Graduate Training in Economics. Report of Panel Discussions at Yale. 1956
During the fall and early winter of 1954-55, Richard Ruggles and colleagues in the Yale economics department organized a series of interviews with representatives of business, government, international organizations, and universities to review the ultimate goals of a graduate education in economics and to identify future desirable directions the evolution of economics training might take. The interviews were followed by panel discussions in the Spring of 1955 attended by, among others, seven future economics Nobel prize winners. Today’s posting is a transcription of the final report printed in 1956.
I came across a preliminary draft of the report in the Milton Friedman papers at the Hoover Institution Archives filed among his correspondence with Richard Ruggles and wondered whatever happened to the project. The report was never really published and survives as part of the “pamphlet literature”. Only recently did I find a printed copy of the final report in John Kenneth Galbraith’s papers in the John F. Kennedy Presidential Library. The relative obscurity of this report can perhaps be attributed to its “Smoothie” style that has managed to blend panel member ideas and opinions into mere minutes of discussions sans quote or illustration. The report’s temporal proximity to the 1953 Bowen report (Graduate Education in Economics, AER, September 1953) could have left journal editors cold as well.
Since the primary goal of Economics in the Rear-view Mirror is to assemble artifacts to help us follow the historical development of the education of economists in the United States, the Ruggles Report of 1956 is worth rescuing from its undeserved obscurity in archival vaults.
GRADUATE TRAINING IN ECONOMICS
A Report on Panel Discussions at Yale
A restudy of graduate education in economics has recently been undertaken at Yale, with the aid of a grant from the Ford Foundation. This study involved two steps. First, economists in universities, government, and business were interviewed to determine what they thought the major problems in training economists were at present. These views were summarized in the form of an agenda, which was then discussed by five panels of economists. This report presents the views of the panel members, as developed in these discussion groups.
The following people participated in the panel discussion and in the revisions of the report.
Robert Adams, Standard Oil Company of New Jersey
Sydney Alexander, Massachusetts Institute of Technology
Kenneth Arrow, Stanford University
G.L. Bach, Carnegie Institute of Technology
William Baumol, Princeton University
E. G. Bennion, Standard Oil Company of New Jersey
Henry Bloch, United Nations
Howard Bowen, Grinnell College
Sune Carlson, United Nations
Gerhard Colm, National Planning Association
Ross Eckler, Bureau of the Census
Solomon Fabricant, national Bureau of Economic Research
Milton Friedman, University of Chicago
Albert Hart, Columbia University
Leonid Hurwicz, University of Minnesota
Dexter Keezer, McGraw-Hill Publishing Co.
Simon Kuznets, Johns Hopkins University
Stanley Lebergott, Bureau of the Budget
Wassily Leontief, Harvard University
Ben W. Lewis, Oberlin College
John Lintner, Harvard Business School
Edward S. Mason, Harvard University
James Nelson, Amherst College
Donald Riley, Bureau of the Budget
Paul Samuelson, Massachusetts Institute of Technology
Robert Strotz, Northwestern University
Clair Wilcox, Swarthmore College
Richard Ruggles, Chairman
The Role of Graduate Education in Economics
THE OBJECTIVES OF GRADUATE EDUCATION IN ECONOMICS which were most frequently mentioned by the panel members were (1) to develop economists who can push back the frontiers of economics; (2) to prepare economists for teaching, not only at the undergraduate level but also in graduate economics departments and business schools; (3) to train individuals who are capable of carrying out research for business, government, labor, and other research organizations; (4) to develop economists who can serve in policy guidance positions in business, government, and labor unions. The panel members agreed that the curriculum of graduate education in economics can no longer be organized exclusively about scholars; it has become essential to produce economists who can do, not just know. Primary emphasis in the past has been placed upon the production of teachers, and although this is an important function, focusing on it may develop a more restricted concept of education than is appropriate today.
The frontier of economic knowledge.
The continual emergence of economists who are capable of contributing to the substance of economics is essential for the vitality of the field. Of course, every student who goes through a graduate school should not be expected to make such a contribution; many are needed to practice the art and science of economics for more immediate objectives in teaching, in applied economics in business and government, and in less basic research in the academic world, business, and government. Nevertheless, the graduate school program should be such as to encourage research of a basic nature and to acquaint students with it. Only by such investment can economics be expected to develop. Such an orientation is useful also for those who do not go on to make substantial new contributions. It provides a  necessary perspective as to the current status of economic knowledge and the bases on which it resets, and points up gaps in economic knowledge and the process by which the evolution of economic thought comes about. Accent on the encouragement of basic research should not be construed, however, as implying that large amounts of learning and scholarship should be the aim. Rather it implies that the creative talents of the individual should be stimulated, and that the individual be trained in the necessary tools to do such research. These aims are complementary to the other objectives of graduate training, not competitive with them.
