations, it seems that only acertain kind of institution is a suitable candidate forGOCO status. It should already have a substantialinvestment in tech base work; it should be able todemonstrate that its operating problems cannot besolved by minor variances from regulations; and itsimportance to its DoD customers should be such thatthey have a stake in improving its operations. In thislight, NRL and some of the larger naval R&Dcenters would appear to be suitable candidates forGOCO status.43

A military laboratory taking this route wouldhave problems to resolve that the DOE laboratorieshave never faced. One would be determining theorganization for which it would be working. It mightbe one of the Services, a Service command, or even

OSD, as is the case with some FFRDCs. Anotherconsideration is that such a conversion could well beirreversible. Because conversion to GOCO could noteasily be undone, making the transition successfullywould require a strong commitment on the part oflaboratory employees as well as DoD. Conversion toGOCO could not occur without the full support ofthe relevant Service, as well as OSD.

Finally, in return for the benefits of GOCO, thelaboratories would also give up something. The newstatus would mean weaker ties with Defense organi-zations, and perhaps a tendency on DoD’s part totreat the reorganized institution as simply anothercontractor. Further, the preceding analysis of DOElaboratories suggests that the benefits from a GOCOoperation tend to diminish over time. In short,everything would depend on the sponsor’s willing-ness to give the laboratory the freedom to strike outin new directions, and to take on work for others thatdrew on its capabilities. Whatever organizationassumed the operation of the laboratory would haveto have specialized management skills that wouldjustify turning the lab over to an outside contractor.

dqon tie c~cumstmces that might justify conversion to GOCO, see ch. 7 below.

ExplmentManage

CONTENTS

OVERVIEW. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89The Changing Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .. .. ., . . . . . . . . . . . . . . . 89OtherExperiences and Concepts-How Applicable? .............. . . . . . . . . . . 90

PLANNING AND PRIORITIES . .. .. .. .. .. .. .. .. ... ...+,..... . . . . . . . . . . . . . . . . . . 90Top-Down v. Bottom-up, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90Balancing Near- and Long-Term Research .. .. .. .. .. . .., ... . .,., 94The Role of Special initiatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95Centers of Excellence . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96MANAGEMENT AND CONTROL . . . . . . . . . . . . . . . . . . . . .. .. ., .. .. .. ... .., .,..... 98

Which Focus: Technologies omissions? . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ....... 98Balancing Intramural and Extramuml Research . . . . . . .. . ........ , .. .. .. ...... 101

GETTING RESULTS-THE “BOTTOM LINE” OF RESEARCH.........,.. . . . 102Applications—Moving Technology From the Laboratory to the Marketplace . . . . . . . 102Collaboration in Research and Technology . . . . . . . . . . . . . . . . . , ., . . ., . 104

Chapter 6

Exploiting Other Management Approaches

OVERVIEWAs the U.S. Department of Defense (DoD) and

Congress look for ways to solve problems associatedwith managing technology programs and researchfacilities, they will find that they are not alone in thisconcern. Other organizations—large corporations,foreign governments, and international groups—face similar challenges, and have taken a number ofapproaches with varying success.

For two reasons, the experiences of these otherorganizations warrant attention. First, they form partof the backdrop against which DoD will operate—the corporate structures and the foreign govern-ments with which DoD will cooperate to implementits research and technology programs. Second, theseorganizations exemplify other approaches to accom-plishing related tasks and addressing similar man-agement problems. They offer models that might beadapted for DoD’s purposes.

This chapter concentrates on organizational andmanagement techniques, but along the way itaddresses the emerging international defense-industrial environment within which DoD willoperate in the next decade.

The Changing Environment

Research managers—both in government andindustry-must find ways to keep up with the rapidpace of science and technology. In industry, timelyapplications of technology are essential to remaincompetitive. In the case of defense programs, theWest must maintain a credible deterrent in a periodof political change and uncertainty, Rapidly increas-ing research costs and diminishing financial re-sources are also causing U.S. Government officialsand their industrial counterparts to rethink theirresearch programs. The overall structure of researchprograms has been under scrutiny, especially themix between long-term research and near-termapplications. The use of new approaches to researchand technology development—such as university-based or industrial centers of excellence—to direct

interdisciplinary resources toward key research goalshas grown in popularity. Policies governing intra-mural v. extramural research—the industrial re-search equivalent of make-or-buy-are being de-bated. On this point, the U.S. Government is uniquewith its large and wholly owned laboratory structure.At the opposite extreme, some governments andcompanies maintain no internal research capabilitiesat all, depending entirely on technology developedby others. Research management methods are alsounder review, with particular emphasis on thequestion of whether centralized management shouldreplace independence at the researcher level.

Whether to collaborate or “go it alone” isbecoming a major issue. In general, full-spectrumlaboratories (i.e., those capable of conducting basicresearch, advanced development, and engineering)appear at odds with those offering specializedcapabilities (e.g., centers of excellence) and focus-ing on selected research topics. The disparity arisesbecause full-spectrum laboratories are often orientedalong mission lines and, for reasons of efficiency orsecurity, prefer to work alone; whereas specializedlaboratories must interact with other organizationsto get the job done. There is a growing attitude ingovernments and the private sector that collabora-tion, with all its inherent difficulties, is perhaps theonly practical way to finance basic and appliedresearch on contemporary topics in science andtechnology. Independence, whether for individuallaboratories, companies, or countries, is becomingfinancially prohibitive, and those who insist ongoing it alone are increasingly at a competitivedisadvantage relative to those who collaborate.

Finally, incorporating laboratory technology inproducts carries a high priority for companies andgovernment officials alike. While different ap-proaches have been taken to encourage bettertechnology transfer, they all involve giving someonethe responsibility and the authority to ensure that theprocess occurs. While this is simple in concept,making it happen is not easy, and few organizationsdo this job satisfactorily.

–89-


Other Experiences and Concepts—How Applicable?

In this chapter, approaches to technology basemanagement employed by other Western govern-mentsl and by the private sector are examined to seeif the successes-or failures-of others can serve asmodels for DoD. Japan, major European govern-ments (those with significant military technologybases), and U.S. and European defense companiesare the primary subjects. The defense sector ofprivate industry was selected because its researchmethods and objectives are viewed as being moreconsistent with DoD’s than, say, pharmaceuticals orconsumer products. The specific approaches varyaccording to the nature of each national or industrialresearch program, the level of financing, and theavailability of skilled researchers to do the work. InEurope, some interesting trends are developing, withgovernments adjusting their research programs andpriorities to changes in world markets, the post-INFEast-West political environment, and the advent ofEuropean economic integration in 1992.2 A fewmajor themes emerged from OTA’s review:

Most governments and companies have someform of “research policy,” which is approvedat the top and promulgated throughout theorganization. While some latitude is still al-lowed and innovation encouraged at the re-search level, projects must be justified on thebasis of their contribution to achieving eitherscience and technology policy objectives orbusiness objectives.

There is a trend toward centralized manage-ment of research programs, with an increasingemphasis on periodic, and relatively frequent,reviews to assess actual v. planned progress.Managers appear to be more willing to cut theirlosses when projects continue to miss mile-stones, and to look outside to acquire technol-ogy developed by others.

Private and public organizations see collabora-tion as a means of affording research programsthat are of significant magnitude and have

promise of meeting overall policy or businessobjectives.

It is tempting to suggest outright that DoD shouldembrace these themes; indeed, DoD has alreadybegun to in some respects. However, their applica-bility to the Department’s overall missions andresponsibilities may not be entirely clearcut. Fur-ther, making sweeping changes to DoD’s structurerequires great caution; it may result in severedisruptions instead of the promised improvements.The magnitude and scope of DoD’s technology baseactivities dwarf those of nearly every other organiza-tion in the world. DoD’s annual Science andTechnology (S&T) Program (i.e., budget categories6.1, 6.2 and 6.3A) is approximately $10 billion,whereas the United Kingdom’s (U. K.) Ministry ofDefence (MoD) equivalent is less than $1 billionand, in West Germany, military S&T is just slightlymore than $500 million. The U.S. Army, Navy, andAir Force each spend more on defense technologybase activities than these other nations.

It may be that DoD’s S&T Program is too largeand diversified to employ effectively the manage-ment techniques of smaller, more manageable or-ganizations. Nevertheless, there do appear to bemethods that could be applied at least in part to DoD.This chapter highlights some promising approachesto the three broad issues that appear to be occupyingthe minds of U.S. Government officials and corpo-rate executives: 1) planning and priorities; 2) man-agement and control; and 3) getting results.

PLANNING AND PRIORITIES

Top-Down v. Bottom-Up

DoD employs a highly decentralized approach toscience and technology planning. (See chapter 4.)The three Services define their research needs withonly minimal direction from the Office of theSecretary of Defense (OSD), and individual re-searchers exert substantial influence over programcontent and priorities, By contrast, most otherWestern governments are involved directly at thehighest levels in setting national research objectives—

*In this chapter, ‘‘Western governments” includes Japan.

‘The “Single European Act,” passed by the European Parliament in 1985 and ratified by European Community (EC) mcmbcr nations in 1986, hasput in mo[ion a set of measures tha[ will Icad to a standardized and largely intcgratcct financial and trade system in Europe in 1992. This should, in turn,result in stronger and more competitive European industries opera[ing m world markets.

Chapter 6--Exploiting Other Management Approaches . 91

in some cases addressing both defense and civilresearch with a ministry review committee. Throughthe use of centralized research committees oradvisory panels, cabinet-level officials set prioritiesand take steps to ensure that the government’swishes are translated into specific programs con-ducted by their laboratories or by the private sector.These priorities strongly influence the content ofresearch programs at all levels. The Japanese Gov-ernment skillfully ‘‘influences” civil research ac-tivities, and the same trend is seen in severalEuropean governments. The European Community(EC) is exerting top-down influence over the scopeand content of the member countries’ researchprograms. This influence is sure to grow as theCommunity works toward its 1992 economic inte-gration.

Japanese GovernmentThe Science and Technology Agency (STA),

Ministry of International Trade and Industry (MITI),and Ministry of Education constitute the threelargest players in Japan’s Government-directed re-search and development enterprise. Much of the sizeand influence of the Ministry of Education isattributable to its responsibility for managing educa-tional research facilities. The other two institutionsare deeply involved in planning and priorities. Thereexists in Japan a broad consensus on the value ofresearch and development (R&D) efforts that pro-vides a stable political and economic environmentfor the pursuit of long-term goals. Bureaucraticorganizations and more politically oriented groupshelp ensure the preservation and continual assess-ment of that consensus. STA, for example, isorganized under the office of the Prime Minister,while MITI’s research programs report directly tothe head of the ministry. Scientific research trendsare monitored and influenced by advisory councilsassociated with the office of the Prime Minister.These councils fulfill multiple roles, includingfacilitating a cabinet-wide consensus on governmentresearch policies, allocating resources, and legiti-mizing initiatives developed in the private or publicsector by publicly endorsing them. Council reports

can have a considerable impact on progress inspecific research fields. Space exploration, forexample, has become a national priority, in partbecause of the role played by these advisory councilsin articulating the government’s objectives andgaining national support for them.

The process is not flawless. Inter-ministryintegration and cooperation in Japan are not alwaysas thorough as they could be. There have beeninstances in which ministries have competed againstone another for prominent roles in research initia-tives, forcing political compromises and wastefulduplication. And important initiatives can fail, evenwhen there is a clear consensus in the governmentand industry. However, Japan’s track record ofsuccessful R&D provides a strong vote of confi-dence for the top-down approach to planning.

British Government

Perhaps the most visible and dramatic movementaway from independence and toward centralizedresearch planning is the one now under way in theU.K. For the past 2½ years, the U.K. ’S policies forR&D have been subjected to intense scrutiny by theBritish Government, Parliament, industry, and thescientific community. In mid-1987, the governmentpublished plans for sweeping changes in the man-agement and funding of R&D in the U.K.3 Theproposals, which emphasized the economic poten-tial of research, were drawn up following sharpcriticism of the government’s annual R&D effort bya House of Lords Select Committee.4 The Lords hadsaid that the government’s R&D strategy lackedcoordination, particularly in the way research wasapplied to industry. If science and technology wereto restore and sustain economic growth and prosper-ity, the Committee said, its promotion should be acentral objective of government policy, with theimpetus coming from the Prime Minister.

On a related issue, a 1987 review of government-funded R&D in the U.K.5 reported that MoD spent52 percent of all government R&D funds in the year1985-86. This high proportion of total R&D dedi-cated to defense generated widespread concern

S“Civil Research and Development, ” Cmnd 185 (London: Hcr Majes[y’s Stationery Office, July 1985).Q“civll Re~arch and ~velopment: Report of the Select Committee on Scicncc and Technology, ” Vol. 1 (HL 20-1), British Parliament, House of

Lords, November 1986.56’ 1987 Annu~ Review of Government Fwded R& D,” Government Statistical Service, United Kingdom, 1987.


among British economists and industrialists thatdefense might be crowding out valuable investmentin the U.K. civil sector. In its 1987 Defence WhitePaper, 6 the government noted this concern andpromised to take a closer look at defense programswith a large R&D content to ensure that governmentfunding was essential. Significant reductions infunding within 2 to 3 years were predicted as defenseR&D became more efficient and competitive, and asBritain reduced its duplication of Allies’ researchefforts through greater collaboration. The aim was torelease more government money to support the civilsector, both in industry and academia. In addition,there was a clear desire (both in the BritishGovernment and industry) for enhanced civil spi-noffs from the R&D carried out by the government’sDefence Research Establishments. Several initia-tives were introduced, both to exploit the Establish-ments’ technologies for the benefit of the civilsector, and to offer selected defense facilities for useby industry. In a potentially dramatic move, theEstablishments may be combined into an inde-pendent Defense Research Agency that must “sell”its research to the Ministry of Defence, industry, andother customers (e.g., universities, European andAmerican industries, and consortia).

In implementing this new R&D policy, theBritish Government outlined two challenges: 1) totarget scientific and technological resources withoutconstraining individual creativity; and 2) to coordi-nate related parallel R&D programs without divorc-ing them from their individual objectives.7 Tosupport this policy there is collective ministerialconsideration, under the Prime Minister’s leader-ship, of science and technology priorities. Also, thegovernment is to be advised by an independent bodythat will comment not only on British scientific andtechnological endeavors, but on international effortsas well. The government’s stated aim is to harnessBritain’s total R&D resources, both civil and mili-tary, in a science and technology program that willenhance both the U.K. ’S economic growth and itsdefense capability. The planning and execution ofthe more-or-less independent civil and militaryprograms are to be coordinated and monitored by a

government committee to ensure ‘value for money”and objectivity, to avoid duplication, and to maxi-mize cross-fertilization between the two efforts.

French Government

In France, top-down research planning has beenthe rule, and appears to remain firmly entrenched.The policies for government funding of French R&Dare highly centralized, but civil and defense R&Dare budgeted and administered separately. Innova-tion and exploitation are encouraged by an elaboratesystem of aids and incentives; economic growth issought through market-driven high technology; anddefense R&D is expected to contribute to the overalleconomy. Policies for nationalized firms and thegovernment-supported research system are framedin the context of long-term plans for R&D andinnovation, with relatively specific priorities andgoals. Science and technology policies (especiallytechnology) are integrated wherever possible withthe government’s industrial and broader economicpolicies.

The stated aim of French R&D policy is tostimulate rapid, science-based economic growth,with a selection of key enabling technologies givenpriority in either national or collaborative programs.These goals have subsequently been reflected inlegislation. The draft 1987 R&D Budget Plan wastouted as an essential element in reviving the Frencheconomy. In the government’s view . . the field ofresearch and technological development is a funda-mental component of that policy, because researchand technological development are seen by everyoneas being a powerful factor for the long-term develop-ment of our economies and providing a decisiveadvantage in present day economic competitionworldwide. “8

West German Government

West Germany presents an interesting contrast.The Federal Government’s philosophy for civilresearch encourages independence. Bonn only pro-mulgates general guidelines while a complex andlargely informal network of Federal and StateGovernment organizations, universities, private re-

~$$statement of tie ~fcnce Estimates 1987,” Cm 101-1 and-n (London: Her Majesty’s Stationery office, 1987).

T’ $1987 Annu~ Review of Government Funded R& D,” op. cit., footnote 5.

~Draft 1987 French R&D Budget pl~.

Chapter 6--Exploiting Other Management Approaches ● 93

search groups, and industries observes priorities andmoves research projects toward applications. In theMoD, however, things are different. The Bundes-wehr Plan, coordinated between all three Services,forms the basis for the MoD annual budget estimate.In addition, there are two MoD agencies concernedwith procurement, but not part of the militarydepartments. The Federal Office for Military Tech-nology and Procurement (BWB) is the principalbody responsible for carrying out procurementplans. The Armaments Division is concerned withprocurement planning and the coordination of tech-nological areas that are considered ‘‘project-free”(i.e., basic research not tied to specific applications).Reporting to the Division head is the Commissionerfor Defense Research, who collates the researchrequirements from all three Services-includinginternational aspects—into the overall research pro-gram.

In 1986, all responsibilities for Research andTechnology (R&T) program formulation and execu-tion were assigned to the Armaments Division. MoDhas defined three categories of research: basictechnology; future technology; and systems technol-ogy, which are roughly equivalent to the U.S. DoD’s6.1, 6.2, and 6.3A. These have been broken intotechnology elements (100 in all) with an ArmamentsDivision Technology Coordinator assigned to each.The Coordinator prepares an annual plan thatincludes overall goals, a survey of the state of the art(in Germany, Europe, and worldwide) and taskdescriptions (with milestones and a 5-year fundingprofile). The Coordinator also prepares bid requests,evaluates proposals, makes awards, and monitorscontracts, Roughly 25 percent of research contractsare delegated to BWB for placement. However,direct control remains in the hands of the ArmamentsDivision Technology Coordinator.

West Germany commits about 15 percent ofgovernment-funded R&D to defense-related R&D.This is spent within the defense-related industries,the national laboratories (not owned by MoD), andthe Fraunhofer Society,9 which has six of itsInstitutes devoted to defense research funded by theMinistry of Defense. The defense research and

technology budget is roughly DMIB (U. S.$550M)annually. For defense research the message is clearthat centralized planning and control has become therule, and that duplication cannot be tolerated. Theresimply is not enough money for laissez-faire.

Civil research presents quite a different picture.The West German Government’s civil R&D budgetis nearly seven times that of MoD. It is augmentedby a nearly equal sum from State Governments, andis spent largely by universities, the national labora-tories, and independent research organizations (i.e.,the Max Planck and Fraunhofer Societies). Throughthis decentralized system, West Germany has beeneminently successful in promoting technology-based economic growth, a fact that seems to call intoquestion the wisdom of instituting a national re-search program based on centralized planning andcontrol as the U.K. ’S appears to be doing. Lookingbehind the scenes, however, German civil researchis anything but laissez-faire. The Max PlanckSociety exerts a major influence on research priori-ties and the Fraunhofer Society, in conjunction withfinancial support from German industry, serves to“‘pull” the products of research out of the laborato-ries in accordance with identified market priorities.The “system,” although not set down in formal,government-wide procedures, is apparently wellorchestrated and effective, as Germany’s record ofindustrial growth and its world leadership in exportswill attest.

Private Industry

Turning to the private sector, in recent years mostEuropean and U.S. companies have instituted top-down planning systems—although specific researchprojects are increasingly set and executed at adivision (or operating company) level. In Japan,top-down planning has always been the rule. Thecommitment of top management in Japanese compa-nies to promoting technological advances withintheir companies is, perhaps, unparalleled. The partici-pation of high-level managers and corporate offi-cials varies from one firm to the next, but there iscorporate-wide awareness of, and support for, re-search. Funding decisions frequently are made atsenior levels, and failing projects are abandoned

gThe Fraunhofer SWiety is a nonprofit society that sponsors and performs applied R&D. Its clients are German industry ~d the Federal ~d StateGovernments, and it is influential in setting the direction of German appiicd research. For basic research, the Max Planck Society pcrforrns a similarfunction.


quickly, usually without their initial supporterssuffering adverse consequences.

Most major companies have elaborate proceduresfor establishing long-range business objectives,including an assessment of the key technologies thatare expected to contribute to their achievement. Forcertain “enabling technologies” (i.e., those tech-nologies with broad, corporate-wide applications),the highest management levels are involved indecisions on projects and funding. Often, corporate-level centers of excellence are established to bring acritical mass of resources to bear on the assignedtasks—a reflection of how capital-intensive researchhas become. These centers generally involve generictechnologies (e.g., advanced materials, artificialintelligence, optoelectronics, and microelectronics),and steps are taken to ensure that the divisions usethe results in applied research, design, and develop-ment programs,

To bring applied research closer to market, theresponsibility for research management no longerrests solely with corporate central research laborato-ries; rather, the operating divisions or companies(where the profit and loss responsibility lies) aretaking charge, The bulk of corporate internal R&Dis thus directed toward achieving near- to mid-termbusiness objectives, usually tied directly to identi-fied customer requirements (e.g., DoD and MoD) orto new products. Whether this trend is good or badwith respect to industries’ contribution to scientificknowledge and national technology bases, it clearlydemands a firm approach, one that focuses on thebottom line and does not accept laissez-faire.

Balancing Near- and Long-Term Research

Defense Research

There is almost universal criticism that defenseresearch programs in the United States and inEurope, both in governments and in the privatesector, focus too much on the near term. This isprobably a fair criticism. Engineering tasks are oftenconducted by DoD laboratories under the guise oftechnology base projects. U.S. companies directtheir Independent Research and Development (IR&D)

toward modernizing programs or improving currentproducts and systems, rather than basic research.10

And as DoD’s IR&D recovery program has comeunder increasing pressure from both Congress andthe Administration, defense companies have re-sponded by focusing even more of their IR&Dinvestment on those areas of technology likely toprovide a near-term commercial payoff. Criticsclaim that the “R” in IR&D is silent. Without agovernment IR&D allocation, European companiesare even more likely to spend R&D funds onproducts, rather than research.

This situation has become further entrenched inrecent years as governments have reduced thepercentage—and in some cases the actual level-ofdefense expenditures for basic and applied research.In the United States, for example, DoD’s research(6.1) budget did not keep pace as defense R&Dbudgets increased under the Reagan Administration.In France, under Chirac, overall research was cut,with civil research taking the biggest “hit”; in theU. K., defense research is being constrained, ostensi-bly to prevent the crowding out of civil research.Defense research in West Germany is really toosmall to make a difference in the ‘‘high-tech” game.

Civilian Research

Overall, it appears that long-term civil researchby European governments—and by the EC—isenjoying a resurgence, with both industry andacademia benefiting from this trend. This is not anaccident. Influential observers argue that the great-er the proportion of a government’s researchbudget spent on civil research, the stronger thatcountry will be in world markets and, therefore,the more prosperous it will become. Japan andGermany are clear examples of this theory. Defenseresearch is viewed as a drag on the economy, andgovernments are being urged not to overspend in thisarea. In curtailing defense research, some governments-notably the U.K.—have put the burden on industry.Companies are being urged to conduct research,either under publicly funded civil programs or withprivate funds, and to apply the results (in a moremature form) to defense needs.

l~sonle Obscwers ~]icvc that dc~rca~s in (j.L funding could bc a cause; 6.2 funding has dccrcascd aboul 1S percent Over the p~sl two decades.

Chapter 6--Exploiting Other Management Approaches ● 95

Japan

Japan is noted for its long-term outlook onresearch and technology. Both the public and privatesectors adhere to this philosophy, with corporatestrategies keyed to the exploitation of future tech-nologies. This is in stark contrast to the situation inthe U.S. and European defense sectors.

Several factors contribute to Japan’s ability tofocus on the long-term. Cultural factors are impor-tant, but there are other reasons, many of themfinancial. For example, in the United States the valueof a company’s stock influences business decisions.U.S. managers rank profits and increased share priceas their primary objectives. Japanese executives alsoview profits and return on investment as important,but put them below market share. Further, Japanesecompanies do not need to worry about the price theirstock commands. Equity remains less important thandebt in corporate financing, and new stock issues arethe exception in raising funds. Also, the lower costof capital in Japan makes long-term projects moreattractive.

A definite shift in emphasis is apparent: Whiledefense R&D was once the “locomotive” foradvancing technology, civil research appears tobe assuming this role in much of the Westernworld. In part, this is because the line between civiland defense technology is fast disappearing; and inpart it is because governments are moving toimprove their industries’ competitiveness in emerg-ing global markets that are technology driven. Indeveloping policies and priorities for defense re-search, DoD officials are sure to come underincreasing pressure to take a wider view of nationalsecurity. Indeed, maintaining an adequate defenseindustrial base may only be possible through main-taining competitive U.S. industries in world mar-kets. The Europeans appear to have offset a near-term focus in defense with support for the long termin civil research. Their industries are finding suffi-cient government- (or EC-) sponsored civil researchprograms to challenge both existing scientific andtechnical staffs and available resources.

The Role of Special Initiatives

Historically, the pace of U.S. science andtechnology has benefitted greatly from a successionof special top-down initiatives, driven either byurgent defense priorities or by political objectives.The World War 11 Manhattan Project resulted notonly in the atomic bomb, but also in an array oftechnologies that served both military and civilpurposes for more than a decade. The ApolloProgram of the 1960s succeeded in meeting Presi-dent Kennedy’s objective to put a man on the Moon;but it also provided breakthroughs in materials,electronics, data processing, guidance and control,and propulsion. Other past technological initiativessuch as Project Sherwood (controlled nuclear fu-sion), Project Plowshare (peaceful applications ofatomic weapons technology), and Vanguard (rocketdevelopment) were not successful, but they alsoprovided beneficial spinoffs in other areas. (TheStrategic Defense Initiative [SDI] program may alsoprovide spinoffs to conventional defense programsand to some civilian fields, but the returns are not yetin.)

While these initiatives created an environmentthat encouraged rapid advancements in science andtechnology, they also disrupted the normal course ofresearch and have, some argue, thereby underminedthe Nation’s long-term technological health. It isstill an open question whether these initiatives haveprovided a net benefit to science and technology, orif S&T would be better left to follow a more naturalcourse.

Until recently, the Europeans have made little useof special technology initiatives. Although concen-trated efforts have been applied in major develop-ment programs [e.g., the Mane spacecraft, theTornado aircraft, the Airbus, and now the EuropeanFighter Aircraft], national research and technologyprograms were relatively well-insulated from politi-cal pressures. But growing concern that Europe isfalling behind the United States and, perhaps moreimportantly, Japan in world markets has changedthis attitude dramatically. Technological initiativesare rapidly becoming the rule in Europe, rather thanthe exception. These initiatives have been mainlymultinational in nature, directed from either the ECor other multinational groups, such as the Independ-


ent European Program Group (IEPG).1l They alsoappear to have provided a “ramp effect” in severalkey technology areas, propelling Europe to a level oftechnology that is close or equal to that of the UnitedStates. While the EC research projects are for civilpurposes, most involve “dual-use” technologies,and many of the results will no doubt find their wayinto Europe’s military systems. The impact of theseinitiatives, therefore, will be relevant to future DoDdefense technology base programs.

The EC is sponsoring several research andtechnology initiatives, headed by ESPRIT (Euro-pean Strategic Program of Research in InformationTechnology). The loosely defined intergovernmentalEUREKA 12 program launched by France in 1985has attracted support from 19 European countries.After early successes—and a lot of publicity—Europe’s governments encountered difficult ques-tions on priorities and funding. The EuropeanCommission proposed a substantial research budgetincrease for the next 5 years. This proposal wassupported by the southern-flank European countriesbut opposed by the major budget contributors (theU. K., France, and West Germany), who urgedfinancial constraint and stringent selection to ensurethat funded projects broke new ground. The mem-bers finally agreed in September 1987 to spend 5.2billion ECU (U.S. $6.8 billion)13 on a “EuropeanFramework” for technology collaboration over thenext 5 years. Within the Framework are severalindividual initiatives addressing, for example, infor-mation technology, advanced telecommunications,biotechnology, alternative energy sources, environ-mental research, and nuclear safety. These initiativeshave been translated into specific research pro-grams, such as ESPRIT, RACE, and BRITE.14 Yet,the Commission does not fund EUREKA, whichcould approach $5 billion in itself.

None of the more than 200 ESPRIT and 165EUREKA projects have as yet yielded break-

throughs, although progress is claimed in manyareas. The most important contribution may bepsychological, with the shedding of isolationistattitudes and inhibitions. Reaction by Europeanindustry and academia to the EC programs variesfrom enthusiasm to open skepticism. To someEuropeans, subsidized EC research collaborationadministered by officials in Brussels is not a cure forEurope’s problems. It might, they warn, evenimpede healthy change by accepting too readily theestablished industrial order. ESPRIT, for example, isdominated by a dozen big electronics groups. Itremains to be seen if these European high-techcompanies can actually cooperate in product devel-opment and marketing, thus capitalizing on the EC’sinvestment in research.

Technological initiatives of some significancemight also evolve from Europe’s defense commu-nity. Driven by decreasing defense budgets, themember countries of IEPG have, after more than adecade of trying, finally begun to develop a coherentprogram for cooperation in research, developmentand acquisition. One of the IEPG’s first actions wasto establish a set of cooperative research projects.Little has come from this effort to date, but muchmore visible progress is being seen on joint develop-ment and production programs. The 1987 report ofan Independent Study Team15 signaled clearly thatthe IEPG would henceforth be a primary forum forcollaborative defense programs within Europe, andthat it would increasingly become the “singlevoice” on acquisition and cooperative issues involv-ing the United States and Canada.

Centers of Excellence

Special research teams or “centers of excel-lence” are becoming a favored means to implementresearch priorities. These groups concentrate oninterdisciplinary research relating to technologiesthat require a critical mass of resources and person-

1 IThe IEPG is comprised of the 13 European members of NATO: Belgium, Denmark, France, West Germany, Greece, Italy, LuxemboWg, theNetherlands, Norway, Portugal, Spain, ‘Ihrkey, and the U.K.

lzThe Europe~ Rese~ch Coordinating Agency is a Europcan program to strengthen non-military technologies, emphasizing JOint industrial mdgovernment funding of civil projects that have clear market potential.

13 EWoPm budgets for rcwwch ~d tw~olo~ Usual]y do not include rese~ch overhead (e.g., general facilities ~d administration). AIso, othergroups (primarily industry) are expected to contribute up to an equal share. Thus, the EC’s $6.8 billion 5-year research budget is actually equivalent toa much larger amount in terms of, for example, a DoD budget.

14 RACE: Rese~ch ~d ~ve]opment in Advanced Communications for Europe. BRITE: Basic Research into Industry Rchnology for Euro?=

lsIndcWndent EuI-opc~ progr~ Group, ‘‘Towards a Stronger Europe. “ Vols. I and 11 (Belgium: NATO Headquarters, 1987).

Chapter 6-Exploiting Other Management Approaches ● 97

nel; these are often corporate-wide activities. In thecase of governments, such centers may serve mili-tary or civil interests or both at once. Some examplesare described below.

European Community

The EC funds four laboratories, known as JointResearch Centers (JRCs), at Ispra in Italy, Karlsruhein West Germany, Petten in the Netherlands, andGeel in Belgium. Whereas the JRCs were once theflagships of the EC’s collaborative research effort,their direction, objectivity and usefulness haverecently been criticized so extensively that the EC isplanning to revamp their management. Under pro-posals adopted by the Commission of the EuropeanCommunity in October 1987, the JRCs are to reducetheir dependence on the EC budget by 40 percent by1991. The proposals envisaged that 15 percent of theJRCs’ resources should come from contract researchfor individual governments and the private sector by1991, with a larger proportion also coming fromother Commission departments. While the plan doesnot call for cuts in the JRCs’ 690 million ECU(roughly $900 million) allocation for the subsequent5 years, the Commission proposed a sweepingreform of the JRCs’ objectives, mode of operationand methods of management.

The nations, however, were unwilling to acceptthe Commission’s proposals; West Germany calledfor more details on how the JRCs’ performancewould be monitored; the U.K. called for bettercontrol on areas where JRC work duplicates otherEC research; and West Germany, the U.K. and theNetherlands thought the 40 percent reduction independence on the EC R&D budget by 1991 did notgo far enough or fast enough. This debate suggeststhat the JRCs are suffering from the same malaiseand lack of relevance in their research that affectedmany U.S. corporate research laboratories in the1960s. In today’s environment, research must beresponsive to the marketplace or to military needs.

European Nations

On the national scene, the intense and publicdebate over the “State of Science” in the U.K.resulted in movement toward the centers of excel-lence concept. In late 1987, two steps were taken forreshaping British science, with emphasis on itsexploitation for commercial purposes. First came the

establishment of the Centre for the Exploitation ofScience and Technology (CEST), based at Man-chester University. Envisaged as a “think-tank”with a Steering Committee headed by the Cabinet’sChief Scientific Advisor, CEST’s role is to helpimprove Britain’s ability to exploit R&D, importedas well as home grown. CEST is to bridge the gapbetween industry and the scientific community; over80 percent of its funding will come from majorscience-based companies and the rest from theGovernment.

The second step was to create a number ofUniversity Research Centres (URCs). The URCs areexpected to have a vital role in the government’splan as ‘‘agents of change. ” Similar in concept tocenters of excellence established in U.S. universi-ties, these laboratories will be devoted to studyingspecific scientific opportunities that hold the prom-ise of being exploitable within a decade. TheNational Committee for Superconductivity choseCambridge University to host the first URC. TheBritish Government’s Chief Scientific Advisor isthought to believe that Britain must quickly estab-lish 30 to 40 URCs to bring about the changes heseeks in British science.

Industry

Industries on both sides of the Atlantic have beenfollowing this trend. In Europe, nearly all majorcompanies and universities have participated in theEC research initiatives or have joined “clubs”(consortia) striving to bring together a critical massof resources. In the United States, major companiesare also beginning to shed their go-it-alone attitudesand are seeking collaboration in key technologies,either with universities or with potential competi-tors. The newly formed SEMATECH, a DoD andindustry consortium created to develop microelec-tronic manufacturing technologies, is the mostrecent example. Earlier examples include the Elec-tric Power Research Institute (EPRI), the Semiconduc-tor Research Cooperative, the Council on ChemicalResearch, the University Steel Resources Center,and the Microelectronics and Computer Corporation(MCC). This trend toward banding together has beenencouraged by a more benign U.S. Governmentattitude toward the antitrust implications of jointventures in advanced technology and by the obvioussuccess of such ventures in Japan, and now in


Europe. While collaboration in the ‘‘pre-competitive phase” of research can result in a degreeof technical leveling and requires a long view ofmarket penetration, this trend holds promise as anaffordable means for U.S. industry to keep pace withinternational competition.

Department of Defense

In DoD, the Services, especially the Army withits University Research Initiatives Program, havebeen instrumental in encouraging the trend towarduniversity centers of excellence. Such centers canand should accelerate the state of the art in technol-ogy, However, time and the danger of technicalleveling could work against what should be a goodidea; clearly such centers need good management. Itwill be important for DoD to set priorities and fostercollaboration with U.S. industry, or perhaps withother governments.

European Defense R&D

Recent trends within major European govern-ments, described earlier, reinforce this argument.For France, Germany, and the U. K., defense re-search activities are being increasingly planned,organized, and managed by central authorities,independently of Service requirements and devel-opment activities. Research organizations are beingset up to serve as sources of technology that, whenmature, can feed into equipment-oriented organiza-tions. This is also becoming the case on a multi-national basis. The IEPG is considering forming aEuropean defense research agency. On the civil side,most EC research projects are directed toward acommon set of enabling technologies, with applica-tions left to industry to determine. In the U. K., evenmature research activities are being consolidated.Under a gradual rationalization policy, MoD’sDefence Research Establishments have been encour-aged to adopt a technology, rather than mission,orientation, 16

MANAGEMENT AND CONTROLJapanese Defense R&D

Which Focus: Technologies or Missions?


DoD’s laboratory structure is primarily mission-oriented, with most Service laboratories dedicated toa particular warfare specialty. While some conduct(or sponsor) generic research, the vast majority areconsidered a dedicated asset for accomplishing oneof the Service missions (e.g., Wright AeronauticalLaboratories, the Naval Ocean Systems Center, andthe Army Missile Command RD&E Center). Al-though a mission focus provides a closer linkbetween technology and military applications, italso encourages overlap and duplication throughoutDoD. At mature R&D stages, a mission orientationmay be appropriate; however, in technology baseprograms, where the key technologies are not yetcoupled to military applications, it can be arguedthat DoD should be organized along more generictechnology lines, with Department-wide prioritiesguiding individual research activities.

Japanese defense technology strategies are inter-twined with an extensive process of technologymanagement within the government and industrythat emphasizes dual-use technologies to assureJapan’s security in the broadest sense into the nextcentury. To understand the direction of defensetechnology management, one must look beyondnarrow definitions of defense and security. One mustexamine the roles and perceptions of a range ofbusiness and government interests in formulatingand implementing technology-management policiesas part of a larger economic strategy. As evidencedby the priority it places on developing dual-usetechnologies with multiple applications, Japan’stechnology policies are generated and implementedin a manner that merges economic, security, andindustrial policy considerations. As a result, govern-ment and industry consciously blur the line betweenpurely defense and civilian technologies to ensuremaximum use of emerging applications and proc-esses. They encourage a flexible approach to apply-ing commercial technology in military systems, with

Ib]f tie proposedconsolidation of ~fence Re~arch Establishments occurs, the U.K. new Dcfcnce Research Agency may find-as has U.S. industrywith its IR&D program--that research which does not directly satisfy a customer’s identified needs will be difficult to justify at the “bottom line. ”


the aim of making Japan equal or superior to othercountries in terms of its defense technology base.

In Japan, there is not necessarily a national orgovernment-wide consensus about the value ofdefense production and research for the overalleconomy. Although Japan has embarked on a policyemphasizing domestic weapons research and devel-opment, that policy is not universally embraced. TheMinistry of Finance argues that virtually any spend-ing on defense comes at the expense of the economy.This attitude is manifested in other ways. A numberof major research efforts within civilian ministriesand agencies have potential military applications(e.g., artificial intelligence, high-performance plas-tics, ceramics, advanced alloys, jet-engine research,and deep-sea mining systems). Although both thepublic and private sectors are examining possiblemilitary applications, the projects nevertheless arejustified primarily because of their expected benefi-cial impact on the civilian economy.

European Civilian R&D

In Europe, the bulk of publicly funded civilresearch is directed toward generic technologies,especially work conducted by universities and byprivate research organizations. To capitalize on thisinvestment in basic research and technology, gov-ernments are also setting up support (or technologytransfer) organizations that work closely with re-searchers and industrialists to move basic researchinto useful, marketable products.

