press enter after type

Harvard. Written vs. Oral Exams. Gerschenkron vs. Chamberlin, 1958



I recall that Robert Solow once remarked that until he encountered psychiatrists of the Freudian persuasion, he had always thought that the opposite of “oral” was “written”. In 1958 the issue of having oral general examinations vs. written  general examinations in economics at Harvard found the economic historian Alexander Gerschenkron arguing the case for “writtens” and the economic theorist Edward H. Chamberlin arguing to continue the tradition of “orals”. Among other things one learns that before there were smart phones to distract examiners in the oral exams, examiners would sometimes read “novels and/or manuscripts” during the parts of the examination that didn’t concern them directly. Especially valuable in Gerschenkron’s memo is his survey of the practice of general examinations in seven graduate programs besides Harvard’s.


Department of Economics
January 1958

Dear Professor [Handwritten: “Galbraith:”]

You will find attached a proposal for a reform of our General Examination procedure which will be discussed at the Luncheon Meeting of the Department of Economics on January 14 at the Faculty Club.

Kindly treat it as confidential.

Sincerely yours,
Alexander Gerschenkron





Introduction of a Written Test as a Part of General Ph.D. Examinations

The proposal for a limited reform of our General Ph.D. Examination which is detailed in the following has emerged from the discussions of the Special Committee nominated by the Chairman of the Department as well as from conversations with a number of colleagues. It does not, however, represent a unanimous view of the Committee, as one member (Professor E.H. Chamberlin) expressed serious reservations which he will communicate in a special statement.

Along with several other members of the Department, I have been concerned for some time with the standards of efficiency and fairness inherent in our requirements for a Ph.D. in Economics. It has seemed in particular that the General Examination as presently given, is deficient in several respects:

  1. An oral examination not supplemented by a written examination places a nervous or slow-thinking student under considerable disadvantage. It must be remembered that the American student, from high school days onward, is trained to take written examinations. In most cases, the General Examination is the first important oral examination of his career.
  2. The appraisal of the candidate varies widely depending on the composition of the board, the mood of the examiners, some interesting, but irrelevant personal traits of the candidate, etc.
  3. As often as not, the examination in the individual subjects touches on limited areas. Either the time devoted to a single question is too short and the student has no opportunity to express himself fully, particularly if communication was slow in being established: or else the whole time is devoted to the discussion of a single question. In either case the student feels that he has not been given the opportunity to display the full extent of his knowledge and understanding of the field. Good and poor students alike resent the certain arbitrariness of our examinations.
  4. The chairman of the Board as a rule recoils from the prospect of having to communicate a negative verdict to a poor candidate. This feeling is somewhat shared, though to a lesser extent, by the other examiners. The results are the numerous examinations graded with a “fair minus.” Their number would be even larger but for the tendency to conceal the governing motivation from the world and from themselves by raising the grade to a straight “fair.”

As is shown in the appended survey of General Examinations for Ph.D. in Economics at leading Universities, it is only Columbia University whose procedure is the same as ours. Every other Department has an extensive written examination for a total of at least 12 hours. In all cases, except Chicago and Michigan, the written examination is followed by an oral examination lasting up to three hours. Michigan has an oral examination for poor students only; Chicago has a thesis seminar in lieu of an oral examination.

At this University, the History Department requires an oral examination only. The Philosophy Department has an extensive written, but no oral examination. The Government Department is seriously debating introduction of written examinations to precede the oral examination. In Social Relations (Sociology) a three-hour written examination preceds the oral examination.

It is fairly clear a) that we are out of step with other Departments at major institutions and b) that the present arrangement is not altogether satisfactory. Nevertheless it is not proposed here to introduce a radical change by adding a written examination in all the subjects which the student has elected to offer. Such a change would no doubt mean a great deal more work for the Department as a whole. Besides, much can be said for gradual change during a period of experimentation. Accordingly, the concrete proposal is as follows:

  1. Students entering in September 1958 must, in addition to an oral General examination in its present form, take a four-hour written examination in Economic Theory and its History. This examination should be given as an integrated comprehensive examination in economic analysis, and be conducted at a level higher than course examinations in Graduate Theory courses. The candidates should be given a generous choice of questions which take into account such special interests in fields closely related to Economic Theory as the students may have.
  2. The examination is conducted by a board of three Department members, two of whom are replaced each year. They are responsible for drafting the questions and grading the examinations. The examination papers will bear a code number rather than the name of the candidate. The three members of the Examining board should be exempt as far as possible from taking part in current oral examinations. In grading they will have the assistance of two junior members of the Department.
  3. The written examination paper will be graded in the same fashion as the oral examination. Candidates who did not receive a grade of at least “fair minus” cannot be admitted to the oral examination.
  4. The written examination will be given once each term. After two unsuccessful presentations to the written examination, the case of the candidate must be reviewed by the Department before another attempt can be made.
  5. The written examination paper will be included in the candidate’s folder and thus made available to the members of the oral examination board.
  6. The final grade for the General Examination as a whole is determined by the members of the oral examination board. In agreeing upon the grade, they should as a rule take the grade received at the written examination into account.
  7. The result of the examination will not be communicated to the candidate at the end of the oral examination, but transmitted to him by mail.

The advantages of the scheme as presented here may be pointed out briefly:

  1. The winnowing effect of the qualifying written examination will tend to raise the standard of a Harvard Ph.D. in Economics.
  2. The requirement of a comprehensive written test in Economic Theory is likely to deepen the intensity and widen the scope of students’ preparation for the examination.
  3. The written examination while not eliminating certain deficiencies of the oral examination would tend to reduce somewhat the inordinate advantage of a nimble-witted but superficial student in favor of a less articulate but more profound student.
  4. The suggested scheme is unlikely to increase the workload of the Department members to any perceptible extent, if at all. The larger claim on the time of the members of the special examining board probably will be offset in the long run by a smaller number of oral examinations in any given year.

The present proposal amounts to a moderate reform which should be adopted as an experiment. After it has been in force for two or three years (reckoned from the first examinations to be given sometime in 1959/60), the Department may well wish to reconsider the problem in the light of the experience that has accumulated in the interval. Alexander Gerschenkron



The following is a survey of the way in which the General examination for the Ph.D. is given at the following universities: M.I.T., Yale, University of California at Berkeley, Johns Hopkins, University of Michigan, University of Chicago, and Columbia University:

Columbia University (Shoup)

A General oral examination very similar to ours. The course work preceding the examination seems lighter than ours. Professor Shoup summarizes his views on the system as follows:

“In general, then, we at Columbia place importance on the oral type of examination, both at the subjects stage and for the dissertation, and do not depend at all on written examinations. The question of whether we should alter our emphasis somewhat has been discussed at some length recently within the deparmentt, but the consensus of opinion seems to be that there is no good substitute for the oral examination and that we do not want the student to spend a great deal of time preparing for written examinations, since we feel these tend to perpetuate somewhat of an undergraduate approach to studies. However, there is some difference of opinion, naturally, on this score—but not enough so that we should anticipate any change in our procedure in the near future.”

University of Chicago (Johnson)

Three written examinations, 4 hours each, for a total of 12 hours. No oral examination, but a Thesis Seminar must accept the proposed topic.

University of Michigan (Ackley)

Four written examinations, four hours each, for a total of 16 hours. Oral examination in doubtful cases only.

Johns Hopkins (Evans)

A comprehensive written examination for a total of 12 hours. Followed by an oral examination, lasting for an hour and a half. As a rule, two students are examined jointly.

Yale University (Reynolds)

Nine to 11 hours of a comprehensive written examination. Followed by an oral examination in two specialized fields, one of them the dissertation field.

University of California at Berkeley (Papandreou)

Four written examinations, three hours each for a total of 12 hours. Followed by an oral examination which lasts three hours.

M.I.T. (Kindleberger)

Written examination in 4 fields, three hours each for a total of 12 hours, given on alternate days from Monday to Monday. Followed by an oral two-hour examination in three fields. Professor Kindleberger summarized his view of the system as follows:

“May I supplement by letter of October 15, replying to your question about the general examinations for the Ph.D., by saying that we feel that the combination of writtens with orals is a desirable one. For one thing, the two examinations complement one another. The oral can be used to follow up answers to the writtens where it was not clear whether the student was fully responsive to the question. Or if a student avoids all theoretical (institutional, empirical, etc. ) questions on the writtens, he can be tested in these areas. With orals alone, he may be tested in his strong or in his weak areas, but the process is somewhat more random.