Research training for business and government.
In recent years, there has been an increasing use of economists for research purposes in business and government. Projections of future demand, analyses of the impact of various market forces, problems of taxation and government expenditure, analyses of productivity changes, studies of business fluctuations, and various international problems related to trade and foreign economic policy all have required that a considerable amount of economic research be carried out. Graduate schools have not generally taken specific cognizance of the needs of these groups so that new Ph.D.’s going into these areas often require a considerable training period before they become useful to their organizations. When the organization does not have available senior staff capable of carrying out such training on the job, the result is that lower grade work is turned out. It is recognized, of course, that schooling cannot entirely substitute for experience, and that some training on the job will always be necessary, but the question still remains whether the present graduate school training is as appropriate as it might be for meeting the research needs of business and government.
Policy and administrative guidance in business, government, and labor.
Besides the technical research uses of economists in business, government, and labor, economists are needed in a more operating  capacity, where day-to-day decisions and advice are required without any formalized research work. Advisors are required at the policy level in large corporations. Banks, insurance companies, large manufacturing firms, and labor unions are employing more and more people in this capacity. Government and international organizations need trained economists to serve as administrators of various programs. These needs are growing in importance as the complexities of economic life increase. Again, most graduate schools have not been particularly attuned to meeting this sort of need.
To a very large degree, teaching is a derivative of the other purposes of economic training. Teachers should be expected to be able to teach those things which are useful in the training of economists. Thus, at the graduate level the objectives outlined above would be pertinent; teachers should be trained to meet these objectives. The problem of undergraduate teaching of economics may at first appear to pose somewhat different requirements, but closer examination indicates that its objectives should be closely allied with the objectives cited above, lest it become too academic and unrelated to the current practice of economics. Undergraduate teachers need to be trained broadly and to have a good general perspective about economics. The development of teachers who are interested in the furthering of economics as a science is necessary in order to prevent the teaching of economics from becoming a sterile academic exercise. The crucial question here is the ability to teach effectively, and to keep on doing it through time—to keep alive, stimulated and stimulating.
Requirements Posed by the Objectives of Graduate Training in Economics
THE OBJECTIVES OF GRADUATE TRAINING IN ECONOMICS are largely complementary in the requirements they pose; there seems little ground for suggesting that individuals expecting to go into different areas of economics should have greatly different and unrelated programs. It was thought that the basic requirements common to all the objectives could be classified into four major categories: (1) a common core of economic knowledge; (2) the ability to present ideas coherently; (3) the ability to do research; and (4) the specialized training in the area of the student’s greatest interest.
No strong line of distinction can in fact be drawn between knowledge, on the one hand, and the ability to present ideas coherently and the ability to do research, on the other hand. A person who does not have the ability to express ideas coherently or the ability to do research cannot be said to possess knowledge of his subject. True knowledge is more than the capacity for parrot-like repetition of what this, that, or the other economist said, and what this, that, or the other formula is, and unless research is narrowly defined as the analysis of empirical data of a limited kind, really operative knowledge is included under either the ability to present ideas coherently or the ability to do research or both. Thus, the teaching involved in imparting the common core of knowledge (as well as that involved in specialized training) should be such as to produce in the student clarity of thinking which should make clear writing a necessary consequence; and, also, the teaching involved in imparting the common core of knowledge (and specialized training) should be such as to leave the student with a clear idea of what research means, and how the interplay of hypotheses with tests based on empirical data results in acceptable knowledge.
In spite of the obvious interrelationship of the four major  categories listed above, however, it will be useful to consider them one at a time.
COMMON CORE OF ECONOMIC KNOWLEDGE
All economists should have a general acquaintance with the basic ideas in economics, and all should be equipped with the tools and the general empirical knowledge about modern economic systems that will provide a basis for economic research, policy guidance, and teaching. The common core consists of (1) a set of analytical tools, (2) a way of handling the tools in research and problem solving, and (3) certain institutional knowledge about the economic world. This common core is necessary not only to meet the above objectives, but also so that economists will be able to communicate with each other, and so that mobility among different uses of economists will be preserved. The substance of economics itself will be enriched if individuals can move freely from one area to another. For example, it is beneficial for the development of the profession if economists can move between business and government, on the one hand, and teaching, on the other. Similarly, research individuals should have the same sort of general background as those who are faced with administrative problems. The existence of a common core helps to ensure this, and is some protection against excessive compartmentalization and overspecialization in the profession. The problem of core training is one of balancing the desirability of having a number of essential requirements included in each student’s program with that of having the minimum amount of formal requirements.
