Judged in such a perspective the distributional emphasis embodied in Table 15 demonstrates an odd assignment of priorities, one that hopefully does not truly reflect our country's hierarchy of preferences. Of the $15 billion attributed to R&D, more than $13 billion goes either for military, space, or nuclear research. Except for medical research many of the problems that most seriously confront civilization and that urgently require scientific analysis receive no significant attention. This simply makes no sense. When boldly confronted by this kind of presentation one suspects that most Americans would assign lower orders of priority to the other identified problem area. And one would hope that Congress would react in that fashion as well-if it were compelled to face the priority requirement in this way, if, to put it differently, it had to decide pointedly whether it wanted to spend $6.7 billion on space research and only three percent of that amount on educational research.

If the federal R&D budget were presented and considered by the Congress along program lines it would also enable the legislature to make discrete comparisons within programs and to effect appropriate reemphasis. For example, Table 15 discloses that most of the money being spent on transportation research is connected with aviation. (16) Very little is used for research in surface transportation, even though virtually all of the country's freight and most of its passenger movement (when allowance is made for urban transport) is accounted for by the railroads, motor carriers, and other forms of surface transit. (17) The authorization of funds by Congress in 1965 for research on improved rail passenger service in the Washington-New York-Boston corridor helps some to right the balance, but it still leaves relatively too much emphasis on air transportation. Either more should be spent on surface forms and less on the airways, or more should be spent on transportation research as a whole.

Similarly, an explicit programmatic approach might also lead to a reappraisal of the relative priorities of manned versus instrumented exploration of outer space. While there are advantages in using men on space missions, the additional cost it imposes is extremely high. It is estimated that as much as two-thirds of the $20 billion to be spent to place a man on the moon is attributable to the fact that a human will make the trip. Again, admitting that there is advantage in using a man on the voyage but also admitting that we have other research needs and given the fact that we have limited resources to apply to R&D, it might be wise to explore space only with instrumented craft and use the amount of money this would free for other problem areas that are now woefully neglected. (18)

While program presentation of R&D requests would facilitate comparison of the worth of one type of research undertaking with another, extension of the idea to the entire federal budget would permit a comparison of research requests with non-research requests. To suggest that what might be involved ponder the implications lurking in the fact that with the $20 billion it is estimated it will require to put a man on the moon we could do any of the following: give $10 million each to 200 small colleges (or, perhaps better yet, give $100 million to each of 20 institutions in different parts of the country, aiming to make them centers of educational excellence); grant a ten per cent raise to every teacher in the country over a ten-year period; create three new, publicly-controlled equivalents of the Rockefeller Foundation; build complete universities for every new nation added to the UN since its founding. (19) Or we could make a good start toward coping with the many serious problems of the cities—in housing, mass transit and public health. Or we could make a really serious effort to deal with poverty and to cope with the implications of automation. You can assemble your own list of possibilities. The important thing to recognize is that the R&D budget reflects choices among competing uses and it should represent a set of priorities that embodies our social preferences. At present it does not.

Approaches to Reorganization. If rational decisions are to be made in the allocation of resources to R. & D.-if the kinds of questions suggested above are to be articulated carefully and considered thoughtfully-important changes must be made in the organization of the government, particularly in Congress. Ultimately one hopes that the membership of both houses will alter the existing machinery for the authorization and appropriation of funds so that programs can be examined along functional lines rather than solely in relationship to their sponsoring agencies. Medical research projects should be scrutinized by one committee (or subcommittee), regardless of the number of departments which sponsor programs in this sector. The same is true of other types of research.

With this critical preliminary examination completed it would be possible, and essential, for another committee (or a parent committee) to explore the relative needs and costs of the various components of the entire research budget.

With this kind of information in hand an appropriations committee could then intelligently assemble a final budget for the complete span of federal operations. The overriding objective would call for the assembly of a budget structured in such a way that it conforms as closely as possible to the society's values-so that the amount spent for R&D, or whatever, in both amount and type, achieve as nearly as possible the nation's goals.

Any such basic reform in the Congress, however desirable and intelligent it would be, remains at best a gleam in the reformer's eye. Less grand modifications must be developed for the short-run, but changes should be made and soon. With the government's dominant place in the country's research assured for many years to come, it is crucial that techniques be devised that will permit intelligent and deliberate decision-making in the shaping of the federal R&D budget and that will also enable Congress to monitor continuously the situation in all of its complexity.

