Fixing the National Laboratory System
Having reviewed the recommendations of outsiders, the Department of Energy responds with a plan of action.
Over the past 20 years, there have been numerous studies of the Department of Energy (DOE) national laboratories. Most of these studies recognize the high technical quality and unique capabilities of the laboratories but also have raised serious questions about the viability of their post-Cold War, post-energy-crisis missions; their cost; their claim on federal R&D funds vis-a-vis universities and industry; and their size, focus, and possible redundancy.
These studies have resulted in a mountain of incompatible recommendations and expectations for action. Some people have proposed new missions for the laboratories, such as industrial competitiveness and industrial ecology. Others have suggested new management structures. One of the most recent and provocative studies, the Report of the Task Force on Alternative Futures for Department of Energy National Laboratories, chaired by Robert Galvin of Motorola, proposed managing the laboratories through a government-owned corporation. In addition, bills were introduced in the 104th Congress that would have established a laboratory-closure commission modeled after the Department of Defense base-closure commission, required specified staff cuts of one-third or more at the laboratories, and dismantled DOE and reassigned the laboratories to other government agencies with related missions.
Having reviewed them, we find that there is much merit in the criticisms and much good intention in the proposals to change the governance and focus of the laboratories. But we believe that the proposals either attack the wrong problems or are simply unlikely to succeed. We have developed instead what we believe is a comprehensive and practical strategy to manage the national laboratory system to achieve national goals and to make the laboratory systems once again recognized as an essential, cost-effective, and well-managed element of the nation's R&D enterprise. We believe that this approach, unlike others tried over the past two decades, is demonstrably working and should be continued.
When we refer to the "national laboratories" in this paper we mean DOE's nine large multiprogram laboratories that serve several DOE missions.
DOE National Laboratories
|Laboratory||Principal Mission Role||FY 1998
|Idaho National Engineering Laboratory||Environment||715|
|Lawrence Livermore||National Security||1080|
|Los Alamos||National Security||1146|
Note: includes funding from non-DOE sources
They have a total annual budget of more than $6 billion and employ some 30,000 scientists and engineers. About $1 billion of this funding comes from outside DOE, primarily from other federal agencies but also from the private industry. The national laboratories play major roles in each of DOE's four missions (national security, energy, science, and environmental quality), but nearly 80 percent of their R&D is concentrated in the national security and science missions.
In addition to the multiprogram national laboratories, DOE also has several specialized laboratories, such as the Stanford Linear Accelerator Center and the Federal Energy Technology Center, that serve a single mission. Most of the debates about the laboratories have focused, as does this article, on the multiprogram national labs, but many of the same issues are relevant to the single-mission labs.
Although owned by the federal government, the national laboratories are operated by nongovernmental contractors. This management arrangement has been followed since their inception, with the intent of giving the laboratories the flexibility they need to attract top scientists and take advantage of private-sector management practices. At present, four of the nine national laboratories are managed by universities, three are managed by a for-profit corporation (Lockheed Martin Corporation), and two are managed by not-for-profit contractors.
Our management approach
External Members of the Laboratory Operations Board
John P. McTague, Vice President, Ford
Robert Bringer, Vice President, 3M (ret.)
Richard Celeste, former governor, Ohio
Paul Fleury, Dean, University of New Mexico
Edward Frieman, Director (ret.), Scripps Oceanographic Institute
Paul Gilman, Executive Director, National Research Council
M.R.C. Greenwood, Chancellor, UC Santa Cruz
Maxine Savitz, General Manager, Allied Signal Ceramics
Robert Wertheim, Admiral (ret.), U.S. Navy
In April 1995, DOE established the Laboratory Operations Board to guide reforms of its laboratories. The purpose of the board is to apply dedicated attention and private-sector expertise to laboratory management on a continuing basis. The board consists of the DOE senior managers who bear responsibility for the laboratories, plus external members with broad expertise in industry and government. The internal members are political appointees with no long-term vested interest in preserving the status quo.