West Germany presents an interesting case wheredecentralization, coupled with adherence to broadnational research goals, has been successful. Whilethe Federal and State Governments have long-standing policies that nurture civil research, theresearchers themselves are free to choose theirsubjects. This constitutionally guaranteed freedomof scientific research is the first of four basic pillarsgoverning research policy in the Federal Republic.The second can be seen as an outgrowth of WestGermany’s federal structure, where the 11 federalstates assume independent responsibility in the areasof education and science. The third pillar is thedeclared intention of the Federal and State Govern-ments to interfere as little as possible with theresearch systems. The fourth pillar is symbolized bythe intention that German research be integrated

closely and effectively into international—specifically European-research cooperation.

The German Research Society (DFG), an autono-mous organization within the scientific community,has great influence over German research programsand policies. Although DFG is funded by the Federaland State Governments [DM1 billion (approxi-mately $750 million annually)], it is not subject todirect governmental influence. It merely shares thegovernment’s goal to seek, realize, and expand upona high standard of achievement in basic research inWest Germany. The DFG’s independent staff ofexperts evaluates research-grant proposals submit-ted by researchers of all disciplines. If their decisionis affirmative, approval of the grant is almostautomatic. The Max Planck Society and the Fraun-hofer Society, both funded largely by the Federaland State Governments, are also independent establish-ments that exert great influence in formulatingresearch policies. The Max Planck Society adviseson what research projects are needed at any giventime, while the Fraunhofer Society serves as acatalyst for technology transfer between the scien-tific and business communities.

Intertwined with this, one finds German industryworking both with the basic research organizations—in search of commercial “nuggets’ ’-and with theFraunhofer Society to smooth the way for technol-ogy transfer. Thus, while the Federal and StateGovernments fund basic research with few‘‘strings” attached, an infrastructure exists toencourage a natural evolution from basic re-search into product-related (or mission-specific)research which is much closer to the market.

Industry

In the U.S. and European industrial sectors, itoften appears that internally funded research islargely conducted in operating companies. Closerinspection reveals that most of this so-called re-search falls into the categories of development orproduct improvement. The really basic researchcontinues to be conducted in central facilities thatconcentrate on specific areas of technology. Asnoted earlier, major corporations have establishedtheir own centers around key technologies, whichprovide a single technology source for the operatingcompanies’ use. These centers of excellence areoften staffed by a combination of permanent re-


search personnel and personnel assigned from oper-ating divisions—the latter tasked to become skilledin the state of the art and to bring that capability intothe division. Also, the trend toward banding togetherin research consortia further emphasizes a growingprivate-sector concern for technology-oriented, ra-ther than mission-oriented, research. Both movesrespond to the skyrocketing cost of research andtechnology.

It can be argued that multiple, mission-orientedlaboratories working along parallel lines will en-courage, or at least create opportunities for, greaterinnovation. It appears, however, that many compa-nies and governments have concluded that thebenefits of duplication are marginal and can even bedetrimental if sufficient funds are not available toconduct in-depth research at multiple locations. Bymerging technical and financial resources, managershope to gain the benefits of new ideas and innova-tion, while maintaining a central focus on selectedareas of technology.

Central v. Local Program Control

In Germany and Sweden, civil research is built ona foundation of independence; the view is thatindependence will encourage innovation, and inno-vation will result in progress. An informal, butinfluential, infrastructure has evolved to link re-search to the market, with researchers having onlyminimal technology -transfer responsibilities. This isfine as long as there is no financial bottom line forthe research establishments, or no military capabili-ties that are needed urgently. In such cases, onecould argue that independence may not be whollyappropriate.

In Japan, government laboratories and researchinstitutes fulfill a variety of roles in the R&Dprocess. It is important to note that they do not serveexclusively as creators of new technologies orinitiators of larger research projects. While theyoften serve these purposes, government facilitieshave an equally important role as neutral testinggrounds to verify results achieved in private-sectorlaboratories, and to carry research to a point whereit becomes more economical to pursue it in private-sector facilities. With such divisions of labor, it isnot surprising that much of the interaction betweenbusiness and government occurs among individual

researchers, their supervisors, and the directors ofresearch facilities.

In Japan’s private sector, engineers, researchers,and other technical specialists are heavily involvedin assigning priorities among potential civil researchprojects, and are active in the design and develop-ment phases of new products. Production andmanufacturing considerations are merged with de-velopment and design stages virtually from the firstconsideration of a promising technology all the waythrough the production phase. These considerationsare incorporated into product design, thus necessitat-ing fewer costly and time-consuming modificationslater. It is still difficult to determine if the same canbe said for defense production in Japan, but similarattitudes and practices probably prevail.

Among Western governments and industry, thereis a notable trend away from independent (i.e.,“project free”) defense research that reflects theneed to get near-term results from shrinking budgets,An example described earlier is West Germany’sconsolidation of its research and technology activi-ties under a single organization within MoD’sArmaments Division. Based on priorities set at theMinister level, MoD “Technology Coordinators”develop, organize, and direct the program. Theseofficials are expected to relate research priorities andresults to future operational requirements and areoften assigned temporarily to concept formulation(or pre-feasibility) study teams to ensure that re-search results will be used. MoD oversight ofresearch activities is maintained at all times.

U.S. and European defense companies are in-creasingly holding their researchers accountable forresults, especially the ultimate applications of theirwork to products or business objectives. As with theWest German MoD, many companies temporarilyassign researchers to project definition studies orlong-term product development activities, empha-sizing that they have a responsibility for the bottomline. The prevailing attitude in some companiesseems to be that ‘research is too important to be leftto researchers.*’ Centralized direction, review, andfeedback are the rule, While central control may tendto stifle innovation, it is becoming a financial fact oflife that the industrial research community mustface.


Balancing Intramural and ExtramuralResearch

U.S. and U.K. R&D

DoD’s extensive network of government-owned and government-operated R&D facilitiesis unique among defense establishments in theWestern world. Perhaps its closest counterpart isthe U. K., with its Research Establishments organ-ized originally to support specific mission areas(e.g., the Royal Aerospace Establishment and theAdmiralty Research Establishment). However, asnoted earlier, these activities have been graduallyconsolidated and rationalized to align them morewith areas of technology. Also, if the ResearchEstablishments are separated from MoD and oper-ated as an independent government research agency,MoD will own no research laboratories and will haveto contract out all of its basic and applied research.MoD would then mirror the West German MoD,which owns virtually no laboratory facilities.

French Defense R&D

In terms of defense research, France fallssomewhere between the United States and WestGermany. The Ministry of Defense owns andoperates, through the Delegation Generale pourl’Armement (DGA), three research laboratories (onein conjunction with West Germany). France is aspecial case; the head of DGA, who reports directlyto the Minister of Defense, has greater control overresearch, engineering, and industrial matters thanany other European-or American-defense offi-cial. In addition to the MoD laboratories, the FrenchGovernment operates development facilities, andowns and controls a large share of the defenseindustry. Under these circumstances, the distinctionbetween government and industrial research isblurred, but it seems that because of budget limita-tions little “project free” research is performed.

Japanese Defense R&DJapanese intramural defense research is directed

by the Technical Research and Development Insti-tute (TRDI). Organized as a division within theJapan Defense Agency (JDA), TRDI is the Agency’sprimary research organization. It is headed by acivilian who oversees three administrative depart-ments along with four uniformed directors, who

supervise research and development in ground,naval, and air systems, as well as precision guidedmunitions. Conceptualization, design, and prototyperesponsibilities occur at this level. Research centerscarry out projects, including surveys, research, test,and evaluation to enable further development onspecific systems. TRDI maintains five researchfacilities in Japan for testing and evaluating a broadrange of weapon systems and technologies. TheInstitute has no prototype manufacturing capabili-ties, relying instead on private-sector capacities.

The government established TRDI as an inde-pendent center for weapons development, as well asto stimulate the growth of the domestic armamentsindustry. It began with a philosophy of limitingdirect participation in defense-related R&D, partlyto minimize government budget outlays; but alsobecause the assumption was (and still is) thatdefense spending constituted a burden on the civil-ian sector. To a large degree TRDI has manageddefense technology according to its impact on thedomestic economic and technological base. TheInstitute does not necessarily target the developmentof technologies to field specific weapons systems; acriterion for selecting and nurturing technologieshas been the expected impact on the commercialsector. The chance that a given technology will betargeted for development by TRDI is greater if itcontributes to the overall industrial base and is likelyto provide commercial opportunities.

Reflecting normal practice in the Japanesecommercial sector, TRDI maintains close relationsbetween government and business. TRDI workswith industry both formally and informally, in manycases simply monitoring research already under wayat private companies. It also carries out preliminaryresearch that it hands over to the private sector, oncethe research reaches the stage where risks have beenreduced and the technology is proven. These pat-terns were reinforced by a reorganization in July1987 that totally eliminated minor research pro-grams that could be pursued more effectively byprivate-sector research facilities. In addition,TRDI’s role was defined to include research thatlacks an immediately identifiable demand in thecommercial sector. This could mask an importantchange in TRDI’s institutional role, and perhapsrepresents a JDA judgment that fielding advancedweapons systems will require selective development


of specialized technologies with primarily militaryapplications.

Civilian R&D

In European civil research, there appear to bemore laboratories that are owned, operated, andsupported by governments than in the United States.France maintains an extensive network of government-owned research facilities performing basic researchin areas such as atomic energy, space, automationand telecommunications. These laboratories, whichare staffed by researchers and administrative person-nel, exert a powerful influence over French researchpolicies. Criticisms have been voiced that these civilservants stifle innovation and serve their ownpurposes, rather than those of the country at large(arguments one often hears regarding U.S. Govern-ment labs). Germany, Sweden, and Italy operate afew laboratories, but rely mainly on the privatesector, especially universities, to provide mostresearch capabilities.

Industrial R&D

Despite the growing number of industrial re-search consortia and, in the EC, common-fundedresearch, most industrial R&D is still performed incompany-owned facilities by company researchers.This is understandable. Industry’s motivations are togain a competitive edge from research, makingcollaboration or contracting out risky. However,financial pressures now make research collaborationacceptable to more industrialists. Some companieshave taken a different course of action by closingtheir central research organizations and concentrat-ing solely on applied research. In these cases, thecompanies sometimes establish small advisory bod-ies to follow worldwide research and invest modestamounts in promising technologies, often withresearch institutes or universities. Such companieseffectively become “technological parasites, ” seek-ing to acquire technologies from any source to applythem to products and systems to obtain a near-termeffect on the bottom line.

Other companies in Europe have establishedhybrid programs in which some basic research isconducted in central research facilities, some inoperating companies (often funded by the centralresearch organization), and a small amount iscontracted out. Rotating personnel between the

research organizations and the operating companiesserves to sensitize researchers and engineers to theirresponsibility for technology transfer. In this way,the companies hope to keep their basic researchorganizations as lean as possible, and yet keep ideasflowing into new products.

Through programs such as ESPRIT, BRITE, andRACE, the EC hopes to establish Europe as a majorcenter for advanced technology, and to use thistechnology to establish European industries asleaders in world markets. Except for the JointResearch Centers, the EC itself has no researchorganizations and relies solely on external resources,e.g., industry, academe, and private research organi-zations. Experts are retained temporarily to helpestablish priorities, set research program goals,prepare bid packages, evaluate proposals, and re-view programs. This raises the inevitable questionregarding the competence of EC staff to makeinformed judgments on advanced technology and toassess which technologies are ready for application.It will be interesting to follow the progress of the ECto see if an organization with such limited internalresearch capabilities can accomplish its ambitiousgoals.

Whether or not the EC succeeds, it appears thatthe overall trend in Europe is toward fewernationally owned research facilities—especiallyon the part of ministries of defense. To retaintechnological relevance in this environment, MODS

will have to coordinate closely with national univer-sities and independent laboratories expert in specificareas of technology. It will also require a muchcloser coupling between civil research and defenseneeds if MODS are to maintain a state-of-the-artmilitary force. The research must be conductedsomewhere and, if it is not done in government-owned defense research establishments, then othereffective mechanisms will have to be found.

GETTING RESULTS—THE“BOTTOM LINE” OF RESEARCH

Applications—Moving Technology From theLaboratory to the Marketplace

For DoD as well as for industry, the payoff fromresearch is its application, whether to a next-generation weapon system for DoD or a successful


new product for industry. If the technology isapplied too hastily, manufacturing can become anightmare and expensive redesign is often needed.Many defense acquisition horror stories are theresult of attempts at concurrency-entering initialproduction stages before a design (or technology) isfully developed and thoroughly tested. But, para-doxically, the Department is often sharply criticizedfor leaving technology in the laboratory too long,thus basing new weapons and systems on yester-day’s technology. U.S. industrialists are also undergrowing pressure because of their apparent inabilityto offer products that capitalize on new technologyin a timely fashion and at competitive prices.European governments and industries appear tobe no better than their American counterparts ateffectively transitioning new technology into weap-ons systems or products. The European problemmay be more the result of a lack of sufficientinvestment—a shortfall the EC is attempting tocorrect through collective investment and “trans-national” research priorities. In the United States, itmay be that too many technological options havebeen available, and that the continued promise of the‘‘next breakthrough” frequently has paralyzed DoD’sdecision process.


Perhaps the toughest problem faced by DoD’sresearch managers is technology transfer-how toinsert research results successfully into weapons andsystems without excessive cost and before thetechnology becomes obsolete. Acquisition programsare essentially risk reduction programs involving asequence of research, development, design, andengineering tasks. New technologies must passthrough a number of phases during which they willbe viewed differently depending on their state ofdevelopment and the skills (and biases) of thepersonnel involved. What might be obvious advan-tages or shortcomings to a researcher might not beappreciated by the development engineer or de-signer. As a result, the technology might be usedimproperly or have too much expected of it, so thatthe insertion effort is deemed a failure. This dilemmaneeds to be addressed. One ambitious DoD ‘ ‘inser-tion” effort is the current very high speed integratedcircuit (VHSIC) program, where existing avionicsand system design programs are converting toVHSIC technology. The VHSIC program brings

significant performance enhancements, but also ashare of start-up problems. None of the Europeangovernments has attempted, or planned, an effort ofthis magnitude.

Japan

At present, Japan seems to be unique in itsindustry-wide ability to move advanced technologyrapidly and effectively from the laboratory to themarket. The current trends in Europe should beexamined in light of Japanese successes, since theEuropeans appear to be trying to apply Japaneseconcepts. What appears to work best is the establish-ment of teams of researchers, engineers, designers,manufacturing specialists, and even marketeers,early in the life of a technology or product. Thisgroup is responsible for ensuring the efficientmovement of the technology through to manufactur-ing. These concepts appear to be under considerationin Europe for EUREKA and for some multinationalprograms sponsored by the EC, such as RACE andBRITE. Within individual European MoDs, thescope of national research may not lend itself to thislife-cycle approach to technology transfer, but ifresearch collaboration grows under the auspices ofthe IEPG, formal technology insertion programsmay soon follow.

European Industry

Private industry in Europe is also struggling withthe transition problem, as several large Europeancompanies are experimenting with new methods ofmanaging R&D. The goal is to concentrate on themost commercially promising areas and to ensure afaster transfer of research results to the market. Inpursuing this goal, R&D responsibilities are becom-ing more closely tied to the marketing and operatingdivisions— a practice that has become the rule inU.S. industry. Scientists, especially the most seniorresearch people, are expected to support the com-pany’s business goals. They attend planning meet-ings and are considered part of the business team,along with the marketing and production personnel.

Examples of the close relationship that isessential between the research staff and those whodevelop technical specifications exist in all success-ful companies. However, in large and complexorganizations the necessary interaction and commu-nication can be jeopardized by interdepartmental


rivalries and parochialism-problems that onlystrong management and a clear set of objectives candispel. Unfortunately, examples of organizationalenvironments conducive to effective technologytransfer are few and far between, with ad hocmeasures often substituting for strong managementand sound policy.

A few specific approaches for assisting the flowof technology can be mentioned. Some companiesrecognize the need to retain continuity of technicalexpertise as a research project moves from researchthrough development into production. One majorEuropean firm, with both defense and commercialoperations, has developed a‘ ‘distributed” technology transfer system involving four types of laboratories:

●

●

●

●

A basic research laboratory that concentrates onlong-term topics (>10 years), working withuniversities and tackling problems of funda-mental interest.Central laboratories that serve major companygroupings through research on areas of com-mon interest and on mission- or business-oriented concepts or systems (5 to 10 yearsbefore product introduction).Site laboratories that work on new products,funded by individual product divisions and bycontracts with outside organizations (2 to 5years before product introduction).Product development within product divisions,working on next-generation products, productimprovement programs, etc. (1 to 2 yearsbefore product introduction).

In this company, technology transfer is effectedthrough a “push-pull” process, with the technologymoving from central to site laboratories, and fromsite laboratories to divisions, in a process thatinvolves the temporary assignment of scientists andengineers to a project for 2 or more years.

In another European example, a corporate re-search center serving the entire company receives itsfunding from a variety of sources (i.e., from centralheadquarters, from product divisions, and fromexternal contracts). A senior scientist within theresearch center monitors all research programs,relating each to possible and actual division inter-ests. He also reviews programs for combinationalpossibilities, commercial leverage, etc., and ar-ranges joint technology demonstrations for the

business areas (divisions) concerned. The center hasalso set aside a budget to fund the business areas thatwill apply new technology, and routinely assignscenter scientists temporarily to the business areas toeffect technology transfer. While this techniqueworks in many cases, in others it may represent toostrong a “technology push,” and can encounterresistance at the division level.

Some defense firms have no central researchorganization at all, with the divisions being solelyresponsible for internal R&D. While the researchfocus of the division is inevitably more near term tomatch the needs of their customers, these companiesdo recognize the importance of acquiring newtechnologies. In one example, an off-line technologygroup in the central headquarters maintains a‘ ‘tech-nology watch” and advises the product groups (ordivisions) when key technologies approach theapplications stage. The team studies research col-laboration possibilities, monitors the introduction ofthe technology into the division’s product line,carries out marketing surveys, etc. The divisions willcoordinate the applied R&D on their own behalf,while the central team then moves on to its nextproblem.

Industry is employing a variety of approaches toencourage the efficient and timely transfer of tech-nology—approaches ranging from secondment ortemporary assignment programs to business devel-opment teams, to formal programs of “technologi-cal parasitism. ” The one common thread is thatsomeone who has both the responsibility and theauthority to make technology transfer work has beenput in charge of the process.

Collaboration in Research and Technology

International collaboration in research is becom-ing a way of life for most Western nations. Not onlyhas the cost of research become prohibitive forindividual organizations, but worldwide competi-tive pressures in defense and civil markets areforcing companies and governments to pool theirresources simply to stay in the game. These factorshave triggered dramatic changes in the operatingmethods of high-technology organizations, includingDoD, During the 1970s and early 1980s, Europeangovernments and companies and, to a lesser degree,their U.S. counterparts began to explore ways to

Chapter Exploiting Other Management Approaches ● 105

cooperate in defense R&D and production programs.This effort spawned a number of European coopera-tive development projects, such as TRIGAT, Tor-nado and Airbus, and transatlantic programs, such asthe AV-8B (Harrier), the Multiple Launch RocketSystem, and the Family of Air-to-Air Weapons.However, little cooperation was achieved in intra-European or transatlantic defense research, eventhough cooperation was growing in civil and spaceresearch activities. In the mid-1980s, this situationbegan to change. Concerns about European competi-tiveness in advanced technology triggered initiativesby governments to promote cooperation in bothdefense R&D and civil research. The Europeanmembers of NATO (the IEPG countries) now acceptthe necessity of giving up a degree of sovereignty tomake more effective use of the $8 billion that theyspend each year on military development, and toreceive through collaboration a better value forexpenditures on major procurement programs. TheEuropean members of NATO, with and without U.S.participation, have been collaborating on defenseprojects on an ad hoc basis for over 20 years, but themood now is to establish a more cohesive, system-atic program of collaboration in all phases ofacquisition, including research.

Steps are also being taken to deregulate the entireEuropean NATO defense industry. The 1987 IEPGStudy Team issued its blueprint17 for a commonarmaments market, or a military EC, which itbelieved could be achieved by “giving greater playto competitive market forces. ” By opening Upfragmented and highly protected national markets,and developing a pan-European competitive andcollaborative environment, the IEPG Study Teamsaid that European NATO members should be ableto reduce the costs of designing and building modemweaponry. The argument went that this would yielda more coherent European defense industry, able tocompete and collaborate on more equal termswith a U.S. defense industry that was twice itssize. IEPG ministers endorsed the report and di-rected their staffs to begin implementing many of therecommendations. While many hurdles remain, thederegulation of Europe’s defense industry appearsincreasingly likely.

It is an open question whether DoD’s researchand technology community is ready to cooperatefully with its European counterparts. Cooperativedevelopment, with the 1985 Nunn Amendment as acatalyst, has gained favor with the Services. Impor-tant programs are now under way. For the technol-ogy base program, however, a “go-it-alone” atti-tude seems to prevail, with Data Exchange Agree-ments dominating government-to-government inter-actions and industrial cooperation discouraged bythe exclusion of foreign firms from many explora-tory and advanced technology development (6.2 and6.3A) programs. The Nunn Amendment succeededbecause it gave the Services a financial incentive tocooperate. Some type of Nunn appropriation mightbe needed to encourage similar collaboration indefense research and technology.

In civil research, the heightened sense of concernin Europe for its technological future is attributableto several factors, especially the scale of modemtechnology and recognition of the severe structuralobstacles to Europe’s international competitiveness.Breaking down the long-standing barriers that haveisolated European companies from each other is anexplicit objective of the Single European Act and theplanned 1992 economic integration. The collabora-tive high-technology initiatives now being pursuedare an important element of this strategy. Europeanindustry also sees other reasons for cooperation inresearch, As technologies converge, companies thatonce were specialists in a single activity now needto draw on a broad spectrum of sciences andtechnologies. Also, with shrinking product lifecycles, there is a need for more frequent introduc-tions of new ideas, thereby increasing the costs andrisks of research. Companies can no longer afford torisk a generation gap in their products as the resultof a research failure. U.S. industrialists also facethese problems.

Other cooperative efforts have grown on anational level. The U.K. Alvey program was di-rected toward developing a capability in informationprocessing that would help British industry keeppace with some aspects of Japanese and Americandevelopments. In Germany, and to a lesser extent inSweden, a collegial relationship has developed overseveral decades among government, industry, and

IT1ndepCndcnt ELUOpC~ Probvw Group, op. cit., foomote 15.


academe in civil research areas. Germany’s strongposition in world trade is partly a result of theserelationships which, without direct governmentintervention, have produced an extraordinary net-work for exchanging information in research andencouraging product development based on thelatest research results. The German system, in whicha key role is played by the independent institutes(e.g., the Max Planck Society for basic research andthe Fraunhofer Society for applied research), maybethe best example outside of Japan of collaborationbetween the public and private sectors. Some arguethat the German system is even better than Japan’sfor stimulating technology-base activities.

Much has been said in this chapter regarding theprogress of Europe’s industries toward cross-bordercooperation. It is certainly true that the nationalindustries have supported their government’s col-laborative civil research within the EC. They havealso accepted the premise that armaments coopera-tion (under the IEPG) has become an economicnecessity. In defense research, however, there hasbeen little industry-to-industry cooperation. This is,in part, a reflection of the meager funding availablefrom Europe’s MODS for basic research-withgovernment research funds being increasingly fun-nelled into cooperative civil projects-and, in part,a reflection of the highly competitive nature of worlddefense markets. While European defense compa-nies are willing to cooperate in development pro-grams (or at least will cooperate when their govern-ments tell them to), their internally funded researchprojects, which are mainly applied research orproduct development, are usually well hidden frompublic view.

In Japan, selective cooperative research, par-ticularly in the pre-competitive phase, plays animportant role in achieving technological gains inthe public and private sectors. Collaborative under-takings, though widespread, are not necessarily therule in Japanese research efforts, and multinationalresearch collaboration is still relatively rare. Thenature, timing, and participants of united efforts varyfrom one field to the next. Informal and formalstructures and processes tend to identify promisingresearch fields or trends. Once a consensus has beenreached between government and industry on spe-cific avenues of research, a joint government andindustry effort or a government-sanctioned research

consortium (involving multiple private-sector inter-ests) is established. As research proceeds, greatercompetition is introduced to hasten the introductionof products to the market.

However, Japanese companies are apparentlyless committed today to the consortium approachthan they might have been in earlier decades. Manyargue that important resources are being divertedfrom corporations to government-sanctioned con-sortium efforts without a demonstration of sufficientpotential for tangible short- or long-term gains.Some firms have suggested that their own resourcesand decisionmaking processes are sufficient forstimulating technological advances. And, while notresenting the government role, these firms believethat it should be reduced or shifted to other forms ofinvolvement in R&D. These same companies, how-ever, continue to participate in deference to main-taining government relations—and out of a com-petitive concern that breakthroughs may be achievedby a consortium to which they would not be a party.

Despite Japanese industry’s broadening disaffec-tion with the status quo, this situation is not likely tochange in the near future. In defense research andtechnology, for example, there are a large number ofindustry consortia, including those in compositematerials, advanced turboprop research, and fighteraircraft. Certain projects, such as the Fighter SupportExperimental (FSX), are seen literally as once-in-a-lifetime opportunities that, if neglected, could leadto the complete loss of important capabilities. Costis another factor favoring cooperation, especially inlarge-scale projects originating in, but not necessar-ily limited to, the defense field.

Many of these same considerations affect U.S.industry as well. Although U.S. companies haverecently begun to band together in the pre-competitive phases of selected technologies, theyare doing so largely because of their fear of foreigncompetitors capturing domestic markets in whichthe company has a stake-not necessarily to boostthe Nation’s overall defense preparedness or com-petitiveness in world markets. Because of financialand competitive pressures, and because “1992” ismaking it a political necessity, one now observes asteady stream of U.S. industrialists traveling toEurope and the Far East seeking to strike deals.While most of these deals are focused on codevelop-


ment or coproduction, American industrialists are U.S. companies (and with foreign companies) maycoming to realize that Europe’s technology is become increasingly attractive if the U.S. Govern-first-rate, and in many areas is on a par with our own. ment provides some incentives. A two-way street inJapan’s technological excellence is, of course, no technology development might then become asurprise to U.S. executives, and the reasons for it arebecoming well known. Research collaboration among

reality for commercial, not just military, purposes.

Chapter 7

Implications for theDefense Technology Base:


CONTENTSPage

INTRODUCTION AND SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .POLICY OPTIONS FOR CONGRESS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

High-Level Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Structure of the Laboratory System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Laboratory Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Funding and Budgeting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

111113113115119122125

Chapter 7

Implications for the Defense Technology Base:Options for Congress

INTRODUCTION AND SUMMARY

Chapters 4,5, and 6 of this report are an integralpackage. Chapter 4 addressed broad managementissues facing the U.S. Department of Defense (DoD)as it seeks to plan, execute and review the defenseScience and Technology (S&T) program. It high-lighted the degree of control-or lack of control——exerted by the research and technology staff of theOffice of the Secretary of Defense (OSD) over thepriorities and content of the S&T program.

Chapter 5 examined how the S&T programoperates at the ‘‘grassroots level’ ’—the DoD labora-tory. This massive and unwieldy structure presentssignificant management and organizational prob-lems that DoD must solve if the defense researchprogram is to become more relevant and productive.Chapter 6 surveyed how other sectors and othernations have tackled the problem of planning andexecuting S&T programs. The objective was todetermine what, if anything, the Defense Depart-ment might adopt from less complex environments.The implications can be summarized as follows.

DoD’s S&T program is basically a bottom-upprocess, with OSD serving largely in the role ofmonitor. While OSD’s research and technology staffoccasionally exerts pressure on the Services regard-ing specific issues, they generally yield to Serviceresearch and development (R&D) personnel on thecontent and direction of research programs. OSD’smain role is effectively to collect the inputs from theServices, correlate them, and defend them beforeCongress for review. With its direct access to theServices, Congress often revises elements of aService S&T program-with or without the agree-ment of the Service, and often without consultingOSD. Once the money is approved, the Servicesexecute their own programs. The other two majorS&T activities are the Defense Advanced ResearchProjects Agency (DARPA) and the Strategic De-fense Initiative Organization (SDIO) (SDI beingfunded under Advanced Technology Development,budget category 6.3A). SDIO is even more inde-pendent than the Services, with the Director report-

ing only to the Secretary of Defense. DARPA reportsto the Director, Defense Research and Engineering(DDR&E), but enjoys a degree of autonomy, onprogram content and priorities, from OSD’s S&Tmonitors.

Different elements of the DoD S&T program arethus managed through three different hierarchies: forthe Services, the laboratories report at a relativelylow level and programmatic decisions work theirway up the chain of command; DARPA reports toDDR&E, three levels removed from the Secretary;and SDIO reports to the very top. Under thesecircumstances, the OSD research and technologystaff can do little but monitor and collect data.

This management structure also encouragesduplication and a degree of inefficiency, especiallyconsidering the vast network of laboratories, centers,and other facilities responsible for research programactivities. Nearly every other organization examinedreflects far more centralized planning and executionof its S&T program. This is especially true inmilitary R&D where a central authority, oftenreporting to the Minister of Defense (or in the caseof industry, to the company president), has both theresponsibility and authority to set research prioritiesand to ensure that the program content meets theorganization’s goals. With goals set from the top,there are means to exert pressure on the performingbodies to make sure that the programs are respon-sive. Further, at least in most governments, militaryresearch programs are generally stable over a periodof several years; thus, researchers in these programsare not faced with annual—and disruptive-changesin funding or priorities. It appears that DoD isfollowing a minority path in its conduct of S&Tprograms, one that is declining in popularity amonggovernments and major companies.

Another unusual feature of DoD’s S&T programis its extensive system of government-owned andgovernment-operated laboratories. No other non-communist nation, let alone private enterprise,operates so many facilities and maintains such alarge research staff. Depending on one’s point ofview, the DoD laboratory system is either a tremen-

–111–


dous asset or a tremendous burden. Whether it is anasset or a burden, the structure is in place today andthe Department needs to address some of the seriousmanagement and organizational problems whichhave beset the laboratory system.

One problem is the nature of most of thelaboratories themselves. In fact, they are not labora-tories; they simply lack the “critical mass” ofmultidisciplinary talent necessary to develop state-of-the-art technologies consistently across a spectrumof areas. Most of these organizations were created,and continue to exist, to support the mission of oneof the military Services. Consequently, developingtechnology per se is not their primary objective.Rather, research and technology capabilities arenecessary to assist a Service in performing itsmissions. Mission requirements can be satisfiedeither by conducting internal research or by contract-ing out to major corporations, universities, or privateinstitutions. Each Service addresses mission supportquite differently. In the Navy, a large share of theS&T budget is spent in-house. In the Air Force, thelaboratories place more emphasis on becoming‘‘smart buyers,’* and contract out the bulk of theirS&T work. The Army falls somewhere in between.With different mission orientations leading to differ-ent approaches to R&D, overlap and other ineffi-ciencies arise throughout the system.

A second major problem with the DoD laborato-ries pertains to their ability to hire and retainqualified researchers. Not only are there significantsalary deficiencies, but the politicized environmentin which research is conducted in many laboratoriesoften discourages qualified scientists and engineers,sending them to higher paying and more rewardingjobs in corporate research laboratories or at universi-ties. This situation might be mitigated if there werea DoD-wide policy to contract out as much researchas possible, essentially letting the research followthe scientist. But policies regarding contracting outare inconsistent, and laboratories are sometimesforced to conduct in-house research with inadequatestaff.

Finally, the government-owned and government-operated laboratories are saddled with proceduresthat often make them less efficient than the betterindustrial and university laboratories. This is anissue that must be addressed because research

budgets are likely to decrease as part of an overallreduction of the defense budget. Other comparableorganizations structure their laboratories differently,and some—with management problems similar tothose of DOD--are currently involved in basicrestructuring. One approach involves aligning re-search according to technology areas, creatingcenters of excellence that assemble sufficient re-sources to make a difference in high-technologyfields. These centers are overseen by top manage-ment and serve as corporate-wide resources. Whilesome observers believe that this approach stiflesinnovation, others argue that it increases efficiencybecause progress is more likely once the organiza-tion decides on the line of research it will pursue.

This assessment raises other issues. Nearly everyAllied government is concerned for the future of itsnational research and technology programs. Politi-cians and the civil service agree that a country’smilitary and economic security depend on thenation’s ability to produce affordable state-of-the-artproducts, including weapons. Most nations alsoconcede their inability to conduct independentresearch programs that are sufficiently deep in morethan a few technology areas to achieve technologicalbreakthroughs sustaining industrial competitivenessin world markets, or deterring aggressors. Whatmoney is available must be wisely spent. To ensurethat this happens, European governments have setpolicies and priorities for research and technology atthe highest governmental level-often at the level ofthe Prime Minister or President.

A related issue is the recognition by Europeangovernments of the “dual-use” nature of advancedtechnology. On one hand, government officials arepainfully aware of the success of the “JapaneseModel” in transferring the results of science andtechnology programs into quality products, therebygiving Japanese companies a competitive advantagein world markets. On the other, they see continueddifficulties in exploiting technology developed underU.S./European cooperative military programs, espe-cially from the standpoint of technology transfer tocivil or to third-party military markets.

There is an international trend toward decreasingthe emphasis on military technology and increasingemphasis on research for enhancing national indus-trial competitiveness. For example, in Europe,

Chapter 7-Implications for the Defense Technology Base: Options for Congress ● 113

military research budgets are declining, while in-vestment throughout the European Community (EC)is growing substantially. Many technologies the ECis funding have clear dual-use applications; govern-ments seem to be expecting that the results of civilresearch will flow into military development andproduction programs. In Japan, while military re-search budgets are growing modestly, the govern-ment is maintaining a close connection between thisresearch and commercial applications.

Additionally, collaboration in research is becom-ing important. Because most nations and privateorganizations find the costs of sponsoring a “world-class” research program prohibitive, they haveconcluded that for all its problems, banding togetheris the only way to go. With the emphasis on civilresearch and the trend toward collaboration, DoDmay find increasingly that it is the “odd man out, ”to the possible detriment of the competitiveness ofAmerica’s high-tech industries. Foreign technologyin the civil fields is approaching, and in some casesexceeding, the quality of DoD’s military technol-ogy. As this trend continues and the line betweencivil and defense research gradually disappears, itmay be necessary to revise policies in such areas asinternational collaboration, technology transfer, for-eign disclosure, and export administration.

POLICY OPTIONS FORCONGRESS

The following policy options are based on thesefindings. While they are by no means exhaustive,they do address several issues that strongly affect the‘‘health” of the U.S. defense technology base. Theseoptions take into account the interdependence ofDoD’s research planning and execution with eventstaking place throughout the S&T community, bothforeign and domestic. In its consideration of theseoptions, Congress should bear in mind that these areextremely complex matters, and that no consensusexists among experts on any of the issues or optionsthat are presented.

The options are grouped under five broad, andunavoidably overlapping categories: 1 ) high-levelplanning, 2) organization, 3) structure of the labora-tory system, 4) laboratory management, and 5)funding and budgeting.

High-Level Planning

Option 1: Establish government-wide priori-ties for defense research and ensure they arefollowed by all DoD components and the privatesector.

This option addresses the need for the FederalGovernment to execute an increasingly complex andexpensive defense research program with con-strained (or decreasing) funding. The prioritiescould be developed in conjunction with DoD’sProgram and Planning Budgeting System (PPBS)process, related specifically to the early planningactivities of the Joint Chiefs of Staff (JCS) andreflected in the Defense Guidance (DG). The privatesector (industry and academia) could be consultedduring periodic technology-assessment exercises(e.g., the planning phase of the PPBS), kept fullyinformed on progress and new directions (e.g.,breakthroughs and political developments), andencouraged to invest in complementary research.

The methods introduced by the European Com-munity for its joint research projects could providesome useful insights. The EC Commission setsresearch priorities. A 5-year budget is adoptedproviding for roughly 50 percent of the needed fundsand industry/academia consortia are invited tosubmit bids. Industry funding of half the work anduniversity involvement are necessary conditions f-ora bid to be considered responsive. Some factors thathave contributed to success include: 1) priorities setat high political levels; 2) long-term (5-year) ECfunding commitment; and 3) research work at the‘‘precompetitive” stage, with applications left up toindustry. Additionally, there is no alternative to jointresearch projects, because individual governmentfunding sources are diminishing rapidly.

This model may not fit DoD’s situation exactly,but it can be made to fit. The key technologiesrequired for defense systems are largely known; andrecent technology assessments, whether made byDoD or industry or both, have much in common. TheDefense Science Board (DSB) could provide theconnection with the private sector for periodictechnology assessments. The necessary planningprocedures are largely in place, i.e., the PPBS, theJoint Strategic Planning Document (JSPD) and theDG. Finally, the independent Research and Devel-


opment (IR&D) program is one source of (partial)industry funding. To implement this, a top-levelcommitment would be needed, with the Administra-tion and Congress providing the political andbudgetary incentives.

Option 2: Reestablish OSD'S corporate over-sight authority for DoD% technology base pro-grams.

It appears that Congress has provided OSD withthe necessary statutory authority to exert strongcentralized guidance over DoD’s technology baseprograms. But as chapter 4 described, OSD—specifically the Director of Defense Research andEngineering (DDR&E)—appears to have relinquishedto the Services much of the responsibility foroverseeing the technology base. This runs counter totrends elsewhere; as chapter 6 pointed out, ourNATO Allies have initiated or further strengthenedthe centralized management of their defense R&D.Congress could insist that OSD, exercising itsstatutory authority, reestablish its corporate over-sight authority for DoD’s technology base programs.

OSD representatives have complained of ever-increasing congressional micromanagement of theirS&T programs. Pentagon officials indicate that theybelieve Congress is overstepping its responsibilitiesin dictating how OSD should structure certaintechnology base programs. Some of the OSDobservations may be valid; nevertheless, it is acongressional perception that OSD is not suffi-ciently exercising its oversight responsibilities thathas led Congress into a deeper involvement inshaping DoD’s technology base strategy. If OSDwere to assert its authority and develop a strategicR&D plan, Congress would probably find it lessnecessary to involve itself in individual programs.

In various discussions with OTA, some Servicerepresentatives expressed their surprise and frustra-tion that OSD had not exerted stronger managementcontrol over DoD’s S&T programs. They contendedthat stronger and more effective OSD technologybase oversight could go far to reduce inter-Servicerivalry and produce a more coherent technology baseinvestment strategy. However, other Service repre-sentatives argued that it would not be appropriate forOSD to exercise greater authority. In their view,because the Services ultimately would use theproducts of R&D programs, they alone know what

is needed. Moreover, they felt that OSD personnelwere too far removed from technology base pro-grams to understand them well enough or to careenough about them to defend budgets successfully.Some argued that if Service control over technologybase programs were to be reduced, the Serviceswould lose interest in—and decrease funding for—the technology base.

