Recruitment, Retention, and Utilization of Federal Scientists and Engineers (1990)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

APPENDIX C WORKSHOP ON RECRUITMENT, RETENTION, AND UTILIZATION OF FEDERAL SCIENTISTS AND ENGINEERS Agenda Participants Proceedings 143 145 147 149

Agenda Workshop on the Recruitment, Retention, and Utilization of Federal Scientists and Engineers National Academy of Sciences Washington, D.C. February 23, 1990 8:00 Continental breakfast 8:30 Opening Remarks and Findings of Previous Studies* Alan K Campbell, chairman, Committee on Scientists and Engineers in the Federal Govemment 9:30 Quantitative Inputs to Federal Technical Personnel Management* Charles Falk retired director, Science Resources Shies, National Science Foundation 10:30 Break 10:45 Organizational and Decision-Making Processes Affecting Recruitment, Retention, and Utilization . 11:30 Lunch, NAS Refectory The Role of OPM in Meeting Federal Work Force Needs with Regard to Scientists and Engineers* John M. Paiguta, deputy director, Office of Policy and Evaluation, U.S. Merit Systems Protection Board 1:00 Differences in Recruitment, Retention, and Utilization Processes: A Comparison of Traditionally Operated Federal Laboratories, GOCOs, and Demonstration Projects* Sheldon B. Clark senior research scientist, Labor and Policy Studies Program, Oak Ridge Associated Universities i:45 The Effects of the Political Appointment Process on Recruitment and Retention of Scientists and Engineers* lames P. Puffer, professor of government and politics, George Mason University 2:30 Break 3:15 General Discussion: Changes in organizational and decision-making processes that might improve the recruitment, retention, and utilization of scientists and engineers Alan K Campbell 5:00 Adjournment *Discussion will follow each 15-~runute presentation. 145

Participants Workshop on the Recruitment, Retention, and Utilization of Federal Scientists and Engineers National Academy of Sciences-Washington, D.C. - February 23, 1990 Federal Agencies U.S. Department of Agriculture: Essex Finney, acting director, Beltsville Area Agricultural Research Service Department of Commerce: John Lyons, director designate, National Institute of Standards and Technology Robert Mahler, deputy director, Environmental Research Laboratories, National Oceanic and Atmospheric Administration Department of Defense: Larry Lacy, head, Civilian Personnel Policy and Requirements Susan Numrich, Office of Strategic Planning, Naval Research Laboratory William B. Porter, technical director, Naval Weapons Center Karl Steinbach, chief scientist, U.S. Army Belvoir Research and Development Center Department of Energy: Norman L. Howton, director, Organization and Personnel Division, Morgantown Energy Technology Center Mary Parramore, assistant to the director, Argonne National Laboratory Department of Health and Human Services: Phillip Chen, associate director for intramural affairs, National Institutes of Health James Eagen, director, Public Health Service Mary Guinan, assistant director for science, Centers for Disease Control Sharon Holston, Food and Drug Administration Department of the Interior: Stephen E. Ragone, acting assistant director for research, U.S. Geological Survey Department of Transportation: Brian Andrews, supervisory computer scientist, Federal Aviation Administration Nancy Mowry, supervisory personnel management specialist Environmental Protection Agency: Clarence Hardy, deputy director, Office of Human Resources Management National Aeronautics and Space Administration: Dale Compton, acting director, Ames Research Center Robert E. Sutherland, manager, Human Resources Division, Jet Propulsion Laboratory 147

Merit Systems Protection Board: Paul van Rijn, research psychologist National Science Foundation: Margaret Grucza, director, Governanent Studies Group Margaret Windus, head, Division of Personnel Management Office of Personnel Management: Jack Curnow, chief, Statistical Analysis and Service Division Leonard Klein, acting associate director, Career Entry and Employee Development Group Paul Thompson, project head, Research and Demonstration Division Dona Wolf, director of policy Carnegie Commission on Science, Technology, and Government Jesse Ausube] David Robinson NAS/NRC/IOM William D. Carey, chairman, Office of Scientific and Engineering Personnel (OSEP) Claudia Dissel, associate executive director, OSEP Alan Fechter, executive director, OSEP Michael Finn, director of studies and surveys, OSEP Steve Merrill, director, Office of Government Affairs Lawrence E. McCray, executive director, Committee on Science, Engineering, and Public Policy Others Mark Abramson, executive director, Council for Excellence in Government Barbara Bailar, executive director, American Statistical Association Sandra Fiske, director, Federal Government Service Task Force Ray Kline, president, National Academy for Public Administration L. Bruce Laingen, executive director, National Commission on the Public Service 148

Proceedings Workshop on the Recruitment, Retention, and Utilization of Federal Scientists and Engineers National Academy of Sciences Washington, D.C. February 23, 1990 Opening Remarks and Review of Previous Studies Dr. CAMPBELL opened the session by explaining that the task of the Committee on Scientists and Engineers in the Federal Government assigned to the Office of Scientific and Engineering Personnel by the Carnegie Commission that is examining major issues concerning scientists, engineers, and technology and the general capability of the United States in those areas- is to look at the issues of recruitment, retention, and utilization of scientists and engineers as they relate to the federal government with an emphasis on organizational and procedural mechanisms. The purpose of this workshop is for the Committee to gather as much information as it can in preparation of a report due to the Carnegie Commission in early summer. He noted that the commissioned papers reveal a lack of consensus in specific matters- for example, evidence that scientists and engineers are overpaid or that they are substantially underpaid in the federal government. The basic issue relates to inadequate disaggregation of the data. Among other matters of interest are the following: "inflexibility" of the civil service system and experimentation in the area of personnel management in the federal government, particularly contrasts between those labs directly operated by the federal government and those to which services are contracted; ethics, an area of counter forces: one dealing with issues that relate to people taking advantage of their federal positions and another advocating enough flexibility so that regulations and laws don't discourage prospective employees; differences in pay based on geographic location (Dr. CAMPBELL noted that, in the private sector, much of this is handled by up-front payments such as bonus sign-one and special bonus opportunities rather than trying to deal with base pay issues); ownership of intellectual property; the quality of the working environment, which varies substantially from lab to lab; and the increase in the number of noncareer appointees in the federal government and their effects on the quality of people that are attracted to federal service and institutional memory. Quantitative Inputs to Federal Technical Personnel Management Dr. FALK opened by asserting, 'Jo say that effective management requires good information input is almost a truism. Nevertheless, that one facet of management frequently does not get the type of attention it should." His presentation was divided 149

into three parts: (~) a description of the types of data needed and sources of relevant data, (2) an analysis as to the extent to which the data are being used, and (3) findings from the data. He divided the relevant data into two broad categories, descriptive and dynamic. Descriptive data provide information on the characteristics of the work force at a given point in time: demography (age, sex, and race); education (highest degree); work activity (occupation, type of work, its relevance to special national interests such as defense or environment, and salary); and quality of the work force. Dynamic data are important to assess what is happening to the work force. providing insights about whv ----rid err , ~ _--- force, providing Insights about why people leave, whether those who stay are satisfied with their jobs, and what is likely to happen in the future. Do Needed Data Exist? The Office of Personnel Management (OPM) regularly gets data from federal agencies about characteristics of their personnel; excepted agencies are the U.S. Postal Service, security agencies such as the National Security Council and the Central Intelligence Agency, and the uniformed services. Data from the Tennessee Valley Authority is not reported regularly to OPM but at certain intervals. OPM's Central Personnel Data File (CPDF) includes most of the descriptive elements and some dynamic ones, primarily those related to acquisitions and separations of personnel, as well as administrative data such as individual pay schedules. The data are compiled quarterly but published only every two years in Occupation of Federal White and Blue Collar Workers, which essentially presents descriptive data (mostly on sex, salary, occupation, and grade) but only in aggregated form by occupational codes. Annual aggregated data are published by the National Science Foundation (NSF) by compiling the OPM data into occupational groups and subgroups-for example, the physical sciences in the sciences and mechanical engineering or electrical engineering in the engineering groups. NSF provides data in 13 different types of tables in its annual Federal Scientists and Engineers. In addition, NSF provides data on the whole national science and engineering labor force, including the federal sector, facilitating comparisons across sectors of employment. For sectors other than the federal government, the NSF data are obtained from the individual 5.5 million scientists and engineers themselves, on a sampling basis. NSF also obtains data from recent graduates, baccalaureates and master's recipients, about a year after they graduate; and the National Research Council maintains a data base on the entire U.S. doctorate population, based on responses to the annual Survey of Earned Doctorates. The NSF numbers and the OPM data vary considerably because different definitions for "scientist" and "engineer" are used by the two organizations. OPM reports data on individuals who are classified in a science or engineering occupation. Based on individual responses to its surveys, NSF classifies as scientists and engineers those who meet two out of three criteria: (~) a degree in a science or engineering field, (2) a job in a science or engineering occupation, and (3) self-cIassification. As a result, NSF data include individuals who are not necessarily working in a science or engineering occupation at the time surveyed for example, an engineer who is now in a nonengineering job but who considers himself or herself an engineer and holds an engineering degree. Discrepancies between OPM and NSF data are inevitable because 150

they come from different sources: OPM data from employers and NSF data from individuals. Two agencies in Washington, D.C., regularly provide projections about the entire scientific and engineering work force: the Bureau of Labor Statistics (BES) and the NSF. A number of BES publications deal with projections, one being the Occupational Outlook Handbook NSF publishes projections of supply and demand both for doctoral scientists and engineers and for all scientists and engineers, showing what would happen · ~ unc er various scenarios. OPM data are primarily used by the central management agencies of the government-OMB and OPM, for example but the dynamic data are not published by OPM. Under those conditions, do the agencies ask for compilations to see how they compare to other agencies? According to OPM staff, this seldom occurs. Therefore, one wonders whether the agencies themselves will use the data that they provide to OPM to analyze their own performance. Findings . . . The Department of Defense (DoD) is the largest employer of federal scientists and engineers, about 50 percent of them, followed by the Department of Agriculture (USDA, 13 percent) and the Department of the Interior (7 percent). In essentially all major occupational subgroups, DoD is the largest employer with two exceptions: USDA is the largest employer of life scientists and the Veterans Administration employs the most psychologists. Only 25 percent of all federal scientists and engineers are actually engaged in R&D, with the rest employed in design, natural resource operation, management, and a host of other occupations. In the federal government, 58 percent of the doctorates in science are in R&D, in contrast to the national Ph.D. total of about one-third. Not all scientists and engineers working on a particular mission are employed in one agency. For example, although the Environmental Protection Agency (EPA) is only the seventh largest employer of scientists and engineers in the federal government, environmental issues are the second largest focus of work activities of federal scientists and engineers. An examination of scientists and engineers by their age group reveals that only 14 percent of federal scientists and engineers fall in the "over 55" group, as compared to 19 percent in the total U.S. work force. Information from a survey by the U.S. Merit Systems Protection Board (MSPB) indicates that members of the Senior Executive Service (SES) leave the government for four main reasons: ceilings on salary, changes in the retirement system, politicization of their organization, and inadequate use of their knowledge and skills. Because salary data are usually presented in terms of median or average salaries, comparisons assume that the distributions in each sector are identical with respect to age, occupation, and highest academic degree, but they are not. 151 a,