The focus of the board's work has been to improve the DOE management processes that shape the laboratories. It is important to understand that Congress, in general, does not directly fund the laboratories. There is no line item or discrete set of line items in the DOE budget for any of the national laboratories. Instead, Congress appropriates funds for DOE in several hundred distinct budget codes. Managers in each program decide (sometimes with some congressional direction) how to allocate funds among laboratories, universities, and companies. It is DOE, not the laboratories, that has public missions, defined by the direction and funding provided by Congress and the White House. DOE must choose how to accomplish these missions most effectively, using the capabilities of the laboratories as well as those of industry and universities.
As a first step in its efforts to reform the laboratories, DOE, under the direction of the board, prepared the Strategic Laboratory Missions Plan-Phase 1, which serves as a baseline for the management reforms. It describes how DOE is currently using the laboratories and defines a vision for the future and a set of management principles. In the vision, which provides the direction for the management reforms, the laboratories will:
Maintain the highest standards of excellence in science and technology. Investments in the laboratories must be focused so that the laboratories are world-class in their specialties.
Have well-defined roles in achieving DOE mission outcomes. Each laboratory must have clear and widely understood roles as well as distinctive technical competencies.
Be well integrated with the nation's R&D enterprise. The laboratories must have strong and mutually supportive links to other agencies, universities, and industry. They should live up to their titles as true national laboratories, serving other customers as well as DOE.
Be recognized as highly efficient and cost-effective research institutions. The laboratories must have substantially reduced overhead costs, and the laboratory complex must be sized properly.
The approach the board has taken to move toward this vision is to ask a series of questions regarding DOE's key management processes that determine the size, focus, and efficiency of the laboratories. The board has been going through a systematic process to address these questions. This work is far from finished but it is making progress. We report here on the results to date and the future directions.
Has DOE defined compelling outcomes in its missions? Our view-and the judgement reached by President Clinton, numerous high-level advisory panels, and Congress as expressed in legislation-is that DOE has vital missions to perform. The end of the Cold War and the absence of an immediate energy crisis do not eliminate a role for DOE and its laboratories. The Galvin task force, although critical of DOE's management of the laboratories, found that they have "a compelling agenda of important work to be performed in their traditional mission areas." It is useful to examine briefly the most significant goals in each mission area.
In the national security mission, the primary goal is to maintain confidence in the safety and reliability of the nation's nuclear deterrent without nuclear weapons testing. This is a highly challenging technical goal, requiring major advances in computer modeling and simulation and nonnuclear testing. President Clinton's commitment to a comprehensive nuclear test ban treaty is dependent on DOE and the laboratories meeting this challenge. Although the budget for defense programs has been cut in half since the end of the Cold War (from $8 billion to $4 billion per year in constant 1997 dollars), these cuts have been concentrated in the weapons production facilities. The need for R&D at the laboratories has remained strong as the program shifts from weapons production to science-based stockpile stewardship. The other main national security goal is to reduce the risk of nuclear proliferation, including helping to secure nuclear materials and technologies in the countries of the former Soviet Union. President Clinton has declared this also to be of the highest national priority.
DOE's science mission provides scientists throughout the country in universities, industry, and laboratories with world-class, large-scale scientific facilities such as synchrotron light sources, neutron sources, and particle accelerators. In the past two years, three such major new facilities-the Advanced Light Source at Lawrence Berkeley National Laboratory, the Advanced Photon Source at Argonne National Laboratory, and the Continuous Electron Beam Accelerator Facility at the Thomas Jefferson National Accelerator Facility-have all been brought on line on time and within budget. DOE also supports the largest federally funded program of research in the physical sciences and is a major part of the national programs in environmental science, life science, and mathematics and computing. The end of the Cold War has not affected the demand for these functions, and there is strong bipartisan support for basic research.
With respect to the energy mission, new technologies are needed to produce energy that is affordable, environmentally acceptable, and secure. The Energy Information Administration projects that in 2015, global carbon emissions will be 50 percent greater than in 1990 and that the Persian Gulf share of world oil exports will rise to over 70 percent, up from less than 40 percent in 1985. The President's Committee of Advisors on Science and Technology, in advising President Clinton on science and technology priorities for the next four years, recently put the development of an energy R&D strategy at the top of their list. In order to reduce the risk of global climate change, reduce projected dependencies on the Persian Gulf, and take advantage of growing worldwide energy markets, they recommended substantial and sustained increases in both private and federal energy R&D, which has declined significantly in real terms over the past two decades.
