The DoD Acquisition System - princeton.edu

Appendix A

The DoD Acquisition System

CONTENTSPage

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Costs of Delay . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Growth in Acquisition Times . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Persistence of Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

PRIVATE SECTOR ACTIVITIES: MODELS FOR DoD ACQUISITION? . . . . . . . . . . .Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Differences in Mission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Differences unaccountability and Oversight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Differences in Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Differences in “Market Forces” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

EFFICIENCYV.EFFECTIVENESS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Level of Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . .Responsibility for the Defense Industrial Base . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Low Production Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Unpredictability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Unacceptability of Failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..

THE ACQUISITION PROCESS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .The Defense Acquisition Executive and Defense Acquisition Board. . . . . . . . . . . . . . . .Program Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .The Planning, Programming, and Budgeting System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Milestones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Appropriateness of Oversight . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... . . . . . . . . . . . . . . . ... . .

ANALYSIS OF THE ACQUISITION PROCESS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Program Variability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Requirements Generation and the Resource Allocation Process . . . . . . . . . . . . . . . . . . . . .Bureaucratic Paralysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Personnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Boxes

3334555777888899999

10111313151525273941

Box Page

A. Concurrency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16B. Multiyear Budgeting and the U.S. Constitution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

FiguresFigure Page

A-l. Cost Growth in Major Projects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6A-2. Components of Defense System Acquisition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13A-3. Life Cycle of Major System Acquisitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14A-4. Chains of Authority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31A-5. Cost v. Regulatory Intensity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

TableTable Page

A-1. Defense Enterprise Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

.

Appendix A

The DoD Acquisition System

INTRODUCTION

The time needed to field new technology dependscritically on the time taken to complete the acquisi-tion cycle, from defining a requirement to deployingoperational hardware. Understanding how new tech-nology becomes incorporated in operational forcesrequires first understanding the acquisition process,which is the subject of this appendix. This appendixis the basis for the concluding section (The DefenseAcquisition System”) of chapter 8 of the mainreport.

The purpose of this appendix is to analyzeacquisition delays that in turn, delay the introduc-tion of new technology into the field. But, since nosingle aspect of the acquisition process by itselfcauses delays, shortening the cycle requires makingthe entire process more efficient and effective.Therefore, the discussion of acquisition in thisappendix takes a broad view.

After first describing the consequences of longacquisition lead times, Appendix A summarizessome studies that have tried to measure the problem.These studies show that the acquisition cycle is notsignificantly shorter now than it was several decadesago, and in some respects maybe getting longer. Thefact that the problem of a long acquisition cyclepersists-even with many studies over time identi-fying many of the same difficulties-is noteworthy.

In light of these findings, the discussion turns tothe issue of why defense acquisition problems are sodifficult to solve. In particular, it examines therelevance of private sector acquisitions to theDepartment of Defense (DoD) environment, in viewof the marked differences between the two ap-proaches. The remainder of the appendix thenanalyzes the major defense acquisition problems—together with proposed solutions-that precedingstudies have uncovered.

Costs of Delay .

For years, defense analysts have been frustratedwith the length of the acquisition process. Delays inacquisition lead to lost time in fielding new systems,which threatens our technological lead over theSoviets. Also, the expense of maintaining extendeddevelopment efforts leads to higher costs. Evenmore serious than the increased time and cost,according to a Defense Science Board (DSB) panel1

that studied the acquisition cycle over a decade ago,are the “second order effects’” of delays:

●

●

●

●

unsatisfactory results, with systems technologi-cally obsolete by the time they are fielded;increased technical risk, since system designersattempt to stretch the state of the art as far asthey can to avoid such obsolescence;loss of flexibility, since the extended approvalprocess makes it difficult to change the designof a system to meet changing perceptions of itsneed; andadded complexity, because delays aggravatethe tendency to want systems to do “every-thing.”

Finally, delays lead to even further delays. Ac-cording to the DSB panel, attempts to fix theproblem by requiring earlier and more frequentreviews only serve to extend the front end of theprocess and make the problem worse. Delays alsotend to be self-reinforcing. Costs escalate. forcingprograms to stretch out to stay within annualbudgetary ceilings. As the expected time for deploy-ment lengthens, planners magnify the anticipatedthreat, up the system’s requirements, and therebyextend development times further.

Growth in Acquisition Times

Several studies of acquisition lead times havetried to determine how serious delays are. andwhether the problem is getting worse. These studies.some of which are summarized in appendix B, havegenerally examined major aerospace systems. To the

-3 -

4 ● Holding the Edge: Maintaining the Defense Technology Base, Volume 2

extent that the acquisition process and/or the level oftechnology of aerospace programs is typical of othermilitary systems, their results may be generalized.The studies indicate that the development cycle—especially the portion preceding full-scale develop-ment (FSD)-has increased somewhat over time.However, this increase accounts for only a smallportion of the program-to-program variation indevelopment time over the years. Moreover, as oneof the studies points out, increases in the pre-FSDphase “should not automatically be deemed undesir-able,” since these increases were consistently ac-companied in the study data by reductions in costgrowth, schedule slippage, and performance short-fall. 2

These studies found that, once decisions had beenmade and approvals to proceed had been given, thefull-scale development period did not generallylengthen. Based on this finding, one can concludethat increases in technological complexity have not,in and of themselves, extended hardware develop-ment. However, production times are increasing asbudgetary limitations, coupled with increasing unitcosts, reduce the numbers of units purchased peryear.

Comparing the increases in decision time to therelative constancy in the length of hardware devel-opment, the 1977 DSB study concluded that “itdoesn’t take any longer to do something; it just takeslonger to obtain the necessary approvals and acquirefunding to do it and to get to the deployment stateonce the development is finished.”3 More recentstudies have-corroborated this finding.

Persistence of Problems

Making the acquisition process more efficient andeffective will not be simple, as the Acquisition TaskForce of the President’s Blue Ribbon Commission

on Defense Management (the Packard Commission). stated in 1986:

. . . present procedures are deeply entrenched. Ac-quisition problems have been with us for decades,and are becoming more intractable . . . In frustration,many have come to accept the ten-to-fifteen yearacquisition cycle as normal, or even inevitable.4

Shortly afterward, the Center for Strategic andInternational Studies (CSIS) concluded that:

. . . the process is in serious trouble . . . [it] hasbecome overburdened with exorbitantly expensivemanagement layers, excessive delays in programdecision approval, inordinate program changes, andcumbersome oversight and regulations

These pessimistic outlooks are not particularlynew. Studies over the previous decades have identi-fied many of the same problems and even proposedmany of the same solutions. However, these solu-tions have not been implemented-or not suffi-ciently to keep the same problems from cropping upin the next study.6

The possibility certainly exists, of course, thatnone of these studies identified the real problems,which therefore remain to be addressed. Alterna-tively, perhaps sheer intransigence and bureaucraticinertia within DoD keep it from substantially im-proving its operation.

More likely, however, is that many difficulties indefense acquisition stem from factors that arebeyond the DoD’s direct control and that no amountof unilateral activity can address. To the extent thatsuch external factors dominate, improving de-fense acquisition will require large-scale struc-tural and institutional changes that would not berestricted to the Defense Department.

Some of these changes are impossible within ourpresent system of government. Others would inter-

.

Appendix A-The DoD Acquisition System ● 5

fere with different objectives that the Nation has sofar-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.

PRIVATE SECTOR ACTIVITIES:MODELS FOR DoD ACQUISITION?

Studies

Given the premise that the private sector canaccomplish tasks more efficiently and cheaply thanthe bureaucracy-encumbered Federal Government,previous studies have looked to the private sector toprovide a model for improving defense acquisitionand shortening the acquisition lead time. In somerespects, defense acquisition compares quite favor-ably to private sector activities. Figure A-1, showingaverage cost growth of major weapons systemsalong with that of large, complex civilian projects,shows that in general the major weapons systems doquite well by this measure. However, this compari-son does not address acquisition time or acquisitionprocedures, which studies and reports have assertedare far more complex and time-consuming in thegovernment than in the private sector.

The 1977 DSB summer study on the acquisitioncycle concluded that the acquisition cycles ofcommercial aviation programs, unlike those ofdefense systems, had not lengthened over thepreceding two decades. The panel attributed thestability of commercial acquisition programs to theirsmaller technical steps, a greater degree of concur-rency between development and production, andcompetitive market forces that place a premium ontimely delivery.

A later DSB summer study also compared DoDacquisition programs with programs of similarcomplexity and size in the private sector, The finalreport of this study7 qualitatively discussed differ-ences in program structure and management be-tween the commercial programs and various DoD

programs, but it did not provide a quantitativecomparison. Nevertheless, the Packard Commis-sion—perhaps relying on interim results or personalcommunications with study panelists-representedthe DSB study as concluding that each of thecommercial programs “took only about half as longto develop and cost concomitantly less” than equiva-lent DoD programs.8 With this premise, along withits own analysis of successful DoD programs thatwere developed under special streamlined condi-tions, the Packard Commission concluded that“major savings are possible in the development ofweapon systems if DoD broadly emulates theacquisition procedures used in outstanding commer-cial programs."9 The study went on to characterizethose features of successful commercial programsthat could be incorporated into defense acquisition:clear command channels; funding stability; limitedreporting requirements; small, high-quality staffs;communication with users; and prototyping andtesting.

While there are certainly lessons that the privatesector can offer the Federal Government-lessonsthat the Packard Commission sought to uncover—fundamental differences between the governmentand the private sector must be grasped before any ofthese lessons can be applied.

Differences in Mission

Private industry exists to make money. Althoughit is too simplistic to assert that the bottom linedominates ail corporate activities-indeed. compa-nies respond to a range of interests and motivationsthat are not adequately described by focusing on anyone measure-the premise underlying the genera-tion of capital for use by industry is that suchinvestments will be profitable. Even if profit andreturn on investment are not the only relevantindicators, at least they are quantifiable measures ofcorporate performance.

On the other hand. the government’s mission ofproviding services such as maintaining the commondefense has no corresponding measure-at least in

7Defense Science Board, “Report of the Defense Board 1985 Summer Study on Practical Functional Performance Requirements,” prepared for theOffice of the Under Secretary for Research and Engineering, March 1986.

8 President’s Blue Ribbon Commission on Defense Management, Op. Cit., footnote 4. p. 11.

9 lbId., p. 12.

6 ● Holding the Edge: Maintaining the Defense Technology Base, Volume 2.

Figure A-l —Cost Growth In Major Projects

Program cost growth (percent)

2 0 0

150

100

50

0 —1960s 1970s Highway Water Public very Non-pioneer Pioneer All

projects projects buildings large processMajor weapon (1972) (1972) (1972) construction plants

systems (1977) (1985)

SOURCE. Michael Rich and Edmund Dews, “Improving the Military Acquisition Process-Lessons From RAND Research” (R-3373-AF/RC), A Project AIRFORCE Report prepared for the United States Air Force, February 1986, p. 11

peacetime. As Secretary of Defense Weinberger ment seeks to satisfy. For example, pursuit ofstated in his fiscal year 1985 Report to the Congress, objectives such as fairness, environmental protec-“We can never really measure how much aggression tion, equal opportunity, and maintenance of Amer-we have deterred, or how much peace we have ica’s economic base may conflict with the ability topreserved. These are intangibles-until they are acquire defense systems efficiently. With govern-lost.” l0

ment purchases of goods and services constituting

Moreover, just as there is no single measure of close to a tenth of the 1987 gross national product ofgovernment effectiveness. neither is there a single— $4.5 trillion, the Federal acquisition budget offersor even a consistent—set of objectives that govern considerable leverage for achieving national socio-


economic goals—leverage that legislators have notrefrained from using.11

Differences in Accountability and Oversight

Government and industry have very differentrelationships with” their sources of funds. A taxpayeris not the same as a shareholder. The two have verydifferent attitudes regarding the uses to which theirfunds are put, attaching to the expenditure of publicfunds a sensitivity-and a standard of accountabil-ity—that goes beyond business accounting prac-tices. Consider how easy it is to set compensation fora member of a corporate Board of Directors com-pared with that for a Member of Congress.

One of the most significant differences betweengovernment and private sector activities, the role ofCongress, has no parallel in the commercial world.Congress serves as the “court of last resort” forsocietal issues that cannot be resolved in any otherforum. Issues of high congressional visibility are bydefinition controversial, and it is unrealistic toexpect the political process by which these issues areresolved to proceed efficiently or directly. Thisprocess, and the annual budget cycle, will alwaysintroduce an uncertainty to defense acquisition thatcommercial programs do not share. As JamesSchlesinger, 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.12

Moreover, the U.S. Constitution gives Congressspecific responsibilities with respect to defense thatextend above and beyond its involvement in other

government activities, making defense activitieseven less comparable than other government pro-jects to private sector activities.

Differences in Size

● Although individual projects in the commercialsector may rival individual defense programs in size,the DoD as a whole is orders of magnitude largerthan most commercial enterprises. The DoD budgetfor fiscal year 1989 was roughly equal to thecombined annual sales of the top four Fortune 500fins-General Motors, Exxon, Ford, and IBM.