Recommendations An examination should be made to determine when data available from OPM and other places are being utilized by the individuals agencies in the management of their science and engineering work forces. There ought to be an evaluation as to whether the dynamic data, those on separations and hires collected by OPM, should be more widely available in published form, at least to other federal agencies. Data on satisfaction or dissatisfaction in the job, as well as on the quality of the federal science and engineering work force similar to that recently compiled by the DoD~hould be generated on a regular periodic basis and made publicly available. Agencies should use projections about the future supply of scientists and engineers in the total U.S. work force to make projections of their own future demand for scientists and engineers. There should be a central body in the federal government that explicitly has the sole responsibility to evaluate the adequacy of the federal science and engineering work force. Discussion Dr. CAMPBELL said that this is a fascinating issue because aggregate turnover numbers are very low: in much of the private sector, turnover rates of 2 percent would be considered too low. He asked the agency representatives whether there would be ways to make the aggregate data more useful to them, or whether the current system of an agency building its own data base for its own management purposes (with the aggregated base being used essentially for broader policy questions) is more preferable. He noted that the Committee will address whether there should be changes in the nature of the federal data-collecting system and the character of what is being collected. Reliability of the Data Inconsistent Definitions: Dr. NORWOOD felt that the basic question really should not be 'why are the data not being used?" but rather 'Mow good are the data?" Having responsibility for the measurement of occupational employment at all levels, BES examined salary data for engineers-learning that those engineers who earn top salaries are managers of engineering work; thus, in the standard occupational classification structure, they are considered to be managers. Yet they are still engineers: they are trained as engineers and many of them are still doing engineering research, but they also are managing. Similarly, she noted that many SES people have to be classified as managers: they manage groups and cannot be classified as experts in a particular area if they are going to reach the higher salary classifications. One also must take into account differences in the people and their skills. At BES, for instance, there are statisticians and mathematical statisticians, each having different kinds of training. For example, although BLS employs three kinds of scientists, data on them is combined with data on scientists in other parts of the Department of 152

Labor (DOL); thus, the picture drawn from OPM data alight reflect what is happening in DOL overall but totally miss what is happening in BES. Other problems arise when terms such as "turnover" are not clearly defined. For example, BES may have a 16-17 percent turnover rate, but if BES employees merely move to other parts of the DOL, the turnover rate for DOL is not affected. Mr. RAGONE commented on the 4 percent turnover rate cited in OPM data for hydrologists: of the 2,170 federally employed hydrologists, less than 200 work in the research program of the U.S. Geological Survev's Water Resources Division: thus. the aggregated data do not reflect the turnover rate in USGS. In addition, he noted that all employees in the Water Resources Division are considered to be hydrologists, even though their degrees may be in different fields, such as geochemistry. The nature of the discussion pointed out that (~) dealing with the personnel system in the federal government is a very complex undertaking and (2) a main underlying problem is definitions and coverages- for instance, whether turnover just involves quits, includes separations, or recognizes transfers and whether the data deal with permanent staff, temporary staff, or seasonal employment. In the aggregate, one cannot deal with the multitude and myriad of problems that are unique to particular areas. That is why, according to Mr. CURNOW, the agencies themselves should conduct self-studies of turnover, separations, and quits. Using its aggregate data base, OPM looks at organizational supplements but not at program offices in any particular area. Furthermore, because the occupational classification that the government uses stays with an individual if he or she moves into a managerial or supervisory position, another definitional problem exists with respect to how managers are classified in the public sector as opposed to the federal sector. Comparability of Data Sets: Mr. FECHTER said that information collected on individuals is good for tracking career trajectories. However, firm-based data, which is based on occupational classes, will not provide very reliable information about movement of individuals. Instead the issues of promotion and of moving into management, which are part of the dynamics of the system, need to be looked at very carefully. Dr. VAN REIN advocated the use of aggregated data available from the CPDF because various agencies have their own data bases, which may be incompatible and based on different definitions for key terminology. Using CPDF data, researchers can present the turnover rate for all occupations and show variations by length of service and from agency to agency. Although agency-specific data will always be more sophisticated and more detailed, based on the agency's purposes, aggregate data are necessary to determine the baseline. He pointed out that once definitional problems are resolved, agency data appear more similar to CPDF data. In fact, people in the agencies are using the aggregate data to stimulate their own internal research and definition of the problem of turnover. Mr. RAGONE cited the danger of using national statistics, which do not reflect what is happening in significant but small parts of the work force. For instance, USGS has difficulty recruiting individuals at GS-12-15 levels for its national research program: it cannot compete with private industry at those levels. He emphasized that national statistics do not reflect the situation adequately; one must look at the agencies, their specific kinds of expertise, and what they are trying to accomplish. 153

Dr. MAHLER cautioned against comparing statistics on the federal work force with those dealing with the national work force because the government, in most cases, cannot hire foreign nationals. He also noted that the more widely available data do not reflect that an agency may employ more scientists and engineers as contractors who are paid on a different scale and do different work than the in-house scientists and engineers. Usefulness of the Data Dr. NORWOOD felt the bigger issue is identifying what statistics are really needed and the extent to which users of the data provide feedback to those who generate them. Mr. HOWTON responded that agencies don't use aggregated OPM data because those data have been massaged so as to be essentially useless at the field operations level. He agreed with others' comments that the federal data collection system does not take into account that private industry moves people into managerial titles out of science and engineering: they cease being engineers about the time they become branch managers, and all of a sudden their salaries are no longer reflected in the data base that OPM or BES uses for its salary curves. However, Dr. AMBLER felt that the collected data are useful to a laboratory director in recruiting scientists and engineers by providing the reasons for leaving and the comparability of pay broken down by different ages and employment sectors. At least three federal organizations Naval Research Lab (NRL), DoD, and Centers for Disease Control (CDC) have designed their own data collection systems to learn more about their own personnel. Dr. NUMRTCH noted that NRL maintains two personnel data bases because it cannot make sufficient projections from the data that it sends to OPM. NRL keeps more detailed, exhaustive information to deal with its own problems, retaining items that have been dropped from the OPM data base. Data disaggregated by NRL have enabled the agency to compare its i,800 scientists and engineers, more than 50 percent having Ph.D.s, with comparable agencies on a quartile basis, using maturity curves. Recruitment and retention data show that NRL's declination rate for firm offers increased exponentially during the last three years. Particularly disappointing is the inability to retain postdocs in the face of outside salary offers of as much as $55,000 (NRL can offer between $35,825 and $46,571 for postdocs entering at the GS-12 level). One solution is to use R&D contracts under which people perform research tasks at the laboratory, but that alternative has two negative results: (~) contractors working next to lower-paid government employees doing the same type of job and (2) government scientists spending their time managing contracts, which is inefficient use of their skills. Mr. LACY said that DoD also has created its own central data base, which covers every civilian and military employee of the department and enables one to distinguish "quits" (departures from DoD) from other types of personnel actions (transfers from the Army to the Navy, geographic relocations, and changes in one's occupation). The actual quit rate for scientists and engineers is quite low, whereas "overall employee turbulence" is much higher, reflecting the large number of people changing jobs within DoD each year. Dr. GUINAN said that because the CDC uses two personnel systems, aggregate data is really meaningless. At least half of CDC's research scientists are members of the 154