In the environment mission, DOE's goals are to clean up the DOE nuclear weapons legacy; to stabilize and safely store or dispose of nuclear waste; to deactivate, decontaminate, and decommission surplus facilities; and to remediate the contaminated environment. The cost of the cleanup has been estimated to exceed $300 billion. Independent reviews of the program, including the Galvin task force, have recommended increased laboratory R&D to develop technologies to reduce the cleanup costs.
In summary, we believe that there continues to be a compelling need for R&D in each of DOE's missions. In spite of proposals to dismantle DOE and a highly competitive budget environment, the administration and the Congress, when they have examined the missions of DOE, have supported stable, and in some cases even increased, budgets.
Is DOE using the laboratories appropriately in accomplishing these mission outcomes? The fundamental principle to be followed is that each program should use the best R&D performer, whether at laboratories, universities, or in industry, to accomplish the mission. This principle is the same for a DOE program manager as for a National Science Foundation (NSF) or Advanced Research Projects Agency manager. If DOE's programs are managing their R&D in the most effective way for the taxpayer, then the work placed at the laboratories is well justified. There are two separate questions that need to be addressed: Are DOE's programs using the right mix of industry, universities, and DOE laboratories? Is DOE using the laboratories in an effective way? The first question relates to the overall funding level of the laboratories; the second addresses the role of each laboratory in DOE's mission and how DOE uses the laboratories as a system.
The Laboratory Operations Board recently asked these questions in a review of the management of DOE's R&D programs. Regarding the use of laboratories versus industry and universities, we believe that universities should usually be selected for basic research that can be conducted by individuals or small groups of investigators and that industry should be selected for the development of specific technologies. DOE's laboratories have a clear role in national security R&D and in operating and maintaining large scientific facilities that are beyond the scope of what industry or universities can afford. In addition, they also excel in attacking large-scale complex problems that require multidisciplinary expertise, such as global climate change or the human genome, and mission-focused research that requires results with more urgency than can be anticipated from other R&D performers.
We found that each program uses a different mix of R&D performers and that these differences are generally appropriate given the different goals. At the same time, we found that there are opportunities in several programs to make greater use of universities. In the national security mission, nearly all of the R&D is performed at the national laboratories, as is appropriate for nuclear weapons work. In the science mission, 70 percent of the work is conducted at laboratories and 25 percent goes directly to university performers. This, however, understates DOE's contribution to university research. In addition to the funding that goes directly to universities, over $1 billion of DOE's $2.5-billion science budget goes toward the operation and construction of major scientific user facilities that are maintained at the laboratories for the broader scientific community. These facilities are used by some 15,000 university and industrial researchers each year. In the energy mission, where the goal is to have industry deploy new energy technologies, 58 percent of the R&D is conducted directly with industry, as compared with 37 percent performed at the laboratories. For the environmental quality mission, most of the $6.5 billion goes directly to cleanup contractors. Of the 7 percent of the program that goes to technology development, about 59 percent was spent in the DOE laboratories and production facilities, with the rest being spent on university and private-sector R&D.
Although we believe that these distributions generally are appropriate, several programs have recognized the need to expand the role of universities. The defense program is bringing universities into the science base of its stockpile-stewardship program, and in FY1996 the Office of Environmental Management and the Office of Energy Research launched a $50-million-per-year environmental sciences program that will be evenly split between laboratories and universities.
The second question is whether DOE is using the laboratories effectively. Some programs, such as defense, focus their activities in a few laboratories; others make use of capabilities in many laboratories. There is no uniformly correct answer. On the one hand, it is desirable to focus work at one laboratory to help maintain a critical mass of expertise and to reduce coordination costs. On the other hand, often the technical capabilities and creative ideas needed for a task may be spread among several laboratories. The challenge is to find the right balance between these needs, and where several laboratories are needed, to use the laboratories in an integrated, systematic way.
We found that several programs that, in the judgement of the Laboratory Operations Board, should focus their work in a smaller number of laboratories. The Office of Energy Efficiency and Renewable Energy and the Office of Nonproliferation and National Security are working to focus several of their programs in fewer laboratories. We also found that many programs can do more to use the laboratories as an integrated system. The Office of Defense Programs and the Office of Energy Research are making major efforts in this direction. The board will review progress in these areas in six months.
