Managing the Space Sciences (1995)

In addressing the issue of alternative organizations for the management of science in NASA, the committee's task group on alternative organizations focused on organizational structures that would produce the best possible quality of science given the present and likely future budget constraints. It identified a number of fundamental principles, or axioms, and goals, that characterize successful science management.

1. Accountability for Performance—Science management must be accountable to higher management for achievement of programmatic or mission goals. Such a goal in the case of exploratory basic research may be a reasoned effort to understand nature. For research in support of a program or project, the timely (i.e., schedule-driven) completion of specific measurements or calculations may be required. Accountability must be accompanied by authority to make programmatic and budgetary decisions in pursuit of the goals.

2. Separation of Program Approval from Program Execution—The responsibility for program approval, including especially determination of whether the research is to be conducted in-house or out-of-house, should not be delegated to performing organizations, which might bias that selection toward in-house sources. If source selection must be so delegated for other reasons, controls to counter any such bias must be put in place.

3. Involvement of Scientists in Management—Scientific expertise is required for the effective management of scientific projects whether conducted in-house or procured under a contract arrangement (the “smart buyer” concept). Effective coordination of scientific goals with technical and engineering requirements is one of the prerequisites for the successful completion of technically complex scientific projects. Scientific expertise is needed to monitor the progress of projects, to evaluate their adherence to program goals, and to participate in trade studies necessitated by conflicts in budget, schedule, and science requirements.

4. Maintaining Effective Relationships—Successful science management requires effective relationships among science partners, performers, and eventual users of the scientific output.

5. . Single Line of Authority—All direction, both program/project direction and ancillary functional direction, should come down to performing organizations from higher management levels through a single management chain. Decision making must be consistent with budget allocation, lest conflicts arise.

1. Quality of Science Output—A fundamental principle of successful science management must be that high-quality science be produced. Factors that contribute to high quality include informed and unbiased selection of the best performers and openness of the program to involvement by the scientific community in program planning and execution. As a corollary to achieving high-quality science, systems to evaluate that quality must be in place and must be used to constructively influence program activities. Evaluation systems include peer review for selection; visiting committees of scientific authorities to evaluate programs, projects, and institutions; performance reviews of individual performers; and such statistics as number of refereed publications.

2. Cost-effectiveness—Not only must the absolute quality of the science be high, but quality must be high in relation to program cost. High quality is achieved by the factors cited above. Low cost follows from reducing infrastructure costs (both workforce and facilities) to the minimum and eliminating excessive oversight of performers. Careful planning and management on the part of both managers and investigators are crucial to achieving high cost-effectiveness.

3. Responsibilities of a National Capability in Space Science—Science programs of an agency like NASA do not exist in isolation but must relate to the corresponding national interest in promoting national security, economic health, and societal well-being. NASA 's programs must address maintenance of key national space science competences and capabilities, including those of the universities (research and education) and industry, as well as government. Planning and decision making must therefore address national—and perhaps even international—interests and capabilities along with narrower programmatic factors.

Chapter 2 briefly describes the way that space science is currently organized and managed in NASA. A number of changes have been or are being introduced concurrently with this study. Because of this, the recommendations that follow are stated so as to be broadly applicable and transcend the specific details of the organization.

The major imperatives for the recent and coming changes are the continuing pressure for budget reductions and the Administration initiative to reinvent government. These imperatives have led to a significant downsizing of NASA, now under way. An internal “Zero Base Review, ”¹ the subject of briefings to Congress on May 19, 1995, concluded that the budget targets of the Administration can be achieved by reducing infrastructure costs (streamlining operations, reducing overlap, and so forth) without cutting programs or closing any of NASA's 10 major field centers.² Cuts beyond those in the budget submission are likely to threaten program content.

The NASA civil service workforce has already been reduced through buyouts and attrition. Actions are being taken to further reduce the NASA Headquarters infrastructure, both civil service and support

contractor, including the transfer of functions and personnel to the field centers. In the view of the “new NASA,” Headquarters would be based on a “corporate management” model and concentrate on policy development and implementation, external relations and liaison, and agency-level and enterprise-level management. Program management would be shifted to the field centers, where project management already generally resides. By this shift, NASA intends to eliminate a layer of management with its attendant personnel.

Changes at the field centers are also under study. One involves consolidating similar activities at a single center and thus eliminating duplicative activities. Another is to establish certain specialized center activities as “institutes” affiliated with a nongovernment organization (e.g., a university). Many of the proposed changes are consistent with the recommendations of the NASA Federal Laboratory Review.3 At present, transforming entire civil service centers to contract operation (as is already the case at JPL) is not under consideration.