One step that Congress can take to this end calls for the establishment of a standing committee, preferably organized on a joint House-Senate basis (in the fashion of the Joint Committee on Atomic Energy), to deal with the entire range of government scientific programs. Such a Joint Committee on Science could perform many valuable functions. For one thing it could, working in cooperation with the Bureau of the Budget, organize the federal R&D budget along program or functional lines and present it in this form to the members and other committees together with its own recommendations as to piority. It would, in short, make explicit the allocative choices in making appropriations. The comprehensiveness of its perspective would be extremely useful to a Congress which now reviews budget requests in the fashion of an auditor rather than a planner.

Suitably supported by its own staff, (20) the Joint Committee on Science could perform other useful services as well, such as in the review of the administration of R&D programs and in the assessment of the social and economic effects of governmental research undertakings. With a broad charter to study and make recommendations as to the R&D budget, to monitor the various endeavors, to assess their implications, and generally to inform Congress on the wide range of questions in the scientific area, such a Committee would be of inestimable value in the legislative process. It would permit Congress to act intelligently in an area where action is now taken typically on faith, with a dose of fear and emotion occasionally thrown in for good measure.

If the allocative process that culminates in the federal R&D budget is to be made rational, changes must not only be made in the legislative branch but in the executive as well. Here the need for organizational adjustment fortunately is less since much can be acommplished simply through improvements in the operations and attitude of the Bureau of the Budget itself.

First, the Bureau can help considerably by presenting the R&D budget in terms of programs. It has taken steps in this direction but they have been tardy and inadequate (program presentation now covers only about half the R&D budget). It should promptly complete its work in this respect for it is essential to meaningful appraisal and choice.

Second, the Bureau must become more deeply immersed in the process which produces fundamental allocative decisions that are reflected in the budget submitted to Congress. Naturally the ultimate decisions should and must be made by the President and his political associates, but presently this is difficult to accomplish as an operational matter given the size and complexity of the budget. Under these conditions it is imperative that the staff of the executive office articulate forthrightly the basic value choices that must be made. (21) If this is not done, decisions will be made in a haphazard fashion that omits a proper understanding of the values that are at stake. Assisted by the Office of Science and Technology the Buraqu is in the best position to identify the issues involved and to guide the President and his immediate policy advisers in their assignment of priorities. Of course, this is a two-way street; the Bureau can help immensely, and more than it has to date, but it talents must be thoroughly and cooperatively exploited in this respect by whoever sits in the White House.

Responsibility for the initiation and conduct of a number of federal research programs should also be consolidated in a single executive agency, whether it be a newly created Department of Science (as has been urged by Hubert Humphrey, among others) (22) or an expanded and revitalized National Science

Foundation. In either case only some of the government's R&D undertakings can be wisely housed in a central agency. Many research programs now under way are only marginally related to the principal function of the responsible agencies; they could be much more efficiently conducted under the supervision of a single administrator instead of being divided up among several departments.

Research in oceanography, meterology, and water resources is conducted by bureaus and offices housed in anywhere from eight to eleven parent departments or agencies, depending on the program. Oceanographic inquiry, for instance, is now being conducted by eight departments and agencies, for none of which it is a significant research or operational factor. Even though this work supposedly is being coordinated by the Federal Council for Science and Technology (23), it is considerably more reasonable to expect that it would be more efficiently conducted if it were under the direction of a single agency. The same is true of the work in meterology and water research where in spite of the amalgamation in 1965 of the Weather Bureau and the Coast and Geodetic Survey into a single Environmental Science Services Administration, responsibility is divided up among a number of governmental units, of which most do not regard the projects as major aspects of their primary mission. Supervision of all such programs could be usefully transferred to an agency principally oriented to research.

However, most of the government's R&D at least in terms of dollar effort— is sponsored by departments and agencies which find that it is intimately related to their major responsibilities. The Defense Department for instance carries on extensive research projects; most (but not all) are inextricably intertwined with the operational requirements of national security. Its R&D could not be efficiently or productively transferred to a separate agency any more than you could place the task of economic analysis in a single Department of Economics or all of the legal work in the Department of Justice.

The great bulk of Defense research is so closely integrated with other departmental missions that it could not be removed without seriously impairing performance of the agency's basic missions. Much the same is true of the Atomic Energy Commission and probably also of NASA which, by virtue of its size alone, may simply have to be left to handle its own research. Yet while this is true it cannot be forgotten that there is much research that could be severed and located in a single department or agency without undermining the ability of the current sponsoring agency to carry out its role. Where this is so the work would be better transferred and centralized.