Bureaucratic complexity increases geometricallywith size. leading to inherent inefficiencies of scale.To some extent, these inefficiencies are unavoid-able. All military/defense work (except that involv-ing nuclear weapons) has been centralized in a singleDepartment of Defense precisely so that all claimson defense dollars could compete against oneanother. One possible way to reduce bureaucracywould be to fence off elements of what is now theDoD budget, eliminating them from the competi-tion.13

This situation existed before 1947, when theDepartment of War and the Department of the Navywere two independent Cabinet-1evel departments.Combining them into a single Department of De-fense made it harder for each component to operate.but in theory the combination benefits the taxpayerby permitting the allocation of available funds wherethey can most effectively be used.14

Differences in “Market Forces”

Ideally in a free market, competition among firmsrewards the most efficient ones and penalizes theunsuccessful ones. Duplication of effort-i.e. com-petition-ultimately serves to improve the quality ofthose who survive. The Federal Government. on theother hand, is a monopoly; there is only one


Department of Defense. To prevent duplication, theroles and missions of DoD components are designednot to overlap. There are, therefore, no correspond-ing “market forces” that provide built-in incentivesfor DoD to improve its performance. Moreover,DoD cannot sell off or disband a military service oragency that does not perform as well as hoped.

EFFICIENCY v. EFFECTIVENESSDefense 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.” 15 The nature of defense acquisitionimposes specific requirements that go even beyondthe disincentives to efficiency facing governmentactivities in general.16

Level of Technology

Much of the technology used in defense systemsis ahead of that in the commercial sector—if indeedany commercial 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-militarytechnology must nevertheless often be developedfrom scratch for a relatively limited production run.

Responsibility for the Defense Industrial Base

Since the Department of Defense is the onlycustomer for sophisticated military systems, produc-ers do not have the option of selling elsewhereshould they not be able to sell to DoD.17 If the

Defense Department wants to maintain a diversity ofsuppliers, ‘t must buy enough from each of them tokeep them in business, even if their products may notbe DoD’s first choice. As analyst Edward Luttwakhas put it:

When I go shopping for shoes, I can select themon the basis of price and quality. I need not buy moreshoes than I want simply to keep shoe-productionlines open. Nor do I have to ensure that this or thatshoe manufacturer has enough profit to pay for thedesign of new shoes. Above all, I have no reason topay more for my shoes to ensure that there is sparecapacity in the industry, to meet a sudden need I mayhave for a hundred pairs of shoes instead of just one.Yet those are all key concerns for defense purchas-ing.18

In-depth examination of defense industrial baseconcerns is beyond the scope of this study. Recentstudies looking at the interrelationship between DoDneeds and policies and the viability of the defenseindustry have concluded that there is cause forconcern.19

Low Production Rates

Aggravating the problem of maintaining a viableproduction base are production rates lower thaneconomically optimal because the required invento-ries are small and must be divided among firms. Topreserve competition, the most efficient producercannot be permitted to drive the others out ofbusiness. Moreover, production rates are typicallydetermined by externally imposed budgetary limits,rather than being derived internally according towhat makes sense for the program.


Unpredictability

There is no way to predict how, where, or whenwar might break out. Procurement needs are there-fore impossible to predict, and shut-down of produc-tion lines is risky.

Unacceptability of Failure

The inevitable consequence of competition in thefree market is the risk of failure, which entrepreneurswillingly accept as the price for the chance to strikeit rich. In the commercial arena failure translates toloss of investment or to reduced earnings. Substan-tial failure on the part of DoD could have far moresevere consequences.

The Department must therefore tolerate a fargreater degree of redundancy and risk aversion thana commercial enterprise would. This degree of riskaversion should not apply to individual defenseprograms; indeed, lack of failures would indicatethat the overall program was far too conservative. Inthe aggregate, however, the Department’s attitudetoward risk must be substantially different than acorporation ‘s.

Summary

In light of the features 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 simplysecond-order expedients to paper over largely insol-uble first-order problems.”20

THE ACQUISITION PROCESSThe Packard Commission’s report was not the

frost attempt to apply lessons from the private sectorto defense management. Seventeen years beforechairing the President’s Blue Ribbon Commissionon Defense Management, David Packard (the Dep-uty Secretary of Defense) established the present

DoD acquisition process to emulate industrial prac-tices of project management and sequential reviewand approval. The basic process is one of distinctphases separated by decision points or milestones.The Office of the Secretary of Defense (OSD)develops policy for major system acquisition pro-grams and conducts reviews to ensure that theseprograms respond to specific needs and are managedsoundly. The military Services and defense agenciesindividually (for the most part) identify those needsand define, develop, and produce systems to meetthem.

The Defense Acquisition Executive andDefense Acquisition Board

Regulations issued by the Office of Managementand Budget and DoD have codified acquisitionprocedures: OMB Circular A-109, “Major SystemAcquisitions”; Department of Defense Directive5000.1, “Major and Non-major Defense Acquisition .Programs”; and various implementing DoD Direc-tives and Instructions. These regulations specify themilestones that major defense acquisition pro-grams-those exceeding certain budgetary limits orhaving particular urgency, risk, congressional inter-est, or other special significance-must pass. TheSecretary of Defense conducts milestone reviews ofthese programs, unless he delegates review authorityto a Service or agency head.

OMB Circular A-109 directs the head of eachFederal agency that acquires major systems to“designate an acquisition executive to integrate andunify the management process for the agency’smajor system acquisitions.”21 The role of DefenseAcquisition Executive is now assigned to the UnderSecretary of Defense for Acquisition [USD(A)], anoffice created on the Packard Commission’s recom-mendation to consolidate responsibility for DoDacquisition. The Deputy Secretary of Defense or theUnder Secretary for Research and Engineering hadserved as Defense Acquisition Executive prior to theestablishment of the USD(A). (The rationale for

10 ● Holding the Edge: Maintaining the Defense Technology Base, Volume2.

establishment of this position is discussed later inthis appendix, under “Bureaucratic Paralysis.”)

The USD(A) chairs the Defense AcquisitionBoard (DAB), a panel of senior defense officials thatassists the Secretary of Defense in determiningacquisition policy and making program milestonedecisions. 22 DAB replaced the Defense SystemsAcquisition Review Council (DSARC), which hadhad a similar function but a somewhat differentcomposition. Upon completing each phase of aprogram’s progress, DAB reviews parameters suchas cost, schedule, performance, and affordability.programs whose oversight is delegated to theServices or defense agencies follow a correspondingseries of milestone reviews at the Service or agencylevel. These service reviews are conducted throughService Systems Acquisition Review Councils, orSSARCs.

Program Management

The DoD acquisition process is based on theprinciple of Program Management, in which oneindividual-the program manager-is responsiblefor integrating in a single office the diverse adminis-trative, professional, and technical capabilities re-quired to manage the development and production ofa major system. This concept was first formalized—

at least within DoD—by the Air Force SystemsCommand in the late 1950s,23 although its basicstructure originated within industry.24 The otherServices have adopted some version of this process.

The size and organization of program offices vary.The larger ones are self-contained, containing up toseveral hundred personnel. Others have “matrix”organizations, in which a small core staff is dedicated to each program, while shared support organi-zations carry out most of the effort.

Under DoD Regulation 5000.1, individual pro-gram managers are to be separated from the USD(A)by no more than two intermediate managementlayers. Program managers are to be given “fullauthority to manage their respective programswithin the scope of established program baselines. ”However, besides the program manager and pro-gram office,

. . . there are many additional senior managers andorganizations who also have management authorityand responsibilities within the overall DOD systemacquisition environment. Programs do not belongexclusively to [program managers]. They are DODand service programs, and represent investmentdecisions by the [Secretary of Defense] and service


secretaries who are also accountable for their man-agement and decisions.25

Study after study has identified the separation ofresponsibility and authority—the control exertedover a program’s outcome by people and officeswho are not directly accountable for it-as amajor problem of the defense acquisition struc-ture. Analysts differ as to the degree to which powerand accountability can be brought back together inthe defense acquisition environment.

The Planning, Programming, andBudgeting System

One of the most important factors external to theprogram manager is the allocation of resources forthe program. Resource allocation throughout theDoD is conducted via the Planning, Programming,and Budgeting System (PPBS) instituted by RobertMcNamara. Prior to McNamara, the Secretary ofDefense exerted little control over the budgetsubmissions of the Services. The result was thatbudget decisions were largely independent of long-term plans, Service budgets were prepared inde-pendently of one another with little rationalizationacross Services. and the Secretary of Defense had noanalytic basis on which to challenge Service re-quests. The situation has been summarized as one inwhich “requirements planning was being donewithout explicit regard to cost, and budget planningwas being done without explicit regard to need. ”26

Although there are formal links between the two,the PPBS is separate from, and largely independentof, the systems acquisition system. In theory, PPBSis supposed to start with assumptions and projec-tions concerning national strategy and future threats(planning) and lead to definition and analysis ofalternative force structures and weapons/support

systems, including resource requirements (program-ming). These programs are then translated intobudgetary terms and submitted to Congress (budget-ing).

In practice, the process has never worked thisclearly. In particular, criticisms leveled at theplanning stage are that the absence of fiscal con-straint makes the process somewhat irrelevant, andthat planning often justifies desired force levels andnew systems after the fact, instead of forming theanalytical basis for setting those levels and initiatingthose systems. The programming and budgetingstages of PPBS, from which the actual fundingrequest and ultimately the funds themselves derive,have a more direct impact on DoD activities than theplanning phase.27

The relationship between the acquisition systemand PPBS has been compared to that betweencongressional authorizations (programmatic review)and appropriations (budgetary allocation). However,this analog fails to recognize that PPBS aloneintegrates programmatic and budgetary considera-tions.% A better model is that acquisition programsproceed along a “dual track.”

This relationship poses complications for pro-gram managers. Under the acquisition system, theyreport through at most two higher officials to theUnder Secretary for Acquisition. Under PPBS,however, their resources are justified through amuch more complicated chain of authority, involv-ing a systems command of their military Service, themilitary headquarters staff of the Service, thecivilian Service Secretariat. and the Office of theSecretary of Defense. The PPBS process is managedand overseen by the Defense Resources Board(DRB), which consists of most of the DoD’s most


senior officials and is chaired at present by theDeputy Secretary of Defense.29 The Defense Re-sources Board, like the Defense Acquisition Board,seines in an advisory role to the Secretary ofDefense, who has final authority over both acquisi-tion and PPBS activities.

The directive implementing the PPBS specifiesthat coordination between the acquisition processand the PPBS is achieved through common mem-bers of the Defense Acquisition and Defense Re-sources Boards and “by the requirement to developan acquisition strategy for all major systems.”30 Theacquisition strategy is the basis of a programmanager’s system acquisition plan. Various mile-stones identified in the acquisition strategy tie intothe PPBS process; approval will not be given for aprogram to proceed to a new acquisition phaseunless its sponsoring Service or agency has plannedfor the program in its budget request through PPBS.

Despite the ties between the two, the relationshipbetween acquisition and PPBS has been controver-sial, especially concerning which path should “havethe last word.” According to a House ArmedServices Committee report, Richard Godwin, thefirst USD(A), proposed that “once a decision todevelop or purchase a system had been made by theDAB it could not be overruled by the DRB,”enabling the acquisition organization to bypass thePPBS.31

This proposal was not accepted. On September 1,1987, Deputy Secretary of Defense William H. TaftIV issued a revised version of DoD Directive 5000.1stating that significant changes in approved majordefense acquisition programs could not be madewithout prior approval of the Defense AcquisitionExecutive (i.e., the USD(A)) “unless made duringthe course of the Planning, Programming, and


"32 Under Secretary God-Budgeting System process.win resigned 2 weeks later.

Milestones

Taken in total, the acquisition process consistsconceptually of the four activities shown in figureA-2. However, only the lower two-definition ofsolutions and production of equipment—are specifi-cally associated with new systems. The two upperactivities of assessing operational needs and advanc-ing the technology base are ongoing and largelysystem-independent.

The formal process for major defense acquisitionprograms-those expected to cross a preset dollarthreshold or otherwise qualify as described in thepreceding section, “The Defense Acquisition Execu-tive and Defense Acquisition Board”-is normallydivided into five phases delineated by distinctmilestones. These phases and milestones are dia-grammed in figure A-3 and discussed in more detailin appendix C.

The diagram and the description are idealized inthat they assume a progressive linear process inwhich each stage is completed satisfactorily beforethe next begins. In reality-no matter what theorganizational structure-activities in each phaseoverlap and interact. Research and development is arisky process. Not only are surprises to be expectedin utilizing new areas of technology, but they oftencrop up in what had been thought to be straightfor-ward applications of established techniques.

Appropriateness of Oversight

Through this series of milestone reviews, the OSDexercises oversight over major Service acquisitions.The degree of OSD oversight in the past has variedconsiderably and remains quite controversial. TheDSARC (now DAB) process was originally justifiedas a means of decentralizing decisionmaking bylimiting OSD involvement in major programs tospecified milestone reviews. However, the extensivebriefings required before DSARC meetings and the

Figure A-2-Components of Defense SystemAcquisition


and PPBS processes are reasonable with respect tothe provision of external review and control overacquisition program budgeting and management.”35

However, the study panel did identify weaknesses inthe process that lead to delays. In particular, itworried that annual budget deliberations, and theease with which unfavorable decisions in one venuecould be reopened in another, had unnecessarilyextended the “front end” of the cycle. Moreover, thestudy predicted that things were likely to get worseunless changes were made.