In assessing organizational alternatives for NASA, the committee first addressed the issue of fundamental responsibilities of the various major elements of the NASA science establishment: NASA Headquarters, the field centers, and the large and diverse outside industrial and academic research community. Management of NASA's scientific research must, of course, remain within NASA. Management functions will be exercised at various organizational levels, depending on scope and function. The following are the committee's findings:

1. NASA's strategic enterprises are evolving as the basic business units of the agency at the highest strategic level,4 including within each enterprise one or more programs. Clearly, enterprise management must be located at Headquarters. At this level, NASA addresses goals, objectives, policy, and budget allocation and thus provides the overall direction, guidance, and constraints to individual programs.

2. Program management (see NASA's definition of “program” in Appendix D) may be either at Headquarters or at a center. Cautions with respect to the assignment of program management are as follows:

   1. Programs with extensive and complex elements at more than one NASA center should be managed from Headquarters. While day-to-day program support might be delegated to a center, program definition, resource allocation, and project review should be Headquarters' responsibilities. The management of space science technology is an example of a complex, multicenter program that should not be managed by a field center (see Recommendation 6-4).

   2. Programs whose elements reside primarily within a single center or that have simple interfaces between participating field centers may be managed by a field center if there is clear separation of the program function from project management. Project management is narrowly focused on project execution, while program management considers more broadly all elements of the program. The authorities, accountabilities, objectives, and motivations of the respective management groups are different, and separation is required. The New Millennium program is an example of a relatively simple, multicenter program that is currently managed by a field center (JPL).

   3. The program management center must rigorously enforce a policy that mandates fairness to program elements outside the center vis-à-vis program elements within.

   4. Any decisions on out-sourcing (make-buy decisions) that may involve the center as performer must be made at Headquarters. This principle applies both for in-house/out-of-house decisions on spacecraft development and for selection of investigations (both for flight projects and for the Research and Analysis program), including oversight of the peer review process.

3. Project management (see the definition in Appendix D) is properly located at a center, even if a center is operated under a contract. Again, some limitations are necessary:

   1. Major make-buy decisions should be approved at higher levels.

   2. Any delegation of project management responsibilities to contract centers must adhere to procurement law and regulations.

4. Because of the existence outside the agency of larger and more diverse industrial and academic research communities and the importance of maintaining that national capability, scientific research should, for the most part, be conducted outside the agency. The committee believes this to be true for flight projects, for the scientific experiments that fly on them, and for the supporting work conducted in ground-based laboratories. There are important exceptions to this rule, and these are discussed under “Balance Between In-House and Out-of-House Research” at the end of this chapter.

In summary, Headquarters should be responsible for enterprise management, including establishment of scientific vision, goals, and objectives; guidance on program content; cost guidelines and constraints; requesting program plans; integrating and approving submitted plans; source selection; scientific discipline support; oversight; and cross-discipline integration. Centers should be responsible for creating program plans in response to Headquarters requests; executing programs and projects; providing program and project oversight; and possibly managing the solicitation process (with selection controlled by Headquarters). Scientific research should be conducted primarily by the academic and industrial communities. The NASA Federal Laboratory Review report5 reaches similar conclusions and recommendations with respect to Headquarters and center roles and missions.

The committee followed closely the charge given it and looked at three alternatives for the Headquarters science organization: an institute approach to science management, following the NIH model, distributed management among the three science offices as exists today, and centralized science management as existed prior to the 1993 reorganization. The committee assumed that a degree of program management would be retained at Headquarters in any of the models. Thus, for example, the Headquarters Office of Space Science (which is responsible for the Space Science Enterprise in its entirety) is assumed to retain a broad management function for the three science programs, Astrophysics, Space Physics, and Solar System Exploration, as well as for the Space Science Enterprise as a whole.

1. Institute for Space Science—Responding to direction in the FY 1994 Senate appropriations report, the committee considered a space science “umbrella organization within NASA to coordinate and oversee all space science activities,” functioning “just as the National Institutes of Health now does within the Department of Health and Human Services.” This was taken to mean an institute constituted with all the science programs of the pre-1993 OSSA, funded separately from the rest of NASA and operating much as does NIH and the many individual institutes coming under its umbrella. The operation of NIH is described in Chapter 2 and Appendix F. (Let it be clear that this model is not what NASA has recently proposed as “science institutes,” which are discussed below.)

   The advantages of such an approach are

   • relative independence of science and

   • relative budgetary autonomy.

