Preface On April 20, 1966, the National Bureau of Standards sponsored a one-day symposium whose purpose was to convene operations research practitioners and users in the agencies of the Federal Government so that they might exchange information on topics of mutual concern. The following considerations underlay the planning of the conference, and our hopes for future related activities: (1) There does not exist, but there should exist, a reference inventory of completed and on-going operations research/systems analysis/multidisciplinary research/scientific planning activities in the Federal Government, so that a technical exchange of methods, techniques, benefit measures, cost methods, etc. could be facilitated. (a) There are methods to be applied that are not unique to one agency, e.g., tests of organizational effectiveness. (b) There are suggestive analogies among problem types from agency to agency, e.g., a network of highways and a network of post offices. (c) There are interaction situations, e.g., the full program benefit to one agency may be incompletely calculated due to excluding benefits arising to other Federal missions. (2) The growth in opportunity to test the usefulness of systems analysis in new non-defense contexts underscores the importance of making sure that any initial study has the fullest possible measure of success at an early date, and this might exclude repeating exploratory work others have done. No reasurring encyclopedic memory bank exists. (3) Each of us knows of a dozen or two other studies or groups, but the creation of new groups is making it more difficult and time-consuming to keep up to date. We need a new one-time updating. (4) The introduction of the new Bureau of the Budget Planning, Programming, and Budgeting System (Circular 66-3) places an explicit requirement on agencies to provide systems analysis, benefitcost studies, and supporting documentation. Some agency chiefs have asked OR groups, "Tell me what you can do to help;" the OR groups need to meet with the PPB originators on a technical basis for a fuller understanding of the substantive requirements and desires so that they can help their parent agency to respond adequately to this new system. (5) The problem of identifying measures of effectiveness for agencies is a common one; some have made good progress in this area, and others are searching for ways to make a good start. Furthermore, many agencies have missions whose ultimate measures of effectiveness can be partly stated in common terms, e.g., stimulate the economy, improve the standard of living, etc. Aside from the recent Brookings Institution publications, Roland McKean's earlier book, the BOB bibliography of benefit-cost studies, and a few other attempts of a more nebulous nature, the best information on measures of benefit is scattered among our individual heads and in individual files. (6) The changing semantics and connotative nuances placed on the terms, "operations research," "systems analysis," "multidisciplinary problem-solving teams," and "planning systems," are further confused by actual practice. For example, OR is frequently defined to be identical with what the statistics group does, or with the problems that are sent to the computing lab, or with the management analysis program of an agency, or with just plain horse sense. Accordingly, there does not seem to be a good common dictionary of what to include, or what to exclude from a conference such as this. Each agency has one or more groups with something significant to contribute or gain; these groups are all talking, with greater or lesser degree of specialization, to a common problem-that of providing a rational scheme for identifying and comparing alternatives in terms of their expected payoff and resource requirements. Similar differences exist in the dimension that starts with pure methods research and ends somewhere with "the back of the envelope calculation" or "the horseback guess." We thought it better to invite all these audiences in the first round, and let the agency decide whether the conference was relevant enough to participate. (7) With non-defense OR getting a stronger impetus, there is an obvious need to profit from the lessons learned in the Defense business. There is also the need to make sure that when defense readiness is a relevant measure of effectiveness of civilian programs, it is included; the reverse is also true. Furthermore, because of the scarcity of professional personnel at the present time, it is important that the growth of non-Defense OR not compromise Defense requirements. (8) Procedural and institutional problems related to starting and effectively using OR groups probably have a large degree of similarity from agency to agency. The lone systems analyst who has been instructed to prepare a plan to start a viable activity in the agency should have access to plans that have a good chance of succeeding. (9) There are agency officials who would like to get OR started, and would like some impartial advice, but who aren't quite sure how to get it. Furthermore, they would like to see at first hand what the nature of the projects suitable for such a group might be. (10) The agency OR groups need to have an inventory of external resources available to help him in his problem solving: universities that want graduate thesis topics and support, professors who want a sabbatical, colleges that are willing