CONGRESS OF THE UNITED STATES CONGRESSIONAL ......NASA's Program and Budget Since the mid-1980s, NASA's program has required consistent growth in its out-year budgets, even after adjusting

CONGRESS OF THE UNITED STATESCONGRESSIONAL BUDGET OFFICE

REINVENTING NASA

The Congress of the United StatesCongressional Budget Office

NOTES

Unless otherwise indicated, all years are fiscal years.

Cover photo of the Hubble Space Telescope is courtesy of the NationalAeronautics and Space Administration.

Preface

T he National Aeronautics and Space Administration (NASA) is confronting thedifficult task of squeezing a program that it anticipated would cost about $95billion for five years into a budget plan allowing just over $70 billion. Thisstudy, which was prepared for the House Committee on Science, Space, and Technology,evaluates NASA's strategy for coping with the expectation of lower funding in the futureand develops a set of illustrative alternatives that would reduce the scope of NASA'smission. In keeping with the mandate of the Congressional Budget Office (CBO) toprovide objective analysis, the study makes no recommendation.

David Moore of CBO's Natural Resources and Commerce Division wrote the studyunder the supervision of Jan Paul Acton and Elliot Schwartz. Molly K. Macauleyprovided extensive comments and helpful suggestions. Useful reviews were also providedby Ronald Konkel, William Lilly, Howard McCurdy, Carole Neves, John Pike, MarciaSmith, Karen Tyson, and Ray Willamson. Other valuable comments came from JohnPeterson, John Sturrock, and William Thomas within CBO.

Leah Mazade edited the manuscript, and Christian Spoor provided editorialassistance. Angela Z. McCollough and Donna Wood typed the drafts. Aaron Zeislerprepared the figures. Kathryn Quattrone, with the assistance of Martina Wojak-Piotrow,prepared the study for publication.

Robert D. ReischauerDirector

March 1994

Contents

ONE

TWO

THREE

FOUR

SUMMARY

THE NATIONAL AERONAUTICS ANDSPACE ADMINISTRATION IN TRANSITION

What Is at Stake? 1The Structure of NASA's Program

and Budget 5

THE RISKS OF A STRATEGY OFMARGINAL ADJUSTMENT

Program Characteristics 11Marginal Adjustment and Problems

with the Current Program 14

CHANGING THE WAY NASADOES BUSINESS

Criticizing NASA's Conduct 19Management Reform 21Procurement Reforms 23A New Relationship with

the Private Sector 25Streamlining 27Intragovernmental Coordination and

International Cooperation 29Total Quality Management 31Can Reforms Make a Difference? 33

A FOCUSED STRATEGY AND ALTERNATIVEPROGRAMS FOR NASA

Program Alternatives 35Comparing Benefits 40

IX

11

19

35

vi REINVENTING NASA March 1994

TABLES

1.

2.

3.

4.

5.

Initial Operating Plan for the NationalAeronautics and Space Administration, 1994

Budget of the National Aeronauticsand Space Administration, 1984-1994

Budget Requests and Appropriations forthe National Aeronautics and SpaceAdministration, 1989-1994

Five-Year Budgets in the 1993 and1994 Plans of the National Aeronauticsand Space Administration

National Aeronautics and Space Administration's1993 Operating Plan and Alternatives

10

10

36

FIGURES

S-l.

1.

2.

Five-Year Budget Requests of theNational Aeronautics and SpaceAdministration, 1990-1995

Budget Shares for Piloted Spaceflightand Space Science, National Aeronauticsand Space Administration, 1984-1993

Five-Year Budget Requests of theNational Aeronautics and SpaceAdministration, 1990-1995

BOXES

1.

2.

3.

Short-Term Economic Effects of Spendingfor the National Aeronautics andSpace Administration

Changes to the National Aeronauticsand Space Administration's AppropriationsProposed in the 1995 Budget

A Brief History of the NationalAeronautics and Space Administration'sContract Preferences 24

Summary

T he National Aeronautics and Space Admin-istration (NASA) is confronting the difficulttask of reinventing its program within theconfines of a dramatically lower five-year budgetplan. The agency has chosen a two-pronged strat-egy: maintaining the broad structure of its programwhile marginally adjusting its content by stretchingout, scaling down, and canceling some of its proj-ects; and buying more program with its appropria-tion by doing business more efficiently. This studyexamines that strategy and a set of alternatives thatwould focus NASA's program more tightly on oneor another of its three major traditional objectives--piloted exploration of space, the generation of newscientific knowledge, or the development of spaceand aeronautical technology-under an annual bud-get of no more than $14.3 billion.

The analysis concludes that improving the wayNASA conducts its business-buying more for less~is unlikely to produce significant budgetary savingsin the next five years. A disproportionate share ofthe burden of living with lower budgets is likely toinvolve adjustments to the content of NASA's pro-gram—buying less for less. If so, the distinguishingcharacteristics of that program (high fixed costs forprojects with long operational lives), coupled withthe agency's tendency to underestimate the cost ofits projects, increase the risk that NASA's strategywill lead to greatly reduced productivity in the formof deferred, diminished, or even lost benefits.

An alternative to the current course would be tofocus the agency's efforts on narrower objectives.Projects in the emphasized areas would then haveadequate budgets, and the chances would be greaterthat NASA would deliver a productive program-one that produces benefits as promised in a timelyway. This strategic alternative would explicitlyforgo other benefits that NASA's program might

deliver, but it would save the costs of pursuingthem in cases in which the risk of failure was high.

NASA's Program and Budget

Since the mid-1980s, NASA's program has requiredconsistent growth in its out-year budgets, even afteradjusting for inflation. By itself, a program planthat requires real increases in funding need notevoke criticism. But in today's environment offiscal restraint, NASA's plan has generated criticismbecause of the agency's recurring problems in esti-mating the costs of its program and because ofshortfalls in the performance of some of its majorprojects.

Concerns about the cost of the NASA programincreased after 1990, when the Budget EnforcementAct (BEA) required a tightening of all domesticdiscretionary spending. As the BEA's caps onspending began to bind, the Congress significantlylowered NASA's budget from the level requested byPresident Bush for 1992 and 1993. In 1992, thebudget request for NASA was $15.8 billion, but theCongress appropriated only $14.3 billion. In 1993,the agency's appropriation was again $14.3 billion,$700 million below the Administration's request of$15 billion.

In this context, the Clinton Administration'sproposal to slow the growth in NASA's budget by$16 billion over the 1994-1998 period represented asignificant change (see Summary Figure 1). Never-theless, the Congress voted a smaller appropriationthan the Administration's request for 1994: it re-duced the Administration's figure of $15.2 billionby $700 million to $14.6 billion. The secondClinton budget proposal for 1995 through 1999

x REINVENTING NASA March 1994

Summary Figure 1.Five-Year Budget Requests of the National Aeronautics and Space Administration, 1990-1995(In billions of dollars of budget authority)

22

20

18

16

14

12

10

Billions of Dollars of Budget Authority

1992Request

1991Request

\

*' '.-•' .--' ,••• 1993Request

1990Request

1994Request

1995Request

Actual

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999

SOURCE: Congressional Budget Office based on Budget of the United States Government (various years) and 1993 projections fromthe NASA Comptroller's Office.

flattens NASA's funding even more and for the firsttime in 21 years requests less for NASA in thecoming budget year ($14.3 billion for 1995) thanwas provided in the current year ($14.6 billion).

To adapt to the new budgetary realities, NASAhas chosen to adjust the content of its programmarginally and improve its efficiency. If successful,this strategy would permit the agency to pursuesimultaneously objectives in piloted spaceflight,space science (using robotic spacecraft), and aero-nautics and space technology useful to both the

public sector and private aerospace industries. Atstake are the benefits of NASA's projects in theseareas-for example, new knowledge about the uni-verse or progress toward the piloted exploration ofMars. Such benefits are directly observable butdifficult to measure and value. Most research onthe effects of past NASA spending and the benefitsof its program does not substantiate the claim thatthe choices among program objectives or fundinglevels for NASA will have significant implicationsfor the U.S. economy.

SUMMARY

The Risks of MarginalAdjustment

A part of NASA's strategy to adapt to new budgetrealities is to delay, scale back, and cancel someprojects but maintain the overall structure of theprogram that the agency has sought to establishsince the early 1980s. That structure includes de-veloping and operating piloted spacecraft (the spaceshuttle and the space station), developing and oper-ating robotic spacecraft (for example, the EarthObservation System and the Hubble Space Tele-scope), and making continued advances in rocketand satellite, aeronautical, and other systems andtechnologies necessary to support the nation's publicand private aerospace activities. Essential character-istics of NASA's current program heighten the risksof the strategy of marginal adjustment. Moreover,long-standing concerns about the productivity ofNASA's overall program will intensify as a strategyof marginal adjustment is pursued.

High fixed costs and support for long-termmission operations and data analysis in order torealize benefits are two characteristics of many ofNASA's projects that complicate a strategy of mar-ginal adjustment. High fixed costs imply that rela-tively large cuts in the activities of a program pro-duce only small budgetary savings. For example,cutting the space shuttle's annual rate of flights by25 percent (two flights of the normally scheduledeight) reduces the operating costs of the shuttlesystem by less than 5 percent. Similarly, reducingthe funding necessary to operate space science mis-sions and analyze the data they produce can inflict adisproportionately high cost in lost benefits, whichlowers the return on NASA's substantial past invest-ment in spacecraft and facilities.