It is certainly true that the DAB oversight processand the PPBS process-both inherently bureau-cratic, both involving the participation of a greatmany people, and both having significant impact onprograms-pose problems for the program manager.Commenting on the 1977 study, one analyst agreedthat “the endless sequence of reviews, interventions,and delays caused by the struggle for access todecisions both within and outside the Pentagon is aprogram manager’s nightmare.”36 However, thisindividual disagreed with the DSB study’s particularrecommendations for streamlining the initial stagesof the cycle. Although “compression of the front endof the acquisition cycle would be a program devel-oper’s dream,” he argued, “to all the other partici-pants in the process-OSD, OMB, and the Con-gress-it would be a nightmarish return to all theevils which brought McNamara to inject OSDforcefully into the process in the first place.”37

ANALYSIS OF THE ACQUISITIONPROCESS

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 selected; bureaucratic paralysis; inap-propriate organization of the defense procurementsystem; and the quality of and incentive structurefacing acquisition personnel.

Program Variability

Constant changes in defense acquisition pro-grams-and the ensuing inefficiencies. cost in-creases, and delays they cause—have become therule. According to a study by the Center for Strategicand International Studies,

Few, if any, defense acquisition programs followeither the course for which they were originallyplanned or any other stable pattern of developmentor production. Many purists refer to this real-worldphenomenon as program instability—a term thatcaptures their frustration, but not the facts of thecomplex legislative/executive system.38

Program variability, the more appropriate term usedby the CSIS study, results from a number of factors:the requirements process; the risks inherent indeveloping new technology; the political/budgetaryprocess; and personnel turnover. While the disrup-tions introduced by these factors can be controlled tosome extent, their underlying causes cannot beeliminated.

Whereas the unwillingness to reexamine require-ments in the light of technological difficulties candrive up the cost and complexity of weaponssystems, changing requirements too frequently canmake sound management impossible. In the past,according to analyst Jacques Gansler, the militaryServices have “felt free to change their mindsfrequently” concerning the requirements and budg-ets for new systems.39 Now, in a process called“baselining,” internal contracts are developed be-tween program managers and the senior manage-ment of their Services concerning the cost, schedule.and performance milestones for new weapons sys-tems. Since changes to the baseline require equallyhigh level review, formalizing a baseline representsan attempt to reduce the amount of change thatprograms undergo within DoD.

In practice, however, baselining requires that theprogram manager have the authority to rejectchanges to his or her program that are imposed from


BoxA—Concurrency 1

One method that has been used to shorten acquisition times is to overlap some of the phases in the process,specifically y those of full-scale development and procurement. In highly concurrent programs, production starts wellbefore fill-scale development is completed on the assumption that, although changes will inevitably be made asdevelopment proceeds, it will be possible to accommodate these changes without disrupting production. Overall,a significant amount of time can be saved.

Besides shortening the time needed to field new systems, concurrency can in principle achieve cost savingsand management efficiencies because reduced development time means lower overhead and more continuity andstability in the labor force. Concurrency can also reduce program changes that would otherwise force cost increasesand delays. However, the principal risk of increased concurrency is that significant problems uncovered afterproduction has begun may necessitate major design changes, forcing extensive rework on completed systems. Thesechanges lead to cost overruns and schedule slippages, countering the very goal of concurrency in the first place. Ifadequate solutions cannot be found, the program must accept diminished technical performance or even facecancellation and the consequent writeoff of sunk costs.

Concurrency has historically been emphasized during wartime or periods of national emergency (e.g., depthcharges developed in World War 1, the Manhattan Project in World War II, the missile programs undertaken in the1950s, and’’ smart” weapons developed for use in Vietnam). But, until the 1960s, concurrency was rare in peacetimedefense acquisition programs. Since then, the practice has gone in and out of favor as the time savings have beenseen to outweigh, or conversely not to justify, the risks. Problems encountered with systems developed in a highlyconcurrent manner in the 1960s led DoD to establish a “fly before buy” system that emphasized prototypedevelopment and testing prior to production decisions. The 1977 DSB study nevertheless concluded that “the policyof ‘no concurrency’ is being applied too rigidly and is inefficient and costly in many cases.”2

Despite this recommendation, pressures against concurrency appear to be increasing after major problems wereencountered with two recent weapons systems, the B-l B bomber and the Division Air Defense (DIVAD) gun,developed in a highly concurrent manner. The Packard Commission has urged that prototypes be built and testedbefore full-scale development, let alone production, begins.3

The current regulatory and legislative environment provides no clear direction concerning concurrency. Whileexisting DoD regulations do not prohibit and in places encourage concurrency, legislation has constrained it. Onone hand, Directive 5000.1 states that “commensurate with risk, such approaches as. . . reducing lead time throughconcurrency. . . shall be considered and adopted when appropriate.”4 On the other, the 1987 Defense AuthorizationAct stated that “a major defense acquisition program may not proceed beyond low-rate initial production untilIOT&E (initial operational test and evaluation) of the program is completed.” This Act also stressed competitiveprototype development that will likely have the effect of inhibiting concurrency.5

In attempting to determine the effects of concurrency, the DSB study found that “there is no convincingevidence that concurrency necessarily adversely affects program outcome in terms of cost, performance, or fieldutility.” Therefore, the blanket ban on concurrency should be eliminated since “the acquisition time span . . . canbe minimized if concurrency is properly employed.’6 A Congressional Budget Office study 10 years later foundthat “no strong relationship exists between concurrency and schedule delay” but that “a modestly strongerrelationship exists between concurrency and cost growth.” The more highly concurrent programs experiencedhigher cost growth.7


sources outside the program. Granting this degree ofauthority would be extremely difficult within thepresent DoD environment. For example, althoughspecified in a program’s baseline, one of the mostimportant program parameters-its budget—is ulti-mately established by a PPBS system external to theacquisition process. Moreover, it is often changed(i.e., annually) by Congress.

Increased emphasis on technology demonstra-tions and prototyping can be expected to help controlprogram changes caused by technological risk. If,however, such demonstrations further extend theentire cycle, they could increase uncertainties due tochanges in the threat and in projected programbudgets.

Changes imposed on defense acquisition pro-grams by the political process-e.g., battles overprogram budgets, policies, and control-originate atevery level of activity within DoD, the executivebranch of government, and Congress. The keydifficulty here is politics-not in the pejorativesense of backroom deals, influence trading, and porkbarreling that the word has come to acquire. but inits true definition as a struggle between competinginterests. Examining major strategic weapons sys-tems such as the MX and Trident, analyst EdwinDeagle illustrates the larger context in which de-fense acquisition fits:

The MX and Trident are not simply expensiveprograms deserving of careful management. Theyare also: major commitments to specific solutions ofthe complex problem of strategic nuclear deterrence;affirmation of roles and missions within and amongthe military services; explicit choices about theimportance of strategic weapons relative to othermilitary activities: explicit choices about the impor-tance of strategic weapons relative to other publicinitiatives such as urban housing, national healthinsurance, energy security or middle class tax relief:and, by no means least commitments to particularcommercial enterprises which, as a result, willemploy people in specific places. In short, theseweapons systems and the R&D process which yieldsthem lie in the center of the competition amongvalues, purposes, and programs inherent in the

process of public choice—by nature a politicalprocess. Organizational and procedural imperativesdesigned to support this political process are likelyto be vastly different from, and perhaps in conflictwith, those designed to yield efficient manage-ment.40

Granted, the programs Deagle has chosen todiscuss are among the largest and most politicallyvisible of defense programs. Nevertheless, the pointhe makes-that political judgments are inherent inresolving competing demands on public resources—applies to all defense programs.

Even without political influences, change isinevitable:

Development will always be difficult, uncertain,time consuming, and more expensive than expected.Threat, doctrine, and resources will change, requir-ing constant reevaluation of the system. That is howit should be, and efforts to isolate the acquisitionmanagement process from such pressures in thename of coherent and sound management aresure to introduce crippling distortions into thepolitical structure of the process.41

The fact that the political process necessarilyintroduces uncertainty into defense acquisition doesnot, however, mean that nothing can be done tomitigate the effects of this uncertainly. Actionswithin both Congress and the DoD can improve thecoupling between the political and the acquisitionprocesses,

Congress

The” level of congressional oversight-manywould say micromanagement—has risen dramati-cally over the past 20 years. A few statistics reflectthis growth: In 1970, the defense authorization actwas 9 pages in length and was accompanied by a33-page conference report. Congress made 180adjustments to the authorization, and 650 to theappropriations bill, during that year’s budget review.By 1985, the authorization act grew to 169 pages andthe conference report to 354; congressional adjust-ments to defense authorization and appropriationlegislation totaled 1,315 and 1,848, respectively.42


Today, 29 committees and 55 subcommittees over-see defense activities in both houses, and over20,000 congressional staffers and employees ofcongressional agencies deal with some aspect ofdefense.

This growth has not occurred in a vacuum. TheDSB Acquisition Cycle task force concluded in1978 that:

A significant portion of the “blame” for thisincreasing congressional “micromangement” canprobably be laid to the fact that the DoD hasexhibited a chronic inability or unwillingness toadequately forecast program, cost, schedule, andperformance information and projections to theCongress.43

The Defense Systems Management College, theorganization that trains DoD program managers,includes the following interchange in a discussion ofcongressional oversight:

Senior DoD acquisition official, appearing beforean authorizing committee:

“Gentlemen, what we’d like to know is when areyou going to stop micromanaging our business?”Senior, veteran professional staff member of thatcommittee:

“Sir, when you start."44

Congressional action on the defense budget isoften an extension of debates conducted in otherarenas.

Much of the so-called Congressional micro-management is, in fact, stimulated by factions withinthe Pentagon hying to reverse, through selectivelyleaked information to Congress, unpalatable deci-sions made within the executive branch. In this sensemany Congressional actions on weapons programsare an extension of internal decision making by theDepartment of Defense.45

Underlying much of the congressional interest indetails of the defense budget is, of course, its directimpact on a great many congressional districts-those having major defense contractors, defensebases, or large numbers of defense workers. Al-though Members of Congress are responsible for

national policy, they are accountable to their indi-vidual constituents. It should not come as a greatsurprise that Members of Congress therefore lookout for their constituents’ interests in the course oftheir legislative responsibilities. To put it bluntly,they have every incentive to pork barrel. Damagingas this practice may be on the national level, it isdifficult to see how changes in congressional proce-dure can substantially reduce it, given the underlyingincentives built into the United States Constitution.

In specific cases where national consensus existson a matter of high priority, Congress has shown thatit can rise above parochial tendencies. A goodexample is the recently enacted legislation that willpermit the DoD to close down unneeded defensebases. Although every Member of Congress wants toeliminate waste from the defense budget, noneconsiders bases in his or her own district to bewasteful. Moreover, many Members of Congress areconvinced that DoD uses base closures to threatenlegislators considered insufficiently “pro-defense.’*Therefore, Congress has enacted legislative road-blocks over the past decade or so, effectively makingit impossible for DoD to close any bases. To breakthis impasse, Congress established a commission todraw up a list of bases to be closed on an “all ornothing” basis, forcing any legislator seeking toremove a particular base from the chopping block totorpedo the entire package. By this means. Congressand DoD cut through the storm of political contro-versy surrounding individual closures.

Congressional review of the defense budgetpresently deals more with artificial accountinginputs (dollars, personnel slots, buildings, etc.) thanwith defense outputs (mission capabilities or strate-gic goals). The inputs are easier to count and tocontrol, and unlike defense mission capabilities theypermit comparisons to other programs across theentire Federal Government.

A report to the Senate Armed Services Committeedescribed how the budget request needed to fulfill aparticular defense mission-developing the capabil-ity to deploy 10 combat divisions to Europe within

Appendix A—The DoD Acquisition System ● 19

10 days of mobilization-was divided by account-ing categories into separate pieces parcelled outamong the committee’s subcommittees. Althougheach different piece—storage site construction,spare part supply, updating airlift capability, trans-porting materials to Europe, etc.-was part of anintegrated whole, each was treated as an independentitem. Each subcommittee compared the items in itsjurisdiction to similar ones wholly unrelated to themobilization mission. “In a short time, the emphasison policy implementation of a major defense com-mitment was lost among thousands of minor deci-sions on accounting inputs.”46

Although many have recommended that Congressserve as a board of directors for DoD as a whole, itspresent budget process tends to push the examina-tion to a much lower level. Moreover, Congress hasdifficulty entering into a dialog with the DefenseDepartment on strategic objectives because there isno clearly identifiable counterpart within DoD withwhom such a discussion can be conducted:

There is no appropriate forum at the OSD [Officeof the Secretary of Defense] level in which strategy,policy, and operational concepts and capabilities arefully debated and translated into specific acquisitionprograms. The thrust of the current process is toconcentrate on procurement, management, and allo-cation of resources for individual systems rather thanon the overarching rationale and purposes that definethe need for and the operational capabilities of thosesystems. . . . Ideally, the Joint Chiefs of Staff (as acollective body) and the Secretary of Defense couldfurnish this longer perspective, but they are ham-pered by process, schedule, and organization fromdwelling on many of these broad operational consid-erations.47

It might be added that much of the “process” keepingthese officials from taking a broad view consists ofresponding to numerous congressional inquiries anddirectives.

Additional factors complicating congressionalreview of the DoD budget are major proceduralchanges that have been introduced over the last 15years. In 1974, the Congressional Budget and

Impoundment Control Act established a new budg-etary process within Congress. Prior to that act, thetwo-stage process of authorization and appropriationdealt with Federal agencies and programs individu-ally. There was no mechanism whereby revenuesand expenditures could be examined across theentire Federal Government. The 1974 Act prefaceda third “budgeting” stage in which Congress estab-lishes income and expenditure targets for the Federalbudget as a whole and specifies spending targets ineach of 15 mission areas. These targets are supposedto guide, but not formally bind, the authorizingcommittees. Later in the budgetary cycle, the budgetguidelines are reviewed and new targets are speci-fied that are binding upon the appropriations com-mittees.