With respect to the nature of the common core, there was fairly general agreement among those participating in the panels, and the conclusions reached are not strikingly different from the current practice in many graduate schools or the objectives expressed in the Bowen Report. There was a general feeling that some reorientation and redesign within the accepted framework might be in order, but that the general framework itself  need not be significantly altered. The content envisaged would include economic theory, economic history, mathematics, and statistics.
The theory requirement in the common core should probably be the most intensive of all the requirements. At least one and probably two full years of formal classwork in economic theory were considered necessary, supplemented by outside reading to fill in gaps not taken up in the formal courses. The courses themselves would not be entirely devoted to a formal presentation of certain specialized areas of theory, but should give students the ability to use theory effectively in handling problems. The work should cover modern theory in most areas of economics, and it should also be tied in with both the history of economic thought in these areas and some of the historical and institutional background that provides the context for the theory.
Economic history as a core component should be distinguished from economic history as a special field. The purpose of the economic history requirement should be one of literacy, to insure that the student has some perspective with respect to how economics is related to various aspects of human development. This requirement can provide the thread of continuity and integration which is normal lacking from work at graduate level. The growth and development of economic institutions in the various specialized areas should be treated in relation to each other, together with the relation of social and political history to economic development and the role of geographic location as a determinant of economic development.
The purpose of the mathematics requirement as a part of the core is partly to serve as a necessary tool for the mathematical economics and statistics, and partly for general literacy. It would  be desirable, of course, for students to have a proper mathematical background when they enter graduate school. Unfortunately, such a requirement is not easily enforced at this time, and it will generally be necessary for this deficiency to be made up either while a student is taking other work in graduate school or during the summers. In view of the specialized nature of the mathematics required for economists, it may well be that a specialized course drawn up specifically for economists or for social scientists would be the most efficient way to meet the need. Such a course would not be intended as a shortcut, but rather would attempt to give the student those areas of mathematics which are relevant to social science and to relate them to problems in economic theory, game theory, statistics, and econometrics. Literacy in the area of mathematics is important so that students will not be frightened by economics which is cast in symbolic terms. If there is to be communication among members of the profession, it is essential that all economists should have enough mathematics so that they can tell in a general way what articles in a mathematical form are about. This does not mean that those students who are not mathematically inclined should be forced to achieve mathematical fluency. However, all students should at least be required to have some minimum competence in mathematics.
As in the case of mathematics, statistics is partly a tool requirement and partly a literacy requirement. As a tool, students should be able to employ statistics for economic research. The traditional topics such as probability theory, statistical tests, and index numbers would all be covered. In addition, however, the student should learn how to handle basic empirical material in a systematic and orderly manner. The uses of accounting data, together with the meaning of various accounting classifications and accounting methods, should be studied. The student should also have a general knowledge of the sources of economic data, such as the kind of material contained in the various censuses of  the U.S., the national income statistics, and the types of economic information provided by the other agencies in the government. They should be familiar with the empirical work provided by non-governmental research institutions such as the National Bureau, and by international organizations. All of these are useful research tools, and they are also required for literacy in this area, so that the student will be able to appraise and evaluate empirical research.
Interdisciplinary training as related to the core of economics.
Considerable attention has been focused recently upon the desirability of having students know about fields other than economics, so that useful cross-fertilization can take place among the disciplines, and so that economics can be used more effectively in helping to handle public and private policy problems. It is argued that training in other disciplines will give the student greater breadth and make his economics training more meaningful. There was a general consensus among the members of the panels, however, that elementary survey courses in other disciplines would be of limited usefulness, and would expand the common core to a point where it would seriously infringe upon the freedom of students to follow lines of their own interest. Undergraduate training supposedly gives a student breadth; if it has failed in doing this the lack should be recognized as a gap in the student’s training. It is questionable, however, whether a graduate school should take formal cognizance of such gaps, as it does in the case of mathematics, and make provision in the graduate school curriculum for filling them. Where the gaps are extremely serious, the student should probably be encouraged to attend summer school, an/or do special reading, to make up the deficiencies. But it does not seem that the subject matter of interdisciplinary training and the deficiencies of preparation in the students are sufficiently clearly defined to make courses in them practical. Experiments might usefully be tried in this area, but they should be regarded strictly as experiments,  which might eventually yield elements that should be incorporated into the common core.
