made during this late period are usually more 'productive' than 'creative' a pulling together of one's life work, guidance of younger professional employees, etc. The older employee is generally secure in his position, has fewer family pressures, and is willing to take risks again .. Older scientists continue to be useful to the organizations in many ways despite their drop in strictly creative scientific achievement. Although they tend to produce a lesser amount of major creative work than their younger counterparts, they do continue to produce over the span of their careers. When the working environment is otherwise favorable, only two conditions can effectively keep older employees from performing worthwhile creative tasks: negative attitudes and failure to try." (ref. 141.)

These conclusions must be regarded with caution. In the literature of the social science, it is rare to find a study unambiguously supporting a theory the effects of school busing, the negative income tax, and medical care policy immediately come to mind for which equal and contradictory data are not available.

Yet there is evidence that some scientists those who are highly motivated to begin with, and (equally important) who are afforded the opportunity to work on important research continue to produce throughout their careers. Management can encourage these people in a number of ways: rotating assignments, offering continuing education, and sponsoring some individuals for leaves to study at a leading management school or the Federal Executive Institute. It is not, then, intuitively obvious that aging and obsolescence are synonymous if the sponsoring agency provides for at least some influx of new staff, if there is a spectrum of opportunities within the organization, if some scientists and engineers can be transformed into managers (a point to which we shall return), and if the agency can devise ways to measure the productivity of research as a step to increasing it.

Research Productivity in the Federal System

In the context of this chapter, productivity implies several not necessarily compatible things. On one level, productivity means the automation of work, the supercession of manual effort by automated instruments, or the use of techniques like simulation programs which reduce the burden on expensive test facilities. So productivity may include any of the following: the use of computer-aided design to custom-design very large scale integrated circuit chips; the use of interactive graphic design systems in preparing production drawings; the programming of wind tunnel tests from startup to shutdown; the development of laser flow diagnostics; and the use of remotely piloted research vehicles for particularly hazardous tests (ref. 142). On a deeper

level, productivity refers to the quality of the research carried out at the laboratory; in turn, this cannot be isolated from the sponsoring agency's mission. Personnel management and a sense of mission are reciprocal; an agency without a strong sense of mission will not keep its best research people for long. The problem in evaluating research productivity is actually three-fold. It is difficult to measure something as intangible as an idea; research scientists, like other professionals, resist attempts by non-scientists to evaluate their output; and any evaluation of research productivity must use multiple criteria, instead of a simple-minded enumeration of research products (ref. 143.)

In any discussion of research management, we are ultimately driven to consider the sponsoring agency's goals. We have discovered that a rise in the average age of professional staff need not be a disaster, that the measurement of research productivity is difficult, and as we must now disclose that there are not simply "scientists and engineers," but professionals of very different temperaments who may not coexist peacefully in the same laboratory. There are, on the one hand, those scientists who go deeper and deeper into a single area of research, and for whom the institutional setting within which the work is done is a matter of indifference. They are as likely to do their best work in a Swiss patent office, a prisoner-of-war camp, or the laboratory of a small college as in a large multidisciplinary research installation. The kind of people who feel at home at NASA or the multiprogram laboratories of the Energy Department are quite different: They are more likely to go from discipline to discipline while fitting their investigations into the content of some larger mission or project. It is this kind of scientist or engineer who, within the same laboratory, will move from aeronautical engineering to an analysis of the physics of lunar cratering; from advanced control theory to design techniques for control systems for powered lift aircraft; or from information theory to developing methods of noise suppression and error correction in deciphering telemetry signals (ref. 144). It is these persons who can move freely between disciplines, who can see practical applications in the most abstruse research, who are the essence of the laboratory's reasons for being. And the number of such researchers is not large — perhaps one or, at most, two dozen in a facility of 3 000 people.

But the distinctions between one kind of scientist and another go even deeper than we have indicated. The scientist who moves easily between disciplines is perhaps more an ideal type than the norm. Instead, scientists seem to fall into one or another category, depending on whether their work is weighted to fundamental science or to the application of their findings to their agency's interests (ref. 145). Scientists in the first category are more apt than those in the other to demand a large degree of freedom in selecting research projects, to identify more with the scientific profession than with the employing organization, and to work on research

activities which are organizationally separate from development.* In the those who have had an other category, the most productive scientists opportunity to translate research findings into useful applications — are more likely to desire promotion into management positions than either the least productive scientists in their own category or the most productive scientists in fundamental research. It is the business of the laboratory's management to accommodate both categories under the same roof.

We seem to have drifted rather far from the nuts and bolts of personnel management, but in fact this is not so. Taking the perspective of a laboratory director, we have tried to analyze personnel issues in their full complexity. After all, it is the director's responsibility to accommodate basic and applied researchers, and to deal with perceptions that one group is gaining at the other's expense; to sponsor discretionary research; to maintain at least the minimum complement of professional staff needed for worthwhile research (the opposite problem - too large a professional staff is seldom a problem nowadays!); to evaluate the work done; and to decide how much of the laboratory's manpower can profitably be diverted to work for others. (To anticipate our conclusions in Chapter XI, work for others only succeeds where it has a particular relevance to the laboratory's mission. Additionally, laboratories with a good record in basic research are better able to diversify than those handling large development projects, because the work of the former is more apt to spill over into a variety of disciplines.) The director can do much to maintain the vitality of the laboratory, whether this involves changing assignments frequently, identifying candidates for management positions, instituting leave programs for the professional development of people, retraining people and the like.