The disadvantages are

• possible disconnection from other elements of NASA that support science (e.g., launches, tracking, facilities operations) and

• budgetary vulnerability.

The committee concluded that the NIH model, while quite effective in its own arena of research with a heavy emphasis on fundamental or basic research leading to the understanding of disease processes, the development of new drugs and treatment protocols, and their clinical evaluation, is not appropriate for the space sciences. NASA space sciences depend on close coordination with other elements of NASA, for example, hardware development, launch services, and tracking and data operations. There are no obvious counterparts in the NIH model, which instead involves primarily self-contained research endeavors. Although NIH research projects are conducted in laboratories, hospitals, and the like, the basically stable infrastructure of these facilities is dealt with simply as “overhead” for the investigation. With a few exceptions, such as establishment of cancer, eye research, and AIDS vaccine development centers, NIH researchers do not generally have to build hospitals or new laboratories from the ground up to carry out their work.

While ground-based laboratory research supporting science conducted in space may resemble NIH research, NASA research carried out on space missions does not. Of necessity, space science proceeds incrementally, flight project by flight project, whereas most other scientific programs proceed by the steady interplay between discovery, new questions, new concepts, new technology, and the changing research interests of scientists. Only creation of a major new particle accelerator, for example, approaches the long lead times, engineering teams, costly and complex infrastructure, and need for schedule management inherent in spaceflight projects.

It also is not clear how such an institute model would deal with budgetary issues. Presumably, an institute could be funded separately from the rest of NASA, perhaps within the same appropriation, but otherwise independently. The committee was concerned about possible mismatches in science funding and related infrastructure funding. Even if science support infrastructure funding were joined to science program funding for the institute, the resulting “fences” between fund sources could constrain trade-offs and thus severely hamper management of the overall NASA endeavor.

Much of the remarkable success of the space sciences since the dawn of the space age has resulted from a tight link between science and space technology. The inspiring goals of scientific discovery and exploration spurred new capabilities in space access and technology, which in turn pushed the scientific goals still further into the unknown. While some technology came from the science programs themselves, much of it came from other parts of NASA. Separating the space sciences from critical space technologies within the agency by forming an independent institute for space science would break this symbiotic link. The committee found the downside risks of breaking this link too large to accept.

In summary, the committee does not favor establishment of an institute for space science or other entity that might be separate or “semiseparate ” from NASA as a whole.

2. Distributed Management (current organization) Versus Centralized Management (pre-1993 organization)—In the current organization, three program offices manage elements of the science program—OSS (traditional space science), OMTPE (Earth science), and OLMSA (life and microgravity sciences). They are directly aligned with the respective enterprises, OSS and OMTPE uniquely, with OLMSA sharing the Human Exploration and Development of Space Enterprise with the Office of Space Flight. The pre-1993 science organization was wholly contained within the Office of Space Science and Applications (OSSA), which consisted of program divisions for Astrophysics, Space Physics, Solar

System Exploration, Earth Science, Life Science (both Space Medicine and Space Biology), and Microgravity Science.

The advantages of the distributed management model over the centralized management model are

• more focused management to achieve enterprise goals (or to implement strategic enterprises),

• simpler management interactions with centers, particularly if centers are realigned as proposed to achieve greater program focus,

• for life science and microgravity communities, the possibility of greater funding attracted through their Human Exploration and Development of Space Enterprise connection, and

• simpler strategic planning because of tighter focus of programs.

Conversely, the disadvantages are

• reduction of budgetary and scientific flexibility for the respective science leaders,

• greater vulnerability of science quality because of program trade-offs necessitated by budget reductions or policy changes independent of the science programs,

• greater vulnerability of smaller programs through their increased “visibility” in the budget process,

• greater difficulty in integrating related science programs, such as life science experiments with planetary spacecraft,

• weaker advocacy for science within NASA (three voices, not necessarily in agreement, heard by the Chief Scientist and the Administrator), and

• larger Headquarters staff overhead.

In conclusion, from a pure science perspective, a centralized-management-type organization appears to be preferable. But NASA is not a pure science organization. Science activities must be closely integrated into the other parts of the space program. The enterprise concept now in place provides a solid basis for that integration in the context of program rationale. In addition, the advantages offered by tighter focus for planning and simplicity of center interactions were given positive weight by the committee. It is also true that the science programs have changed dramatically since the days of the previous organization (for example, the advent of the Space Station and the great expansion of the Earth observation program). The committee therefore endorses the present approach, which distributes science among three separate enterprises.