The extent and timing of the common core.
In terms of formal requirements, the common core should probably not exceed four or five year courses, depending upon whether or not the student can anticipate the mathematics requirement. In addition to this formal work, however, it might be desirable to provide for some sort of tutorial instruction to fill in gaps not covered in the courses and to follow up lines of special interest to the individual student. Such tutorial instruction would provide an element of flexibility not obtainable in formal classwork. With respect to timing, it seems logical that the major portion of the core would be covered in the first year, inasmuch as it provides tools used at later stages in graduate work. On the other hand, some time should be left in the first year for students to take courses of their own selection. Students should have an opportunity to sample several specialized areas before finally determining the area in which they are most interested.
The Ability to Express Ideas Coherently
The economist should have the ability to express his ideas coherently, and to move easily between the abstractions posed by economic analysis and the empirical elements of the problems with which he deals. This requirement is more than that of being able to write grammatical English; it involves training in the organization of ideas and the development of perspective. Rigor and clarity is essential if the profession is to serve its many potential functions. One of the major complaints of people who hire economists in business and government is that the products of graduate schools whom they hire do not have this ability to present their ideas coherently. They often express the opinion that economists who are intending to go into business and government should receive special training in this respect. However,  it is not any less important that individuals going into pure research or teaching should be trained to express their ideas coherently. Perhaps the reason teaching and academic research have not appeared to suffer as much in this respect lies in the lack of direct supervision of such individuals by supervisors who bear the responsibility for their written and oral presentations.
As already indicated, the ability to express ideas coherently is not merely a problem of correct grammar, but rather involves the organization of ideas in a meaningful manner. Unless a student can express an idea clearly, he does not really understand it. Thus, the ability to express ideas coherently is highly related to the problem of substance, and is properly the responsibility of a graduate school. Some students have difficulty in writing because they have little or nothing to say. They have not developed habits of creative thinking, and do not know how to approach a subject.
Because the economist usually crystallizes the results of his work in written form the writing itself is a tool, and is part of the basic methodology of the profession. In other disciplines such methodological tools are given explicit consideration. For example, in the sciences, students are thoroughly trained in laboratory work. In mathematics, students are drilled in working through problems. In law, briefs and case studies are written. In medicine, the internship trains the student in the handling of actual medical cases. Few graduate schools of economics, however, have considered writing explicitly as a tool of the profession, and therefore relatively little accent has been placed upon training the student adequately in this function.
The Ph.D. thesis, traditionally the masterpiece of a student being trained for the doctorate, does not fulfill this need. All too often it is instead a traumatic experience which leaves the student scarred but untrained. In a great many instances, furthermore, the thesis is done by the student out of residence, and the supervision of the writing of it leaves much to be desired. The student often attempts to write the thesis while he is pursuing another job on a full-time basis, and the writing may take  a period of five or six years. The hurdle is so great, as a matter of fact, that a large proportion of students who have completed everything but the thesis never finish it. Also, the moral pressure on professors to approve theses of students who have spent a large number of years on them is very great, with the result that the thesis itself need only show effort and length to be acceptable. In other words, the Ph.D. thesis is quite unsatisfactory for teaching students how to write, and because of the institutional considerations involved this failure cannot be corrected merely by exhorting students and teachers to greater effort and higher standards.
The members of the panels believed that the solution to the problem of training students to write coherently lies in the direction of more writing practice early in the graduate training program, and reliance on a larger number of shorter papers (5 to 10 pages) rather than a small number of major papers. This process should intimidate the student less, offer him more practice in organizing material, and make the task of criticizing and evaluating any given paper simpler.
One important aspect of training students to write, now largely neglected, is provision for revising and reworking papers. So much effort goes into the original writing of a lengthy paper, and the task of reworking it is so great, that most of the student’s writing tends to be a single-shot experience. In many cases the student never even seriously re-reads what he has written after he finishes it. In order to promote the reading and criticism of papers, it was suggested that some of the papers be duplicated and discussed in essay seminars attended by both students and faculty. Students should learn from such a procedure not only when their own work is presented but also from the problems encountered by other students. In this connection also, all papers need not be written in the confines of formal courses. The tutorial function spoken of in the previous section might well bear some of the brunt of criticizing short papers.
Courses involving group research would provide an opportunity for students to prepare papers in conjunction with each  other. Such joint papers would force the students to discuss the organization and presentation of the material, so that an agreed-upon version may be arrived at. This practice will prepare students for the sort of writing experience they are likely to encounter in business, government, or other group research.