commissioned corps of the Public Health Service and are eligible for bonus pay not available to veterinarians and physicians whom CDC hires. Because of the way the system works, the aggregated data do not reflect this phenomenon. Dr. GUINAN posed another problem: unless the original purpose of OPM's data collecting was to help recruit and retain new employees, it is both useless and invalid to use the collected data for that purpose. Only the agencies themselves can define both what their needs are and how a system should be designed to respond to those needs which are so agency-specific, especially in critical areas, that no aggregated data would be useful to a single agency. Dr. GINZBERG asked if new and additional data would help agencies respond to their most urgent problems of trying to manage, attract, retain, and utilize effectively their scientific and engineering personnel. Dr. CAMPBELL doubted that such data would make any difference to internal management but could make a case for improvement in the broad employment system. Dr. NUMRICH responded that agencies would like market-based pay, pay banding instead of automatic promotion steps, and removal of the pay cap. When Dr. GINZBERG asked whether she would trade off some personnel slots to gain these changes while staying within a total financial budget, she responded that, as necessary, NRL's senior management would trade seniority rights and positions; however, the concept of total financial budget is meaningless for Navy labs, which are run under the Naval Industrial Fund. Labs are not funded as line items, but compete for research and development funds which include money for salaries. If their work is not cost-effective, the funder could choose to support the research elsewhere. As a senior computer scientist, Mr. ANDREWS views himself as a "trench warfare scientist" who writes major contracts and is involved in computer development on highly sophisticated, highly reliable systems. He questioned the usefulness of data on new engineers and new scientists: his experience is that it takes about five years for new employees to learn the culture of the Federal Aviation Administration in order to be productive. Dr. AMBLER cautioned against statements that the aggregated data are useless. The larger issue is the dynamics of the organization that make it what it is. He felt the main point is that the data are useless in hiring scientists and engineers at the agency level. Mr. CURNOW noted the differences between micro- and macrostatistics. OPM data cover all scientists and engineers, but the problems cited by workshop participants deal with a small subset, those engaged in R&D in labs. Dr. GUINAN concurred that information on who is doing R&D and statistics on that particular group of scientists and engineers is more germane than knowing what all scientists and engineers are doing in a particular area of research. Quality of the Work Force: Several workshop participants emphasized that the most useful data would be those that indicate the quality of the federal scientific and engineering work force. Dr. MAHLER was concerned about the quality of the future work force because (~) the best undergraduates are foreign nationals who cannot be hired and (2) in spite of changing demographics, agencies are not addressing the people coming up in the system, concentrating their efforts only on the averages. Dr. FALK concurred, noting that 50 percent of engineering Ph.D.s graduating in the United States 155

are foreign, but most government agencies cannot hire them for security clearance purposes. Dr. ROBINSON stated that improving the quality of the federal science and engineering work force is probably the most important issue for improving the way the government deals with science and technology, and Mr. PALGUTA concurred but added that, in spite of this, we have found no acceptable way to measure quality. Dr. FALK responded that the educational institution attended provides some indicator of the quality at the entry-level, but determining quality at m~d-level is difficult. Based on studies conducted by the National Academy of Public Administration, Dr. KLINE said that most National Aeronautics and Space Administration (NASA) personnel officers believe they are not having great difficulties in hiring engineers of the quality they have had in years past: average entry-level engineers have GPAs around 3.2, just as they did during the height of the Apollo program. Federal agencies have seldom attracted people with GPAs in the 3.6-4.0 range, although people use numbers like those to show declining quality. Neither Dr. STEINBACH nor Dr. NORWOOD considered GPAs to be a very good measure of quality without reference to the major area of studies and the school. Negative Impacts: Dr. COMPTON said that data not only are not useful at the field laboratory level, but they are used against agencies. For instance, low turnover rates are used to prove that the federal government does not have an attrition problem: whenever an agency asks for something special, it must turn in turnover numbers to prove that it has a problem. He felt that although turnover numbers are low, a serious problem exists, particularly after 5-7 years of employment. Of 50 new scientists and engineers hired by Ames Research Center a year ago, Dr. COMPTON projected that 30-40 percent of them will leave within the first five Years. ~ ,, Mr. HARDY said the EPA has problems with the data because they presume more stability and less diversity than what the agency is experiencing. He noted that in implementing EPA's Superfund program, the aggregated data were used to keep EPA from getting the salary flexibilities and other requisites for recruiting and retaining needed people. In some cases, the data need to be disaggregated by agency, by individual agency within a larger group, and by individual programs in the agency. Timeliness of the Data Dr. NORWOOD noted that obtaining personnel data from the federal government is the most difficult problem of BLS, which gets good response from industry to surveys whose data are incorporated into its standard occupational classification. Coordination of Data-Collection Efforts Dr. GINZBERG, who agreed with the view that no office in the federal government has overview responsibilities for scientists and engineers, voiced concern with the weaknesses and quality of the data systems. He asked the agencies if a lack of data ever affected the planning of their operations and whether any particular kind of data was needed "to improve their management of scientists and engineers." 156

Conclusions Mr. CAREY said that some of these points of "information" have secondary and tertiary effects throughout the decision system. The data need to be examined much further, not to question their quantitative accuracy but to see whether they reflect positively or negatively on government's ability to utilize and manage its very critical human resource. If government retains only 25 percent of its scientists and engineers in actual R&D, how does that compare with other sectors that are heavy users for example, manufacturing or some of the Federally Funded Research and Development Centers (FFRDCs)? He felt that if a number like that is isolated in terms of the way government behaves, it could suggest things that are either true or false relative to the productivity and utilization of scientists and engineers; such data just cannot be left hanging. Before trying to make a case for higher compensation for scientists and engineers, isolating that population of specialists away from the general population of career people in other government occupations we midst hefter ~ln~er.ct~nr1 the `1~t~ points. Mr. PALGUTA mentioned the difficulty of getting Congress to look at even the aggregated data on federal civil service issues, let alone disaggregated data. He felt the most appropriate use of disaggregated data (e.g., data unique to an agency or agency subunit) is at the agency level. He suggested that the best solution is to allow greater discretion for each individual agency to tailor its personnel system to any unique needs identified in the data. This may require convincing Congress of the need for major changes in the civil service system to allow that discretion. Reminding participants of the Committee's charge, Dr. CAMPBELL asked if there is a sufficient set of problems and difficulties with scientists and engineers to rec- ommend the establishment, some place in the government, of a collection of data that is more disaggregated and thereby more usable for management. He wondered whether there is something so unique about scientists and engineers that there should be recom- mended a special set of institutions to deal with both data collection and policy recom- mendation in that area. Dr. NORWOOD made two points: (~) at least 100 other areas would request that same thing, and (2) before one can argue that the federal govern- ment has to do something, one must deal with the broader problem of a potential short- age of scientists and engineers in all sectors of the work force (although it is true that the private sector is much more flexible than the federal government). Dr. MESSNER agreed but noted that the federal government has a problem beyond the general condition. Meeting Federal Work Force Needs With Regard to Scientists and Engineers: The Role of the U.S. Office of Personnel Management Mr. PAL~GUTA explained that the MSPB is interested in recruiting and retaining not just scientists and engineers but the federal work force as a whole. As part of the civil service Reform Act enacted in 1978, the MSPB was established to hear and adjudicate appeals and to conduct periodic studies of the health of the federal civil service system. Scientists and engineers are an extremely vital and necessary resource 157

for government to carry out its assigned role of service to the nation, and it is in the national interest to recruit and retain scientists and engineers who are among the most capable. Yet there is growing evidence that federal agencies are not very successful in this regard. Policy-makers have two concerns in this area: (~) the quality of the individuals coming into government service, especially compared with those who are leaving, and (2) whether interventions might reduce any undesirable turnover. As to turnover, MSPB has found that nationally only about 5 percent of engineers and scientists leave the federal government each year, and this is split about evenly between retirements and resignations. Because many federal employees retire as soon as they are eligible. one goal might be to induce some of them to stay a few more years before retirement. Among the 2-3 percent of engineers and scientists who resign each year, Mr. PALGUTA noted, some of them undoubtedly are not well matched to their jobs and, therefore, agencies would not want them to stay. Other resigners are among the "stars" who leave for salaries of $150,000 and higher. Although these are the extremes, it is clear that difficulties in meeting current federal work force needs with regard to scientists and engineers watt not be overcome solely or even primarily by reducing turnover. On the national level, therefore, Mr. PALGUTA felt emphasis should be placed on recruitment interventions. The government simply is not effectively filling its vacant engineering and science positions. Recruitment is a multifaceted function, having three interrelated aspects, each of which is necessary for good recruitment but none of which alone is sufficient: (~) an agency's job opportunities have to be known; (2) the jobs, as well as the agency itself, have to be attractive to potential applicants; and (3) the hiring process must operate in a timely manner and also allow employers to make qualitative distinctions among the candidates for employment. The question is '~hat should the government do to improve its recruitment process?" To answer that question, one needs to understand the changes that have occurred in the environment within which the government recruits. Twenty years ago, the hiring process for federal scientists and engineers was largely centralized in the U.S. civil service Commission. This policy was partly justified by a large influx of applications from individuals who found the image of government as an employer fairly good: we had recently put a man on the moon, the EPA had just been established to improve and protect our environment, world hunger might be conquered through federal agricultural research, and so on. The federal government was viewed as an exciting, interesting place to work and had the reputation of seeking to hire the best and brightest. Federal pay, while still not a match for many private-sector jobs, was closer than it is today. The weak spot 20 years ago was the government's centralized screening process: it could take six months or longer for a person to even have an application acknowledged and to get a numerical rating, and then even longer to be referred to an agency having a vacancy. Further compounding the recruitment process was the "rule of three": after screening the applicants available, the Civil Service Commission referred the three best applications to the agency, but the agency often rejected all three, believing that better applicants were available. The current situation and conditions are much different. Recruitment today rarely involves screening massive numbers of applicants but rather tries to find a 'dive" applicant willing to accept a federal position. Today approximately 75 percent of engineering and science recruiting is done on a decentralized basis, in large part through ~, . 158

