Senator NELSON. It has also been suggested here that it might be possible to set up by region some teams of systems analysts who would be available as advisers, consultants, and so forth, in areas broader than one State.

Does that idea seem to have any validity?

Mr. GEYER. I think when the problem crosses State boundaries it is almost necessary to do this. There is no point, really, in studying air pollution in Manhattan unless you look at what is going on in New Jersey and Long Island and pretty soon may be you will have to look at what is going on in Philadelphia and Baltimore to see what the effect on the character of the air is on up the line. Regional studies are very desirable.

Senator NELSON. I appreciate very much your excellent testimony and appreciate your taking the time to come over here and give us your help this morning.

Mr. GEYER. I thank you for the opportunity. It has been a pleasure.

Senator NELSON. Thank you.

We will take a 5-minute break for the reporter.

(Whereupon a brief recess was taken.)

Senator NELSON. Our next witness will be Prof. Kathleen Archibald, assistant director, Public Policy Research Organization, University of California at Irvine, Calif.

Professor, we appreciate very much your taking your time to appear before the committee today.

You have a prepared text, do you?

STATEMENT OF MISS KATHLEEN ARCHIBALD, ASSISTANT DIRECTOR, PUBLIC POLICY RESEARCH ORGANIZATION, UNIVERSITY OF CALIFORNIA, IRVINE, CALIF.

Miss ARCHIBALD. Thank you very much, Mr. Chairman, for the opportunity to appear before the subcommittee.

I thought I would just summarize the main points of my statement in speaking before you. I think probably the two points that may be of most interest in connection with the Public Policy Research Organization at University of California, Irvine, are

Senator NELSON. What does that mean, by the way, "Public Policy Research"? What is the area-everything?

Miss ARCHIBALD. In general, what we are talking about is an interdisciplinary, systems approach to problems of public policy. By this we mean a careful look at the alternative choices in public policy, including the reformulation of policy problems and possibly the invention of new alternatives. So, in other words, generally, systems analysis and, in particular, systems analysis as that term has been used by the Rand people.

In terms of the areas we plan to work in, this is open at the momentprimarily in domestic areas. We also expect, though, to do some work in international security and international development.

I am afraid I cannot give you information derived from prior operating experience because we really are just beginning to get going, but in planning the organization

Senator NELSON. This is one department within the institution. Miss ARCHIBALD. Right; it is not connected with any particular academic department, it is an all-campus organization and actually our charter suggests we also draw on faculty resources from the other campuses of the University of California, so in that sense we are all-university as well as all-campus.

Senator NELSON. You may read and/or extemporize, proceed as you would like to do.

Miss ARCHIBALD. Fine; I think I will start by reading and shorten it somewhat.

People interested in the problems and potentials of policy-oriented research have often noted that there may be a need for new organizational forms to encourage and conduct such research and to aid in the dissemination of its results.

For instance, this was mentioned in the Bell report, prepared by the Bureau of the Budget several years ago; it was also mentioned in the testimony before this subcommittee in May of 1966.

What may be of particular interest to you is the fact that the public Policy Research Organization is a new kind of structure. It differs from other existing research organizations in several ways. It is a hybrid, falling somewhere between the traditional university research institute and the nonprofit advisory corporation.

It has two primary objectives: (1) to undertake long-term programs of policy research and analysis in several substantive fields, such as, education, international development, and social welfare; and (2) to establish educational and training programs in the field of policy research and analysis. The training program will be connected both to the academic curriculum within the university and to ongoing policy research programs within the Public Policy Research Organization.

I mentioned earlier what we mean by policy research and analysis, so I won't go over that again. I might add, however, that there is a problem of labeling. The field is still young enough that people use different labels to refer to it, or sometimes the same label to refer to different things. I noticed in listening to Professor Geyer's testimony that he was referring more to what I would call systems engineering and I think a distinction can be made between systems engineering and systems analysis. Systems engineering is a much more established field. For instance, I will mention later some of the questions we have about how you train people in systems analysis or policy research, whereas in systems engineering they know how to train people and they have been training systems engineers for some time. The more social science side of the systems approach is a newer field and it is a little bit more difficult to see how you should go about training.

This brings me to another point of difference between PPRO and existing organizations. This is in the approach we hope to develop and the way it will differ from at least the past work of existing organizations. The disciplines most involved in the development of systems analysis have been economics, engineering, statistics, and, to a somewhat lesser extent, mathematics and some of the physical sciences. The social sciences, other than economics, have been very little involved in the systems analytic approach although they have

been developing their own distinctive approaches to applied work, particularly to practical problems at the level of the community and smaller groups.

We hope it will be possible to develop a better articulation between systems analytic approaches and the social sciences, since the special expertise of, for instance, sociology, anthropology, and social psychology, seems to be particularly relevant in understanding and coping with many of our domestic problems.

To give you a little background on this program: The idea of establishing a new kind of structure to conduct policy research and analysis developed out of a feeling shared by many people within the University of California that the university had both increasing opportunities and increasing responsibilities to provide expert advice on public choices between and among policy alternatives.

Ex-President Clark Kerr set up an ad hoc committee, composed of faculty members from all campuses of the university, to explore the desirability, feasibility, and possible design of a policy research organization. In his letters of appointment to committee members he stated:

A structure for policy research should be complementary to, rather than competitive with, the teaching and basic research functions of the University. It may need to be quite different from our usual research institute in its relations, both to the world outside the University and to the people within the University. At the same time, it should be designed in such a way that it can help and not hinder the research of individual faculty members and the education of graduate students.

This committee was chaired by Roger Ravelle, who is now at Harvard and was then the dean of research at the University of California, and I served as a senior staff member to that committee.