NASA's strong tendency to underestimate thecost of its projects is a third characteristic that com-pounds the risk of the agency's marginal adjustmentstrategy. Extensive documentation compiled by theGeneral Accounting Office (GAO) and the Institutefor Defense Analysis attests to NASA's poor recordin this regard. The prospect that large numbers ofprojects in NASA's program will cost more thananticipated complicates decisions about which pro-grams to downgrade, delay, or cancel, and further

increases the possibility that the benefits of NASA'swork will be deferred, decreased, or lost.

Concerns about the content and worth ofNASA's program might well arise even if cost andbudgetary problems were not evident, but thoseconcerns are strengthened by the adjustments thatNASA is making in the content of its program toreduce its budgetary requirements. First is thequestion of people in space. On the one hand,supporters of piloted spaceflight and human explora-tion are unhappy with the slow pace of these activi-ties. On the other hand, critics argue that NASA'sdecision to spend more than 50 percent of its budgeton piloted spaceflight crowds out more worthyscience and technology projects.

A second content issue is that NASA's scienceprogram is dominated by projects that critics labelas too big, too expensive, and too long-lived. Forexample, the Hubble Space Telescope cost billionsof dollars to build and operate; the life span of theproject, from the beginning of development to theend of operations, is expected to be at least 20years. Critics contend that "cheaper, better,quicker" missions are preferable: although suchmissions are less ambitious than recent large proj-ects, more of them can be supported, and they in-flict a lower cost in lost scientific benefits if theyfail.

Third, the content of NASA's program has beencriticized as unresponsive to the economic chal-lenges facing the nation. This viewpoint calls formore emphasis on projects to increase private pro-ductivity-for example, research and developmentsupporting U.S. aircraft, rocket, and satellite manu-facturers.

Changing the Way NASADoes Business andReducing Program Costs

Changing the way NASA does business, the secondpart of the agency's strategy to adapt to lower bud-gets, may offer improvements in program manage-ment and technical performance and some reduc-

xii REINVENTING NASA March 1994

tions in costs. But the associated budgetary savingsare uncertain and unlikely to be realized in the nearterm. Accordingly, the first element of NASA'sstrategy to adapt to lower budgets in the future-making adjustments in the content of its program-will have to bear more of the burden of loweringcosts. The Congressional Budget Office (CBO) hasreviewed six types of proposals for improving theway NASA conducts its business.

Ongoing Management Reforms

Proposals for reforming NASA's management em-phasize better planning, uniform and more central-ized review of projects, improved cost estimatingindependent of program advocates, and developmentof measures of contractor and program performance.If successful, the proposals could allow the agencyto control its costs better, but the effects of theproposals are more likely to be felt in the futurebecause improved management and planning willinfluence new programs more than current ones.Making a success of these efforts will require asteadfastness among NASA's leaders not alwaysevident in the past. For example, funds for plan-ning projects carefully early in their life cycle havebeen cut in difficult budgetary times, despite theacknowledgment by senior management that suchfunds are necessary to avoid future problems.

Ongoing Procurement Reform

NASA proposes three major changes in its procure-ment process: modifying the agency's proceduresfor incentive contracting, placing more weight on acontractor's past performance when awarding newcontracts, and streamlining midrange procurement(purchases between $25,000 and $500,000).

Formal evaluations of the Department ofDefense's (DoD's) use of incentive contractingsuggest that incentives helped to hold down growthin the costs of developing strategic missile systemsand satellites. Because NASA has long practicedincentive contracting, the changes currently beingconsidered are unlikely to lower costs significantly.But NASA may be able to improve its technical

results by basing a contractor's incentive fees moreon the performance of finished systems than onmeeting interim schedules and cost goals, and em-phasizing a contractor's past performance whenawarding new business.

NASA spent only 13 percent of its 1992 fund-ing for procurement under contracts covered by itsMid-Range Procurement Initiative. Thus, even anextremely successful reform effort that reducedcosts by 5 percent would save only about $85 mil-lion annually. The initiative might yield additionalsavings by decreasing the number of NASA em-ployees needed in the procurement area. However,increased productivity in procurement activities ismore likely to allow the agency to make do with asmaller increase in personnel for that area than wasrecently recommended by examiners for both theexecutive branch and the Congress.

A New Relationship withthe Private Sector

The possible relationships between NASA and theprivate sector span a wide range. At one extreme isNASA's traditional mode of acquisition, which ischaracterized by extensive and direct involvement ofthe government in all phases of activity. At theother extreme is purchasing data and services pro-vided by private firms that are wholly responsiblefor the design, production, launch, and operation ofthe spacecraft necessary to provide those products.Among the candidates for purchases on commercialterms are NASA's communications satellites or theservices they provide, data needed for research onthe global climate, and launch services for smallscientific payloads.

The vision that underlies suggestions to buymore on commercial terms emphasizes two points.First, the aerospace industry can produce the techni-cally sophisticated products that NASA needs morecheaply without government supervision than withthat oversight. Second, NASA has a self-defeatingtendency to drive up the cost of the hardware itbuys in the traditional manner through excessiveoversight, overly detailed design specifications, andtoo many contract changes.

SUMMARY

Similar concerns have been raised about thedefense acquisition process. A 1993 study by theDefense Science Board, for example, argues thatDoD's acquisition costs could be reduced by asmuch as 20 percent, largely by applying commercialpractices. That conclusion should be treated astentative, however, because it is based on a smallnumber of cases and expert judgment. The boardused those factors to develop rules of thumb that itthen extrapolated to the entire defense budget. Inaddition, the study concluded that the savings itprojected were likely to accrue only after five yearsof determined reform, a point as applicable toNASA as to DoD.

Purchasing on commercial terms has drawbacksand limitations. In some cases, the government'spotential savings from commercial purchasing maybe offset by the higher relative costs of privatefinancing and insurance, which are included in theprices that the government pays for commercialproducts and services. In other cases, the substan-tial risk involved in developing the new technol-ogies necessary for some NASA programs makescommercial purchasing inappropriate. Finally, therisk of loss of human life in piloted spaceflight maypreclude applying the more hands-off governmentposition typical of commercial purchasing to thoseprograms.

Streamlining

Proposals for streamlining overlap with the call forNASA to buy more on commercial terms. Theyfocus on increasing the authority and responsibilityof program managers and prime contractors byloosening procurement and acquisitions regulationsand decreasing the role of NASA's field centers inprogram management.

Advocates of streamlining NASA's procurementprocess point to two examples to support their case:classified military space projects and the recentexperience of the Strategic Defense Initiative Orga-nization (SDIO). Yet no public studies have shownthat the universe of classified military projects hasdemonstrated superior cost, schedule, and technicaloutcomes compared with NASA or open militaryacquisitions. The SDIO claim also seems unsup-

ported by serious analysis. Moreover, regarding thefederal acquisitions regulations in particular, theGeneral Accounting Office's "High-Risk Series"review of NASA's contract management found thatthe agency has often failed to comply fully withprocurement requirements. That failure has ledNASA's field centers to approve contract changeswithout adequate technical evaluation and to allowunpriced contract changes to persist. Such factorshave contributed to cost overruns and unsatisfactoryperformance.

New Cooperative Ventures

The Cold War prevented NASA from taking fulladvantage of joint ventures with the U.S. military orwith foreign governments. Now that that conflicthas ended, many observers have suggested thatNASA could increase its productivity by enteringinto new cooperative efforts. The agency has takenup that suggestion and is aggressively pursuing newinternational cooperation in piloted spaceflight withthe Russian Space Agency. The Administration isalso examining the prospect of integrating NASA'sEarth observation efforts with the polar satelliteprograms of DoD and the National Oceanic andAtmospheric Administration.

The major focus of the new cooperation withRussia is to develop and subsequently operate aspace station, an effort that carries both risks andrewards. The evolving and preliminary plan for thenew international station would restore some of thecapability lost in the earlier redesign and virtuallyall of the lost schedule. The costs of the venturewould be lower than those estimated for Freedom(an earlier design) and would not exceed the $2.1billion cap that the Administration has placed onannual spending for the station for the next fiveyears. Whether the current estimates of costs holdup will not be known until late summer 1994, whenfinal contracts with the prime contractor, Boeing,are expected. Integrating U.S. and Russian hard-ware, computer software, and operating procedurescould prove difficult, however. In addition, politicaltensions could always stop the project in its tracks,forcing NASA to either cancel the station (and losethe chance of a return on its past investment) orredesign the program yet again.

xiv REINVENTING NASA March 1994

Total Quality Management

Total quality management (TQM) is a managerialphilosophy whose objective is to achieve customersatisfaction through continuous improvement ofproduction processes. Customer satisfaction and thepositive performance indicators that go with it areachieved by committed managers and empoweredemployees seeking to continuously improve theirproducts by applying empirical data and analysis toproduction processes. First adopted by privatemanufacturing firms in Japan, the approach spreadto private manufacturers in the United States in thelate 1970s, achieved wide acceptance in the 1980s,and by late in that decade was being adopted bylarge parts of the federal government. AlthoughTQM originated in manufacturing, it has spread tothe service sector, where it has gradually won ad-herents.