At the heart of this issue is the quality of the R&D program managers. Strong, technically competent program managers will put together world-leading R&D programs using the best capabilities of the laboratories, industry, and universities. Without this leadership, programs will tend to be collections of individual laboratory initiatives. With the current downsizing, DOE is losing some of its best and most experienced managers. The Laboratory Operations Board will address ways to strengthen the quality of DOE technical leadership over the next six months. One approach may be to use temporary appointments of technical experts from laboratories and universities. Recent legislation gave DOE the same authority as NSF to use temporary appointments. Another approach may be to delegate more management responsibility to the laboratories, where much of the technical management expertise currently exists.
Does DOE have mechanisms in place to ensure the quality of R&D at the laboratories? This is a subtle issue and one that the Laboratory Operations Board is currently examining. It is sometimes asserted that the work of the laboratories suffers because DOE does not use the same peer-reviewed grant selection process used by NSF and the National Institutes of Health (NIH). However, DOE has extensive, if often unrecognized, mechanisms for evaluating the quality of work at the laboratories. These mechanisms include review and advisory committees at the laboratories, advisory committees to the DOE programs, and the extensive use of the National Research Council and JASONs (a group of scientists that conducts reviews and studies for the government under the auspices of MITRE Corporation).
It also is appropriate that the DOE system of ensuring quality differs from that used by NSF and NIH. The peer-reviewed grant selection process as practiced by NSF and NIH is excellent at evaluating the scientific merit of proposals, but because "peers" typically are selected for their scientific expertise rather than their broader expertise about the agency's mission, it is less effective for evaluating relevance to mission and overall system effectiveness. Moreover, a number of critics of peer review have noted that it can exclude the risky, unconventional work that often leads to scientific breakthroughs.
We believe that the overall quality of work at the laboratories is high, as demonstrated by numerous awards and some comparative analyses. Scientists working at the laboratories or using laboratory facilities have won 34 Nobel Prizes, and the laboratories have won 415 prestigious "R&D 100" awards, which is twice the number awarded to all other federal agencies combined. The issue for DOE is not whether there is a merit review system, but the quality of the reviews and how they are used by DOE. This is the focus of the Laboratory Operations Board's present review.
Is DOE managing the laboratories in an efficient manner? The concern emphasized most in the Galvin report is the efficiency of DOE's management system. The Galvin report focused on the myriad rules and regulations and the micromanagement that DOE, often following congressional direction, has imposed on the laboratories.
The Galvin task force recommended a new operating mode for the laboratories by proposing to put them in a government-owned corporation governed by an appointed board of directors. Funding for this corporation would be provided by Congress in four or five blocks, one for each mission area, with no congressional direction allowed. Although we are sympathetic with the intent of this proposal, it is not clear that it is feasible, desirable, or appropriate to remove the Congress so completely from funding decisions. There are many areas where the trade-offs between research fields rest on judgments made more properly through political rather than technocratic processes. Balancing funding between nuclear power and renewable energy, between the human genome and high-energy physics, or between weapons design and arms control verification primarily involves national policy judgments rather than technical judgments. These decisions will have more legitimacy if made by elected officials than by an appointed board. Not surprisingly, few members of Congress have supported the Galvin task force's proposal for a government-owned corporation supported by block funding.
The Galvin task force recognized the uncertainty of this proposal's being implemented, and identified as an alternative a long list of changes that should be made under the current system of governance. This is the direction DOE has followed. In the past two years, under the review of the Laboratory Operations Board, DOE has undertaken a series of efforts to reduce the burden put on the laboratories.
DOE has revised its procurement procedures. DOE's M&O[Ed: Spell out M&O] contractors had previously been expected to conform to federal purchasing principles and practices. In 1995, DOE allowed contractors to use best commercial practices instead, which greatly improved efficiency.
DOE has streamlined its oversight of business practices at the laboratories. During the first year of these reforms, DOE reduced the number of business practice reviews from 343 to 21, person-days of effort from 28,547 to 9,472, and costs from $10.2 million to $2.8 million.