Recommendation 4-1: The committee found no compelling reason to establish an independent institute for space science modeled on the National Institutes of Health. Rather, NASA should retain its science programs in the present three-enterprise distributed form, but also take action to provide the necessary integration among the programs (see Recommendation 4-2).

With this distributed management model, the three science associate administrators must coordinate and integrate their respective programs into a single, coherent NASA science program. The Chief Scientist can exercise the strong, central coordinating function that is required to ensure that this integration occurs, but the current authorities and responsibilities of the Chief Scientist position are not, in the view of the committee, explicitly sufficient. In addition to its advisory, coordination, and interface responsibilities, the position should include a formal concurrence on planning, programming, and budget matters of all the sciences and authority to resolve conflicts among the science offices. The NASA Science Council, currently chaired by the Chief Scientist and including as members the science associate administrators, would provide a mechanism for the Chief Scientist to use in meeting these expanded

responsibilities. The responsibility for formal concurrence on planning, conflict resolution, oversight of science quality, and representation of NASA to the outside world exceeds the capabilities of any one individual. For this reason, recommendations for strengthening the role of the Chief Scientist will require creation of a small office staff of experienced and technically astute personnel to support the Chief Scientist.

In NASA, as in other R&D organizations, the strongest control of prioritization lies with officials who control budgetary decisions, in this case the science associate administrators. If the Chief Scientist were given general budgetary authority over these offices, the effect would be tantamount to re-creation of a single science office, like OSSA. What is being recommended here instead is the establishment of a

formal planning and budgeting concurrence by the Chief Scientist; in practice, the coordinating influence of the Chief Scientist would be enforced by the location of that position within the office of the Administrator. The budget development and advocacy process would continue to reside with the line associate administrators.

Recommendation 4-2: The position of Chief Scientist should be strengthened to ensure full integration of the space sciences. A proposed functional statement that encompasses the necessary authority is provided, followed by a proposed functional statement for the NASA Science Council (see box).

With the establishment of enterprise management, NASA seems to have created another management layer to add to those already in existence. There is an overlap in the functions of enterprise management and program management. The committee recommends that program management functions be subsumed by the enterprises and appropriate elements then delegated to the centers.

Recommendation 4-3: NASA should eliminate the separate “program management” layer, assign to the enterprises the program management functions properly located at Headquarters, and delegate to the field centers those program management functions properly located there.

With enterprise management at Headquarters, the major functions of field centers will be management of out-of-house projects and execution of in-house program elements (i.e., projects and investigations).

The committee considered two primary alternative organizational arrangements for a science-oriented field center: (1) the current civil service arrangement for centers (except JPL) and (2) centers wholly operated by an appropriate university, consortium, or joint university/industry partnership as a government-owned, contractor-operated (GOCO) installation (the JPL model). Because only GSFC and JPL are primarily science centers, the first alternative is equivalent to making no change and the second corresponds to converting Goddard to a GOCO; the committee considered the latter option even though it was not recommended by the Zero Base Review report. No consideration was given to converting JPL to the civil service.

The committee also considered establishment of “institutes,” defined as university- (or consortium-or university/industry partnership-) staffed and operated organizations to conduct certain specific science operations now residing at a nonscience center.⁶ This approach was considered not as an alternative to the two preceding options for a whole center, but in the more limited sense considered by the Zero Base Review for collecting the scientific activities of a center that are not closely aligned to the primary mission of that center. Examples currently under consideration are a life science institute at Ames Research Center, primarily an aeronautical research center, and biomedical and planetary science institute(s) at Johnson Space Center, primarily a human spaceflight center. (Advantages and disadvantages for such focused “institutes” for discrete scientific activities in a predominantly nonscience-oriented center are discussed below.)

Table 4.1 displays the advantages and disadvantages of the two organizational concepts for entire centers.

Current Civil Service Arrangement—Perhaps the principal advantage to retaining the present civil

service system for GSFC is that no action is required; there would be no change to the present system and thus no lengthy transition to undergo. The last advantage, stability of staffing, results from the present NASA budgeting system, which provides funds for civil service staff from one account (Research and Program Management—R&PM) and program funds (contract costs and in-house project costs other than staff salary) from another account (R&D). The R&PM account has been relatively stable over time, linked closely to staff size for the agency as a whole. Thus staff can be maintained during short-term gaps in program funding (such as deferral of new project starts) and a center director has the discretionary authority to assign staff to high-priority activities.

The chief disadvantage of retaining the present system at GSFC is the inflexibility associated with the civil service system, such as hiring, firing, and salary rigidity. This may negatively affect staff quality over the long term, especially in a period of downsizing, as is the case at present. The committee considered this the principal reason for favoring a change.