Under this new process, review and appropriationof the DoD budget takes significantly longer. Finaldecisions for the defense budget are made bycongressional conference committees as (or, inmany recent cases, after) the new fiscal year starts,late in the executive branch’s budget-formulationprocess for the following fiscal year. Last-minutechanges in the appropriated funding levels requirelast-minute changes to next year’s request andinfluence planning estimates for the following fiscalyear. The Packard Commission found that the timingand scope of changes introduced late in the appro-priations process “prevent the DoD from makingcoherent linkages among the three defense budgetsthat it manages at any one time—the budget beingexecuted, the budget under review by Congress. andthe budget that DoD is developing for the upcomingfiscal year.”48

The Balanced Budget and Emergency DeficitControl Act of 1985 (Public Law 99-177, morecommonly known as the Gramm-Rudman-HollingsAct) placed further constraints on congressionalbudgetary process. Besides reinforcing the “zero-sum” nature of the Federal budget, the major impactof this legislation was its emphasis on Federal“outlays,” or money actually spent, during a fiscalyear. The appropriations process prior to Gramm-Rudman-Hollings dealt not so much with actual


spending but with “budget authority,” or permissionto enter into contracts that obligate the FederalGovernment to future expenditures.

Deficits are created not by budget authority but byactual expenditures. Since funds appropriated fordifferent purposes are spent at very different rates,the relation between budget authority and outlaysdepends on the purposes for which the funds are tobe used. Salaries, for example, are essentially spententirely within the year for which they are appropri-ated; cutting one dollar of budget authority forsalaries will reduce that year’s outlays by a dollar.Funds for building ships, on the other hand, can bespent more than a decade after their appropriation;cutting a dollar off a ship procurement appropriationtrims as little as two cents off that year’s outlays.

Due to this variation in outlays versus budgetauthority, Gramm-Rudman-Hollings outlay controlsmake it even harder to review the defense budget.Increasing appropriations in one area may require fargreater cuts in another to keep outlays from chang-ing.

As a final point concerning the role of congres-sional “micromanagement,” the prospects forchanging the relationship between Congress andDoD to one of greater strategic oversight weredamaged by the years of tension and confrontationthat existed between Congress and DoD in the early1980s.

Department of Defense

Actions within the DoD contribute as much toprogram variability as do those by Congress. Al-though congressional line item changes certainlycomplicate program management, changes gener-ated within the many layers of DoD managementadd significantly to the problem. The cuts that arepassed down are due to DoD’s inability to forecastprogram costs accurately, to defer new starts untilsufficient funding to cover the actual (rather than theoriginally estimated) costs is available, or to elimi-

nate programs-rather than stretch them out—in theevent of funding shortfalls.

In 1981, then-Deputy Secretary of Defense FrankCarlucci offered 32 management initiatives to im-prove the defense acquisition process, with number4 being to “increase program stability in theacquisition process.” Program stability was also oneof six areas cited for high-level management atten-tion in 1983. Nevertheless, a General AccountingOffice (GAO) review of the Defense AcquisitionImprovement Program in 1986 found that “despitelarge budget increases, DOD has reported essen-tially no progress in stabilizing major weaponprograms. “49 GAO found that although the impact ofunderfunding programs is “well-recognized anddocumented, a workable and effective method formatching DoD’s needs with budgetary constraintshas not been developed.” 50 The Office of theSecretary of Defense, according to GAO, “hasreported that the inability to cancel low priorityprograms continues to be a fundamental obstacle toimproving program stability.”51

Limiting the number of new programs and termi-nating low priority ones will be required in order toprevent the remaining programs from being under-funded. Although DoD has claimed progress inlimiting the number of “major new programs,’” GAOfound this reduction to be due in part to a doublingof the minimum cost threshold that defines a majorsystem. “Consequently, fewer new starts are consid-ered major under the revised higher thresholds.”52

These funding issues are discussed further in the“Affordability” section of chapter 8 of the mainreport.

Too often, desire for “funding stability” waneswhen the possibility of funding growth presentsitself. According to the Comptroller General of theUnited States. the boom and bust cycle that the DoDbudget experiences “encourages managers to pro-cure as much as possible when funding is relativelyplentiful and not attempt to develop a stable and


realistic procurement plan.”53 Although he doubtedthat “defense budgets will ever be as stable as DoDmanagers would like,” the Comptroller Generalnevertheless argued that steps could still be takenwithin DoD and Congress to “create as muchstability as possible in an environment which willalways be uncertain to some degree.”54

Personnel Turnover

Another contributor to program variability isturnover in acquisition personnel. Although typicaldefense programs have lifetimes measured in dec-ades, the average tenure of program managerssurveyed by GAO in 1986 was less than 2 1/2 years.Such short tenures make it difficult to increase theauthority of program managers because they hinderany attempt to assign accountability. Moreover, theycan generate pressures to sacrifice long-term qualityfor short- term results.

The contribution of high turnover rates to programvariability is only one of the many issues concerningthe acquisition workforce. Additional issues arediscussed in the section on “Acquisition Personnel,”below.

Reducing Program Variability

Analysts disagree as to which of two managementfailures is the more serious in the light of unexpectedchange: failure to plan and budget flexibly, or failureto hold to a fixed schedule in the light of perform-ance and budgetary uncertainties. Writing for theCSIS Defense Acquisition Study, Leonard Sullivannotes that acquisition plans during the ReaganAdministration “have gone through a boom and bustcycle that totally defies rational planning.” “Thesegyrations . . . make fixed planning for ‘stableacquisition’ an unachievable ideal.” His conclusionis that “inescapable fluctuations in White House andCongressional budget expectations and tactics de-mand the development of an acquisition system that

responds resiliently to the inevitable changes inlong-range projects in America’s short-range politi-cal environment.”55

On the other hand, participants in the DSBsummer study on Practical Performance FunctionalRequirements believed that every effort must bemade to hold to a fixed schedule in the event ofunexpected changes. “Upon commencement ofFSED [Full Scale Engineering Development],schedule should be considered as the dominantprogram driver and the program contracted andfunded accordingly . . . In the event that technologi-cal opportunities or operational requirements war-rant change, block upgrades [deferring a set ofchanges for a later production series rather thanredesigning and/or retrofitting changes into theentire production run] should be the primary solu-tion to avoid schedule delays.”56 If a block upgradeis not acceptable, “it’s probably better to terminatethe program and begin the process over again.”57

Several techniques have been proposed to lessenprogram variability and/or plan in the face ofuncertainty, among them baselining, multiyearbudgeting, and increasing management flexibility.

Baselining-The Defense Acquisition Improve-ment Act of 1986 (Public Law 99-661) mandatedbaselining for major acquisition programs-a re-quirement incorporated in DoD Regulation 5000.1as of September 1987. The Act also requires eachmilitary Service to designate selected high priority.major acquisition programs as “Defense Enterpriseprograms” having streamlined reporting proceduresfor program managers. For these programs. congressmay authorize funding for the full-scale develop-ment or production stages “in a single amountsufficient to carry out that stage, but not for a periodin excess of five years . . .“58 Such multiyear authori-zations would eliminate annual congressional re-views for these programs, at least by the Armed


Services Committees. Four programs-the Army’sMultiple Subscriber Equipment (MSE) and Ad-vanced Tactical Missile System (ATACMS) pro-grams and the Navy’s Trident II missile and T-45Trainer System programs-have been given Mile-stone Authorization status by Congress. However,Congress has neither received nor approved theactual authorizations for these programs.

Multiyear Budgeting-Multiyear budgeting goesbeyond multiyear authorizations for selected pro-grams to provide authorization and appropriation ofthe entire Department of Defense budget for periodslonger than one year. With a longer planning horizonand less frequent congressional review, the hope isthat programs can enjoy greater stability, withcongressional oversight directed more towards stra-tegic guidance and away from individual line items.Following the Packard Commission’s strong recom-mendation, the Department of Defense submitted a2-year 1988-89 budget to Congress early in 1987.

Biennial budgeting has not been readily acceptedby Congress. One reason is obvious—one of thepurposes behind biennial budgeting is to lessencongressional influence. The matter is not thatsimple, however, since Gramm-Rudman-Hollingsplays a role. Facing fixed deficit targets for futureyears, Congress is reluctant to commit itself to futureoutlays when it has no firm idea of what thecorresponding revenues will be. Estimating reve-nues for the ongoing fiscal year is difficult enough,given their dependence on economic conditions thatcannot be predicted in detail. Doing the same thingfor future fiscal years becomes highly problematic.

The CSIS Defense Acquisition study points out anumber of other practical disadvantages and compli-cations of the biennial defense budget.59 Unlessimplemented government-wide, it would give DoDa preferential status within the executive branch.Government policies and procedures-especiallywith regard to personnel—emphasize uniformityacross the board. If the 2-year cycle enabled the

Defense Department to resist budget cuts, it couldcome under strong political attack.60

Second, while a 2-year budget reduces opportuni-ties for congressional micromanagement. it alsorestricts DoD’s flexibility. A supplemental appro-priation adjusting the second-year amounts wouldrestore some of this flexibility, but it would thereforealso reintroduce opportunities for congressionalintervention.

The CSIS study goes on to describe timing difficulties that the 2-year terms of members of theHouse of Representatives and the 4-year term of thePresident present when coupled with a biennialbudget. If the 2-year budget were submitted andapproved during a presidential election year, theincoming President would not be able to execute hisor her own defense budget for 20 months after takingoffice. If the budget were submitted and approved inodd-numbered years, Members of Congress runningfor reelection would be doing so on the basis ofdefense budget votes made more than a year ago.The CSIS study believed that Members would notwant to put themselves in this position. However, thereport does not make clear why a Member ofCongress would find it much harder to justify adefense budget decision made the previous year onthe basis of conditions at the time, than it would beto defend any other decision made during the firstyear of a 2-year term.

The CSIS study does not discuss the root causeof these timing problems, which is that absoluteprogram stability is fundamentally incompatiblewith holding elected officials accountable atperiodic intervals for their actions. Every time anelected official is replaced, there is—and mustbe—the opportunity for the new official to changethe way things have been done.

There are two ways to handle the timing problem.One is to permit a new President to make significantchanges in a previously submitted 2-year budget,


thereby vitiating the advantages of long-term budg-eting. The other is for the new President to leaveessentially intact the budget he inherits upon assum-ing office, concentrating instead on the budgets thathe will submit This latter approach, of course,counts on the new President’s successor to respectthose future budgets when they extend into asubsequent Administration.

For whatever reason, the first submission of abiennial budget in 1987 for fiscal years 1988 and1989 did not go far. Neither the House nor the SenateArmed Services Committees authorized very muchof the DoD 1989 request. More significantly, the1988 authorization act specified that “authorizationsof appropriations and of personnel strength levels inthis Act for fiscal year 1989 are effective only withrespect to appropriations made during the firstsession of the One Hundredth Congress’ ’-i.e.,appropriations made that year. The appropriationscommittees of the House and Senate, even lessenthusiastic about multiyear budgeting than theArmed Services Committees, did not appropriateany 1989 funds.

Although no funds were appropriated for 1989during the 1988 budget cycle, one effect that the2-year submission did have was to make DoD’sout-year plans more visible to Congress than theyhad been before. Although some might fear that thisvisibility just gives Congress that much moreopportunity to meddle, it is also plausible thatimproved communication between Congress andDoD might give Congress the confidence in DoDplanning it needs to relax its level of “micro-management. ” Good communication extends theplanning horizon, enabling both parties to take alonger view.

Management Flexibility—A further way to reducethe variability of DoD programs is to increase theability of the DoD to adjust to changing circum-stances without having to come back to Congress.Such techniques-which would make DoD manage-ment easier just as they would lessen congressionalinfluence-include increasing DoD’s ability to

transfer money from one program to another (i.e.,reprogramming); using funds in one appropriationtype (e.g., procurement) for another purpose (e.g.,research and development); and including unspeci-fied management reserves in program budgets.

Congress now grants DoD the ability to makesome such changes without prior notification of orapproval by Congress. However, in other cases.either notification or prior approval of the Appro-priations and/or the Armed Services Committees is


required. 61 Furthermore, each fiscal year’s budgethas associated with it a maximum amount of transferauthority. If the sum total of all reprogrammingssubject to the transfer authority limit reaches themaximum, no further reprogrammings can be made.

These requirements, along with the emphasis inDoD budgeting on specific program elements,restrict DoD’s ability to respond to changing cir-cumstances. The DSB Acquisition Cycle Task Forcepointed out several DoD needs that do not neatly fallwithin existing line items and that therefore requireadditional funding flexibility to address:

●

●

●

The

Getting started with technology and systemexperiments in areas that DoD has alreadydecided to submit to Congress in the followingyear’s budget. (This problem would be aggra-vated by a biennial budget cycle.)Purchasing good ideas from the losing biddersin competitions.Providing extra support to programs perform-ing better than expected.