NASA was one of the first federal agencies toadopt TQM during the late 1980s. According to a1992 GAO survey, eight NASA installations em-ploying roughly 20,000 people have adopted TQM.GAO asked respondents to place themselves in oneof five phases of TQM. Four of the NASA installa-tions placed themselves in the second phase, "justgetting started"; three in the third phase, "implemen-tation"; and one in the fourth phase, "achievingresults." (The first phase is "deciding whether toimplement TQM," and the final phase is "institu-tionalization.")

The GAO survey reported two sets of results:external organizational performance—the implement-ing agency's assessment of its relationships with itscustomers~and internal operating conditions. ForNASA installations and for a larger survey sampleof more than 2,200 other federal installations, self-reported improvement was correlated with progressalong GAO's five-phase scale. Those improvementsincluded reductions in costs, although GAO did notreport the size of the reductions or the categories ofeffort in which they occurred.

Even if NASA's adoption of TQM is ultimatelysuccessful, it is unlikely to lower the cost of theagency's program or to have a significant budgetaryimpact, at least in the next several years. The pri-vate sector's experience with TQM indicates that it

is most effective when consistently practiced over along period and when improved quality precedesreduced costs. The federal experience with TQM,including NASA's, is relatively limited. Thesefindings should create skepticism about claims thatimmediate cost savings will follow the decision toemploy TQM.

Alternative Programsfor NASA

An alternative to NASA's strategy of adjusting tolower future budgets is to narrow substantially thefocus of the agency's activities. If NASA's prob-lem is trying to do too much with too few dollars,one solution is to do less. Narrowing NASA'sfocus directly addresses the issues of cost and pro-gram content and might even provide more opportu-nities for effective reform of the way NASA doesbusiness. By explicitly forgoing some benefits,budget costs could be reduced. Moreover, the like-lihood would be increased that NASA could actu-ally achieve results and obtain benefits in the areasin which its resources were concentrated.

CBO has developed and evaluated three alterna-tives to NASA's current program, each of whichillustrates the option of a more focused strategy.Each emphasizes one of the three major objectivesthat NASA historically has pursued, although noalternative is a specific proposal of NASA's critics.The annual budgetary cost of each alternative is$14.3 billion or less, ranging from a program fo-cused on piloted spaceflight for the full $14.3 bil-lion to a program emphasizing technological devel-opment and robotic space science for $7.0 billion.The budgetary cost of each of the programs rangesdownward from the current level of funding becauseof the national emphasis on deficit reduction and theprospect of diminishing returns to dramaticallylarger investments in programs with more limitedobjectives than the current one. The budgetarycosts of the second and third alternatives are pre-sented as point estimates. Actual costs could varyas much as a $1 billion above or below the esti-mates.

SUMMARY

The three alternatives presented here are broadoutlines that might be better viewed as end points ina process of adjustment rather than starting points.They neither include strategies for transition fromthe current program, nor do they take account oftransition costs. The three alternatives are:

o A program that emphasizes piloted space-flight at a sustained budget of $14.3 billionannually. To fund this option, plans for roboticspace science would be cut. This alternativeconcentrates on the space station program andon new technology to support future pilotedexploration of the solar system. It would fundthe space station program at a higher level toensure its timely completion and secure thebenefits of the program, including improvedrelations with Russia. This alternative alsoresponds to the criticism that NASA's currentprogram does not give a high enough priority tofuture human exploration of the solar system.Spending for space science and technologyactivities in areas that do not directly supporthuman exploration would be reduced dramati-cally. Yet the pace of piloted exploration islikely to be slow, as most estimates of the costof a base on the Moon or a piloted mission toMars make such activities difficult to affordwithin a constrained budget.

o A program that emphasizes robotic spacescience at a budget of $11 billion a year andincludes piloted spaceflight only for scientificpurposes--a criterion under which the spacestation would be canceled. This alternativeemphasizes the creation of new knowledge,including that gained in piloted spaceflight, butit does not support piloted spaceflight for thepurposes of improving relations with Russia orpreparing for future piloted exploration of theMoon or Mars. This content mix addresses

the criticism that NASA's program places toomuch emphasis on piloted spaceflight when theagency's major contribution is creating newscientific knowledge. This alternative does notdirectly address the "cheaper, better, quicker"criticism of shuttle-era space science. It should,however, permit experimentation with smallsatellites within the space science program andthe execution of long-planned, large-scale mis-sions.

o A program budgeted at $7 billion annuallythat eliminates piloted spaceflight and insteademphasizes robotic space science and devel-oping new technology for both private indus-try and public missions. This alternative,which would effectively end the current era ofpiloted spaceflight by the United States, ad-dresses the criticism that NASA's activities donot contribute to the competitiveness of U.S.industry. The alternative would concentrateresources in areas that are most likely to pro-duce tangible payoffs—technology developmentdirected toward specific industries and spacescience activities with significant applicationsvalue. For example, funding would be availablefor refurbishing aeronautical facilities, includingnew wind tunnels, which was included in thePresident's request for 1994 but dropped fromthe 1995 plan.

Posing alternatives to NASA's current programand providing estimates of their cost do not solvethe problem of valuing what NASA produces. Theycan, however, illustrate that the balances struck inthe current program-between piloted and unpilotedactivities or between science missions and the de-velopment of new technologies-are neither the onlyoptions nor necessarily the best ones for NASA as itattempts to adjust to the lower budgets that it antici-pates in the future.

Chapter One

The National Aeronautics and SpaceAdministration in Transition

T he Administration's five-year plan for theNational Aeronautics and Space Adminis-tration (NASA) requires that the agencycarry out its program with a nearly flat budget.This prospect marks a significant change for NASA:since the mid-1980s, the agency has planned itsprogram with the expectation of continuous in-creases in funding. NASA has developed a two-pronged strategy to maintain productivity in the faceof its constrained budget outlook. First, the agencyis making marginal adjustments to the content of itsprogram by stretching out, scaling down, and evencanceling some projects. Second, NASA is seekingto improve the way it does business to get more inreturn for the money that it spends. This studyevaluates NASA's strategy and an alternative onethat would radically change the agency's programby emphasizing one or another of the broad objec-tives that NASA has historically pursued.

NASA bears major responsibility for the na-tion's space and aeronautics activities. Its mostvisible efforts are the flights of the piloted spaceshuttle. It also develops, launches, and operatesunpiloted spacecraft whose purpose is to increaseknowledge about the Earth, the solar system, andthe universe. To accomplish those missions, NASAconducts research and develops supporting technolo-gies for piloted and unpiloted missions alike. Theagency also plays a key role in supporting researchand providing facilities to meet the nation's civiland military aviation needs.

In 1994, NASA was permitted over 24,000 full-time-equivalent workyears (figured in time spent byfederal workers) to accomplish its objectives. Per-sonnel were located at the agency's Washington

headquarters and nine major installations, or centers(for example, the Johnson Space Center in Hous-ton). NASA's federal employees bear responsibilityfor all aspects of the agency's activity, but privateindustry executes most of NASA's program. Theagency's procurements from industry typically totalmore than 90 percent of its annual spending.

What Is at Stake?

NASA's total funding for 1993 was $14.3 billion,which constituted less than 1 percent of all federalspending. The agency's budget for research anddevelopment (R&D), however, accounted for 5 per-cent of total national investment in R&D, over 10percent of federal funding for R&D, and almost 25percent of nondefense federal funding for R&D.1

NASA applies these resources to objectives in threemajor areas: the piloted exploration of space, scien-tific research on space-related topics, and the devel-opment of space and aeronautical technologies forcarrying out future public missions in space and forserving the technological needs of private industry.NASA's supporters contend that accomplishment inthese areas improves the nation's self-image, en-hances its international prestige, furthers certainforeign policy objectives, creates new scientificknowledge, quickens the pace of technologicalchange, and contributes to economic productivity in

For total federal and NASA research and development for 1993,see Budget of the United States Government, Fiscal Year 1994,pp. 44-45; for total national R&D spending, see National ScienceFoundation, "U.S. Expenditures on R&D Expected to Increase in1993," SRS Data Brief, September 24, 1993.

2 REINVENTING NASA March 1994

the aerospace industries and perhaps, indirectly, inthe larger economy. Ultimately, it is these contribu-tions that are at stake when decisions are madeabout NASA's program or budget or when ques-tions arise about how well the agency manages itsresources.

The Agency's Objectivesand Their Value

Although the objectives of the NASA program areeasily listed, measuring progress toward those goalsand valuing that progress in dollar terms haveproved to be extremely difficult.2 The lack of ob-jective evaluation leaves substantial room for equal-ly supportable but very different opinions about theappropriate mix of activities in NASA's program orthe potential benefits from increasing the efficiencyof the agency. For example, advocates for contin-ued emphasis on piloted spaceflight place a highvalue on such activities and attribute to them bothpast and potential future successes in U.S. foreignpolicy. Advocates for the agency's space scienceefforts argue that most of the important benefits thatNASA has produced in its 37-year history havebeen generated by the less than 20 percent of itstotal budget devoted to scientific enterprises. Noobjective measure exists to compare these contrast-ing visions of what NASA has accomplished andthe value of its current activities.

