Author: Irwin Collier

The significance for the history of economics of the following three memos is that they provide an illustration of the diffusion (infiltration?) of the M.I.T. canon to other departments. Albert Ando taught a few years at M.I.T. before coming to Penn and Phoebus Dhrymes (M.I.T., Ph.D., 1961) wrote his dissertation under Kuh and Solow.  The memos were sent to the curriculum committee of the department of economics at the University of Pennsylvania in January 1965 (at least the Ando memo is dated January 14, 1965 and it explicitly refers to the Phoebus memo and their recommendations to the Mathematics Committee that are undated).

Obituaries for both Ando and Dhrymes have been added to this post and precede the three memos.

Economics in the Rear-view Mirror thanks Juan C. A. Acosta who found these memos in the Lawrence Klein Papers at the Duke University Economists’ Papers Project and has graciously shared them for transcription here. 

Addition to post: At Banca d’Italia, N. 7 – Albert Ando: a bibliography of his writings.

_______________________________

Albert Keinosuke Ando
1929-2002
Obituary

Dr. Albert Ando, professor of economics, SAS and professor of finance, Wharton, died on September 19 [2002] at the age of 72.

Dr. Ando was born in Tokyo, Japan in 1929 and came to the United States after World War II. He received his B.S. in economics from the University of Seattle in 1951, his M.A. in economics from St. Louis University in 1953, and an M.S. in economics in 1956 and a Ph.D. in mathematical economics in 1959 from Carnegie Institute of Technology (now Carnegie Mellon University). Dr. Ando came to Penn in 1963 as an associate professor of economics and finance and became professor of economics and finance in 1967. He held this position until his death.

Dr. Lawrence Klein, Nobel laureate in economics and professor emeritus of economics wrote the following about his colleague.

After World War II many Japanese scholars visited the United States for general education and to modernize their training in some key subjects. Albert Ando, Professor of Economics and Finance, who died of Leukemia last week was an early arrival in the 1940s. He was educated at Seattle and St. Louis Universities and often expressed gratitude at the career start provided by his Jesuit teachers in an adopted country.

He completed the doctoral program in mathematical economics at the Carnegie Institute of Technology, where he was strongly influenced by Herbert Simon with whom he collaborated in research papers on aggregation and causation in economic systems. He also worked closely with another (Nobel Laureate to be) Franco Modigliani on the life cycle analysis of saving, spending, and income.

Dr. Ando was on the faculties of the Carnegie and of the Massachusetts Institutes of Technology before moving to the University of Pennsylvania, where he remained since 1963. He had visiting appointments at universities in Louvain, Bonn, and Stockholm. He consulted with the International Monetary Fund, the Federal Reserve Board, The Bank of Italy, and the Economic Planning Agency of Japan. He held many positions as an editor of scholarly journals and wrote numerous articles and books.

The main contributions of Professor Ando were in econometrics (theory and applications), monetary analysis, demographic aspects of household economic behavior, economic growth, and economic stabilization. His work on the Massachusetts Institute of Technology, University of Pennsylvania, and Social Science Research Council (MPS) model was of great benefit for the research department of the Federal Reserve Board, and his more recent work on econometrics for the Bank of Italy had been very fruitful.

He served as chairman of the graduate group in the economics department, 1986-1989, and developed excellent working relationships with many advanced students. He set very high standards, and those he worked with as thesis supervisor benefited greatly. He was extremely loyal and dedicated to their work, maintaining close connection with them after they departed from the University.

During his long and fruitful career, he earned many honors–as Fellow of the Econometric Society, as a Ford Foundation Faculty Research Fellow; as a Guggenheim Fellow, and a Japan Foundation Fellow. He was given the Alexander von Humboldt Award for Senior American Scientists.

Albert Ando is survived by his wife of 35 years, Faith H. Ando, two professorial sons, Matthew and Clifford, and a daughter, Alison, who has just been admitted to the New York Bar. His mother, sister, and brother, live in Japan.

–Lawrence Klein, Professor Emeritus of Economics

Source: University of Pennsylvania. Almanac. Vol. 49, No. 6, October 1, 2002.

_______________________________

Phoebus James Dhrymes
(1932-2016)

Phoebus J. Dhrymes (1932-2016), the Edwin W. Rickert Professor Emeritus of Economics, was a Cypriot American econometrician who made substantial methodological contributions to econometric theory.  Born in the Republic of Cyprus in 1932, Phoebus Dhrymes arrived in the United States in 1951, settling with relatives in New York City. After a few months, he volunteered to be drafted into the US Army for a two-year tour of duty; afterwards he attended the University of Texas at Austin on the GI Bill. In 1961 he earned his Ph.D. from the Massachusetts Institute of Technology under the supervision of Edwin Kuh and Robert Solow (Nobel Laureate 1987).  After a year-long post-doctoral fellowship at Stanford, he began his professorial career at Harvard, then moved to the University of Pennsylvania, and then UCLA.  In1973 he joined the Department of Economics at Columbia University; he was named the Edwin W. Rickert Professor of Economics in 2003 and retired in 2013.