Government-Owned, Contractor-Operated (GOCO) Centers (the JPL model) —This alternative envisions converting GSFC into a university-operated GOCO center like JPL. If this could be accomplished, the principal advantage would be the flexibility in personnel management inherent in the private sector. Through this mechanism, staff quality in the long run could be enhanced. Further, increased flexibility in personnel management would facilitate adjustments in staff capabilities to meet changing requirements, such as infrastructure reduction.

The principal disadvantage of converting GSFC to a GOCO center would be the difficulty of the transition itself. Conceptually, conversion would be quite straightforward: disestablishing the center as an organization of federal employees and establishing the GOCO under federally funded research and development center (FFRDC) rules, establishing a competition for the operation of the center as a GOCO, and closing out the civil service complement effective the date the contractor assumes responsibility. The difficulties focus on the conversion of civil servants to private sector employees, particularly with respect to retirement programs, but also involving other facets of personnel management, such as

salary scale, seniority, union representation, and benefit programs. The committee believes it is feasible but recognizes the potential for upheaval. Indeed, the committee was briefed on a similar plan prepared for the Naval Research Laboratory (NRL) some years ago. Though the Navy has not implemented the plan, it appears feasible.

A second disadvantage of conversion is the transition cost and duration; cost was estimated for NRL to be in the range of $137 to $162 million, depending on the generosity of severance and other payouts, and three years was the estimated time required, provided no legislation was necessary. Corresponding cost and time for NASA conversion of GSFC would likely be of the same order of magnitude. Finally, as a contract organization, the center could not be responsible for certain inherently governmental functions (e.g., fundamental make-buy decisions), and these would therefore have to be retained in Headquarters.

The committee does not list recurring costs as either an advantage or disadvantage in comparing the present civil service system with a GOCO system. At first look, one might conclude that total costs for the GOCO might be higher because of a higher salary scale and the contract fee. But other efficiencies inherent in a contract arrangement might compensate, and which system has greater cost is not clear. Given the likelihood of little difference, the committee concluded that cost was not a decisive factor. However, it is important to recognize that any long-term cost comparisons between in-house and out-of-house alternatives must include all program costs, both R &D and R&PM. Considering only R&D costs would make the civil service in-house alternative look far lower in cost, because in-house R&PM costs are not currently charged to R&D programs.

Later in this chapter, the committee recommends the adoption by NASA of an industrial funding approach for its field centers, whereby all costs of executing a program, including the cost of civil service staff, would be charged to that program. This full-cost accounting, toward which NASA is already moving, would considerably reduce any advantage in responsiveness and accountability held by a GOCO management approach.

The flexibility in personnel management and attendant positive impact on staff quality drive the committee to the conclusion that the GOCO approach offers the greatest long-term benefit. There has been concern ⁷ that the technical competence of government laboratories employing civil servants erodes over time; this possibility provided some of the motivation for the proposal to convert NRL. Despite the formidable difficulties of conversion to GOCO status, a decline of personnel quality at GSFC would favor such a move. However, even within the present civil service system, GSFC has been able to maintain a high staff quality. The committee concluded that GSFC should be left as is for the present, but that this conclusion should be reexamined periodically.

Recommendation 4-4: The Goddard Space Flight Center should remain a civil service center for the time being. However, NASA should establish periodic external reviews of GSFC to determine whether new consideration should be given to conversion to a GOCO arrangement.

As remarked above, NASA is also considering establishment of a class of university-, consortium-, or partnership-operated institutes. Not an alternative for operation of an entire center, this institute concept is being considered for a focused science activity within a center having broader, and generally nonscientific, responsibilities. Such institutes would be operated by a nongovernmental entity (e.g., a university) under contract. A number of such arrangements have been proposed in the Zero Base Review plan.

Some advantages and disadvantages of this scheme are displayed in Table 4.2. NASA's stated intent in creating such institutes is to upgrade the quality of the science programs concerned, bringing them to a

level from which they can compete successfully for support, and then to reduce the “guaranteed” or “base” NASA funding level. The institutes would then have to survive (or shrink and “sunset”) on the basis of merit competition for funding.

Synergy of an institute's science with that of its adoptive parent is an especially important feature of the concept. In most similar current arrangements, both at NASA and elsewhere (e.g., DOE and NSF), the contractor is a nonprofit entity, usually a university or a consortium of universities, and that is the arrangement assumed here. It is further assumed that the institute's scientific staff are contractor employees, hired by the university or consortium, either after conversion from civil service or based on criteria established in the contract. There are several especially positive features to this approach:

• If the scientific and geographical match is carefully chosen, the staff of the institute and the faculty and research staff of the university or consortium will have overlapping scientific interests that should result in scientific collaboration, benefiting both parties. The university should contribute something of value (e.g., facilities, faculty participation, or student involvement) to enhance the focus on common goals and problems.

