Although this function is no less important than the first three, we shall not say much about it here. First, evaluating results depends on the goals set, the game plan under which the laboratory functions, and the structure drafted for it by its senior management. It is not something independent that can easily be treated from a theoretical viewpoint. Planning, operations, and evaluation merge into one another. The same budget which maps the center's strategy for the coming fiscal year is also a commentary on those activities which executives think should be terminated, sustained at current levels, or augmented.

But there is a more important reason why the evaluation of results cannot be treated here. No Federal laboratory is completely independent in choosing what it does. If its work is of national importance — and if not, what is its reason for being? - it will be under pressure from its sponsoring agency, from Congress, from the scientific community, from industry, and from any number of other outside players. It is useful to treat the laboratory as a closed system, but it is not realistic. In the next chapter, we set matters right by analyzing the relations between laboratories and their sponsoring agencies. And we shall look especially closely at a single question: How do laboratories interpret their mission to apply their research to “national" needs? In accordance with our practice throughout this book, we shall approach an answer by way of two case studies, one of the multiprogram laboratories operated for the Department of Energy, and the other of the National Bureau of Standards. We will then be in a better position, through these examples, to show how the work done in a laboratory is actually evaluated in practice.

CHAPTER X

The Laboratory and Its Sponsors

"Can two walk together, except they be agreed?”

The Role of the Sponsoring Agency

-Amos, III, v. 3

There is a delusive simplicity to studying Federal laboratories as though they were closed systems. What we propose is to get behind appearances and review some of the many ways in which laboratories interact with their external environments. As a general rule, research and technology development laboratories tend to be more stable than the agencies that justify and provide the funding for the laboratories. While it is expensive to set up a research institution, the reorganization of an agency in Washington does not automatically require the construction of large new facilities. A good laboratory is a more or less permanent institution and it may shift between agencies as the political climate dictates. Examples abound: the transfer of the National Bureau of Standards from the Department of the Treasury (1901) to the Department of Commerce (1903-1913); the transfer of segments of the Bureau to the National Oceanic and Atmospheric Administration; the organization of the multiprogram energy laboratories, first under the Atomic Energy Commission (1946), followed by the Energy Research and Development Administration (1975) and the Department of Energy (1977); and the selection of the research centers of the National Advisory Committee for Aeronautics to be the core of NASA when it was created in 1958. There is nothing immutable about the ties of a laboratory to its sponsoring agency. The agency may be target as well as shield.

It often happens that the reason for being of a technology development agency becomes politically irrelevant, once the objectives for which the agency was created been achieved. Since 1967, NASA has had to face this problem, shutting down its Electronics Research Center and transferring it to the Department of Transportation, as well as cutting back support to other centers. Similarly, the Army transferred part of the Fort Detrick Chemical Warfare facilities to the National Cancer Institute when the Nixon administration terminated all biological warfare research in 1969. In both cases, there was a sponsor able and willing to take on a

new facility. In other cases, a laboratory, while retaining a primary loyalty to one agency, tries to pick up additional support by diversifying or by arranging to do work for other Federal agencies. We will call this the "resource sharing" model. Mention was made in Chapter IV of the location of the Army Air Mobility Development Laboratory at NASA's Ames Research Center. Another example of resource sharing is the location of the Army Corps of Engineers Nuclear Cratering Group at the Lawrence Livermore National Laboratory, which is operated under contract to the Energy Department. In each case the agency operating the laboratory is funded by another that wants work done there. Almost every large laboratory operated by a given agency does some reimbursable work for other agencies.

More will be said about resource sharing in the case studies included in this chapter, but certain general rules (which account for successful resource sharing) can be mentioned here. The first is that "user" groups should be small compared to the "host" laboratory probably no larger than 20 percent of the host. Second, the host laboratory must be extremely careful not to interfere with the programmatic function of the user group. Resource sharing will not work if the program of the user group is so skewed that it only benefits the host laboratory and its parent agency. Finally, the user group should operate under roughly the same personnel, procurement, and fiscal regulations, if the host laboratory is to execute these functions properly. Unless some administrative uniformity exists, it is possible for purely administrative problems to wreck a relationship that otherwise makes sense on technical grounds.