This analysis does not solve the problem of howto value NASA's activities. Instead, it emphasizestwo points. The first is that the balance thatNASA's current program strikes among piloted ex-ploration, space science missions, and technologydevelopment is only one of several possible choices.As funds become scarcer, the agency may eitherrearrange its priorities under the current level offunding or focus on a more limited set of objectives(and accomplish them for less than it now spends).

2. A report by the Office of Technology Assessment, FederallyFunded Research: Decisions for a Decade (May 1991), discussesthe general issues of measuring and valuing the output of R&Dagencies. The difficulties in measuring and valuing NASA'soutput are not unique; similar problems exist in measuring thevalue of most federally supported R&D efforts.

The second point is that a scaled-down versionof NASA's current program plan (the first part ofits strategy for dealing with its constrained means)may not be the best use of the agency's resources.Although the benefits of NASA's activities cannotbe quantified, common sense suggests that under-funded or poorly planned projects will not accom-plish the objectives that ultimately produce thebenefits associated with NASA's program. A majorquestion arises about whether the current programcan be rationally downsized and avoid the trap offunctioning as a level-of-effort enterprise-one thatis mired in stretched-out, overbudget projects thatdo not meet their objectives and fail to deliver theirultimate benefits.

NASA and the Economy

How NASA affects the U.S. economy is likely toconsume a large part of any debate about theagency's program. The problems involved in as-sessing the direct benefits that NASA provides haveled some advocates of continued increases in spend-ing for the agency to claim that the indirect influ-ence of NASA's program on the economy is suffi-cient to justify its cost. The more general issue ofwhat the federal government—and taxpayers-arereceiving from their R&D investments has alsofocused attention on the economic consequences ofNASA's spending.

The balance of the evidence does not supporthigher levels of funding for NASA as a means toincrease economywide productivity. In the shortterm, NASA's spending affects the economy in thesame way that other government spending does-andis properly viewed as a cost rather than a benefit ofthe program (see Box 1). Over the longer term,NASA's contribution to the economy does not ap-pear to be large when measured by the most objec-tive standard. Studies that employ other approachesrequiring large measures of judgment by the analystand examine NASA's contribution within particularmarkets can be used to bolster the argument thatpast spending by NASA has led to increased pro-ductivity. But for the most part, economists haverejected the argument that would justify NASA'sprogram based on its contribution to the economy.

CHAPTER ONE THE NATIONAL AERONAUTICS AND SPACE ADMINISTRATION IN TRANSITION 3

Box 1.Short-Term Economic Effects of

Spending for the National Aeronauticsand Space Administration

NASA's spending does not have a uniquelylarge short-term effect on the U.S. economy.All federal spending for goods and servicestends to stimulate the economy temporarily,increasing growth and employment for a shorttime (provided that the economy is not alreadyat full employment). Such spending also tendsto increase inflation and interest rates.

Under certain conditions, NASA's expendi-tures could have a slightly larger or smallershort-run effect on the growth of the economyand on employment compared with federalspending overall. For example, if an unusuallylarge proportion of NASA's spending was di-rected toward industrial sectors or regions of thecountry that were experiencing much higherunemployment than the nation as a whole, theeffects of the spending would be slightly larger,although still temporary. If spending for NASAwas concentrated in industries that had a highvalue added per worker, the effect on employ-ment would be slightly smaller than federalspending overall. On balance, nothing suggeststhat unique aspects of NASA's spending causeit to affect the economy differently from othertypes of federal spending for goods and ser-vices.

Production Function Studies. A production func-tion is a mathematical formulation that relates thevalue of output to the value of inputs. Comparedwith other approaches (for example, cost-benefitanalyses or case studies), production function stud-ies require the analyst to make the fewest assump-tions and subjective judgments.3 The private firm

3. At the economy wide level, many studies have found a strongrelationship between spending for R&D and productivity. Butwhen the contributions to economic growth of private R&D andpublic R&D are evaluated separately, private spending remains astrong positive factor, and public investment in R&D is not corre-lated with growth in productivity. See, for example, FrankLichtenberg, "R&D Investment and International ProductivityDifference," Working Paper 4161 (National Bureau of Economic

Chase Econometrics Associates prepared two suchstudies under contract to NASA; the General Ac-counting Office (GAO) conducted another that eval-uated the first Chase study. All of the studies illus-trate the issues associated with attempting to dis-cover NASA's contribution to private productivity.4

The first Chase study found a substantial contri-bution by NASA to productivity—indeed, one largeenough to explain all of the productivity growth inthe U.S. economy over the 1965-1974 period. (Theequivalent return on NASA's research and develop-ment spending would have been 43 percent.) TheGAO critique showed that small changes in theperiod covered in the estimate or in assumptionsabout capacity utilization or labor quality reducedthe estimate of NASA's contribution to a level in-distinguishable from zero.5 The second Chase re-port confirmed the GAO finding.6

Research, Cambridge, Mass., September 1992), p. 24. The authornotes that from the mid-1960s through the late 1980s there was a"negative, large, and highly significant" relation between govern-ment-funded R&D and output.

4. Michael K. Evans, "The Economic Impact of NASA R&D Spend-ing" (Chase Econometrics Associates, Inc., Washington, D.C.,April 1976); General Accounting Office, NASA Report May Over-state the Economic Benefits of Research and Development Spend-ing (October 1977); and David M. Cross, "The Economic Impactof NASA R&D Spending: An Update" (Chase Econometrics Asso-ciates, Inc., Washington, D.C., March 1980).

5. One of the most frequently quoted estimates of NASA's contribu-tion to economic growth-for every $1 of NASA R&D spending,$9 will be returned to the economy over a roughly 20-year period-relies on a production function approach. (See MidwestResearch Institute, Economic Impact and Technological Progressof NASA Research and Development Expenditures, vol. 2, Eco-nomic Impacts of NASA R&D Expenditures, Kansas City, Mo.,Midwest Research Institute, 1988, pp. II-2 through II-3.) Thestudy by the Midwest Research Institute makes two debatableassumptions that are unsupported by other research: that NASA'sR&D is as productive as the average of all publicly and privatelyfunded R&D, and that NASA's R&D investment falls into thesame category of federal R&D as health and agriculture (for whichpositive returns have been found) rather than into the categorywith "purely military projects" (for which positive results have notbeen found). In contrast to the latter assumption, most analystsargue that NASA's activities are similar to those on the defenseside of government, exhibiting the same mission orientation andrelying on the same contractor base.

6. Another study by the Department of Labor's Bureau of LaborStatistics (Impact of Government and Private R&D Spending onFactor Productivity in Space Manufacturing, July 1980) alsofound no measurable relation between R&D spending by NASAand changes in productivity in industries that are directly affectedby NASA procurement.


The conclusion of such studies-that NASA'sR&D spending has not had a significant effect onnational productivity-is neither surprising nor indic-ative of a waste of resources. The case can bemade that these results fail to capture a positive ef-fect that is actually occurring.7 The benefits of newproducts that come from R&D activities are moredifficult to assess than the reductions in costs per-mitted by innovations in production processes. Tothe extent that all federal R&D and NASA effortsare biased toward innovation in products rather thanprocesses, the contributions from those efforts couldbe understated.

Other Studies. Cost-benefit analyses (many ofthem supported by NASA) have been undertaken todetermine the effect of the agency's programs onconsumers and producers.8 Studies of this typehave characteristically produced large ratios of ben-efits to costs or large estimates of the benefits fromspecific innovations generated by NASA funding.These studies have considered innovations that werespun off from NASA's spaceflight and science pro-grams as well as those supported by its general re-search and technology programs.

Studies of this kind constitute a microeconomicapproach that requires the analyst to make numerousassumptions about where the credit for an innova-tion lies, the period over which benefits should beassessed, conditions of supply and demand in di-rectly affected and related markets, and the possibil-ity that in the absence of the innovation being eval-uated an alternative might have been devised. Con-sequently, cost-benefit studies of past NASA effortsinclude a large measure of subjective judgment.Nevertheless, the studies demonstrate that placingthe contribution of NASA's R&D at zero, a resultthat could be implied by the production function

7. Henry R. Hertzfeld, "Measuring the Returns to Space Researchand Development," in Joel Greenberg and Henry Hertzfeld, eds.,Space Economics (Washington, D.C.: American Institute of Aero-nautics and Astronautics, 1992), pp. 153-155.

8. See, for example, Mathematica, Inc., Quantifying the Benefits tothe National Economy from Secondary Applications of NASATechnology (National Aeronautics and Space Administration,March 1976). The study found that NASA's role in four innova-tions (a computer program used to analyze large structures, inte-grated circuits, insulation for supercooled materials, and the gasturbine engine~an effort begun by NASA's predecessor in the1940s) produced $7 billion (in 1974 dollars) in net benefits.

studies, is probably as mistaken as attributing largeparts of past productivity growth in the nationaleconomy to NASA's program.