authority delegated to individual federal agencies to directly hire any qualified applicant. OPM maintains a list of eligible candidates. Thus the question has changed from how to best screen a large number of applicants to whether the federal government can .. , "~e . . ~ ~ .. .. ^. ~ .. . An. __ _ attract a suttlclent number ot well-qualliled applicants in the first place. We have gone from a centralized process to a decentralized one. We have seen an evolution and a shifting of roles between Congress, OPM, and the agencies, but the changes in government's organizational or procedural mechanisms that should accompany such a shift have not all been made. For instance, because of a traditionally rigid compensation system, salary remains a major issue: OPM's data show that an engineer hired at a GS-5 level in 1986 would be offered $~S,700 versus a private industry offer of $27,000. More revealing is that four years later, in January 1990, those two individuals the one who entered government as a GS-5 and the one who chose private industry employment for the additional $8,000 or $9,000 annual starting salary- end up earning about the same, about $39,000 a year. Government salaries rise more rapidly in the early years, but industry had learned that it could "front-Ioad" salaries to attract employees. The government currently does not have this flexibility. Once someone has entered the work force, however, inertia sets in: as long as an employee is being treated decently and is making an acceptable wage, there is a certain tendency not to shift to other employers. This may account partly for the low turnover rate in engineering and science occupations, as compared with other occupations that are affected by different dynamics. In addition, although the government could take care of any turnover problems by hiring people that nobody else wants, worse problems would obviously arise. The goal should be to recruit a good applicant pool at the outset. The individuals hired may be in government for a long time. With regard to OPM's role in the process, MSPB has looked at delegation and decentralization as part of its review of efforts to simplify personnel management. MSPB supports OPM's efforts to delegate more authority to the agencies to recruit, select, and hire. The agencies, in turn, must take the initiative for developing their own college relations programs to recruit well-qualified employees and accept greater responsibility for the process. To assist in this regard, OPM and Congress must address the issue of salary through legislative change. OPM is aware that it should provide leadership to address not only the pay issue but also the process of recruitment, and it is attempting to do so. OPM should provide the tools for the agencies to use to recruit and retain well-qualified engineers and scientists and, through oversight, assure that they are properly used. OPM need intervene only where there is a problem, and then usually only on an agency-specific basis as opposed to a government-wide level. Finally, although it is not OPM's charter to be a champion of the federal employee per se, OPM is expected to be the champion of effective and efficient human resource management. In the long run, good human resource management involves treating employees and applicants for employment fairly and earning a reputation as a good employer. Discussion Need for More Personnel Workshop participants stressed the need to look at the work activity and maturity curve to make the case for more personnel. Mr. HARDY noted that nothing in the 159

current system will help EPA fly its Superfund problem-total comparability, facilities, equipment, etc. short of some statutory or legislative remedy to help recruit and retain people. Although the recruitment, retention, and pay problems are significant, EPA generally does not experience these problems. However, because the Superfund program is a new activity that uses scientists and engineers in nontraditional ways, turnover is as high as 25 percent in some cases; people leave for higher pay, doubling the average $37,000 a year salary paid by the government. Competition with Other Employment Sectors Mr. ANDREWS said that because scientists and engineers have very marketable skills, the government must be very careful in how they manage them: there is a lot of parallel contractor activity, and contractors are willing to pay much more money for these marketable skills. Mr. HARDY said that in a new and emerging industry, EPA finds itself competing with its own contractors. This led EPA to undertake a self-study, concluding that the agency was creating its own problem and that less work would be done by contract. Ms. MOWRY noted that the Department of Transportation (DOT) cannot compete not only with private industry but also with federal agencies such as National Institutes of Health (NTH) and NASA, which have more visible programs. Thus, DOT is not able to attract the higher level, quality applicant. As a result, the agency has little attrition at the m-level, but instead is stuck sometimes with lower quality employees who really cannot compete in private industry because they lack the necessary talents and background. Compounding the situation is that the department has little training money to advance mediocre applicants. Thus some hope that the proposed defense cutbacks and current hiring freeze will result in gravitation and cross-line transfers from DoD to DOT. Finally, Ms. MOWRY cited the effects of the Gramm-RuUman five-year plans: DOT is analyzing how to operate with less money and still improve programs in all hard-to-fill occupations. When Dr. LYONS expressed concern about the loss of National Institute of Standards and Technology (NIST) employees to academe, Mr. CAREY asked why that movement causes more concern than movement to industry: is there not a value in strengthening academic flow, too? Mr. LYONS replied that the flow used to go the other way; the National Bureau of Standards (NBS, predecessor to NIST) would lure people from academe. The real problem is not that the agency's top scientists and engineers are moving to academe but that the agency is losing them to second- and third-rank schools. Mr. RAG ONE asserted that (~) although the agencies have good theories about how to better recruit, retain, and pay scientists and engineers, the decision on imple- menting these theories is a political one determined by Congress, and (2) the three big research sectors of this country- the federal, private, and university sectors are not really engaged in competition but actually reinforce each other. Other Problems Ms. MOWRY said that OPM has made it a lot easier for agencies to recruit but cited some factors detracting from effective recruitment: 160

(id Most recruiters are untrained, having recruiting as a collateral duty. (2) Most recruiters are much older than the college applicants whom they face at career fairs or on college campuses. (3) Most agencies cannot compete at career fairs with private industry because the monies are tight. Agency recruiters do not have the glitzy displays or the number of recruiters actually needed to do a one- or two-day career fair. (4) There is also much confusion among the different agencies as to what the direct hire authorities are: because they differ agency to agency, applicants become confused. She also said that most college graduates are looking for program stability; salary is not their primary consideration. In fact, many college counselors advise young people to work for the federal government to get the experience of beginning a project and seeing it through for several years. However, the federal government may not continue to offer such stability. Because of cuts to many programs, a lot of employees are being shifted or moved geographically. Mr. KLEIN noted that disillusionment sets in among entry-level scientists and engineers when they find they won't be doing real engineering work but rather preparing to be contract managers because so much federal work is contracted out. This has led to congressional examination of how government work is conducted, to questions on the role of government R&D, and to discussions of what should be done in-house versus what should be done by contractors. Several workshop participants said that scientists and engineers in the government want to do things, not just manage the work of others, although that experience often serves as a means for them to acquire even more market- able skills. As mentioned earlier, this disillusionment is heightened when contractors sitting next to civil service people earn almost twice as much. This as a problem of some importance in terms of the health of federal science and engineering and the R&D establishment: How will federal agencies maintain the intellectual, managerial, and technical strength needed for government to do what will be required of it? Decentralization Dr. NORWOOD observed that although OPM attempted decentralization, the Office of Management and Budget (OMB) advocated centralization. She wondered whether there is any conflict between the decentralization required, since both OPM and other agencies to whom these tasks are decentralized have had their staffs reduced markedly because of OMB directives. Mr. PALGUTA responded that problems do arise because agencies are given more work but their staffs are reduced. He noted, however, that agencies are directed to cut overhead by a given percentage but are free to determine the job classifications that such cuts will affect. Because of decentralization and tighter budgets, individual managers, not personnel staffs, find themselves tasked with greater responsibility for recruitment, which is probably as it should be (although many agencies and agency managers do not recognize that they are being tasked with a greater responsibility). Some activities must be centralized because agencies need basic tools such as a flexible. realistic compensation system. OPM has tried to assist agencies in two areas: 161

(1) Because many students do not know where they might apply and how to find a federal job, in September OPM set up the college hotline: by calling a 900 number and answering a series of questions about one's specialties, degrees, and college(s) attended, a person can learn how to apply for a job and is sent the appropriate federal forms for applying. A prerecorded voice explains the process and the basics of the system. If an individual wants to find out about specific agencies, he or she may call a given number to talk to an actual recruiter. (2) OPM has replaced the FS-171 with an automated form processed within 24 hours at its Macon, Gal, facility. The rest is left up to the individual agencies. For example, if an agency wants to recruit an engineer, it uses a specific code to hook into this automated system; after specifying the series, grade, and specialty wanted, within 13 minutes the agency will receive either a particular application or a referral of all the candidates who qualify for that job. Mr. KLEIN enthusiastically noted that many agencies find this system useful, after expressing disbelief that the federal government had created it. Besides its facility and speed, this system benefits agencies because the referrals allow direct hire of anyone on the list. A disaggregated system could never be imposed on the total federal government, for the resulting paperwork would be mountainous. He felt that uniform definitions are not the solution: some agencies would find that those definitions do not work for them. Furthermore, if the general movement during the past 10-15 years of granting greater authority to the departments and agencies should continue, one must be careful not to impose from the center even data requirements that can influence how management operates. Determining how much flexibility and decentralization are needed in government is very difficult because of the overwhelming pressure of the centralizing site, whether it is the central agencies, the data collectors, Congress, or the General Accounting Office. All of those things tend to force centralization, and maintaining a decentralized operation in that context is very difficult. In spite of Mr. KLEIN's enthusiasm, most workshop participants said they were unaware of this automated system. Hiring Authority Dr. CAMPBELL reinforced the idea that, if there is one major cultural difference between the federal government and the private sector, it is the degree to which the chief personnel officer is the line officer. In the private sector, the personnel people provide assistance but also get involved in every personnel decision because the success of their unit depends on those people who are recruited and given training. It is impor- tant not only that line managers and senior executives go out and recruit but that estab- lished relationships with universities continue between recruiting trips. Dr. NORWOOD thought many agencies have done that. Dr. CAMPBELL added that agencies should be given the authority for hiring on the spot and thus avoid the situation whereby desired individuals have taken employment elsewhere. Many participants were grateful that agencies are allowed to hire college graduates off the campus. As a result of decentralization, 70 percent of all new federal employees are hired through delegation, as opposed to the 14 percent under a more centralized system in 1981; 95 percent of scientists and engineers are employed through direct-hire 162