Perhaps enough has been said to show why easy generalizations about the proper roles of government laboratories are implausible at best, misleading at worst. Government laboratories exist in every phase of dependence or freedom. A laboratory may work exclusively for one sponsor or for several; perform reimbursable or non-reimbursable work for other agencies; be a joint venture of an agency and one or several universities and be operated as a distinct organizational entity;* do fundamental research or perform work very closely tied to an agency mission; or even carry on work which is not closely tied to any agency mission.** No generalizations are adequate to encompass such varied institutional possibilities, and so we have elected to explain matters indirectly, by detailed case studies of two very different kinds of laboratories: the National Bureau of Standards in Gaithersburg, Maryland

*An example of this is the Joint Institute for Laboratory Astrophysics operated by the National Bureau of Standards and the University of Colorado.

** A possible example might be the National Center for Atmospheric Research in Boulder, Colorado, which is funded by the National Science Foundation and operated by a consortium of universities.

and the nine multiprogram laboratories operated under contract to the Energy Department. Once the operations of these laboratories have been reviewed, it will become easier to understand the complexity of the transactions between a laboratory and its clientele, or the reasons for the difficulties encountered by Energy laboratories in moving into work in nonnuclear energy research, development, and demonstrations. A laboratory may be deemed "national" because it is supported out of public funds and supplies certain public goods in, for example, national security, space exploration, or the maintenance of a national measurement system. But such an assertion tells us almost nothing about the way these laboratories operate or why they choose certain courses of action over others apparently as rational. Only selected case studies can bring out the reality behind a laboratory's organic legislation.

But even the case studies set out later in this chapter need something by way of a preface to make them intelligible. We need to explain two of the most important features that make technology development laboratories what they are. First, we will set forth in general terms the functions performed by a parent agency for its laboratories. Then, we shall briefly examine those conditions which militate for or against the independence of laboratories vis-a-vis their sponsors.

Justification of the Technology Development Function. The most important function of the parent headquarters organization is to justify the research and technology development functions carried out by the laboratories and to see to it that the necessary funds are appropriated. Relationships between laboratory and sponsor are complicated and decentralized; only in this way can all the talents, both at the laboratory and headquarters, be brought to bear productively on the problem of how to set the right amount of funding. Figure 47 illustrates these relationships schematically.

The top laboratory managers establish their primary relationship with the second tier of managers at headquarters, and so on down the line. The laboratory director will normally deal with the program associate administrators (or assistant secretaries), the department heads of the laboratory with the various headquarters division heads, and the laboratory division chiefs with the group leaders in the headquarters organization. There is, of course, considerable cross-talk between the various tiers, and independent negotiation regarding programs and funding. These negotiations are keyed to the Federal budgetary cycle and the three main groups involved: the agency itself, the Office of Management and Budget, and the congressional authorizing and appropriations committees. The Federal budgeting cycle is illustrated in figure 48.

The agency formulates a budget which is presented to the President through the OMB sometime in the fall of the calendar year, usually in

September. (Preliminary budgets may be submitted to OMB earlier in the year as a basis for subsequent negotiation.) OMB then reviews the agency's budget, compares it with those submitted by other agencies, and responds by giving a "mark" to the agency in October or November. Final negotiations then begin, and by the following January a final budget for the executive branch is developed. This budget is part of the President's annual message, and by law must be submitted to Congress before January 20.

Once the budget is presented to Congress, it is referred to the appropriate authorizing committee. The distinction between authorization and appropriation is subtle, but it is absolutely crucial. An agency whose appropriations have run out is normally kept alive through a continuing resolution, which permits the agency to continue spending at the funding level of the prior fiscal year until a new appropriations bill is passed. But an agency whose authorization expires ceases to exist. What has made this process especially important has been congressional insistence on periodic, or even annual, authorization. Prior to 1959, most agencies — the Army Corps of Engineers was a notable exception permanently authorized by their organic legislation. In that year Congress added a rider to the NASA appropriation bill, requiring that agency to seek authorization before it could request appropriations. The annual ! authorization requirement was subsequently extended to all new military research and development programs and to the Atomic Energy Commission.

were

FIGURE 47. This chart shows the formal relationships and lines of communication between the headquarters and a large research institution. The most important point to recognize is that these communications take place at many different levels.