advance in our system for storing and communicating information of all kinds. I refer, of course, to all those kinds of things that we would bring into this conversation if we were talking about computerized information nets, automated libraries, electronic data centers, and so forth. And, of course, I include in this development ordinary communications as represented by publication, both in the traditional sense and, in the future, in new ways through the information transmission net. These developments may even affect the nature of our conferences and meetings in the future, if and when we can afford to interact frequently and effectively over a communications system and without need to gather together physically for such interchanges. I shall narrow my remarks even further, to a very special aspect of this newly developing national and international information network. I refer to the storage and transmission of raw data, or summaries of data, or formal models that represent large bodies of data compactly, and all other "facts" that can be stored. and made accessible quickly and efficiently through such information network facilities. Important among these many forms of stored models are the increasingly powerful computer programs that we can all share in use while profiting from the systematic synthesis they provide, often as a result of the work of a quite small group of persons. More specifically, I propose as well worth your careful consideration, the possibility that all operations researchers throughout the federal government can look toward the development of great information files that will be accessible for all who have a real need. This constitutes a very important new form of communication, since the information stored by one group need be made available to others only after they recognize a real need for it, and then in a form where it has been precisely described and is easily used. Some may say that my proposal is equally applicable to many other kinds of persons and activities. Although this is true to some extent, and is certainly true very importantly in some instances, I am arguing that the analysts throughout the government are in the best position to develop these information stores so that they are indeed reliable and can indeed be used easily. Nor do I mean to limit this to relatively simple tasks such as the standardization of accounting data, or programming and budgeting data, important as these are; rather, the intent is to include as data such things as a highly perfected and validated simulation of a major agency or operation within the government. I hasten to remark for the benefit of any who may feel that I am far too optimistic about such hopes, that I would expect this to be a gradual development but will be very surprised if it is not clearly significant within ten or fifteen years from now. Let us consider the proposal from a rather different point of view, stressing the impor tance of working toward a systematic understanding through modelling, of larger and larger portions of the Federal government. It is probably impractical currently to construct a valid mathematical model, or useful computer simulation, of the entire Department of State or Department of Justice, or even of major subordinate bureaus or agencies. But I would not consider it naive, or undesirable, to begin the attempt immediately. I am reminded of the story about President Jefferson when he was impatient with his gar dener about planting a tree and immediately responded, when reminded that the tree would not reach maturity for 20 or 30 years, that this made it all the more imperative to plant it that same day. It is in this spirit that I urge the importance of keeping this larger goal in mind over the coming years, as you all meet with success in analyzing and understanding larger and larger segments of Federal governmental activities. " For example, and from a purely practical standpoint, I would think it not too soon now to begin the development of models that might eventually enable us to evaluate tradeoffs between such major activities as education and justice. More specifically, it should be very instructive and helpful to make a study that attempts to evaluate the tradeoffs between another million dollars for the FBI in contrast to another million dollars for educational research, or for care of the aged, or for foreign aid, or for any one of the other major national programs that surely are all directed toward the same set of national goals. This is not at all to say that such work will be easy, or yield imme Idiate results, but it will happen sooner and better if an early attempt is made to develop the study activities in a way that will increase the likelihood of eventual success. How might some relative simple steps be taken now to promote progress of the kind I have proposed? Obviously there are many pres ent activities that are substantial and already making contributions in this sense. In par ticular, the activity that Mr. Haldi represents, I will surely yield very important ingredients in Ideed for just such tradeoff evaluations. J.C.R My primary suggestion is that an explicit attempt be organized and made to accelerate this advance by cooperation among all of you in the development of a computerized informa tion network that will store such information and make it much more readily accessible) throughout the Federal government for use by en operations researchers, and for others having need for such a resource. As J.C.R. Licklider often puts it, you can create your own online intellectual community. My own first association with an operations research type of effort within the Federal government was a little over 30 years ago when I worked with a distinguished panel of technical experts on an advisory task for Attorney General Homer Cummings. Our panel was part of a broad study under Dean Wayne Morse, now Senator Wayne Morse, that was intended to bring expert advice to the Attorney General in connection with his broad responsibility for parole and release procedures. This panel was chaired by the late Professor Samuel Stouffer, eminent sociologist, and included among others, Samuel Wilks and Frederick Stephan, both distinguished statisticians. The panel also included criminologists, economists, psychologists, mathematicians, and lawyers and was supported by a strong staff group. We could barely discuss basic questions such as the negative social value of a released murderer, or other criminal, as compared with the social value of keeping him in custody or under parole supervision. It was my impression then, at a time when I was engaged in studies of this kind in New Jersey, that we knew essentially nothing about major social questions of this type, nor did we have the talents or the trained personnel to tackle them. From what little I know at present about this field, I suspect that we remain pretty much in that same position today, except for the fact that we certainly have new talents and much more expert personnel who could tackle such problems-perhaps with considerable success. The newly established National Commission on Crime, and similar commissions that have been established recently in some of