Task Force recognized the belief within Con-gress that too many reprogrammings were alreadybeing used to evade congressional intent. “Negotiat-ing new and higher thresholds will thus require arestoration of DoD’s credibility with the Con-gress.” 62

The decomposition of the defense budget intodifferent accounting categories restricts DoD flexi-bility. It also can impede efficient program manage-ment objectives such as maintaining a smoothtransition from development to production. A DSBTask Force concluded in 1983 that “the Design toProduction transition is a process and not a fixedevent,” and that DoD funding rules prohibiting theuse of R&D funds for production make it “very

difficult to apply resources [during full-scale devel-opment] to producibility, manufacturing planning,tooling and test equipment and other actions leadingto production.’"63

A Final barrier to sound program management,and the biggest obstacle towards giving programmanagers greater authority over their own programs,is the lack of management reserves. Although thefunds required to fix unexpected problems obvi-ously cannot be estimated for any particular pro-gram, they can be determined statistically in theaggregate. Supervisors who oversaw several pro-gram managers, if provided with these reserves andthe authority to allocate them, would be able toaddress problems as they arose. According to theAcquisition Cycle Task Force,

, . . the important thing to keep in mind is that this isnot “contingency” money that is simply budgeted“in case something happens.” It is a necessarymanagement resource that should be provided be-cause it is well known, and experience amplydemonstrates, that something will happen and it mustbe fixed quickly if the program is to remain onschedule and within “planned for” costs.64

However, the intense competition for funds wi-thin DoD, as well as the degree of scrutiny appliedto defense budgets by Congress, both mitigate against providing such reserves. In an environmentwhere there are already far more claims on defensedollars than available funds, there is every incentiveto underestimate the costs of programs when Servicebudgets are prepared. Even if contingency reservesare initially provided for, they are one of the firstitems to be trimmed.

Were management reserves somehow to surviveDoD’s internal budget preparation process, they


would probably not fare well on Capitol Hill. from which the company’s chief executive officerAccording to Leonard Sullivan,

Reserve funding wedges, if identified in thebudget so that congressional staffers can find them,end up spotlighted, renamed “slush funds,” to protectthe taxpayer from waste, fraud+ and abuse.65

Requirements Generation and the ResourceAllocation Process

Description of the Problems

In 1985, a DSB summer study examined theprocess by which requirements for new militarysystems are generated. The task force concluded that“although promising efforts are underway in all ofthe Services to improve their requirements proc-esses, deficiencies in this process are still likely to besignificant contributors to continuing increases inboth the cost and the length of time required to fieldnew defense systems."66 The report identified threeproblems in particular:

. Users are not involved directly and continu-ously in determining and ranking their militaryneeds.

. Requirements are expected to be observed toorigidly.

. Acquiring organizations do not go over theirrequirements often enough with their suppliers,before making them formal.

The study proposed emulating the organizationalstructure of successful commercial programs tostreamline DoD acquisition, a proposal whichformed the basis of the Packard Commission’srecommendations regarding acquisition organiza-tions.

According to the DSB report, deciding what toacquire in the commercial world—at least for thehighest priority, “bet-your-company” programs ex-amined by the study panel—is essentially a one-stepoperation. Balancing requests from users againsttechnological opportunities and available resources,the program manager advances realistic proposals

(CEO) can select. -

The PM [Program Manager] is motivated to berealistic about performance, cost and schedule, bothbecause he will have to carry out the program if it isapproved and because his job is dependent on themerits of the proposal and not simply on whether itis accepted.67

The DoD, on the other hand, decides what to buyin two stages. First, a highly political competition forfunds involves the military Service, the OSD, OMB,and Congress. After funds are reserved. as denotedin the milestone process by a DRB decision toinitiate a new program, a second stage of competi-tion selects the actual supplier.

There are great pressures to overpromise in orderto survive the [funding] competition. Since thedecisions are made by political processes among alarge and diverse group of people, there is littlepressure to discipline the process and to enforcerealism. Clear-cut designs to meet the requirementsare not allowed because they would interfere withthe next step—competitive source selection. Theresult is a firm over-stated requirement which toofrequently can neither be met nor changed.68

Leonard Sullivan describes a little more bluntlysome additional factors within DoD that lead tooverstated requirements:

The twin siren songs of “nothing is too good forour boys” (sung by the Services) and “nothing isimpossible” (crooned by the technological commu-nity) have produced a deeply embedded Americandefense culture and guarantee the perpetuation of amilitary force that is at or beyond the leading edge oftechnology in the factory, and at or behind thetrailing edge of any realistic sustainable warfightingcapability .69

“Another myth popular among amateur ‘require-ments’ generators,” Sullivan adds, “is that since thedesired system is going to be expensive anyway, themarginal costs of adding a few more capabilities will

26 ● Holding the Edge Maintaining the Defense Technology Base, Volume 2

be small . . . [However, these] add-ens become ‘thestraw that breaks the camel’s back’ in terns ofdesign complexity, {development scheduling, andproduction costs.”70

Sullivan, the DSB panel, and the Packard Com-mission all attribute much of the pressure foroverstated requirements to insufficient interactionbetween those who know what is needed and thosewho know how to provide it. According to thePackard report, “Generally, users do not havesufficient technical knowledge and program experi-ence, and acquisition teams do not have sufficientexperience with or insight into operational prob-lems.’71

The DSB panel recommended that the Command-ers in Chief [CINCs] of the operating forces be givena more significant role in requirements generation.72

The CINCs “do not participate with the services inmaking requirements tradeoffs even though theymay be the most qualified to judge the trueoperational value of a particular requirement.”73 ThePackard Commission agreed that much greateremphasis should be placed on “an informed trade-offbetween user requirements, on the one hand, andschedule and cost on the other.”74 The DSARCprocess. according to the Packard Commission, wasunable to strike this balance. Although DSARC wasable to determine whether proposed new systemswould meet the requirements set for them, it “lacksa viable mechanism for challenging those require-ments.”75

The 1985 DSB panel on requirements that recom-mended the CINCs play a greater role in generatingrequirements also called for them to be moreinvolved in-or at least, more aware of—the subse -

quent development process. Admiral W.J. Crowe—at the time of the DSB study the Commander inChief of the U.S. Pacific Command-believed thathis input into the requirements process was suffi-cient until the system entered development. “Fromthat point, however, I have little influence over theprocess because feedback on affordability, priori-ties, and any tradeoffs made by the developingService is almost non-existent. I do not want thecapability to design or build systems, but I do needsufficient involvement in the development processto be able to point out major design changes ofomission or commission which would affect mycapabilities and/or strategy.”76

According to the DSB, even the program manag-ers who are immediately responsible for developingmajor systems do not have sufficient ability or desireto reexamine requirements once development hasstarted. Should meeting a particular requirementprove more difficult than expected, leading to costgrowth or schedule slippage, program managers alltoo often fail to reconsider the need for it. Moreover,since program offices are established after therequirements have been ratified, managers generallyarrive too late to affect requirements at ail.

Improving the Requirements Process

Changes in the requirements process can come intwo different areas. One is in the process by whichthe Services first establish requirements for newsystems or upgrades. The existing two-stage processis a recipe for producing the wrong system too lateand at too high a cost. Moreover, Services may notfully evaluate non-traditional means of meeting theirrequirements, especially if they involve changingthe respective roles and missions of the Services.

. . .


Improvements to the initial requirements genera-tion process involve strengthening the role of theUSD(A) in these early stages, ensuring an objectiveevaluation, and preserving a role for Defense Ad-vanced Research Projects Agency (DARPA) toexplore nontraditional solutions outside of Serviceprocesses.

The report of the Project on Monitoring DefenseReorganization, charged with reviewing the im-plementation of the Packard Commission’s recom-mendations, concluded that “although the PackardCommission’s objectives pertaining to ‘require-ments’ are far from fulfilled, there has been materialprogress.”77 Most important, according to the study,was the establishment of the Joint RequirementsOversight Council (JROC) under the newly createdpost of Vice Chairman of the Joint Chiefs of Staff.This council is charged with reviewing all programsthat are candidates for joint development becausethey can be used by, or affect the operation of, morethan military Service. JROC has also served toincrease the role of the CINCs in decisions onweapon characteristics, according to the study.

With respect to the Packard Commission’s recom-mendation that requirements be better balancedagainst cost and schedule and that affordability betaken more seriously, the implementation studyfound that “the organizations and procedures thatcould make possible such a change have been setup,” but “their effective operation will requirecontinued high-level attention.”78

The requirements process can also be improved atthe point where program managers review them withthe ultimate users: the CINCs and others serving inoperational capacities. Here, managers can bringconsiderations of cost. schedules, and technicaldevelopments into play to change those require-ments.

Bureaucratic Paralysis“When I took over procurement responsibility for

General Motors, the guidelines for running theacquisition activities was 154 pages. I gave them atarget of 10. We ended up with 13 pages to run allGeneral Motors acquisition efforts.

“I was interviewing a General from the Air Forcefor a job and he said, ‘You cannot run an organiza-tion with only 13 pages. ’ I said, ‘We are. He said,‘1 have 3,650 pages,’ and I said, ‘General, you cannotrun an organization with 3,650 pages. “

—Robert Costello, Under Secretaryof Defense for Acquisition79

Documentation

Perhaps the most discussed problem with defenseacquisition is the bureaucratic burden that individu-als and companies involved in defense acquisitionmust carry to do their jobs. On the way to a DABmilestone review, a program manager may have tomake as many as 100 briefings. Attention must bepaid to thousands of regulations, specifications, andstandards. As the Packard Commission described,

The program manager finds that, far from beingthe manager of the program, he is merely one of theparticipants who can influence it. An army ofadvocates for special interests descends on theprogram to ensure that it complies with variousstandards for military specifications. reliability,maintainability, operability, small and minoritybusiness utilization, and competition, to name a few.Each of these advocates can demand that theprogram manager take or refrain from taking someaction, but none of them has any responsibility forthe ultimate cost, schedule, or performance of theprogram. 80

Increasing complications in the job of the programmanager have been accompanied by lengthening thetime needed to complete contracting actions andincreased regulation, oversight, and auditing ofdefense contractors.


A recent RAND Corporation study tried toquantify both the increased regulatory activity inrecent years and the effects of those regulations oncost, schedule, and performance.81 They found nearunanimity among those who work in acquisition thatcomplying with regulations, management review,audits, etc. is much more difficult now than in thepast. However, the indicators RAND chose tomeasure that difficulty-growth in staff sizes, re-quests for DoD testimony, numbers of DoD regula-tions, numbers of GAO reports, etc.-did not clearlyconfirm the increase.

Of the indicators sought to identify the effects ofthe regulatory burden-cost, schedule, and perform-a n c e -RAND found that cost shows the clearesteffects:

We conclude, on the basis of the sparse dataavailable, that the sum of all incremental costs whichcan reasonably be charged to regulatory controlsprobably amounts to between five and ten percent oftotal program costs.82

These numbers are lower than some that havebeen cited by defense contractors, possibly becausethey address only the incremental effects of recentregulation and not the cumulative effects. Onecontractor in a dual-use (military and commercial)business told OTA that the constraints imposed bydoing business with the DoD are responsible for 20to 50 percent of the total price of the defense product.Other estimates go even higher. The president ofGrumman Corporation has stated that “only about athird of the time and money spent in developing newweapons systems has anything to do with design,development, and testing. The rest of it is the cost ofreview and oversight.83

This estimate is almost certainly high, sinceGrumman Corporation would surely conduct somereview and oversight activities for its own use evenif DoD did not mandate them. In fact, according to

a senior executive at another aerospace corporation,DoD imposes no administrative burden above whatthe company would want to do anyway. Accordingto Albert D. Wheelon, for 16 years the head ofsatellite production at Hughes Corporation,

Our experience is that similar spacecraft costabout the same, whether they are bought undermilitary or commercial arrangements . . . Complyingwith DoD systems for cost and schedule control,contract management and quality control was notparticularly burdensome. In fact, we used theirprocedures in our commercial programs by choice.In essence, DoD asked us to do no more for itsprograms than we would want to do for ourcommercial customers and ourselves.84

Even if cost penalties can be unambiguouslyattributed to regulation, it is hard to consider them asmeasures of government waste. As the RAND reportmakes clear,

. . . to sustain an interpretation that all, or even most,of these costs are “wasted” money would requiredemonstrating that no benefits derive from thereporting and oversight activities that account for thebulk of the cost.85

For reasons discussed earlier in this chapter, defenseacquisition is clearly not managed solely to mini-mize cost and maximize efficiency. Congress, theServices, the OSD, and the regulatory agenciesapparently have found the value of their respectiveinvolvement in defense acquisition to be worth theadditional cost.

Analysis

Whether or not red tape can be quantitativelyshown to affect defense procurement, and regardlessof the degree to which it has increased over the years,it is unambiguously greater in government than inthe private sector. The RAND study noted that:

Military program managers are frequently sepa-rated from the senior OSD-level acquisition execu-

●


tive by five or six administrative layers. Each layerdemands a right to review all progress reports andmajor program change proposals. Not so apparentfrom the literature is that some of those layers havean extensive horizontal structure, so that the views ofseveral different offices must be accommodated inorder to pass through a particular layer or “gate.’86

Not only do program managers devote much of ‘theirtime towards preparing for these reviews, but theregulations-and their increasingly strict interpreta-tion, a point not amenable to RAND’s analysis-have the effect of limiting the initiative and discre-tion that program managers are allowed to exercise.

Note, however, that if the present hierarchyrequires five or six management layers between aprogram office and the senior Defense AcquisitionExecutive, any compression of the command chainwill be accompanied by increasing the burden onthose at the top-unless the total number of acquisi-tion programs is cut proportionally. Bringing anyone program to the attention of the most seniormanagement will ensure that it moves rapidly ahead.Bringing every program to that level. without someway of ranking them to determine which ones trulydeserve the attention, will create grid lock.