authority-that is, the agencies find potential employees and hire them. OPM has taken the stance that if individuals are qualified and agencies are satisfied with those qualifications, the agencies should go ahead and hire them. Dr. NORWOOD asked, 'Mow hard is it to get OPM to declare a particular occupation in shortage?" Mr. KLEIN responded that it is becoming easier because there are more shortages. Dr. CAMPBELL was alarmed that the delegation of hiring authority results from a shortage of candidates rather than from a change in philosophy in the central personnel agency. But Mr. PALGUTA explained that the greater delegation of hiring authority during the transition from the civil service Commission to OPM was a planned action, although it was followed by a rescission of those authorities along with much negative feedback from the agencies that had enjoyed them. Redelegation and re-decentralization are supply and demand problems as well as a resource problem, since OPM cannot handle the hiring of government employees by itself. Essentially, all engineers, "hard" scientists, medical specialists, mathematicians, and computer scientists are direct hire. Dr. STEINBACH said that he had been informed that direct hiring authority extends only to engineers and not to physicists or statisticians, but Mr. KLEIN noted that that was false information. Flexibility Dr. GINZBERG asked whether anybody has been clever enough to structure the job classification system to have less trouble hiring at the entry level. Dr. CHEN responded that NIH uses a special authority of the Public Health Service Act, its enabling legislation, to bring in young scientists primarily in the biomedical area without being restricted by the civil service classification system: a tenure-type system similar to that of a university enables NIH to keep these individuals for up to seven years in a temporary appointment. Because salaries are negotiated and set administratively, supervisors feel they have more power at the entry level. At the end of seven years, either these individuals receive tenure, usually as a GS-13, or they leave; about 10 percent achieve tenure. Mr. HOWTON added that agencies have some hiring flexibility at the entry level-based on GPA, one can be hired above step one. However, when a OS-7, step 10, can only be offered only $27,000 by the government but can earn $40,000 in private industry, problems arise. Concurring with others about the importance of initial recruiting, Dr. AMBLER emphasized that Congress, OPM, and the departmental personnel offices should have no role in that process after granting agencies the flexibility, but Mr. PALGUTA concluded that a higher level decision-making, that of Congress and OMB, currently preempts OPM and the agencies from achieving greater flexibility. Both Dr. GINZBERG and Dr. CAMPBELL thought the agencies' most important need is flexibility to respond to the market in real dollars and other benefits. What it would take for an agency to tell Congress that, instead of providing a personnel budget, 100 percent in terms of jobs, it wanted a 5-10 percent conversion possibility of its personnel budget to establish super-grades at the bottom and at the middle levels? The different agencies have different problems, and with a little flexibility on the dollar side and benefit side, they could better handle many of the problems they face. However, Dr. NORWOOD said that the term "agency" should be clearly defined. 163

The Federal Employees Retirement System Many participants felt that the federal retirement system also stymies recruitment and retention. Dr. GUINAN advocated a federal retirement system compatible with that in most universities; it might lure a scientist from academe to contribute five years of very productive time in federal service. Dr. NUMRICH wondered about the effect the Federal Employees Retirement System (FERS) will have on attrition in about 10 years, when m-career people will be on FERS, highly mobile, and facing college tuitions of their children. Mr. PALGUTA responded that the conventional wisdom is that FERS we have a negative impact because the very concept behind FERS was portability. However, he pointed out that in its periodic government-we samplings of 20,000 federal employees, the MSPB consistently finds that federal employees like their jobs, by and large, although they may be dissatisfied about particular conditions of employment such as pay. Once individuals enter government, agencies have a strong ability to keep them if they are given at least a living wage, meaningful work, and good management and if agencies use their skills. Dr. GUINAN felt that most agencies would advocate more flexibility in the retirement system as a mechanism for enhancing their recruitment and retention capabilities, facilitating movement from academe to government and back. Dr. CAMPBELL asked how much difference it would make if the government were able to make contributions to TIAA-CREF when it attracted academics enrolled in the program. Dr. GUINAN responded that it would make a big difference in a small percentage of cases for a director of a center, for example. Such an individual would be willing to go for five years, and the agency would have five years of leadership in a very important position. Dr. WINDUS indicated that NSF does this now but cautioned that, besides TIAA-CREF, some visiting scientists to NSF belong to retirement systems- such as state systems that prohibit contributions by NSF during the individual's tenure at NSF. Dr. ROBINSON asserted that this can be mandated legislatively; some states do it. Dr. WINDUS noted the problem posed by the restriction on the amount that can be paid to visiting scholars, congressionally limited at $95,000. Future Projections Dr. GINZBERG asked whether people from DoD saw any change in the labor market in terms of "the changing outlook for the defense budget." Individuals may not want to get involved in the defense area, which is expected to be downgraded among federal priorities and lead also to less money for the private sector in the defense- aerospace area. He wondered whether the federal government has made any hypothetical adaptations to this changed labor market and asked how quickly in the short run the tightness on the scientific and engineering front can change. He noted that when the NASA and defense budgets went down simultaneously in the early 1970s, both that market and the federal government's position toward it changed rapidly. Dr. STEINBACH responded that DoD has not noticed any changes because it has a hiring freeze. Dr. NORWOOD added that the proposed cutback in defense is really a cutback from levels that have not been achieved, that is, future levels. There- fore, DoD will not really be affected for some years. Supporting that idea, Mr. LACY said that data on successions reveal that 1989 was a banner year for hiring new 164

engineers: DoD added about 7,000 new engineers as civil service employees, compared with about 4,300 in 1988. However, he added that DoD is looking at what effect the hiring freeze we have in the long run: although DoD expects only a limited problem with the retention of scientists and engineers because they turn over so slowly, soon everybody will be doing their own loping because of a shortage of secretaries. Dr. CAMPBELL disagreed, expecting very substantial changes in terms of employment for engineers and physicists both in the broader defense industry as well as in DoD. Mr. PORTER said that because the broader defense industry has begun to react to the proposed cutbacks, the Naval Weapons Center is able to use the advantage of its demonstration program to offer higher salaries than other federal agencies and to still offer jobs, with acceptance rates jumping to 43 percent since 1987. Differences in Recruitment, Retention, and Utilization Processes: A Comparison of Traditionally Operated Federal Laboratories, M&O Facilities, and Demonstration Projects Dr. CLARK discussed some imaginative and different approaches to the management of the federal government's R&D activities, looking at labs directly operated by the government, contracted labs, and some of the civil service demonstration projects undertaken to provide increased flexibility to federal agencies. Most of the Department of Energy (DOE) laboratory systems are contractor operated. The Department of Commerce's NIST and the Navy labs examined are demonstration projects. The Tennessee Valley Authority is special in that it is a federal lab but not subject to the same policies and procedures as most civil service labs. Talking about DOE laboratories as a group is difficult because they are more like the private sector, the types of contractors that run these labs are quite varied, and the personnel policies and procedures that they use also vary significantly. Even though DOE has approval authority of the personnel systems employed in these labs, they still are negotiated with DOE's nine operations offices, each of which has responsibilities for several DOE labs. Recruitment and retention are interrelated, and separating cause from effect is not easy. The organizational factors that may cause a scientist or engineer to leave a federal lab after X number of years are the same ones that may keep a well-informed candidate from joining the lab initially. Dr. CLARK listed the kinds of problems experienced in federal R&D labs and discussed some of the techniques that labs use to resolve them: Problems Compensation Solutions One compensation issue of concern to most lab directors is the federal pay cap and its effect on the ability to retain senior administrators and outstanding researchers. Alternatives include nonsalary compensation, pay banding, recruitment bonuses, relocation bonuses, and occupation- specific pay scales. 165

Time required to extend Direct-hire authority, simplified hiring procedures, and an offer of employment increased personnel authority for line managers. Difficulty of promotion Pay banding, increased personnel authority for line after reaching the GS-12 managers, and occupation-specific salary schedules. level Restricted role of line managers in personnel , · ~ c ,eclslons Excessive paper work Questionable tie between performance and pay Personnel ceilings and reductions-in-force (RIF) Flexibility in increasing salary without promoting, increased personnel authority for line managers, and occupation- specific salary schedules. Direct-hire authority, computer-assisted classifications, and more generic classifications. Performance appraisals, multiple components of pay increase that are not mutually exclusive, and bonuses and awards. Use of adjunct personnel such as postdocs, flexibility in considering things other than seniority, and simplified classification systems that enable the labs to retrain RIFed staff. The aging work force Well-thought-out institutional planning is necessary to ensure long-term health of the lab, but not much is being done lob satisfaction, morale See above. Lab directors and managers, in general, seem to fee} that the civil service system is external to them. It operates outside the lab that it is serving so its accountability is less than would be desired. In DOE, for instance, each lab has its own personnel system. Also, the civil service system is designed to be an efficient, effective system for the "average" federal employee in the "average" federal agency. Discussion Demonstration Projects Mr. THOMPSON cited the measurable benefits of the Navy demonstration projects as identified by OPM: (~) An increase in quality of people who have been recruited, even in less advantageous geographical settings, as indicated by increases in GPAs; (2) Easier recruitment because of the ability to offer starting salaries reasonably close to the industry average; (3) Pay progression more similar to that employed in the private sector (start higher and advance sIower); 166

(4) Pay for performance: high-quality people are often attracted by a system that will reward them differently from people performing less well; (5) Satisfaction with the revised job classification system, which is as accurate and certainly more expeditious than the old system; (6) Improved attitudes of managers, who feel significantly more empowered to run the personnel system through the various flex~bilities built into it; and (7) Increased job satisfaction. China Lake: According to Mr. PORTER, the Navy's demonstration program at China Lake has been considered a real success since its establishment in 1980. It began as a five-year experiment, was extended for a second five years, and now has been extended indefinitely by congressional legislation backed by OPM. Although lacking authority to grant recruitment bonuses, China Lake does have direct-hire authority and the authority to adjust beginring salaries beyond the broad pay band to be competitive in the marketplace. Assignment of position classification is the responsibility of line managers, with audits by the personnel department after the fact. The personnel office, instead of being in an adversary relationship, is supportive, helping China Lake to provide better training of staff. China Lake also employs a dual-career ladder, whereby top technical people are promoted based on their technical skills and can earn as much as or more than the managers. Mr. PORTER said that the annual additional cost for implementing these changes during the project's first six years was only 1 percent more than the expenditures of similar labs that operate under the traditional civil service system; that cost difference appears to have stabilized. Although the Navy believes a 6 percent change over a 10-year period is not all that significant, OMB staff, looking at the possibility of extending the project government-wide or laboratory-wide, have a different opinion. Dr. KLINE advised that concerns about budget neutrality should be balanced by recognition of the lab's improved performance, which the Packard Commission report and other studies have pointed out. NIST: Mr. LYONS explained that in 1982 the White House Science Council asked David Packard to convene a subcommittee that visited a number of national labs and made four recommendations, one of which dealt with personnel problems and focused on the China Lake experiment. Presidential interest in those findings led the science adviser to convene a Committee on Federal Laboratories, chaired by Jim Ling, to look into the Packard report. That committee responded to its charge by drafting a bill that proposed even more flexibility in regard to personnel issues than the China Lake experiment allowed. NBS had one goal in mind-to maintain and increase the quality of its staff. However, under review by OMB, OPM, and the agencies, the proposed bill engendered stiff opposition, but Dr. Packard personally delivered the draft to Don Fuqua, then chairman of the full House Science and Technology Committee. Congress eventually passed that legislation as a part of the reauthorization bill for the NBS, detailing line for line what its demonstration project could do. One provision is direct-hire authority for the whole agency, except for blue-collar workers: even a division chief can make a hire, offering what he or she thinks the market demands up to the 75th percentile of the top salary stated in the DOE salary surveys. In addition, 167