Case studies of NASA's role in influencing thedevelopment of three industries—aviation, satellitecommunications, and materials processing in space-also provide data about the economic effects ofNASA's programs. The agency's historical role inthe aviation industry is generally viewed as posi-tive;9 its support of the communications satelliteindustry is usually but not always assessed as a pos-itive contribution.10 NASA's more recent attemptsto create new industries requiring piloted spaceflighthave been failures to date.11

A 1992 study employing a historical approachand a large measure of subjective judgment re-viewed the concept of the "spin-off'—a product orprocess developed by NASA for its purposes thatfinds its way into the larger economy and leads toincreased productivity.12 The spin-off occupies acentral place in the mythology of NASA's relationto the private economy and, accordingly, in the ar-gument that secondary economic benefits might jus-tify spending for NASA. The study's critique ofspin-offs as an organizing principle for technology

9. David Mowery and Nathan Rosenberg, Technology and EconomicGrowth (New York: Cambridge University Press, 1989), pp. 181-184; and George Eberstadt, "Government Support of the LargeCommercial Aircraft Industries of Japan, Europe, and the UnitedStates" (Office of Technology Assessment, May 1991), pp. 63-87.

10. Recent studies of satellite communications and NASA's role increating that industry do not agree fully about the significance ofNASA's activities. Linda Cohen and Roger Noll, "The Applica-tions Technology Satellite Program," in Linda Cohen and RogerNoll, eds., The Technology Pork Barrel (Washington, D.C.:Brookings Institution, 1991), pp. 149-178, offer a generally favor-able appraisal of NASA's role in the private satellite communica-tions industry, citing the agency's development of satellite tech-nology and its role in providing launch services. A contrastingview is offered in Peter Cunniffe, "Misreading History: Govern-ment Intervention in the Development of Commercial Communi-cations Satellites," Report 24 (Program in Science and Technologyfor International Security, Massachusetts Institute of Technology,Cambridge, Mass., May 1991). The author accepts the importanceof NASA's role in providing launch services but finds that mostof the significant technical innovations in the industry were pri-vately financed and developed.

11. See Chapter 4 in Congressional Budget Office, Encouraging Pri-vate Investment in Space Activities (February 1991).

12. John Alic and others, Beyond Spinoff: Military and CommercialTechnologies in a Changing World (Boston: Harvard BusinessSchool Press, 1992).


policy noted that the technology and institutionalarrangements necessary for success in the missionsof federal agencies were diverging from the charac-teristics necessary for success in private markets. Itconcluded that "grandiose projects patterned on theApollo moon landing or the Strategic Defense Ini-tiative will be increasingly irrelevant to world tech-nological competition."13 The analysis suggests that,even if spin-offs from NASA's program were im-portant in the past, they are unlikely to be as impor-tant in the future. The production systems thatNASA requires and those that serve the private mar-ket follow different paths.

The Structure of NASA'sProgram and Budget

In 1994, NASA received $14.6 billion in budgetauthority, which it allocated as shown in Table 1.Funding for the space shuttle system, including bothoperations and continuing investment, was the larg-est single item at $3.8 billion-over 25 percent ofNASA's total appropriation. Funding for the spacestation at $1.9 billion was the next largest singleitem and accounted for 13 percent of the agency'sbudget. The space science and applications pro-gram, which supports the robotic spacecraft thatNASA uses to gather information about the Earthand space, received $3.3 billion in 1994. Thatfunding was concentrated in three areas: physicsand astronomy ($1.1 billion), planetary exploration($650 million), and Earth science ($1.1 billion).NASA allocated $1.4 billion to efforts to advanceaeronautics and space technology, with more than70 percent of the total going to aeronautics. Thebulk of NASA's remaining funds were dividedamong the accounts that pay federal employees andthat support the construction of facilities.

Funding Trends During the 1980s

NASA's current funding is about twice the $7.2billion granted the agency in 1984 but only a small

Table 1.Initial Operating Plan for the NationalAeronautics and Space Administration, 1994(In millions of dollars of budget authority)

Category of Spending Amount

Research and DevelopmentSpace stationSpace transportation capabilitySpace science and applications

Physics and astronomyPlanetary explorationLife sciencesMicrogravityEarth science (Mission to Planet Earth)Other8

Subtotal

Advanced concepts and technologyb

Aeronautical research and technologyTransatmospheric research and technologySafety, reliability, and quality assuranceAcademic programsTracking and data advanced systems

Total

Space Flight, Control, and Data CommunicationsShuttle production and operations capabilityShuttle operationsSpace and ground tracking systemsLaunch services

Total

Construction of FacilitiesResearch and Program ManagementInspector General

Total, All Categories

1,946663

1,068654188177

1,068162

3,307

4331,007

20348624

7,529

1,0352,744761313

4,853

5181,636

15

14,551

13. Ibid., pp. 12-13.

SOURCE: Congressional Budget Office based on data fromNational Aeronautics and Space Administration,"Initial Operating Plan for 1994" (1993).

NOTE: Numbers may not add to totals because of rounding.

a. Includes some spacelab costs for life sciences and micro-gravity experiments and $50 million for U.S.-Russian coopera-tive activities.

b. Formerly Space Research and Technology, and CommercialPrograms.


increase compared with NASA's 1991 funding levelof $13.9 billion. Between 1984 and 1991, theNASA budget increased at an average rate of 9.7percent annually. Between 1991 and 1994, annualaverage growth in the agency's budget fell to 1.5percent.

Traditionally, NASA presents its budget to theCongress as the sum of four major appropriations(see Table 2). (NASA's 1995 budget request in-cludes a change in its appropriations from the tradi-tional accounts shown in Table 2 to a new schemediscussed in Box 2.) The Research and Develop-ment category supports development of aeronauticsand space technology, and development and opera-tion of both piloted and robotic spacecraft-with themajor exception of the space shuttle. Spending forthe Research and Development component morethan tripled over the past decade, increasing itsshare of the agency's budget from about 30 percentin 1984 to around 50 percent in 1994.

The increase in NASA's research and develop-ment spending was driven by the piloted space sta-tion program and large-scale robotic space scienceprojects. In 1984, spending for the space stationwas less than $100 million spread throughout theagency. By 1994, annual funding had reached $1.9billion. Large robotic space science missions alsocontributed to the growth in NASA's research anddevelopment spending. In the mid-1980s, spendingfor physics and astronomy projects (the HubbleSpace Telescope and the Compton Gamma Ray Ob-servatory) and spacecraft for planetary exploration(the Galileo probe to Jupiter, the Venus Radar Map-per, and the Mars Observer) accounted for the in-crease. By the late 1980s and early 1990s, thoseprojects were on the downward sloping tail of theirbudgetary lives, but spending was on the rise in theEarth science area, primarily for the Earth Observa-tion System (EOS). Under the Bush Administra-tion's 1993 budget plan, the space station and the

Table 2.Budget of the National Aeronautics and Space Administration, 1984-1994(In millions of dollars of budget authority)

Category 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994

Research andDevelopment

Space Flight,Control, and DataCommunications

Construction ofFacilities

Research andProgram Management

Inspector General

Total

2,064 2,468 2,619 3,154 3,255 4,238 5,228 6,024 6,827 7,089 7,529

3,772 3,594 3,666 6,000 3,806 4,452 4,625 5,124 5,385 5,086 4,854

156 158 138 169 178 282 411 498 531 525 518

1,256 1,332 1,341 1,452 1,762 1,926 2,023 2,212 1,576 1,615 1,636

n.a. n.a. n.a. n.a. n.a. n.a. 9 H 14 1j> 15

7,248 7,552 7,764 10,775 9,001 10,898 12,296 13,869 14,333 14,330 14,551

SOURCE: Congressional Budget Office based on National Aeronautics and Space Administration, Budget Estimates (1984-1994).

NOTE: Numbers may not add to totals because of rounding, n.a. = not applicable.


Box 2.Changes to the National Aeronautics and Space Administration's

Appropriations Proposed in the 1995 Budget

The National Aeronautics and Space Administration's(NASA's) budget request for 1995 includes a proposalto redefine NASA's appropriations. Instead of thecurrent division into five appropriations (Research andDevelopment; Space Flight, Control, and Data Commu-nications; Research and Program Management; Con-struction of Facilities; and the Inspector General),NASA would divide its budget into four categories:Human Space Flight; Science, Aeronautics, and Tech-nology; Mission Support; and the Inspector General.(See the table below, which provides a "bridge" be-tween the current and proposed categories.)

The appropriation structure proposed for 1995highlights the division of NASA's program betweenpiloted and unpiloted activities in a way that the currentstructure does not. For 1994, the Research and Devel-opment appropriation stood at $7.5 billion, including$1.95 billion for the piloted space station program.The Space Flight, Control, and Data Communicationsaccount was funded at $4.9 billion, of which $3.8 bil-lion was allocated to the piloted space shuttle. Underthe classifications proposed for 1995, NASA wouldcombine spending for development of the space stationwith funds for the operation and continued developmentof the space shuttle. The combined spending wouldconstitute the Human Space Flight appropriation, whichin 1994 would have totaled slightly more than $6 bil-

lion (with the addition of $300 million for other pilotedspaceflight activities). The proposed Science, Aeronau-tics, and Technology appropriation-which would havebeen $5.8 billion in 1994, had the new categories beenin effect-is essentially the sum of funding for currentrobotic space science, aeronautics, and technologyprograms, or the current Research and Developmentappropriation minus the funding to develop the spacestation.