Econometrics refers to that aspect of the economist’s work concerned with quantifying and testing economic trends. Phoebus Dhrymes‘early research focused on problems of production and investment, but he soon turned to more methodological work and produced important results on time series and on simultaneous equations.  Throughout his career, Phoebus Dhrymes placed much emphasis on the dissemination of scientific knowledge. In the early 1970s he helped found the Journal of Econometrics, which has become the leading journal in this field.  He was also on the advisory board of the Econometric Theory, and was managing editor and editor of the International Economic Review.He was a fellow of the Econometric Society and the American Statistical Association.Dr. Dhrymes was also one of the founders of the University of Cyprus, from which he was later awarded an honorary degree.

He wrote a series of influential textbooks including Distributed Lags:  Problems of Estimation and Formulation. This work was translated into Russian and published by the Academy of Sciences of the Soviet Union, and in the 1970s Dr. Dhrymes was invited to visit the (now former) Soviet Union, specifically Moscow and Novosibirsk. At the time such visits were unusual events for westerners, requiring rarely-issued visas and security clearances, particularly for centers of research such as Novosibirsk.

In a 1999 interview he characterized his books as “filters that distill and synthesize the wisdom of many contributors to the subject.   On this score, I was influenced in my writing by the way I learn when studying by myself.”  (Econometric Theory, 18, 2002)

Dr. Dhrymes is survived by his daughter, Alexis, and his sons, Phoebus and Philip. In his personal life, he was regarded as a generous, kind and gentle man, always there for his family. He came from humble beginnings, and garnered great respect from his family and friends for his achievements. He spoke often of how much he enjoyed teaching. He was always available to his students.He encouraged individualized thinking and understanding of processes rather than rote memorization in learning. He had a warm and affable demeanor, recalled fondly by former students and family members. He will be sadly missed.

Source: Obituary for Phoebus J. Dhrymes at the Columbia University Department of Economics Website.

_______________________________

Memorandum

To: Herbert Levine, Chairman, Curriculum Committee
From: Albert Ando
Subject: Offerings and Requirements in Macroeconomics, Monetary Theory, and Related areas in General Economics Ph.D. Program

1. Macroeconomics

Enclosed herein is a copy of the outline and references of Economics 621 [The outline and references will be posted later] as I am offering it this fall. It is fairly similar to [the] one year course in macroeconomics which is required of all Ph.D. students at MIT. I am sure that opinions would vary on details, but it is my view that this represents more or less the topics and literature that all Ph.D. students in economics should be familiar with. Ideally, I think there should be another major topic at the end of the outline dealing with current problems and policies.

It is fairly clear that this outline could not be covered in one term, particularly under our present system in which there are only 13 to 14 weeks of classes for a term. As a matter of fact, this fall, with a great deal of rushing throughout the term, I will be able to finish the static part of the outline by the end of the fall term, but certainly no further.

This suggests that the required macroeconomics for Ph.D. students should be two term sequence of courses, the first term dealing essentially with the Keynesian static analysis, and the second term with dynamics, i.e., business cycles and growth models.

2. Monetary Economics

I have just discovered that Economics 622 is taught without any prerequisite, and that there will be some students in 622 who have not had any macroeconomic theory this spring. I am somewhat stunned, and do not see how I will be able to teach a satisfactory course under the circumstances. This situation is indicated by the fact that 622 is required not only of Ph.D. students in economics but also of master’s candidates, and therefore it is apparently impossible to exclude the students from 622 who have not had 621. An obvious temporary solution is to make those students who have not had 621 wait until next year to take 622. In my view, elements of monetary problems should be included in the first term of the required macroeconomics course, and courses in monetary theory should be made elective. The course in monetary theory should then be taught assuming that students have had adequate preparation in macroeconomics and microeconomics, particularly the theory of general equilibrium, at the level where we can discuss the research and developments in the past dozen years or so, bringing students up to a point where they can draw a thesis topic from their work in the course. There is a room for an argument that there should be another course in addition to the advanced theory course, which deals with more traditional money and banking material. As a matter of fact, I offered two courses in monetary economics at MIT for several years, one dealing with traditional money and banking material taking the one term each of macro and micro economics as prerequisites, and another highly theoretical and advanced course taking two terms each of macro [and] micro economics as prerequisites. It seems to me, however, that Economics 639, Monetary Problems and Policies, should serve as the good traditional money and banking course, so that only one additional course seems to be needed.