In a proper conception of the government's organizational involvement with science, a new Department or a rejuvenated National Science Foundation would have much to do besides the conduct of a number of research programs. Perhaps most importantly it would have to become a critic of and spokesman for the nation's science. As indicated earlier it is vital that the executive and legislative branches come to recognize that in approving the federal R&D budget they are engaged in a complex process of resource allocation. Choice is an inherent feature of this process, which means that it cannot be intelligently conducted if the complete range of the country's research needs is not portrayed. At the present time agencies such as the Defense Department, NASA, AEC, and the National Institutes of Health articulate their research requirements and then defend them forcefully in and out of the public forum.

But for many kinds of problems that warrant the application of scientific skills there is now no official advocate. This is true of the many social problems of man as well as of certain areas, like earthbound transportation, where the voices heard in Washington (such as that of the Federal Aviation Agency) address themselves to only a small part of the nation's transportation (and this is why the government had until 1965 invested practically nothing in research related to the railroads, even though they remain the single most important form of freight transportation). (24)

Research deficiencies of all sorts should be systematically identified and, where possible and otherwise desirable, programs should be prepared to cope with this kind of need and considered vis-a-vis other research projects. Until some agency begins to perform this job neither the Congress nor the President can make intelligent choices as respects the elements in the federal research effort.

One might think that the National Science Foundation could meet this need. Yet this is not the case as things now stand. First, the Foundation is limited by statute solely to the development and support of basic research, (25) but most of the country's unmet research deficiencies call for a sizable dose of applied research and development. If the Foundation were to become a prolocutor for

unfulfilled scientific requirements its charter would have to be amended to include the authority to sponsor all kinds of research. Second, many of the country's research needs call for work in the social and behavioral sciences, precisely the areas which the Foundation has grievously neglected ever since its creation.

The Foundation's statutory mandate calls for it to promote "basic research and education in the sciences"-all the sciences, not just the physical and life sciences. Yet of the research it financed in fiscal 1964, only five percent of its effort was devoted to the social sciences. (26) It spends nary a penny in political science. (27) With this record of neglect there is reason to doubt whether the Foundation can fulfill the role which has been carved out for it here. Perhaps with a broadening of its statutory charter and with an enlarged staff and larger budget, it could, with active leadership, meet the challenge.

All things considered, however, it probably would be wiser to create a new Department of Science and transfer to it the new responsibilities outlined in this chapter along with the duties now carried out by the Foundation. With Departmental status, the creation of a new agency offers more hope that it could handle the large job before it than there is in the resuscitation of the Foundation. Other Changes in Policy and Programs. Much more remains to be done, of course, than to effect changes in the organization of the federal government as it pertains to science. The review of the government's R&D effort in Chapter V suggested a number of specific steps that should be taken.

First, the narrow distribution of federal funds among the nation's universities demonstrates that there are far too few outstanding institutions of higher learning. To broaden the educational base the Congress should provide substantial additional aid to the country's colleges and universities with a view to increasing significantly the number of distinguished educational institutions. This should not be confined to producing centers solely of scientific excellence (as past proposals of the National Science Foundation contemplated), but of excellence in all fields. Second, the government should encourage and assist in developing new nonprofit research institutes, particularly in geographic regions which have fallen behind the rapid technological pace of the 1960's. Similarly, the government should endeavor to induce companies to industries which do not now engage in any significant research (like textiles and construction) to establish cooperative research institutes.

Third, the government should seek to encourage more civilian research through special tax treatment of expenditures that reflect stepped-up effort. Section 174 of the Internal Revenue Code currently provides some inducement to privatelyfinanced R & D but it is an inadequate stimulus; (28) further, its benefits are available to a firm even if it devotes no more to research than it ever had. If additional research is to be encouraged special tax concessions, such as a credit against tax liability, must be linked only with an increase in expenditure above that reflected in a firm's past research outlays. (29)

Other remedial measures must also be undertaken. As Chapters V-VII reveal, many policies and practices applicable to research and development demand a thorough overhaul if the public interest is to be served. Among other things. efforts must be made to reduce the degree of needless concentration of funds in a handful of large corporations; to do this the extent of competition must be increased and greater contract opportunity given to more business firms. The patent policies of certain government agencies, most notably the Department of Defense, should be altered. And the information stemming from governmentfinanced projects should be more promptly and more fully disseminated throughout the economy, with special efforts mode to insure that the fruits of government research are thoroughly exploited. To achieve these objectives calls for a radical overhaul of existing practice and, in some cases, for new legislation: the rudiments of the requisite reforms are described in the paragraphs which follow.