NIST can make an on-the-spot offer and hire in a matter of days, offer $10,000 recruitment and retention bonuses, and pay whatever starting salary seems valid. Another important aspect of the demonstration project is having line managers do the classification and the qualification check, with personnel officers authorized to audit the programs. As a result, ent~y-leve! problems experienced by NBS seem to have been resolved, as have those associated with recruiting. Retention problems have arisen, however, as division chiefs and senior scientists accept professorships at m~-level- uni~rersities for more than $100,000 (but aggregate statistics do not show that loss, which only concerns the top 5 percent of employees). Some participants believed that some turnover is necessary to prevent a laboratory from becoming static. Dr. ROBINSON expressed concern about the mechanism by which the operations of a federal demonstration project achieve full-scale implementation. He described, from his own experience, the difficulty in moving from research into production, particularly when there is an institutional barrier between the research lab and the production facility; the major role of the manager is to overcome these barriers before they arise. Similarly, he believed, before beginning a demonstration project, there should be a mechanism for determining what one hopes it will achieve and criteria on which to base decisions about implementing the demonstration on a permanent basis, as well as a mechanism for providing continuing oversight. Both Dr. AMBLER and Dr. CAMPBELL asserted that just getting authority to establish a demonstration project was a major feat and that taking steps during the planning stage to ensure its life beyond the experimental period would be too time- consum~ng and involved. Nonetheless, Dr. CAMPBELL, concurring with Dr. Robinson's analysis of the barriers to implementation, agreed that a major question is how to overcome those barriers to make the leap from the current personnel management system in the federal government to a more effective one: How do you select mechanisms for deciding to go beyond the experiment and implement its policies and procedures on a permanent basis? Discussion among participants provided no answer to this question. Dr. GINZBERG felt that if Congress will not give the executive departments more leeway in spending their allocated funds, implementation of demonstration projects will be difficult; he advocated substantial decentralization of the personnel budgets of the agencies. Dr. AMBLER added that one must recognize from the beginning that designing a demonstration project is a lot of work for senior managers, who must plan and clear it with the relevant department and OPM and get the technical people enthusiastic about it. Dr. AMBLER asked how OPM and the congressional committees would react if the President's science adviser said, '] have read about these demonstration projects. They are the best thing since sliced bread." Dr. CAMPBELL felt that OPM would support demonstration projects, but Dr. NUMRICH believed OMB would oppose increasing the number of projects involving scientists and engineers; furthermore, OMB has criticized the existing demonstration projects as having not proved success through the use of acceptable metrics such as increased productivity. Dr. FINNEY said OMB's support would depend on whether the project is revenue neutral. However, Dr. NORWOOD said that even more important is the political reality of concern to 168

OMB: when the President says we are going to put the government on a diet but the result is not fewer people, it becomes a political issue. Dr. MESSNER offered two other relevant points: (1) OMB has become less aggressive about setting arbitrary numbers as controls, focusing mainly on a guarantee that the demonstration project not exceed its funding level set by OMB; and (2) it takes a long time for Congress to move. Dr. KLINE said that when he attended Cabinet meetings during the 1980s, support for demonstration projects that would enhance the recruitment and retention of scientists and engineers was not forthcoming because the data showed an attrition rate of only 5 percent for scientists and engineers, half of the government-wide rate in general. He felt that instead the focus should be on how a demonstration project can enhance the performance of government laboratories. However, Dr. AMBLER mentioned the lack of a productivity measure for an R&D institution. M&O Facilities For the management-and-operating-contractor (M&O) facilities, the DOE operations offices negotiate personnel policies and salaries with the contractors at each lab. Although salaries vary from one lab to another, the contractor itself must conduct market survey to Justin the proposed salary schedules. On the other hand, because DOE has final approval authority, staff in the operations offices are aware of the relationships between the salary schedules of their contracted facilities and the GS schedules. Mr. HOWTON said that DOE does operate two GOGO (government-owned, government-operated) research laboratories, one in Morgantown, W.Va., and one in Pittsburgh, Pa., as well as the on-site M&O contractors. In a broad sense, the contractors' salaries are significantly (10-15 percent) above the federal salaries, although the GOGOs and M&Os are almost mirror organizations. There are differences, however, in the deferred compensation costs associated with each type of facility; such costs for the M&Os show up in contract costs to DOE, not in base salary. M&Os also have stock ownership plans and other benefits not offered by the government. Dr. SUTHERLAND described the strong working relationship between the Jet Propulsion Laboratory (JPL) and Cal Tech, whose faculty serve as principal scientists or investigators on JPL projects and whose students are involved in the lab's research projects. All of JPL's policies and procedures are Cal Tech policies, not federal ones. JPL advises NASA of planned changes, and NASA can object within a certain period of time, but the relationship with NASA is superb. In fact, many JPL staff are detailees to NASA headquarters on a non-conflict-of-interest basis, and some Cal Tech faculty are on loan to NASA through JPL. Other Possible Solutions Dr. COMPTON said that the NASA Ames Laboratory had two options to the civil service system: to fix it or to get out. The use of demonstration projects is an example of fixing the problem. Becoming an FFRDC or an M&O is a way of moving out of the civil service system; JPL is a fulfillment of that option. He described the difficulty in acquiring university sponsorship of Ames activities. Respected universities 169

were not interested in picking up a federal research center as part of their institution for two reasons: Ames was too big an operation, and universities want more control over the lab's activities, particularly because of adverse publicity that might occur if something went wrong in the lab's operations. Dr. COMPTON observed that to establish a strong relationship, a university and a federal agency must "grow up together": they are difficult marriages to make once two mature institutions are involved. Mr. CAREY, mentioning his membership on the University of Chicago's Board of Governors for Argonne National Laboratory, agreed that the university places a high value on its relationship with the laboratory. They have established a positive, mutual relationship, although the inevitable tensions between the parent government department, DOE for example, and the laboratory are sometimes difficult for the trustees of the university to accept. In other words, constraints exist, even in the flow of appropriation funding to the laboratories. Mr. HARDY observed that some federal organizations such as the Atomic Energy Commission (AEC) offer a number of the flexibilities associated with the demonstration experiments. However, the AEC opted for contractor-operated labs because of the frustrations associated with hiring and the desire to use their work force effectively. Dr. CAMPBELL asked whether managing a contractor requires special skills, whether it is more difficult or easier to manage a contractor than it is a federal work force. Mr. HOWTON replied that comparisons are inappropriate because each involves an entirely different set of laws and procurement regulations; management problems that do arise tend to be legal and liability issues associated with nuclear energy: people are scared of what the costs will be for them. However, Mr. HARDY noted that in the Clinch River Breeder Reactor project, which involved both contractor and federal employees, dealing with the federal part of the system was more frustrating and more difficult, having more "traps" and rules. Responding to Compensation Concerns Entry Level: Dr. VAN RIIN was concerned about morale problems that may arise when scientists working in different divisions receive different entry-level pay because they have different managers. Dr. LYONS did not consider that a serious problem: the NIST demonstration project resembles an industrial personnel system. Dr. STEINBACH pointed out that the entry pay in the government is low. For example, hiring a physicist in the area of electromagnetics is difficult not only because there are so few available but also because their initial pay is less than that of a secretary in the Washington area. He also observed that the median salaries indicate insufficient flexibility in the government: just about everybody gets the same pay. Mr. FECHTER noted that a recent review of DoD's science and engineering work force reinforces the impressions that the salary disparity is quite large at the entry level but narrows dramatically after four to five years of employment. Promotion is a mechanism for managing one's work force, although it may not be the right mechanism. The DoD study also examined movements of individuals within the federal government, between positions and across grades. Dr. NUMRICH presented data based on the national Hay Survey of R&D 170

organizations contracted by DOE. On a maturity curve basis, NRL lags about 40 percent at the 90th percentile for entry-level engineers (special rates are included in the NRL figures). At the higher experience levels, the differences are about 20 percent for nonsupervisory positions. However, when the comparison is based on a comparable private-sector organization, salaries for all types of degree recipients (B.S., M.S., Ph.D.) working in either supervisory or nonsupelvisory roles lag by a figure closer to 25 percent, representing dollar lags of as much as $25,000 for lead researchers. Fully aggregated data show only a 2-5 percent lag. She emphasized that analyses of data relating to federal recruitment and retention should include examinations of private-sector organizations doing similar work. Mid-Career and Senior Levels: Agencies-particularly demonstration projects-have devised special provisions to address issues dealing with supervisors and senior scientists. At NIST these included higher salary offers, a program of NBS fellows, and supervisor differentials of 3 percent for group leaders and 6 percent for division chiefs (NIST has no employees in the super grades GS-16 to -18, about two dozen on the ST salary schedule, and slightly more than 100 in the SES). Dr. MAHLER noted that although both NIST and the National Oceanic and Atmospheric Administration (NOAA) have facilities in Boulder, Colo., the initial concern-that the higher salaries offered by NIST would encourage NOAA employees to transfer to NIST-has not been realized. He felt that NOAA was able to retain its staff because salary is not the only issue: nonsalary items such as the work performed are important, too. CDC has no trouble recruiting young scientists, both because of the bonus pay option and because of its research orientation: the excitement of research institutions in the federal government is a lure for young scientists, who are willing to put up with a lot of things to work in an exciting environment. However, Dr. GUINAN noted CDC's difficulty in recruiting at the middle and upper levels because the excitement of the environment is overshadowed by prospective employees' concerns about money and sending their kids to college. A middle-level scientist with two children in college cannot survive on a federal salary. In spite of the difficulty of recruiting at the mid-level, Dr. GUINAN felt there is not enough turnover once career scientists reach the upper levels: the lack of new blood coming into the upper strata could lead CDC away from the cutting edge. She concurred that there is a particular shortage of scientists in any emerging field: CDC needs molecular biologists to conduct research on retroviruses but cannot compete with the higher salaries paid in the private sector. Mr. RAGONE found it difficult to justify noncompetitive salaries at a time when the research work force is aging dramatically. At GS-5-9, the U.S. Geological Survey is being "eaten alive" by the consulting firms that offer hydrologists much higher salaries. He also noted the lack of discretionary funds that would enable federal employees to attend professional meetings. Pay for Performance: The criteria for evaluating performance is based on comparing an individual with his or her peers in the scientific area-how many papers they write and how much basic research they do. In general, scientists seem satisfied with these criteria. Mr. PORTER said that China Lake has been able to establish a link between 171