But even the most capable director can do only so much. To repeat: Personnel policies cannot be appreciably changed or even understood apart from other policies which determine how the Federal Government gets its research and technology development done. And dissatisfaction with these policies is growing. From the Bell Report of 1962 to the reports of the Defense Science Board in the 1960s and 1970s to the recent report (May 1983) of the White House Science Council, the diagnoses have been much the same, although the more recent the report, the more

*These attitudes are more characteristic of highest productivity scientists within each category. Thus while in one survey 66 percent of the most productive scientists in the first category regarded selecting research projects as "extremely important," only 12 percent of the least productive scientists agreed. And while only 19 percent of the most productive scientists desired promotion into management positions, 52 percent of the least productive did so. See Howard M. Vollmer, "Evaluating Two Aspects of Quality in Research Program Effectiveness," in M. C. Yovits et al., eds., Research Program Effectiveness (New York: Gordon & Breach, 1966), pp. 160-161.

caustic the criticism: too much red tape and too little discretionary authority at the workbench level; salaries low in comparison to salaries paid by industry for comparable work; inability of many laboratories, particularly those under civil service rules, to attract, motivate, and retain the most qualified staff; and too much direction of the laboratories by their sponsoring agencies (ref. 146).* And up to a point, the proposed remedies have been the same: make government salaries "comparable" to those offered by private industry; eliminate excessive layers of management; delegate more authority to laboratory directors to make decisions and control funds; and give the directors of government-owned, contractor-operated facilities complete authority to set and carry out personnel policy (ref. 147).

Considering this unanimity of opinion, this dissatisfaction with things as they are (or have been), it is surprising that so little has changed in twenty years. At contractor-operated facilities especially, the freedom from civil service requirements seems not to have produced the intended good effects. There seems to be a rule that, with time, contractoroperated and government-operated laboratories tend to become more like each other. The reason is quite simple: In each case, operating funds ultimately derive from congressional appropriations, and the senior officials of sponsoring agencies are held accountable for their proper use. The failure to understand this is at the root of the conflict between many contractor-operated facilities and their sponsors — for example, NASA and the Jet Propulsion Laboratory (JPL). When JPL's functions were transferred to NASA in 1958, JPL officials assumed that the Laboratory would remain a quasi-independent institution working for one principal client in this case, NASA rather than Army Ordnance. But NASA officials saw matters very differently. To them, JPL would take on the functions of a NASA center, although with a freedom not available to centers staffed entirely by government employees. These two views were irreconcilable and, under pressure from NASA, JPL did become more like a government laboratory, especially after 1964. JPL created the position of general manager, dropped the clause from its contract with NASA requiring tasks to be set by mutual agreement, and agreed to be paid on a cost-plus-award-fee basis rather than by an annual lump sum. The case of JPL can stand for many others. In practice, no Federal agency has been willing or able to give its contractor-operated facilities complete

One criticism found in the Bell Report seems to have dropped out recently namely, that the Government was contracting out essential functions that were properly its responsibility. That less is heard of the problem twenty years later may mean one of two things: that the problem has been addressed and resolved, or that the fusion of government and contractor work has advanced so far that no separation of functions is really possible.

independence to set policies within the framework of its mission, even when there were no specific regulations to prevent this.

These conclusions will strike readers as pessimistic. But there's the rub: Federal laboratories, like other large organizations, do not change quickly. Without recognition of this fact for fact it is there is not the slightest possibility that the recommendations of the latest panel will have any greater effect than its predecessors'. Within the interstices of the Federal personnel system, laboratory directors can do much to start or redirect work, move younger staff into management positions, and, through a variey of tools, breathe life into the laboratory and maintain its technical competence. To understand the limits and possibilities of personnel management in a government environment, we shall look at the subject from an agency perspective. Specifically, we shall examine NASA in the 1970s and a recent experiment by two Naval centers to simplify their personnel management systems by introducing more flexibility. In this context, flexibility includes the ability to link pay to performance, drop outdated position standards, increase turnover of low performers, use the agency's block of senior level and excepted positions to retain the most desirable people, and move people between different divisions or even different laboratories within the same agency.

One word of caution. Many of the methods that make for flexibility are available. Many technology development agencies already have authority to retain certain appropriations until they are spent, reprogram within their accounts, use a percentage of laboratory funds for discretionary research, and terminate unsatisfactory employees. But as mentioned earlier, the theoretical freedom to do certain things is usually limited by the accountability of Federal agencies for the ways in which public monies are spent. They are limited still further by the authority of the Office of Personnel Management to set policy for executive agencies and by the authority of the Office of Management and Budget to set personnel ceilings for the same agencies. At every turn it seems as if the checks and balances of the Federal personnel system work against the interests of the government laboratory. Let us see if this is the case.

Two Case Studies in Personnel Management

NASA in the 1970s. One of the severest tests to which a large mission-oriented agency can be put is how it reacts to the completion of its original mission. Confronted by this problem as early as 1967, NASA responded by closing its Electronics Research Center, cutting back at most of the other centers, and creating a Personnel Management Review Committee of senior employees to advise the agency on personnel matters. But as NASA moved into the 1970s and cutbacks instigated by