The proposed Mission Support appropriation wouldinclude all of the current Research and Program Man-agement appropriation (more than $1.6 billion in 1994for NASA's federal employees). The category wouldalso include funds currently appropriated under theConstruction of Facilities accounts and funding forNASA's ground and space tracking system, which isnow divided between the Research and Developmentappropriation and the Space Flight, Control, and DataCommunications appropriation.

The budgetary presentations and analysis in thisstudy use the current appropriation categories but trans-late well into the proposed new structure. This is par-ticularly true of the discussion in Chapter 4 of alterna-tives to the current NASA program that either dramati-cally increase or decrease the share of NASA's budgetdevoted to piloted spaceflight.

Current and Proposed Appropriation Structure for the 1994 Fundingfor the National Aeronautics and Space Administration (In millions of dollars)

Proposed Categories

Current CategoriesHuman Space

Flight

Science,Aeronautics, and

TechnologyMissionSupport Total

Research and Development

Space Flight, Control,and Data Communications

2,435

3,601

4,725

860

369

392

7,529

4,853

Research and ProgramManagement

Construction of Facilities

Total

n.a.33

6,070

1,636222

2,619

SOURCE: Congressional Budget Office based on Budget of the United States Government, Fiscal Year 1995, Appendix, p. 821.NOTES: The Inspector General category is not shown because it remains the same under both structures.

Numbers may not add to totals because of rounding.n.a. = not applicable.


EOS were the two largest civilian R&D projects inthe federal budget.14

Funding for the space shuttle, traditional un-piloted rockets, and the orbital communications andtracking network is included in the Space Flight,Control, and Data Communications category ofaccounts. This aggregate has grown more slowlythan spending for research and development, asshown by its decline from about 50 percent of theagency's spending in 1984 to around 35 percent in1994. The spike in funding for spaceflight in 1987represented the purchase of a replacement orbiterfor the Challenger lost in 1986. The other conse-quences of the Challenger accident for NASA'sprogram do not show up as clearly in budget data.In 1984, the agency received about $3.1 billion tosupport between 12 and 14 shuttle flights annually.Ten years later, in 1994, the shuttle system wasfunded at a modestly higher level of $3.8 billion butplanned to support only eight flights a year.

The last two major categories of NASA fundingare Construction of Facilities and Research and Pro-gram Management. The former has more than tri-pled in size over the past 10 years, but it still ac-counted for only 4 percent of NASA's spending in1994. The latter increased from $1.3 billion in1984 to more than $2.2 billion in 1991 but then fellto $1.6 billion in 1993. (The drop was a conse-quence of NASA's redefining its accounts to shiftabout $400 million in funding for maintainingNASA centers and installations from the Researchand Program Management accounts to the Researchand Development and Space Flight, Control, andData Communications accounts.)

Several other trends can be identified. The pro-portion of NASA's budget devoted to piloted space-flight has remained constant at about 50 percentthroughout the decade, as measured by spending forthe space station and the space shuttle, developmentof space transportation capability, and associatedspace science projects (see Figure 1). The share of

14. Congressional Budget Office, "Large Nondefense R&D Projects inthe Budget: An Update," CBO Staff Memorandum (March 1992),p. 2. For a general approach to measuring the budgetary effectsof large projects, see Chapter 2 in Congressional Budget Office,"Large Nondefense R&D Projects in the Budget: 1980-1996,"CBO Paper (July 1991).

Figure 1.Budget Shares for Piloted Spaceflight andSpace Science, National Aeronautics and SpaceAdministration, 1984-1993 (In percent)

60

50

40

30

20

10

Percent

Piloted Spaceflight

Space Science

1984 1985 1986 1987 1988 1989 1990 1991 1992 1993

SOURCE: Congressional Budget Office based on data fromthe National Aeronautics and Space Administration.

NOTE: Piloted spaceflight includes spending for the spacetransportation system and the space station only.

space science and applications, a subset of the Re-search and Development category that is dominatedby the development and operation of robotic space-craft, increased from 16 percent of NASA's budgetin 1984 to 20 percent in 1994. If life sciences andmicrogravity research-activities associated withpiloted spaceflight-were excluded, however, thegain in the share of space science would be moremoderate. Finally, spending for aeronautics, asmeasured by program expenditures, has roughlytripled over the past 10 years, and the share of theseactivities has increased from 4 percent to 7 percentof NASA's budget.

The Breaking Bow Wave

What is not apparent in reviewing NASA's fundingover the past 10 years is the expectation of futuregrowth that permeated the agency's planning. Since


Figure 2.Five-Year Budget Requests of the National Aeronautics and Space Administration, 1990-1995(In billions of dollars of budget authority)

22

20

18

16

14

12

10

Billions of Dollars of Budget Authority

1992Request

1991Request

\'* '• *

' ,'* *

1993Request

1990Request

1994Request

\ ,X"1995

Request

Actual

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999

SOURCE: Congressional Budget Office based on Budget of the United States Government (various years) and 1993 projections fromthe NASA Comptroller's Office.

the mid-1980s, NASA's program has required in-creases in its annual budget above the rate of infla-tion (see Figure 2).15 The force driving NASA'sfuture budget requirements upward in the late 1980swas the anticipated cost of first developing and thenoperating new spacecraft that were to be integratedin a low-Earth-orbit infrastructure. NASA envi-sioned the infrastructure as including the space shut-tle, the space station, tracking and data relay satel-lites, and several large satellites carrying instrumentsthat looked outward to the stars or back at theEarth. Once in place, this investment was to pro-duce near-term benefits measured in scientific ad-vances, new technologies, and contributions to eco-nomic growth. Over the long term, the low-Earth-

orbit infrastructure was seen as a stepping-stone forpursuing NASA's long-held goals of a Moon baseand a piloted mission to Mars.16

The plans for NASA's program outlined in itsbudget requests for 1990 through 1993 continued toshow increasing funding requirements, despite thecompletion of major parts of the infrastructure. In1992, GAO testified that NASA's program plan for

15. Congressional Budget Office, The NASA Program in the 1990sand Beyond (May 1988), pp. xi-xiv.

16. See Chapter 2 in Howard McCurdy, The Space Station Decision(Baltimore, Md.: Johns Hopkins University Press, 1990).


Table 3.Budget Requests and Appropriationsfor the National Aeronautics andSpace Administration, 1989-1994(In billions of dollars of budget authority)

1989

1990

1991

1992

1993

1994

Request

11.5

13.3

15.1

15.8

15.0

15.3

Appro-priation

10.9

12.3

13.9

14.3

14.3

14.6

Differ-ence

0.6

1.0

1.2

1.5

0.7

0.7

SOURCE: Congressional Budget Office based on NationalAeronautics and Space Administration, Budget Esti-mates (1989-1994).

crating current projects remained high and the costof projects in development continued to increase.

NASA's program plan has attracted the attentionof critics in part because of the recent focus on thenation's budget deficit. Concerns about the cost ofthe NASA program grew after 1990 and the tighten-ing of all domestic discretionary spending requiredby the Budget Enforcement Act. When the caps inthe act began to restrain spending, the Congress sig-nificantly lowered NASA's budget from theamounts requested by the President in 1992 and1993 (see Table 3). For 1994, the Congress againappropriated less than the President requested, de-spite the Administration's proposal to slow thegrowth in NASA's five-year program plan by $16billion compared with the plan included in PresidentBush's last budget (see Table 4).

Further reductions could be in the offing. TheOmnibus Budget Reconciliation Act of 1993 con-tains a series of caps on appropriations for the nextfive years that will essentially freeze all discretion-ary spending at 1993 levels. The caps have led theAdministration to scale back NASA's budget evenfurther.

1992 through 1997 would require almost $13 billionabove the Congressional Budget Office baseline forthe agency.17 This finding was consistent with theconclusion reached two years earlier by a federaladvisory committee convened by President Bush.The Advisory Committee on the Future of the U.S.Space Program, better known as the AugustineCommittee, found that NASA was over-committedin terms of the scope of its program and would re-quire annual increases of 10 percent above the rateof inflation to realize all of its objectives.18 As theClinton presidency began, NASA's budget stillrequired strong growth because the cost of op-

17. Testimony of Neal P. Curtain, Director of Planning and Reporting,National Security and International Affairs Division, General Ac-counting Office, before the Subcommittee on Science, Technology,and Space, Senate Committee on Commerce, Science, and Trans-portation, March 17, 1992, pp. 1-3.

18. National Aeronautics and Space Administration, Report of theAdvisory Committee on the Future of the U.S. Space Program(December 1990), pp. 1-9.

Future budget requirements are easier to scaledown than the programs that underlie them. Theexpectations represented by NASA's plans in the1980s may be equally difficult to deflate.

Table 4.Five-Year Budgets in the 1993 and 1994Plans of the National Aeronautics andSpace Administration (In billions of dollarsof budget authority)

Plan 1994 1995 1996 1997 1998

1993

1994

Difference

17

15

1,

.0

.3

.7

18.6

15.7

2.9

19

16

3.