DOE has committed to move from a system of self-regulation to a system of external regulation. Since the creation of the Atomic Energy Commission, DOE and its predecessor agencies have been responsible for regulating environment, health, and safety matters at the laboratories. In December 1996, DOE announced that it would propose legislation to turn regulation of worker health and safety over to the Occupational Safety and Health Administration and regulation of its nuclear facilities over to the Nuclear Regulatory Commission. It is not clear if this change will result in savings, but it is likely to be reassuring to the public that DOE's facilities follow the same regulations and have the same public accountability as other organizations.
DOE is improving its own internal methods of environment, safety, and health management. A set of generally accepted standards will be tailored to each facility at a laboratory site, rather than imposing the most restrictive standards required for a specific facility to the whole site. This process will result in reduced administrative oversight and its associated costs, as well as better focus on areas needing special attention.
DOE has also streamlined its directives and orders, which are the means by which DOE establishes requirements and guidance for work by employees of DOE and its contractors. During 1995 and 1996, DOE reduced the number of orders by 50 percent (from 312 to 156) and revised the 100 most burdensome orders into more user-friendly documents. Much of DOE's regulatory culture remains, however, and senior management will have to be vigilant to ensure that the physical reduction in requirements actually results in reduced compliance burdens and greater efficiency.
DOE has reformed its process for allowing other federal agencies to use its laboratories and is in the process of achieving similar improvements for nonfederal customers. This involves delegating most elements of the approval process to the laboratories; having standard, pre-approved contract documents; and implementing a parallel review process to speed review by all elements concerned. DOE has also been reducing its staff, which will help to reduce the oversight burden on the laboratories. DOE set a goal in FY 1995 of reducing staff by 27 percent by FY 2000; it is currently significantly ahead of schedule in meeting these reductions
DOE is developing performance-based contracts for its laboratories. Instead of compelling laboratory behavior through regulations and directives, DOE is shifting toward specifying performance criteria and fees for performance in the contract. This approach gives laboratory directors more ability and responsibility to achieve results.
Feedback from the laboratories indicates that progress in some areas, such as the reform of procurement and oversight of business practices, is substantial. In other areas, such as environment, safety, and health management, although pilot projects have been successful, changes have not yet resulted in a significant reduction of work throughout the system. Management reforms must continue vigorously in these areas.
In the aggregate, however, these actions are having an effect. In part due to changes or promised changes described above, the DOE laboratories have eliminated unnecessary administrative functions and reengineered processes to cut costs. As of December 1996, cost cutting at the DOE laboratories was expected to result in productivity improvements worth over $2 billion over a 5-year period. This corresponds to about 5 percent of the laboratory budget over that period. Between 1994 and 1996, the number of support personnel dropped by 3,600 full-time equivalents, while the number of research personnel dropped by 1,100. The ratio of research personnel to support personnel increased from 1.46:1 to 1.65:1.
More needs to be done. The recent reviews highlighted the fact that there are inefficiencies in the operation of DOE due to its complicated management structure, involving both headquarters and field elements. DOE has large field offices around the country that oversee DOE's contractors. Although this structure may have been appropriate before technological advances made long-distance communication much simpler, it is not clear that this structure is desirable today. There is a clear need to rationalize the headquarters and field elements of DOE, resulting in further streamlining of functions and organizations. This will be a major priority for DOE over the next year. The reviews also revealed that many unnecessary reports are being required of the laboratories. These will be stopped.
Is DOE making sound decisions about capabilities and infrastructure for the laboratories? Is there a sound process for making sure that the laboratories have the right capabilities needed for the future and that unnecessary capabilities are divested or closed? Individual programs have plans for the capabilities they need at the laboratories, and there is an institutional planning process for the laboratories. We found, however, that the program plans are uneven and often are not adequate bases for planning. We also found that the institutional planning is not as effective as it needs to be. For example, DOE currently does not have a long-term plan for major scientific facilities. It will now develop one. The Laboratory Operations Board will review the overall planning process to determine how to make it more effective.
Should any of the multiprogram laboratories be closed or consolidated? We believe not, at present. Each multiprogram laboratory plays a key role in at least one mission. As program missions and funding change and as programs focus their work in fewer laboratories, this could change in the future. Laboratories should be closed when they fall below the critical mass of competency and are no longer competitive performers in DOE's research.
