(most of whom are now involved in defense work). Senator Nelson's Scientific Manpower Utilization Act (S. 2662) appears to be designed for just that purpose. To make such a bill effective, other things need to be done. I'll conclude by mentioning three relatively minor things, and one major factor.

(1) Systems analysis and systems engineering are concepts with good reputations, which should be guarded and not oversold. There is the present danger of routine programs and proposals being dressed up by throwing in an undefined "systems analysis" phase. One remedy is for the recipient of such a proposal to require a detailed description of this phase; this should at least assure a disciplined method of measuring and comparing alternative methods of achieving the objective. Another is for the recipient to find advisors, possibly a committee of people with some systems experience in defense, electronics, or computer firms.

Eventually, though, the salary structures of government agencies will have to provide for knowledgeable and experienced systems people somewhere in the chain of program or proposal development and approval. Consultants may furnish much of this help, but some in-house competence is essential.

(2) The prospective sources of systems services should prepare now to market or supply them to state and local government customers. The primary source presently appears to be defense industry, although other firms (or non-profit groups) may acquire or develop key systems men and offer as good or better services. The universities are a potential source. Their difficulties, mentioned earlier, may be partially overcome by research institutes or interdisciplinary centers, each maintaining a nucleus of people accustomed to working with each other (and with government agencies) and calling on other interested faculty members as needed. However, few if any of these sources are apt to devote resources to accumulating experience unless support is in sight.

The

(3) State and local governments should prepare themselves organizationally to use a systems approach. It is obviously desirable that the problem-solving governmental jurisdiction cover the physical limits of the problem area. Model Cities Act is somewhat concerned with this; the same philosophy should pervade other urban and regional planning activities. With fragmented local governments, some substantial institutional changes may be required.

The major institutional factor, however, is timing. A change like the proposed Nelson Bill must be adopted well before its benefits are expected or required. Substantial time lags are inevitable between deciding to try systems approaches and achieving their benefits.

All the above-mentioned minor institutional changes requires time. It also takes time to plan for problem-solving capital investment programs at the Federal level (e.g., the ARA and OEO experiences). It will take time for state and local government agencies to learn to use or buy systems approaches to their problem solving. It will take time for many of the sources of systems help to develop interdisciplinary systems teams, whether these sources are industrial or academic. It will take time for defense firms and universities to learn to market civil systems services. It will take time for firms without systems experience, including universities, to put together systems teams.

In conclusion, money needs to be made available for work on systems approaches to problems well before money is available for capital investment in problem solving.

It appears that problem-solving capital may be available in a few years. Either the current Vietnam costs will decline, or the economy will grow to a point where we can afford both Vietnam and some new civil problem-solving investment.

The announcement time of availability of capital will be too late to start applying systems approaches, particularly if it is a sudden announcement, e.g., one seeking to offset an unexpected cut in defense expenditures. The time lags listed above mostly involve getting ready to use the systems approach. The actual systems analysis and much of the systems engineering must also be carried out before CA Day (for Capital Availability Day). If CA Day is several years ahead of us, we need to start funding and accumulating experience on civil systems approaches now.

PREPARED STAtement of Don A. GOODALL, LEGISLATIVE ACTION GENERAL MANAGER, CHAMBER OF COMMERCE OF THE UNITED STATES

The Chamber of Commerce of the United States is gratified by the interest you have demonstrated by holding hearings on proposals which, if adopted, will lead to the application of "systems analysis techniques" to the solution of public problems.

Congress over the past decade has enacted a host of programs designed to solve public, social, and economic problems. But the dimensions of these problems are staggering and becoming more complex. This is readily observable in problems relating to air and water pollution, urban transportation, core-city housing, and the like. And, it is becoming clear that these problems are not susceptible to the traditional solutions. New ways to manage the public business must be found.

For a number of years the private sector of the economy has successfully applied the techniques of management analysis to problem solving. And most recently, as the language of S. 430 indicates, the same technology has been applied to the defense sector of the Federal Government.

There is great potential for bringing the experience and expertise of the private sector to bear on the solution of social problems in the public sector. Demonstrating support of this concept, the Board of Directors of the Chamber of Commerce in November, 1966 gave their support to H.R. 17310, a bill introduced in the 89th Congress, which would establish a National Commission on Public Management. We, therefore, support an identical bill introduced in this session, H.R. 20.

At a time when financial pressures of the war are growing and the whole grantsin-aid system is being seriously questioned, the establishment of a National Commission on Public Management, rather than the enactment of yet another grant-in-aid program, would be the preferable manner for promoting the development of the application of systems analysis to public problems.

PREPARED STATEMENT OF DR. ROBERT W. KRUEGER, PRESIDENT, PLANNING RESEARCH CORP., LOS ANGELES, Calif.-WASHINGTON, D.C.