.5

.1

4

20

16

3,

.3

.5

.8

21.0

16.8

4.2

SOURCE: Congressional Budget Office based on data from theNational Aeronautics and Space Administration.

Chapter Two

The Risks of a Strategyof Marginal Adjustment

T he National Aeronautics and Space Admin-istration has chosen to make marginal ad-justments to its basic program as one wayof coping with the prospect of flat budgets in theforeseeable future. The agency has scaled backindividual projects, stretched them out, and in somecases even canceled them, but it remains committedto a program structure that includes the developmentand operation of major piloted systems (the spacestation and the shuttle), the development and opera-tion of major robotic space science missions (forexample, the Earth Observation System and theHubble Space Telescope), and support for newaeronautical and space technology.

Although NASA has adjusted its program ineach of the past several years, when the Congressprovided less funding than the Administration hadrequested, adjusting the agency's program to flatout-year budgets (those for the four years beyondthe fiscal year of the budget request) is a moredifficult and riskier exercise.

The strategy of marginal adjustment, and itscomplement of improving the way that the agencydoes business, may be successful in accomplishingNASA's planned missions and delivering their ulti-mate benefits. But success is not assured. Theattempt to fit a program that was projected to costmore than $20 billion a year in the late 1990s intoan annual budget of $14 billion risks delay, missionfailure, and the loss of anticipated benefits. Essen-tial characteristics of NASA's program increase therisk of failure associated with a strategy of marginaladjustments. Moreover, that type of strategy mayexacerbate perceived problems with the currentprogram. This chapter explores both of those fac-

tors in assessing the risks of NASA's plan for mar-ginally adjusting its program.

Program Characteristics

Three characteristics of NASA's current program in-crease the risks associated with marginally adjustingit to fit the smaller future budgets that NASA ex-pects. First, many of NASA's programs have highfixed costs. Second, the agency must allocate sub-stantial funding for mission operations and dataanalysis late in a project's life cycle to realize areturn on the sizable investment it has already madein the spacecraft. In addition, successful small-scaleprojects have led to the spread of larger-scale effortswith high fixed costs and long-term and substantialoperating funding to many areas of the agency'sprogram. Third, NASA has consistently underesti-mated the costs of its projects.

High Fixed Costs

Economists sometimes characterize high fixed costsas "lumpiness." A lumpy expenditure or cost isessentially an all-or-nothing proposition. The goodor service desired cannot be purchased in smallerquantities, even if the buyer has no use for the fullquantity. This characteristic applies to large parts ofNASA's program and budget and limits the areas inwhich adjustments can be made to bring projectsinto line with a no-growth budget. The space shut-tle and space station programs-the mainstays ofNASA's activities in piloted spaceflight-and large


space science missions are all examples of projectswith high fixed costs.

The shuttle system best illustrates the concept oflumpiness and the problems it presents to a strategyof marginal adjustments. A recent analysis by theGeneral Accounting Office estimated that the totaloperating cost of flying eight shuttle missions in1993 was about $3.3 billion, or 23 percent ofNASA's 1993 funding.1 Adjusting NASA's pro-gram by scaling back the rate of shuttle flights bytwo missions (25 percent of the planned annual rate)would save less than $100 million a year accordingto NASA, or less than 3 percent of the shuttle'stotal annual operating costs. Once the agency de-cided to engage in piloted spaceflight and to adoptthe shuttle as its primary flight system, it accepted a"lump" or fixed cost of between $3 billion and $4billion in its annual budget.2 Marginally adjustingthe rate of shuttle flights will not generate signifi-cant budgetary savings.3

A less strict example of high fixed costs in-volves the development phases of the space stationprogram and large space science missions. Accord-ing to NASA, the space station requires a minimumof $2 billion annually to make progress toward actu-ally developing and launching the facility. Thefixed cost of maintaining project teams-funding forpersonnel and overhead within NASA and amongits contractors-is a substantial part of this annual

1. General Accounting Office, Space Transportation: The Contentand Uses of Shuttle Cost Estimates (January 1993), p. 4. Becausethe cost of a shuttle flight is incurred over three fiscal years, thecost of flights flown in 1993 does not equal the 1993 appropria-tions for shuttle operations. However, when the annual flight rateof the shuttle is roughly constant, as is currently the case, theannual appropriations for the shuttle and the annual cost of theshuttle system are roughly equivalent.

2. In addition to expenditures to operate the shuttle, NASA spendsaround $1 billion annually on improvements to the system. Thisspending is more amenable to budgetary reductions that result incomparable reductions in program activities. Most recent de-creases in projected funding for the space shuttle are attributableto the canceling of planned improvements—for example, the Ad-vanced Solid Rocket Motor, the extended durations orbiter kit(proposed), and spare parts for the shuttle orbiters.

3. National Aeronautics and Space Administration, Budget Estimates,Fiscal Year 7993, p. SF 2-3, projected the costs of future shuttleoperations by reducing the fixed costs of operations by 3 percentover each of the following five years. The agency has beenlargely successful in achieving this goal, although further reduc-tions are likely to be more difficult.

expenditure. As a result, with the current programplan and hardware design, it would be unproductiveto attempt a program funded at $1 billion annuallybecause that level would be insufficient to supportthe fixed cost of the program, let alone make prog-ress in actually building a space station.

The case is much the same for major spacescience missions in development—for example, thedual-spacecraft Comet Rendezvous Asteroid Flyby(CRAF) and Cassini missions approved for devel-opment in 1989. A 1993 GAO report examined thecancellation of the CRAF spacecraft. After spend-ing $700 million of a projected $3.7 billion totalcost for both spacecraft, NASA canceled the CRAFproject in 1994. But it reduced its total projectcosts by only $700 million, $535 million of whichwas accounted for by reduced costs for the launchand operations.4 The saving in development costswas only about $165 million because the fixed costof developing the spacecraft to be used for bothmissions was relatively high.

Although not all of NASA's activities can becharacterized as high-fixed-cost projects, many haveaspects that would permit such a characterization,particularly projects in which the agency seeks tomaximize productivity by completing them withinschedules that allow development at close to mini-mum cost. The implication of high fixed costs inthe context of NASA's current budgetary realities isthat large parts of NASA's program are not candi-dates for marginal reductions. However, otherparts-largely in the operation and actual use of thehardware and systems developed over the past de-cade—could be disproportionately cut under a strat-egy of marginal adjustment.

Mission and Operations FundingLate in a Project's Life Cycle

The life-cycle characteristics of a typical NASAproject could also limit the effectiveness of thestrategy of marginal adjustments. Specifically, theywould inflict a high price in the form of lost bene-

4. General Accounting Office, Space Science: Causes and Impacts ofCutbacks to NASA's Outer Solar System Exploration Missions(December 1993), p. 20.

CHAPTER TWO THE RISKS OF A STRATEGY OF MARGINAL ADJUSTMENT 13

fits as a result of reducing funding in the operationsphase of a project. In some senses, this outcome isanother manifestation of high fixed costs becausethe additional cost of operating a spacecraft is smallcompared with the cost of developing, producing,and launching it.

The typical NASA project incurs large annualcosts early in its project life but delivers most of itsbenefits later during the operating and data analysisphase, a period of relatively smaller annual costs.Adjusting NASA's program to fit within smallerfuture budgets by reducing spending for missionoperations and data analysis could significantly de-crease the benefits of past investments. As a con-crete example, it would be difficult to produce areturn on the nation's past investment in the HubbleSpace Telescope if its current operations were notfunded.

In the space science and applications area, fund-ing for mission operations and data analysis for thephysics and astronomy, planetary exploration, andEarth science programs totaled $728 million in1993, or 25 percent of the $2.9 billion allocated tothe area.5 The best examples of long-term opera-tional costs are found in the physics and astronomyprogram. The Hubble Space Telescope cost $1.7billion to develop. To reap the full benefit fromthis past expenditure, spending in excess of $200million annually for servicing, operations, and dataanalysis was necessary in 1992 through 1994. Fiveother astrophysics missions now in the operationalphase, the most prominent being the ComptonGamma Ray Observatory, required a total of $85million over each of the past three years for opera-tions and data analysis.6

If NASA continues on its present course, thespace station and Earth Observation System willalso require annual operating support to secure thebenefits of the nation's current investment. NASAestimated in September 1993 that the space stationwould require a minimum of $1.5 billion annually.According to data that NASA furnished to the

Office of Technology Assessment, also in 1993, theEOS could require as much as $500 million annu-ally for operations, data analysis, and management.

Progress in achieving space-related goals has ledto larger-scale projects in a widening array of scien-tific disciplines and subdisciplines. As with theproblem of high fixed costs, the demand for ongo-ing operational expenditures across a number ofprogram areas increases the risks of a strategy ofmarginal adjustment.

Cost Overruns

The problems that NASA has experienced in esti-mating the cost of its projects are not inherent to itsmission. Unlike high fixed costs and postdevelop-ment operating expenditures, underestimated costsare not a necessary condition of the NASA enter-prise, although they have been a pervasive charac-teristic. As the agency's decisionmakers strive tobring the cost of its program down to a level thatcan be productively supported by a flat budget, costoverruns represent a significant risk to their success.