If the writing of papers is to be stressed as a part of the graduate training program, it is only proper that it should assume a more significant role in the grading system. The student who can produce a first-class report at this own leisure, using the materials freely available to him, may well be a better economist than one who is more facile in showing his learning well in an examination but who may also be less proficient in turning out an independent piece of research. Present grading systems rely heavily upon examinations, which may test the student’s leaning ability but do not ordinarily test his ability to produce a well-conceived and well-executed report. The comprehensive examinations weigh very heavily in determining whether students are permitted to proceed and what kind of financial aid they are given. At both the course level and at the comprehensive examination level, it would be possible to give greater weight to written reports in the grading scheme. For the comprehensive examination, the student might be required to present what he considered the best two or three papers he had written. An evaluation of these papers would add a significant new dimension to the judgment of the abilities of students at this stage. By giving reports and papers a significant weight in the grading structure of the graduate school, students would be encouraged to revise and rework their manuscripts to a greater extent than they now do. Originality would be rewarded just as learning ability is now rewarded.
Because so many economists are required to do research of some sort in their work, and because all economists must be able to analyze and evaluate the results of such research, research  training is essential. The tools of economic research are, of course, necessary at least in some degree, but fully as important as the teaching of tools is the actual training of students to do research by doing it. The student emerging from graduate school should be able to carry through a piece of research in a systematic and meaningful manner. Students must be trained to set out a problem, design their work program with reference to this problem, carry out the basic work utilizing pertinent sources and appropriate methods, and finally, evaluate the results of this research, relating them to the original problem and appraising their validity.
A number of members of the panels felt that economic research generally suffered from a lack of respect for discipline and rigor. Casual empiricism, rather than scientific testing of hypotheses, is all too frequent. In many major pieces of research the sources and methods behind the results are not indicated adequately. These faults, they believed, are the result of inadequate teaching of research methods.
The misapplication of research tools, or the failure to apply suitable tools, is also widespread in much current economic research. The research worker may carry extremely unreliable estimates out to a number of decimal places, causing an inordinate amount of computational effort and lending a spurious appearance of accuracy. At the same time, this same research worker may gloss over important characteristics of his material which should have been tested for bias or general inconsistency by the use of fairly ordinary and straightforward statistical testing procedures.
The lack of research competence is also evident in the formulation of research problems. Often the reader of a research paper is at a loss to discover just what is being undertaken, and whether it was in fact achieved. This confusion often stems from a lack of clarity on the part of the original research worker in the conception of his problem, even more than from his presentation of it. It is very important that those embarking upon research recognize the importance in the research process of the original  conception of the problem and the design of the research to fit the problem.
These faults in economic research, combined with indecisiveness on the part of the individual research worker, lead to a considerable amount of floundering and waste motion. It is frequently necessary to re-do a piece of research because the formulation of the problem was inadequate. The failure to apply the proper tools at the proper time in the research process also may require that much of the work be redone, to make adjustments the need for which becomes obvious at a later stage in the research process. The prevalent lack of discipline and rigor makes all these revisions of portions of the research process extremely difficult, so that in fact the work usually must be completely redone, very often with quite different results.
In the light of these difficulties, research training should start early in the student’s graduate career and continue throughout its duration. Although in his first year the student will not have the necessary background and tools to do very much economic research, even at this early date practice with simple research problems would be useful in acclimating students to the various problems that research poses earlier in their careers rather than later. More of the student’s time can then be focused at a later stage on problems of a more substantive nature. It is well known that the greater part of time now spent on the Ph.D. thesis is spent in floundering around trying to select a problem and decide just how to carry it out. More and earlier practice in research might avoid much of this floundering.
The assignment of a larger number of short research subjects seems generally preferable, at least in the earlier part of the graduate training, to concentration on a few more substantial topics. If a number of different subjects are assigned, the student is faced again and again with the problem of how to formulate the research objectives and how to design the research. A larger number of projects also will serve to introduce the student to a number of different areas of economics, rather than to concentrate his attention solely in one direction. The question of  whether specific research topics should be assigned or whether the student should be allowed to choose his own is not an easy one to answer. Probably some of each approach should be used. Assignment of topics has the advantage of training the students to write for a customer. Freedom of choice in topics, on the other hand, has the advantage of allowing students to follow areas of special interest—and also gives them practice in arriving at a decision.
One of the major objectives of research training should be practice in the handling of empirical material of all sorts. The student should become used to dealing with historical material, economic statistics from all kinds of sources, and also material from other disciplines. He should gain experience in the critical evaluation of definitions and concepts, and in the manipulation and recasting of material.