A. Purpose

I. INTRODUCTION

The United States faces problems of unprecedented complexity as we enter the last third of the twentieth century. The swift pace of technology, population growth, and social change at home and abroad impose extraordinary problems and threats but, equally, they present entirely new horizons of human opportunity. Both the promises and dangers transcend the marketplace and require greater and greater government involvement. Governments at state and local levels as well as at the Federal level are responding-somewhat unsurely and with frequent backward glances-by expending vast sums. The sums now being spent by governments would, only a third of a century ago, have exceeded the Gross National Product. Further, governments have accepted entirely new responsibilities-to reduce poverty; to provide education with less regard to ability to pay; to reduce illness, unemployment, discrimination.

It is essential, with governments spending nearly one-quarter (about 22 percent) of the Gross National Product, that the resources be wisely used, both to avoid waste and to assure that the nation's objectives be effectively achieved. It is believed that Senate bills 430 and 467 will greatly assist in achieving this goal at the state as well as the Federal level.

As suggested in Senator Nelson's letter to me of March 13, 1967, I am submitting these written comments to the Special Subcommittee on the Utilization of Scientific Manpower for consideration of the members of that subcommittee and for inclusion in the hearing record.

My purpose in desiring to be heard is to lend support to Senate bills 430 and 467 by a group of companies that are a major factor in the performance of systems analysis and which to this point have not been heard. These companies

are the for-profit, systems analysis firms whose main function is the application of systems analysis to the problems of our day for clients in government and industry.

With regard to the strong role of systems analysis in these for-profit firms, there is an important distinction between them and the non-profits and between them and the other organization from whom you have heard. One distinction is that, certainly in the manufacturing industry, in government, and in the universities, systems analysis involves only a very small portion of the people in those organizations. Only in the case of certain non-profits such as THE RAND Corporation, from whom you have heard, does systems analysis play a pervading role.

B. Distinction among terms used

In past testimony you have heard very fine definitions and discussions of the meaning of systems analysis from representatives of several organizations. However, I would like to provide some clarification and perhaps point out some differences in meaning which I think are important and which I believe have not been presented previously.

As a general definition, systems analysis involves the search among a great number of alternatives or combinations of different designs for that alternative or combination that best accomplishes a certain established objective, purpose, or goal according to a certain criterion (or criteria), such as a dollar cost or the cost in terms of consumption of some other valuable resource. It is important to note that systems analysis often involves the search for the criteria and the methods for measuring cost and effectiveness in the first place.

Systems analysis, as employed by aerospace firms and by many non-profit firms that are contractually tied to the military, involve the application of techniques that are extremely valuable in problem solving. But the problems solved invariably are associated with the uses of military hardware, the selection of military hardware, and the operation of the organizations which employ them. The objectives which guide these military systems analyses are seldom, if ever, as complex as the objectives of state and local government. For the military, effectiveness rather than benefit is measured, and the goals of the research are most often clearly understood by the service or agency asking for the analysis. In contrast, the goals of government are more complex, are less often easily quantifiable, may be multiple, and are often inconsistent with each other. In addition, it is my belief that not infrequently in state and local government the agency responsible for a given function (and its members) are not always aware of what the goal is, or whether a goal exists, but instead continue to perform a function which is anachronistic, inefficiently performed, and often dispensable. A systems analysis which emphasizes the alternative designs of equipment as distinguished from the other aspects of a total system, while also certainly including them, is sometimes called systems engineering, and indeed this is why this is why this term is more commonly used among manufacturers in the aerospace industry. However, some people in that industry use the expression 8y8tems engineering synonymously (in the broad sense) with systems analysis. A systems analysis which emphasizes the procedural aspects is sometimes called, particularly in the military, operational analysis. A great deal of what is done in the military in systems analysis is of this nature, being concerned with how best to use equipment that already exists.

Cost-effectiveness studies are a kind of simplified version of systems analysis in which the criterion chosen is most definitely cost and in which nonquantifiable considerations of, say, a political or social nature (such as acceptability) are treated separately on the side.

Planning-Programming-Budgeting (PPB) systems involve a special application of systems analysis-especially of cost-effectiveness analysis to the development of efficient, integrated programs covering a number of years. Here, the program includes the entire stream of related equipment, people, and procedures required to accomplish a certain program goal or effectiveness, with the cost of the program intimately related to the description of it.

Systems analysis could a priori be considered in the very broad sense (that has generally been used in these hearings) or in a very narrow sense, such as applied to a very limited field of a particular discipline (as previous testimony has noted)—e.g., the problem facing a spectroscopist in his laboratory of how best to arrange, assemble, and use the equipment he has in order to best photograph the spectra necessary for his theoretical investigation. However, here I will continue to use the broader definition of a systems analysis, wherein usually the analysis does involve a combination of at least technical, economic, and operational considerations and increasingly also involves social and political considerations; I will refer to lesser systems analysis by the expression subsystems analysis. Indeed, one can define various levels as subsystems analysis; in the example I gave earlier, the spectroscopist might be a "sub-sub-sub-sub-systems" analyst.

In the professional for-profit, service type of organization we all are engaged in the broad type of systems analysis like those in the non-profit organizations and we view most of the systems analysis done, say, by elements of the manufacturers in the aerospace industry, as at least one order of systems analysis below that which we do. (Some people in those industrial organizations view what we do as "super" systems analysis.) Indeed, organizations like ours are often hired by aerospace firms to complement their own subsystems analysis, in which they emphasize engineering, with our "super" systems analysis.