pay and performance that works better than the old system; although there is not universal satisfaction with the new system, external studies indicate that, in general, it is accepted and people are happier with the system now than they were before. According to Dr. AMBLER, pay for performance was implemented at NIST before the demonstration project; it is a lot of work but is considered to be much better than the alternative. In fact, NIST hesitates to give retention bonuses because the whole system is based on pay for performance, and it is better to give employees permanent adjustment to base salary. Dr. AMBLER also clarified that its authorizing statute says nothing about revenue neutrality, but the NIST administration wanted it. (After the authorizing act was passed, NIST was told to be budget neutral and not to exceed the civil service pay cap. However, because half of NIST's income comes from other federal agencies, Dr. LYONS wondered whether anybody knows what "budget neutral" is.) He concurred with Dr. CAMPBELL that, in order for a pay-for-performance system to work, there must be clearly established objectives to be achieved in a particular time period. Responding to a question of Dr. Numrich, Mr. THOMPSON said that OPM considers the Performance Management Recognition System (PMRS), established in 1984 by Congress, to be a system based on pay for performance. However, he noted two important distinctions between pay for performance in the Navy and NIST demonstration experiments and the PMRS: (1) the potential for large rewards under the demo systems has been much greater because of greater funding, and (2) the demonstration project systems have won greater employee acceptance because of the way they were implemented, beginning with communication of objectives and follow- through dialogue between employee and supervisor. Dr. CAMPBELL questioned whether promotions in the federal government are a result of performance or simply a way of catching up and making the government competitive. Dr. NORWOOD responded that federal scientists almost automatically progress from GS-S to GS-11; after reaching GS-11, it becomes harder to get a promotion. Dr. CAMPBELL said that a variety of gain-sharing programs exist in private industry but are primarily restricted to mid-level management as opposed to being total- employee programs. Dr. MAHLER cited the inequities in the federal government's bonus systems: A scientist could receive a bonus under the PMRS, while the bonus system for nonmanagers is the General Workforce Performance Appraisal System (GWPAS). The PMRS has a finite pool of money that can be used for bonuses, whereas GWPAS has limitations only on the size of each individual award. Thus an outstanding manager might receive an $800 bonus while an outstanding performer under GWPAS could receive $2,000. Dr. CAMPBELL contrasted this policy with that in the private sector: companies want everybody to maximize their bonus because if that is true, the corporation is doing great. Dr. GINZBERG said that this awkward, inflexible, complicated system of personnel management in the federal government is closely connected with less-than- optimal performance. That fact should be the basis of any case for greater agency discretion. In response, Dr. NUMRICH declared that NRL has historically performed excellent research and continues to do so: '~e are in a situation of diminishing returns and maintain our influx of talent only because of heroic efforts in recruitment and retention and the presence of excellent scientific leadership." The real issue is that in 10 172

years the federal government can expect catastrophic failure because senior leadership will be eligible for retirement, and new employees (those under FERS), although they may be well-trained and have sound scientific reputations, will leave because of being in high demand and facing the financial problem of meeting college tuition payments: the federal government Will be like a university without sernor professors. Dr. NORWOOD asked whether demonstration projects have been tried in an atmosphere where at least the m~ddle-level employees are unionized. Dr. CLARK responded that Tennessee Valley Authority used a pay-for-performance system for its nonmanagerial, white-collar workers for seven years but abandoned it because of perceived inequities, both individually and across organizational units. The Political Appointments Process and the Recruitment of Scientists and Engineers Dr. CAMPBELL said that certain facts and observations prompted the Committee to consider the relative significance of noncareer appointments to the recruitment, retention, and utilization of scientists and engineers. Of concern were the increase in the number of political appointments; the number of vacancies in those positions; their influences on the culture, performance, and productivity of the federal organizations that make extensive use of scientists and engineers; and the slowness of the appointment process. Dr. PFIFFNER began by describing the pressures of a Presidential transition: at a time when policy, power, and position are "up in the air," one must handle personnel recruitment, likened by some to "trying to take a sip from a fire hydrant." Of about 550 PAS positions (Presidential appointments with the advice and consent of the Senate) in the executive branch, the subset of scientists, engineers, and people that supervise them equals about 250 (according to National Research Council calculations). These numbers exclude noncareer SES and Schedule C appointments. Recruiting the best and brightest people for these positions is not difficult at the Cabinet level because of the prestige and power associated with Executive Level I appointments. However, recruitment of the subcabinet-that is, Executive Levels II through V and noncareer SES is difficult for a number of reasons: noncompetitive pay, ethics requirements, financial disclosure, post- employment restrictions, the short tenure in office, and the daily risk of being dismissed (because one serves at the pleasure of the President). Determining how the increased numbers of political appointees affect the career force may be difficult, but the size of the increase is significant: PAS positions increased from 152 in 1965 to 527 in 1985; the number of noncareer SES rose from 582 in 1980 to 658 in 1986; and Schedule Cs increased from 911 in 1976 to 1,665 in 1986. These increases reflect a deeper penetration of political appointees into the career bureaucracy. In addition, control of political appointments has been centralized in the White House. Although PAS's are all presidential appointments, in the 1950s and 1960s, most sub-Cabinet appointments were determined by Cabinet Secretaries. But beginning in the Nixon administration, Presidents have felt that they gave away too much of their appointment power, leading the Reagan administration to centralize all political appointments, including noncareer SES and Schedule C, in the White House. The potential implication is that agency heads and cabinet secretaries might have 173

J different criteria in mind than the White House staff Certainly an agency head watt Took for somebody who has expertise and competence in management because that person watt make the agency work or not work. The White House, on the other hand, may very well have a different perspective and be especially sensitive to political pressures for rewarding the party faithful and appointing those with certain ideological values. Dr. PFIFFNER agreed with a VoIcker Commission task force chaired by Elliott Richardson: the problem is systemic such that higher numbers in combination with deeper penetration and centralization do have some relationship to the diminishing quality of political appointees. Also important to scientists is the leadership and vision of the President. In addition, the President's science adviser plays an important symbolic role: if it appears that person has to compromise his or her professional ideals to do something political, fewer scientists watt be welling to enter public service, even for a few years. Scientists and engineers said that it is important that they be able to respect the technical competence of their boss and fee] comfortable with his or her ability to evaluate their work. Finally, scientists and engineers expect political appointees to duly consider their work and to buffer these technical people from the whims of the political wind. Discussion Dr. CAMPBELL asked for comments on the degree to which it makes any differ- ence whether assistant secretary or deoutv assistant secretary appointments which have , ~, . · · ~ · · ~. ~ ~ ~^~ ~ some supervisory responsibility In relation to the K~L' sloe or government, t1) are Ellen or left vacant and (2) affect the quality and tenure of people appointed to those jobs. Problems Associated with Unfilled Positions Dr. CHEN pointed out that the NTH director was not a presidential appointee until passage of the National Cancer Act in 1972. Since that time the job has become quite political, to the point that for the last six months, NTH has not had a director, resulting in a certain loss of momentum, morale, and overall sense of purpose. A search committee suggested individuals to serve as NIH director, but there were sufficient political ramifications that none of those candidates were interested in the job. The secretary of the Department of Health and Human Services has convened a second high-level committee to determine how to make the job more attractive. Once issues such as salary and authority are worked out with the Secretary, it is hoped that the job can be made sufficiently attractive to initiate a second search. Dr. CHEN felt that such situations would be less prevalent if the position were not a political appointment. Dr. ROBINSON believed that having an assistant secretary for health who can override decisions of NIH staff, including the director, is a problem as is the fact that the director is many levels removed from the actual running of a $6 billion agency. Quality of Presidential Appointees Many participants said that having a boss who has a reputation in his or her respective field is important, even more important than salary levels, all the way down the line, even in hiring decisions. 174