The form of research training should probably differ at different stages of the graduate training process. In the early stages it may well take the form of special workshop courses, together with some for the work done for tutorial purposes. At a later stage, internship in various research projects within the university might be advisable. If possible, summer internship programs with business, government, or economic research foundations would also be desirable. Finally, individual research relationships with the faculty members on the basis of research assistantships or apprenticeships would serve a valuable role.
The Ph.D. thesis should serve a major function in research training, and should provide a test of whether the student has achieved research competence. But the primary research training should be begun much earlier in the student’s career; it should not fall upon the thesis alone. The thesis may well emerge as an outgrowth of some earlier research project.
Specialized training in specific fields is necessary so that economists can usefully bring to bear both the more detailed knowledge  of the institutions pertinent to the special area and the latest developments of economic analysis in this area. Without special field training, a student will not approach the frontier of any field, and will not have any training in depth. Specialized training, therefore, not only serves to equip a student to handle problems in a special area, but it also gives him training in depth as a background for understanding the process of research and appreciating the development of economics in general. In many special fields, economics alone will not be sufficient. Other disciplines are often required to enable the economist to deal with the specialized problems. In the area of corporate finance, law and accounting may be necessary. Law may also be necessary for public finance, labor, and international trade. Psychology or sociology may be pertinent to studies of consumer demand and labor. Each special field will necessarily entail the study of those portions of other disciplines which are germane to the set of problems encountered.
Under present circumstances specialization often tends to be somewhat superficial. The first year of graduate work is usually spent on the basic tool courses or general survey courses, and specialization is possible only during the second year of course work. A cumulative build-up of work within a special area is often impossible since the student finishes his term of residence at the end of the second year. Specialization may thus consist of one or two courses taken concurrently in the second year of graduate study.
The charge is often made that the areas of specialization offered tend to be too academic. Theory is extolled, and the actual work done by the student is largely confined to the library. Knowledge of the institutional setting of the special field tends to be slighted. There is little or no opportunity for internship in the special field during the period of graduate work.
Specialization may be conceived of as a highly detailed study of some small segment of economics or it may be conceived of as embracing a general area of problems for which other disciplines besides economics may also be relevant. Unfortunately,  present graduate training seems to emphasize only the first conception of specialization, but if the products of graduate schools are expected to serve as professionals in these areas the narrow concept of specialization must give way to the broader concept.
Finally, it is argued by representatives of both business and government that graduate training does not prepare students for the kind of work required in business and government. Unlike the conclusion in the previous sections with respect to the common core of economics, the ability to express ideas coherently, and the ability to do research, where it was concluded that the requirements are the same irrespective of whether the student wants to go into academic work, business, or government, additional training will depend upon the field the student decides to enter. The criticism that graduate schools at the present time do not offer appropriate specializations for students interested in business and government in the role of professional economists appears to be justified. The kinds of courses that would be required for such a specialization would cover such topics as projections, studies in demand and cost, and general economic accounting.
In order to correct the tendency toward superficiality, the student should customarily take two or three courses in a given special area, over a period of at least two years. This would provide the student with an opportunity to work in the area over a longer period, and so would permit a cumulative build-up.
Research work involving the handling of empirical material and/or field work should be undertaken simultaneously with the course work. Such research work might be part of an internship program, a workshop course, or an apprenticeship as a research assistant. In some cases, suitable summer employment might serve as part of the program.
As already indicated, training in related disciplines should accompany the work in the special field. Generally speaking, survey courses in related disciplines will not meet the need. Either courses especially designed to suit the area being studied or relatively advanced work within the other disciplines would be  appropriate in giving greater breadth to the program of specialization.
In order to meet the needs of business and government, a number of courses in fields not now generally offered could usefully be added. Such things as the problems of making projections, studies in cost and demand analysis, operations research, and economic accounting are all appropriate subjects, which could serve either as specialties in their own right or as valuable tool adjuncts in such fields as industrial organization, labor, and international trade.
The Role of the Ph.D. Thesis
In viewing the Ph.D. thesis as both a test of and a means of acquiring core knowledge, clarity of expression, and research competence, the panel members felt that the form of the thesis required some reconsideration.