• The institute staff will be an additional resource for, and will benefit from, the interaction with graduate students.

• By retaining the institute, with its economic and cultural benefits to the community, it might be possible to secure some support from the state government(s).

• As noted for a GOCO-type center, establishment of a personnel system that is outside the civil service should increase the flexibility of NASA in establishing and retaining a high-quality staff that can adapt to fit a NASA mission that evolves in size, scope, and focus.

Some negative features of such institutes exist, as well:

• Funding will be problematical; first, the institute will have to compete successfully for NASA program funds. Then, if not fully successful, the institute will need to seek outside funding to maintain itself, and, because of limited outside interest in its research areas, such outside funding might be difficult to attract.

• The host NASA center is likely to be less motivated to provide support to the institute than to

support its own elements. Of particular concern is the danger that the institute might lose the coupling between its scientific research and NASA engineering support.

• There will be a need for a somewhat increased NASA infrastructure to see to the “care and feeding” of the institute.

• The tendency for such institutes to proliferate, sometimes in response to political pressures, must be controlled. Similar pressures will exist to maintain individual institutes indefinitely, irrespective of merit.

The committee believes that the institute model for NASA scientific staff can, in appropriate cases, help to preserve and improve the quality of some focused science activities, while providing other potential benefits in flexibility and NASA-university collaboration. However, success of these institutes will depend, in large measure, on the not-yet-developed specific agreements between NASA and the operator. In the long run, if budgets continue to decline, quality will be maintained only if the less-competitive institutes are reduced in size or terminated as a consequence of a continuing process of peer-reviewed scientific competition.

Recommendation 4-5: Creation of contractor-operated institutes (a focused science activity operated under contract within a center having broader, possibly nonscientific, responsibilities) may be advantageous for specific science functions or facilities. The selection and creation of such institutes should be limited to special circumstances, specifically,

• to be responsible for a narrow, bounded area of science,

• to operate a special facility for the scientific community, and

• to maintain a unique national facility or staff capability,

but

• NOT just to lower civil service headcount and

• NOT if the entity is an integral part of the larger organization (the host center).

The committee recommends that, if NASA proceeds to establish pilot science institutes, it give due attention to the out-years process by which these institutes will have their guaranteed base budgets ramped down, and to the competitive process under which they will be expected to compete successfully and maintain or increase their size, or to compete less successfully and shrink or terminate in an orderly fashion. The committee recommends that other institute initiatives be deferred until the success of initial pilots can be evaluated.

A number of management issues that bear directly on one or another of the center alternatives were identified and are discussed in the following sections.

NASA requires in-house scientists for its space research, exploration, and technology programs. These scientists coordinate science and operations on larger missions, guide development and utilization of unique research facilities, assist outside scientists and technologists to effectively use NASA facilities or flight opportunities, and enable NASA to act as a “smart buyer.” The number of in-house scientists

should be determined by the extent of these support functions and not by a desire to exploit perceived flight opportunities. Space science leadership and the generation and testing of new ideas should be the domain of the broader scientific community, of which the NASA scientists are only a part. As noted earlier, the committee believes that scientific research should, for the most part, be conducted outside the agency.

The in-house scientists should be of the highest quality. To maintain that quality, they should be encouraged to practice their science at the cutting edge of research and instrument development. However, their research should be thought of as being “derived” from the NASA mission and not their sole duty. For the most part, NASA scientists should be expected to devote a significant fraction of their time to support functions (e.g., as project scientist) and a commensurate fraction of their funds should come from the programs and projects that benefit from these efforts. Additional research funds should be earned through open competition among the broader scientific community. By linking the responsibilities of the scientific staff directly to programs and projects, NASA can establish and maintain a positive feedback between the size of its programs and the size of its scientific staff.

Recommendation 4-6: NASA should conduct scientific research activities in-house only to the extent required to compete successfully for high-quality staff, to maintain capability to manage and interface with corresponding activities on the outside (e.g., as a “smart buyer”), or to maintain and operate an in-house capability that is unique in the nation. Otherwise, if a valid out-of-house capability exists, the work should be done there.