Dr. NUMRICH felt that having both political appointees and vacant positions at the top leads to less stable funding. When department heads lack institutional memory and knowledge of what ought to be going on, particularly in the administration of funding, fewer risks are taken. She said that R&D labs perform risk-oriented work that is possibly high gain, but such work is not now regarded as primary. Another issue brought to the Committee's attention is the inadequate management preparation of career executives and career managers. More than once, a newly appointed SES member admitted to attending his or her first management training session after being appointed. Dr. PFIFFNER said that the uniformed services develop their executive talent by sending them to special schools, but less is done on the civilian side. Dr. CAMPBELL agreed that moving people from technical specializations into managerial positions has always created problems; he prefers to take a specialist and give him or her managerial training rather than assume that a manager can manage anything. Tenure of Presidential Appointees Dr. MESSNER stressed that the engineering and scientific community in the government cannot exist separate from the political process; it would not be a realistic goal to build a wall around the scientific engineering community, treating it differently from other government employee groups. Without political participation, however, there is no way to prevent the erosion of the attributes of the scientific and engineering existence in the federal workplace. If an agency does not have an advocate at the table when budgets are reviewed, an agency not only will not get training money, but also probably will not get facility money, parking places, or health care for its occupational safety program. MS. MOWRY said that the current system makes a big impact. During the first year political appointees must develop loyalties with the career employees. A PAS member must develop a better image of the federal employee and be open-minded and willing to cooperate with the career staff. Since the average stay of a political appointee is 18-24 months, there is constant change at the top, with subsequent change in loyalty as well as much reorganization. Dr. AMBLER said that when Presidential appointees are not allowed to do the job they are supposed to do, they start to micromanage. They are appointed because, presumably, they are of the same opinion as the President on political affairs; however, scientists and engineers become nervous by the implication that ideology would overwhelm scientific objectivity. Advocacy, in the budgetary sense, will go to agencies whose programs fit what the President is trying to do. Fixed Term of Office Workshop participants offered several insights and suggestions: (~) According to Dr. KLINE, a fixed term is not a bad feature but does not guarantee that the appointee will hold the position for the entire term. (2) Dr. ROBINSON advocated a fixed term for the directors of NSF and NTH so that when the time comes for appointment, consideration can be given to reappoint 1 7C

ment of the person in a technical position who has done a good job. Some thought that fixed terms established SO as not to expire around the time of a presidential election would be beneficial. While not disputing this point, other participants considered having the confidence of the Administration more important to agency staff: if an appointee has a s~x-year term but not that confidence, he or she has little besides an office. There are inconsistencies within the system: the commissioner of Labor Statistics has a four-year term, but the director of the Bureau of the Census does not: therefore, the latter position is considered a political appointment. (3) Designated Positions Dr. CAMPBELL asked where political appointees should be in the system. The civil service Reform Act, which eliminated the designation of positions as either political or career and specified that no more than 10 percent of the SES appointments could be noncareer, attempted to open up higher level positions to career people, thereby making it possible for them to be appointed to assistant secretary positions and the like. He wondered whether the government should return to a system of designated positions, rather than leaving the classifications open. According to Dr. PFIFFNER, after lengthy examination, the head of personnel at the Department of Health and Human Services has concluded that we might as well use the former system, because as soon as a political appointee is placed in a job that was formerly a career position, the job becomes politicized; the next administration thinks that is its slot. Dr. CAMPBELL said that, if that occurs, career people can never aspire to become assistant or deputy assistant secretaries: taking a political appointment would eliminate their rights back into the civil service system. Mr. CAREY said it is important to give career people the opportunity to accept political appointments with some falIback protection. Dr. NORWOOD said that during her long tenure in a PAS position, even though it has a fixed term of office? she has seen a continuing erosion: many positions that could be filled by either a careerist or a presidential appointee have been designated PAS unless a career position has been clearly protected. The 10 percent limit on the number of PAS positions has been · ~ maxlmlzec .. Dr. MESSNER observed, based on personal experience with the White House personnel office under five different administrations, that the political appointment process as it regards scientists and engineers is nonexistent. He wondered, if one really cannot resolve this issue by designating jobs as political or career, if a solution might be to advocate identifying positions that should have professional criteria, limiting the President's appointment power. Dr. ABRAMSON added that staff in the White House personnel office complained that they did not have enough names of scientists and engineers for PAS positions. He suggested that national groups supply the personnel office with names of qualified people in the scientific community. In addition, the personnel office staff find that the individuals whom they do contact think the pay offered is not worthwhile and the ethical requirements are bothersome. Based on his own experience, Dr. CAMPBELL said that the White House personnel office does not try to identify and recruit people except for the Secretary 176

positions. Dr. MESSNER agreed and added that the source of the name that is sent to them is screened by the personnel process; one must have certain political prerequisites to have his or her suggested nominee even considered. Mr. HOWTON agreed that a long-term civil servant with the right political connections will not get through the personnel office because the Hatch Act prohibits such contacts. He said that political appointees can be good or bad, but if an agency has no political appointees, it will instead have a congressional oversight committee a situation that creates its own set of complications. General Discussion Dr. MESSNER opened this session by saying that in a buyer's market, one can afford to be much more sloppy in dealing with human resource issues, but in a shortage situation, one can get in trouble quickly. The engineering community is concerned about the potential lack of human resources in the near future partly because of the changing demography: There are fewer young people. During the next 15 years, 85 percent of the entering work force will be women and minorities, but historically the engineering profession has been unsuccessful in attracting women and even less successful in attracting minorities. Enrollment in engineering schools is down, even for women, who for a short time were enrolling in engineering schools in greater numbers; 50 percent of the students now in U.S. engineering schools are foreign and could be a resource if they choose not to return to their own countries. Statistics show that the United States will have a shortage of about 400,000 engineers as we enter the 2Ist century. Dr. MESSNER questioned the effectiveness of the U.S. education system, stating that some scientific and engineering specialties are more successful in preparing students for the work force. As an employer, he is concerned about investments in the engineering person power and felt the federal government must engage in forward thinking about the human resource pool and how it will attract that pool to its programs. Dr. NORWOOD asked whether the supply of scientists and engineers is actually a bigger problem for the federal government than for the private sector. In response, Dr. FALK said that his industry contacts say they can get the people they want, as long as the pay is high enough. Mr. RAGONE said that federal agencies provide a training ground for industry, hindering the agencies' abilities to fulfill their missions, but the loss of small groups of highly trained scientists and engineers is not reflected in aggregated statistics. Mr. ANDREWS said that he had had the same experience: the Federal Aviation Administration hires engineers who, after a training period of three to four years, take positions in industry at salaries $10,000 higher. Dr. STEINBACH noted similar problems in the fields of electromagnetics and computer software and predicted that the current national shortage will get worse. Dr. NORWOOD concurred that many agencies face similar problems: scientists and engineers come in, get valuable experience, and then go on. 177

Dr. MESSNER said the federal government must plan with greater because its infrastructure causes the government to move so slowly. Industry can move more quickly because firms can operate without adjusting to the labor market, either through pricing or, more importantly, through strategies to recruit from underparticipating groups. For instance, industry can attract more women by responding to special problems that women have in the work place. Constraints Mr. FECHTER asked participants to define the real constraints. He noted that mechanisms are being used to get around pay caps, and promotion is a mechanism to get around pay inequities. However, what other constraints exist and what are the mechanisms for getting around them? Dr. CAMPBELL pointed out that today OPM may delegate most personnel authorities other than pay caps and some other limitations. Dr. CHEN said that NIH, because of the Public Health Service Act and some of its special authorities, does not have the same constraints as many agencies about hiring foreigners. As a result, about one-third of its doctorates are foreign. In ensuing discussion, participants suggested that language in appropriation acts might be changed so that other agencies could hire foreign nationals. Two other points raised by Mr. ANDREWS are that technology now has a more international focus, and the rate of technology changes so quickly that engineers must spend much time to even keep pace with it. Past Initiatives To Recruit Scientists and Engineers Dr. NUMRICH suggested looking at various initiatives, including DoD's attempt to extend the Naval Ocean Systems Center/Naval Weapons Center demonstration project to the rest of the department's laboratories, and evaluating why they have failed. To move forward we must understand how and why those failures happened so they will not recur. Dr. CAMPBELL said that some initiatives fail because of opposition based on the proposition that they should be made government-we rather department-specific. He surmised that legislation now before Congress to grant special salary pay rights to specific departments and agencies may, in fact, be attractive but lead to further fragmentation of the total federal employment system. Dr. CHEN cited an example to support that belief: NTH developed a legislative proposal for higher salaries for its scientific faculty (equating its scientists with comparable ranks in medical school faculty) based on an annual salary survey conducted by the Association of American Medical Schools. Another aspect of the proposal was to link the retirement system to the TIAA- CREF, allowing faculty from medical schools to come into the government for a brief period, remain in the same retirement system, and then maybe move out of government. The proposal was not successful because of objections by OMB and by other sectors that wanted to have similar special legislation. Dr. MAHLER mentioned NOAA's Environmental Research Laboratories' use of a large part of its global change money to finance six joint institutes done cooperatively with universities and co-Iocated in its laboratories. ERL does not pay the principal 178

investigators' salaries but does pay postdocs' and graduate students' salaries, thus promoting the study of science and engineering by providing monetary support. There are not many environmental scientists, and the supply has been going down. Thus NOAA is trying to encourage potential environmental scientists by putting more money into the universities, where the training is available and these cooperative institutes work. NOAA hopes not only to encourage undergraduate students but also to broaden the spectrum of new entrants. This discussion led Dr. GINZBERG to ask: To what extent does the goverrunent really get in its own way in terms of attraction, retention, and utilization of people by underinvesting in the continuing education of its scientific people? Participants described the effective programs of DOE, NRL, and CDC but noted that training money is being cut back for 1991, and providing such training does not ensure that those in agency-sponsored education programs will not use that training to acquire higher paying jobs in industry or academe. Future Initiatives Dr. MAHLER believes there is a new mode of operations whereby the government runs programs across agencies, not through agencies. An intergovernmental committee on earth sciences, for instance, is advocating the global change program now: . . . . . c;~11 a'~t;ll`;y 1~lVulVea naS agreed on ltS ContrlOUtlon to the program and the ~nmmitt~ ^~- It ~1..~4 ~_~ ~_~1 _ :~ '1~ . _ .1 has created realistic budgets and is selling the program to Congress. Such programs are more salable, but they take some control from the agency heads and put it in the program areas. These interagency committees have broad scientific and management support of attempts to find solutions across the board, a far more effective mechanism than single-agency programs. Dr. FALK said that although such committees could be effective, the real solution is establishing one central organization responsible not only for the welfare of science but also for the effectiveness of the federal science and engineering work force; such an organization could be the one to fight the political battle. . r ~--, ~_ 179