NASA's problems in estimating costs, althoughneither unique nor limited to this period of theagency's history, are nevertheless quite serious. A1992 study by the Institute for Defense Analysis(IDA) examined NASA's record and concluded thatthe agency had enjoyed considerable technical suc-cess but that its record in meeting schedules andgoals related to costs was "considerably worse eventhan the DoD's experience."7 A recent GAO studyfound that 25 of the 29 projects with initial costestimates above $200 million that NASA startedbetween 1977 and 1991 cost more than originallyestimated.8 The range of overruns stretched from14 percent to over 400 percent, with a median ofabout 75 percent. Of the four projects that did notexperience overruns, two were significantly reducedin scope from their original conceptualization.

5. National Aeronautics and Space Administration, Budget Estimates,Fiscal Year 1994, pp. RD 3-1, RD 4-1, and RD 6-1.

6. Ibid., p. RD 3-21.

7. Karen W. Tyson, J. Richard Nelson, and Daniel M. Utech, APerspective on Acquisition of NASA Space Systems (Alexandria,Va.: Institute for Defense Analysis, December 1992), p. 79.

8. General Accounting Office, NASA Program Costs: Space MissionsRequire Substantially More Funding than Initially Estimated (De-cember 1992), pp. 1-4.


The GAO study reported both general and spe-cific reasons for the overruns. Insufficient studiesto define the projects, instability in the programsand their funding, overoptimism on the part of pro-gram officials, and unrealistic estimates by contrac-tors were noted as general causes. Specific factorsincluded program redesigns, technical complexities,incomplete cost estimates, shuttle launch delays, andunanticipated inflation. The IDA study noted mostof these factors as well and stressed the roles ofunderestimating the technical difficulty of projectsand inadequate planning.

The budgetary appetite of the whole NASAprogram and the agency's failure to estimate thecosts of its projects accurately have reinforced oneanother. In the past, the out-year projections of theagency's overall program that showed growingbudgetary requirements in future years must cer-tainly have understated the actual cost of completingthe program as planned. (Those estimates, after all,were no more than the sum of costs for individualprojects that the GAO and IDA studies showedwere consistently underestimated by NASA.) At thesame time, the requirement for growth in theagency's total budget placed pressure on programmanagers and contractors to be overly optimistic inmaking cost estimates, a factor identified by GAOas contributing to NASA's consistent underestima-tions.

The strategy of marginal adjustment will notchange this relation. Indeed, margins for error maydecrease and even disappear. Reducing NASA'stotal budget could intensify pressure to underesti-mate the costs of individual programs. In a generalclimate of cost cutting, project managers and theircontractors might be tempted to accept overly opti-mistic cost estimates, realizing that projects thatcould claim cost reductions but preserve their poten-tial output would fare better in a demanding fiscalenvironment. If help for a project experiencing anoverrun is sought by restricting other projects oper-ating on slimmer-than-usual margins, problems inone project or area can be transmitted to other proj-ects or areas. For the program as a whole, theprospect of cost overruns distorts the choices madein marginally adjusting the content of NASA'sprogram and increases the risk that the benefits of

NASA's activities will be diminished, deferred, orlost entirely.

Marginal Adjustmentand Problems with theCurrent Program

Even when NASA's out-year budgets arched up-ward, critics questioned the agency's priorities andspeculated that its program was not deliveringenough benefits to justify its cost. Marginally ad-justing the content of the current program is likelyto intensify concerns about the agency's prioritiesand the value of the benefits it provides.

Winning the race to the Moon in 1969 isviewed by some observers as the peak of theagency's accomplishments and usefulness. Sincethe Apollo era, the political system has been unwill-ing to fund fully the agency's overriding objective--piloted exploration of the solar system. The end ofthe Cold War eliminated any lingering reason tosupport NASA's emphasis on piloted spaceflight asa demonstration of the superiority of democraticcapitalism over totalitarian communism.9 Althoughthe Administration's initiative to include Russia inthe international space station has resurrected for-eign policy as a primary reason for piloted space-flight, other justifications for the agency's programhave grown in importance. These include NASA'scontributions to the advancement of science, to theunderstanding and monitoring of the global environ-ment, and to the activities of aerospace industries.

Specific dissatisfactions with the content andpotential benefits of NASA's program are many, butthree are of particular importance because they arelikely to be aggravated as NASA adjusts its pro-gram. First is the question of people in space.Enthusiasts of piloted spaceflight and human explo-ration are unhappy with the slow pace of NASA's

9. Vice President's Space Advisory Board, A Post Cold War Assess-ment of U.S. Space Policy (December 1992), pp. 1-11.

CHAPTER TWO THE RISKS OF A STRATEGY OF MARGINAL ADJUSTMENT 15

activities in this area. Other critics contend that themore than 50 percent of NASA's budget spent onpiloted spaceflight is too large a share.10

Second is the criticism that NASA's space sci-ence program is too focused on large-scale, expen-sive projects with long operational and budgetarylives. For example, the Hubble Space Telescopecost billions of dollars to build and operate and isexpected to enjoy a project life span of at least 20years, from the beginning of development to the endof operations.11 Critics argue that such projects ex-tract too large a cost when they fail and are overlysubject to bureaucratic inefficiencies.

Third, and finally, the content of NASA's pro-gram has been criticized as unresponsive to the eco-nomic challenges facing the nation. From this pointof view, NASA should place more emphasis on ac-tivities to increase private productivity-for example,research and development supporting U.S. aircraft,rocket, and satellite manufacturers.

by crowding out more worthy science projects.From this point of view, reducing federal spending,investing in other space science projects, or support-ing other scientific enterprises that do not involvespace are likely to produce a higher level of bene-fits.

For many years, rising budgets and the expecta-tion of future increases in funding muted the con-flict between advocates and opponents of pilotedspaceflight within the community of interests thatgenerally supported spending for space. For exam-ple, the Augustine Committee assigned its highestpriority to NASA's largely unpiloted scientific ac-tivities. But the assumption that NASA's budgetwould continue to grow by 10 percent a year afterinflation allowed the committee to downplay thefriction between NASA's budgetary tilt toward pi-loted spaceflight, the desire for even more spendingto support the future exploration of Mars byhumans, and the committee's own observation thatthe scientific benefits of piloted spaceflight werelimited.13

People in Space

Putting people in space is costly, a point on whichmost critics agree, whatever their position. Advo-cates of spending more on piloted spaceflight viewits benefits as sufficient justification for those highcosts.12 Moreover, they argue that investing in newtechnology that reduces the cost of having people inspace will drive the benefit-cost ratio even higher.Opponents of piloted activities counter that suchprograms do not produce sufficient benefits to just-ify their high costs and that in a constrained budgetenvironment they impose an unacceptably high cost

The progressive constriction of NASA's five-year budget outlook and the strategy of marginaladjustment have now brought more of the tensionsabout piloted spaceflight to the surface. PresidentBush's proposal to commit the United States to ahuman outpost on the Moon and a piloted missionto Mars by early in the next century was rejected bythe Congress, largely because of its expected cost.The Clinton Administration's initiative to lower thecost of the space station and at the same time re-duce the growth in NASA's total budget has leftboth advocates and opponents of piloted spaceflightonly partially satisfied. Further downward pressure

10. The Congressional Budget Office's (CBO's) estimate of the shareof piloted spaceflight is based on National Aeronautics and SpaceAdministration, Budget Estimates, Fiscal Year 1994, p. AS-8.CBO's estimate includes 1993 funding for the space station, spaceshuttle, and life sciences and microgravity projects, and a proratedshare of research and program management.

11. Development of the Hubble Space Telescope began in 1978 withfunding of $36 million; see General Accounting Office, Status ofthe Hubble Space Telescope Program (May 1988), p. 18. In De-cember 1993, NASA serviced the telescope, which allowed itsorbital life to be extended until at least 1996. An additional nineyears of operation are possible, according to NASA's plans, ifservicing missions are undertaken every three years. Thus, frombirth to death, the Hubble could "live" 24 years.

12. General Accounting Office, Space Projects: Astrophysics FacilityProgram Contains Costs and Technical Risks (January 1994),sheds light on the benefits of piloted spaceflight in space scienceenterprises. The success of the repair mission on the HubbleSpace Telescope demonstrated that the risk of failure for a spacescience mission could be decreased by developing systems thatcould be repaired by astronauts. However, the redesign of theAdvanced X-Ray Astrophysics Facility (AXAF) described in theGAO report illustrates the costs of lowering risk by providing forrepair by astronauts. GAO found that without a link to pilotedspaceflight, the AXAF would cost less than half as much anddeliver roughly the same scientific contribution as the alternativethat provided for repair by astronauts.

13. National Aeronautics and Space Administration, Report of theAdvisory Committee on the Future of the U.S. Space Program(December 1990), pp. 5-8.


on NASA's budget will increase the tension overthe content of NASA's program as the Administra-tion and the Congress confront choices between ma-jor piloted programs (for example, the space station)and unpiloted efforts (for example, the Earth Obser-vation System).

Too Big, Too Expensive,and Too Long?

The thrust of NASA's pr

Documents

CONGRESS OF THE UNITED STATES CONGRESSIONAL ......NASA's Program and Budget Since the mid-1980s, NASA's program has required consistent growth in its out-year budgets, even after adjusting