The presence of a few individuals of exceptional talent has, to a very large degree, been responsible for the success (and even the existence) of outstanding research and technology development organizations. It is not the function of the laboratory to be "representative," to be a cross section of the population, but to nurture exceptional talent wherever it may be found. A technology development organization that cannot tolerate and nurture a few eccentrics is halfway toward rigor mortis. In many Federal laboratories, the best fundamental research projects are almost always built around one outstanding individual. This person has demonstrated that he is capable of performing basic research of high quality, and he is accordingly granted the freedom to pursue his interests. It should be noted that in these smaller projects the question of duplication of effort usually does not arise, since usually no large resources are involved.

The most talented individuals will make their way, but what of the majority of professionals? For them, continuing education is essential not because, as some allege, the exponential growth of new technology is pushing back the age of staff obsolescence, but to improve peripheral skills and to make employees familiar with subjects outside their areas of expertise. Note, also, that continuing education does not necessarily improve the performance ratings of people whose skills are becoming obsolescent.* Studies of continuing education programs in laboratories indicate that the highest performers took the fewest continuing courses; that a staff member's level of initial education is more important in determining his or her rating and salary than subsequent course work; that the younger the staff member at the time of receiving his highest degree, the higher the rating; and that there is little evidence of improvement in relative performance as a result of participating in continuing education courses.

These findings imply that the cost to an organization in man-hours lost and dollars spent may not pay off in improved staff performance. Alternatively, the findings may be interpreted as indicating that the particular programs of continuing education covered in these studies were

* Based on an admittedly limited survey taken by the General Accounting Office at two Air Force laboratories, it appears that many scientists and engineers consider management training more likely to lead to reward than technical training. More than half of the interviewees "considered that maintaining technical expertise is not an effective way to enhance promotion prospects. Promotions and awards are based upon job performance and, while special recognition may not be given to skill enhancement through education, there would be an indirect benefit when such activities improve performance upon which rewards are based. Nevertheless, their perception of nonrecognition might discourage some individuals from participating in continuing educational activities." Letter, R.W. Guttman, Director, Procurement and Systems Acquisition Division, U.S. General Accounting Office, to Secretary of the Air Force, March 6, 1978.

poorly designed or administered. Continuing education for mature scientists and engineers should not attempt to repeat or update graduate courses. The mature staff member needs a clear presentation of material he can use on the job, preferably presented by successful professionals within his own organization. The fact that so many outstanding research and technology development organizations with tough-minded managements provide continuing education, often wholly or partly during working hours, indicates a prevailing belief in continuing education as a means of raising the general standard of staff performance.

Finally, we assert that, in principle, simple measures exist for evaluating the aggregate productivity of professional staff and that they need not exclude one another. Productivity can mean the number of publications produced by an individual within the past five years, the frequency with which these publications are cited by the author's colleagues, or the extent to which the staff translated their own or others' research into applications (ref. 159). Such measures are essential for planning and budgeting purposes but, even more important, they are the yardsticks by which the laboratory's reasons for being may be judged. And it is surprising, but also encouraging, that productivity does not seem to be clearly related to staff age, age distribution, tenure, or turnover rate. Perhaps it remains only for research administrators in some laboratories to recognize this.

CHAPTER VIII

Supporting Functions and Personnel

Supporting Functions Defined

Whatever the technology development laboratory may be, the last thing we can call it is a one-man show. The presence of a few exceptional individuals may account for the quality of the laboratory's work, but hardly explains how the work gets done. The exceptional scientist or engineer can accomplish little by himself. He needs colleagues with whom to exchange ideas, as well as to serve as collaborators in the working out of those ideas. But even after we subsume all the scientists and engineers at a single installation, there remain several categories of workers without whom the laboratory could not function at all. These perform what are called, for want of something better, "support" functions. There is no really adequate definition of support, but provisionally, we can say that whatever is not included in scientific and engineering work at one end and administration and clerical work at the other is a supporting function.* Support activities include everything from mowing the lawn or carting trash to writing sophisticated computer programs or running a tracking station. The more complex these support functions are, the less distinguishable they are from research and technology development.

But even this way of putting matters scarcely elucidates the significance of support functions for a research installation. Consider any large NASA center. It may have a photographic laboratory or image-processing facility for converting digitized information transmitted by satellite into pictures, storage facilities for holding expensive

* The functions we will discuss are covered by NASA Occupational Codes 100 and 300.

100-Wage System- (Trade and Labor Positions): Includes trade, craft, and general laboring positions (non-supervisory, leader, and supervisory), compensated on the basis of prevailing locality wage rates.

300-Technical Support Positions: Includes scientific and engineering aid, technician, drafting, photography, illustrating, salaried shop superintendents, quality assurance specialist, production planning, and inspecting positions.

one-of-a-kind equipment, tracking or telemetry stations for communicating with satellites designed by its staff, and (almost certainly) a data-processing facility for supporting the center across the board. While all of these facilities "support" the center's mission, we need to go somewhat deeper in order to understand the problems supporting functions pose. There are really three questions we have to consider: Can anything useful be said about functions as diverse as lawn mowing and providing ground support for a deep space probe? How can productivity or efficiency in general support functions be first measured and then evaluated? And what are the public policy implications of the contracting out of many support activities by Federal agencies?

To answer our first question: There are two broad classes of support activities. The first comprises those functions that cannot be managed and controlled as a direct function of the levels of primary technology development tasks of the laboratory. Suppose we call these general support functions (table 7). Included in general support are the personnel office, procurement and supply, financial management, administrative computing, and a number of more specialized staff offices. General support activities are typically organized along functional lines; that is, there exists a single personnel office or procurement office serving the entire installation. General support is also designated "research administration," "base support," and "indirect support."

The second kind of support can be closely adjusted to the laboratory's primary work. Under this head we include graphics, reproduction, publications services, technical computing, photographic services, and shop services (table 7). It is clear that costs for a part produced in a shop, for example, can be charged to the technology