Option 3: Institute a strategic planning proc-ess within DoD that will lead to a coordinatedDepartment-wide technology base investmentstrategy.

Currently, the Services dominate planning forDoD’s technology base program. If it wanted toprovide more centralized control, Congress couldconsider requiring the Secretary of Defense to begina DoD-wide technology base strategic planningprocess directed by the Undersecretary for Acquisi-tion [USD(A)]. Without the endorsement of theSecretary of Defense and USD(A), it will not bepossible to implement a strategic plan successfully.

Such an investment strategy could: 1) facilitateOSD-directed strategic decisions, 2) diminish thelikelihood of technological surprises, 3) reduceduplication of effort, 4) encourage short- and long-term technology base planning, 5) enable OSD andthe Services to examine the outputs of the S&Tprograms and not just the inputs, 6) enhance theunderstanding within DoD of the importance oftechnology base programs within DoD, and 7)provide Congress with a more coherent defense S&Tstrategy.

However, in the view of some, centralized controlwould only add another layer of bureaucracy be-tween the invention of new military technologiesand the managers in the Services who will ultimatelyhave to acquire the technology for weapon systems.From this perspective, OSD staff would be too farremoved from the technology base to understand theneeds of the user, and protect the funding forcritically important technology base programs.

USD(A) could direct DDR&E to initiate astrategic planning process that would involve theparticipation of the three Services, the defenseagencies (including DARPA), and SDIO representa-tives from the JCS, the unified and specifiedcommands, and the intelligence community. One

Chapter 7-lmplications for the Defense Technology Base: Options for Congress . 11.5

official within the Office of the DDR&E (ODDR&E)would be responsible for developing, implementing,and directing a DoD-wide technology-base strategicplan. Developing a strategic technology base plan-ning process would require the full-time attention ofthe OSD official responsible for its success. ’

Any strategic technology base plan will have tobe tied closely to DoD’s national security objectives.The principal aim of the strategic plan should be toestablish near- and long-term S&T objectives, lead-ing to the achievement of the Department’s opera-tional objectives.

Once the DoD-wide strategic technology baseplan is completed and approved by DDR&E andUSD(A), the Services, DARPA, and the otherdefense agencies can use it to develop their owntechnology base programs. In turn, OSD can use theplan to evaluate the Services’ and DARPA’s tech-nology base investment strategies and determine theextent to which each technology base programmatches the plan.2

Finally, in view of the high turnover rate of OSDpolitical appointees, an accepted strategic planningprocess should help new appointees to draft acoherent technology base investment strategy. Thisexisting process should reduce the need for each newtop-level civilian manager to “reinvent the wheel. ”

Organization

Option 4: Establish a central coordinatingactivity within the Administration to ensure thatdual-use technology is exploited in the bestinterests of the nation as a whole.

A serious imbalance is emerging between theUnited States and its allies with respect to dual-usetechnology. Japan spends very little on defenseresearch and technology programs; there, civilresearch is the main focus. This approach helps toaccount for Japan’s enviable record of success inworld trade. In Europe, the EC has committed $6.8billion over the next 5 years to civil research, much

of which also has potential military applications.The European national governments are supportingthe EC and other joint civil research projects andappear to be pulling out of defense research. TheU. K., for example, has limited its defense researchbudget so as to prevent “crowding out” of civilresearch—the government openly encourages do-mestic industry to bring the results of research to theMinistry of Defense (MoD) “when it’s ready.”

In such fields as high-temperature superconduc-tivity, high-definition television, microelectronics,fiber optics, supercomputers, and telecommunica-tions, the United States is competing with countries(or blocs) with whom we are allied in the East-Westpolitical competition. The United States, with itsfocus on the Soviet threat, has placed its industriesat a potential disadvantage in world markets throughmeasures such as restrictive export and technology-transfer policies. Operating under less stringentrestrictions, our allies see their industries enjoyingsignificant sales growth in market areas previouslydominated by U.S. companies.

The Administration must take these trends intoaccount as it considers the future health of America’sindustrial base (recognizing that the defense indus-trial and technology bases are only two elements ofthe Nation’s industrial base). A balance should befound between the need to protect defense technolo-gies (largely at the applications stages) and thegrowing need for industry to exploit the same (orsimilar) technologies in U.S. and foreign markets.Accomplishing this tricky balancing act will requirethe full support of the President, Congress, andindustry. It will also require the appointment of aresponsible and independent official with statureand authority.

This official could be located in DoD, but if so,interagency coordination (e.g., with State, Com-merce, NSF, and NASA) should have high priority.Alternatively, the Office of Science and TechnologyPolicy (OSTP) could perform this function. Inconjunction with government agencies, industry and

I The President’s 1983 private Sector Survey on Cost Control, Task Force on Research and Development, indicated that it could t~e S to 5 yems toimplcmenl a strategic planning process for DoD’s R&D programs. Consequently, top DoD management will have to be persistent in its support if strategicplanning is to be implemented.

2A recent Instltu[c for ~.fenw Analy~s (DA) task force rccomnlend~d 17 [e~hnology p~els (O improve the coordination of DoD’s S&T prO~~S.

OSD should bc able to determine the extent to which each panel ac[ivitics would bc linked to tic strategic plan; if some S&T projects have poor linkage,they could be redirected or canceled. See ‘‘Report of the Task Force for Improved coordination of DoD Scicncc and Tcchno@~ Programs” (Alexandria,VA: Institute for Defense Analyses, July 1988).


academe, broad national technology goals couldevolve, with a “crosswalk” described betweentechnology investment plans and various applica-tions, both military and civil. Joint research projectsbetween agencies, with industry and/or academe, orwith other nations could be encouraged. Applica-tions would be left to individual agencies; in the caseof DoD this would normally be when mission needsare matched with technologies, i.e., at the advanceddevelopment (6.3A) stage. Most projects could beunclassified, with the results flowing into civil, aswell as military, applications.

Option 5: Develop a coordinate, Administration-wide program of collaboration and cooperationin defense research and technology.

This option includes two kinds of action. First,better inter-Service research cooperation would bepromoted, with some consolidation of responsibili-ties and lines of authority to improve communica-tion and minimize duplication. The goal should be tobring to bear on key technologies sufficient re-sources to “make a difference. ” Second, a dedicatedbudget might be established for cooperation inresearch and advanced technology, involving jointresearch projects, technology demonstrations, andperiodic high-level reviews to assess opportunitiesfor cooperation, monitor progress, and set priorities.

Research collaboration within DoD (i.e., all threeServices, DARPA, and SDIO) is widespread today,although much is ad hoc and conducted at aresearcher-to-researcher level. This approach shouldbe retained, but augmented with senior-level coordi-nation on priorities, the assignment of lead organiza-tions for key technologies, and a secondmentprogram through which special skills are assembledinto multidisciplinary research teams. Over time,this combination of approaches could encourage a‘‘natural rationalization” of DoD’s laboratory struc-ture. The U.K. Research Establishments haveevolved this way in the face of serious budgetreductions. The Establishments were first broughtunder a single authority-the Controller, Establish-ments, Research and Nuclear (CERN)---separatingthem from the previous direct Service orientation.Lead Establishments were then assigned to areas oftechnologies with the other Establishments ● ’en-couraged” to follow. There was no need for

draconian measures. Instead, a gradual rationaliza-tion occurred.

From this perspective, DoD and the U.S. defenseindustry need to exploit foreign technology far morethan they do today. To achieve major gains, how-ever, will require the easing of restrictions imposedon industry by strict technology-exchange andexport-administration regulations as well as a reduc-tion of time-consuming procedures that governindustrial collaboration. There is another imperativeto increase cooperation on international defenseresearch: our European allies are developing acoherent program of intra-European cooperation incivil and military research and technology. Unlessthe United States develops a policy for transatlanticresearch cooperation, we may become ‘locked out”of their plans. NATO armaments cooperation wouldsuffer a severe blow, and U.S. industry might welllose existing competitive advantages in world mar-kets.

The Nunn Amendment to the fiscal year 1986Defense Authorization Act (and its subsequentcontinuation) has been a “shot in the arm” forNATO armaments cooperation. It has given both theServices and U.S. industry incentives to pursueNATO (and now non-NATO) cooperative programsin systems development. A simple extension of theNunn legislation might provide a similar incentivefor research cooperation. A specific budget (6.1/6.2), obligated for cooperative research, wouldundoubtedly result in increased interest on the partof Services and our Allies.

Option 6: OSD could establish DoD-widesystematic guidelines to enhance the transfer oftechnology into new or existing weapon systems.

Congress might recommend that DoD developguidelines for selecting, planning, managing, andevaluating all advanced technology demonstrationprojects. OSD could develop these guidelines withthe participation of the three Services, DARPA, andSDIO. It is important to have such guidelines, sincethe purpose of DoD’s technology base programs isto insert new technology into weapon systems asrapidly as possible.

Because DoD’s current technology -transfer proc-ess relies heavily on individual initiative, it isinconsistent and haphazard. Developing a DoD-


wide advanced technology demonstration processcould provide a more rational basis for settingpriorities, expedite the rate at which new technolo-gies are adopted, and provide consistent guidancefor evaluating the success of various advancedtechnology development projects.

Option 7: Appoint a civilian research advocatein DoD with oversight of all technology baseprograms (6.1, 6.2), and a role in coordinatingadvanced technology development (6.3A) withService research heads.

The overriding task would be to ensure that theresults of DoD’s technology base programs areexploited by the Services as soon as possible. Threesubordinate functions are critical to this task: 1) anoversight process that augments normal ‘‘peer”review with a management review focusing onnon-scientific factors, such as priority, applications,opportunity costs, and cost-benefit; 2) a means forDoD-wide dissemination of data on technologiesdeemed ready for transition; and 3) procedures formonitoring the efficiency of the transition (i.e.,technology transfer) process.

The Administration could use parts of severalmodels; although no single system covers all aspectsof this option. In the U. K., the Chief of DefenceProcurement (CDP) and in France the DelegueGeneral pour l’Armement (DGA) have full authorityover all defense R&D, and procurement. Each has adeputy for land, air, and naval systems, and a deputyfor R&D, who is responsible for all “project-free”research. This approach works fairly well for thesenations; however, in each case the scope of theirtechnology base programs is less than one-tenth ofDoD’s—and even less is actually “project-free.’* Itmight not be appropriate to adopt these models intoto, but they do make a case for OSD to assert moreauthority over the content of technology baseprograms.

Neither the U.K. nor France has appointed atechnology transition authority, and both have asmuch difficulty in this area as DoD. Some lessons,however, can be found in West Germany and Japan,especially in the civil fields. In West Germany, onlybroad civil research goals are promulgated fromBonn, and the private sector is organized to effect

transfer. As previous chapters described, two influ-ential private (but largely government-funded) so-cieties are central: 1) the Max-Planck Societyperforms basic research and serves as a “locomo-tive” for other research institutes and universities,and 2) the Fraunhofer Society performs appliedresearch and couples closely with industry to effecttechnology transfer into the marketplace. While noteffecting direct control, a number of government-sponsored groups provide oversight and advice. InJapanese industry, teams are formed at early stagesof research that consist of researchers and expertsfrom engineering, design, manufacturing, and mar-keting. Their basic mission is to ensure the fastestpractical transition from research to a marketableproduct.

DoD is well positioned to adopt this option. TheGoldwater-Nichols DoD Reorganization Act estab-lished the USD(A) to oversee all defense R&D andacquisition. The DDR&E, reporting to USD(A),looks across all of DoD’s technology base programs.DARPA’s new role in prototyping, and its recentlyexpanded involvement with technological initia-tives, provide the framework for a “transitioning”authority. However, budgeting and review authorityover most of the technology base program rests withthe Services and other DoD components. Congressand the Administration could bring these elementstogether under DDR&E and charge this seniorofficial with exploiting the results of the govern-ment’s $10 billion annual investment in research andtechnology. However, this could result in greatertechnology push, which some believe could bedetrimental to U.S. defense efforts.

Option 8: Streamline the current OSD organ-izational structure for RDT&E programs.

Peter F. Drucker has discussed the importance ofsound organizational structure:

Few managers seem to recognize that the rightorganization structure is not performance itself, butrather a prerequisite of performance. The wrongstructure is indeed a guarantee of nonperformance; itproduces friction and frustration, puts the spotlighton the wrong issues, and makes mountains out oftrivia. 3

3U.S. Senate, “Defense Organization: The Need for Change,” Staff Report to the Senate Committee on Armed Services, Oct. 16, 1985, p. 92.


The current DoD organizational structure forresearch, development, test, and evaluation (RDT&E)appears ill-suited to its role. Both civilian andmilitary representatives have argued that if DoD’stechnology base programs are primarily responsiblefor maintaining DoD’s scientific and technologicalsuperiority, then the official in charge of the RDT&Eprogram should report directly to the Secretary ofDefense. The Goldwater-Nichols Act puts primaryresponsibility for the technology base program withthe DDR&E, who reports to the USD(A). Unfortu-nately, DoD’s general preoccupation with procure-ment issues has diverted the attention of the USD(A)from important technology base issues.

Within ODDR&E, the Deputy for Research andAdvanced Technology (R&AT) is responsible foroversight of the Services’ S&T programs, while theDirector of DARPA reports directly to the DDR&E.This organizational arrangement has made it diffi-cult to coordinate DARPA’s activities with those ofthe Services fully. Although DARPA’s mission isdifferent from the Services’, its ultimate responsi-bility is to support the development of high-risktechnology for the Services. This activity could befacilitated by requiring DARPA and the Services toreport to the same office.

Finally, any organizational review focusing onthe technology base should include SDIO. Since theDirector of SDIO reports to the Secretary of De-fense, there is no formal technology base coordi-nation with the Services and DARPA. If OSD is todevelop an effective technology base investmentstrategy, much closer coordination will be neededbetween SDIO, DARPA, and OSD.

Option 9: Improve DoD’s ability to attracttop-quality political appointees and high-levelcivil servants.

Current and former DoD S&T personnel assertthat DoD is unable to attract individuals of highscientific and managerial talent. They contend thatthis problem must be solved if competent civilianleadership is to be restored within OSD and theServices.

There appear to be three specific actions thatCongress could take to help resolve this problem:

1. increase salaries of DoD science and engineer-ing personnel,

2. amend the conflict-of-interest statutes, and3. amend the Federal tax laws with respect to the

forced divestiture of assets.

These recommendations are not new, but are pre-sented as options to highlight a problem that appearsto be deepening.

Numerous studies conducted by the FederalGovernment and the private sector have documentedthe growing disparity between compensation for topFederal S&T personnel and that of their universityand private-sector counterparts. Congress couldexamine the possibility of eliminating pay caps forsenior executives within DoD and instituting com-pensation that reflects the current market for suchindividuals.

DoD should also have a pay structure thatcompensates officials on the basis of their S&Tmanagement responsibilities. Unlike the privatesector, the Federal pay cap fails to distinguishbetween a laboratory director who manages theactivities of 3,500 people versus a director whooversees a 500-person laboratory-or for that matteran OSD Senior Executive Service (SES) managerresponsible for supervising a 12-person staff. DoDcannot pay top scientists and engineers what theprivate sector can.

Past and current DoD civilian personnel assertthat potential top-level political appointees—andscientists and engineers—are often reluctant tomake the financial sacrifices required under theFederal conflict of interest or “revolving door”statues in order to accept a high-level DoD position.As the Senate Armed Services Committee study onthe DoD organization observed, rather than alteringthe divestiture requirement, Congress could alterFederal tax law with respect to the forced sale ofassets.4 This would still protect the objective ofmaintaining public confidence in OSD officials, butit would reduce the onerous financial consequencesof accepting public service.

dIbid.


Structure of the Laboratory System

Option 10: Restructure the military’s RDT&Eorganization by establishing corporate laborato-ries for each of the military services and creatingsome full-spectrum weapons development cen-ters.

The creation of “corporate” laboratories dedi-cated to individual Services would rationalize theconduct of DoD’s RDT&E program. Corporatelaboratories could perform the bulk of each Serv-ice’s technology base work, by generating researchconcepts and bringing them to the demonstrationphase. The laboratories’ mission would be to mar-shal the technical resources of their Services toattack new objectives. This would require DoD toincrease its investment in technology base programs,rather than have the corporate laboratories competewith academe for a shrinking 6.1 budget.

The existing engineering and development cen-ters would have to continue to establish priorities intheir development programs, pursue a dialogue withthe corporate laboratories, and position themselvesto transition technology. At their discretion, theywould also compete for that portion of the corporatelaboratory’s funds that would sponsor externaltechnology base work.

The creation of corporate laboratories wouldinvolve more than establishing the equivalent of theNaval Research Laboratory for the Army and AirForce. Corporate laboratories could receive fundingfrom a variety of sources, under procedures analo-gous to those by which the Department of Energy(DOE) national laboratories-and even some ofDoD’s federally funded R&D Centers—receivetheir funds. They would receive multiyear blockfunds to cover the programs authorized by theirrespective Service, plus reimbursable funds to sup-port work for others. In this context, “others” wouldinclude the other military Services, independentDefense agencies like DARPA and SDIO, andcivilian Federal agencies.

The Services might create semiautonomous unitswithin their corporate laboratories for certain kindsof high-risk, high-payoff programs. A Service might

consider a particular discipline or mission area soimportant that it would justify the creation ofspecialized units working on them, The Army hasused this approach in setting up its Night Vision andElectro-optical Laboratory and the Life-cycle Soft-ware Engineering Center. Even research in highlyspecialized areas requires collaboration by expertsin several disciplines. And if the work is well done,the results of such specialized research can flow intoother areas.

Corporate laboratories could create a much richernetwork of external relationships----comparable tothose enjoyed by DOE’s national laboratories. Thelaboratories’ work in basic research and exploratorydevelopment could make them more attractivepartners for collaborative ventures. These relation-ships could include: technology transfer mandatedby the Stevenson-Wydler Act; work for non-Federalsponsors like that done by the National Institute ofStandards and Technology (NIST); and the buildingof formal and informal communication networkswith universities and industry. Far from precludingcollaborative work, the corporate laboratories’ mis-sions would virtually require it. The point is to avoidthe two extremes: laboratories serving as “pass-throughs” for development money, on the one hand,and on the other, the inbreeding that results whenlaboratories try to do everything in-house.

Congress and DoD might also create weaponsdevelopment centers to pursue work on significantmilitary systems problems, as was suggested in a1966 Defense Science Board (DSB) report.5 Whileprivate industry would continue to do virtually all ofthe engineering and production work, these centerswould encompass the full spectrum of activitiesfrom advanced development (6.3B) to the creation offeasibility models to demonstrate ‘‘proof-of-principle” in a military environment. These centerswould be project-oriented research and engineeringinstitutions working in broadly conceived weaponsareas.

As the DSB defined it, weapons developmentcenters would have certain family resemblances.They would have a critical mass of at least 1,000scientists and engineers; the center director wouldhave direct control over all the necessary resources;

sDefense Science Bo~d, “Dcp~ent of Defense In-House Laboratories, ” report prepared for the Office of the Director of Defense Research &Engineering, Oct. 31, 1966, p. 9.

95-677 - 89 - 5


center specialists would participate in determiningmilitary requirements associated with its mission;and the center would be involved in the initialprocurement of equipment. Instead of serving aspass-through agencies, each center would do most ofits development engineering in-house, with con-tracts serving to support such work.

The advantages of this approach are straightfor-ward. A weapons development center would havethe critical mass to work on a range of problems,clear responsibility for its end products, and theability to respond quickly to military emergencies.The existence of such centers would enable DoDengineers to work on military problems and to bringtogether specialists in many disciplines.

In creating such centers, certain problems wouldhave to be solved; for example, how would a centerconcentrating on aeronautical development relate toone whose mission encompassed missile design?Further, each center would inevitably be biasedtoward its own system, even if another kind ofweapon or platform would provide a better militarysolution. As the DSB panelists were well aware, bytheir nature such centers would tend to committhemselves to long-term projects, even in the face ofevidence that other approaches might work better.Such centers could easily reduce their contractors tosuppliers of narrowly specified equipment andservices, with nothing to offer to the center’sportfolio of ideas.

Option 11: Consolidate some military labora-tories and close others.

A case can be made that there are too many DoDfacilities whose contribution to the defense technol-ogy base is difficult to discern. In the currentenvironment, both Congress and DoD should ex-plore merging some facilities that can no longerstand on their own and closing others. This optionappears radical only if one assumes that Federalfacilities are permanent. There is no definite Federalpolicy on the closure of government facilities,although something can be gleaned from Office ofManagement and Budget (OMB) circulars barring

agencies from performing activities more suitablefor the private sector. Agency officials have assertedprinciples that might justify closing a substandardlaboratory: if it has served its initial purpose, if thereis no likelihood that a new role for the laboratory canbe found, or if closing the laboratory would not leavea significant gap in the national capability to performR&D.6

Consolidation and closure may be more palatableoptions now than at any time since the mid-1970s.The closing or merging of R&D facilities has notalways been unthinkable. In the early 1970s, forinstance, the Air Force undertook a major reorgani-zation of its laboratories, converting its CambridgeResearch Laboratories from a basic research to an“exploratory development” institution, closing theAerospace Research Laboratory at Wright-PattersonAFB, and delisting one of its contract researchcenters. This was also the period when NASA closedits Electronics Research Center and transferred thefacility to the Department of Transportation; andwhen part of the Army’s Fort Detrick became acontractor-operated facility working for the NationalInstitutes of Health.7

The present budgetary environment will probablyencourage the Services to make difficult but neces-sary choices. Short of actual closure, the Servicescould employ a number of strategies to keep theweaker laboratories going—authorizing them toseek support from other sponsors, clarifying theirroles, and redirecting them. But at some point, theServices may decide that they can no longer carry allof the research centers they currently support. For aService, there may well be a bigger payoff in cutting,say, 20 percent of its laboratories than in slicing 20percent from each laboratory’s budget.

The advantages of this approach are threefold:first, fewer laboratories would make the remainingones more visible to their sponsors; second, morefunding available to the remaining centers wouldstrengthen them and probably produce more worth-while research; and third, closing some laboratories

~Arnold S. Levine, Mamging IVASA in the Apollo Era (Washington, D C : N a t i o n a l A e r o n a u t i c s a n d S p a c e A d m i n i s t r a t i o n , S c i e n t i f i c ~dIniorrnation Branch, 1982), p. 137.

TFor ~ex ad other Cxmp]es, see TOJ, wilb~ks, ‘Domestic Models for National LaboraiOV utlll~a[lon~ “in Energy Research Adviso~ Board, TheDeparrmenr ofEnergy A4ultiprogram Laboratories, Vohune //: Speciul ,’hdies (Oak Ridge, TN: Oak Ridge National Laboratory, September 1982), pp.66-67.

Chapter 7-lmplications for the Defense Technology Base: Options for Congress ● 121

is a necessary step to consolidating disciplines thatshould go together.

DoD could possibly strengthen its technologybase with fewer and larger laboratories and engi-neering centers, because they would have the criticalmass of professional staff to move on severalresearch fronts. On the other hand, the question ofwhich facilities to merge or close is exceedinglycomplex and highly political, and changes wouldrequire several years to implement.8 A usefulprecedent for an approach might be found in therecent base closure legislation. Adopting a similar“package deal” might ease the process.

Option 12: Promote the sharing of “national”facilities and the interchange of personnel be-tween government laboratories.

In the current budgetary environment, fewagencies have the luxury of duplicating existingfacilities. The sheer expense of building a new windtunnel or particle accelerator is forcing agencies toturn to collaborative ventures—a tendency thatshould be encouraged. At the same time, no agencywill willingly depend on another to accomplishsome of its most important programs. The creationof national facilities available to all qualified usersis one way out of this impasse.

The multiprogram DOE laboratories are not theonly ‘‘national” entities that the government spon-sors. Since 1980, NASA has opened two nationalfacilities at its research centers: the National Tran-sonic Facility at Langley Research Center, and theNumerical Aerodynamic Simulator at the AmesResearch Center. Both are world-class facilities thatkeep the United States in the forefront of aeronauti-cal research and serve all U.S. commercial, military,and scientific requirements.

The creation of such national facilities bears onthe DoD laboratories in several ways. First, and mostobviously. they would obviate the need for DoD toduplicate—at great expense—facilities that alreadyexist, Second, as resources available to all qualifiedusers, they actually make the military laboratoriesmore productive, with no investment beyond thatrequired for covering their share of facility opera-

tions. And third, the NASA facilities offer theoptions of either onsite use of facilities or remoteaccess, via data communications networks that arenow in place.

The same principle of shared use applies to NISTand the multiprogram DOE laboratories. The mis-sion of NIST demands extensive work for otheragencies in a variety of areas. At the same time, theDOE weapons laboratories are seeking a broaderdefense role in nonnuclear weapons research. Giventheir capabilities, one might expect that the DOElaboratories’ multidisciplinary strengths could be-come a major resource for DoD.

Program budget pressures will no doubt forceDoD laboratories to work more closely with eachother and with those of other agencies. Congresscould explore the possibility of giving corporatelaboratories created by each Service the freedom totake on the work of others. Like the DOE laborato-ries, a certain portion of each laboratory’s operatingbudget would include work undertaken for anotherService, another Federal agency, or even State andlocal governments. By working for others, labora-tory scientists and engineers would acquire more ofa ‘‘hands-on” acquaintance with dual-use technolo-gies than an individual Service might be able tofund. A broader base of interests would, in turn,allow researchers from all the Services to work onmany generic technologies (e.g., software engineer-ing, weapons simulation, and high-speed process-ing) that could lead to Service applications.

It might also be to the DoD laboratories’advantage to promote an exchange of personnel withother facilities working in similar areas. There areprecedents for such assignments. For many years,NIST has had a Research Associates program,whereby scientists and engineers from industry cancome to NIST at their company’s expense to workfor a specified period on projects of mutual interest.The DOE laboratories have had even closer ties withoutside organizations, not least because their con-tractor status virtually demanded it. These arrange-ments include joint ventures with industry, summerstudy programs, joint appointments with the operat-

8Ano~er option, not fufier ~on~idcred here, would ~ t. keep ce~aln la~ratories in he ‘‘doub~ul” ~atego~ open, while leaving them free to s~k

support from any sponsor willing to provide it. A military Service would not be responsible for assuring a total ICVCI of support. Instead, the lab wouldbe placed on a footing analogous to Naval Industrial Funding, with military and other customers paying for much of the cost of operations.


ing contractor, and the creation of university consor-tia like Oak Ridge Associated Universities.

A program to promote short-term exchanges oflaboratory personnel would serve DoD aims in manyways. It would give DoD scientists and engineers abetter idea of the research being sponsored at DoDand other government laboratories. It would enableprofessionals from different services to work ongeneric, or cross-cutting, technologies. Finally, itwould promote the idea that DoD laboratories—particularly the corporate institutions—are resourcesthat should be freely available to all of DoD, as wellas some of its industrial contractors.

Laboratory Management

Option 13: Convert some government labora-tories to government-owned, contractor-operated(GOCO) status.

From time to time, government panels like theWhite House Science Council and DSB havesuggested that some government-operated labsshould convert to GOCO status. Experience in suchconversion is limited to the partial 1983 conversionof a small DOE technology center to privateoperation.9 All other GOCO laboratories have hadthis status since their inception. The issue that DoDand the Congress should consider is what, if any,advantages would flow from a GOCO conversionthat could be achieved in no other way. It issignificant that, in its 1987 summer study, the DSBproposed such a conversion for existing laboratoriesmainly as an alternative to improving their operationwithin the system. As the study group put it, ‘‘whereexisting government laboratories are not performingwell, conversion to a GOCO laboratory has someattractive properties. ” But it also added that “suchconversion would involve significant disruption andpolitical opposition. ”10

Based on the DOE’s experience, the FederalR&D community knows something of the advan-

tages and disadvantages of the GOCO approach. Thegreatest of these advantages is flexibility in person-nel management: flexibility in developing personnelsystems; flexibility to set salaries at levels compara-ble to those in the private sector; and flexibility tomove staff from one activity to another on shortnotice.11 And provided they comply with Federalnorms, GOCOs face a somewhat lighter regulatoryburden than do their government counterparts.

GOCO arrangements also carry disadvantages.Some analysts claim that GOCO status reduces alaboratory’s commitment to its sponsor’s mission,and that there may be a perceived conflict of interestif the contractor is a for-profit corporation. Other,more fundamental criticisms are that the systemfosters a lack of accountability and that, by turningtechnology development over to a contractor, thegovernment loses control of the operations of itslaboratories. Nor are GOCOs free from the moreburdensome kinds of oversight. If anything, theseinstitutions tend to impose on themselves the kindsof burdens from which their status as GOCOssupposedly exempts them.

In sum, GOCO status may be an option undercarefully specified conditions: if an agency isconsidering a new facility; if government operationforecloses the possibility of improving a labora-tory’s operations; or if the sponsoring agency wantsthe expertise of an industrial contractor for produc-tion facilities or of a university for research anddevelopment. At this time, there may not be enoughhard evidence either way to justify the conversion ofa government laboratory to GOCO status.

Option 14: Eliminate institutional barriers tothe effective operation of DoD laboratories.

Congress could facilitate change by extendingpractices at certain facilities to the rest of DoD’sR&D community. The measures discussed beloware in line with the DSB’s 1987 recommendationthat each Service select at least one ‘‘representa-

g]n 1983, tie Energy ~p~ent tr~sferred responsibility for its Bartlesville Energy Technology Center to the Illinois Institute of TechnologyResearch Institute (lITRI). Under a cooperative agreement between DOE and IITRI, the center, renamed the National Institute for Petroleum and EnergyResearch, would work for both government and indusuy. IITR1 is responsible for the facility and shares operating costs, but receives no fee. The contractprovides that fees earned from industrial clients revert to DOE, and that for basic research, IITRI must write an annual work plan for DOE approval.

IODcfcnse Science Bowd, 6‘Sunlmer Siudy on ‘Rxhnology Base Managcmcnt: How to Improve the Effectiveness ~d Efficiency of the R&D ~oces%”report prepared for the Of’fice of the Under Secretary of Defense for Acquisition, December 1987, p. 15.

I Ion t~lew ~d Other GOCO feat~es, we Office of Scicncc ~d Tcchno]c~W p~li~y, Executive Office of the President, ‘ ‘Final Rcpofi of the Working

Group on Federal Laboratory Personnel Issues, ” July 1984, p. 24.


tive” laboratory, and so alter its management that itcould attract the highest quality staff, improveoperations, and provide management with “author-ity and accountability. ”12

The most immediately obvious changes would bein personnel management. For example, Congresscould extend the approach embodied in the ChinaLake experiment (described in ch. 5) to all DoDlaboratories. This would give technical directors andtheir division managers added flexibility to recruitand promote effectively: broad pay bands thatincorporate a simplified classification scheme; anemployee appraisal system that links pay to per-formance; and an emphasis on performance as aprimary criterion for retention. If anything, theChina Lake approach could be more carefullytailored to the problems of professionals working atgovernment laboratories. DoD might combine fea-tures of both the China Lake and NIST demonstra-tion projects: the consolidation of 15 grades into afew broad pay bands and the delegation of classifica-tion authority to line managers that is common toboth projects; and NIST’s direct-hire authority andthe ability to offer its professional employees totalcompensation “comparable” to that offered in theprivate sector for the same positions. In addition,Congress might consider allowing laboratories topay exceptional scientific and engineering talentmarket rates above civil service ceilings, and paycompetitive salaries for all technical employees.

This option raises two opposing questions. Whyshould an approach tried at three facilities beextended to the rest of DoD? Conversely, if theChina Lake/NIST approach has been successful,why not extend it government-wide? China Lakedemonstrated that a simplified personnel systemcould raise employee morale and lead to higherretention of more capable professionals, even if itdid not automatically lead to government pay scalesthat were more competitive with the private sector.In answer to the question, “Why not government-wide?” it can be argued that the current DoD (andcentral oversight agency) approach to personnel andfinancial management violates a basic rule of equity:Do not treat unlike institutions as though they werealike. A scientist at the Air Force Wright Aeronauti-cal Laboratories should not be covered in quite the

same way as a Treasury official who maintains thegovernment’s central accounting system or a Gen-eral Services Administration official who managespublic buildings. The merit of the China Lake andNIST approaches is precisely that they recognizethat different groups deserve to be treated differ-ently.

The China Lake experiment may give technicaldirectors and division chiefs an irreducible mini-mum of authority in hiring, promoting, and firing.Similarly, a laboratory technical director shouldhave direct authority over all of his organization’sfunctional offices, such as personnel, procurement,and data processing. Lacking such authority, notechnical director can be fully responsible for hislaboratory’s operations.

Congress should also consider reforming theways by which the laboratories receive and spendtheir operating funds. Studies have shown the effectsof overmanaging and underfunding DoD laborato-ries. In particular, laboratory officials have to copewith uncertain funding—so uncertain that fundsoften do not reach them until late in the fiscal year.

Multiyear and no-year funding might give DoDlaboratories the same kind of institutional stabilitythat the DOE’s national laboratories enjoy. Espe-cially where technology base work is involved,technical directors need the assurance that work willbe both fully and continuously funded, that fundswill cover all expenses, and that funding will beassured over the life of a project. Block fundingcould very well provide this assurance. Under thisapproach, a laboratory would receive a lump sumsufficient to cover the full costs of technology basework, without the need for allocating funds underexisting DoD budget categories.

Laboratory directors also need discretionaryfunding to start new work, to sustain projects whereother funding is incapable of carrying them tocompletion, and to encourage cooperative venturesbetween the laboratory, universities, industry andother Federal agencies. The 1983 Packard Reportrecommended that between 5 and 10 percent of alaboratory’s annual budget be reserved for inde-pendent R&D at the director’s discretion—a rangethat would permit potentially important work that

lzDcfense Science Board, op. cit., fbOmOte lo, p. 19.


now goes unfunded. ’3 While officials will disagreeon the appropriate size of the discretionary pot, thismuch must be said: if a Service considers alaboratory’s mission worth doing at all, it shouldaccord a certain percentage of discretionary funds asa matter of right.

The acquisition process is another area ripe forreform. Evidence mounts that the length and com-plexity of acquisition cycles impose tremendouspaperwork burdens on military laboratories. Con-gress is aware of these problems, and has put in placemechanisms that have somewhat eased the laborato-ries’ burdens. These include the use of BroadAgency Announcements for research and explora-tory development, the exemption of certain kinds ofscientific computers from the ‘full and open compe-tition” provisions of the Competition in ContractingAct, and the use of a simplified procedure for SmallBusiness Innovative Research (SBIR) procurements.These approaches could be extended to other opera-tions, for example, the acquisition of office equip-ment and general-purpose computers.

A major reform in acquisition must reflect aproper sense of the laboratories’ missions. A labora-tory, most of whose personnel monitor contracts,cannot easily carry on its inherently governmentalfunctions, act as a smart buyer, and serve as a centerof technical excellence. And yet the majority ofmilitary laboratories are conduits through whichbuying commands funnel money to industrial con-tractors. Instead of the laboratories acting as pass-throughs for development work, it may be that suchprocurements could be handled directly by theService commands, with the laboratories providingsupporting research before a buy occurs and techni-cal consultation afterwards.

Option 15: Allow DoD laboratories to contractfor those services that are not inherently govern-mental.

As an alternative to GOCO conversion, DoDlaboratories might elect to contract for those servicesthat are not essential to the conduct of R&D. Theprincipal guidance on acquiring commercial prod-ucts and services needed by the government iscontained in OMB Circular A-76. Although that

Circular specifically exempts R&D work from itscoverage, it does include “severable” commercialactivities in support of research and technologydevelopment. Given the blurring of lines between,say, scientific programming and the work of in-house researchers in artificial intelligence, it is oftendifficult to distinguish between activities that areand are not covered by A-76.

The important issues, though, concern efficiencymore than policy. A facility that contracted out allsupport services would achieve a status somewherebetween government operation and GOCO. Suchcontracting out would serve several purposes. Itwould enable laboratories to pay market rates forsupport services; give laboratory executives greaterflexibility in hiring workers and dismissing themwhen they were no longer needed; and bring inprofessionals who would not work directly for thegovernment. Under such a system, a laboratorycould, for example, contract out facility manage-ment, supply operations, and financial and adminis-trative processing. Scientific and engineering pro-fessionals would remain government employees,either under a reformed personnel system based onthe China Lake model or some special system, likethe one used to pay faculty of the UniformedServices University of the Health Sciences.

The best example of this hybrid system can befound in the NASA centers. Since its establishmentin 1958, NASA has routinely contracted out almost90 percent of its total budget, with much of thatgoing for center operations. NASA sponsors twokinds of contract support. First, an agency installa-tion may be managed by government employees,with NASA awarding a master contract for house-keeping and base support and separate contracts formore specialized functions. This is the arrangementat the Kennedy Space Center, where EG&G pro-vides base support, and at the Johnson Space Center,where Rockwell International is the prime contractorfor mission support. Second, a NASA installationmay be government-run, but without a mastersupport contract. Instead, the center would letseparate contracts for services such as technicalwriting, janitorial services, image processing, com-

l~E~ecutivc Office of the president, “Report of the White House Science Council, Federal Laborato~ Review Panel,” Office of Science andTechnology Policy, May 1983, p. 8.

Chapter 7-Implications for the Defense Technology Base: Options for Congress . 125

puter programming, or the operation of trackingstations.

It is important to determine if a hybrid systemalong these lines could work at DoD. The advan-tages listed above seem clear enough. The disadvan-tages are not nearly so. Nevertheless, the bifurcationof support and essential functions might be difficultfor a laboratory to sustain over the long run. Further,there are legal questions relating to the supervisionof contract employees by government officials. It isthe Office of Personnel Management’s position thatsuch supervision constitutes a personal servicecontract and is illegal. It could be argued, though,that so long as the sponsoring agency simply laysdown a general requirement—for example, “Weneed someone to run this facility ’’—it could removesuch contracts from the prohibited category. So longas a few military laboratories are candidates forGOCO status, the hybrid arrangement could be anattractive alternative-provided the legal and otheruncertainties surrounding it are removed.

Funding and Budgeting

Option 16: Institute multiyear budgeting forDoD’s RDT&E program.