The 1985 DSB Summer Study on PracticalFunctional Performance Requirements devoted aconsiderable amount of analysis to the differencesbetween the organizational environment of a DoDprogram manager and that of an equivalent positionin non-defense-related private industry. Successfulcommercial programs examined during the summerstudy shared a number of features:87

●

●

A Program Manager (PM) who has continuity,authority, flexibility, accountability for deci-sions, and direct access to the key decisionmaker (CEO).A powerful executive (sometimes the CEO)who has authority to make unchallengeabledecisions, settle disputes, and allocate addi-tional resources. The CEO can directly support

the Program Manager and insulate him or herfrom external pressures as critical needs arise.

● Active user involvement. The commercial user,not committed to a single supplier, is free topurchase from other producers. Therefore, theProgram Manager has a strong incentive toinvolve the user throughout product develop-ment, and emphasizes adherence to schedule(e.g., by modifying requirements with userconcurrence) in the event of difficulties.

There are many “minor players” in this commer-cial model, including inside staffs, governmentregulators, consumer groups, etc., but “one of themajor advantages of the Commercial Model is thatthe minor players play a minor role.”88

In its planning stage, according to the DSBsummer study, the commercial model is essentiallya one-step procedure. The Program Manager, bal-ancing user needs, foreseeable resources, and avail-able technology, prepares a realistic proposal for theCEO to consider. The CEO, weighing this proposalagainst other alternatives such as having the pro-posal revised or rejecting it in favor of other uses forcorporate resources, makes the decision to go ahead.“His future depends on whether programs he ap-proves are ultimately successful, not on whether ornot he goes ahead with them.”89

The plan’s execution is marked by a close, directworking relationship between the PM and the CEO:

The CEO must be kept informed and the PM mustbe able to get help rapidly and reliably if he needs it.The principle is one of a joint activity towards acommon goal. A program failure is a failure of bothCEO and PM.

The staffs and inspectors, test groups and “ilities”[reliability, maintainability, supportability, etc.],groups exist, but are insulated from the PM by theCEO. The staffs can talk to the PM and comment andadvise but cannot direct the PM without goingthrough the CEO. Only the PM and the CEO canmake decisions; they have the responsibility andtherefore the authority.90


The summer study sought to emulate thesepractices within DoD acquisition programs. Specifi-cally, they recommended that DoD establish whatthey called “surrogate CEOs’’—individuals whohave been delegated authority and responsibility toserve as the ultimate decisionmakers for one or a fewprograms. To implement this recommendation, theMilitary Departments would have to reduce thenumber of people involved in the decision processes,reduce the number of layers through which theprogram manager reports, and reaffirm programmanager responsibility for all phases of programexecution. They would also have to provide programmanagers with access to those senior managers (thesurrogate CEOs) who would have the authority andresources sufficient to “make and enforce decisionsregarding tradeoffs between performance, schedule,and cost. ”

The Packard Commission cited this DSB study asthe basis for its recommendations to streamline theacquisition process. In particular, the Commissioncalled for “unambiguous authority for overall acqui-sition policy, clear accountability for acquisitionexecution, and plain lines of command for those withprogram management responsibilities."91 At the topof the acquisition structure recommended by thePackard Commission would be a new position, theUnder Secretary of Defense for Acquisition(USD(A)) who would serve as the Defense Acquisi-tion Executive. Reporting to the USD(A) would becomparable to Service Acquisition Executives(SAEs) in the Army, Navy, and Air Force andequivalent positions in the defense agencies. EachSAE would appoint and oversee a number ofProgram Executive Officers (PEOs), who in turnwould oversee Program Managers. The PEOs, “likegroup general managers in industry, should beresponsible for a reasonable and defined number ofacquisition programs. Program managers for theseprograms should be responsible directly to theirrespective PEO and, on program matters, report onlyto him."92 It would be the responsibility of the UnderSecretary for Acquisition to ensure that “no addi-

tional layers are inserted into this program chain ofcommand.”

Through the Defense Reorganization. Act of 1986and concomitant Executive Orders and DoD Direc-tives and Instructions, the organizational structurerecommended by the Packard Commission wasestablished. However, the new structure supple-mented-and did not replace—any existing chainsof authority and command. According to a study ofthe implementation of the Packard Commissionrecommendations and associated legislation,

, . , the purposes of the legislation have not been met.Our sense is that the new positions were simplysuperimposed on top of the existing structure.93

The new acquisition chain is at present a communi-cations link, and does not control funds. Figure A-4shows the new acquisition lines of authority alongwith the existing organizations for command andbudget.

Regardless of how effectively the implementationof the Packard Commission recommendations wi-thin DoD captured the intent of those recommenda-tions, it is clear that the actions taken to date do notaddress the original concerns of the DSB summerstudy.

Nor is it clear that they could. At the same timethat it recommended changing DoD practices to putthem more in line with commercial ones, the DSBsummer study also acknowledged that:

There are inherent and basic differences betweenthe DoD and non-DoD processes which certainlyinhibit and may even prevent the direct mapping oflessons learned [from the commercial examples] intothe DoD requirements process. For example, there isno counterpart to the role of Congress in industry,nor are there any unifying quantitative measures ofsuccess in DoD corresponding to profit or [return oninvestment]. Furthermore, some personnel con-straints in DoD have no counterpart in industry.Finally, DoD does not operate in a free market asbuyer or seller, and can only imperfectly approxi-mate free market competitive conditions.94


Neglecting the inherent and essential involvement ofthe political process, in particular, will lead toinappropriate solutions. Edwin Deagle’s analysis,cited in the previous section of this appendix on“Program Variability,” is particularly relevant todiscussion of the acquisition (or, more accurately,the military R&D) process:

. . . organizational and procedural designs are unusu-ally important. . . since they determine the structurewithin which massive managerial and politicalcontrol problems intersect.” Moreover, there can beconflict between organizational strategy designed toproduce efficient political decision processes andmanagerial strategy designed to achieve coherentcontrol of weapon system development. Yet theorganization for control of military R&D inevitablyis a mixture of both purposes. It is argued here thatthe failure to cope explicitly and well with thisparadox is the central public policy problem ofmilitary R&D.95

The DSB summer study acknowledges the impor-tance, but not the inevitability, of political influ-ences within DoD in a passage on DoD decision- ‘making that could equally well describe Congress:

No one person has the authority to make firmdecisions. Decisions are made by a large, diffusegroup that acts something like an extended commit-tee and that lacks clear-cut responsibility and ac-countability. The DoD itself exists in a politicalenvironment that further smears out the decisionmaking process. As a result, decisionmaking islengthy and uncertain. The players change and thedecisions tend to change with them. The programManager is separated from the top level of the DoDby many intermediate layers, all of whom must bedealt with, none of whom can say yes, but most ofwhom can say no. Decisions are late, inconsistentand untrustworthy.96

And in an earlier passage,

Although the DoD is nominally a hierarchicalauthoritative organization, it is very difficult in a

democracy for anyone to make a controversialdecision stick.97

The key to the direct decisionmaking processesand lines of authority in the DSB summer study’scommercial model is the close and direct linkbetween the program manager and the CEO. How-ever, the commercial programs analyzed by thesummer study—the ESS-4 automated electronicswitching system for long-distance communicationdeveloped by Bell Labs, the Boeing 767 airliner, aSatellite Business Systems communication satellitesystem, the IBM System 360 computer series, andthe Federal Aerobatics Administration national airtraffic control system-were not run-of-the-millactivities. They . . .

were of great importance to the companies involvedand therefore to the CEO. There is hardly any singleprogram in DoD of equivalent importance to ServiceSecretaries, let alone to the Secretary of Defense.DoD has too many important programs for suchofficials to keep track of them in detail.98

***Increasing the authority of the PM alone will not

solve the problem. Attempts to streamline theprocess and to connect the PM more directly to thetop of the DoD have not been successful except inextraordinary cases. There are too many programsfor the top level to understand in detail. They mustrely on their staffs and authority rediffuses in thebureaucracy. 99

This was to be the role of the “surrogate CEOs”which the DSB summer study called for establishingwithin DoD. The success of the Surrogate CEO . . .

will depend on how much authority he really has toadjust performance and schedule, provide additionalresources if needed, make or approve tradeoffs.100

It was this recommendation that led the PackardCommission to call for the establishment of ProgramExecutive Officers. However, since the acquisitionchain of authority established by the military Serv-ices in response to the commission’s recommenda-


tion has no real control over resources, it isquestionable how well it fulfills the Commission’sintent. In the Navy and the Air Force, the Com-mander of the System Command product division towhom a program manager reports has been desig-nated as his Program Executive Officer—despite theconclusion of the DSB summer study that:

A supervisor or commander in the current DoDstructure is not equivalent to a Surrogate CEObecause he does not have the necessary delegatedauthority . . . He does not have any more authorityover performance, cost, and schedule of his pro-grams than his PMs do. He cannot transfer fundsamong programs and he has almost no discretionarymoney under his control. His control of staff andmonitoring groups is minimal. He is overcommittedand has almost no flexibility.l0l

In the Army, PEO offices have been establishedseparate from the commanders of the System Com-mands within Army Materiel Command, but eventhese offices have no real control over resources.

Truly implementing the recommendations of theDSB summer study and the Packard Commissionwould require drastic changes in the operation ofDoD. Given the inherent involvement of the politicalprocess within defense acquisition, true implemen-tation may not be possible at all. The essence of theSurrogate CEO/Program Executive Officer conceptlies not in rearranging who reports to whom, but inconcentrating real authority in an individual posi-tioned to make decisions about a program and seethat they are implemented. However,

The law of conservation of authority says that thisdelegated authority must come from somewhere andit must come, in fact, from the Surrogate CEO’ssuperiors and from the staffs and regulatory bodiesin the government. These people, in the manner of allhuman beings, will resist giving up authority evenwhen they understand that their previous activitieshave been harmful rather than helpful. If the mostsenior people will really delegate their authority andinsist that it be further delegated to Surrogate CEOs,there is a chance the idea will succeed. There willstill be plenty of other things for the senior people todo.102

“Successful” DoD Models

Certain programs within the Department of De-fense—in particular, highly classified “special ac-cess” or “black” programs,103 and high-prioritystrategic programs such as the Minuteman missile,the Air-Launched Cruise Missile, and the Navy’sStrategic Systems Program Office that developedthe Trident system-have been held out as modelsthat have successfully conquered DoD bureaucracy.Special access programs, due to extreme securityrequirements, bypass much of the review andapproval process that ordinary, “white” programsmust contend with. Exempt from normal procure-ment and oversight operations, they are significantlyless encumbered with bureaucracy.104

According to Bernard McMahon, former Execu-tive Director for the Director of Central Intelligence(responsible for reviewing all intelligence programsand operations) and subsequently staff director for


the Senate Intelligence Committee, special accessprograms do have a number of advantages:105

●

●

●

speed of deployment-equipment is generallydeveloped and deployed faster than in normalprograms;exceptional stability, both in personnel and inconcept andbetter program managers and personnel thannormal programs of the same cost have.

Many senior officials with experience in bothspecial access and ordinary program managementreport that the streamlined management approachesand freedom from bureaucracy that characterizespecial access programs make possible the speedwith which these programs can field hardware.Others, however, argue that the advantages pos-sessed by black programs are not necessarily due tobureaucratic shortcuts. McMahon argues that sincemanagement and oversight of these programs aretightly restricted, those who perform these functionstend to be the most senior management of themilitary Services:

Because special access programs are reviewedonly by top management-their review boards arecomposed of flag officers and senior DoD civilianexecutives-they tend to get “special status” whenfunding priorities are established. Top managerstend to view the programs as their own, sponsorthem, defend them, protect them in the competitionfor dollars with regular programs, and favor them insetting priorities. Seldom are they terminated, re-duced, or stretched out nor is the economic rate ofproduction considered.106

The exceptional stability enjoyed by these programsis therefore due, at least in part, to their high priorityand the high level at which they are reviewed.“Management obstacles are cleared for specialprograms in ways normal program managers neverexperience.” 107 Similarly, their advantages in per-

sonnel are partly due to their priority. Admittedly,managers also have the advantage of being able tospend more of their time managing and less handlingbureaucratic overhead and advocacy.

The advantages enjoyed by special access pro-grams also come at a price. Procedures used inspecial access programs “significantly increase therisk of failure, both of program hardware and ofaccomplishing what we paid the money to do.”108

Part of the increased risk reflects the fact that specialaccess programs tend to be technically riskier.However, risk is further increased by eliminatingreviews and by short-circuiting the political processin which normal DoD programs operate:

The short cuts taken in the special access pro-grams , . . are dangerous. In the special access worldone hears horror stories of equipment that was tooexpensive, did not meet design expectations, was notsupported, was unreliable, and duplicated othercapabilities. 109

Those who attribute some of the successes ofspecial access programs to their management ap-proaches argue that these approaches should beextended to other DoD procurements. McMahon,however, argues that the model offered by blackprograms should not be extrapolated to the rest ofdefense procurement.

We simply cannot conduct a defense wideprocurement system using special access programprocedures. Top management does not have time toreview all programs with the degree of oversight itmust give to special access programs. Programs thathave succeeded have done so because they weresmall and few in number. . . .Efficiency alone is notsufficient. In rare, important cases we may choose totake risks and skip important steps; it should notbecome general defense practice. 110

The strategic systems also held out as examples ofsuccessful management share some of the same

1.


characteristics of successful special access pro-grams: viz., high priority and high visibility to seniormanagement. According to a critique of the PackardCommission report by an ad hoc committee of theAmerican Defense Preparedness Association, thesestrategic programs use “high quality but rather largestaff"—as opposed to the Packard Commission’srecommendation for small, streamlined staffs—andthe programs have “established sufficient priority toavoid the normal budget drills and priority-settingdisruptions.” The committee’s critique “questionsthe feasibility of achieving these objectives on allprograms.” 111 In other words, given a long line ofclaimants, those at the head of the line move faster.This does not mean everyone should be at the headof the line.