“We find that the students prefer writtens with orals, since orals by themselves increase the emotional tension. Most of our students have not had any experience with this form of examinations, all are tense about it, and some perform less well under tension than others. Our grade for the generals a a whole (which is not revealed to the student) is an average of written and oral performance, and probably runs somewhat higher than if it were based on the latter alone.”




Department of Economics

Appendix B to letter from Professor Alexander Gerschenkron dated January 1958.


My chief reason for opposing the introduction of a written test as a part of the General Examination for the Ph.D. is that it will add substantially to the burdens of the Department and to the Departmental machinery, without any appreciable gains that I am able to discover. It will also add to the burden of the students, who already take 24 hours of written course examinations per year, or a total of 48 hours in the normal case of two years residence for the Ph.D. We already have ample evidence of the quality of written work of which our students are capable, and this written work is in fact given substantial consideration, both as a standard of qualification for the oral exam and in the total evaluation of the student’s record.

I simply can see no case for changing our present system. Is there any evidence that our students are generally inferior; or are inferior to those of institutions with written exams, or are suffering in competition with them? If so, it has not been presented, and I do not know of any.

My esteemed colleague, Alex G., lists four respects in which our present General Examination is allegedly defective, and I offer the following comments on them:

  1. Of course some students do not do as well in orals as in written work. The reverse is also true. Both types of ability (or its lack) should be taken into account, as I believe they now are in the student’s whole record.
  2. The personal element in the composition of the boards, etc., is an old old story. It used to be much more of a problem when there were several well-known cases of members of the Department who were noted for being far out of line with the general Departmental standard. I believe that such problems are not a present factor of any importance, and that the grading of orals is more uniform now than it has been for a long time. Orals boards should, of course, constantly strive to achieve better standards of uniformity in judging candidates. Standards in this respect could be improved far more, and with far less grief than by adding written exams, if examiners would refrain from reading novels and/or manuscripts while their colleagues examine the candidate, and concentrate on evaluating the candidate’s performance in all fields instead of merely in their own. (I may add that this sloppy practice, which must be highly disillusioning to the candidate and destructive of the morale of the examination, is entirely a development of the last ten or fifteen years.) Perhaps a longer period of more detailed discussion after the exam would also help to achieve greater uniformity of standards.
  3. What student is there who, after any examination has not reflected on all that he knew or had studied, that was not asked for? Of course each of us should strive to give some breadth to his own part of the oral examination. On the other hand, the oral exam gives an opportunity to probe for depth as well as breadth, and this is also important. There will always be arbitrariness in any exam; and incidentally, a new element may well be added in the proposed written ones if the questions are made up by a board of three members of the Department without consultation with others—a board which will have a different orientation as it change every year. Uniformity will be achieved only within any one year.
  4. The fourth point is controverted by the brutal facts. It present a hypothetical argument leading to a conclusion of “numerous examinations graded with a fair minus” which would be even more numerous were it not for a second alleged wave of self-deception which raises may fair minuses to a straight fair. Last year, 1956-57, out of a total of 33 examinations there were one fair minus and two fairs. There seems to be no reason to suppose that any who received as high as a fair plus or better should have been failures, an no reason on the face of it to think that the three who received fair or fair minus should have failed. (Incidentally, Kindleberger comments that the grades at M.I.T., which are a combination of written and oral, probably run somewhat higher than if they were based on orals alone. Perhaps with written generals nobody would be failed at all!)
    The entire distribution for last year is:
Excellent minus
Good plus
Good minus
Fair plus
Fair minus
Failed—no bar
Passed for M.A. only


My esteem for my esteemed colleague, Gerschenkron, is unbounded, but in the present matter, I am unable to see any case whatever for changing our present system, which seems to be dispensing justice about as well as one has a right to hope for in this imperfect world. Perhaps it is Columbia (and Harvard) who are right.

E.H. Chamberlin


Source:   John F. Kennedy Presidential Library, Papers of John Kenneth Galbraith, Series 5. Harvard University File, 1949-1990. Box 525, Folder “Harvard Department of Econ.: [Departmental documents].

Image Source: Alexander Gerschenkron from Harvard Class Album 1951 and Edward Chamberlin from Harvard Class Album 1946.

Irwin Collier

Posted by: Irwin Collier