DoD first submitted to Congress a 2-yearRDT&E budget request for fiscal year 1988 and1989. Congress approved a 2-year authorization, butappropriated no funds for the second year. Congressmight consider reviewing the feasibility of provid-ing multiyear appropriations for DoD’s technologybase program.14

Multiyear appropriations should decrease theamount of time OSD personnel spend preparing,reviewing, and defending annual budgets. It wouldalso add stability and efficiency to technology baseactivities by providing known funding levels forfuture S&T programs. By reducing the number ofprograms that have to be acted on in any one year itcould also provide Congress with time for morethorough oversight activities, such as giving theAppropriations Committees more time to study therecommendations of the authorizing committees.

Certainly, there are some disadvantages tomultiyear funding. Congress would be giving upsome of its annual oversight powers. Further, ifbudget projections proved to be inaccurate it couldbe difficult to make mid-cycle revisions, or toaccommodate changes in budget priorities. Yet,multiyear budgeting could give OSD and Congressadditional time to consider technology base activi-ties in terms of strategic options. Combined with astrategic technology base plan, a multiyear budgetcould improve the ways in which Congress reviewsDoD’s technology base programs. Lacking a coher-ent technology base strategy, OSD now presents itsS&T budget to Congress primarily as the sum ofindividual program elements. An overall strategicbudgeting approach would help Congress under-stand the trade-offs and implications of differenttechnology base funding options.

Finally, multiyear appropriations could facilitateDoD’s ambitious goals for allied R&D cooperation.By 1994, 10 percent of DoD’s RDT&E budget is tobe committed to joint R&D projects with NATO andother allies. Many Pentagon officials believe thatthis goal is not attainable under the present annualbudgeting cycle. They argue that the EuropeanAllies earmark funds for 3 to 5 years for R&Dprograms, and that European officials may bereluctant to enter into numerous high-risk, coopera-tive R&D programs with the United States unlessCongress is willing to guarantee funding for morethan one year.

Option 17: Separate the technology basebudget from the development, test, and evalu-ation portion of the RDT&E budget.

Although the ultimate success of many develop-ment programs may depend on underlying technol-ogy base projects, the 6.1-6.3A portion of DoD’sbudget is often overlooked in the “high-stakes”game of RDT&E budgeting. The Pentagon’s top-level budget review committee, the Defense Re-sources Board (DRB), seldom considers individualtechnology base programs or priorities; rather, itusually addresses only broad issues of spendinglevel.

141n this case, mu]tiyew appropriations could mean a congressional funding commitment Of from ~ tO ~ years, witi Congress rc~rving tie ‘ghtreview the program at the conclusion of its second or third year of funding. Roughly half the Federal budget is permanently appropriated. (A permanentappropriation is budget authority that became available as a result of previously cnactcd legislation and docs not require annual action by Congress.)


USD(A) could provide Congress with anRDT&E report that clearly highlights the achieve-ments of the Department’s research, exploratorydevelopment, and advanced technology develop-ment programs. The report could summarize currentand future major thrusts of the technology baseprogram, demonstrating the linkage between theseactivities and future military capabilities. It wouldalso be useful if the report were to address potentialcivil applications of selected technology projects, inrecognition of the increasingly dual-use nature ofadvanced technology.

The funding portion of the report should clearlyseparate technology base funding trends from theremaining “DT&E” portion of the budget. Thisbreakdown could provide Congress with a clearpicture of DoD’s RDT&E funding priorities-andthus the “health” of the defense technology base.For example, if the report were produced today itwould reveal that research (6.1) and exploratorydevelopment (6.2) programs have suffered signifi-cant declines in recent years. If Congress wished todo so, it could instruct USD(A) to halt this fundingdecline.

Chapter 8

Lab to Field: Why So Long?

CONTENTSPage

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .. .. .. .. .. .. . . . . . . . . . . . . . . . . . . .AFFORDABILITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .. .. .. .., .. .. .. ... ... ..... . . .

Funding Shortfalls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Aging Inventories .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Policy Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ... . . . . . . . . . .

TECHNOLOGY INSERTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .. .. .. .. ... ... .....The Technology Insertion Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Previous Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Analysis and Policy Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ....

THE DEFENSE ACQUISITION SYSTEM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Introduction . . . . . . . . . . . . . . . . . . . . . .Comparison With the Private SectorEfficiency v. Effectiveness . . . . . . . . .Analysis of the Acquisition ProcessPolicy Options . . . . . . . . . . . . . . . . . . . .

Figure4.Technological had Times . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..’

Figures

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5. Projected Future Costs of Tactical Aircraft . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6. Projected Average Age of U.S. Army Tank Inventory . . . . . . . . . . . . . . . . . . . . . . . . . . . .7. Projected Average Age of USAF Cargo Aircraft Inventory. . . . . . . . . . . . . . . . . . . . . . .8. Projected Average Age of U.S. Army Attack Helicopters . . . . . . . . . . . . . . . . . . . . . . . . .9. Cost v. Regulatory Intensity

Table5.Navy Advanced Technology

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Table

129131131131133136137139141145145146146147156

130132133133133152

PageDemonstrations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143

Chapter 8

Lab to Field: Why So Long?

INTRODUCTIONThis nation’s military strategy depends upon

maintaining a technological lead in fielded militaryweapons systems, both to take advantage of thestrong U.S. technological capability and to compen-sate for a numerical disadvantage relative to theSoviet Union in many categories of weapons.Nevertheless, leading defense officials are con-cerned that the technological level of operationalU.S. weapon systems lags considerably behind thestate of the art.

Some of this discrepancy is unavoidable. Manysystems now in the field have been there a long time.Those just now entering service had their designssubstantially frozen years ago, while the level oftechnology in the laboratory has continued toadvance.

The inherent time lag between lab and fieldnotwithstanding, the length of time it takes for newtechnology to be fielded in U.S. military systems isdisturbing. According to the Department of Defense(DoD),

The Soviets are methodically and efficientlytransitioning new technologies into their vast arse-nal, oftentimes more rapidly than the West . . .Consequently, the Soviets, although lagging theWest in technology, frequently field systems that aresufficiently well-engineered to meet or exceed thecapabilities of counterpart Western systems. 1

A 1987 study by the Defense Science Board(DSB)—a panel advising the Secretary of Defenseon technical matters-found that the inability tomove technology rapidly from research and devel-opment (R&D) programs to systems and products‘‘is a primary contributor to the growing crisis inmilitary competition as Soviet weapons systemperformance approaches and, in some cases exceeds,that of U.S. and Allied forces. ”2

Figure 4 shows when several technologies now inuse in Air Force systems first started to be developedin the laboratory. Some of the apparent lead timesare exaggerated, since the Air Force systems shownare not necessarily the first ones to use the technol-ogy. (For example, since the B-1 was not the firstplane with a variable swept wing, the 20-year leadtime shown in figure 4 is not an accurate measure ofthe time needed to get this technology into the field.)Nevertheless, this illustration does suggest thattypical technologies now being fielded in militarysystems began their development 10 to 15 years ago.

The lead time needed to field new technology canbe reduced in three ways: a nation can, by spendingmore or by spending more efficiently, increase therate at which military systems are modernized; it canhasten the rate at which new technologies areincluded in proposed system designs; and it canspeed up the acquisition process by which anyparticular new system gets into the field. Althoughthese different aspects—affordability, insertion,and acquisition—are discussed separately in thischapter, it is important not to treat them in isolation.Even though the strategy of the United Statesdepends upon maintaining a technological advan-tage, that advantage can be realized only whentechnology leads to increased military capability.Introducing new state-of-the-art technology into amilitary system has no benefit if the system cannotbe developed, if it cannot be supported and main-tained in the field, or if it is prohibitively expensive.

This chapter first looks at the affordability issue,one which cuts across all activities of DoD and is acritical determinant of the rate at which forces aremodernized. It then examines factors that influencethe selection of new technologies when upgradedecisions are made, and it concludes with a discus-sion of the DoD acquisition system itself—theprocess by which decisions to modernize are imple-mented.

I LI,S. Dep~ment of Defense, “Soviet Military Power: An Assessment ol the Threat, 1988,” 1988, p. 149.

2Defensc Science Board, “Report of [k Defense Science Board 198”/ Summer Study on Technology Base Management, ” prepared for the Officeof the Under Secretary of Defense for Acquisition, December 1987, p. E 2.

- 1 2 9 –


chnological Lead Times

On-board oxygen generation●

Electronically agile radar●

Monopulse ECM●

Engine components

Year

Year

Year

Variable sweep wing●

-20 -15 - l o -6 IOC

Guidance seeker●

Laser device theory

Atmospheric transmission codes●

-20 -16 -10 -6 IOC

Second color IR detector●

IR background models●

Laser communications●

IR detector arrays●

Computer-driven displays

-20 - l5 -10 -5 IOC

Structured turbine blades●

inlet/forebody aerodynamics

Look-down, shoot-down radar●

Engine structural integrity●

Fly-by -wire flight control

Year -20 - lb -10 -6 IOC

SOURCE. U.S. Air Force, Headquarters Air Force Systems Command, Depu!y Chief of Staff/Technol~y and Requirements Planni~, The Air Force Science & Technology andDeve/ogxrrerrt P/amring Program, June 22, 1988,

Chapter 8-Lab to Field: Why So Long? . 131

AFFORDABILITY

Funding Shortfalls

The biggest impediment to fielding state-of-the-art technology in future weapons systems may not begetting the technology into the design; it may noteven be getting the design through the acquisitionprocess and into the field. The biggest problem maybe finding the money to buy the new system in thefirst place.

After undergoing unprecedented peacetimegrowth during the early part of the Reagan Admini-stration, the DoD budget faces equally unprece-dented shortfalls in future years as existing plans farexceed likely available funding. Two factors areleading to this squeeze. One is the “bow wave,”representing the bills yet to be paid for weaponsystems that are now undergoing development orentering production. The second, termed the ‘‘sternwave, ” represents the rising cost of supporting andmaintaining weapons that have already been deliv-ered. DoD data show that operations and support(O&S) expenditures for new generations of weaponsystems often exceed those of the systems that arebeing replaced. Although technological improve-ments sometimes actually reduce O&S costs, theComptroller General of the United States has statedthat expectations to this effect generally “are notbeing fulfilled. ”3

Then Secretary of Defense Carlucci stated thatbetween $174 billion and $300 billion will have tobe cut from the planned DoD program for fiscalyears 1990 to 1993,4 assuming that the defensebudget will rise at a rate of 2 percent over theinflation rate. Given the present $140 billion budgetdeficit, the Gramm-Rudman-Hollings spending lim-its, and other Federal obligations such as cleaning upyears of environmental neglect in the nuclearweapons production complex, these increases in theDoD budget may not be realized. A DoD budget that

only keeps track with inflation will fall short of onewith 2 percent real growth by another $36 billionover the next 4 years; one that only remains level incurrent (not constant) dollars falls short by muchmore. Clearly, as the Comptroller General has said,“‘the services have too many systems chasing toofew dollars. ”5

Much of the problem is that the cost of newsystems is increasing at a rate that consistentlyexceeds inflation. This does not necessarily meanthat the money is being wasted, since the quality andperformance of these systems is going up as well.However, given fiscal constraints, this cost growthwill severely limit the quantities of new systems thatcan be purchased. Norman Augustine, president andchief operating officer of a major aerospace firm,drives this point home in a striking way. Extrapolat-ing current trends in tactical aircraft cost growth(figure 5), he finds that the U.S. defense budget willbe able to afford only one plane in the year 2054, andthat the plane’s successor some 75 years later willconsume the entire Gross National Product (GNP).

Aging Inventories

Inability to complete ongoing modernizationprograms at planned rates----even given the recentbudget buildup-aggravates what is already a slowrecapitalization rate within DoD. According toLeonard Sullivan in an analysis conducted for theCenter for Strategic and International Studies ’(CSIS)Defense Acquisition Study:

The total fiscal year 1986 replacement value of allDoD facilities and properties ran just under $3trillion-about 75 percent of the U.S. GNP. Based oncurrent [in 1985] acquisition plans, DoD is ‘‘rollingover” its weapon and support systems roughly onceevery 25 years and its fixed facilities once every 50years. No commercial enterprise operates with suchslow turnover. It would appear difficult if notimpossible to keep defense at high readiness andnear the leading edge technologically with this poorreplacement rate.6

~Ch~]es A. Bowsher, C~mp~~lier General of the United States, quoted in George C. Wilson. “Pentagon Bracing for Two Waves: Rising CostsThreaten Weapons, Readiness, ” The Washington Post, Nov. 13, 1988, p. Al.

‘$The SecrctW of Def~nse’$ rem~ks are referred to in I.IK Slalemcnl by Charles A. Bowshcr, Comptroller General of the CJnitcd StatM, before theSenate Committee on Arrncd Services, Mar. 14, 1988, p. 9.

SIbid.~~on~d Sullivan, Jr., “Charactcrizing the Acquisition Process,” piqxr presented at the Center for Strategic and International Swdics Confcrencc

on U,S. Defense Aequisi(ion, November 1986, Washington, DC, pp. 2-3. ((’omrnissioncd for U.S, Defense Acquisition: A Process in Tmublc, the CSISDefense Acquisition Study ).


OneQuintillion

OneQuadrillion

OneTrillion

OneBillion

OneMillion

OneThousand

Figure 5-Projected Future Costs of Tactical Aircraft

1

1900 1950 2000 2050 2100 2150

Year of Initial Operation

SOURCE: Norman R. Augustine, Augustine’s Law (New York, NY: Penguin Books, 1983), p. 142,

Sullivan also points out that major systems—atleast platforms such as aircraft and ships-can easilystill be in service 40 years after they enteredfull-scale development. With systems replaced, onaverage, every 25 years, aging systems remain inactive service for a long time.

A study done by the DSB in 1984, during the peakof the Reagan buildup, concluded that:

. . . many major equipment inventories will experi-ence a steady aging during the remainder of thiscentury, [and] an increasing share of the necessary

Then-Year Dollars

force modernization of the future must occurthrough the upgrading of equipment already ininventory or already committed to production.7

Figures 6, 7, and 8 show the increasing averageage of Army tanks and attack helicopters and AirForce cargo aircraft.8 Many other weapons categories-although certainly not all—also show increasingaverage ages.

The DSB study, while basically optimistic aboutthe potential for upgrades, did identify some areas

7u.s. Dep~men[ of Delknse, Irnprove(illefense Thro@ Equipment Upgrades. The U.S. and Its Securit)%Partners, Final Report of the 1984 DefenseScience Board Summer Study on Upgrading Current Inventory Equipment, November 1984, p. 2. (Emphasis in original.)

xAlthough the DSB study did not give the source of the data from which lhcsc graphs were derived, the office of Donald Rice (president and ChiefExccutivc Officer of [he Rand Corporation), who chaired the study, confirmed [hat they were calcuhued from the long-range Extended Planning Annexesof the Services. In the past, such long-range plans have tended to ovcrcslimatc future weapons purchases, duc both to underestimating weapon cost andto overestimating available funds.

Chapter 8-Lab to Field: Why So Long? ● 133

Figure 6-Projected Average Age ofU.S. Army Tank Inventory

12

10 /

2-1---- , ,1975 1980 1985 1990 1995 2000

Year

SOURCE: U.S. Department of Defense, Improved Defense Through EquipmentUpgrades: The U.S. and Its Security Partners, Final Report of the 1984Defense Science Board Summer Study on Upgrading Current InventoryEquipment (Alexandria, VA: Defense Technical Information Center, Novem-ber 1984).

Figure 7—Projected Average Age ofUSAF Cargo Aircraft Inventory

Average age in years

Year

SOURCE: U.S. Department of Defense, Improved Defense Through EquipmentUpgrades: The U.S. and Its Security Partners, Final Report of the 1984Defense Science Board Summer Study on Upgrading Current InventoryEquipment (Alexandria, VA: Defense Technical Information Center, November 1984).

for improvement .9 It found that the Services seemreluctant to pursue major upgrades for weaponssystems they are trying to replace, and that a“systemic bias” against upgrades results from

Figure 8-Projected Average Age ofU.S. Army Attack Helicopters

10

5

0 , , , , I1975 1980 1985 1990 1995 2000

Year

SOURCE: U.S. Department of Defense, Improved Defense Through EquipmentUpgrades. The U.S. and Its Security Partners, Final Report of the 1984Defense Science Board Summer Study on Upgrading Currant InventoryEquipment (Alexandria, VA: Defense Technical Information Center, Novem-ber 1984).

consistently underestimating system lifetimes. Sincethe Services are reluctant to upgrade systems thatthey expect to retire soon, underestimating servicelifetimes thwarts upgrades.10

The DSB study also concluded that upgrade plansshould be part of a comprehensive modernization foran entire equipment category, including upgradesand new starts. Moreover, upgrading is much easierif provided for in the original design of the system tobe upgraded.

Policy Options

The future shortfall in procurement funding can bemet in the short term only by reducing procurementexpenditures or by making cuts elsewhere in theDoD budget. The magnitude of the task, involvingcuts of hundreds of billions of dollars from futureDoD budgets, will certainly curtail our ability tosustain a technological advantage through forcemodernization. Moreover, cuts of this magnitudewill have effects that go far beyond hindering theintroduction of new weapons systems and theupgrade of older ones. They will affect overallnational security policy, strategy, and goals that lie

~Defcnse Science Board, op. cit., footnote 7, P. xii.I~,vcV Onc of tie almost 40 helicopters, fighters, attwk aircraft, and wttiwibmarinc warfare aircraft fielded by the Navy since the carlY 195~s has

remained in active service longer than planned, some by over 20 years. The study concluded (p. 2) that there is ‘‘every reason to believe that this picturerctlects the experience of the other Services, too. ”


far outside the scope of this study. Therefore, thisstudy does not attempt to present a completediscussion of the options facing military planners,but will instead sketch out the implications of someof the choices.

Distinctions must be drawn between short-termand long-term solutions. Solutions that might bestimprove the situation in the long run, such asimproving the efficiency of the acquisition processor restricting the number of new starts, will takeyears to produce substantial savings and will nothelp the short-term problem. At the same time,short-term fixes such as deferring or stretching outweapons acquisitions will only make the long-termproblem worse.

Short-Term Measures

To balance the procurement budget in the shortterm, either the cost of new systems must be reducedor else more procurement funds must be madeavailable by cutting other areas. Options for reduc-ing the aggregate cost of new systems involve threedifferent approaches: funding deferral, cancellation,or upgrading existing systems.

Stretchouts or Funding Deferrals-This optionhas been the traditional choice for handling fundingcrises in the defense budget. It has the advantage ofbeing politically much easier than canceling pro-grams outright, and it avoids having to write offprevious investments. However, it is one of the leastattractive solutions for the long run. Not only arecosts deferred, rather than eliminated, but thosedeferred costs are increased due to keeping infra-structure and support on standby, inefficienciesimposed by lowering production rates, changingprogram plans, and inflation. Stretching out someprograms can provide room in the budget for otherimportant modernizations to proceed. However,stretchouts exacerbate program variability, one ofthe most-cited problems with defense acquisition.

Canceling Programs—Although canceling pro-grams forces writing off sunk costs, at least thosecosts do not come back to haunt budget planners infuture years. Moreover, some of the investment canoften be recouped in future programs that draw upontechnology developed in the canceled program. The

earlier in development a program is canceled, theless the sunk cost will be and the sooner thoseresources can be directed to other goals.

However, program cancellation is extremelydifficult, considering the balancing act of negotia-tion and compromise within DoD and between DoDand Congress required for programs to be approvedin the first place. Ideally, those programs judged tohave the lowest military utility of all active programsshould be the first ones to be eliminated in times offiscal constraint. But, there is no universally ac-cepted, objective measure that can help make thisdetermination, Program cancellations-like pro-gram approvals—inevitably involve political judg-ments.

Upgrade Rather Than Replace—The 1984 DSBstudy cited above recommended that system up-grades, rather than replacements, be emphasizedmore heavily in the future. To the extent that presentsystem design makes this possible, increasing em-phasis on upgrades is likely to be an attractive optionfor permitting modernization of systems we cannotafford to replace. This option will not work, how-ever, if the military Services see upgrades as threatsto their long-term plans for future acquisitions. Morerealistic estimates of the service lifetimes of existingsystems will be needed for making valid upgradedecisions.

To promote upgrades, proposals for new acquisi-tions could be required to include detailed compari-sons of the relative merits of replacing v. upgradingan existing capability. The office of the UnderSecretary of Defense for Acquisition [USD(A)]would be an appropriate place for such a review tobe conducted, and it could provide inputs independ-ent of the requesting Service.

Besides reducing spending on new systems,funds could be devoted to procurement by makingcuts in other areas. Options for cuts elsewhere in theDoD budget include:

Reducing Research, Development, Testing, andEvaluation (RDT&E)—As was pointed out inOTA’s previous report on the defense technologybase,11 R&D is always vulnerable to budgetary cutsbecause its benefits are difficult to measure. More-

1 IU.S. Congress, office of Technology Assessment, The Dejknse Technology Base: lntrodu~’tion and @emiew+l .Tpecial Report, OTA-ISC-374(Washington, DC: U.S. Government Printing Office, March 1988), especially pp. 3S-36.

Chapter 8--Lab to Field: Why SO Long? ● 135

over, cutting R&D appropriations in a given fiscalyear reduces actual spending that year by much morethan the same size cuts in other areas, such asprocurement. Cuts in RDT&E funding at first glancewould seem to threaten the U.S. strategy of compen-sating for quantitative inferiority by technologicalsuperiority, since that technological superiority hastraditionally arisen from the DoD technology baseprograms. Upon further examination, however, amilitary strategy that depends on increasing thetechnological sophistication of weapons systemsto the point where they can no longer be affordeddoes not provide a sound foundation for nationalsecurity. Reevaluating the role that RDT&E plays innational security is a long-term, rather than short-term, measure; accordingly, it is mentioned again inthe discussion of “Longer Term Measures. ”

Reducing the Operations and Support Budget—Cuts in O&S budgets, like cuts in RDT&E, have theadvantage of yielding relatively larger reductions inoutlays for that year than cuts in procurement. Thesecuts are therefore attractive in the short run. How-ever, making effective use of our substantial invest-ment in defense systems and personnel requires thatsystems be maintained and supported and thatpeople be trained. Therefore, reductions in O&Sfunding would probably not be the most cost-effective way to reduce the DoD budget in the longrun. Components of the O&S budget, however, cancertainly be reduced. The General AccountingOffice (GAO) has identified improvements thatcould be made, for example, in logistics and sparepart inventories. ’2

Reducing Military Forces—in testimony beforethe Senate Armed Services Committee, the Comp-troller General stated that budget restrictions mayforce the United States into reducing its level ofmilitary personnel. “We may also have to rethinksome of our worldwide commitments in light of ourbudgetary resources. ” This study will not presumeto speculate as to which commitments this countrycould afford to cut back on. However. any reductionsin personnel, operations and support, and procure-ment might have the effect—planned or otherwisc—of limiting this nation’s ability to fulfill its commit-

ments. Reevaluating those commitments in the lightof budgetary pressures represents not so much adecision to let budgets drive policy, as a recognitionthat they do so whether that is desirable or not. Itwould be preferable to start with the decision to limitobligations and reduce spending accordingly, ratherthan let budget cuts limit those commitmentsarbitrarily.

One difficulty with reducing forces to savemoney is that personnel reductions could involveoffering early retirements and redeeming accruedleave, which might actually cost more money in theshort run than retaining people on full salary.

Reducing Civilian Personnel—DoD employsover a million civil servants. Without doing abottom-up review as to how all these personnel areemployed, it is difficult to specify where reductionscould be made. However, many have suggested thatsuch reductions not only would save money but alsowould improve DoD operation. The Packard Com-mission recommended ‘‘a substantial reduction inthe total number of personnel in the defense acquisi-tion system, to levels that more nearly compare withcommercial acquisition counterparts. ”14 However,the likelihood that personnel reductions may notsave much in the short term applies to civilianpersonnel as well as military.

Longer Term Measures

Reexamine National Defense Commitments—This option is the long-term continuation of theshort-term option of “Reducing Military Forces. ”This nation’s long-term defense needs must-bydefinition-meet its long-term defense budget. Whe-ther the adjustment is made by lowering commit-ments or by raising additional funds, a deliberate,well thought-out examination of national prioritiesmay be required. Like any other consensus-buildingprocess that sorts out competing interests amongconstrained resources, this process is inherentlypolitical. It would require a continuing effort.

Improve Acquisition Efficiency-Although thedefense acquisition system probably does spendmore than an acquisition system designed for

l~chwlc$ A. Bowshcr, op. CIL, f~~tnote 3, p. ‘ 1‘

‘~lbid.

lqIbid.

136 • Holding the Edge: Maintaining the Defense Technology Base

optimum efficiency would, those excess costs areoften inherent in the political process surroundingdefense acquisition and in other cases are the pricewe pay for pursuing national goals unrelated todefense. Reducing many of these costs could requireCongress and the American public to reexamine thevalue they currently attach to oversight and review,as well as the cost they are willing to pay to pursuea clean environment, fair labor practices, equalopportunity, and many other objectives.

If savings in the acquisition process could beidentified-either through eliminating waste or bychoosing to relax various requirements that drive upcosts—it would take many years for those savings toresult in substantially lower system costs The vastmajority of the total life-cycle cost for systems nowin development has already been determined.

Reduce or Reevaluate Research, Development,Testing, and Evaluation-The President’s BlueRibbon Commission on Defense Management (thePackard Commission) recommended that “DoDshould place a much greater emphasis on usingtechnology to reduce cost—both directly by reduc-ing unit acquisition cost and indirectly by improvingthe reliability, operability, and maintainability ofmilitary equipment. ” 15 Cuts in the RDT&E budget,were they selected to address the affordabilityproblem, could be associated with a reevaluation ofhow well the DoD technology base serves the goalof cost reduction, in addition to-or instead of—themore traditional goal of enhancing performance.Note also that increasing the emphasis placed onsimulation, as opposed to hardware development,can reduce RDT&E costs to the extent that thesimulations are valid. Increased computational capa-bility, along with growing experimental databases,can improve the validity of simulations.

Enforce Budgetary Discipline-One policychoice here could be to require the Services to makelife cycle cost estimates of new systems for longerterms than they do today, and to prevent them fromstarting new programs unless they provide room inthese longer-term budgets to develop, produce, andsupport the future systems. However, not only wouldsuch a requirement demand accurate cost estimates

for the operation of systems that have not yet beendeveloped—almost a contradiction in terms—but itwould also require dependable projections of futureService budgets, a task that has proven no easier.Moreover, this exercise would be of little use unlesspressures within government and industry to under-estimate the costs of new systems in order to fit theminto future funding requests can be mitigated. Theseissues are discussed further in the section onacquisition.

Consolidate Missions of Weapon Systems—According to the Comptroller General, greaterefficiencies will have to be obtained in a number ofareas such as families of equipment that now fulfillcommon missions. For example, several differenttypes of weapon systems, from shoulder-mountedrockets to tanks to aircraft, in the past have beendeveloped to attack tanks. “While some variety ofsystems is probably desirable, we must exercisegreater restraint in the future because we cannotafford to replace weapon systems on a one-for-onebasis. ” 16

Realigning the assignments of weapon systems tomissions will involve substantial analysis on the partof the military Services. It may even requirereadjusting the Services’ respective roles and mis-sions, if it is determined that tasks presently assignedto one Service will in the future be accomplished byupgrading or replacing a weapon system operated byanother Service. Firm guidance from the Office ofthe Secretary of Defense (OSD) and the Joint Chiefsof Staff (JCS) will be required to make the necessarytrade-offs.

TECHNOLOGY INSERTION

Making room in the budget to update or replace asystem does not automatically ensure the introduc-tion of the latest technology. In fact, relatively fewsystems developments or upgrades are undertakensolely to exploit a specific new technologicalcapability:

Of the many scores of major acquisitions currentlyin progress, fewer than a handful are responding togenuinely original military needs (such as ASAT[anti-satellite weapons]) or to a truly revolutionary

1566A Quest for Excellence: Final Report (o the President,” by the Presidcn(’s BILK Ribbon Commission on Defense Management, June 1986, p. 56.l~ch~]es A. Bowsher, op. cit., footnote 4, P. 11

Chapter 8-Lab to Field: Why So Long? • 137

Soviet threat that challenges U.S. technologicalprowess. Possibly 95 percent occurrent acquisitionprograms are basically aimed at making marginalthreat-related improvements at the same time thatthey offset depreciation of aging inventories withsomething new.17

In the majority of cases where the primarymotivation for an upgrade or replacement is moderni-zation, the introduction of new technology is neithereasy nor automatic. Although there have beensignificant exceptions, DoD has traditionally notbeen very successful at taking advantage of newtechnologies that were promoted by their developers(“technology-push”) in the absence of an interestedconstituency among the technologies’ eventual users(“requirements-pull”).

Those responsible for planning and developingmilitary systems should ensure that the potentialincreased capability made possible by new technol-ogy justifies the risks-in cost, schedule, andpossibility of failure—inherent in that technology’sdevelopment. In the case of obsolete equipment, forexample, putting any replacement at all in the fieldis usually more important than including the latesttechnological features. This conservatism posesbarriers that must be overcome before new technolo-gies can be fielded.

To the degree that proven technologies are notfielded, or promising technologies are not investi-gated, those barriers are inappropriate. However,they should not be eliminated completely. It is not,after all, the mission of DoD to deploy newtechnology for its own sake. Unproven and high-risktechnologies that cannot be developed successfullywill not improve our military capability no matterwhat their ultimate potential may be. Moreover, justbecause a technology is new and effective does notmean that it is the best solution to any particularproblem.

The Technology Insertion Process

Technology insertion depends, of course, on theentire acquisition process, which is discussed moregenerally at the end of the chapter. It refersspecifically to the process by which technicaldevelopments in the laboratory are selected for usein new weapon systems.

The office of the USD(A) was established in partto combine jurisdiction over research and develop-ment with that over production. However, there is avery significant discontinuity between technologybase activities and the later stages of full-scaledevelopment and production. Technology base ac-tivities are undertaken with potential military rele-vance or application in mind, but they are generallynot targeted specifically towards a particular systemrequirement. Instead, they are managed and directedaccording to their field of science or technology, andthey serve to stock the shelves of the “technologysupermarket” from which designers of new systemslater draw.

When a requirement for a new military systembecomes formalized, at least for major systems,funding and responsibility for that system is as-signed to a System Project Office (SPO) dedicatedto satisfying that particular requirement. It might beexpected that developing a major new weaponsplatform-ship, aircraft, land vehicle, or spacecraft—would ease the introduction of advanced technologythrough new generations of subsystems and compo-nents; in fact quite the opposite can occur. In today’spolitical environment, where a conspicuous failurecan be used to delay or scuttle a new program,proponents may choose to outfit an entire platformwith existing systems to minimize the risk of failure.Then to take full advantage of the capability offeredby the new platform, its component systems must beupgraded with new ones after the platform becomesoperational. Providing for upgrades in advancemakes those upgrades easier and more effective.However, technology might be introduced still fasterif new platforms were designed to take betteradvantage of new components and systems from thebeginning.

When a new system or subsystem is undergoingdevelopment, its funding is generally in budgetcategory 6.4, engineering development, and respon-sibility for the system lies primarily with theindustrial contractor or contractors that won thedevelopment contract. Thus, detailed design ofmilitary systems, including the selection of tech-nologies for use, is primarily the responsibility ofdesigners in private industry. Of course, thesedesigners do not work in isolation; their bids must

IT~onard sulliv~, Jr., op. cit., footnote 6, pp. 18-19.


respond to government request, and the bids areevaluated by government employees. SPO obvi-ously has overall direction and responsibility for theproject. However, it is significant that the projectoffice personnel are largely separate from the peoplewho fund and execute R&D within governmentagencies and laboratories.

Several mechanisms help bridge the gap betweentechnology base activities and the design andproduction of particular military systems. The mostindirect might be termed technical diffusion, bywhich findings and results of (unclassified) technol-ogy base funded activities appear in the openliterature and become available for use.18 Interactionbetween those doing R&D in a generic field oftechnology and those responsible for designingparticular systems is an important transfer mecha-nism, as is the actual transfer of personnel fromtechnology base activities to systems engineering.Although technical interchange is essential in promot-ing the development and application of defensetechnology in this country, there is concern thatincreased diffusion could also allow this informationto pass to potential adversaries. Therefore, thegovernment has attempted to restrict export oftechnical information, and there is considerablecontroversy as to the net benefit to the United Statesof these restrictions.19

More direct mechanisms to bridge the “transitiongap” include Independent Research and Develop-ment (IR&D) conducted by industry and (in thecase of defense contractors) partially reimbursed bythe government as art allowable charge on govern-ment contracts. Through IR&D, industrial scientistsand engineers—with feedback from governmentevaluators-can explore technologies and gain suf-ficient expertise with them to feel confident enoughto prepare bids proposing their use in new systems.Industry retains ownership of intellectual propertydeveloped through IR&D.

In contract research and development, thegovernment funds and retains ownership of thedevelopment of a particular technology, component,or subsystem. The research findings and technical

data resulting from such contracts may be madeavailable to others, subject to classification andexport control restrictions. Even without proprietaryrights, the contractor winning such a contractbenefits directly by developing “hands-on” experi-ence with the technology; other companies benefitindirectly from the reports and technical data andmay find themselves forced by competitive pres-sures to develop an equivalent capability. Much ofthis type of development is funded through budgetcategories 6.2 and 6.3.

Perhaps the most direct means for transferringtechnology from the laboratory into systems isbudget category 6.3A, advanced exploratory de-velopment. Category 6.3A includes funding fornon-system-specific prototypes or technology dem-onstration experiments intended to validate tech-nologies to the satisfaction of those—either withinthe system project offices or private industry-whowill ultimately recommend or select those technolo-gies for use in future systems.

None of these transfer mechanisms resembleswhat one government laboratory official charac-terized as the fictitious “midnight loading dock”approach by which a government lab develops aprototype and leaves it out overnight for an indus-trial contractor to pick up, duplicate, and churn outmany identical copies. In reality, the relative roles ofgovernment scientists, government project officesponsors, and industrial developers are far morecomplex. Since the path by which technologiesdeveloped in government laboratories end up insystem designs is so indirect, it can be difficult totrace the contributions of the labs. Technologiesdeveloped in, or whose development is sponsoredby, the government laboratories are picked up byindustry, where they are further developed, refined,perhaps put to new uses, and eventually incorporatedinto system designs. By the time they end up in bidproposals, their origins in government-conducted orgovernment-sponsored research may no longer beapparent.

The preceding discussion of technology insertionapplies to new system developments in which

Igclassificd findings we a]so disseminated through classified journals and seminars. However, the audience is restricted to those holding appropriateclearances who can demonstrate a ‘‘need to know” the classified information.

lgThe exP~ con~o] con~oversy is discussed in depth in a recent study by the National Academy of Scicnccs: National Research Council, Balancingthe National Interest: U.S. National Security Export Controls and Global Economic Competition (Washington DC: National Academy Press, 1987).


industry designs and builds a system to meet specificmilitary requirements. To the extent that the militaryis able to use commercial products, either as they areor with minor modification, it can bypass the lengthydevelopment process and proceed to apply thetechnology embodied in those commercial productsdirectly to military use. In many areas, commercialtechnology leads that available in the defense sector.Such an emphasis on non-developmental items isdiscussed in other chapters of this report that analyzedual-use technologies and their relevance to militaryneeds.

Previous Studies

Several prior studies have addressed difficulties infielding state-of-the-art technology in military sys-tems. The same factors often crop up in analysesdone years apart, showing that understanding aproblem does not automatically lead to a solution inthe face of unwillingness or inability to makechanges. In other cases, problems identified indifferent studies appear to contradict each other.

1981 DSB Study on Technology Base20

In 1981, the DSB issued a report on the technol-ogy base. In addition to identifying crucial technolo-gies to be emphasized and evaluating the currentgovernment technology base investment and opera-tion, this study identified a number of barriersinhibiting successful transition of technology intosystems:

●

●

●

●

Discontinuity of funding, indecision, and theshort-term orientation of many key decisionmakers.The organizational and physical separationwithin DoD of technology base activities andsystem development.Little emphasis on technology demonstrationsthat can illuminate risks, costs, and payoffs ofusing new technology.Little emphasis on “test marketing,” or devel-oping a constituency among the system devel-

opers for using new technological develop-ments.

The study found that “there is a strong incentiveto pursue low risk options” and that “incrementalimprovement is one of the biggest enemies ofinnovation.” 21 It recommended creating an “Ad-vanced Projects Agency” separate from the DefenseAdvanced Research Projects Agency, or DARPA.This proposed new agency, to be staffed by person-nel from the military Services, would developexperiments to quantify the maturity of emergingtechnology, conduct the “test marketing” experi-ments mentioned above, and protect funding forthese experiments from being tapped for otherneeds. In the absence of such an agency, the studystrongly recommended that more funding be allo-cated to category 6.3A, in any case, to conduct theseexperiments. After concluding that DoD does notmake effective plans for inserting technology through-out the life of a system, the panel recommended thattechnology insertion plans be made a basic andfundamental part of program planning.

1985 DSB Summer Study on Practical,Functional Performance Requirements22

This study examined a number of DoD programs,concentrating on the earliest parts of the acquisitionprocess during which the requirements for systemsare determined. In apparent contrast to the 1981DSB study, which accused system developers ofbeing overly conservative in their choice of tech-nologies, this study concluded that developerstended to reach too far. “The foremost factorassociated with unsatisfactory program outcomeswas that the technology, usually after the fact, wasassessed as being unready for entry into engineeringdevelopment. ” Like the 1981 study, however, thisDSB panel also highlighted the need for objectivemeasures of maturity. “It is likely that in almostevery case of failure the project’s initiators believedat the time of initiation of engineering developmentthat the technology was, in fact, mature. ”23

20 Dcfcnsc Scicncc Board, “ReXJ~ of the Defense Science Board 1981 Surnmcr Study Panel on the Technology Base, ” prepared for tk Office of tieUnder Secretary of Defense for Research and Engineering, November 1981.

~lrbid., pp. IV-3, IV-5.zzDcfcn\c Sclencc B(}ard, ‘LRcPfi of the Defense Science Bo~d ~9~5 Summer Study on prac~ica] Functional perfOrma.ncc Requirements,” preparCd

for the Office of the Under Secretary for Research and Engineering, March 1%6.2~1bido, p. 20. (Emphasis in original.)