Overregulation and Public Opinion

Those who decry the inefficiencies imposed byregulation, audit, and oversight must realize thatthese penalties may be intentional; taxpayers placestringent requirements on expenditure of publicfunds. Figure A-5 illustrates the cost of doingbusiness as a function of regulatory scrutiny. Withminimal regulation or oversight, the government isdependent on the goodwill of contractors and publicofficials. Honest officials and corporations couldoperate very efficiently in this region, but dishonestones would take advantage of the lack of oversightto defraud the government.

At the other end of the spectrum, tight regulatorycontrols deter or detect those defrauding the govenr-ment, but they also drive up the cost of doingbusiness for everyone else. As was noted earlier,analyses by the RAND Corporation and othersimply that the existing regulatory regime imposesadditional costs of between 10 to 50 percent on thecost of doing business with the Department ofDefense. How much fraud this regulation deters isimpossible to estimate, but it must certainly be lessthan the $15 billion to $75 billion represented by 10to 50 percent of the procurement budget.

Most likely, the current regulatory regime isconsiderably more stringent than that which, accord-

Cost

Figure A-5-Cost v. Regulatory Intensity

I

I

ing to strict economic considerations, would resultin minimal cost. It may be the case that the publicwould not demand such stringent controls if it fullyunderstood the costs. If so, making the costs ofoverregulation clearer could lead to a relaxation ofunnecessary constraints. It may be possible, how-ever, that the American taxpayer prefers to pay the“tax” that overregulation imposes rather than permitthose in positions of public trust to misappropriatelesser amounts. If public demands for overregulationconstitute avoidable waste, then perhaps waste mustbe considered the price of curbing fraud and abuse.

Reducing Paperwork and Bureaucracy

Arbitrary measures to cut red tape or streamlinethe bureaucracy will fail unless they take intoaccount the reasons for establishing a bureaucracy inthe first place. For one thing, regulations area meansof preserving institutional memory in an environ-ment where presidential appointees have a medianlength of service of just over 2 years 112 and wheremilitary personnel are regularly rotated. They incor-porate the political oversight and review procedures

111“Quick Reaction Assessment of the President Blue Ribbon Commission on Defense Management,” an ad hoc Study conducted under the auspicesof the Undersea Warfare Systems Division of the American Defense Preparedness Association, October 1986.

112’’Leadership in Jeopardy,” National Academy of Public Administration, November 1985, p. 4. This figure applies to the entire Federal Government.


that come with the expenditure of public funds. Theycodify management procedures for large and un-wieldy organizations. Finally, they further impor-tant policy objectives that may be in the Nation’sor DoD% collective best interest even though theymight interfere with the most efficient executionof individual programs. As has been stated before,the government has many goals-environmentalprotection, occupational health and safety, fair laborpractices, equal opportunity, etc.-that may conflictwith any individual program manager’s ability torun a program. Just because a program manager doesnot believe his or her program is the appropriatevehicle to implement national policy does not meanthat that policy should be ignored. Although regula-tions have been criticized as attempts to solveyesterday’s problem by impeding today’s progress,those problems are certain to be repeated in theabsence of some way of institutionalizing thelessons learned.

A number of different approaches can be taken toreduce bureaucracy and regulation within DoD.Implementing any of them, however, presumes anatmosphere of trust among the DoD, the rest of theexecutive branch, and Congress. Our political sys-tem guarantees that the executive and legislativebranches will compete for power and influence.However, this competition can be carried out inmore or less confrontational terms. The relationshipbetween DoD and Congress in the early 1980s wasone of confrontation, substantially aggravating thelevel of mistrust.

In such an atmosphere, Congress chooses tolegislate rather than persuade because it has noassurance that persuasion will have any effect. TheDoD prefers to err on the side of strictness, for fearof incurring a congressional investigation and stillstricter legislation.

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 severe

punishment for those who get caught. Enacting sucha system would involve a major overhaul of theexisting defense acquisition system and the environ-ment in which it is conducted. Moreover, it wouldrequire (and also follow from) reducing what manyin government and industry see to be the existingadversarial relationship between the two.

Bottom-Up Review-Since regulations (or atleast guidelines) are inevitable in so bureaucratic aninstitution as DoD, one approach to alleviating theregulatory burden might be a bottom-up review of allregulations to ensure that only absolutely necessaryones are retained. However, the definition of “abso-lutely necessary” is highly subjective, and differentgroups or factions within the Department of De-fense, the executive branch, and Congress areunlikely to agree. Every DoD regulation was origi-nally instituted for what seemed to someone to be aworthy purpose. This point is acknowledged by thePackard Commission in describing the “army ofadvocates” for various special interests that belea-guer program managers:

None of the purposes they advocate is undesirablein itself. In the aggregate, however, they leave theprogram manager no room to balance their manydemands, some of which are in conflict with eachother and most of which are in conflict with theprogram’s cost and schedule objectives. Even moreimportantly, they produce a diffusion of manage-ment responsibility, in which everyone is responsi-ble, and no one is responsible.113

Before any of these advocates or excess regulationsare eliminated, those who instituted them will haveto be satisfied that the goals they advocate will bepreserved. Moreover, those with the time to reviewthe regulations would most likely not be the onesadversely affected by them, and it is unlikely thatthis approach would effect significant change.

Evolutionary Review—The DoD is testing anevolutionary process to relax unnecessary bureau-cratic requirements. Pursuant to the Defense Acqui-sition Improvement Act of 1986, DoD has desig-nated selected high-priority, major acquisition pro-grams to be “Defense Enterprise Programs” havingstreamlined reporting procedures (table A-1 ).


Each Enterprise program is being reviewed to seewhat regulatory relief would be useful. As soon asthese reviews are completed, it is expected that theServices will request waivers of certain regulationsfrom the USD(A). Complicating these reviews,however, is the scale of the problem. Programofficials can find it more trouble to petition forwaiver of the numerous regulations that are thoughtto be inappropriate, inapplicable, or obsolete than itis simply to ignore them and see if anybody noticesor cares.

Evaluating the success of these programs may bedifficult because some of them are already amongtheir Service’s highest priorities. At least one (theC-17) was the Air Force’s model program for aprevious initiative on Acquisition Streamlining, andhas therefore already received special attentiontowards streamlining.

The same approach of setting up a structure bywhich waivers to particularly obnoxious regulationscan be solicited and acted on is used in two otherDoD efforts, the Model Installations Program (MIP)and the Pilot Contracting Activities Program(PCAP). In each of these-one aimed at DoD basesand installations and the other at organizationsengaged in significant amounts of contracting—requests for waivers are forwarded to the individualswho can approve them, and if appropriate they aregranted on an experimental basis. If the experimentshows that the waiver should be extended in time orto a wider audience, proposals recommending theappropriate change are prepared.

Note that none of these processes has the power toremove constraints originating outside DoD-suchas legislation—because nobody within DoD has theauthority to waive those constraints. However, incases where outside constraints are identified, DoDcan request relief from the outside agency or fromCongress. Waivers to such outside constraints areencouraged so that the ones most limiting DoDactivities can be identified.

Shifting the “Burden of Proof"—Another possi-bility, more along the lines of the Packard Commis-sion and the DSB summer study recommendations,is to shift the “burden of proof" from the programmanager to those who wish to overrule the program

Table A-l-Defense Enterprise Programs

Army. . . . . . . . . . . . . . . Multiple Subscriber Equipment (MSE)communications system

TOW II missile● ATACMS missile

Navy . . . . . . . . . . . . . . . Trident II missileT-45 trainer systemSSN-21 submarine

Air Force . . . . . . . . . . . Medium Launch VehicleC-1 7 TransportSRAM II missileTitan IV booster

● The program has also been granted Milestone Authorization status byCongress. See preceding discussion of “Baselining,” beginnlng on pg. 21of the appendix.

SOURCE: Office of the Secretary of Defense.

manager. In this approach, most regulations wouldbe made advisory, rather than mandatory. Programmanagers would be free to decide which ones couldbe overridden in their particular circumstances. The“special interests” and “advocates” would still existand would still be free to make recommendations tothe program manager. However. the program man-ager would be free to disregard their advice—unlessthey could persuade the program manager’s supe-rior.

This system could only work if program managersand their superiors were evaluated not only on howwell individual programs fared but also on how wellthe programs on balance supported the intent of theregulations—which, after all, serve to incorporateDoD and national policies that senior policy makershave decided are important. Program managerswould have to realize that their goal is not simplydevelopment and deployment of a weapon systembut furthering national policy as well.

It is not clear that this approach could be pulled offsuccessfully. First of all, it requires a stable andhighly professional work force. Government by fiatand decree removes individual initiative. and for thatreason can compensate to some extent for anuntrained work force. The requirement for restoringinitiative is having people capable of exercising it.

Another, more intractable, problem is deciding onthe irreducible core of regulations that would remainmandatory. Discretion cannot be permitted in areasaffecting safety, for example, or in regard to matters


that are specified by law.114 It is not clear thatdeciding on an irreducible set of minimum, absoluteregulations would be any easier or more effectivethan the “bottom up” review of all regulationsdiscussed above.

Any implementation of a program of this sortwould have to be flexible. As time progressed,feedback as to which mandatory regulations neededrevisiting or which advisory guidelines were beingsystematically ignored would be used to makeadjustments. Every level of authority would have tosupport the program and cooperate to make it work.In an environment where tensions exist betweenCongress and the executive branch, between DoDand industry, between the military Services and theOSD, and within the Services, that maybe too muchto ask for.

Reducing Delays

Many of the delays built into the acquisitionprocess follow from the implementation of regula-tions and the operation of the bureaucracy asdescribed above. No particular delay can be ad-dressed in isolation. However, two problems inparticular seem to be mentioned frequently. They aresingled out for discussion below.

Reducing the Delays in Contracting-Much ofthe time and complexity of the contracting processstem from requirements and regulations that serve toenhance competition, to ensure that all potentialbidders capable of doing the work are given anopportunity to bid on it, and to support socioeco-nomic goals. The last two of these items-fairnessand socioeconomic goals-are policy goals thatCongress has found worth pursuing even if theyimpede defense acquisition. Like any other politicaldecisions, these judgments could be reversed ifCongress were to find that the benefits of pursuingthese goals did not justify their cost to the acquisi-tion system.

The first factor, however, stems not so much froma political judgment that competition is inherently

good as from the fact that competition-at least in acommercial market-is the mechanism that pro-vides the buyer better quality at a lower price.Competitive purchasing in defense procurement isoften misinterpreted to mean competition on thebasis of price alone. While this might have been truein the days of “formal advertising” or sealed bidsthat used to characterize government procurement.passage of the Competition in Contracting Act of1984 extended the concept of competition to includenon-price factors. Some argue that price is still tooheavily weighted, but it is clearly not the only factorthat can be considered.

The debate concerning competition in defenseprocurement concerns how well the concept can beextended from the free market-where it clearlymakes sense—to the highly regulated defense indus-try, which is characterized by few sellers, a singlebuyer, and the requirement to create new systemsthat press the state of the art.

The Packard Commission very strongly endorsedthe concept of competition:

Commercial procurement competition simultane-ously pursues several related objectives: attractingthe best qualified suppliers, validating productperformance and quality, and securing the bestprice . . . we believe that DoD should greatly in-crease its use of truly effective competition, using asa model the competitive buying practices of majorcorporations and their suppliers. 115

However, 2 years later, Commission chairmanDavid Packard appeared to have changed his mind—at least as far as competing major acquisitionprograms is concerned—when he said, “One coulddo as good a job awarding major contracts bythrowing darts at the names of qualified bidders.’’116

The contracting and bid award process has comeunder increasing scrutiny recently amid allegationsof serious improprieties in bid preparation andselection. This area will certainly be looked intofurther. However, nobody has yet come up with amechanism by which all the benefits of competing

114 Note that this statement does not imply that all existing laws should necessarily be retained under this approach. Indeed, to alter the present regulatoryregime, substantial legislative change would be required. Nevertheless, those laws that remain in force cannot be waived at the discretion of a DoDofficial.

115 President's Blue Ribbon Commission on Defense Management, op. cit., footnote 48, pp. 64-65.116 Quoted in James Flanagan, “Competition in Defense Buying Costly to U.S.,” Los Angeles Times, July 31, 1988.

Appendix A—The DoD Aquisition System ● 39

major acquisitions can be preserved in a lesscumbersome process.

Contracting mechanics should pose less of aproblem in procuring research than in procuringsystems. The Competition in Contracting Act ex-empts “research” form many of its provisions, andDoD had previously taken this exemption to applyonly to budget category 6.1. However, a letter fromMembers of Congress to the Secretary of Defensemade clear that this exemption applies to technologybase activities-research and exploratory develop-ment—in general. 117

Reducing the Delays in Review-Considerabletime is taken in preparing for oversight reviews bythe DAB or Service equivalents. With poor plan-ning, activities of the program under review grind toa halt while the necessary documentation is preparedand analyzed. Appropriate planning should providefor delay, using the span between submission ofdocumentation (3 months prior to the DAB meeting)and the review’s outcome for work that does notcommit large sums of money to anticipated out-comes of the review.

These reviews almost never lead to programcancellation, so in practically every case, programofficials can foresee activities to be conducted afterthe board review no matter what the review outcomeis. Obviously, major full-scale development con-tracts should not be let pending the decision toproceed to full-scale development. However, manyactivities that would facilitate the FSD process-orthat might occur during full-scale development butdo not involve commitment to a major FSD con-tract—could be conducted while awaiting an FSDgo-ahead.