In discussing the two alternatives for a science field center, civil service versus GOCO, it was noted that NASA funds operation of the institution (personnel, facilities, and so forth) from the R&PM budget, determined chiefly by staff size and operations costs of the installation, while program costs (primarily procurement costs) are funded by the relevant program's R&D budget. This is termed “institutional” funding. Conversely, “industrial” funding provides that all costs of a specific program (including personnel) be charged to the sponsor of that program. Although all the NASA civil service centers are institutionally funded, JPL, as a contract center, is essentially industrially funded, with personnel as well as program costs charged to the responsible program office.

Advantages of employing institutional funding, rather than industrial funding, are

• flexibility in assigning people,

• ability to maintain a stable workforce, and

• ability to conduct possibly high-payoff, speculative research.

Disadvantages of institutional funding, compared to industrial funding, are

• less accountability,

• tendency for workforce factors to drive program decisions,

• disincentive for detailed workforce planning,

• greater difficulty in managing staff size, quality, and skill mix, and

• distorted make-buy decisions.

An additional disadvantage of industrial funding is that research sponsors may tend to use it as a management tool, that is, “buying ” research in little bits and pieces, thus creating a paper problem and stimulating researchers to spend excessive time seeking funding.

Overall, the committee strongly favors the concept of industrial funding because it reflects total program costs and permits ready assessment of comparative costs that might otherwise be hidden by an institutional funding approach. The committee recognizes as a potential problem the assignment of responsibility for maintenance of agency infrastructure but believes that higher priority must be given to applying the budget discipline of industrial funding. The committee expects that senior management will be able to resolve these infrastructure issues even under the discipline of industrial funding. The committee endorses NASA's decision (stated in the Zero Base Review briefing to the Congress) to identify, budget, and manage by total program costs, including the costs of civil servants.

Recommendation 4-7: NASA should adopt an industrial funding model for field centers. Decisions on program priorities and budget would be more rational if based on full-cost accounting, and program accountability and discipline in personnel management would be thereby enhanced. The principal objective is that cost accounting should be based on full program costs, including civil service salaries, to achieve the advantageous features cited above. A similar recommendation was made in the NASA Federal Laboratory Review report.⁸

Currently, each NASA field center reports institutionally to a NASA Headquarters program office that has major, but not necessarily exclusive, program responsibility at that center. Center institutional costs (salaries and other personnel costs, facility operations, and other nonprogram specific support costs) are funded from the R&PM budget. Program direction and funding, however, come to the center from Headquarters program offices. For example, although the Office of Space Science funds a large fraction of the programs located at GSFC, institutional management is provided by OMTPE, the institutional “parent” of GSFC, and institutional costs are charged to the R&PM account. Likewise, institutional management of the atmospheric research program of OMTPE at the Langley Research Center is provided by the Office of Aeronautics, Langley 's “parent” institutional office, with institutional costs again charged to R&PM. Thus programs at a center “owned” by a different institutional office operate in a dual (in some cases multiple) management and funding environment. If the respective office priorities are significantly different, programs may be impaired.

An alternative institutional reporting arrangement, that is, with all centers reporting to one Headquarters institutional manager, has been employed by NASA in the past⁹ (see Table 2.2). Under this arrangement, all programs executed by field centers would face split funding, as all institutional costs would be provided by this separate office. An advantage of this system is that institutional resource priorities would be established for the agency as a whole and would presumably directly reflect overall program priorities of the Administrator, to whom the single institutional director would report. A disadvantage is that it would be difficult for a program associate administrator to redirect institutional resources within his or her program area even at a center that dominantly supports that program.

Three different kinds of requirements can be imposed on a program, that is, programmatic, institutional, and regulatory (e.g., environmental, safety, procurement). Neither of the alternatives for institutional management can completely eliminate this complexity. Therefore, whatever the system, it is important that all potentially competing requirements and directions flow down to center program/ project management through a single channel. Given the trend toward concentration of center capabilities toward a primary mission for that center, coupled with more focused enterprise management, the committee concluded that center reporting arrangements should remain as they currently are, with each

center reporting to an appropriate enterprise and all direction channeled through that enterprise. Some split reporting will result, but the alternative of having centers report to a central office would exacerbate that split. If some form of industrial funding is put in place (by R&PM allocation, budget structure change, or center GOCO transition), the problems resulting from split reporting will be further reduced because program budgets would then include staff and other institutional costs. In those cases where a significant split in center responsibilities exists (e.g., at GSFC, where astrophysics and space physics programs constitute a major part of the center 's activities, though it reports institutionally to OMTPE), NASA must actively work to ensure that program and institutional needs are well served.

Recommendation 4-8: Center reporting arrangements should remain as they currently are, with each center reporting to an appropriate enterprise. All requirements and directions addressed to a center from any source should be directed through that enterprise.