GAO Letter on Technology Transition,January 198724

Upon concluding its review of the transition oftechnology base activities into weapons acquisi-tions, GAO did not issue a formal report or makerecommendations. Its Associate Director for Na-tional Security and International Affairs did, how-ever, write the Secretary of Defense expressingconcern that “early demonstrations of advancedtechnologies have not received adequate manage-ment attention at the Office of the Secretary ofDefense level. ” GAO found that the “most signifi-cant barrier” to effective transition is the lack ofemphasis on such demonstrations, and it citedrecommendations of the Packard Commission high-lighting the benefits of early prototyping. GAOcalled attention to the low budget priority anddecentralized decisionmaking approach given tosuch demonstrations. In response, the USD(A)agreed with the importance of early technologydemonstration, conceding that the budget for suchactivities had remained level in constant dollarsduring the period reviewed by GA0.25 He noted thatfunding for technology demonstration was projectedto double over the next 5 years.

1987 Defense Science Board Summer Study onTechnology Base Management26

This DSB panel found that “both the DefenseDepartment and commercial industry are seriouslydeficient in rapid technology transition from R&D tosystems and products. ” Like the GAO and the twopreceding DSB studies, this DSB panel concludedthat the “greatest opportunity to improve the rateand effectiveness of this transition process is byincreasing focus on the early advanced developmentphase of the S&T [science and technology] program,that is, Budget Category 6.3 A.” According to thepanel, 6.3A activities should include building andtesting experimental systems in field environments

to establish feasibility and utility before a commit-ment is made to full-scale engineering development.

Army Science Board Summer Study onTechnology Insertion27

The Assistant Secretary of the Army for Research,Development, and Acquisition asked the ArmyScience Board (an advisory body to the Secretary ofthe Army analogous to the Defense Science Board)to survey the Army, DoD, and industrial technologybases to identify candidates for insertion into Armysystems, to evaluate the cost and effectiveness of theArmy technology insertion process, and to reviewthe Army acquisition process to recommendchanges. The panel found that:

●

●

●

●

●

New technology will have to be inserted in atimely manner into fielded systems. Introduc-tion of new systems will be severely limited byfuture funding pressures and (particularly forthe Army) by delays or cancellations of majorsystems, such as the LHX (Light HelicopterExperimental) and the DIVAD (Division AirDefense gun).“To understand how technology insertion canaddress cost and system effectiveness, tech-nologists must understand operational prob-lems . . . The payoffs from the technology baseusually come from combining of technologiesby system developers who know availabletechnical options and can see how to usethem.” 28

Basing technology selection on acquisition costalone will always result in selection of the ‘lowrisk, low cost, low technology approach. ” Newtechnologies have their biggest payoff in life-cycle, not acquisition, costs.29

Acquisition personnel are insufficiently experi-enced.The budget process is a problem.

zQMichae] E. M~[ley, Associate Director, National Security and International Affairs Div ision, General Accounting Office, letter tO CaSpiW Weinbergcr,Secretary of” Defense, Jan.16, 1987.

zsRichmd G~wln, Under SccretW of ~fcnse for ~qulsitlon, letter to Mi~hael Mo[lcy, Associate Director, National !jecurity ~d [ntematiOndAffairs Division, General Accounting Office, May 18, 1987.

z~Report of the Dcfcn= Science Board, op. cit., footnote 2, Dcccmbcr l’~~7.

27 Amy Science Board, ‘‘Army Science Board 1988 Summer Study on Technology Insertion in Army Systems,” prepared for the Assistant Secretaryof the Army for Research, Development, and Acquisi(lon, in press.

z~[bid,, p. 33. (Emphasis In original.)

291bid., p. 57.


Analysis and Policy Options

The problems identified in the studies cited abovefall into three general categories. Those pertaining tothe discontinuity of funding, the short-term focus ofdecision makers, the budget process, and personnelaffect the entire acquisition process and are dis-cussed in the concluding section of this chapter (see‘‘The Defense Acquisition System”). Another set ofproblems relates to technological overoptimism orextreme conservatism and the consequent need forobjective assessments of the maturity of a technol-ogy. These issues can be addressed by increasing theemphasis put on prototyping and technology demon-stration experiments, as well as by building productimprovement cycles into system design. Finally, athird set of problems addresses the organizationalseparation between technology base activities andsystems developers, the lack of “test marketing”new ideas, and the lack of a constituency fortechnological advances within the “user” commu-nities. These issues can be addressed by an organiza-tional structure that attempts to bridge the gapbetween the laboratory and the system developer,placing the ultimate users of a technology in more ofan “ownership” position and therefore makingthem more receptive to the use of that technology.

Prototyping and Technology Demonstration

Most of the studies cited above argued forincreased reliance on prototyping and technologydemonstration. The Packard Commission foundthat making trade-offs between the risks and benefitsof state-of-the-art technology requires reliable infor-mation, and that “the only consistently reliablemeans of getting such information is by buildingprototypes that embody the new technology. ” Itrecommended that ‘‘prototyping, either at the sys-tem or critical subsystem level, be done as a matterof course for all major weapon programs. ”30

Earlier studies had cautioned against overem-phasizing prototypes. Almost 10 years before thePackard Commission reports were issued, a DSBsummer study analyzing the acquisition cycle con-

cluded that “the widespread or mandatory use offull-scale system prototypes for all programs up tothe production prototype level is frequently wastefulof critical national resources-dollars and man-power as well as time. ”31 This panel was particu-larly opposed to the contemporary practice offorcing industrial contractors to fund large costlyprototypes out of their own resources. However, atthe component or subsystem level—rather than thesystem level—the panel concluded that competitiveprototyping could significantly reduce the cost andtime needed to make a full-scale developmentdecision. In summary, the report found that proto-typing could be ‘‘a sound and useful practice inmajor system acquisitions provided that the candi-dates for the use of prototypes are carefully selected,that only those things are prototype which reallyneed verification, and that prototypes are not consid-ered to be some form of free lunch’ for the procuringagency [e.g., by forcing contractors to pay forthem]. ”32

Advanced Technology Transition Demonstra-tions-The 1987 DSB report on technology basemanagement placed a heavy emphasis on AdvancedTechnology Transition Demonstrations (ATTDs),which it saw as an extension of the PackardCommission prototyping recommendations to in-clude technologies that are not necessarily commit-ted to defined system developments. This distinctionis important. Prototypes are test versions of militarysystems that have been designed to meet particularmilitary requirements. ‘‘ Demonstrations,” on theother hand, provide opportunities to test technolo-gies that are militarily relevant; but they do not inthemselves represent designs of specific systems.The technologies they demonstrate, if successful,could be implemented in future systems. (Note thatif a technology demonstration were realistic andsuccessful, there would be less need to prototype afollow-up system using that technology.)

ATTDs, according to the DSB panel, shouldfollow several basic guidelines:

qopresident’s Blue Ribbon Commission on Defense Management, ‘*A Formula ~or Action: A Report to the President on Defense Acquisition,” April1986, pp. 18-19.

slDefensc Science Board, “Report of the Acquisition C’yClc Task Force 1977 Summer Study, ” prepared for the Office of the Under kret~ forResearch and Engineering, Mar. 15, 1978, p. 53.

321 bid., p. 54.


●

●

●

●

They should reduce technical risk by dem-onstrating a technology’s potential and matur-ity in an ‘‘operational,” rather than ‘ ‘labora-tory, ” environment.They should show a potential for new orenhanced military capability, or for a signifi-cant improvement in cost effectiveness.They should be accompanied by a technologytransition plan at the outset of the demonstra-tion. That is, potential applications and oppor-tunities to implement the technology should beidentified at the start, rather than the conclu-sion, of the demonstration process.They should involve the participation of boththe developer of the technology (typically aService Systems Command) and the system’sultimate user (an Operating Command). Theuser should serve as sponsor, with the devel-oper as project manager.33

According to the DSB panel, a successful ATTDwould clarify the definition of the military need thatthe technology is to meet; stimulate strong accep-tance and sponsorship of the innovation among itsultimate users; combine viewpoints of the research,development, production, and operational commu-nities; clearly prove both the maturity of thetechnology and the satisfaction of a perceivedmilitary need; provide visibility to those higherlevels within DoD and the Congress that willultimately approve subsequent developments; andensure adequate financial support to meet the goalsof the project and initiate follow-on development.

Given a limit on resources allowable for suchdemonstrations, together with the need to provideenough funds to do a meaningful experiment (esti-mated by the DSB panel as typically $10 million to$100 million over 3 years), ATTD candidates wouldhave to be selected competitively. This competitionshould ensure that the best ideas get funded. TheDSB panel urged that funding for these ATTDs be“fenced off” from other R&D needs so thatoverruns on large, more immediate demonstrationsdo not threaten the many smaller, longer-term R&Dprojects. (What this means in practice, of course, iseither that provision should be made in advance foroverruns when preparing project budgets, or else

that overruns should be covered from somebodyelse’s pot.34)

The panel urged that these ATTDs be conductedwithin the existing military Service and defenseagency acquisition procedures, and not centralizedDoD-wide. The various Services now have some-what different practices concerning their 6.3A budg-ets. Most of the $2 billion now spent within 6.3A isless focused, less field-oriented, and longer-termthan the proposed ATTDs would be. The DSB panelrecommended that, by 1991, each Service devotehalf its 6.3A budget to ATTDs, sufficient to fund atotal of 20 to 30 projects.

Existing Technology Demonstration Pro-grams— At present, the Navy and the Air Force eachhave a program embodying many of the principlesrecommended by the DSB for ATTDs. In essence,both involve establishing an agreement betweenthe developer and the user that if the technologyis successfully demonstrated, it will be used; thecriteria for success are jointly developed at theoutset. Prior agreement is required both to establisha sense of sponsorship in the user and to ensure thatthe user reserves sufficient flexibility in its out-yearbudget requests to make funds available for theprogram once it has been successfully demonstrated.

DARPA, for its part, has significantly increasedits role in prototyping technologies. This increasedrole has proven controversial.

Navy----The Navy has the smallest 6.3A programof the three Services, totalling $189 million in fiscalyear 1989. Part of this 6.3A program representsgeneric technologies—such as explosives develop-ment—that contribute to many weapons systems.The remaining part of the 6.3A budget providescandidates for Advanced Technology Demonstra-tions (ATDs), which formed the model for the DSBrecommendation regarding ATTDs (see table 5).

Navy ATDs are funded through the Navy-wide6.3A account and are not funded or managed by thecommands responsible for the development ofparticular new systems. ATDs therefore provide anopportunity to demonstrate a high-risk technology toa skeptical customer—a system development com-

33Dcfense s~iencc Board, op. cil., footnote 2, pp. 22-23.34~oviding ~ontingcncy fwlding in DoD budgeting is discussed ]a[cr in [his ~hap[cr under ‘ ‘Reducing program Variability’. ”


Table 5-Navy Advanced Technology Demonstrations

FY 1987 FY 1988 FY 1989

Advanced Fiber Optic All-Optical Towed Surveillance IRSTTechnology Array (infrared search

and track)

SEA RAY (fiber optic Unified Network MADOM (magneto-tether) Technology acoustic detection

of mines)

Undersea Weapons Airborne Transient Quiet Weapon LaunchTechnology (heavy Processor (signal (undersea heavy-torpedo propulsion processor) weight weapons)improvement)

Fiber Optic ADCAP Adaptive Monopulse(heavyweight Countermeasurestorpedo)

Ultra-Low-NoiseCrossed FieldAmplifier a

aAdded to replace the canceled BRIGHT EYE.

SOURCE: Office of the Secretary of Defense.

mand-without making the customer pay up front.They are not appropriate for high-payoff, but low-risk, projects that users are willing to fund withoutany additional incentive. If an ATD proves to besuccessful, according to criteria the user has agreedto in advance, the user agrees to pick up futurefunding. Even with future user support assured, thenew technology cannot be incorporated into newsystems unless the industrial contractors providingthose systems are involved. Typically, about half theeffort on an ATD is performed by industry. More-over, once an improved technological capability hasbeen demonstrated to the Navy’s satisfaction, theNavy will provide incentives for contractors to useit. For example, the Navy may establish performancerequirements that cannot be achieved with oldertechnologies.

Sources for Navy ATDs come from Navy andother DoD labs, DARPA, and industry, In 1988, 55proposals were submitted that were ultimatelywinnowed down to 7 new starts. Projects take amaximum of 3 years and cost about$12 million eachover that time. The total ATD budget is projected togrow to about $65 million per year. In fiscal year1989, the ATD budget was $32 million, whichrepresented about 17 percent of the total Navy 6.3Abudget. However, the Navy is moving towardsapplying ATD management techniques to a muchgreater fraction of its 6.3A activities; it is estimated

that some 50 to 60 percent of the Navy’s 6.3A budgetcould be managed under the ATD model.

Budgets for individual ATDs are protected unlessand until they run into problems. Since the projectsselected are all high-risk, technical problems areexpected; however, to prevent other projects frombeing dragged down, projects that get into troubleare killed. For example, BRIGHT EYE, an elec-tronic countermeasure program scheduled to start asan ATD in fiscal year 1989, was terminated when itappeared that it would not be able to meet itstechnical objectives. Budget cuts are not distributedproportionately to all ATDs, but rather are absorbedby canceling the lowest priority projects in theirentirety.

OSD, following up on the DSB 1987 summerstudy that recommended use of ATTDs, is trying toapply this management technique to 50 percent ofthe 6.3A programs across the Services.

Air Force—The Aeronautical Systems Division(ASD) of the Air Force Systems Command hasinstitutionalized a technology transition processbetween the Air Force laboratories, which controlmuch of the Air Force’s technology base activities,and SPOs within ASD, which are responsible fordeveloping new systems. The objectives of the newprocess are to bound and focus activities at thelaboratories, and to enhance the involvement of


acquisition managers within technology base activi-ties, i.e., to narrow the gap between the originatorsand users of technology.

When an Air Force laboratory proposes a new 6.3activity, part of the budget submission processinvolves preparing a technology transition plan.This plan is presented to a panel composed ofrepresentatives from the engineering support direc-torate (EN) of ASD. This panel--called SENTAR,for Senior EN Technology Assessment Review—evaluates the program’s objectives, recommendsmodifications, compares the program’s schedule toits need in the field, helps determine the program’spriorities with respect to the lab’s other 6.3 work,guides the development of criteria that will denotewhen the activity is ready to be picked up by asystem project office, and determines when theproject meets those criteria. A major goal of thisprocess is to identify system project offices-theusers—that can benefit from the new development.In doing so, the process generates customers forthese innovations who have an interest in seeingthem through to completion. Interested programoffices commit to a “strong moral obligation” topick up support for the activity, should it meet thegoals identified in its technology transition plan.

This technology transition process also estab-lishes incentives for industry to incorporate newtechnologies in their bid proposals. The EN of ASDreviews all Requests For Proposals issued by ASD.In these reviews, EN checks to see that the govern-ment requester will be receptive to companiesbidding technologies that have successfully passedthrough the SENTAR process. If a company issatisfied that use of a new technology will not beconsidered too risky by the proposal evaluators, itwill be much more likely to incorporate thattechnology into its bid.

DARPA-The Packard Commission urged thatDARPA, which was at the time charged withconducting research and exploratory development inhigh-risk, high-payoff technologies, also put empha-sis on prototyping defense systems. DARPA hassince been given an expanded mission in this area.For fiscal year 1989, technology demonstrationswere funded at a level of $237 million, or about 42percent of DARPA’s 6.3 budget. Prototype fundingwas included in the fiscal year 1989 budget within

technology demonstrations, and totalled $43 mil-lion. For fiscal year 1990, prototyping funds willmore than double to $94.7 million and will beseparated from demonstrations; the remaining tech-nology demonstrations will be funded at $167million, $27 million below their fiscal year 1989level.

Given that the military Services at present largelyhave control over their own research, development,and acquisition programs, DARPA is perhaps theonly agency where a revolutionary new technologythat may not fit within the perceived missions of theServices--or that might be seen as threatening thosemissions-can be explored. However, preciselybecause DARPA is outside the existing Serviceacquisition chains, it has in the past faced difficultyin turning technologies over to the military Servicesfor implementation. Giving DARPA a greater role inprototyping will aid the transition of DARPA-sponsored technology from the laboratory to a majorfield experiment. However, without participation byor interest within the military Services, the problemof turning the technology over to the Services fordevelopment into systems might remain.

An additional concern raised over giving DARPAa greater role in prototyping is the degree to whichit will retain its original mission of exploringhigh-risk, basic technology. If the expensive proto-type demonstrations siphon funds from these activi-ties, DARPA’s original mission could be endan-gered.

Preplanned Product Improvementand System Upgrades

In addition to the increased use of prototypes anddemonstrations, another solution to the problem ofattempting too large a technological leap is theconcept of preplanned product improvements. If asystem is designed from the start with the intentionof periodically upgrading its capability, its operatorscan be assured that they will be able to addstate-of-the-art technological capability in the futurewithout demanding it all at once.

Product improvements, or system upgrades, offera lower-cost and faster alternative to new systemsdevelopment for getting new technology out into thefield. However, since they do provide an alternative,upgrades may be resisted by the Services as posing


a threat to new system development. For example, inthe past the Navy has been reluctant to proposeupgrades to its existing fleet of Los Angeles-classsubmarines because those upgrades might be seen asreducing the rationale for the Seawolf, a major newsubmarine that the Navy sees as essential to counterthe increased Soviet threat. Moreover, more realisticlifetime estimates for deployed systems are neces-sary if upgrades to those systems are to receiverealistic consideration.

Summary

Technology demonstrations have the potential forsolving two seemingly contradictory problems:overemphasis on what later turns out to have beenunproven technology, and unwillingness to acceptwhat later turns out to have been viable technology.By convincing the skeptics that a technology canwork, and at the same time disabusing the optimistsof the notion that it can do everything, objectivelyevaluated technology demonstrations enhance thetechnology insertion process.

Prototype development is thought by some to bean important aspect of realistic program planningand cost evaluation. However, others caution thatexcessive prototyping can impede the very processthat it is supposed to enhance.

THE DEFENSE ACQUISITIONSYSTEM

Introduction

For years, defense analysts have been frustratedwith the length of the acquisition process. Delays inacquisition lead to lost time in fielding new systems,and threaten our technological lead over the Soviets.These delays also result in higher costs due to theexpense of maintaining extended development ef-forts. Even more serious than the increased time andcost, according to a DSB panel35 that studied theacquisition cycle over a decade ago, are the “secondorder effects” of delays: technological obsolescenceby the time new systems are fielded, increased riskas designers stretch the state of the art to avoid this

obsolescence, and added complexity as delaysaggravate the tendency to want “everything.”

Moreover, delays beget additional delays. Costescalation due to delays, together with budgetaryceilings, leads to program stretchouts that compoundthe original delay. Extending the expected time fordeployment also causes planners to magnify theanticipated threat, upping the systems’ requirementsand lengthening the development time still further.

No single aspect of the acquisition process isresponsible for schedule delays. To prevent delays,and to shorten the acquisition cycle, the overallacquisition process must be made more efficient andmore effective, Therefore, the following discussionof acquisition, along with Appendix A upon whichthis discussion is based, takes a broad view. Itexamines several systemic difficulties with acquisi-tion, each of which can lengthen the acquisitioncycle or drive up its cost (which, as stated above, canamount to the same thing).

These problems are not new. The foreword to arecent compilation of six major studies of defenseacquisition over the past four decades states that‘‘the bulk of the cures proposed as far back as 1948were still being proposed in 1983 because they hadnever been implemented. ”36 The possibility cer-tainly exists, of course, that none of these studiesidentified the real problems, which therefore remainto be addressed. Alternatively, perhaps sheer intran-sigence and bureaucratic inertia within the Depart-ment of Defense keep it from substantially improv-ing its operation.

More likely, however, is that many difficulties indefense acquisition stem from factors that arebeyond the Department’s direct control and that noamount of unilateral DoD activity can address. Tothe extent that such external factors dominate,improving defense acquisition will require mak-ing large-scale structural and institutionalchanges that would not be restricted to DoD.

Some of these changes are impossible within ourpresent system of government. Others would inter-fere with various objectives that the nation has so

gSDefense science Board, op. cit., f(X)mOtC! ~1, pp. 38-39.

3~David ~ckw~, Andrew Mayer, and c~eryl crow, Library of Congress, Congressional Research SeWkC, “Defense Acquisition: Major U.S.Commission Reports (1949-1988), Vol I,” prepared for the Defense Pollcy Panel and Acquisition Policy Panel of the Committee on Armed Services,House of Representatives, Committee Print No. 26, Nov. 1, 1988, p. V.


far-explicitly or otherwise-decided are at least asimportant as efficient defense acquisition. And stillothers involve resolving longstanding political dis-agreements and identifying common ground in theface of seemingly incompatible positions.

Since the constraints within which defenseacquisition must operate are so important, thediscussion that follows begins with a description ofsome of these constraints and their effects. Next, theanalyses of particular acquisition problems, andspecific options that have been proposed for amelio-rating some of them, are presented. The chapter thenconcludes with a more general discussion of fourdifferent overall approaches that can be takentowards defense acquisition reform. Depending onwhich overall approach one selects, different spe-cific options make sense.

Comparison With the Private Sector

One of the most important features of defenseacquisition is that it is conducted by the government.Since the premise is widely accepted that the privatesector can accomplish tasks more efficiently andcheaply than the bureaucracy-encumbered FederalGovernment, previous studies have looked to theprivate sector to provide a model. A 1977 study bythe DSB found that, while the portion of the defenseacquisition cycle preceding full-scale developmenthad lengthened over the previous two decades, thecorresponding interval for commercial aviation pro-grams had not (see Appendix B of Volume 2).37

More recently, the Packard Commission concludedthat “major savings are possible in the developmentof weapon systems if DoD broadly emulates theacquisition procedures used in outstanding commer-cial programs. ”38

There are certainly lessons that the private sectorcan offer the Federal Government, lessons that thePackard Commission sought to uncover. However,fundamental and inherent differences between thegovernment and the private sector must be under-stood before any of these lessons can be applied.

These differences-described more fully inAppendix A of Volume 2-concern factors such asthe inability to measure government effectiveness inthe same way that profit, or return on investment,provide figures of merit for the commercial world.They involve the standards of accountability de-manded by the taxpayer-and imposed by Congress—on the expenditure of government funds, as well asthe pursuit of national goals such as fairness,environmental protection, and equal opportunitythat may interfere with the ability to acquire defensesystems efficiently.

Other important differences between the govern-ment and the private sector include the role ofCongress and the political process, which has noparallel in the commercial world. DoD’s sheer size(its budget is several times larger than that of thelargest U.S. corporation) imposes inefficiencies ofscale not shared by smaller private-sector opera-tions. Market forces that reward efficient companiesand punish inefficient ones have no counterpartwithin the DoD, which cannot simply sell off ordisband a military Service or agency that does notperform as well as hoped. As James Schlesinger,former Secretary of Defense, has stated;

This is a society that based its system of govern-ment on the Constitution, which calls for a disper-sion of powers. That means that everybody has toagree, and under normal circumstances, most peopledon’t agree, As a consequence, we are never goingto have the kind of model efficiency in the Depart-ment of Defense, or in government generally, thatsome kind of theorist would want. 39

Efficiency v. Effectiveness

Defense analyst Edward Luttwak has stated that‘‘The great irony is that the defense establishment isunder constant pressure to maximize efficiency, andthat its leaders believe in that goal when they oughtto be striving for military effectiveness—a conditionusually associated with the deliberate acceptance ofinefficiency. “4° The nature of defense acquisitionimposes specific requirements that go beyond even

sTDefense Science Board, op. cit., fOOtI’IOtc 31.

s~prcsidcnt’s Blue Ribbon Commission on Defense Management, op. cit., footnote 30, April 1986, p. 12.

3~6 ‘The Second Annua] Report of the secret~ies Of Defense, ” edited transcript of a conference held by the Southern Center for International Studies,at Gaillard Municipal Auditorium in Charleston, SC, Sept. 30, 1988, p. 24,

dOEdwwd LuttwA, ‘‘The Price of Efficiency, “ Military logistics Forwn, July/August 1984, p. 22.


the disincentives to efficiency facing governmentactivities in general. Much of the technology used indefense systems is at a level of sophistication aheadof that used in the commercial sector—if indeed anycommercial analogs exist at all. Although thedefense lead is not as pronounced as it has been—and several areas of defense technology now lagbehind their commercial counterparts-military tech-nology must nevertheless often be developed fromscratch for a relatively limited production run.

Since DoD is the only customer for sophisticatedmilitary systems, producers do not have the optionof selling elsewhere should they not be able to sellto DoD.41 If the Defense Department wants tomaintain a diversity of suppliers, it must buy enoughfrom each of them to keep them in business-evenif their products may not be DoD’s first choice. Themost efficient producer of a military system cannotbe permitted to drive the others out of business.Aggravating the problem of maintaining a viableproduction base are annual purchase sizes—typically determined by externally imposed budget-ary limits—that mandate suboptimal productionrates. 42

Entrepreneurs in the commercial sector willinglyaccept the risk of failure—in the form of a loss ofinvestment or reduced earnings-as the price for thechance to strike it rich. Substantial failure on the partof DoD, however, would have consequences thatcould be far more severe. Therefore, DoD practicesa far greater degree of redundancy and risk aversionthan a commercial enterprise does. Such risk aver-sion also extends to proposals for reform, which facea stricter ‘burden of proof” than might be expectedfor corporate reform.

In light of the factors that characterize govern-ment activities in general and defense acquisition inparticular, it may well be true, as defense analystLeonard Sullivan has concluded, that ‘‘many effortsto make acquisition more efficient are simply

second-order expedients to paper over largely insol-uble first-order problems. ”43

Analysis of the Acquisition Process

The President’s Blue Ribbon Commission onDefense Management was not the first attempt toapply lessons from the private sector to defensemanagement. Seventeen years before chairing theCommission, David Packard, then Deputy Secretaryof Defense, established the present DoD acquisitionprocess to emulate industrial practices of projectmanagement and sequential review and approval.The basic process is one of distinct phases separatedby decision points or milestones. OSD developspolicy for major system acquisition programs andconducts reviews to ensure that those programsrespond to specific needs and are managed soundly.The military Services and defense agencies individu-ally, for the most part, identify those needs anddefine, develop, and produce systems to meet them.

DoD acquisition programs are run according tothe principle of Program Management, in which oneindividual, the program manager, is responsible forintegrating in a single office the diverse adminis-trative, professional, and technical capabilities re-quired to manage the development and production ofa major system. However, many people and organi-zations inside DoD, but outside the program office,have considerable influence over the program’soutcome as well. The separation of responsibilityand authority—whereby people with no directaccountability for a program’s outcome never-theless exert control—has been identified bystudy after study as a major problem of thedefense acquisition structure. Analysts differ as tothe degree to which power and accountability can bebrought back together in the defense acquisitionenvironment.

The review and oversight that acquisition pro-jects receive at all levels, from commands withinindividual military Services through OSD to Con-

q~c~mpmies can produce for expo~, but such expofls must bc approved by the U.S. Government and are not usually approved for technO@es ator above the state of lhe art available to U.S. forces. Moreover, as the abortive F-20 fighter program demonstrated, foreign governments may not wantU.S. systems that the U.S. DoD is unwilling to buy.

~21n-dep~ examination of dcfen= indus~a] base conccms is beyond the scope of 111)\ study. For treatment of this subject, scc ‘‘Bolstering DefenseIndustrial Competitiveness, ” Report to the Secretary of Dcfcnsc b) Ihc Under Sccrclary of Dcfcnsc for Acquisition, July 1988; and the report of theDefense Science Board 1988 Summer Study on the Industrial Base, 1989. Weapon systcm production rates arc discussed further in Eflkcts of WeuponsProcurement Strek.%otm on Costs and Schedtdes (Washington, DC: U.S. Congrcssicmul Budget Office, 1987).

qq~onard Sul]ivan, Jr., op. cit., footnote 6.


gress, has also attracted considerable attention fromanalysts of the acquisition system. Many criticsdecry what they see as excessive bureaucraticlayering and micromanagement. However, otherspoint out—as did the GAO-that such critics ‘‘failto realize that program managers are responsible forexpenditures involving billions of dollars in publicfunds and that a system of checks and balances isessential.” 44 The level of scrutiny needed to ensurean appropriate level of checks and balances remainscontroversial.

Problems in defense acquisition can be separatedinto a number of categories, including: programvariability (sometimes called program instability);the requirements generation process, including theprocess by which resources are allocated and weap-ons systems are selected; bureaucratic paralysis;inappropriate organization of the defense procure-ment system; and the quality of and incentivestructure facing acquisition personnel.

Program Variability

Sources of Program Variability-Perhaps themost significant difference between defense acquisi-tion programs and commercial activities is thedegree and the unpredictability of year-to-yearchange in defense programs. Constant variationmakes sound management impossible. As a result,studies of the defense acquisition system alwayshighlight variability as a major problem.

Many pressures for changes in defense acquisi-tion programs are peculiar to government procedure,originating from every level of congressional andexecutive branch operation. Other stimuli forchange, shared by both government and privateactivities, are changing threats (or market demands,in the commercial world) and the inherent uncer-tainty of the technology development process. Evenif the changes due to governmental procedurecould somehow be eliminated, these latter sources--which no amount of planning or acquisitionreform can remove-would remain.

A key source of self-imposed change is politics,not in the pejorative sense that the word has acquiredconnoting back-room deals, influence peddling, andpork barreling, but in its original definition as a

struggle between competing interests. Decisions tobuild multi-billion dollar weapon systems do notmerely follow from technical or strategic analyses.They also represent choices concerning the relativeimportance of certain military needs over others, andof those military needs over other public needs (e.g.,housing, health care, economic security, tax relief,and deficit reduction). Finally, these decisions alsoultimately represent commitments to specific manufac-turers employing people and purchasing goods inspecific congressional districts. These are inherentlypolitical decisions, and in the United States, nopolitical decision is final.

The political process involves constant competi-tion and interaction among many different actors:the military Services against one another and againstOSD, DoD against the rest of the executive branch,the executive branch against the Congress, andvarious committees, subcommittees, and Membersof Congress against one another. When the interestsof many of these parties align, differences betweenthem can be resolved. However, in the face offundamental disagreement, the competition for in-fluence and control can make it very difficult tomaintain continuity.

The struggle between Congress and the executivebranch leads to what is generally referred to as‘‘legislative oversight responsibilities” within Con-gress and as ‘‘micromanagement” within the ex-ecutive branch. It results in hundreds of budget lineitem changes and other legislative restrictions andrequirements each year. Although congressionalmodifications to the DoD budget request certainlycomplicate program management, changes gen-erated within the many layers of DoD managementadd significantly to the problem. Many, if not most,of the budget cuts imposed upon or generated withinDoD are due to DoD’s inability to forecast programcosts accurately, to defer new starts until sufficientfunding to cover the actual (rather than the originallyestimated) costs is available, or to eliminate programs—rather than stretch them out—in the event of fundingshortfalls. GAO has found that, although the impactof underfunding programs is ‘‘well-recognized anddocumented, a workable and effective method for

‘W-J.S, General Accounting office, “A Critique of the Performance of the Defense Systems Acquisition Review Council: Billions in Public FundsInvolved,” PSAD-78-14, Jan. 30, 1978, p. i.


matching DoD’s needs with budgetary constraintshas not been developed. ”45

Reducing Program Variability-Although meas-ures can be taken by both Congress and DoD toreduce the number and effects of program changes,changes cannot be eliminated. Analysts disagree asto which of two management failures is the moreserious in the light of unexpected change: failure toplan and budget flexibly, or failure to hold to a fixedschedule. Efforts to reduce program variabilityinclude reforming congressional budget review pro-cedures, multiyear budgeting, program “baselin-ing, ” increasing DoD management flexibility, andreducing personnel turnover.

Reforming the Congressional Budget ReviewProcedure-The current congressional budget proc-ess, involving three levels of review between thebudget committees, the authorizing committees, andthe appropriations committees, takes too long tocomplete. Final decisions on the defense budget aremade by congressional conference committees as (orin many recent cases, after) the new fiscal year starts,late in the executive branch’s preparation of thefollowing fiscal year’s budget. Last-minute changesin the appropriated funding levels require last-minute changes to the next year’s request--changesthat can be difficult to accommodate in a rationalmanner.

Changing the congressional budget processwould require a major revision in congressionalprocedures that would involve either cutting downthe number of committees having a significant roleor sharply delineating committee responsibilities.However, Congress is a highly pluralistic institution,and there is no single individual or organization thatcan mandate these changes. Enacting them wouldtherefore require either widespread agreement wi-thin Congress or the unilateral abdication of author-ity on the part of committees that are now involved.

Even if the number of actors reviewing the budgetis reduced, the structure of that budget may not beoptimally suited for evaluating defense roles andmissions. Congressional review of the defensebudget now deals more with accounting inputs(dollars, personnel slots, buildings, etc.) than with

defense outputs (mission capabilities or strategicgoals). The inputs are easier to count and to control,and unlike defense mission capabilities they permitcomparisons to other programs across the entireFederal Government. However, they also focuscongressional attention on funding for individualprogram elements, whereas many argue that a moreappropriate role for Congress would be a high-levelstrategic review,

Multiyear Budgeting-Lengthening the budgetcycle would provide a longer planning horizon andrequire less frequent congressional review. Congres-sional oversight would be directed more towardsstrategic guidance and away from individual lineitems, offering the hope that programs could enjoygreater stability. Although there are constitutionalrestrictions on appropriations longer than 2 years forcertain military purposes, legislative and executiveprocedures could be changed to permit budgeting in2-year intervals, and program authorizations couldbe even longer-term. However, this approach haslimits, because absolute program stability is funda-mentally incompatible with holding elected officialsaccountable at periodic intervals for their actions.Every time an elected official is replaced, thereis—and must be—the opportunity for the newofficial to change the way things have been done.

Although biennial budgeting was attempted forDoD for fiscal years 1988 and 1989, no funds wereappropriated for 1989 during the 1988 budget cycle.One effect that the experiment did have, however,was to give Congress more visibility into out-yearplans of DoD than it had previously had. Inparticular, for the 1990-1991 budget submission,Congress will for the first time be given access toDoD’s Five Year Defense Plan. Although somemight fear that this visibility would simply giveCongress that much more opportunity to meddle, itis also plausible that improving the communicationbetween Congress and DoD in this manner can helpgive Congress the confidence in DoD planning thatis needed before Congress can relax its level ofoversight and micromanagement. It extends theplanning horizon, enabling both Congress and DoDto take a longer view.

45 LJ.s. (icneralkcoumjng Office, “MajorAequisitions: Summwy of RcwrringProblcmsand Systemic Issues 1960- 1987,” GAO/NSIAD-88- 135BR,September 1988, p. 10. (See also the previous section of this ch~pter on ‘ ‘Affordability.”)


Baselining--A “baseline” is an internal contractbetween a military program manager and the seniormanagement of his or her Service concerning thecost, schedule, and performance milestones for anew weapon system program, Since changes to thebaseline require equally high level review, formaliz-ing a baseline represents an attempt to reduce theamount of change that programs undergo withinDoD. In practice, however, baselining requires thatthe program manager have the authority to rejectchanges to his or her program that are imposed fromsources outside the program. Granting this degree ofauthority is extremely difficult within the presentDoD environment. For example, although specifiedin a program’s baseline, one of the most importantprogram parameters, its budget, is in the finalanalysis established externally. Moreover, it is oftenchanged annually by the Congress. Fully realizingthe benefits of program baselining requires extend-ing it, or some equivalent, to Congress. It alsorequires providing program managers or their supe-riors sufficient authority to resist or accommodatechanges imposed by other DoD organizations, suchas testing and evaluation offices. Changes of thisscope would go against recent congressional initia-tives that strengthen independent auditing and evalu-ation functions within DoD.

Increasing Management Flexibility-Anotherway to reduce the variability of DoD programs is toincrease the Defense Department’s ability to adjustto changing circumstances without requesting con-gressional approval, DoD's ability to accommodatechanges—whether imposed by Congress or result-ing from changing threats or unanticipated techno-logical difficulties-is also limited by the absence ofreserve funds. Unless a means is available foraddressing unforeseen problems quickly, it is oftenimpossible to meet expected costs and maintainschedules.

Although no individual program’s requirementfor such reserves can be predicted, the amount likelyto be needed by a group of programs can bestatistically estimated in aggregate. However, theintense competition for funds within DoD and thedegree of scrutiny applied to defense budgets by

Congress both mitigate against providing reserves.Indeed, in an environment where there are alreadyfar more claims on defense dollars than there areavailable funds, there is every incentive to underesti-mate the costs of programs when Service budgets areprepared. So even if contingency reserves areinitially provided for, they are one of the first itemsto be trimmed. And were management reservessomehow to survive DoD’s internal budget prepara-tion process, they would probably not fare well onCapitol Hill, where they—referred to as “slushfunds’’—are usually eliminated to protect the tax-payer from waste, fraud, and abuse.

Members of Congress, on the other hand, canpoint to instances where they believe DoD has usedinternal fund transfers to evade congressional re-striction or to protect programs that Congress hassought to delay or cancel. Providing DoD withmanagement reserves and raising the thresholds forinternal funding transfers will therefore requireestablishing a relationship of greater credibility andtrust between DoD and Congress.

Personnel Turnover—Another contributor to pro-gram variability is turnover in acquisition personnel.Although typical defense programs have lifetimesmeasured in decades, the average tenure of defenseprogram managers surveyed by GAO in 1986 wasless than 21/z years. Such short tenures make itdifficult to increase the authority of program manag-ers, because they hinder any attempt to assignaccountability. Moreover, short tenures can generatepressures to sacrifice long-term quality for short-term results. (See the section on ‘‘AcquisitionPersonnel” below.)

Requirements Generation andResource Allocation

Weapons systems are procured by “buyingcommands” within the military Services that are notdirectly tied to the Commanders in Chief (CINCs) ofthe operational forces, who (or, more often, whosesuccessors) would have to use those systems incombat.46 As a result, many studies have found thatthe operational users are not sufficiently involved inthe acquisition process, including the establishment

46For operational Puvoscs, [he Arrncd Forces are org~i~cd into military commands that report through hC chairman of lhc Joint Chiefs of Staff ~dthe Sccrctary of Defense to the President. The military Scrviccs thcmsclvcs, each headed by a civilian Sccrctary, arc responsible for training and equippingmilitary forces, but not for commanding them operationally.


of the military requirements that initiate new acquisi-tion programs. Furthermore, requirements, whenestablished, tend to be observed rigidly rather thanbeing reexamined in light of new circumstancessuch as schedule and cost overruns. While changingthe requirements too frequently does exacerbateprogram variability, as described above, unwilling-ness to make changes in light of changing circum-stances can force cost, schedule, and complexityupwards.

Requirements tend to be overstated due toinsufficient interaction between those who knowwhat is needed and those who know how to provideit. Further pressure for exaggerating military re-quirements stems from the process by which DoDdecides to develop new systems. This process isconducted essentially in two stages. Once a militaryrequirement has been established, funds to meet thatrequirement must be found in a highly competitiveand political environment involving the militaryService, OSD, the Office of Management andBudget, and Congress. After funds are reserved, asecond stage of competition selects the actualsupplier.