Some funds might be jeopardized because manag-ers conducted activities judged inappropriate in thelight of subsequent oversight board decisions. How-ever, these expenses would almost certainly beoutweighed by the savings made possible by permit-ting large development teams to do useful work,rather than wait idly by, during the period pendingan oversight review.

Organization

. . . good organizational design alone will not exor-cise all the demons in the weapon system acquisitionprocess, but the lack of it is almost sure to keep themthere.

—Edwin A. Deagle l18

So far this appendix has discussed acquisition‘ procedures within the existing DoD organization.

However, there are other organizational models,some of which were proposed in various pieces oflegislation introduced in the 100th Congress. Thesebills run the gamut of acquisition structures fromthose similar to current practice to substantialdepartures from it:

●

●

●

●

Not

H.R. .3898 (Kasich): Gives the USC(A) prece-dence over the Service secretaries. This prece-dence is asserted by DoD regulation in acquisi-tion matters, but regulations do not make clearwhether the Service Acquisition Executivesreport directly to the Under Secretary of De-fense for Acquisition through the ServiceSecretaries.S. 2621 (Dixon): Centralizes procurementauthority under the USD(A) but permits it to bedelegated back to the Services.

S. 2732 (Roth) and H.R. 4950 (Hertel): Estab-lishes under the USD(A) a Defense AcquisitionAgency or Corps that receives requirementsfrom the Services and then completes theacquisition process, giving the USD(A) finalauthority over procurements. Terminates theprocurement authority of the Service Secretar-ies and prohibits delegation of certain USD(A)authority back to the Services.

H.R. 5048 (Boxer): Establishes an IndependentProcurement Corps outside the Department ofDefense to research. develop. and producemajor weapon systems for DoD.

included in this list—yet—are even morefar-ranging ideas such as regulating the defenseindustry as a public utility, or even nationalizing it.


The approach suggested under the third of these:alternatives—consolidating all procurement activityunder the USD(A)-was considered but rejected bythe Packard Commission.

. . . such centralization would not serve the cause ofreducing the bureaucracy, because it would tend toseparate further the acquisition staff from the mili-tary user. We believe that it is important to maintainthe Services’ traditional role in managing newweapons systems.119

The program manager, argued the Commission,must understand the operational uses to which thesystem will be put and the environment in which itwill operate.

However, some analysts share the viewpoint ofLeonard Sullivan, a civilian writing for the CSISDefense Acquisition study, who argues that militaryinvolvement in acquisition is far too extensive:

The U.S. acquisition system is laced withusers . . . they are almost anyone in uniform exceptthe equipment operators in the field. And they havedone a poor job keeping the acquisition process onthe straight and narrow.

A military person’s judgment about technicalfeasibility, costing and budgeting, quantitativeanalysis, affordability, and supportability is no betterthan, and may be worse than, that of a professionalcivilian . . . The role of user is a convenient mythperpetuated by the military to increase its presenceand by civilians to rationalize dubious decisions.120

Proponents of a centralized civilian acquisitionagency argue that only such a mechanism can fosterthe professional, stable, qualified work force neededto implement true reform.

Taking acquisition away from the Services andturning it over to a civilian agency would representa radical change. Most individuals involved indefense procurement-within DoD and in industry,military and civilian-do not favor such a sweeping

change at present. Most studies of the issue have,like the Packard Commission, recommended againstit. One major exception is the President’s PrivateSector Survey on Cost Control, or Grace Commis-sion. The Grace Commission recommended that“consolidation of the management of the acquisitionprocess within the Office of the Secretary of Defense(OSD) would improve efficiency and provide oppor-tunity for significant cost savings.”121

A somewhat more tenuous endorsement of theidea was provided by the Project on MonitoringDefense Reorganization, a study of the implementa-tion of the Packard Commission recommendations.This study stated a preference for leaving acquisitionauthority with the Services, but recommended con-sideration of an independent organization under theUSD(A) in the event that the Services refused tocreate specialized “acquisition corps.” The studyconcluded that “radical steps, such as the establish-ment of a single procurement organization withinthe department [of defense], should not permanentlybe ruled out.’’ 122

The GAO found that the prevailing opinion itencountered in a study of centralized acquisition layagainst establishing such an agency.123 Some of theadvantages to such an agency cited by GAO were

. reducing Service parochialism and fosteringmore common/joint system development;

● improving the quality and continuity of theacquisition work force; and

. reducing the size of the work force and elimi-nating administrative layers by consolidatingduplicate acquisition functions.

Some of the more significant disadvantages were:

. Inability to address acquisition problems thatwere not organizationally related. Many prob-lems with the existing system were thought tobe in this category, such as those involving


●

●

●

If a

identifying what weapons to buy and tradingoff military requirements against cost.Possible disregard of military operational expe-rience that could support claims that the newequipment is operationally suitable and effec-tive for military use.Adding an additional layer of bureaucracy.The potential large size of such an agency,which could render it unmanageable.

centralized acquisition agency were formed,GAO recommended that it remain within the DoD.GAO reported the “overwhelming opinion” of thosewith whom it spoke that the Secretary of Defenseshould be accountable for all resources dedicated todefense.

A RAND Corporation study concluded that thereis no reason to believe a centralized acquisitionagency would operate more effectively than theexisting system. Inputs from military users “prob-ably receive insufficient attention even today, and itis difficult to believe that the interests of the userswould be better represented by a more civilianizedmanagement.” 124 The study recommended changesin the acquisition process, rather than the acquisitionorganization.

Although study of European nations that usecentralized procurement systems might illuminatethe successes or failures of such a plan, factorsbesides their centralized procurement systems makesuch analyses difficult. One important difference isthat their defense programs are small compared tothat of the United States. Other differences, aspresented in a recent study of European weaponsacquisition practices by The Analytic SciencesCorporation, 125 are that:

●

●

European military Services do not dominateacquisition.Multiyear defense plans dominate fiscal plan-ning in Europe and make it impossible to obtainprogram funds not in the multiyear plan.

. The annual defense procurement budget isapproved by the legislature with minimalchanges.

. The government imposes minimal “how-to”requirements on the defense industry.

. Industrial policy is a major consideration indefense contracting.

According to this study, the U.S. approach toacquisition, when compared to the European one,results in considerably more sophisticated andcapable weapons developed over a shorter period athigher cost, but with lower cost per unit perform-ance. The advantages of the European model-earlyanalysis of cost v. performance, adherence to long-range fiscal plans, and concern for affordability--donot require a centralized acquisition agency toachieve. Moreover, if U.S. acquisition activitieswere centralized in a single agency, that agencywould have about 15 times the staff and budget ofthe largest European acquisition agency.

Personnel

There has always been an implicit assumptionwithin the Defense Department that people with littleor no advanced training and experience in themanagement of large industrial programs couldfunction effectively at any management level. Thisassumption has been a key factor leading to thedisappointing results of virtual] y every improvementprogram in the last twenty years.

—J. Ronald Fox, with James L. Field 126

Documentation

Successful implementation of many recommen-dations for improving defense acquisition-severalof which have been cited in previous sections—requires a high-quality, stable, and well-trainedacquisition work force. In a letter to PresidentReagan one year after the publication of the PackardCommission report, 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.127

Improvements recommended by the PackardCommission included reducing the barriers to re-cruiting senior-level executive branch personnel,128

attracting qualified new personnel, improving thetraining and motivation of existing personnel at themiddle management levels, and continuing therecent improvements in defining military careerpaths in acquisition.

The Commission thought that civilian acquisitionpersonnel needed much more attention than military,and cited many of the deficiencies of the federalCivil Service system that are described in the contextof national laboratory personnel in chapter 5 of themain report. Recommendations particular to theacquisition work force included enhancing the statusof the contract specialist job classification. Atpresent, this classification is an “administrative”series position, prohibiting establishment of anybusiness education requirement; the Commissionrecommended moving this position to the “profes-sional” series. The Office of Personnel Manage-ment, which classifies Civil Service positions, hasresisted this change on the grounds that DoD is freeto require a business-related college degree for anyparticular contract specialist position, but that re-quiring such a degree for all such jobs is arbitraryand unnecessary.

In a major study of defense acquisition, ProfessorJ. Ronald Fox of the Harvard Business Schooldistinguished between two prevailing attitudes to-wards the government’s role. Those holding what heterms the liaison manager view believe the govern-ment program manager seines primarily to promotea program, prepare progress reports, negotiate withvarious parties within DoD, and resolve conflictsbetween these parties and the contractor. Costcontrol is solely the responsibility of the contractor,and there is no need for the program manager to haveextensive training or experience with industrialmanagement and cost control methods. Program

management is therefore a reasonable rotation formilitary officers between operational assignments.Those holding the liaison manager view, accordingto Fox, are widespread in both government andindustry. They see the present acquisition process asessentially well managed, with few problems.

Fox himself believes very strongly in an alterna-tive that he terms the active manager view. In thisformulation, the program manager’s role is one ofplanning, rigorous oversight, negotiation with, andcontrol over the contractors. Responsibility for costcontrol is shared between government managers andthe contractor; by establishing and implementingincentives, both formal and informal, the programmanager has significant opportunity to reduce coststhroughout the life of the program. The existingsystem of staffing and training military programmanagers cannot produce individuals capable oftaking this role:

As in industry, the development of highly quali-fied program managers requires focused careerpaths, progressing from technical work to assign-ments at laboratories, program offices, and plantrepersentative offices, to full program managementresponsibility for small programs, and ultimately forlarge programs. There is no time left to becomeexpert in a military operational specialty as well. 129

Civil Service personnel share few similaritieswith military officers in acquisition assignments,according to Fox. Whereas the short tenure ofofficers in acquisition rotations severely impedestheir ability to match their industrial counterparts,many civil servants “remain for so long that theyresist innovation and change.”130 Fox recommendsreforming civil service regulations to establishhigher standards and permit removal of mediocreperformers. Absent these changes. “defense acquisi-tion programs will appeal primarily to those satisfiedwith the present low level of responsibility. ”

The DSB 1987 Summer Study on TechnologyBase Management recommended establishing a trial“Senior Scientific, Technical, and Acquisition Exec -


utive Initiative” to investigate means of improvingthe quality of personnel involved in defense acquisi-tion and DoD technology base program executionand management. This program would provide up to100 non-tenured positions for senior managersserving 3-year, renewable terms. One of the keyfeatures of this program would be to providecompensation comparable to equivalent positions inacademia or industry through a special mechanismthat would be outside conventional Civil Serviceregulations and limits. Poor performers would not berenewed. The summer study saw conflict of interestregulations, which restrict interchange of seniorpersonnel between government and industry (the“revolving door”), as the most serious impedimentto instituting such a program. “Some form of conflictof interest waiver-requiring legislative action—will be required to make the demonstration trulyeffective. ”131

Analysis

All proposals for reforming personnel policies runinto conflicts between competing objectives. Sig-nificantly increasing the tenure of military personnelin acquisition assignments, and weighing thoseassignments more heavily in promotion reviews,would probably improve acquisition. However,those actions would require making significantchanges to what senior military officers now con-sider to be requirements for successful militarycareers.

Making fundamental reforms to Civil Serviceprocedures— or even exempting groups fromthem—would also pose substantial political difficul-ties. Federal employees already feel as if they have240 million supervisors, and it sometimes seems—at least while reading “Letters to the Editor”columns when civilian pay raises are debated inCongress-that there is nothing so despised as acivil servant. proposals that would increase compen-

sation or other benefits of Federal employment in aneffort to attract more senior and more highlyqualified employees would be seen by others asadding slots to the Federal trough.

Conflict-of-interest regulations provide a case inpoint. Some argue, as did a panel of senior industryofficials advising the Senate Armed Services Com-mittee, that:

There can be no question about the need to attractcompetent industry-trained men and women intovital upper-middle level appointee positions in thePentagon. “Revolving Door” legislation, howeverwell intended, defeats this need. The stigma of evilassociated with the “revolving door” issue is mostunfortunate and largely unwarranted. 132

Contrast that attitude with the following:

Weapons makers and weapons buyers shouldhave different perspectives, and therefore differentskills. Thus. there should be no tendency to share thesame labor pool . . . Whether or not these people[who go back and forth between government andindustry] are bribed, or promised future employ-ment, they will be caught up in a loyalty to theproject(s) they work on. They have lost theirconsumer’s perspective.133

It will be difficult, if not impossible. to reconcilethese two points of view. Those insisting on strict“revolving door” legislation to prevent officialsfrom consciously misusing their public office forprivate gain might be satisfied that extraordinarilysevere penalties could deter blatant conflict ofinterest violations. However, those more concernedabout the “loss of perspective’ ’-the suspicion thatthe interests of government and those of industryshould not be so closely aligned that individualswould be able to work just as effectively in one as inthe other-would probably not agree that tougherpenalties for violations of law would help clarify thismore ambiguous situation.

131 U.S. Department of Defense, “Report of the Defense Science Board 1987 Summer Study on Technology Base Management,” prepared for the Officeof the Under Secretary of Defense for Acquisition, December 1987, p. 21.

l32"Report to the Subcommittee on Defense Industry and Technology, Senate Armed Services Committee,” by the Ad Hoc Defense Industry AdvisoryGroup (13 senior defense industry officials), p. A-1.

The DoD Acquisition System - princeton.edu

Documents

Transcript of The DoD Acquisition System - princeton.edu