The desirability of having the thesis written in residence is well recognized. Furthermore, the panel members generally agreed that it would seem sufficient as a requirement if students could turn out an article-length paper which would be of publishable quality. Such a short thesis could be examined and criticized in greater detail by the faculty, and, if needed, revised more often and more basically by the student. This does not mean that long Ph.D. theses should be prohibited; a student should have the right to undertake any task he wants to. Still, it does not seem unreasonable to require that even in the case of a long thesis the student shall, in order to meet the thesis requirement, present some piece of material not longer than 30 to 50 pages which can stand as an independent piece of writing, aside from possible appendices on sources and methods. Whatever he wants to do over and above this, of course, he can. It may well be argued that the short thesis should not be compulsory, but that it may be enough to announce to students that short theses are not only acceptable but encouraged. Several panel members felt that the short thesis might be inappropriate  for specific topics, and that the way should be left open so that the student could write a longer thesis if he chose to do so. There is danger in this approach, however, in that students may take the safe way out and write a long thesis much on the same basis that they write long answers to exam questions covering every possible facet of the question. In such a case the tendency to judge theses by the pound might continue.
If the requirement that the thesis be of publishable quality is seriously intended, it might be desirable to consider having the university undertake the actual publication, in the form of an annual series. If the theses are in fact held to a length of 30 to 50 pages, the cost of publishing them would not be excessive. Such an arrangement would have several advantages. First, it would tend to make the students more careful of what they offer, since in most instances it would represent their first published work. Second, it would provide the student with copies of his thesis at nominal cost in the form of reprints. This would be very useful for job applications. Even when prospective employers were not sent a reprint by the student they would be able to obtain the thesis series from most libraries, and so could have access to a sample of the student’s work. Furthermore, the faculty would feel more conscientious with respect to the supervision of theses, since it would be evident to other institutions and members of the profession generally what caliber of work was being done. Finally, the work involved could be arranged to accord the students themselves with experiences in publishing in much the same way a law review does in law school. The argument against such a series is that the better theses or redrafts of them will be worth publication in the regular professional journals, and that this would be much preferable. There is also no guarantee that the university series would offer any substantial incentive to high quality, but may well have the opposite effect.*
The General Form of Graduate Instruction in Economics
These requirements partially dictate the general form of graduate education in economics. For one thing, a certain degree of formality will be required in education at the graduate level. This formality comes about because the entering graduate student usually does not possess the background necessary for graduate work in economics. Unlike the sciences and medicine, it is not practical to require that all entering students possess training in specific areas. The decision by students to become economists almost invariably is made very late in their undergraduate careers, so that it is usually impractical for them to acquire more advanced training in this area while they are undergraduates. Students should, of course, be encouraged to acquire the background at the undergraduate level insofar as possible, and the graduate curriculum may be modified to accelerate students who are adequately prepared. Nevertheless, there will still be a considerable area of the common core to which almost all students should be subjected.
Students who are capable of good work in one direction but find some other area extremely difficult may perhaps be permitted to waive certain of the requirements. The exceptional students, furthermore, need not necessarily be only those brilliant students who excel in economic theory. Students of more specialized interests, such as those primarily interested in the filed of labor, economic history, or corporation finance, should be given consideration fully as much as the theorists.
To a considerable extent, flexibility of graduate training can be secured by more individual attention in the form of some sort of tutorial and/or internship training in graduate school. Such a tutorial and/or internship would make the individual needs of the students known to the faculty, and it would give the student more opportunity to go his individual direction, either filling in gaps in his knowledge or pursuing lines of special interest. It would not always be necessary that senior faculty members be used as tutors. Younger staff members who  were themselves more recently graduate students may make more suitable tutors, in that they are closer to recent graduate training and are generally freer with their time.
Finally, it seems necessary to maintain some form of certification as a function of graduate education, as long as the number of students trained is substantial. People hiring students will want to know the kind and caliber of work done by the student in question. It has been suggested that the certification problem can be lessened by relying for purposes of recommendation and scholarship evaluation on more lengthy comments written by the student’s supervisors.
The Period of Graduate Training
It is the present practice of many graduate schools to concentrate the tool courses in the first year of graduate studies. Such an arrangement tends to make a somewhat regimented, formal, and uninspired first year of graduate work. The beginning student is left little room to follow lines in which he is interested or to explore areas to see whether he would find them interesting.
The specialization that takes place in the second year, as noted in the preceding section, often means only a single course in the special field. As a result, a survey course within an area is considered advanced work in that area. This specialization, furthermore, occurs at the same time the student is preparing for his comprehensives, and usually more attention is given to the comprehensives than to the specialization.
The thesis is often not started until after the student has finished his second year of graduate work and passed his comprehensive examinations. As a result, not only the writing of the thesis but the conception of it as well may be done after the student has served his time in residence and left. The consequent lack of supervision, the relegation of the thesis to a part-time task, and the prolongation of the thesis period to a number of years all tend to reduce the quality and usefulness of the thesis.