The funding competition imposes great pressureto over promise capability while underestimatingcost; there are few incentives to enforce realism.Specific designs cannot be offered at this stagebecause they would interfere with the ensuing sourceselection competition. Program managers, whoshould be in the best position to weigh the militaryrequirements for a system against the technologicalprospects for satisfying those requirements, aregenerally brought into the acquisition process toolate to have a significant impact on requirementsgeneration. It also appears that they are too seldomable to modify requirements in response to subse-quent events.

A recent study aimed at monitoring the reorgani-zation of DoD, which drew upon the PackardCommission’s recommendations and the Goldwater-Nichols Defense Reorganization Act, noted progressin enabling trade-offs to be made between require-ments, cost, and schedule and in taking affordabilitymore seriously. “The organizations and proceduresthat could make possible such a change [in acquisi-

tion procedures] have been set up,” the studyconcluded, but “their effective operation will re-quire continued high-level attention.”47

Bureaucratic Paralysis

Causes and Effects—A constant complaint ofthose involved throughout the defense acquisitionprocess concerns the increasing bureaucratic burdenthey must struggle through in order to do their jobs.This bureaucracy manifests itself in multiple levelsof approval, the diffusion of responsibility andauthority, the lack of individual accountability, andthe profusion of auditors, inspectors, specifications,and regulations. It is blamed for causing excessivedelay, stifling innovation, suppressing initiative, andincreasing costs.

Although these perceptions are widespread, theyare difficult to validate objectively. Analysis at-tempting to quantify trends in regulatory activityfound some indicators that showed increases andothers that did not. The effects of governmentalbureaucracy and regulation are even harder tomeasure than the trends in regulatory activity, giventhe absence of a standard for comparison. Althoughprivate sector activities are often held out as modelsfor defense acquisition, it is not clear how relevantthey are to government operations (see sections on“Comparison With the Private Sector” and “Effi-ciency v. Effectiveness” above). Therefore, theusefulness of studies measuring how much lengthieror more expensive government programs are than‘‘equivalent” private sector ones is limited. Givenall the uncertainties and difficulties of estimating thecost penalty imposed on defense acquisition byexisting defense acquisition regulations, it is notsurprising that such estimates are widely divergent.They range from a few percent to more than 50percent.

A simple model of the cost of excessiveregulation is shown in figure 9. With minimalregulation or oversight, the government is dependentupon the goodwill of contractors and public offi-cials. Honest officials and corporations could oper-ate very efficiently in this regime, but dishonest oneswould take advantage of the lack of oversight todefraud the government. At the other end of the

dTH~o]d Brown and Jamcs Schlesinger, co-chairmen, ‘‘ M,aking Dcfcn.sc Reform Work: The Project on Monitoring Defense Reorganization.” a jointproject of the Johns Hopkins Foreign Policy Institute and the Center for Strategic and lnterna[ional Studies, Washington DC, November 1988, p. 49.

95-677 - 89 - 6


spectrum, tight regulatory controls deter or detectthose defrauding the government, but they also driveup the cost of doing business for everyone else.

Much of the political debate concerning “waste,fraud, and abuse” concerns where on this curve thedefense procurement system lies. Analyses of de-fense procurement consistently indicate that thesystem lies somewhere on the side of excessiveregulation, at least in terms of strictly economicconsiderations. However, the public—to which Con-gress responds very effectively—may well believethat the system is not yet regulated enough, espe-cially in the wake of recent reports of procurementscandals and defense contractor fraud. It may be thatthe costs of imposing stricter controls are not wellunderstood by the public, and that if these costs weremore widely recognized, calls for additional regula-tion would be moderated. However, it is alsoconceivable that the American taxpayer prefers topay the high costs of overregulation rather thanpermit even lesser amounts of public money to gounearned into someone’s pocket. If public demandsfor overregulation can be thought of as a source ofavoidable waste, then perhaps some waste must beconsidered the price of curbing fraud and abuse.

Some argue that the present approach of legislat-ing strict oversight and accountability requirementshas the effect of penalizing everyone in defenseacquisition instead of just those individuals who aretruly guilty of violations of ethics or law. Onealternative system, according to this line of reason-ing, would be one in which people are trusted to becapable of doing their jobs without intrusive over-sight and indeed are allowed to do so. However,those found guilty of violating this trust would bepunished severely. While the relaxed oversightmight reduce the probability of detecting illegal orunethical activities, those actions could neverthelessbe deterred by the increased severity of the punish-ment if caught. This approach would replace thecurrent adversarial relationship between govern-ment and industry with a more collaborative one.

Reducing Paperwork and Bureaucracy— Meas-ures to cut red tape or streamline the bureaucracywill fail unless they take into account the reasonswhy the bureaucracy was initially established. Regu-

Figure 9-Cost v. Regulatory Intensity

Cost rises dueCost rises due to bureaucraticto fraud i n e f f i c i e n c y

Degree of regulation


lations and guidelines are a means of preservinginstitutional memory in an environment wherepresidential appointees have a median length ofservice of just over 2 years48 and where militarypersonnel are regularly rotated. They incorporate thepolitical oversight and review procedures that comewith the expenditure of public funds. They codifymanagement procedures for large and unwieldyorganizations. Finally, regulations and guidelinesfurther important policy objectives that may bein the nation’s or DoD’s collective best interesteven though they might interfere with the mostefficient execution of individual programs. As hasbeen stated before, the government has manygoals—environmental protection, occupationalhealth and safety, fair labor practices, equal opportu-nity, etc.—that may conflict with any individualprogram manager’s ability to run a program effi-ciently. Just because a program manager does notbelieve his or her program should be the vehicle toimplement national policy does not mean that thatpolicy should be ignored. Although regulations have

A~Natj~al Academy of Public Administration, ‘‘Leadership in Jeopardy,’” November 1985, p. 4. (This figure applies to the entire FederalGovernment.)

Chapter 8--Lab to Field: Why So Long? ● 153

been criticized as attempts to solve yesterday’sproblems by impeding today’s progress, those prob-lems are certain to reappear in the absence of somemeans of institutionalizing the lessons learned. Inother words, much of the bureaucracy andregulation surrounding defense acquisition hasresulted from the political environment—reflected in public opinion and in legislation—within which defense acquisition is done.

Studies such as that of the Packard Commissionhave recommended changes in the DoD bureaucracythat would have the effect of delegating authority tolower levels. Program managers and their immediatesuperiors would be freer to do their jobs, and theadvocates for interests such as competition, smallbusiness, equal opportunity, testing and evaluation,etc., would be relegated to advisory roles. Inparticular, the Packard Commission recommendedsetting up a streamlined acquisition chain of com-mand in which program managers would reportthrough no more than two levels of command to thesenior procurement executive in DoD (the UnderSecretary for Acquisition). Much of this structurehas now been established. However, the newstructure supplements-and does not replace—the existing chains of authority and command.According to the study monitoring implementationof the Packard Commission recommendations:

. . . the purposes of the legislation [implementingsome of the recommendations] have not been met.Our sense is that the new positions were simplysuperimposed on top of the existing structure.49

The new acquisition chain is at present a commu-nications link, and does not control funds. Trulyimplementing the Packard Commission’s recom-mendations would require substantial changes in theoperation of DoD.

The Packard Commission and other studies heldout certain programs within DoD-in particular,highly classified “special access” or “black”programs50 and high-priority strategic programs—asmodels that have successfully conquered the DoDbureaucracy. Special access programs, due to ex-treme security requirements, bypass much of the

review, approval, and bureaucracy that ordinaryprograms must contend with. However, those samesecurity constraints make these programs difficult toanalyze in general.

Some officials with extensive experience in bothspecial access and ordinary program managementsay that the approaches used in the special accessworld enable equipment to be fielded much morequickly, and at lower cost, than do standard acquisi-tion programs. Other officials say that the highpriority, high-level review, and high-quality staffs ofspecial access programs, more than their manage-ment techniques, are responsible for their success.Moreover, they point out that bypassing checks andreviews—although sometimes necessary in the nameof security-adds considerable risk. While some saythat extension of special access contracting proce-dures would improve acquisition, others say it couldnot provide a general solution.

A number of different approaches can be taken toreduce bureaucracy and regulation within DoD.Implementing any of them, however, presumes anatmosphere of trust among DoD, the rest of theexecutive branch, and Congress; many reformsrequire the same degree of trust to hold betweenDoD and the defense industry,

Major Legislative and Administrative Reform-One approach would be to replace the existingstatutory and administrative framework, in whichfraud and abuse are deterred by extensive reportingand auditing requirements, with one in which greaterresponsibility is placed on voluntary compliancecoupled with vigorous enforcement and severepunishment for those who get caught breaking a law.Enacting such a system would involve a majoroverhaul of the existing defense acquisition systemand the environment in which it operates. Moreover,it would require (and also follow from) reducingwhat many in government and industry see as theexisting adversarial relationship between the two.

Bottom-Up Review The opposite approach is tostart with the present system but examine eachregulation, directive, and specification to ensure that

@“ M~ing Defense Reform Work: The Projwt on Monitoring Defense Reorganization, ” op. cit., fOOtllOte 47, p. 50.s~echnically, DOD does not ~= the tem L‘b]ack”progm$A‘‘specia] access program” is onc in which additionat restrictions beyond hose available

through the normal Confidential-Secret-Top Secret classification system are deemed required. The budgets and existence of such programs mayor maynot bc classified. “Black” programs generally refer to those whose existence is kcpl secret.


it is still relevant and appropriate. Such a task wouldbe a mammoth undertaking. Moreover, those withthe time to review the regulations would most likelynot be the ones adversely affected by them, and it isunlikely that this approach would effect significantchange.

Evolutionary Review-Another approach, whichis being implemented in a number of DoD activities,is to establish a mechanism by which those ad-versely affected by a regulation can petition for itswaiver, with all such petitions examined to seewhich regulations should be waived or modifiedacross the board.

Those affected by a regulation can already seekits waiver from the issuing authority without anyspecial program. However, DoD programs such asthe Pilot Contracting Activities Program and theModel Installations Program have been establishedto provide waivers in a more systematic way.Requests for waivers are tabulated to identify thoseregulations that seem to provide the greatest barriers.If approved, waivers are evaluated on an experimen-tal basis to see if they should be made permanent oreven extended DoD-wide. These programs at pre-sent cannot waive regulations imposed externally onDoD (by legislation, for example). However, theycan identify those external regulations and laws thatparticipants find to be particularly onerous, and theDoD can then propose legislation or regulatoryreform at higher levels to ameliorate the problem.

One drawback to this approach, from the point ofview of those seeking major reform, is that it is anevolutionary process. Another problem is that in atleast some cases, individuals having to put up withobsolete, ineffective, or inapplicable regulationshave found it far easier to ignore them than topetition for their waiver.

Shifting the ‘ ‘Burden of Proof ‘—In this approach,the “burden of proof” is shifted from those seekingto waive regulations to those seeking to enforcethem against the objections of the program manager.Essentially, it consists of pre-delegating waiverauthority all the way down to the program manager,who could decide which regulations are appropriate.The “special interests” and “advocates” wouldstill exist and would still be free to make recommen-dations to the program manager. However, theprogram manager would be free to disregard their

advice, unless they were able to persuade theprogram manager’s superiors. To the extent thatregulations are simply being ignored today, asdescribed in the preceding paragraph, this approachis in essence being taken now—without officialsanction.

Such a system could only work if programmanagers and their superiors were evaluated notonly on how well individual programs fared but alsoon how well the programs, on balance, supported theintent of the regulations—which, after all, serve toincorporate DoD and national policies that seniorpolicymakers have decided are important. Programmanagers would have to realize that their goal is notsimply development and deployment of a weaponsystem but furthering national policy as well.

True implementation of this approach would alsorequire congressional action to relax statutory con-straints, since those could not be waived by programmanagers. Moreover, the problem of identifying theessential core of laws and regulations that wouldremain mandatory-ineligible for waiver at theprogram manager’s discretion—re-creates the origi-nal problem. If it were easy to identify the irreduc-ible core of regulations and laws in the first place,this approach would not be necessary.

Organization of the Defense Acquisition System

So far this chapter has discussed acquisitionprocedures within the existing organization, inwhich OSD establishes policy and participates inmilestone reviews for major programs, but acquisi-tion is executed (and for programs other than themajor ones, reviewed) by the Services. However,there are other organizational models, ranging fromgiving the USD(A) the acquisition authority thatpresently rests within the Services all the way tocreating a civilian acquisition agency outside DoD.

Most studies of defense acquisition argue that themilitary Services must have primary responsibilityfor acquisition to ensure that the needs of theoperational user are met. However, some civiliananalysts argue that much of what goes on inmanaging acquisition programs does not require,and may not even be greatly aided by, militarycontrol. They argue that the professional, stable, andhighly trained acquisition work force needed toimplement procurement reform can be created only


in the context of a civilian acquisition agency. Evenso, few proponents of a civilian agency call for it tobe outside DoD; most believe that the Secretary ofDefense must be responsible for national resourcesdevoted to defense.

Although a study of European nations that usecentralized procurement systems might illuminatethe successes or failures of such a plan, there aresignificant factors that make such an analysisdifficult. One important difference is that Europeandefense programs are small compared to that of theUnited States. There are other differences, too:European military services do not dominate acquisi-tion, European defense plans are done on a multiyearbasis, the legislatures make minimal changes toannual defense procurement budgets, the govern-ment imposes minimal “how-to” requirements onthe defense industry, and industrial policy is a majorconsideration in defense contracting.

One compromise position, adopted by the studygroup that examined the implementation of thePackard Commission recommendations, would beto encourage each of the Services to create aspecialized “acquisition corps, ” but to considercreating an independent acquisition organizationunder the USD(A) in the event that the Servicesbalk. 51 Although the study stated a preference forleaving acquisition authority with the Services, itwent on to conclude that ‘‘radical steps, such as theestablishment of a single procurement organizationwithin the Department, should not permanently beruled out. “52

Acquisition Personnel

Improving defense acquisition depends on ahigh-quality, stable, and well-trained acquisitionworkforce. In a letter to President Reagan one yearafter the publication of the Packard Commissionreport, David Packard stated that:

Personnel policy is the keystone of virtually all ofthese reforms. With able people operating them,

even second-rate organizational structures and pro-cedures can be made to work; and without ablepeople, even first-rate ones will fail.53

Improvements recommended by the Packard Com-mission included reducing the barriers to recruitingsenior-level executive branch personnel,54 attractingqualified new personnel and improving the trainingand motivation of existing personnel at the middlemanagement levels, and continuing the recent im-provements in defining military career paths inacquisition, Members of the Commission thoughtthat civilian acquisition personnel needed muchmore attention than military personnel, and theirreport cited many of the deficiencies of the federalCivil Service system that are described in the contextof national laboratory personnel in chapter 5.

As was noted in the previous section, the ‘Projecton Monitoring Defense Reorganization” recom-mended establishing within each of the militaryServices a professional ‘‘acquisition corps. ” Withinthese corps, military officers who wished to special-ize in acquisition would be able to pursue a careerpath that did not constantly rotate them out ofacquisition billets. They would also receive thetraining necessary to do their jobs and compensationcomparable to their private sector peers. Officerswith operational experience would still be assignedto acquisition jobs, but in fewer numbers than now.Although the Services have long resisted establish-ing such corps, the study concluded that the in-creased professionalism that this approach wouldbring is essential for effective and efficient acquisi-t i o n .5 5

All proposals for reforming acquisition personnelpolicy run into conflicts among competing objec-tives. Creating a military acquisition corps couldimprove acquisition but it would also create amilitary career path unlike any that the Services nowbelieve to be appropriate. Making fundamentalreforms to Civil Service procedures—or even ex-empting significant groups from them—would pose

51‘ * M&ing ~fense Ref~rm Work: ne project on Monitoring Defcmsc Reorganization,” op. C1l., fOOmOte 47, p. 59.521 bid., p. 51.sJDavid packard, letter to the president of the United States, July lo, 1987; cited by J. Ronald Fox and James L. Field, The Defense ~aW?ement

Challenge: Weapons Acquisition (Boston, MA: Harvard Business School Press, 1988), p. 315.sqAnlong tic changes specified Were slmp]ifylng finan~ia] di,sclosurc forms and allowing appointees 10 defer capital gains tax liability incurred in

divesting assets so as to satisfy conflict-of-interest provisions.55’ ‘M&ing ~fensc Rcfom Work: me project on Monitoring ~fcnsc Rcorg~ni~ation, ” op. cit., ~“oomotc 47, p. 59.


substantial political difficulties. Federal employeesalready feel as if they have 240 million supervisors,and it sometimes appears, at least while reading“Letters to the Editor” columns whenever civilianpay raises are debated in Congress, that there isnothing so despised as a civil servant. Proposals thatwould increase compensation or other benefits ofFederal employment in an effort to attract moresenior and more highly qualified employees wouldbe seen by many as adding slots to the Federaltrough.

Conflict-of-interest regulations provide a case inpoint. Many individuals with experience in defenseacquisition argue that ‘‘revolving door” legislationthat erects barriers to the interchange of individualsbetween government and industry prevents skilledindividuals with hands-on technical or managerialexperience in the industrial world from contributingtheir skills to DoD. On the other hand, a significantsegment of public opinion—shared by a significantsegment of Congress—sees the interchange ofindividuals between government and industry asproviding inherent conflicts of interest. The politicalreconciliation of these two points of view will bedifficult.

Policy Options

Trade-offs, inefficiencies and problems in thedefense acquisition system stem from a wide varietyof interrelated causes. Some of them are due tostructural limitations of the United States Constitu-tion and our resulting political system. Others resultfrom the relationship between the Congress andDoD, and would be amenable to congressionalaction or clarification. Many problems arise fromconscious choices that have been made to emphasizesome national goals over others, choices that couldbe reversed if the political mood of the nation wereto shift. And still others are unintended conse-quences of aggregating many individual actions,each of which may be widely accepted.

Solutions almost always involve trade-offs.Should the government relax its controls overindustrial performance, or should they be strength-ened? One point of view is that of the PackardCommission, which believed that although majorimprovements were essential, “self-governance isthe most promising mechanism to foster improvedcontract compliance. ”56 Quite a different viewpointis provided by the Project on Military Procurement,which argued that “as expensive as it is to hirelegions of auditors, it is even more expensive toallow contractors to continue to steal and goofoff. ”57 Although this picture of contractor behavioris not supported by analyses of defense procurement—which generally find that fraud, while certainlypresent to some extent, does not consume a signifi-cant fraction of the defense budget—it does repre-sent the attitude of a substantial fraction of taxpayersand therefore of Congress.58 Regardless of its merit,reformers of the defense acquisition system ignorethis public sentiment at their peril.

Has the overhead that comes with governmentprocurement (viz., accountability trails and socio-economic goals) impeded defense procurement sobadly that we should be willing to trade off thesegoals to obtain a more efficient system? Is the risk ofa visible and, in hindsight, preventable failure worsethan the risk of quashing individual initiative byimposing regulations? Are we willing to assignindividual accountability and responsibility, know-ing that the price of allowing star performers to excelis the risk that incompetent and even criminalactions may take place as well? What are the costsof delaying a military capability v. the benefits ofdelaying an expenditure? These are difficult butcrucial questions.

Incentives

The best, and possibly the only, solutions toacquisition problems involve changing the incentivestructure facing people and organizations, ratherthan imposing additional regulations. The present

sb~esiden~’s Blue Ribbon Commission on Defense Management, op. cit., footnote 15. P. 84.

57D~a Rasor, et d. Defense procurement Papers: Campaign ’88 (Washington, DC: Project on Military procurement, September 1988), P. 45.58A public opinion smey of 1,500” Americ~s t~en for tie pa~k~d C(Jnlmi\~ion fo~d that the public believes $45 of each $100” in thc defense budget

goes to wrote (poor management) and fraud (illegal activities), with that $45 about evenly split between the two. Money lost through waste and fraudis thought to end up primarily in defense contractors and individuals’ pockets. See Appendix L, “U.S. National Survey: Public Atti[udcs on DefenseManagement,” prepared by Market Opinion Research; in A Questjor Exce//ence: Appendix, Final Report by the Prcsidcm Blue Ribbon Commissionon Defense Management, Exccutivc Office of the White House, Washington, DC, June 1986, pp. 217, 219.

Chapter 8--Lab to Field: Why So Long? ● 157

acquisition system offers incentives, according to J.Ronald Fox and James Fields, but these often act thewrong way. They argue that no lasting improvementis likely unless an appropriate system of incentivesand disincentives is formulated and enforced:

Unless changes are made in the contractor sourceselection process, which makes optimistically lowcost estimates a significant advantage in competingfor a contract, it is useless to discuss realisticcontractor proposals. The source selection processmust give far more weight to realistic cost estimatesand the contractor’s record of past performance.

Unless changes are made in the current profitsystem that demands higher costs as a prerequisitefor higher profits, it is futile to expect lower costs.Because profits are largely based on cost, there islittle economic motivation for contractors to reducedirect or indirect costs. The profit system needs amajor overhaul to relate profits more to contractperformance than to the level of costs.

Unless changes are made in the current militarypersonnel system that makes short-term assignmentsnecessary for military officers to acquire the numberand variety of assignments required for promotion,any significant reduction in personnel turnover indefense program offices is unlikely.

Unless changes are made in the current OSD andcongressional practice of routinely accepting pro-gram stretch-outs as a tactic for funding newprograms, it is unrealistic to advocate economicalproduction rates.

Unless changes are made in the current DoDpractice of waiving training requirements and offer-ing only short training courses, which limit coverageto introductory rather than in-depth treatment ofimportant subjects, it is unrealistic to expect im-proved training for acquisition managers.

Unless changes are made in military careers thatcurrently provide few opportunities beyond age 45or 50, it is unrealistic to expect military officers notto seek a second career in the defense industry. Inaddressing this problem, DoD needs to listen tolieutenant colonels and colonels and Navy com-manders and captains to learn their views on theadvantages and disadvantages of the acquisitioncareer field.

Without genuine promotion opportunities forthose who make the difficult decisions associated

with successful negotiating and wise buying, it isunrealistic to expect to retain in government serviceexperienced program managers able to do muchmore than promote their programs, prepare progressreports, and conduct briefings.59

Approaches

Although there are innumerable specific changesthat could be made to the way defense acquisition isdone, the discussion in this chapter suggests thatmany policy choices concerning defense acquisitionfundamentally rest on where the balance is estab-lished between efficient defense acquisition, on theone hand, and furthering national goals such asfairness and accountability, on the other. Fouralternate approaches, each establishing this balancein a different way, are presented below. In Approach1, the balance is tilted sharply towards efficientacquisition. Approach 2 has the same objectives asApproach 1 but pursues them in a more gradualmanner. Approach 3, ratifies the present choice ofthat balance, and Approach 4 takes the position thatnon-defense-related national objectives should beemphasized even more than they are today.

Approach 1: Enact major structural andlegislative reforms of the environment withinwhich defense acquisition takes place, emphasizingefficient procurement over other national goals.

Selecting this approach would represent a conclu-sion that the existing procurement system places toomuch emphasis on non-procurement-related objec-tives. Many laws and regulations mandating, forexample, procedures for competitive bid solicitationand award, barriers to conflict of interest, and thepromotion of minority-owned and small business,would have to be reviewed and revised to giveindividual contracting officers and program manag-ers greater authority to do as they see fit for the goodof their programs. They would be less involved withjustifying every action, establishing audit trails,complying with accounting standards, and fosteringfull and open competition. Instead, the system wouldrely more on individual responsibility, which wouldhave to be measured both by the success of theprograms and by the necessarily subjective evalu-ation of their manager’s superiors as to how well

S9J. Ronald FOX Wd J~cs L. Field, op. cit., fOOmOte 53, pp. S 1~-319.


they protected the public trust.60 Careful studywould be needed to arrive at a new balance betweenacquisition and these other goals. Successful accom-plishment of this approach would also require somedegree of societal consensus, and the continuingcooperation of Congress, so that the balance be-tween procurement and these other goals would notbe reexamined every budget cycle.

Approach 2: Preserve the basic structure ofthe defense acquisition system, but pursue evolu-tionary changes that would emphasize efficientacquisition over other goals.

This approach is similar in underlying intent tothe preceding one, but would not try to do everythingat once. To the extent that individual regulations,laws, policies, or procedures could be shown toimpede acquisition efficiency, the regulation, law,etc., could be individually evaluated to see whatwould be lost if it were changed. By proceeding ata much more measured pace than the previousapproach, some would argue that it has a much betterchance of being implemented and would pose lessrisk On the other hand, those totally dissatisfiedwith the present acquisition system would probablyfind anything less than a total overhaul insufficient.

Approach 3: Decide that the current balancebetween efficient acquisition and other nationalgoals is more or less appropriate, and in so doingrecognize that acquisition will not be as efficientor as effective as it would be if it were conductedin isolation from those other goals.

This approach, too, would seek evolutionaryimprovement to the acquisition system when spe-

cific impediments could be identified. However, itwould not presume that, given a conflict betweenacquisition and other values, acquisition shouldnecessarily win. Selecting this approach essentiallycodifies the status quo--it would be assumed thatthe environment surrounding the acquisition processis shaped by a compromise between competinginterests and has led to the creation of a system thatperhaps pleases no one, but is preferable to anysignificant alternative.

Approach 4: Extend regulation and oversightof the defense acquisition system under thepremise that it is not yet sufficiently responsive tonational needs.

Those who see recent press accounts of procure-ment improprieties and contractor fraud as indicat-ing a lack of supervision and oversight wouldrecommend an approach diametrically opposed tothose discussed above. instead of favoring acquisi-tion efficiency over other objectives, they wouldseek changes—such as more stringent accountingrequirements and conflict-of-interest standards—that would have the effect of increasing the bureau-cratic overhead of the acquisition process. Propo-nents of this approach argue that more stringentregulation would save taxpayers funds on net—i.e.,that although the extra controls might cost money toestablish, they would prevent an even greateramount of fraud and abuse, or that the price of thefurther standards is worth paying to ensure publicconfidence in the acquisition process.

~his Iattercriterion is not easy to memure. Is it good enough for a contracting officer to consistently award contracts to companies who do good work?What if he or she then retires to lake a high-paying job with one of the companies that he or she had favored? What if another company, which had rea..onto believe that it could have done a job better or cheaper or both, was not allowed to bid’! In both these cases, the appearance that the most efficient useof taxpayer funds may not have been made must be considcrcd, even il” the reality is that the product obtained was as good as any.

Chapter 9

Civilian Technology andMilitary Security

PageMILITARY ACCESS TO CIVILIAN TECHNOLOGY . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161HEALTH OF THE DUAL-USE INDUSTRIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164

Fiber Optics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165Polymer Matrix Composites . . . +. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167

CONVERGENCE OF MILITARY AND CIVILIAN TECHNOLOGIES . . . . . . . . . . . . 168Fiber Optics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Polymer Matrix Composites . . . . . . . . . . . . . . . . . . . . . . . . . . .

BARRIERS BETWEEN THE CIVILIAN AND MILITARYFiber Optics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Polymer Matrix Composites . . . . . . . . . . . . . . . . . . . . . . . . . . .

SUMMARY OF CASE STUDIES . . . . . . . . . . . . . . . . . . . . . . . .FINDINGS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Overall Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Findings Relevant to the Department of Defense . +.. . . . . .Findings Relevant to the Defense Industrial Sector .....,.Findings Relevant to Congress . . . . . . . . . . . . . . . . . . . . . . . . .Findings Relevant to Civilian Industry . . . + . . . . . . . . . . . . . .

POLICY CONSIDERATIONS FOR CONGRESS . . . . . ..+ ,Military Dependence on Foreign Technology ., ...., ... , .Institutional and Administrative Barriers . . . . . . . . . . . . . . . .

Figure

10. Military Dependence11. Military Dependence

onon

ForeignForeign

Industry-OwnershiTechnology Located

. . . . , . . . . . . . . . , , . . . . . . . . , . 177

. . . . . . . . . . . . . . . . . . . . . . . . 178

. . . . . . . . . . . . . . . . . . . . . . . . 178

. . . . . . . . . . . . . . . . . . . . . . . 178

. . . . . . . . . . ... . . . . . . . . . 183

‘in Foreign Countries . . .. 182

Chapter 9

Civilian Technology and Military Security

MILITARY ACCESS TO CIVILIANTECHNOLOGY

This assessment was prompted by a concern onthe part of the Congress that the defense technologybase might be eroding in the United States. As aresult, the Armed Services might be unable to retaina technological advantage over the Soviet Union andother possible adversaries in the future. This concernis closely related to two observations. First, certainhigh-technology industries-such as semiconduc-tors and numerically controlled machine tools—have lost domestic market share to foreign competi-tion. Some appear to have entered a cycle ofdecreased capitalization, weakened innovation, fur-ther loss of market share, and eventual loss ofleading-edge capabilities, in both design and processtechnologies. As a result, U.S. military dependenceon foreign civilian technology is increasing. Thesecond observation is that the U.S. military appearsless and less able to acquire the leading-edgetechnology that does exist. Such technologies maybe available in the civilian sector of the domesticeconomy, but they are somehow beyond the reach ofthe Department of Defense (DoD).

Because of these concerns, OTA conducted policy-oriented case studies of three dual-use1 technologies—fiber optics, software, and advanced polymer matrixcomposites (PMCs)—to assess the availability ofcivilian technology for military purposes and toanalyze difficulties in the transition of technologybetween the civilian and military sectors of the U.S.economy. The case studies avoid extended technicaldescription in favor of policy-relevant analysis ofmajor issues confronting these high technologyindustries. For each case, the inquiry has addressedthree central questions: 1) Are civilian high-

technology industries (those critical to the mili-tary) eroding in the United States? 2) Do militarytechnologies and their applications diverge sig-nificantly from their counterparts in the civiliansector of the economy? 3) What are the principalbarriers, both technical and institutional, thatinhibit military access to civilian technology andvice versa? Each of these questions requires someamplification.

The rationale for the first question depends on theextent and nature of military dependence on civiliantechnology, both domestic and foreign. If it ispossible and desirable for the defense industrialsector to develop and produce all technologiesnecessary for the Services, then the competitivenessof the civilian sector of the economy, while impor-tant for other reasons, would not be related tonational security considerations, But if the defenseindustrial sector relies to any substantial extent onresearch and development (R&D), innovation, andproduction conducted in the civilian sector, then lossof commercial competitiveness in industries thatdevelop technologies that are pervasive or enablingfor military systems would have serious policyimplications. A recent report on defense industrialcompetitiveness by the Under Secretary of Defensefor Acquisition took this position.2 It was also theprimary concern of the Defense Science Board(DSB) when it recommended an industrial policy,which amounted to a strategy for economic defenseof the domestic semiconductor industry.3 In the caseof semiconductors, the DSB argued in its report thateroding domestic capacity seriously jeopardizes ourability to build and field major weapons systems—unless the United States is willing to depend onforeign firms for strategically important technolo-gies, materials, and devices in the future.

I ]n this chapter the term ‘‘dua]-u~” refers to technologiestha[ can have multiple, significant applications to military systems, ~d that can bc employcdextensively in civilian industry as well.

z’ ‘As a nation and as a continent, we no longer are total]y sc]f+ufficicn[ in all essential materials or industries required to maintain a strOng nationaldcfcnsc . . . . Clearly, the Department of Defense cannot provide massive financial assistance for every American industry characterized by a lack ofInternational competitiveness, nor can we effectively prowdc incentives Ii)r every manufacturing industry critical to our dcfcnsc. ” (Department ofDcfcrwe, ‘‘ Bolstering Defense Industrial Competitiveness, ” Report to the !$wrctar-y of Defense by the Under Sccrctary of Defense for Acquisition, July1988, p. v.)

~Dcfensc Scicncc Board, “Rcpo~ of wc Defense Science Board Task Force on D“efense Semiconductor Dependency, ” prepared for lhc Office of theUnder Sccrctary of” Defense for Acquisition, February 1987.

–161–


These concerns have equally significant impli-cations for emerging and fast-moving, high-technology industries such as fiber optics, software,and advanced composites. There is substantialevidence and widespread agreement among experts,for example, that the Japanese government andindustry are already deeply engaged in the process offunding, researching, planning, and designing market-oriented applications for a wide range of fiber opticand photonic technologies. By contrast, many Ameri-can firms appear to be waiting in the wings, bidingtheir time, hoping for Federal assistance, and unwill-ing to commit substantial corporate funds forproducts that might not reach the market for 10years. If fiber optic and photonic technologies doindeed supplant large portions of the consumerelectronics and computer industries-as some ana-lysts argue they will—the implications for thecivilian side of the U.S. economy and its ability tosupply the military will be enormous. For thisreason, the case studies, which are located in aseparate volume of appendies to this report, alladdress the question of the health of domesticindustry, specifically analyzing the threat of destruc-tive foreign competition now and in the future.

The purpose of the second question -concerningdivergence and convergence in dual-use technology—is to make explicit the differences and similaritiesbetween military and civilian technologies and theirvarious applications. As technological divergencebetween the two sectors increases, it becomes moredifficult for the military to draw on the resources ofthe civilian sector of the economy. Absolute diver-gence between military and civilian technologieswould mean that by far the largest portion of thetechnology and industrial bases in the United Statesand around the world would be unavailable to themilitary. This is clearly not the case. Indeed, a basicstrength of the West is that technological innovationis rooted in society and not in the military. Neverthe-less, there is a widespread belief among manydefense planners and technologists that military andcivilian technologies are inherently different, be-cause weapons systems must push the outsideenvelope of performance, must be built to sustainbattlefield environments, and are expected to sur-

vive up to several decades of readiness and trainingmissions.

Critics of this position contend that it is possibleto plan for convergence of military and civiliantechnologies, even in early development, by harmo-nizing otherwise divergent standards and specifica-tions. They believe that much divergence is theresult of a military fixation on achieving technicalperformance levels that may or may not be decisivein warfare or desirable from a training and mainte-nance perspective. They believe that civilian prod-ucts can be as rugged as those built to militaryspecification. There are, for example, few environ-ments more hostile than the one under the hood of acar, where semiconductor devices are hard-mountedto the engine block.

Many observers agree that in any dual-use indus-try there will be strong convergence between themilitary and civilian technologies themselves, andpotentially significant divergence when it comes toend-use. Clearly, there are military applications thatwill never find expression in civilian life, but at thesame time it appears to be possible to substitutecommercial off-the-shelf (COTS) products for Mil-spec items4 in some of the most advanced weaponssystems. But the relationship between civilian andmilitary technologies is far more complex thanarguments about the adoption of finished pieces ofhardware can suggest. It is evident that civiliantechnology that could be important to the ArmedForces will be irrelevant if military officials—fromgeneral officers to program managers—reject suit-able civilian parts in favor of items developed undercontract with the government, And it is likely thatService acquisition managers will continue to resistcivilian technology and components until the incen-tive system is changed. At present, many managersavoid new technologies available in the civiliansector, because the acquisition system forces them toassume responsibility when such a part or compo-nent fails, thereby jeopardizing their careers. Mostprefer to use older technologies, waiting until thespecifications, standards, research, and testing arecomplete—and the risk of failure is eliminated orcan be transferred to another program.

4M11sPc itcms MC Ptis (Jr systems fomally specified in the kfy of ~!i]it~ry Spwifications and St~dwds published by the Dep~ment of Defense.

Chapter 9-Civilian Technology and Military Security ● 163

In the sphere of complex, information-intensivemilitary systems—whether involving weapons ornot—the utility of civilian technology and applica-tions must be addressed case by case. For each of thecases in this report, an effort was made to analyze theextent of convergence and divergence of civilian andmilitary technology, and to explore existing andpotential areas of overlap. The case studies attemptto sort out the extent to which divergence istechnologically necessary and the extent to which itis a product of military culture as well as economicincentives for the contract industries on which theServices depend.

The final question asked earlier, regarding im-pediments to military use of civilian technology,requires an analysis of the barriers to the fluidexchange of technology between the military and thecivilian economies. When OTA began to investigatequestions of dual-use technology, complementarycomplaints surfaced from within DoD and fromcivilian sector managers. Military technologistsasserted that while there were many advanced anddesirable technologies in the civil sector, the mili-tary encountered difficulties in gaining access tosome of them. In addition, senior defense officialsindicated that some military contractors had diversi-fied into the civilian sector of the economy with theintention of getting out of the defense businessaltogether. These officials cited a need to reduce redtape and to make doing business with DoD morestable and predictable. The other side of the complaint—which first emerged from managers of small, en-trepreneurial firms-was that doing business withthe government involved difficult adjustments, in-cluding compliance with outdated military specifi-cations, cumbersome auditing and reporting proce-dures, and possible compromise of proprietaryinformation. They considered DoD to be a verydifficult customer, often too costly to pursue forcommercial purposes. The bottom line is that DoDmay have to buy advanced technology from compa-nies that do not need, or even particularly want, itsbusiness.

Accordingly, it is important to investigate theprocesses by which specific technologies movebetween the civilian and military sectors. Theexistence of significant barriers-whether technicalor institutional in character-points to severe defi-ciencies in national security policy. If security reststo a large extent on the health of the domesticeconomy and its ability to produce technology andmateriel for the national defense, then any substan-tial inability of the military to draw on civiliantechnology and industrial resources constitutes aserious and presumably unnecessary liability. Ifcivilian high-technology industries erode signifi-cantly in the United States, or if the governmentloses the ability to exploit the civilian technologybase efficiently, it amounts to the same thing. TheOTA case studies specifically investigate the extentto which and under what conditions existing barriersare technical in nature, and the extent to which theyare due to institutional considerations, includingregulatory, legal, and administrative factors.

The focus on the problem of the transition oftechnology between the civil and military sectors ofthe economy served as a principal criterion for theselection of the three technology case studies in thisreports Fiber optics is a high-technology industrythat is well-established in the civilian sector, bothdomestically and internationally, but which haslagged significantly in the military sector where ithas numerous potential applications. On the otherhand, advanced polymer matrix composites weredeveloped in the aerospace industry for militarypurposes. They have only begun to find markets inthe civilian sector, even though they could beemployed extensively in the construction, civilianaerospace, automotive, and medical instrumentationindustries, among others. Finally, software is a casewhere the technology is advanced and resident inboth sectors of the economy, but where each sectorstill encounters difficulty in drawing on the re-sources of the other.

sThis chapter employs a rese~ch strategy based on [he method of structured, focused comparison. SCC Alexander L. George, ‘‘Case Studies andTheory Development: The Method of Structured, Focused Comparison, ” In Paul Gordon Lauren (cd.), Diplomacy. New Approaches in History, Theory,and Policy (New York, NY: The Free Press, 1979), pp. 54-59. Scc also, Alexander L. George, “Case Sludies and Theory Dcvclopmcnl, ” a paperpresented to the Second Annual Symposium on Information Processing m organi~ations, C’arncgic Mellon University, Oct. 15-16, 1982, pp. 25-34.