What we set out to do was first to attempt to establish the conditions that have facilitated the conduct of policy research and analysis by looking at the past experience of other organizations. We looked at programs that had been successful, and programs that had been unsuccessful, and tried to account for some of the differences.

We paid particular attention to the nonprofit research corporations which had established the field of systems analysis and had done the best work in the area. The preeminent example here seems to be the Rand Corp.

We then looked at the problems of conducting this kind of work within traditional university structures. Finally, we attempted to design an organization that would be linked closely to the university but would also reflect some of the best features of the nonprofit corporations.

I will outline the basic requirements we identified as being important to the conduct of good policy research, and then mention some of the difficulties we felt would be encountered in meeting these requirements within a standard university setting.

1. Independence. Independence for a research organization is important to best serve the client's needs, to obtain a good staff and maintain their interest and morale, and to contribute to the long-run welfare of the Nation.

Independence is not simply the freedom to tell a client something different than he wants to hear, nor is it necessarily the freedom from

having any client. It is the freedom to say "No" to some jobs and some clients and "Yes" to others, as well as the ability to initiate research when no client is currently interested in supporting it.

It should be mentioned that having independence is not quite enough, a policy research institute also has to be seen as being independent, to have an image of independence.

2. Stability. Systems analysis or policy analysis has been at its best when performed within the context of a long-term, continuing research and advisory operation. People from different disciplines have to learn how to work with each other-or, perhaps more accurately, whom they can work with. The organization has to develop some depth of expertise in its areas of specialty. A working relationship with appropriate policy and operating agencies has to be developed; and decisionmakers have to develop their own methods of utilizing analyses. All this takes time. The one-shot project is seldom much help to anyone. Thus, a policy research organization under ideal conditions should have the kind of stability needed to think in terms of 5- or 10-year programs in particular areas.

3. Flexibility. Probably the most essential aspect of flexibility is being able to modify the size and composition of the professional staff to meet the needs and demands of changing research commitments. With a very large staff you get into a problem of getting enough contracts to meet the payroll. If you have a very small staff you become very limited in the kinds of work you can take on. We thought the ideal arrangement might be to have a small versatile staff, particularly skilled in the methodology of policy research, and then a larger pool of people with diverse expertise that can be drawn upon fairly flexibly. We hope to draw upon the expertise of faculty members on all campuses of the University of California in this way.

Senator NELSON. You are planning to do contract research?

Miss ARCHIBALD. This is planned to be primarily a contract research corporation, the initial support, the "seed capital" comes from the University of California.

Senator NELSON. Does it remain an integral part of the university? Miss ARCHIBALD. Well, the word we have been using to describe it is to say it is a semi-independent part of the university. Perhaps if I could leave that for the moment-I do get into it more specifically later, as to how it looks different from the standard university institute which is our standard research unit.

Senator NELSON. All right, if I don't understand then I will ask this question again.

Miss ARCHIBALD. Fine, I think that will be answered later.

4. Significance. It perhaps goes without a saying that one requirement of a policy research program is that it do significant work, that is, work that contributes something of importance to the policy issues of our day, or of tomorrow.

There is no way of guaranteeing significance but there are ways of encouraging it: by having first-rate research people; by maintaining good external relations, particularly with responsible decisionmakers; and by providing a setting that facilitates significant work. Two facets of such a setting are (a) to have the flexibility to move into new lines of inquiry when problems are anticipated but before they are

salient as policy issues, and, (b) to maintain several major interdisciplinary programs concerned with long-term policy issues or problem

areas.

This is not meant to suggest that short-term projects are intrinsically of lesser significance, nor that a policy research program should avoid working on short-term projects. Both short- and long-term policy research is important, but long-term work takes on greater significance because it is more difficult to handle intellectually and to sustain financially and thus tends to lose out by default.

5. Interdisciplinary work. For most policy problems, interdisciplinary research is required and this commonly means a team approach. Although the interdisciplinary approach is often accepted in principle it is difficult to realize effectively in practice. Because it "does not come naturally," a policy research program has to pay some attention to the means of facilitating interdisciplinary work.

6. Access. The main purpose of policy research is to contribute to policy and this means having access to policymakers. A policy research program should have both formal and informal avenues of access: formal, through doing at least some cooperative research, that is, research on contracts; and informal, through individuals on the staff who are known and respected by policy officials. Time and effort must be expended on both internal standards and external relations to maintain effective and mutually beneficial working relationships with clients.

When we looked at the difficulties or ease in meeting these requirements within a university setting it very quickly became clear that standard academic organization and academic tradition poses problems for policy research.

It was not by accident that systems analysis developed within the nonprofit corporations rather than within the universities. The key problem is that the organization of universities is very largely determined by the traditional disciplines and the disciplinary boundaries can often be quite rigid.

Further, the incentive structure of the university primarily rewards those faculty who contribute, through publication in scholarly journals, to the advancement of basic knowledge in their own discipline. Not only does policy research tend to be interdisciplinary but also it tends to have a lower publication-to-effort ratio than pure or basic research.

Research within universities is typically conducted under the administrative aegis of a research institute or center. These organized research units do not reflect disciplinary boundaries in the way that departments do; nevertheless, they do not seem to provide the requisite conditions for the conduct of interdisciplinary research and analysis. The majority of university research institutes and centers put primary emphasis on individual research and scholarly publications. They often become the center for multidisciplinary discussions, but they are seldom the source of interdisciplinary teamwork. And research that combines the efforts of both the physical and social sciences is almost nonexistent in university settings.

This suggested to us that an organization designed for the conduct of policy research would have to deviate from the standard institute pattern.