Successful execution of a project, whether in-house or through a contract, depends on meeting simultaneously the established schedule, budget, and scientific requirements. These three sets of requirements will come into conflict repeatedly during the course of any project. Therefore, it is essential that the management structure permit trade-offs throughout the life of the project. Science advocacy is usually embodied in the position of the NASA “project scientist,” who specifies scientific performance, leads the science working group, and provides an ongoing assessment of whether or not the project will achieve that performance. This function is independent of whether the center is a civil service or GOCO center and applies both for projects executed in-house and those procured under contract.

The project scientist is responsible for documenting the scientific requirements and for evaluating engineering designs and other documentation in order to determine whether the requirements will be met. In order to fulfill this responsibility, he or she must normally be co-located with the project team and must be given the opportunity to concur in all significant project decisions that will affect scientific performance. The project scientist should participate in all major project reviews, including preliminary and critical design reviews. While the project scientist is responsible for scientific performance, the project manager is responsible for overall mission performance, including keeping the project on schedule and within budget. When they cannot agree on a course of action, resolution must be sought at a higher level.

Committee members involved with NASA science projects have observed instances of what appears to be a diminished role of the project scientist in science project management. Fundamental decisions bearing on the ability of the project to meet its scientific objectives have sometimes been made without adequate project scientist (or investigator) input and involvement.

Recommendation 4-9: Renewed emphasis should be given to the essential role and responsibility of the project scientist in project trade-offs and decisions through the life of the project.

A recurring theme heard from interviewees from other organizations and from some committee members was that NASA imposes an excessive degree of oversight on its contractors. While the discussion centered on anecdotal evidence, the committee was struck by the high frequency with which the subject was raised. The issue arose regarding oversight of contractors by field centers, of investigators by Headquarters scientific staff, and of field centers by Headquarters. This heavy oversight has two important detrimental effects: performers may be required to devote unwarranted time to satisfying oversight requirements at the expense of project performance, and managers spend excessive time in

questionable oversight, including heavy attendance of monitoring staff at project meetings. Redundant oversight in a period of constrained budgets can impose especially severe penalties.

Historically, a hallmark of NASA management has been rigorous engineering oversight and penetration into all high-value space flight projects extending well beyond the safety-critical human flight programs. This management practice was established early in the Apollo Program and grew in intensity with mission complexity, becoming identified within NASA as a key component and major contributor to early agency success. High-visibility accidents such as the Apollo Command Module launchpad fire, Skylab meteoroid shield failure, and the 1986 Challenger accident only reinforced the belief that active independent program oversight was necessary for success. The Rogers Commission Challenger report¹⁰ was particularly critical of decentralized NASA management, so that attention was further focused on strong Headquarters oversight, short rein on center authority, and an independent review process to balance the decision-making process. Delegation of program execution exclusively to project managers at field centers was not viewed as compatible with the oversight requirement, and program management was centralized at NASA Headquarters with a Levels 0, 1, 2, 3, and 4 chain-of-command architecture (Administrator's Office, associate administrator, lead center, center project office, and contractor management, respectively). A strong argument can be made that this heightened oversight worked, but not without costs. The practice has penetrated most of NASA' s programs, bringing with it the bureaucracy necessary for its implementation. The Space Shuttle and Space Station programs have clearly been subject to this renewed management style fashioned during the Apollo era; however, science programs within the agency have also seen tighter control and oversight from Headquarters, especially since the Hubble Space Telescope mirror problem.

For a number of reasons, management decisions being made at NASA over the past several years are reversing this trend. Reasons include declining budgets, substantial NASA staff reductions, and an emphasis on shorter development cycles and smaller-scale projects. Greater responsibility and accountability for project execution are being given to contractors. The large human space flight programs, the Space Shuttle and Space Station, have also seen decreased centralization and delegation of authority to the field sites. These delegations are consistent with the movement of Total Quality Management principles to the forefront of both government and industry, where the intent is to build in quality from the beginning, rather than try to achieve it through intensive oversight.

Recommendation 4-10: NASA should maximize delegation of project responsibility to the executing agents of its space science programs (center, laboratory, university, or industry) and minimize redundant NASA oversight. This practice can yield excellent results at an affordable cost. High-value, one-of-a-kind space science missions may justify increased independent oversight. If necessary, this increased oversight should be clearly established in the program definition stage and should not be permitted to creep or evolve with program maturity. The committee makes no recommendation regarding oversight, delegation of responsibility, or privatization in the human spaceflight program.