~Thc fol]owi[lg ~ree ~ctions present summaries of t.ttc results of each of the cases as [hey apply to the three principal questions addressed in thischapter. The reader is referred to Appendices D, E, and F in Volume 2 of” this report for more in-depth analysis and documentation.


HEALTH OF THE DUAL-USEINDUSTRIES

Are civilian high-technology industries, thosethat are critical to the military, eroding in theUnited States’? The story is somewhat different foreach of the three high-technology industries consid-ered in this study.

Fiber Optics

Over the past decade, the fiber optics industry hasrealized tremendous growth, not only in productionand sales, but also in the scope of the technologyitself. It is a vital technology whose future patternsof development and diffusion have strong economicas well as military implications. Military plannersrecognize the technical superiority of fiber-basedcommunication systems over those that employcoaxial cable or twisted copper wires. Many analystsbelieve that fiber optic and related photonic tech-nologies will eventually exert an impact on theworld economy comparable to that of the electronicsrevolution of the 1970s and 1980s.

The worldwide fiber optics industry is character-ized by overcapacity and intense competition, withmany advanced industrial nations already designat-ing fiber optics as an essential national capability.By 1980, definite patterns had emerged in the waythat OECD7 member governments would respond tothe strong growth potential of fiber optics markets.In the United States, large, vertically integratedfirms like ITT and AT&T had begun to investheavily in fiber optic R&D.8 An early lead in fiberdevelopment was established by Coming GlassWorks, which holds many of the most importantpatents in the field. Major cable makers weretargeted for takeover by firms seeking to positionthemselves for future fiber optics business that hadnot been principally associated with the telecom-munications industry.

In contrast to the United States, the Japanesegovernment pursued a deliberate strategy of spon-

soring a domestic industry, insulating home marketsfrom foreign competition, building up a highlycapable, vertically integrated industry with signifi-cant overcapacity, and encouraging export of qualitysystems to Europe and the United States. NTT,MITI, and KDD9 (the Japanese international com-munications agency) initiated a carefully orches-trated campaign. NTT (then an official governmentagency) led the effort, conducting and promotingfiber optic and optoelectronic research, workingprincipally with three companies, Sumitomo, Furu-kawa, and Fujikura. At the same time, KDD set upa long-term program to develop all aspects of thetechnology necessary for submarine fiber opticsystems. And MITI sponsored two substantial re-search projects, the Hi-OVIS program and theOptical Measurement and Control System R&Dprogram. By the mid-1980s, Japanese optoelectron-ics companies had developed technology on a parwith the best in the world, and had established amajor position in world markets for fiber opticsystems.

European countries generally appeared to take amiddle ground, with the national PTTs (state-runpublic telecommunications monopolies) establish-ing R&D programs (such as BIGFON in WestGermany) and actively seeking to promote theinterests of their domestic industries. In Sweden andthe Netherlands, the private sector appears to havetaken a stronger role. Most European Community(EC) member states have designated fiber optics asa critically important technology, and the nationalPTTs have tended to favor a few domestic suppliersof equipment and cable. The PITs provide central-ized planning and control of the telephone networkand have supported the introduction of new technol-ogy into that network by sponsoring trials anddemonstration projects. The present configuration ofnational policies would change dramatically if apan-European policy develops in the future.

As a result of these differences in policy andapproach, U.S. firms face stiff competition at

6The fo]]owlng three ~ctions Present ~~marics of the results of cxh of the cases as they apply to the three principal questions addressed in thischapter. The reader is rcftxred to Appendices D, E, and F in Volume 2 of this report for more in-depth analysis and documentation.

7OECD: @-ganila~ion for Economic Cooperation and Dcveloprncnt.

81ntemationa] Tclcphonc and Tclcjytph, and American Tclcphonc and Telck~aph.

gNipWn ~leph~nc and Te]e~aph, Ministry of International Trade and Industry, wd Kokusai Denshin Denwa.


home—while they are effectively barred from sub-stantial penetration of some important foreign mar-kets. Nevertheless, representatives of some Ameri-can fiber and optoelectronic companies believe thatthe United States presently maintains a technologi-cal lead in virtually every area of fiber optics, butthat this lead is eroding. The American position wasestablished and is still based on intense competitionfor sales to American telephone companies. Someanalysts believe that because the industry is robust,officials in Washington should stay on the sidelinesand allow market forces to continue to strengthen anemerging industry in which the United States hasalready proven itself to be particularly sturdy andcapable.

Many analysts are less optimistic about thecompetitive status of the fiber optics and optoelec-tronics industries in the United States. They believethat the success of the U.S. industry is by no meansassured, but instead will hinge on a variety of criticalfactors. The most important of these are discussedbriefly.

First, the future health of the U.S. fiber opticsindustry largely depends on the extent to which itcan sell fiber and optoelectronic devices to thetelecommunications companies. That business, inturn, depends on building fiber optic links toindividual homes across the United States, Legisla-tors and regulators have tended to shift responsibil-ity for the national telecommunications infrastruc-ture to market forces and to the courts. Someanalysts believe that the present regulatory structure—one that effectively separates telephone from televi-sion delivery systems and inhibits the spread oftelematic (online) services—retards the develop-ment of the optoelectronics industry in the UnitedStates. At the same time, huge, vertically integratedJapanese and European firms are gaining experiencein the production and commercialization of large-scale fiber optic local area networks (LANs) in theirhome markets.

A second area of concern focuses on the lack ofinternational standards for fiber optic systems andassociated optoelectronic devices. While interna-tional standards are developing, especially for inte-grated services digital networks (ISDN), progress inthis area is slow for an industry that is innovatingquickly. Different countries have tended to adopt

different standards, and standards have sometimesbeen used as non-tariff barriers to protect homemarkets for developing industries. Some industryrepresentatives believe that Japan and the Europeannations have advanced farther towards setting stan-dards than has the United States, and that they maysucceed in imposing de facto standards on thecompetition in the future.

Third, penetrating foreign markets-especially inJapan but also in some European countries-is stilldifficult for American firms. This disadvantage forU.S. companies is compounded because futureexpanded demand for fiber optic systems is expectedto occur first in foreign markets, where domesticmanufacturers are favored.

Fourth, most European producer nations and theJapanese Government have designated fiber opticsas an essential technology of the future and theysubsidize R&D in the optoelectronics field. In theUnited States, government assistance has beenconfined largely to the military, and U.S. companieshave tended to pursue research and development onan ad hoc, isolated basis.

Finally, the United States continues to maintain aregime of export controls for fiber optics that is morerestrictive than that of its CoCom partners andnon-CoCom nations such as Sweden and Finland.U.S. unilateral controls have tended to exclude U.S.firms from participating in some markets that areopen to the European and Japanese competition. Inaddition, some foreign firms are reluctant to buyU.S.-made optoelectronic parts and components,because they fear that U.S. Government prohibitionsagainst exporting goods to third-party countries willapply to them.

Software

Although the U.S. software industry currentlydominates world markets, both technically andeconomically, its continued superiority will dependon a number of complex factors. The industry facesdifficulties in meeting growing demand for all typesof software-packaged, integrated systems, andcustom-built. International competition is increas-ing as other nations—particularly Japan, France, theU. K., Korea, and India—establish software produc-tion capacity and seek to penetrate global softwaremarkets. U.S. software firms increasing y face tariffs


and foreign trade policies that restrict imports ofU.S.-developed software. Enforcement of intellec-tual property rights in international software trade islax. And finally, as the world market continues togrow, its composition will undoubtedly change, andthe demand for new types of software may createadvantages for companies in foreign nations. Thesefactors are addressed below.

The ability to meet the growing demand forsoftware, and the ability of the United States tomaintain its dominance of the software market,depends on the supply of computer programmersand the technology available to them. U.S. compa-nies cannot meet the demand for software with thepresent number of computer programmers. Theshortfall of software professionals in the UnitedStates is estimated at 50,000 to 100,000 and isforecast to grow steadily over the next decade. Thelack of qualified software developers maybe part ofa larger shortfall in trained science and engineeringprofessionals in the United States. Beyond anydoubt, there is a serious shortage of rigoroussoftware engineering programs at U.S. colleges anduniversities.

Many programming methods and practices usedin U.S. industry today are primitive when comparedto sophisticated software engineering techniques.The software development process can be improvedthrough the use of formalized and automated engi-neering techniques. These support the iterativebuilding and testing of software prototype systems,allow for the reuse of software components, andaccommodate the complexity of software systems.Widespread use of these technologies in the UnitedStates is impeded by the existence of a large,embedded, heterogeneous software base.

The growing cost of software maintenance isdirectly related to the failure to recognize softwareengineering as a scientific discipline and to the lackof trained software engineers. Software maintenance—the modification of software to correct errors and toincorporate changes or enhancements—has becomethe primary cost in most software systems. Presentestimates indicate that in fiscal year 1990, DoD will

spend 80 percent of its $20 billion software budgeton maintenance.

Approximately 40 percent of the packaged soft-ware10 revenues earned by U.S. firms come from

outside the United States. This share is threatened bythe software industries in Japan, France, the U. K.,Korea, India, Taiwan, and Singapore. Japan is thestrongest competitor primarily because of its ad-vanced hardware industry and the propensity to takeadvantage of standardized technologies and developmarketable products from them. A principal strengthof the Japanese is the ability to close large portionsof their domestic market to foreign products, andsimultaneously to penetrate U.S. markets withsystems software developed using U.S. standardizeddesigns.

A comparison of the U.S. and Japanese industriesshows that, while the level of software technology inboth countries is similar, Japanese firms create moredisciplined software engineering environments inwhich the use and production of tools is morewidespread. As a result, Japanese programmers aremuch more productive than their U.S. counterparts.In contrast to the U.S. industry, Japanese softwarecompanies tend to invest more money in basictechnology and to distribute this capitalizationacross the entire firm, rather than limiting it toparticular software projects. Many Japanese compa-nies view programming as an applied science. Their‘‘software factories, ” which reuse approximately 30percent of previously developed software, have anerror rate one-tenth that of U.S. companies, and havethe potential to produce lower cost and higherquality software.

As U.S. software companies operate in worldmarkets, they are increasingly subjected to intellec-tual property violations and infringements. U.S.domestic intellectual property protections (copy-rights. trademarks, trade secrets, and proprietarydata) are insufficient to protect U.S. interests inmany foreign nations, where the penalties forintellectual property infringement can be less thanthe resulting profits. This problem is most pro-nounced in less developed countries, which havelittle to lose and much to gain by not honoring U.S.

lop~~k~gc~ soflwwc is soflwwc that is ~Onlmercial]y cicvcloped an(j broadly marketed, as opposed to custom software, which is dcvclopcd to meetthe particular needs of a specific user.

Chapter 9—Civilian Technology and Military Security ● 167

regulations. Japan is also cited frequently for viola-tions.

Additional economic loss for U.S. industry isattributed to the restrictive trade policies of manyforeign nations, which serve to foster native soft-ware industries at the expense of U.S. firms. Importquotas, discriminatory taxes, local ownership re-quirements, embargoes, and preferential treatmentfor locally produced goods are among the commonpolicies that discourage or preclude U.S. firms fromseeking business. These practices are most pro-nounced in Brazil, India, Mexico, and Korea.

Polymer Matrix Composites

Although the U.S. DoD drives the development ofcomposite materials technology (historicallythrough its R&D funding and now through itsaircraft/aerospace purchases), advanced compositesis a global business conducted by companies (U.S.and foreign) with broad international interests.11

Large chemical and petroleum companies are sup-pliers of fibers and composite parts around the globe.

The world PMC industry is extremely intertwinedin terms of corporate vertical integration, integrationwith its major end-use market (the military aero-space prime contractors), and with multinationalchemical and petroleum interests. Advanced com-posites are formed in a series of stages, each ofwhich corresponds roughly to a different industry—raw materials, fiber preparation and shaping, andcomponents for end-use.12 In recent years, rawmaterial suppliers such as Amoco, British Petro-leum, Phillips, Shell, BASF, Ciba-Geigy, Du Pent,and Hercules have moved downstream into fibersand shapes, where there is more value added in theproducts. Most of these companies buy from, sell to,and compete with each other for business frommilitary prime contractors. At the same time, de-fense aerospace companies, which had relied onspecialized companies for part forming services,have moved upstream, making parts in-house andbuying only the raw materials.

In this discussion, two distinct stages of corporateintegration can be defined: material suppliers andend users (including intermediate material suppli-ers). Fibers are sold as standardized commoditymaterials. End users (and intermediate suppliers)develop individually tailored structures for eachapplication. Because of this dichotomy, fiber suppli-ers conduct a different style of business, withdifferent issues and concerns, from that of the endusers.

Material Suppliers

Carbon fiber is a principal ingredient in theproduction of advanced PMCs. About 65 percent ofthe U.S. carbon fiber market is in the aerospaceindustry. Over half of the U.S. aerospace market forfiber is military. Defense applications are projectedto grow by as much as 22 percent annually in thenext few years. The U.S. military market is a primarytarget for foreign companies producing carbon fibercomposites, because it is the largest, most advanced,and most attractive in terms of sales and profitabil-ity. The second largest market is in the Far East,where carbon fiber products are used to makesporting goods.

Worldwide, carbon fiber capacity is twice thecurrent market volume. Japan and the United Stateshave about equal capacity. Japanese companiesmanufacture a carbon fiber precursor, which is thensold to U.S.-based carbon fiber suppliers (mainlyHercules), which is in turn the major supplier of fiberfor military programs. At present, no Japanesecarbon fiber is supplied directly to U.S. militaryprograms.

U.S.-based industry is continuing to add carbonfiber capacity—about 1 million pounds in 1988.Accordingly, there is and will continue to be a greatdeal of excess capacity both in the United States andin world markets. While the United States has a largefiber overcapacity compared to domestic marketrequirements, most of the world excess capacity isconcentrated in Japan.

1 IThe ~l~cussion of pMCs draws on ~ previous OTA a s s e s s m e n t . Scc ~J, s , CongCsS, offi~c o f ‘TcchrIo]ogy A s s e s s m e n t , Advance

Design: New Sfrucruraf Materials Technologies, OTA-E-35 1 (Wush]ngton, LX: U.S. Government Prin[ing Ofilcc, June 1988).121n ~hc ,Ilmufacturc of PMCs, highly ~r~cs~d ~wbon fi~rs xc ~hcm ically ~C~tCd ~nd ~nded With a m~trix matcri~], The malcria] is shaped during

this process which involves heating and compressing it into d mold. Thcw shapes arc ihcn Iinishcd by machining, and bccomc final products such asairplane wings and tennis rackets.


Although Japan is the largest manufacturer ofcarbon fiber in the world, it has been only a minorparticipant to date in the advanced compositesbusiness. Japanese companies have been limited bylicensing agreements from participating directly inthe U.S. market. In addition, Japan does not have adomestic aircraft industry to which advanced PMCscould be sold, although it is trying to establish onethrough a joint venture with Boeing and through itsdecision to build the FSX fighter. Japanese compa-nies are building a strong position worldwide inPMC technologies.

End Users

The United States leads the world in developingand using advanced PMC technology, based largelyon the strength of its military aircraft and aerospaceprograms. Nevertheless, foreign commercial endusers outside the aerospace industry are more activein experimenting with these new materials than aretheir U.S. counterparts. Western Europe leads theworld in composite medical devices, partly becausethe regulatory environment controlling the use ofnew materials in the human body is less restrictivein Europe than in the United States. The EC is takingadditional steps to commercialize advanced PMCs.For example, the EUREKA Carmat 2000 programproposes to spend $60 million through 1990 todevelop advanced PMC automobile structures.

The U.S. market on the whole is projected to growfaster than the world market, based on the assump-tion that the military demand for PMCs will expandrapidly over the next 5 years. Although the numberof U.S. military aircraft being built is declining,composites are replacing much of the metal onairplanes. For example, the F-16 has 260 pounds ofadvanced composites per aircraft, while the V-22,which recently moved into production, will havefrom 8,000 to 9,000 pounds per aircraft. However,growth is expected to level off in the middle 1990sas advanced PMCs move into all of the structures forwhich they are suited.

Foreign production of U.S. aircraft components isincreasing, and manufacturing of composites forcommercial aircraft is moving offshore in manycases. A significant number of foreign companiesfabricate parts for U.S. aircraft manufacturers. Thisis largely the result of economic offsets that are usedto secure sales of aircraft by offering portions of the

aircraft fabrication to companies from the buyingnation. Such sales enhance technology developmentin, and the potential economic competitiveness of,foreign-owned advanced composites businesses,possibly at the future expense of U.S.-owned firms.

In the past few years, participation of WesternEuropean-owned companies in the U.S. advancedPMC market has increased dramatically. This haslargely taken the form of acquisitions of U.S.-ownedcompanies. Industry analysts indicate that U.S.carbon fiber facilities have been sold, due tocorporate ‘‘impatience” resulting from the need toreport favorable quarterly earnings. In general,foreign corporations tend to be more patient. Despiteexcess worldwide capacity and profitability prob-lems, the Japanese have not sold any carbon fiberfacilities. As a result of extensive acquisition of U.S.firms, foreign makers of advanced materials haveentered the U.S. aerospace market and share thetechnology leadership that participants enjoy.

CONVERGENCE OF MILITARY ANDCIVILIAN TECHNOLOGIES

Do military technologies and their applicationsdiverge significantly from their counterparts inthe civilian sector of the economy?

Fiber Optics

The distinction between tactical and fixed-plantfiber optic systems is important. Tactical systemsrequire rapid mobility. Although fixed-plant sys-tems are installed directly in the ground or inconduits, most tactical systems must be placed onthe ground, strung above the ground, or deployed atsea. While there are no significant limitations oncable length for fixed systems, tactical systems mustbe configured so that they can be set up and retrievedquickly. In addition, cable used in tactical communi-cations must be more flexible and durable than thatused in fixed-plant systems. While optical splicingmay be used for many fixed applications, connectorsare necessary due to the requirement for mobility ina tactical environment. And finally, batteries orother local sources of power are usually required todrive sources and repeaters in tactical systems.Generally speaking, shipboard fiber optic systemscan be considered as fixed plant.


Do such differences between military andcivilian applications translate into differences in thetechnology itself or in the way that R&D for fiberoptics must be conducted? The answer is a qualified“No.” For fixed-plant systems, military require-ments would differ only marginally, if at all, fromthose used in private-sector businesses or for localarea subscriber networks. For a large percentage ofmilitary applications—wiring the Pentagon, theDoD laboratories and R&D facilities, and themilitary bases—the technology is broadly availablefrom the civilian sector. In addition, fiber opticsystems deployed on ships would be similar toLANs now undergoing trials in the private sector inJapan and in the United States.

Optical sensors have enormous potential in a widerange of applications both military and civilian.Many of the major sensors used by the military-orunder development—are analogous to those used inthe civilian sector. One fiber optics group in theNavy has tested 54 different sensors developed forcivilian purposes and found that most of them do notperform adequately in a military context. Theyconcluded, however, that the civilian sensors shouldnot be discarded and replaced by sensors built tomilitary specification. Such specifications do not yetexist, and the process of writing them and gettingthem approved would take years. Instead, the grouptakes the approach of addressing the military re-quirement by modifying commercial products sothat they are suitable for the particular military taskfor which they are envisioned.

Their objective is to use the existing technology—which they believe is far more advanced than thatwhich the Services presently need. For example,industry already has endoscopic devices for lookinginto machinery and into places where electronicscannot be placed. This is not a new or radicaltechnology. These devices represent basic technol-ogy with new applications. In this approach, DoD’schallenge is to figure out how to take the technologythat is available-not a radical departure fromit—and use it in a military setting.

Despite the decidedly military character of theFOG-M missile,13 its designers indicate that the

Army has been able, for the most part, to use opticalfiber that can be produced on modified commercialmanufacturing equipment. The fiber companieshave entered into earnest discussion with the FOG-M program, because they anticipate a run of fiberthat might reach a volume of up to 2 millionkilometers. There are special military requirementsin the way that the fiber is wound on the spool, in thefiber design, and in the materials that are used toattach the fiber to the spool. But these do nottranslate into large differences from civilian technol-ogy, nor do they require significant changes in theway that R&D) is carried out. The military require-ments can be met if civilian fiber companies arewilling to develop the modifications.

Software

The software industry is increasingly divided intotwo camps, one that is dedicated to military interestsand another that supplies the commercial world.These two sectors have been present since the birthof the industry, and exchange between the two wasassumed to be the norm, not the exception. But thereare significant indications that divergence betweenthese groups is increasing, which may contribute toa weakening of the U.S. software technology base.

The underlying software technologies are verysimilar in both the military and civilian sectors, anddivergence becomes noticeable only in the detailedrequirements for specialized applications. Conver-gence between civilian and military software indus-tries is most noticeable in the small-scale applica-tions and systems software areas. Both sectors usepackaged COTS software for the majority of theirsmall-scale software applications, such as personalcomputer (PC) based programs and office automa-tion products.

Similarities in the applications of software are notlimited to PC-based and systems software. Analo-gous applications of large-scale software systemsalso can be found in both sectors, including softwaredeveloped for avionics, telecommunications, andembedded systems. But while the applications aresimilar, military and civilian environments placedifferent, sometimes opposing requirements on the

]3Thc Fibcr @tic Guided Mi\~ile (FOG. M), now in fu]].sca]c Cnginccring dcvelopmcn(, pays OIJ( optical fiber from a bobbin, enabling the battlclicld

operator to target the missile with a real-time video image cmanating from a camera in the nose of [he missile.


software that controls these systems. This is particu-larly true for large-scale, mission-critical applica-tions in the DoD.

Thus, different requirements, as well as differ-ences in scale, create two distinct software industriesin the large-scale applications area. The industrydivergence is illustrated in avionics systems soft-ware, where military requirements for high-performance avionics are exchanged for high surviv-ability and safety in civilian avionics. The signifi-cance attached to software requirements by eachsector, and whether they become rigid specificationsor economic trade-offs, partially explains why thereis little transfer of software between the military andcivilian sectors in the embedded and large-scaleapplications.

Military requirements for custom-built and em-bedded software are generally far more rigid thancivilian requirements. Once documented and ap-proved in the design stage, specified requirementsgovern the subsequent development of the software.

The need for specific performance and opera-tional characteristics is evident in many DoDmission-critical systems. It is necessary to requirenearly 100 percent reliability for a missile guidancesystem or multi-level security in a networkeddefense communications system. But when theserequirements are transferred unnecessarily to othermilitary systems, the cost of development increasesand the ability to use analogous civilian applicationsor commercially developed software diminishes.

Many of the requirements often identified asunique to military applications—for example, multi-level security, data encryption, interoperability,survivability, and high reliability—are equally ap-propriate in banking, insurance, commercial flightcontrol, and other civilian applications. Indeed,many features incorporated into military systemscould be transferred to civilian applications and viceversa. But while these features are desirable andappropriate in civilian applications, their implemen-tation would be based on economic and risk analysis.In the civilian sector, if the cost of implementing arequirement exceeds the expected return, then therequirement is usually deleted or deferred. Thisanalysis and design-to-cost approach rarely occursin military software acquisitions, although similar

accommodations will be more likely if militarysoftware costs continue to escalate.

A more recent divergence between the militaryand civilian sectors of the software industry relatesto the military’s mandated use of a single high-orderlanguage, Ada, in its mission-critical software sys-tems. DoD’s sponsorship of Ada began in 1974when the ‘‘software crisis” was first recognized andacknowledged to have potentially serious conse-quences for the military’s ability to maintain andoperate its many computer systems. In 1983, Adawas approved as a standard by the AmericanNational Standards Institute (ANSI) and by DoD asMilitary Standard (MIL-STD) 1815A. By 1987, Adawas approved as an International Standards Organi-zation (IS0) standard.

The DoD Directive that Ada shall be the singlehigh-order language used in command and control,intelligence, and weapons systems has no counter-part in the commercial environment. With theexception of civilian avionics systems, Ada is notwidely used in U.S. commercial applications. In-stead, civilian-based software continues to be imple-mented in the language considered to be best for thatparticular application—whether it be COBOL, afourth generation language, or any other computerlanguage, As new DoD computer systems aredeveloped, the convergence of new software tech-nologies and the ability to transfer software betweenthe two sectors will depend a great deal on severalfactors: first, the civilian sector’s acceptance of, anddemonstrated use of, Ada; second, DoD’s willing-ness to grant waivers to its Ada mandate; and finally,the military’s acceptance of, or ability to, incorpo-rate commercially developed, non-Ada softwareinto its computer systems.


There is both convergence and divergence inmilitary and civilian applications of advanced PMCtechnology. In general, military and civilian marketshave different technical and cost criteria for theselection of materials and process technologies.Convergence and divergence occur simultaneouslyin different aspects of the PMC industry and itsmarkets.

Various segments of civilian and military marketsplace different emphasis on performance and cost. In

Chapter 9---Civilian Technology and Military Security ● 171

the commercial aerospace, military non-aerospace,automotive, and construction markets, for instance,acquisition costs and operating expenses are themajor purchase criteria, with a progressively lowerpremium placed on high material performance. Inmilitary aerospace, biomedical, and space markets,on the other hand, functional capabilities andperformance characteristics are the primary pur-chase criteria.

Although general functional requirements (e.g.,low weight, high strength for primary structures,lower strength for secondary and nonstructuralparts) lead to convergence between the military andcommercial aircraft sectors, the stringent missionrequirements for military aircraft drive the use ofadvanced composites in the military. For spaceapplications and fighter aircraft, advanced PMCs aremore than just one of many competing materials.They can be the enabling technology for missionrequirements because of their high strength-to-weight ratio.

The use of lower cost materials (such as glass-reinforced composites, or fiberglass) in generalmeans more weight and lower performance in thetraditional aerospace sense. Industry representativesassert that battlefield conditions require that weap-ons systems weigh less. That was the initial reasonfor the attractiveness of composites, particularlygraphite-reinforced composites. While lower costsare desirable in the military aerospace sector,performance remains the main driver.

According to advanced PMC industry representa-tives, cost currently limits market growth and thetransfer of high-performance military PMC technol-ogy to the commercial sector. Carbon fiber is pricedat about $15 to $20 per pound. Chemically treatedfiber, called prepreg, sells for $35 to $40 per pound;and the cost of finished aircraft structural compo-nents is between $250 and $600 per pound. Alumi-num structures cost about $85 a pound, including 2hours of labor and $5 of material. Some 70 to 80percent of the cost of a finished advanced PMC partis due to fabrication costs.

Many developments have wide applicabilityacross both the civilian and the military arenas.There is synergism between military and commer-cial aircraft production in resins and fibers, the waymaterials are stitched together, and the way they are

used. For military and commercial aircraft, thestructures made from composites (e.g., wings, tail,and empennage) are similarly complex to fabricate.The basic method of production of aircraft parts isalso similar: coating of continuous fibers with resin,careful placement of fibers, and application of heatand pressure to form the structure.

However, military requirements may make itnecessary to modify the fabrication process. Forexample, pultrusion is typically used in the commer-cial market to form beams. Military applicationsneed superior load-carrying capacity, so that formilitary applications the pultrusion process must bemodified to impart different properties to the fabri-cated part.

From abroad perspective, the military communityoften requires custom-made hardware, while com-mercial industries look for off-the-shelf productscombining low cost and high quality. Many militaryand space hardware applications are very specializedand require low production volumes. The automo-tive industry, on the other hand, is driven by lowcosts and high production rates. Between the aero-space and automotive advanced PMC markets, avariety of other market applications (including thenon-aerospace military market) have productionrates higher than military aerospace, cost objectivessimilar to automotive applications, and moderateperformance requirements.

Military and commercial aircraft both experiencesimilar environmental conditions, and thus requiresimilar lightning protection, corrosion resistance,fatigue resistance, and material toughness. While thetechnical requirements for PMCs in commercialaircraft are comparable to those for fighter aircraft,there are some major differences related to peakG-loading and maneuverability, repair strategies,stealth, radiation hardening, and design tempera-tures.

Military and commercial aircraft have inherentlydifferent duty cycles. Military aircraft are on theground a significant portion of the time, whilecommercial airplanes are in the air much of the time.Commercial aircraft designers are concerned withstructural fatigue, and with takeoff and landing dutycycles. The dominant factors for maintenance ofmilitary airframes are ground temperature, corro-sion, and exposure.


BARRIERS BETWEEN THECIVILIAN AND MILITARY

SECTORSWhat are the principal barriers, both technical

and institutional, that inhibit military access tocivilian technology and vice versa?

Fiber Optics

In order to do any substantial amount of businesswith the DoD, fiber optics companies have found itnecessary to create a separate corporate division. Tomeet government regulations and specifications,fiber optics businesses must organize many of theirprincipal functions differently—including account-ing, personnel, auditing, R&D, production, advertis-ing, marketing, and management information sys-tems. They must also adjust their business psychol-ogy and profit orientation, Successful fiber opticsand optoelectronics companies invest heavily inresearch, develop a superior product, realize largeprofits, and plow their earnings back into the R&Deffort. This business environment contrasts sharplywith one of government-subsidized research andregulated profit margins.

The question of how to specify fiberoptic systemsand devices for the military poses what amounts toa paradox, both for the industry and for thegovernment. The problem is that optoelectronic andfiber optic technologies are changing so rapidly thatno one can agree on standards. DoD is confrontedwith the problem that, by picking a standard, it maylock itself into an obsolete technology or an applica-tion that no one in the civilian sector is willing tobuild at a reasonable cost. This is because themilitary wants to nail down prescriptive standards14

in a field that is changing from month to month. Thealternative is to adopt performance standards essen-tially specifying, in a general way, the characteris-tics that a part or component must meet, and thenleaving it to industry to figure out the specifics. Thiswould, however, make it more difficult to conductcompetitive procurements. The range of competingdesigns might be very wide, and it would benecessary to trade off price against quality.

Industry executives suggest that the militarygenerally does not recognize the capabilities of thecommercial sector. From the industrial perspective,this is due to “the momentum factor” and “culturalconservatism” in the military, two substantial barri-ers to the large-scale introduction of fiber optictechnology. The former proposes that the Serviceshave committed themselves to older communica-tions and sensing technologies, many of which arenot compatible with fiber optic systems. Accordingto the latter, there is little incentive for programmanagers to seek out a new technology and put itinto a weapon system, particularly if the technologyis changing rapidly and proposed parts or compo-nents are not fully specified.

The lack of industry standards exacerbates thisalready difficult internal problem. From the DoDperspective, there is no way that acquisition manag-ers can make mass-scale purchases from civilianindustry—and this is where the technology resides—in the absence of performance, design, and testingspecifications. These are considered essential to theacquisition process.

By insisting on the use of existing specifications,the military can create barriers to the introduction ofa new technology—for example, when a largecivil-sector company attempts to install a standardfiber optic telecommunications system for a militarybase. DoD could procure regular commercial prod-ucts, since there are no special military require-ments. But it is very difficult to install such a systemon a base. If there are no existing military specifica-tions and standards, DoD is reluctant to buy asystem. If there are military specifications, they areunlikely to correspond to existing commercial prod-ucts, because civilian technology probably advancedwhile the military specifications were being writtenand approved. In this case, the Defense Departmentwill end up paying more for a less capable systemthan would a commercial purchaser. Somehow,DoD must learn to make decisions about what itwants—either by writing specifications, modifyingspecifications, or carrying out procurements withoutspecifications-in a matter of a few months.

1dln ~cncra], ‘Prescriptive stmdards” speci fy how somc~ing is M bC made or what it is to be made from. [n contrast, ‘pcrfonmccstandards” sPecifYonly lhc resulting capabi Iity or pcrfw-rnancc Icvel to be achicvcd.

Chapter 9--Civilian Technology and Military Security ● 173

Industry executives and analysts point to sev-eral key reasons why some optoelectronics and fiberoptics firms have difficulty in selling their productsto DoD, and others are reluctant to do business at all.Those most often cited include: 1) DoD cannotguarantee firms that funding will be available forauthorized projects; 2) DoD seeks to acquire datarights that would compromise large R&I) invest-ments; and 3) to do business with DoD, a firm mustfundamentally alter its corporate structure, policies,and overall intentions. Each of these problems isdiscussed below.

In a somewhat ironic case, a fiber optics companylicensed its technology from a university researchprogram funded by DoD, and is now unwilling to dobusiness with the government. It is a small, highlyprofitable company that is limited in the extent ofmoney and technology that it can leverage for anygiven purpose. Its executives are very reluctant totake contracts with DoD, because they cannot affordto hire specialists who can respond to DoD regula-tions, contracting procedures, auditing practices,and other requirements. They are unable to supportthe cost of research and gearing up for production,unless there is u definite market for the product inquestion and the opportunity to realize substantialprofits.

In the civil sector, a company can developlong-term relationships with its suppliers and cus-tomers. Government notions of fairness and compe-tition rules make it difficult to sustain such relation-ships, a$ does the turnover of contracting personnel.

A second major problem, cited by some industryanalysts, is that government procurement officersand regulations do not recognize the extent to whichfiber optic and optoelectronic technologies aredriven by R&D activity. Government agents tend todemand as many data rights as they can get in anygiven contract. Most fiber optics firms are unwillingto share their data, because they believe that suchdata can be used to reveal a core of proprietaryprocess information.

For optoelectronics and fiber optics companies,the problem of protecting proprietary rights comes at

the very beginning of a decision to take a gover-nment contract. If the company has not worked withDoD in the past or if the military segment of thebusiness is small, many executives tend to skirt theproblem by avoiding government contracts.

A third major impediment between DoD andcivil-sector fiber optics firms is the perception on thepart of industry executives that they are simplyill-equipped to do business with DoD. This is in parta consequence of the divergence of business prac-tices in the military and civilian sectors of theeconomy. and partly a result of inflexibility on thepart of government. To do substantial business withDoD, managers would have to learn to live with andrespond to regulatory, reporting, accounting, andauditing requirements that are largely incompatiblewith their own systems, and that do not make sensein the context of civil-sector business.

Software

Despite similarities in the technologies availableto the civilian and military software sectors, differ-ences in their respective acquisition strategies ob-struct the exchange of software technologies andapplications. Persistent barriers to the transfer oftechnology, methodologies, and products betweenmilitary and civilian interests are identified below.

In 1987, a DSB task force reported that bothtechnical and management problems are evident inmilitary software development, with the latter beingmore significant.15 These management problemsrelate to the manner in which DoD procures soft-ware, and they represent major barriers to theexchange of software technology between the civii-ian sector and DoD.

According to industry representatives, the princi-pal problem is the bureaucracy and administrativeoverhead associated with DoD acquisition proce-dures. The requirements regarding procurement,design, development, and maintenance of DoDsoftware are set forth in DoD-STD 2167A. As agovernment review mechanism, DoD-STD 2167Areferences and directly or indirectly requires compli-ance with many additional standards, directives,data item descriptions, and Federal Acquisition

lsDcfcnsc science Bowd, “Repo~of~e DSB 1987 Summer Study on Technology Base Management,” prepared for the Officcofthc Under Sccrctqof Defense for Acquisition, December 1987.


Regulations (FAR). As is true of fiber opticscompanies, in an attempt to comply with contractualobligations, commercial software vendors mustemploy specialists who are fluent in military regula-tions, government reviews, documentation, andaccounting procedures. These requirements andassociated legal issues have forced many DoDcontractors to establish autonomous divisions forconducting business with the government. As aresult, few civilian software firms regularly contractwith DoD.

Defense Depw-tment acquisition procedures andcontracting practices limit the number of potentialvendors and discourage established contractors whoalready work for the military. Civilian firms thatcontract with DoD receive no guarantee of acontinued relationship with the government, achievepoor profit margins, and often lose the rights-in-datato their software.

Although software firms guard proprietary infor-mation closely, this property is often transferred—by contract—to the government. Despite the flexi-bility allowed government contracting officers tonegotiate less-than-exclusive rights to data in soft-ware acquisitions, commercial venders generallylose most, if not all, of their intellectual propertyrights to the software they develop. The gover-nment’s claim to unlimited data rights is based on thenotion that these rights protect the government andensure public dissemination of publicly sponsoredresearch efforts. In negotiating for unlimited rightsto data for its software, the government achieves theability to maintain and modify its software systemsin the future. This practice is intended to ensure faircompetition for future software maintenance andreprocurement contracts. Some analysts assert, how-ever, that such policies weaken DoD’s ability tonegotiate for the best software at competitive prices,because they drive away potential bidders.

Ada has been cited by some civilian softwarefirms as a barrier to doing business with DoD. Thedirective stating that Ada shall be the “single,common, computer programming language” used incommand and control, intelligence, and weaponssystems may help in the long run to alleviate themilitary’s software crisis. But because of its relativeimmaturity, the number of software firms proficientin Ada is limited. The mandate to use Ada appears

to reduce the already limited number of firms willingand able to contract with DoD.

Some experts cite Ada as an example of thegovernment’s tendency to standardize too much, tooearly. Although the requirement to use Ada formission-critical applications was arguably prema-ture in 1983, developments associated with Adahave advanced significantly since that time. Butmany commercial vendors, with the exception ofthose in the avionics industry, still take a wait-and-see attitude about Ada.

The merits of a single, standardized language,such as Ada, will continue to be debated. Ada’sbenefits include its embodiment of engineeringtechniques essential to the development of maintain-able software, its support for modular (and reusable)components necessary in the development of large-scale, integrated systems, and its portability amongdiverse computer architectures. Additionally, be-cause it was standardized early and trademarked,there are no incompatible dialects of the language;such dialects tend to decrease the reliability andcomplicate the maintenance of software systems.These characteristics have the potential to bridgesome of the technological differences between thecivilian and military sectors.

Whether Ada becomes an area of convergence,rather than a barrier, remains to be seen. BecauseDoD is the single largest consumer of software andis committed to the use of Ada, the language will bean important factor in future software technologies.Its potential, though, conflicts with the currentsituation in which many military mission-criticalapplications are required to be implemented in Ada,while similar civilian applications will continue tobe developed in the language deemed best for eachparticular project.


Advanced composites technology was first ap-plied in the military sector. Although the PMCindustry envisions a very large commercial marketfor advanced composites in the future, it sees limitedcommercial opportunities today. PMC suppliers feelthat commercial development is the key to profit-ability in advanced composites, and that sustaininga presence in the military marketplace is a way topursue it.