II. DEVELOPMENT OF SYSTEMS ANALYSIS

A. History and types of firms

Operations research, particularly operations analysis, evolved during the war and its features were used and developed, particularly by major non-profit organizations, immediately following the war into what is now known as systems analysis. The RAND Corporation was not only the principal developer, but also the earliest developer, of this approach.

Other non-profits attached to certain departments of the Government and some individual ones such as Stanford Research Institute (SRI) soon followed in this effort. About seven years after the war there emerged a number of organizations like ours, private, for-profit, purely professional-service organizations not engaged in production or laboratory development, and emphasizing the application of the systems analysis approach to the solution of important governmental and industrial problems.

The origin of the non-profits in general, as most of you know, was related to the desire to detach work done for the government from day-to-day happenings. There was also the necessity to create a vehicle for hiring the scientific-engineering talent in a highly competitive market at salaries higher than the government civil service structure would permit. These early non-profits were constituted without any attachment to hardware, and were considered, rightfully so, to have the necessary objectivity with respect to government work. However, I want to emphasize, in representing our group of for-profit, purely professional-service firms, that the latter criterion of objectivity applies equally well to for-profit firms, and that, in addition, the motivations produced by the normally constituted private enterprise ownership produce efficiencies in this work that government has come to recognize.

Development of the broader definition of systems analysis followed later in industry, and today systems analysis is primarily done by three major groups: industry, the non-profits, and the private, for-profit organizations of which we are one. Two other groups engage in a very modest amount of systems analysis— first, within the government where, of course, many people understand the systems analysis approach, but where even they, like most government people, are so saddled with day-to-day operating problems that they have little time for the more detailed analysis, and, because of this, hire for-profit and non-profit professional service firms to do this work.

The second other type is the university; it has been, I believe, generally the least likely place for systems analysis work to be done. This is because the

one principal requirement for this kind of work is the existence of a potential for easy development of a well-knit interdisciplinary team to apply this approach; the university consists primarily of individuals concerned almost entirely with their individual disciplines. Indeed, those who seek interdisciplinary activities in universities often leave to join organizations like ours and the non-profits. B. Systems analysts available in the United States today

In order to provide the subcommittee with an approximation of the sources and number of available systems analysts in the United States, I have prepared several tables that represent our best estimates at this time. We have used the best available sources in constructing these estimates, including information developed by the National Science Foundation in American Science Manpower, 1964 and in Scientists, Engineers, and Technicians in the 1960's and also by the Los Alamos Scientific Laboratory in its 1966 National Survey of Professional Sicentific Salaries.

Some information has been obtained by direct contact with several of the firms engaged in this work. I want to point out, however, that these are only estimates, developed primarily to indicate the relative distribution of systems analysts by type of employer and that, in developing these estimates, much subjective judgment had to be exercised by the Planning Research staff.

Examination of these exhibits (see especially Exhibit 1) indicates that there are between 17,000 and 20,000 systems analysts in the United States, providing a great resource for the study of the problems faced by governments.

These systems analysts are well spread throughout the several types of employers, with non-profit organizations accounting for approximately 3,900 (about 22 percent) of the total. Of those in the non-profit organizations, an estimated 1,620 are employed by the non-profit Operating Federal Contract Research Centers (see Exhibit 2), which include Aerospace Corporation, Center for Naval Analysis, Institute for Defense Analyses, MITRE, Research Analysis Corporation, and RAND.

Over 60 percent of the toal systems analysts are employed by the for-profit type of organizations, that is, outside of government, educational institutions, and non-profits. While most of the systems analysts in the "All Other Industry and Business" category are primarily engaged in analysis related to their own industry (e.g., chemical, petroleum, automobile, etc.), the group we have shown as "Business Service Firms, For-profit" have significant-sized staffs that have had experience with a broad range of diversified studies for government and industry susceptible to the systems analysis approach.

These latter companies, most of which are shown in Exhibit 3, have staked their free enterprise existence on the conduct of the systems analysis type of work, and have grown by applying systems analysis to a broad spectrum of problems. Systems analysis is the major function of their activity. Each of the other major groups of companies we have shown in Exhibit 1 has either a minor interest in systems analysis or is supported primarily by the government. As an example of one of these systems analysis companies, I take the one I know best-my own organization—Planning Research Corporation. Although it is one of the largest of the 21 companies listed in Exhibit 3, its growth pattern is illustrative for most of them. Planning Research Corporation was founded in 1954 with a very minimum of capital and was then regarded as a spinoff of The RAND Corporation, since all five of the founders of Planning Research, including myself obtained their experience in systems analysis at RAND. The cumulative growth rate of Planning Research over the past ten years has been 42 percent, and each year the firm has shown a profit. The corporation stock was listed on the American Stock Exchange in 1964 and currently there are more than 3,000 shareholders. At the end of its last fiscal year, the Corporation's contract revenues were $15,000,000 and current revenues, on an annualized basis, are running in excess of this figure. Planning Research employs 850 people, 600 of whom are professionals. Approximately half of the professionals are qualified systems analysts.