to carry out a class project on a systems problem for an agency, etc.; contractors and their strong points and experience; expert personnel resources; data sources that could be relevant and could be used; etc. We recently compiled a list of contracting agencies with a Washington office and which advertised OR or systems analysis as a specialty, and ended up with 40 on the first trial. RFP lists should be complete and relevant, and there is no good way of being reason ably sure of this at present. It is not common knowledge how much or on what basis one agency can call on another (e.g., Bureau of the Budget or Bureau of the Census) for substantive help or data exchange. (11) Currently employed methods of quality control on OR studies could be improved if experts in other agencies were known and could be asked to participate in project design or review. (12) There is a need for an experimentation laboratory in systems analysis in addition to the specific agency opportunities. Field experimentation has always been part of a complete analysis, yet for most purposes, a truly experimental opportunity does not exist. (13) There will soon be a need for some analytic scheme to tackle the inter-agency program balance problem, be it a system of rebuttal and debate with common reference terms, or a government management game, or an inter-agency analysis group, or some other method. (14) There is a need for an OR textbook suitable for Government agency use. (15) There may be a need for an explicit mechanism to rotate OR personnel among agencies, to universities and back, etc. etc. (16) There is a need for a census of resources including hardware, software, programs, computers, The recent rapid growth in operations research/systems analysis/multi disciplinary problem-solving teams in the non-defense agencies of the Federal Government has opened a challenge of unusual proportions to the O.R. community. At issue are questions concerning the development of analytic methods, test methods, data systems, and means of drawing inferences in studies whose purpose is to improve the operations of the nondefense agencies of government. Questions related to the definition, the criteria, the measures, and the means of measuring the effectiveness of non-defense programs are but the beginnings of the work ahead. There is the problem of determining the best program mix within an agency whose missions are stated in many different ways and which sometimes appear irreducible to a single common scalar purpose. There is the question of how to determine the effectiveness of a Federal program when it is designed to assist in improving individuals, communities, states, and regions, especially since the instruments of bringing about that improvement are in the hands of many sources of power in the democratic system of rebuttal and debate that characterizes the American society. Accordingly, operations research in the non-defense agencies is confronted with the unavoidable task of discussing issues that are at the heart of our democratic society. It is important that analytical, methodological, and substantive discoveries made in one agency that are suitable for use by another agency be made a portion of the Government memory bank as easily and quickly as possible. One of the aftermaths of the O.R. Conference was an expressed need, as articulated by many groups, for the following two kinds of activity as a minimum. (1) The convening of small informal ad hoc discussion sessions by agencies of the Government to which the O.R. people from other agencies are invited. (2) The maintenance by some agency of an up-to-date mailing list of all those government agencies O.R. groups, universities, and contractors, who should be informed of relevant items, including an up-todate inventory of on-going O.R. projects and people finders in Government. It appeared that each agency might want to take its own initiative in convening informal group meetings of interest to itself. The Technical Analysis Division of the National Bureau of Standards volunteered to perform the second function insofar as it was able to do so. In the course of preparing for the symposium, we had already circulated questionnaires inquiring as to O.R. activities in various agencies. The wide variations in respondents' interpretations of "operations research" precluded the replies' being sufficiently consistent to warrant inclusion in these Proceedings, but the experience gained should prove valuable to us in designing a more carefully phrased survey. Prior to the Conference, a number of agencies mailed us summary statements about the contents of their operations research programs. These were posted at the entrance to the auditorium at the time of the meeting. These forms appear to be a good starting point for the creation of an informal and frequently updated notebook of agency O.R. work that might be circulated widely, both among the agencies and to external institutions such as consulting firms and universities which might assist in solving agency problems. Our current plan is to issue that first set of reports as a separate document, and then to undertake the administrative chore of issuing updated versions of it. We deeply appreciate your participation in this conference and sincerely hope that this volume will serve a continuing useful purpose for you. We hope that it will be but the first of a long series. W. E. CUSHEN, Chief |