 The panel was generally agreed that the distinction in timing between tool courses, specialization, and the thesis should be less sharp than is current practice. In the first year, the student should be allowed to do some browsing. Some of the tool courses should be postponed until the second year, so that more of a cumulative development in the tools themselves would be possible.
The preliminary work on the thesis should not be put off until the third year of graduate work, and the thesis itself should be completed while the student is in residence. Initial work might start in a thesis seminar in the second year of graduate study. Rather than spending full time on the thesis at any point in his graduate work, the student would be expected to work on his thesis along with other course or seminar work.
Internships, research assistantships, and other such programs may mean that the student will interrupt or prolong the period of graduate work, or he may spend some of his summers in such activities. Programs such as these, however, should be planned in terms of the student’s total graduate training, and should be carried out as part of it. They should not be devised solely in terms of the faculty’s manpower needs—as at present is sometimes the case.
These requirements indicate that a minimum of three years in residence will be required by graduate students to complete the work. Generally speaking, four years will be more usual, so that the student can get practical experience as well as formal training into his graduate training. For the student’s own good, a period of more than five years in residence between entrance and the obtaining of the doctorate is probably undesirable. Should the student contemplate a more ambitious program than this, it should be of a post-doctoral nature. It would be useful for this purpose if universities could set up programs whereby post-doctoral students could obtain internships in business and government for a year, and then return to the university in a teaching position for a year following the internship. Such an arrangement would encourage business and government to take  students on an internship basis, and would at the same time give the individual student an opportunity to get established after having served his internship.
Summary and Conclusions
- The familiar concept of giving all graduate students in economics basic training in a common core appears to be a useful device, and should be kept as an integral part of graduate training in economics. This common core, if properly conceived, has the advantage of providing some breadth to the student’s training, not only making him more literate, but also giving him a better perspective within which to place his more specialized training. The common core also makes it easier for economists to communicate with each other insofar as they have had the same type of general training. Finally, mobility within the profession is promoted, so that it is possible for economists to move between business, government, and academic work to a much greater extent than might otherwise be so.
- The inadequacy of the current training of economists in writing and research was considered to be one of the greatest gaps in graduate training. The ability to express ideas coherently and the ability to carry through research work in a skillful manner should both be considered major tools of the economist. The graduate program, therefore, should take account of both these needs early in the period of graduate training, and attention should continue to be directed to them throughout the graduate program. Both writing and research should be weighted more than is done at present in the grading structure of the graduate program. One of the primary objectives of graduate schools should be to produce people who do not just know, but who can do as well, and the grading structure should be changed to assist in bringing this about. Special programs to promote research training, such as internships in the university or outside of it, should be developed to give the student more research experience under supervised conditions.
- Specialization in graduate school should equip the student  with more advanced training in various areas. It is important that this training not be too narrowly conceived nor too superficial. Instances where a single advanced course and little outside work is supposed to make a student a specialist are all too frequent. Specialization requires a longer build-up of cumulative work, and may involve going into related areas outside of what is generally considered to be economics. Graduate schools should give more careful attention to the specialized training students receive and whether this training does in fact meet the requirements for genuine specialization.
- Graduate training normally takes place over a very extended period. Students often work part time while trying to get their doctorate. It is thought that much would be gained if, as in the case of the professional schools, graduate training in economics could take place in an unbroken period of concentrated effort. If the common core is to be retained as is suggested in item 1 above, and more emphasis is to be placed upon writing, research, and specialization, as suggested in items 2 and 3 above, it seems very probable that the total effort going into graduate training in economics by the student will have to be increased. The concentration of studies into a period of three or four consecutive years on a full-time basis will do much to increase the efficiency of the students’ training and permit these objectives to be met. Summer programs of research or internship training may also be of considerable aid in fulfilling these objectives without extending graduate training further.
- The present form of the Ph.D. thesis is not an optimal device for achieving these objectives. It was thought that short theses, which could be reworked more easily and which could generally be made available in published form, would be more manageable and would provide a more effective training device. Such a thesis could be integrated into the graduate training program, and could generally be expected to be written while a student was still in residence; the doctorate would be granted directly upon completion of the period of residence and the thesis.
*One panel member has suggested that in cases where a mediocre short thesis is written only an M.A. be granted, and the Ph.D. reserved for theses of exceptional quality.
Source: John F. Kennedy Presidential Library. Personal Papers of John Kenneth Galbraith, Series 5. Harvard University File, 1949-1990. Box 517, Folder “General Correspondence 8/7/56—12/10/57”.