There is little obvious benefit to consolidation for its own sake It is not clear that a few very large laboratories are more efficient than a larger number of medium-sized laboratories. To the contrary, larger laboratories often have higher overhead rates. In addition, there are high costs for moving facilities and even higher costs for reconstituting a closed laboratory if it is determined to be needed in the future. President Clinton, in his statement of September 25, 1995, noted that "While it would be easy to destroy premier federal laboratories through severe budget cuts or senseless closures, that is not a path this administration will follow."
As laboratories shrink because of changing mission needs, however, they become candidates for privatization or closure. In the past year, DOE has taken several actions to consolidate or privatize laboratories. For example, it consolidated the management of the Morgantown and Pittsburgh Energy Technology Centers and privatized the National Institute for Petroleum and Energy Research. This year, the Laboratory Operations Board is reviewing DOE's small, mission-specific laboratories to validate their roles or determine whether they are candidates for privatization, alternative contracting mechanisms, or closure.
We believe that the process described above is making unprecedented progress in improving DOE's management of the laboratories. There is, however, much left to be done, andit will take several years of sustained effort to carry these efforts to fruition. The following are key elements of the agenda ahead.
Continue to reduce management burdens on the laboratories. The reforms that have begun must be fully implemented, and streamlining and reengineering of DOE's management processes must continue. An important part of this process will be, as described above, to rationalize the work of DOE's headquarters and field structures. Another important element is to continue to reform contracts with the laboratories to achieve good performance through performance metrics and incentives and the leadership of laboratory management rather than through regulations and oversight.
Strengthen DOE R&D management. Some programs need to be focused among few R&D performers, and several need to use multiple laboratories as well as industry and universities in a more coordinated and integrated way. In addition, many areas need better long-term program plans and technology roadmaps. DOE also needs better long-range plans for institutional investments at the laboratories, especially scientific user facilities. The foundation of R&D program management is the quality of the managers, and in some areas this needs to be strengthened; and attention must be paid to areas in which downsizing is hurting quality.
Improve the integration of laboratory work with universities, industry, and other government agencies. Although the laboratories do about $1 billion worth of work for non-DOE customers and several hundred million dollars' worth of cooperative R&D with industry, there are opportunities to continue to improve these relationships. DOE must continue to remove the barriers, some of which are self-imposed and some of which result from arcane government accounting practices, to use of the laboratories by other government agencies and industry.
We believe that this process, which is deliberate and time consuming, is the correct, responsible approach to reforming the management of the laboratories. Many of the problems have resulted from the fact that DOE managers' attention has been focused primarily on budget issues rather than on issues that affect the management of the laboratories and their contribution to DOE missions. The Laboratory Operations Board has succeeded in focusing the attention of DOE's senior managers on these issues. We are aware that to some this approach may seem too modest. Many people are properly skeptical, because many studies over the years have identified similar problems but failed to generate adequate action. We believe, however, that our approach is the first that is demonstrably working. If this agenda is sustained, we are confident that the laboratories will once again be viewed as cost-effective, well managed, and essential elements of the nation's R&D enterprise.
As the Galvin task force noted, "The laboratories' research role is a part of an essential, fundamental cornerstone for continuing leadership by the United States." It is our obligation to future generations to manage these laboratories prudently to ensure that their capacity for scientific and technological leadership is maintained and that our investment in the future is well spent.
Robert W. Galvin, "Forging a World-Class Future for the National Laboratories," Issues in Science and Technology, Fall 1995.
Report of the Task Force on Alternative Futures for Department of Energy National Laboratories (Galvin Task Force). Washington, D.C.: Secretary of Energy Advisory Board, February 1995. (Available at http://www.doe.gov/seab/seab.html)
Steven H. Schiff, "Future Missions for the National Laboratories," Issues in Science and Technology, Fall 1995.
Strategic Laboratory Missions Plan-Phase 1. Washington, D.C.: U.S. Department of Energy, Laboratory Operations Board, July 1996. (Available at http://www.doe.gov/seab/seab.html)
Charles B. Curtis is the deputy secretary of energy and is the chairman of DOE's Laboratory Operations Board. John P. McTague is vice president for technical affairs of the Ford Motor Company and is the vice-chair of the Laboratory Operations Board. He was formerly the deputy director of the Office of Science and Technology Policy and the acting science advisor to President Reagan.