As the statistics cited earlier vividly demonstrate, the flow of research funds from the principal government agencies to industry is highly concentrated (with respect to the Defense Department three firms account for 23 percent of the total. ten firms for 53 percent). To offset this trend and to broaden the procurement base demands a variety of changes in existing techniques so as to afford competition a hospitable environment. Within the existing framework improvements can be made that will increase the degree of competition and encourage a wider distribution in the award of research contracts.

The single step most likely to accomplish this goal calls for the separation of large research projects into functional pieces that individually are made the subject of competition. This would permit the participation of many small research

organizations that are capable of handling scientific assignments at the early basic and applied research stages even though they lack the capacity to assume responsibility for an entire project. Today they are effectively excluded by virtue of the propensity of the principal government agencies to contract with a single firm that can perform an entire mission of vast size. By breaking such undertakings into their integral parts and permitting competition at each juncture, defense and space projects would no longer be the exclusive province of a few large corporations.

In the last analysis, however, any effort to alter prevailing contract practices by simply tinkering with the procedure involved is probably doomed to failure, just as Congress has been frustrated over the years in its efforts to assure better opportunities for smaller business. The Department of Defense represents the biggest challenge. Here prevailing attitudes are so thoroughly engrained that the only effective answer appears to lie in shifting the responsibility for awarding research contracts out of the hands of the individual services and placing it squarely in a special agency coming directly under the civilian supervision of the Secretary of Defense. (30) This would have several advantages, but among the more important it would immerse the politically accountable officials of the Defense Department in the contracting process; at present the bulk of contracts are awarded (and modified) at subordinate levels within the services, and only rarely does the civilian hierarchy's top echelon become meaningfully involved. (31) By changing this situation the task of broadening the procurement base can be laid on the shoulders of those who are politically responsive.

To urge that the responsibility for establishing and effectuating procurement policy be assigned to the highest civilian officials in the Defense Department is, in essence, to suggest that until it is recognized that in contracting for the performance of R&D (and the rest of procurement) we are dealing with one of the most powerful forces in the economy, fraught with serious social implications, it is idle to talk about the refinements of a change in policy.

As things now stand the job of awarding billions of dollars in research contracts within the Defense Department is widely diffused among uniformed officers of the armed services and associated civilian employees who are buried so deep in the bureaucracy that they are largely immune from the occasional formal admonitions which call for a broadening of the procurement base. Administrations, departmental secretaries, and Secretaries of Defense come and go; but procurement policy continues in its established course, with most of the money involved going to a coterie of giant armaments makers.

In this environment any concern for large public policies is completely absent. Until this environment is radically altered no change of any consequence in present operations can be expected. Real reform can only come when the highest levels of executive leadership in the government face up forthrightly to the problem and make it a primary issue of administrative policy. (32)

Any effort to improve the decision-making apparatus as it pertains to the award and administration of research contracts is placed in jeopardy, whatever its character, if the government is not placed in a position to attract and retain sufficient number of competent scientific and managerial personnel. At the moment those agencies most heavily engaged in research, specifically the Defense Department and NASA are able to accomplish too limited an amount of the requisite work in their own laboratories, even though this might frequently be more efficient than contracting it out to independent performers.

This is bad enough, but in the short run what is even more critical is that the government lacks the human talent essential to responsible supervision of the research contracting process itself. (33) By setting unrealistic pay ceilings for its own workers and by awarding cost-plus contracts to outside organizations the government has so frustrated its contract operations that in several instances it has had to hire private firms to perform its managerial functions. (34) All too often the government's contract administrators are inexperienced, underpaid, and ineffective, using their period of public service as a means of gaining a more lucrative job with industry. Although there are many able, dedicated individuals engaged in contract work on the government payroll, they are too small a group, excessively burdened with detailed work, to correct the situation.

The key remedy lies in a sizeable increase in authorized manpower and a sharp rise in pay levels for personnel engaged in research whether as scientists or as administrators. Without explicit recognition that the conduct of a multibillion dollar operation requires a sizeable, competitively-compensated work force, possessing both technical and managerial skills, the federal government's R&D