our states, reflect a growing public awareness of the importance of dealing with major social problems more systematically and more effectively. will all move unless steps are taken by the Federal government and industry to anticipate and correct some of the glaring deficiencies in our communications system and in our laws and administrative practices. I suppose that none of the difficulties I foresee has already been observed and discussed somewhere within government or industry, but I concluded on the basis of my study that a great deal could be done to improve the situation. My most recent operations research kind of service for the federal government was a study that I completed this past September for the National Commission on Technology, Automation and Economic Progress. My assignment was a study of the national information network development, or the "computer utility" as it is sometimes called at MIT by Professor Fano and others. My report is just now being published by the National Commission, but Dr. Cushen has asked me to comment regarding Some of its conclusions pleased to have this chance to do so. -I am, of course, - Briefly, I am impressed by the great opportunity afforded our nation in terms of the rapidly developing information network, which is a good and exciting prospect indeed, but I am disappointed by the slow pace at which this For example, the common communications carriers, such as the American Telephone & Telegraph Company and the Western Union Telegraph Company, are moving very vigorously to develop and construct a national communications network that will permit the speedy and efficient transmission of data as well as voice messages. However, it is generally agreed by experts on communications systems, who are also knowledgeable about data transmission systems, that this must be a "store and forward" system, rather than a system which merely provides an exclusive line connecting two users for some period of use. In other words, data will originate at many individual nodes on the network, then be transmitted to intermediate nodes where it will in turn be sorted and re-routed, perhaps after temporary storage, and eventually reach "terminal" nodes, where it will often be processed and sent back through the nodes of the system, often through quite complex paths. All of this requires new kinds of electronic switching centers, computerized to handle, store and forward activity of this type, and industry is moving ahead to provide this type of service. Unfortunately, data transmission demand will surely exceed transmission capacity in the very near future. The need is to take steps soon which will minimize this disparity between capacity and demand in future years; my report suggests a few moves that might be made toward this goal. For example, there seems now to be no single Federal agency that is clearly and obviously charged with responsibility for insuring that this important development does take place within our nation at the proper pace. Another example, and a lesson learned from experience with some of our major time-shared computing systems, such as one at MIT and another at the System Development Corporation, is the importance of bringing the future users into the system early to help ensure that its developing direction is a good one. Toward these ends, the report proposes that at least one major national information network be sponsored by the Federal government in the near future and that several major systems be made available in the immediate future for use by individual citizens, as students, or housewives or merchants, and so on. The report also stresses a number of social problems that require changes in legislation, or in regulations, or in administration, if we are not to suffer from too leisurely development of the national information network facility. The report also urges that the Federal government look very carefully at its own internal operations to ensure that this type of network capability is made available and used appropriately soon by Federal agencies. It is in this spirit, within the Federal government, that I believe there is an unusual opportunity for analysts to cooperate in specifying the nature of such a system in order that it meet their needs later. Meanwhile, there will be many opportunities for all of you to exert your influence through usual channels as this developing information network progresses within the government, and for your agencies. It may strike you as strange that I have devoted most of my time to a proposal that rests so heavily upon recent technological developments in information networks within the Federal government and elsewhere. I have done this partly because of my recent experiences with such activities, but primarily because I have been deeply interested in these developments for I feel that they are so centrally important for the future of scientific and analytical efforts to improve operations and management within the Federal government and elsewhere. I close by repeating my opening remark, that the special opportunity that seems to me to be available to all of you working within the Federal government on problems primarily within the civilian sector is the relative openness of this system and the opportunity you have to communicate with each other regarding your work both while it is in progress and after it is completed. Furthermore, I see the developing information networks providing you with a truly advanced method of communications, or information transfer, since you can in the future share your experiences and your data by storing them in the information network and taking advantage of that store as you each proceed with your own work. If the information network facility will be with us in the near future, then I predict that you will be sorely pressed to develop good methods in time for preparing your experiences for storage in that system, so that they can in fact be useful and be used by others working on similar problems. Just as President Jefferson remarked with respect to his tree, since it will take so long to finish, you should begin immediately. The Planning-Programming-Budgeting System Dr. John Haldi Program Evaluation Staff I welcome this opportunity to discuss the Planning-Programming-Budgeting (PPB) system with you. Historically, suggestions for Federal program budgeting date from 1912, when a Commission reporting to President Taft proposed that a "budget bureau" be established and that Government agencies present their budgets along program lines. More recently, two Hoover Commissions have recommended the improvement of Federal management through the adoption of program budgeting. Before Mr. McNamara came to the Department of Defense, many agencies already had some form of this type of budgeting. Many facets of PPB are in use in certain Government agencies. It is probably the combination of Planning, Programming, and Budgeting into an integrated system which is most new to all Government agencies. The system, which most economists view as a straightforward application of common sense, has a few fundamental concepts. All agencies are being required to: (1) define their activities by objective within what is called a