Chapter %--Civilian Technology and Military Security ● 175

As in the other industries under review, militarycontracting and accounting procedures, and thepotential loss of proprietary rights and patentability,may be the costs of participation in the militarycomposites market. Loss of data rights is consideredby some commercial sector companies as a threat totheir survival in a competitive marketplace. Forfeit-ing proprietary rights goes against the ‘‘corporateculture” in many non-defense companies and fear ofsuch losses inhibits the flow of technology betweenthe defense and commercial sectors. Indeed, tech-nology developed in the commercial half of acompany may not be shared with the military halfdue to proprietary concerns.

These barriers represent inhibitions, but not prohi-bitions, to the transfer of technology between thecivilian and military sectors. Participation by com-mercially oriented companies in recent defenseprograms, such as the Low Cost Composite WeaponProgram and C-17 subcontracts, indicates that suchcompanies are willing to engage in military pro-grams.

Government business rules and regulations haveinhibited the transfer of PMC technologies from thecommercial sector into military applications. Forexample, in 1978 ACF Indusrnes successfully de-veloped an inexpensive glass-fiber composite rail-road car based on aerospace technology. DoDrepeatedly approached ACF to use this technologyin an ongoing defense program. ACF managementdeclined to work with the government, becauseputting up with government audit procedures wasmore trouble for the company than it was worth.

Similarly, the teaming arrangement for the LowCost Composite Weapon Program was designed toaugment a military aircraft manufacturer’s capabili-ties with the lower-cost commercial technology ofnonmilitary subcontractors. The lack of simplepurchase orders for commercial sector contractorsand the government accounting compliance require-ments met with stiff resistance. The commercialsector subcontractors expressed reluctance to par-ticipate on this project, because of the requiredforms, audits, and the justification of overheads.

In addition, personnel working on highly clmsi-fied programs sometimes cannot obtain clearance toshare nonsensitive information such as genericmaterials and process technology data. This infor-

mation is often embedded in classified reports. It iscostly for the military or the contractor to employpersonnel to extract generic types of informationfrom classified reports.

The DoD has similar problems internally. Someanalysts are concerned that there may be technologyunder development in the “black world” that therest of DoD could build on but does not know about.PMC industry representatives have indicated thatmore attention should be placed on the transfer of“black” technology into the “white” technologybase.

Industry representatives indicate that the pressureto share data in military markets to reduce costsconflicts with their competitive instincts. Somecompanies feel that information disclosed to the~overnment would become public and might be usedby their competitors in a different market. Neverthe-less, some sharing of materials databases is neces-sary to reduce the currently excessive costs of R&Dand processing.

Aircraft manufacturers, parts fabricators, andmaterial companies that contract directly with DoD(or that take subcontracts) often set up separatedivisions to comply with government regulationsand procedures. Although personnel can be trans-ferred from the commercial divisions or hired fromother defense contractors, industry analysts state thateverybody in the defense division eventually thinks“government contracting. ” The overhead chargedby that division is typically a great deal higher thanthat charged by the rest of the company.

SUMMARY OF CASE STUDIESThe case studies send mixed signals about the

overall health of the three industries. Each exhibitsdifferent strengths and vulnerabilities. The U.S.optical fiber industry is strong today, but it isconcentrated in two large companies. These compa-nies face daunting competition in the future, bothfrom the EC nnd from .?apan. In software, U.S.companies clearly lead the world in both sales andtechnology leadership. But the competition—especially from Japan—is closing the gap, and theUnited States is experiencing a growing shortage ofsoftware engineers. The U.S. PMC industry is strongand thoroughly internationalized. It is also ex-tremely dependent on the Defense Department. DoD


has supported PMC technology because it is bothpervasive and enabling for a large number ofmilitary systems. But the cost to DoD is very great.To date, there is little indication that U.S.-basedproducers are willing or able to diversify intocivilian markets, so long as a lucrative militarymarket exists.

For each of the three technologies, a high degreeof technological convergence between civilian andmilitary applications is evident. There are also someapplications that are unique to the military for eachof the technologies. There is, however, significantdivergence between the military and civilian eco-nomic sectors for each industry, and this occurs forreasons that are not directly related to the technologyitself. In the fiber optics industry, the civilian sectoris far in advance of the military in most areas. Thesoftware industry is increasingly divided into twocamps--one that serves the military and one thatdoes not. Such divergence is rooted largely indifferences in the way that military and civilianbusiness is conducted. Of the three technologies,advanced materials shows the greatest divergencebetween military and civilian applications. There aresignificant differences in the molecular structure ofeach end-product because each PMC material mustbe individually designed. However, automation ofthe production process and dissemination of databases would certainly reduce costs and aid indiffusing PMC technology into a variety of civilianapplications.

In each of the cases reviewed, the barriers thatexist between the military and civilian sectors of theeconomy are due largely to differences in organiza-tion, administration, and business practices, ratherthan to differences in the technologies themselves.Indeed, this is a principal finding of this study. Thereality—that the military often buys less thanstate-of-the-art technology—is disturbing, both froma national security perspective and in terms of thekind of efficiency that is associated with goodgovernment. As we have seen, many companies thatproduce fiber optic and optoelectronic technologiesare reluctant to do business with the government.Barriers to the participation of civilian sector com-panies appear to be largely generic across manyindustries. In the case of software, DoD simply doesnot have access to the best and most advanced

civilian talent and products, and there is very littlesynergy between the military and civilian compo-nents of this critical high-technology industry.Advanced composites offer a different view of thesame sorts of problems. Because the industry wasdeveloped largely by military aerospace companiesand other DoD contractors, it makes sense to look fora ‘‘spin-off” effect. However, U.S.-based firms thatdo significant business in advanced composites withthe DoD have generally not been successful inmarketing their products in a nondefense context.

FINDINGSThe general findings of the case studies are

presented below. While they are based on the threecases, they have wider implications for the defensetechnology base and for other high-technologyindustries. Detailed findings that are more specific tothe individual technologies are presented in theparticular case studies, located in Volume 2 of thisreport.

Overall Findings

1. Two relatively separate economic sectors haveevolved in the post-World War II period, onemilitary and the other commercial. Business prac-tices in the two diverge significantly, and substantialbarriers impede the transfer of advanced technologybetween one sector and the other.

2. Nevertheless, the ability of the military toachieve and maintain leading-edge technology in thefuture will depend in many cases on the health of thecorresponding indust~ in the commercial sector ofthe economy. Machine tools and semiconductors arewell-documented examples.

3. The barriers that stand between the military andthe commercial high-technology sectors are largelydue to legal, institutional, and administrative factors,and are not inherent in the technologies themselves.

4. The United States is failing to develop and/ormaintain a competitive commercial base for sometechnologies that are important or even essential tomilitary procurement. It is likely that DoD either willhave to turn increasingly to foreign suppliers toachieve or maintain state-of-the-art capacities insuch areas, or will pay a high price to maintainin-house capacities.

Chapter 9---Civilian Technology and Military Security ● 177

5. Ixmgstanding industrial and trade policies mayhave to be reformed if the United States is to achieveand/or maintain world-class industrial capacity insupport of certain essential dual-use technologies.

Findings Relevant to theDepartment of Defense

1. DoD faces two central problems in the area ofdual-use technology. First, government procurementpractices make it increasingly difficult for DoD toobtain state-of-the-art technology in areas where theprivate civilian sector is leading. Second, certainessential high-technology industries are weak in theUnited States, and others may not be located hereinthe future.

2. Due to the magnitude of the investment that isrequired to create advanced technological capabilityin a number of critical areas, DoD cannot afford tofinance advanced technology and product develop-ment across the full spectrum of technologies thatare important to the military. Instead, it must rely oninnovation and R&D in the civilian sector to pullsome technologies forward.

3. Initiatives to increase DoD’s access to and useof commercial technology have tended to fail in thepast, largely because they did not address institu-tional and structural factors.

4. Many firms are reluctant to do business withDoD because they consider the government to be abad customer. Some commercial firms cite exces-sive regulation, burdensome auditing and reportingrequirements, damaging competitive procurementpractices, rigid military specifications, compromiseof proprietary information, loss of data rights, andcorporate “culture shock” m reasons not to seekDoD business. Some of these problems might beresolved through DoD or congressional action.Others are probably inherent in any industry-govemment relationship,

5. The recent expansion of special access or“black” programs reflects both an effort to increasetechnological security and an attempt to circumventburdensome regulation and congressional oversight.Because they are highly classified, such programspresent additional barriers to companies that are notordinarily engaged in defense work.

Findings Relevant to theDefense Industrial Sector

1. In many high-technology areas, the defenseindustries no longer lead the commercial sector, andthe disparity may be increasing.

2. The defense industrial sector has been shapedby an ad hoc—yet extensive—system of regulationand defense industrial policies that has tended tostifle innovation and creativity over time. Theseinclude nearly 400 different regulatory requirementsin the FAR alone, extreme and uncoordinatedgovernment auditing activity, and pervasive over-specification of developmental items. Some wereestablished pursuant to acts of Congress; others arethe result of internal DoD practices.

3. Congressional attempts to reform DoD and thedefense industries may be inappropriately aimed atfixing an archaic military-industrial structure that isout of step with a world economy radically trans-formed by intense international competition. Com-prehensive restructuring and elimination of ineffi-cient elements and practices within DoD may benecessary.

4. Strategic planning in the largest defense primecontractors is based on the assumption that doingbusiness with the government is a slow and ponder-ous process, and that it is not likely to get better. Forbetter or worse, corporate planning is married to theDoD planning and budgeting cycle.

Findings Relevant to Congress

1. Congress plays a major role in shaping theacquisition process, and with it, DoD’s access toboth the defense-specific and commercial technol-ogy bases. But the acquisition system has built upover time, and overriding national goals oftenconflict with the particular program-specific objec-tives of defense acquisition. These national goalsinclude efforts to ensure fairness, access for smalland minority firms, environmental protection, com-petition, and the best product obtainable with thetaxpayers’ money.

2. The opportunity to interpret and amplify theintent of Congress exists at many levels—in OSD, inthe Services, in the large prime contractors, and atthe sub-tiers-and is often acted on, particularlywhere criminal sanctions might be imposed. The


result is a risk-averse, highly conservative defenseindustrial sector that has trouble taking advantage ofrapid technological change in the commercial sector.

3. The structure of the large defense companiescomplements the legal, administrative, and bureau-cratic form of the government. Such structures areimposed on the defense firms by acts of Congress, byDoD regulations, by military specifications, and byauditing requirements. Every aspect of business inthese companies conforms to and is enforced by suchbureaucratic and administrative controls. Makingthe defense industries more efficient and account-able might entail radical alteration of the legal andinstitutional structures that shape the DoD.

Findings Relevant to Civilian Industry

1. In many dual-use, high-technology industries,the civilian sector leads the defense industries. Thiscivilian capacity may or may not be located in theUnited States.

2. A company can organize to do business ineither the military or in the civilian sector of ahigh-technology industry, but it is extremely diffi-cult to do both under one administrative roof.Companies that work in both sectors typically haveseparate divisions that are organized, administered,and staffed differently. In that case, the two divisionsusually cannot share staff, production lines, labora-tory facilities, data, research, accounting procedures,and other administrative systems.

3. These differences are profound. In large aero-space companies, for example, the commercial sideof the firm responds to market conditions, whereasthe military side of the house responds to the natureof the threat, to government directives, and to thefederal budget. Executives and engineers transferredfrom a military to a commercial division oftenexperience a prolonged period of culture shock andsome are unable to make the adjustment. This is theopposite of the situation in civilian industry, wherethe chief executive officers of Fortune 500 compa-nies are increasingly interchangeable.

4. Many entrepreneurial civilian companies—large and small alike-are unable and/or unwillingto conduct business with DoD because of the heavyinvestment and reorientation in business practicesnecessary to meet DoD requirements.

5. Antitrust policy and a rigid regulatory framew-ork in some high-technology areas is adverselyaffecting the competitiveness of U.S. industry.Faced with foreign competition—specifically, withgovernments that act to create advantages for theirfirms in the U.S. market-U.S. firms may well failto compete successfully with foreign businesses thatare presently gaining experience in these areas.

POLICY CONSIDERATIONSFOR CONGRESS

The policy discussion that follows is divided intotwo areas. The first deals with the question ofmilitary dependence on foreign civilian technol-ogy. It is addressed because an increasing number ofdomestic high-technology industries that are impor-tant to the military are losing technology leadershipand market share to foreign competition. Themilitary response has been to buy materiel fromhigh-technology firms located abroad. The secondarea focuses on the problem of inadequate militaryaccess to civilian technology in the United States.In many industries, the military could improve itsaccess to civilian technology substantially, butCongress would have to make changes in theprocurement system to stimulate DoD demand forcivilian products and to make it easier for civiliancompanies to do business with DoD.

Military Dependence on Foreign Technology

The U.S. strategy of developing and fieldingbetter military technology than that of potentialadversaries requires that the DoD have access tomany technologies that are sold primarily in civilianmarkets. The technologies of microelectronics, forexample, and those contemplated for optoelectron-ics, require enormous and continuous investments inR&D and production facilities. These investmentswill be made only by companies that expect to sellthe resulting products in a civilian market that ismany times larger than defense purchases. Thatmarket is generally one that innovates more rapidly,because intense competition has compressed theproduct life cycle, forcing the incorporation oftechnological advances at the earliest possible date.In most cases, DoD cannot afford to pay the priceassociated with this kind of R&D and is unable toinduce private industry to develop it with thepromise of future orders. Therefore, if the Defense


Department wants to acquire state-of-the-art tech-nologies in these fields, it must purchase them fromcompanies that are producing for, and selling to, alarge civilian market.

Although it may be very important to U.S.national security that the Defense Department haveaccess to such technologies, DoD can do very littleto influence the location, ownership, capitalization,and fundamental directions of the commercial tech-nologies and industries that it needs. These aredriven, instead, by domestic and international mar-ket forces, financial opportunity, and the trade andindustrial policies of OECD and other producernations. The corporate structure that prevails willultimately be determined by fierce internationalcompetition for civilian markets and by the nationaltrade policies of self-interested individual tradingnations.

Two policy problems arise in this context. First,there are unmistakable signs that participation byU.S. companies in the high-technology sector of theinternational economy is weakening in importantrespects. Significant loss of capacity by U.S. compa-nies in these dual-use industries could ultimatelyundermine our basic military strategy of counteringsuperior numbers of enemy troops and equipmentwith superior technology. The second problem,which follows from the first, is that DoD isbecoming increasingly dependent on advanced tech-nology and products that are developed abroad or byforeign-owned companies located in the UnitedStates. Although DoD does not maintain systematicrecords on the amount of foreign content in U.S.weapons systems, military and civilian officialsagree that it is significant and growing. Failure toaddress these problems would eventually leave theU.S. military vulnerable to the self-interested ac-tions of other nations upon whose technology theU.S. may depend.

There are no easy solutions. DoD is a small andrelatively insignificant customer when compared toaggregate consumer demand in most high-technology, civilian-based industries. Accordingly,a requirement that DoD systems contain productsmade only by U.S.-owned companies located in theUnited States would exert little influence on theinternational marketplace. Instead, such a mandatewould be likely to produce a number of unintended

and dangerous consequences. It would, for example,limit access to advanced foreign technology, heigh-ten tensions with our allies and trading partners, andcreate financial and administrative havoc in thedefense sector of the U.S. economy. The resultingindustries would have small, assured markets, andaccordingly, little incentive to press toward thecutting edge of international technological competi-tion.

If Congress wishes to address this problem, it willfind that the issues of eroding dual-use industriesand military dependence on foreign firms extendwell beyond the purview of the DoD as well as thejurisdictions of the Armed Services committees ofboth Houses. This is because the environment inwhich policy must operate is the civilian sector ofthe international high-technology economy. DoDdoes not have the competence, the resources, or thepolicy levers to approach the situation in a compre-hensive manner. At best, it can attempt band-aidsolutions, such as funding Sematech and researchinto high definition television. What DoD mightusefully do is to help establish parameters for thekinds of dependence that would be more or lessacceptable, even if it cannot take meaningful steps toaddress the underlying causes of foreign depend-ence. Similarly, the Armed Services committees areconstrained by custom and by the limits of theirjurisdictions. As a result, few if any policy choicesare available without the cooperation of the tax andtrade policy committees which most affect the rulesby which companies compete in America. Neverthe-less, the nation and the Congress have a nationalsecurity interest at stake, because it is unlikely thatDoD, acting alone, will be able to keep pace withworldwide technological developments and avoidforeign dependence.

Policy makers will have to start with a clearconception of what a U.S. corporation or industry is,and will have to distinguish between the question ofownership and that of location of manufacturingfacilities. Figure 10 indicates four different combi-nations, each of which suggests the need for adifferent policy response. In case 1, the simplestcase, a corporation or industry is largely owned byU.S. interests and conducts most of its R&D andmanufacturing operations in the United States. Froma military perspective, this is the ideal situation, andit was, in fact, a characteristic condition in the


immediate post-WWII era. At the other extreme,case 4, a corporation or industry is owned by foreigninterests and is physically located abroad. Here, aflexible policy is needed that can calibrate thedegrees of risk associated both with different sup-plier nations and with specific technologies ofmilitary significance. It would, of course, be neces-sary to make adjustments as international tradingand military relationships evolve.

Cases 2 and 3 suggest intermediate possibilities.From a military perspective, case 2 (U.S.-based andforeign-owned) generally represents the more ac-ceptable condition, because the corporation andmost of its employees would be subject to the lawsof the United States and could be required to givepriority to U.S. national security needs in a crisis.But from an economic perspective, creating incen-tives for foreign-owned companies to locate manu-facturing and R&D facilities in the United Statesmight enhance or detract from the competitivenessof U.S.-owned companies with plants in foreignnations. There are competing interests at stake inthese situations, but it is not necessarily a zero-sumgame.

It is necessary, moreover, to distinguish betweencompanies that merely assemble parts into finishedproducts and those that actually conduct R&D andmanufacturing operations in the United States. Theformer contribute far less to the U.S. defensetechnology base. In addition, DoD may need accessto R&D and manufacturing facilities. While manyanalysts contend that the most efficient businessescolocate R&D and production facilities, the two arequite clearly separable. For example, Honda makesand sells cars in the United States but conductsresearch in Japan. And Coming Glass Works manu-factures optical fiber in Australia but designs it in theUnited States. It would be important to consider thatforeign-owned firms might produce less advancedtechnologies in the United States, saving the leading-edge R&D and production for their home bases. Carewould have to be taken in connection with anypolicy that seeks to encourage foreign-owned firmsto establish R&D and production facilities in theUnited States.

To the extent feasible, DoD would naturallyprefer to minimize foreign dependence in dual-useindustries that are important to the military (as in

case 1). However, doing so is not a simple matter; itwould require a variety of congressional actions, theconsequences of which would extend far beyondtheir impact on foreign dependence for defensetechnology. Therefore, Congress will want to takemany economic security considerations into accountin considering whether or how to act.

The basic strategy for minimizing military de-pendence on foreign technology would be to extenda variety of incentives to U.S. companies to carry outR&D and manufacturing in the United States. Asecond, and perhaps complementary, course wouldbe to establish incentives and sanctions to encourageforeign (and U. S.) companies to locate their R&Dand manufacturing facilities in the continental Uni-ted States (case 2). In high-technology industrieswhere these two approaches might be unavailable, arealistic policy would rank technologies (accordingto military necessity) and countries (according togeopolitical factors). It would then be necessary toproceed on an industry-by-industry basis, weighingthe risks of foreign dependence against the cost andfeasibility of maintaining a particular capability inthe United States. A policy framework for each ofthe four cases is shown in figure 10.

Case 1: If the goal is to promote military security,then it makes sense to establish policies to enhancethe dual-use portion of the defense technology basethat is U.S.-owned and located in the United States.The question of foreign dependence arises when acritical industry is failing in the United States orwhen U.S. companies fail to enter the competition ina particular technology at all. The policy problem ishow to stimulate and otherwise assist corporationsthat produce technologies and products that are (orcould be) important to the U.S. military. Policies thatare intended to improve the defense capacity of theselargely civilian companies may simultaneously af-fect their economic competitiveness. What is goodfor the military may or may not contribute to thehealth of any particular dual-use industry. It may benecessary to set up one group of institutionalmechanisms to assist U.S. dual-use corporations andanother to enhance military access to, and procure-ment of, technology and products developed in thecivilian sector of the economy. These mechanismsare discussed at the end of this chapter.

Chapter 9-Civilian Technology and Military Security . 181

Figure 10-Military Dependence on Foreign industry Ownership v. Location

OWNER SHIP

u.s. Fore i g n

u.s.

LOCATION OFMAN U FACT U R i N Q /

R&D CAPACITY

For e i g n

1. Pro motes

mi I i t a r y

s ecu r i t y

3. Risk depends

on speci f ic

technologies

a n d n a t i o n s


Case 2: There will be some dual-use industriesof military significance that are located in the UnitedStates, but which are partly or largely foreign-owned. The advanced composites industry is a goodexample. Roughly half of these companies areforeign-owned, but their R&D and production facili-ties are mainly based in the United States. Thiscritically important technology is enabling for manydefense aerospace applications. DoD has tended notto discriminate among such companies on the basisof national ownership, and has not promoted astronger U.S. presence in this industry. In mostcases, it makes little difference to the militarywhether or not a corporation or industry is domi-nated by foreign interests—so long as the criticalR&D capacity and production facilities are main-tained at state-of-the-art in the United States. Yet,distinctions would be necessary. For example,would it be acceptable for a company to manufactureproducts in the United States if all the machineryused in the plant were foreign-built? Similarly, incomplex weapons systems, how far down theparts-supplier chain should a requirement to manu-facture in the U.S. reach’?

2. Acceptab le , sub jec t

to U.S. priori ties

i n a crisis

4. L e a s t a c c e p t a b l e

i n terms of m i I -

i tar y security

(#3 also app l ies)

Case 3: Cases 3 and 4 can be combined, with thecaveat that case 3 companies (those that are U. S.-owned and foreign-located) would presumably bemore receptive to making concessions in the na-tional interest, although still subject to host countrycontrols on their operations. In general, thesecompanies do not contribute significantly to thedomestic defense technology base (although profits,if repatriated, may produce economic benefits for theUnited States). However, some U.S. companiesconduct R&D in the United States, but havemanufacturing and assembly operations overseas.Policy should be sensitive to this situation, acknowl-edging that these companies do make a contributionto the U.S. technology base beyond that of case 4 andsome case 2 companies. Policy may seek to removebarriers that impede competitive domestic manufac-turing.

Case 4: The defense implications of dependenceon industries that are foreign-based and foreign-owned are more complex. Policy will have to besensitive both to the geopolitical relationship be-tween the United States and the particular foreignnation and to the specific technologies under consid-


eration. Although it is an oversimplification, figure11 displays some of the factors that would have to beevaluated on a case by case basis. The United Statesmight, for example, tolerate foreign dependence forsome key technologies and products if they weremade in Canada, which shares technology base andfree trade agreements with the United States. Con-trast this with an extreme case, in which the U.S.military depends on a Warsaw Pact nation for atechnology that is enabling for a major weaponssystem. While it is easy to discriminate betweenfriendly neighboring countries and some EasternEuropean states, it is a matter of extreme delicacy toassess the security risk associated with technologicaldependence on a variety of nations, ranging from theEC to the Persian Gulf and the Pacific Rim. There isthe further consideration that, for some technologies,the United States might be forced to accept a foreignsupplier or do without. One alternative is to createthe capacity domestically, using grants, tax incen-tives, guaranteed low-interest loans, R&D contractswith the government, and other schemes—possibly,

but not necessarily, at expense to the taxpayer. Thereare also models from other nations, including Japan,where a portion of some domestic markets, bothcivilian and defense, is reserved for domestic firms.

Congressional Action andInstitutional Mechanisms

Some analysts believe that new institutions willbe needed to address these problems. They think itunlikely that Congress can effect such policies bydelegating the task of implementation to existingagencies; no agency presently has the necessarycapacities or powers. From this perspective, ifCongress is interested in pursuing a policy ondual-use technology and foreign dependence, itcould invest extraordinary powers and independenceof action in a high-level council or agency createdfor that purpose. Such an agency would take steps to:1) gather data on such essential items as foreigncontent in defense systems and foreign investment inhigh-technology companies, 2) assist U.S. dual-useindustries that are essential to U.S. military security,

Figure n-Military Dependence on Foreign Technology Located in Foreign Countries

GEO-POLITICAL FACTORS:

. a l l y

● s ustaine d supply

● cont iguous

TECHNOLOGY

FACTORS:

. stategic

. enabling

. pervasive

Possibly

acceptable

(-)

Probably

U n a c c e p t a b I e

(‘)

AcceptabIe

● noncr i t i ca l

● substitutes exist

(+). Iimited miIitary

applicationsProbably

accept a b I e

(-)

. hostiIe

● interruption of suppIy IikeIy

. distant



3) induce foreign-owned companies that are locatedin the United States to conduct R&D and manufactur-ing operations here as well, and 4) develop or attractan indigenous capacity for dual-use technologies inindustries where foreign dependence is unaccept-able, and the domestic private sector is unwilling orunable to withstand or enter the competition.

As an alternative, Congress may wish to considermandating coordination between existing agenciesand offices such as the Department of Commerce,DoD, and the United States Trade Representative.There can be no assurance that such an approachwould work. Each agency has its own establishedareas of business and expertise, and debilitatingbattles over leadership, functions, and turf could beexpected. Congress might, of course, opt to requirefurther studies of this problem and various ap-proaches to it.

If Congress decides that an institutional approachto this policy area is inadequate, an array ofstrategies, used by other nations with varyingdegrees of success, is available. These includeincentives and sanctions, both positive and negative.For example, Congress could require foreign-ownedcompanies to locate manufacturing and R&D ca-pacities in the United States if they intend to selldual-use, high-technology products in this country.In addition, Congress could substantially strengthenthe U.S. defense technology base through increasedfunding of graduate education for scientists andengineers, and by targeting the funds for Americancitizens, who are more likely to make their careers inthe United States. Congress might require a policy ofreciprocal dependence—for example, the UnitedStates might depend on Japan for DRAMs16 and inreturn, Japan would agree to depend on the UnitedStates for jet engines, with the intent of establishinga regime of equivalent dependencies. Or Congresscould change the structure of the tax system toencourage U.S. companies to make longer terminvestments, and change the tax code so that it nolonger favors speculative investment by increasingtaxes on short-term capital gains.

These options require painstaking analysis that isbeyond the scope of this particular assessment. Theyare raised here because they illustrate the point that

the solutions to the problems of eroding high-technology capacity and increasing military depend-ence, while critical to the national defense, falloutside the usual jurisdictions of the requestingcommittees of Congress. If the Armed Servicescommittees believe that there is a national securityinterest in conserving the health of the defensetechnology base in the United States, it may benecessary to reorient the way in which the businessof the committees is conducted—i.e., to focus lesson the internal structure of DoD and more on takingsteps to build a consensus within the Congress thatcan place these problems centrally on the nationalagenda.

Institutional and Administrative Barriers

Unlike the problems discussed above, the issue ofmilitary access to domestic technology fallssquarely within the purview of DoD and thejurisdictions of the Armed Services committees ofCongress. DoD and the military-industrial sectorhave become insulated from the rest of the economyin ways that tend to weaken military access toleading-edge civilian technology. This is largely theresult of a gradual accretion of regulations, auditingrequirements, paperwork, detailed specifications,and inefficient business practices that constitutesubstantial barriers between the military and civiliansectors of the U.S. economy. Most of these ruleswere instituted for good reasons and in response toreal problems. But the cumulative effect has been tomake defense procurement cumbersome and toconcentrate military buying in a relatively smallgroup of companies that have learned to conductbusiness according to government rules and norms.

If Congress wishes to improve military access tocivilian technology, it will have to make someextremely difficult choices. Congress has played anintegral role in establishing the structure of themilitary-industrial sector, as well as the rules andregulations under which it is run. To a large extent,Congress already approves or disapproves manyimportant decisions that affect the defense technol-ogy base in the United States, The problem is thatmany isolated decisions and actions—taken not onlyby Congress, but also by DoD and the executive

16Dynmic r~dom.access IIlemO~ chips.


branch—have built up over time, and the resultingsystem is a patchwork of conflicting requirementsand goals. To remove barriers that impede militaryaccess to the civilian sector, Congress will have tobalance competing interests—many of which aresupported by basic notions of fairness and othervalues that have shaped the present system. Themost important barriers are outlined below, togetherwith discussion of the relevant policy choices andproblems.

Procurement Reform

In recent months, interest in the defense procure-ment process has focused on allegations of criminalactivities by officials in the Pentagon and amongsome of the large defense prime contractors. Whileit is important to discourage such misconduct, anexclusive focus by the 101st Congress on issues ofmalfeasance will do little to address underlyingstructural problems that inhibit DoD’s access toadvanced civilian technology. Indeed, if Congressmandates several new layers of regulation andauditing in response, it may inadvertently createadditional barriers.

Many civil-sector companies are already reluctantto bid on contracts with DoD because they are notorganized and staffed to comply with the FAR, andbecause they do not need government business.Generally speaking, DoD has structured its procure-ment process to deal with corporations that areprimarily or exclusively engaged in work for themilitary. These regulations inhibit access to compa-nies whose technology and business is largelyresident in the civilian sector of the economy. Oneresult is that DoD is often forced to pay a premiumfor the development of a range of technologies andproducts that already exist in the civilian sector.

If Congress acts to increase regulation and audit-ing requirements, it may be able to reduce theamount of fraud and misconduct within the procure-ment system. But in so doing, it will almost certain] yalso reduce the efficiency of existing defensecompanies, because they will have to increase theirpaperwork load and internal audits to meet the newrequirements. At the same time, tightening up thesystem to eliminate malfeasance would result in aneven more complex regulatory environment forcivilian firms, increasing the probability that suchfirms would not choose to work with the government

in the future. One alternative is to do nothing, whichis a possibility when combined with rigorousenforcement of criminal statutes already on thebooks.

The opposite course is also a viable policy option.Congress could take steps to reduce paperwork,regulation, and auditing, with the intention ofincreasing overall efficiency and DoD’s access tohigh-technology companies and products in thecivilian sector of the economy. Such action might ormight not result in an increase of fraud andmismanagement in the defense sector. But it mightexpand the interaction with civil-sector firms thatare now reluctant to do business with DoD. Someargue that the complexity of the defense regulatoryenvironment encourages defense contractors to findways to skirt the rules, simply because the prolifera-tion of such regulations over time has made itextremely difficult to conduct business in an effi-cient and rational manner. Congress will have to sortout these issues as it seeks a policy that discouragesmisconduct without increasing the barriers of ineffi-ciency and complexity that afflict the presentsystem.

If Congress is reluctant to act, it may wish to studythese problems in greater depth. To do so it couldestablish an independent commission to explorethe effects of: 1) reducing or expanding procure-ment regulations, and 2) exempting high-technology civil-sector firms from some procure-ment regulations. This commission would examinethe difficulties that civilian companies face in doingbusiness with DoD. It would evaluate the ways inwhich the procurement system itself mitigatesagainst military access to civil-sector technology.And it would weigh the costs and benefits ofexpanding or reducing procurement regulation. Acentral purpose would be to recommend changes inthe procurement system that would induce civil-sector companies to sell their products to DoD or tomodify them to meet DoD’s needs.

A complementary and probably subsequent ap-proach would be to create a market for civilianproducts in DoD by mandating a preference forcommercial items that are not developed undercontract with DoD, using simplified and expe-dited contracting and acquisition procedures. InDoD, commercial products are referred to as non-

Chapter 9--Civilian Technology and Military Security ● 185

developmental items (NDI) and commercial off-the-shelf. The purpose of this preference would be toensure that program managers give careful consid-eration to existing NDI/COTS technology beforebeginning new development. Despite repeated cabinet-level memoranda and executive direction to thecontrary, DoD rarely substitutes commercial prod-ucts for milspec developmental items. This is duepartly to a historical bias in favor of contracting fordevelopment, partly to a reluctance of programmanagers to risk using items not designed to militaryspecification, and partly to the regulatory structurethat governs the procurement process. If Congressdecides to increase government access to civilian-based technology and products, it will have to usethe blunt instrument of a direct and unambiguouslegislative mandate to overcome a large measure ofresistance within DoD and the military sector of theeconomy.

It is likely that any reform of the procurementprocess that encourages the use of civil-sectortechnology and products will also require fundamen-tal changes in the relationship between DoD andsome of the firms with which it does business.Civilian sector executives who do not need DoDbusiness are unlikely to tolerate the heavy hand ofgovernment regulators and auditors changing theadministrative policies and practices of their compa-nies. Regulated profits, detailed military specifica-tions, set-asides for small and minority firms,affirmative action, specialized auditing procedures,and competitive bids from sub-tier suppliers maysatisfy government notions of fairness and socialresponsibility. But when they are imposed oncommercial operations, they tend to weaken theefficiency and competitiveness of companies thatdepend on sales in consumer markets for survival.These companies are in business to make a profit.Their managements are responsible to stockholders,who expect shorter-term results than are typicallyenvisioned by government programs or internationalcompetitors. Congress will have to make allowancesfor the fundamental differences in the way businessis conducted in the defense and civilian sectors of theeconomy, if it wants to increase military access toadvanced, commercially based technologies andproducts.

Specifications Overhaul

In general, the DoD specifications process is toocumbersome and too rigid to establish reasonableand realistic standards for technologies that arechanging rapidly. Too often, prescriptive standardsare mandated where performance specificationswould be more appropriate. Although they arenecessary if DoD is to conduct procurement at all,specifications can lock the military into a develop-mental mode and block access to existing civil-sector technologies, products, and systems. TheOTA case studies found instances where the militarycould not use existing superior civilian technologybecause DoD regulations mandated compliancewith outmoded specifications. Civil-sector compa-nies, military prime contractors, and milspec spe-cialists in the Services all agree that there are toomany specifications, referencing too many addi-tional documents, and that there is no effectiveprocess for eliminating outdated and unnecessarydocumentation. Military specifications can exert thelargely unintended effect of creating divergencebetween military and civilian applications of thesame technologies.

If Congress wishes to address this problem, thereare a number of steps it could consider. In high-technology fields where civilian products and tech-nologies are clearly at the leading edge, Congresscould require the harmonization of militaryspecifications with best practice in the civilian,high-technology sector. Such a requirement wouldapply to dual-use technologies where meeting ad-verse training and battlefield conditions is not atissue. In its study of fiber optics and software, forinstance, OTA found numerous examples where themilitary could benefit from adopting best practice infast-moving civilian technologies. Such action mightencounter resistance from quarters within DoDwhere there is a strong belief in the necessity ofdesigning systems specifically to meet user require-ments and battlefield conditions. Nevertheless, wherethe underlying technologies are similar in thedefense and civilian sectors, great cost savings mightbe realized by tailoring military specifications totake advantage of existing civilian products andapplications —as opposed to designing divergenceinto the specification and codifying it.


In addition, Congress may wish to mandate theuse of performance-type specifications in fast-moving, high-technology fields. The alternative isto nail down prescriptive standards that tend tofreeze military systems at or below present-daylevels of technological development, Performancespecifications have the advantage of introducing ameasure of flexibility into the system itself, and intothe procurement process, that could substantiallyenhance military access to advanced technologies inthe civilian sector. They could be written broad] y toencourage substitution of NDI or COTS productswhere possible,

It is unlikely that specific policy options can beimplemented until the system of writing militaryspecifications and the culture that sustains it arechanged. With this caveat in mind, Congress mightrequire the review and consider the revision ornullification of existing prescriptive specifica-tions at some specified interval. The purpose ofsuch a review would be to make the process ofwriting specifications more responsive to techno-logical advances in the civilian sector, and to doaway with the proliferation of unnecessary oroutdated requirements and documentation. By thetime they are written, many military specificationsno longer reflect the state of the art in dual-usetechnology industries. This is partly because theproduct cycle is so much shorter in the civiliansector. In order to reduce the multiyear process ofspecification writing, it may be necessary to intro-duce greater flexibility by reducing inter-Servicecoordination and permitting the different Services touse different specifications to meet their specificneeds.

Restructuring Data Rights

It is extremely difficult to strike an equitable andappropriate balance in allocating data rights incontracts between the government and the privatesector. Government agents tend to demand as manydata rights as they can get in any given contractbecause they are under a fiduciary obligation toprotect the interests of the government. In manycases, the contractors lose most, if not all, of theirintellectual property rights to the technology andproducts they develop. DoD negotiators typicallydemand the right to duplicate, use, and disseminatesuch data without restriction. The claim to unlimited

data rights is based on three important considera-tions. The first is that by securing complete datarights, DoD is in a position to foster competition bysharing the data among potential contractors. Sec-ond, unlimited rights protect the government fromfuture costs and claims of infringement, and help toensure dissemination of publicly sponsored researchefforts. And third. full data rights to software ensurethat DoD will be able to modify and maintainsoftware in the field.

This orientation contrasts sharply with practice inthe civilian sector, where R&D and process data arecarefully guarded and no company could expect anyrights to another’s proprietary information. Manyexecutives of civil-sector firms believe that govern-ment procurement officers and regulations do notrecognize the extent to which high-technologyindustries are driven by R&D activity. These firmsare typically unwilling to share data, because theybelieve it can be used to reveal a core of proprietaryinformation. In some cases, software and fiber opticscompanies invest tens of millions of dollars todevelop a process or series of products. Conse-quently, these civil-sector companies are unwillingto contract with DoD, because it insists on extensivedata rights and may even set a competitor up inbusiness.

In the case of software, the most recent directivesand regulations enable DoD to accept limited rightsto intellectual property. For software developedwholly with private funds, the contractor can negoti-ate restricted data rights giving the government theability to modify software and make backup copies,but allowing the developer to incorporate a typicallicensing agreement. The government hopes toretain the ability to maintain its software systemsand to ensure that future maintenance and reprocure-ments will be competitive. Despite these acknowl-edgments and the flexibility granted governmentcontracting officers to negotiate less-than-exclusiverights to data, DoD still insists on full transfer of datarights in most cases.

If Congress wants to increase government accessto civilian high-technology firms, it will have toreevaluate the principle of exclusive and unlimiteddata rights. In part, this may require that DoDdistinguish companies that have developed a prod-uct or process with private funds from those that


have done so under contract with the government or private sector. Congress may wish to review existingwith funds that are reimbursed by DoD. In addition, data rights policies and procedures mandated inter-Congress might require DoD to create different nally by DoD directives, and to assess the level ofcategories of data rights—ranging from the unlim- training that would be necessary to enable contract-ited to the narrowly specified-when it is buying ing officers to negotiate data rights and still protectfrom firms that do most of their business in the the legitimate interests of the government.