program structure, (2) prepare a multi-year program and financial plan, (3) translate program decisions into the conventional appropriation format, and (4) provide for periodic reappraisals of their major programs and objectives. The first step in the installation of a PPB system is definition of outputs and drawing up program structures or outlines grounded on those outputs. For the last several months all major agencies have been developing program structures focused on what the agency produces, whether for the benefit of the public or of other Government agencies. The program structures include categories, subcategories, and program elements. Program elements are the smallest units in the program structure of an agency. is to serve as a tool for top management review. The major purpose of the program structure We hope that it will lead managers to ask the "right questions"-as President Johnson has said, to reexamine the programs to see if we are really doing the right thing, or to see whether we are actually funding obsolete programs. The primary focus of this exercise is to make sure that we are selecting the right alternative among plausible policies, and directing resources to their best use. A program structure, intended primarily for consideration by the Secretary, the Director of the Budget Bureau, and the President, must be aggregated into a few meaningful program categories. Here are illustrations of program categories. Example 1: A program category in the Department of Defense is "Strategic Retaliatory Forces." This program consumes billions of dollars. Subcategories within it are "Offensive Forces" and "Defensive Forces," and Offensive Forces are broken down into sub-subcategories, such as "Manned Bomber Systems" and "Missile Systems." Program elements in the Defense structure are B-52's and Polar Missiles. In the Defense Department structure, the element "Polaris Missile System" is the lowest level of detail shown. Yet this element has a budget which is bigger than many total agency budgets. Included in the Polaris Missile System element are not only the missiles, but also the submarines needed to keep these missiles on station, the men that man the submarines, and the tenders and other supply facilities necessary to keep these submarines operating. Thus the program element is an aggregate of many things. This illustrates the total system concept. It enables management to see at a glance how much of its resources are devoted to this specific element. The program element indicates the purpose for which these resources are brought together, and it provides a focus for the measure of performance. Management's attention is thus focused on the performance of the entire weapons system and not on the performance of support items. Example 2: In many instances, Government agencies do not produce tangible products. The Department of Justice, for example, produces no hardware. Yet here, again, program budgeting means deciding what the money should be spent for. That is, the purposes for which the program is engaged, and what the program accomplishes. The Department of Justice program structure would include programs dealing with anti-trust declaratory policy, investigations and prosecutions under O.R. relating to other laws which are the direct responsibility of Justice, investigations for other Government agencies, etc. Example 3: The program structure for the Forest Service includes such major categories as timber production, wildlife, grazing, and recreation. Within the Department of Agriculture also, there are services directly providing outdoor recreation facilities, and also two more services indirectly providing recreation through conversion of crop land. This year, for the first time in Federal budgeting, the Department of Agriculture has pulled together all its efforts on recreation and reviewed and analyzed them at the Secretary level. This review is one immediate benefit of program budgeting. This Recreation example shows how program budgeting can shed light on Government programs. For example, the Recreational Advisory Council produced a study which showed that about 80 percent of all Government expenditures on recreational activities are made west of the Mississippi whereas about 75 percent of the people live east of the Mississippi. Because of PPB, we have now changed the emphasis to providing more recreation facilities in the East. Example 4: Some agencies, like those just a mission-oriented program discussed, use structure. The Post Office Department, on the other hand, has a program structure oriented toward function, because that seems more appropriate for their needs. The four most important categories in the Post Office program structure are: (1) Getting Mail into the System, which consists of two primary subcategories window operations and mail collection services; (2) Processing the Mail; (3) Transporting the Mail; and (4) Delivering the Mail. In addition, there is the catch-all "General Support" category for overhead, "Processing the Mail" is an intermediate operation which might appear to be merely part of a broader system. Why was it made a major category? The reasons are: ---Over $2 billion is spent to process the mail -Mail processing involves many interest- Although some agencies have outputs which cannot be measured in physical terms, almost all programs should be output-oriented. This method shows what is to be accomplished, not merely the old accounting of object classes, or Setting up a structure requires rigorous con- Once the structure is set up, the agency begins developing a multi-year Program and Financial Plan (PFP), which specifies the outputs and cost estimates for each program element. The PFP time span can be as short as five years beyond the current Fiscal Year and the next Fiscal Year, or as long as ten years for the Corps of Engineers' water projects, or even 80-100 years for the Forest Service planting program. In the PFP the agencies are asked to reflect a comprehensive plan, corresponding to the program structure. The PFP is the inte grated financial document for the agency. Three sets of tables make up the PFP: Part I sets forth outputs for each year by program element; Part II shows dollar costs associated with each program element; Part III shows whatever other information the agency determines is relevant and necessary, including alternative measures of outputs. Agencies that have not been obtaining such data all along may soon be making somewhat greater demands on field personnel than in the past. Analysis is the cornerstone of the PPB system. We are giving this major emphasis forming planning and analytic staffs at the Secretary level where they do not exist, and improving those which already exist. Without much better analysis than we have previously had, this whole effort would fail. This year the budget submission for the spring preview will consist solely of the Pro gram and Financial Plan (PFP) and Program Memoranda (PM) in support of the PFP. Th Program Memoranda are to be hard analyti documents which deal with the issues confron ing the agency and which propose alternative for study during the year, before critical de sions are made. I stress the word "altern |