3. Microeconomics and Mathematics

After some discussion with Dhrymes, it is fairly clear that microeconomics should also be taught as a two term sequence. A possible division between two terms would be to deal with partial equilibrium analysis of consumers and firms during the first term, and with the general equilibrium analysis and welfare economics in the second term.

During this fall term, Dhrymes and I found it necessary to conduct a few special remedial sessions in mathematics so that some rudimentary notions of calculus and linear transformation will be available in the discussions in theory courses. The idea, of course, is to arrange so that all students are equipped with minimum of mathematics by the beginning of the second term. If the recommendation of the committee on mathematics is adopted, so that students will learn elementary calculus and the matrices and linear transformation, including rudiments of linear differences and differential equations at the level suggested by the committee it is possible to synchronize it with theory courses so that theory courses will be using only those mathematics students are learning in mathematics remedial courses. For instance, the first term of macro theory would not require too much mathematics except the notion of the systems of equations and their solutions, and the first term of micro theory not much more than the condition of extremum in a fairly informal manner. In the second term, on the other hand, theory courses will require conditions of stability in the general equilibrium analysis, and the difference and differential equations in dynamic models in macroeconomics.

4. Overall First year program and Second year fields of specialization.

In addition to micro and macro theories and mathematics required for these theory courses, students should be asked to learn minimum of statistics and econometrics. The level of statistics and econometrics should be maintained at the level of text books such as Frazer, Brunk, or Mood plus Johnston.

The implication of the above statement is that the course schedule for typical first year Ph.D. students should look as follows:

First term:

Microeconomics I (Partial equilibrium analysis)
Macroeconomics I (Static Keynesian analysis, including some monetary considerations).
Mathematics I (Elementary calculus)*
Mathematics II (Elementary Linear Algebra)*
Economic History (For those with Adequate mathematical training)

*For the suggested content of mathematics courses, see recommendations of Mathematics Committee.

Second Term:

Microeconomics II (General equilibrium analysis and welfare economics).
Macroeconomics II (Dynamics, business cycles and growth)
Econometrics (6 hour course)

This schedule, of course, would be subject to variations depending on the background and preparations of students. For instance, students who already have sufficient mathematical training might be encouraged to take a course in economic history and a course in somewhat more advanced mathematics, such as mathematical theory of probability or a course in topology in the first term in place of Mathematics I and II.

_______________________________

Lists of Topics for Mathematics for Economists
[Recommendations of Ando and Dhrymes submitted to the Mathematics Committee]

(Mr. Balinski is to suggest some alternative text books)

1. Calculus
   1. Sets and Functions.
      1. Definitions
      2. Operations on Sets and Subsets.
      3. Relations, Functions.
         K.M.S.T. Chapter 2, Sections 1 through 6, possibly Sections 10 through 13.
   2. Functions, Limits, and Continuity.
   3. Differentiation and Integration of Functions of one variable.
      1. Concepts and Mechanics.
      2. Infinite series and Taylor’s Theories.
      3. Extremum Problems.
   4. Differentiation and Integration of Functions of many variables.
      1. Concepts and mechanics.
      2. Extremum problems, nonconstrained and constrained.
      3. Implicit Function Theorem.
         Any elementary text book in Calculus (e.g. Thomas; Sherwood and Taylor), Supplemented by some sections of a slightly more advanced text on Implicit Function Theorem and La Grange multipliers.
2. Linear Algebra and others.
   1. Vector Spaces and Matrices.
      1. Vector Spaces and Matrices, Definitions, and Motivations.
         Perlis, Chapters 1 and 2.
      2. Linear Transformations.
         K.M.S.T., Chapter 4, Sections 7 through 12.
      3. Equivalence, Rank, and Inverse.
         Perlis, Chapter 3.
         Perlis, Chapter 4.
      4. Quadratic Forms, Positive Definite and semi-definite Matrices.
         Perlis, Chapter 5, Sections 1, 2, and 5
      5. Characteristic Vectors and Roots.
         Perlis, Chapter 8, Sections 1 and w[?], Chapter 9, Sections 1, 2, 5, and 6.
      6. Difference and Differential Equations; Linear with Constant Coefficients.
         Goldberg, Chapters 1, w, e, and Chapter 4, Sections 1 and 5; Perlis, Chapter 7, Section 10. Some reference to two dimensional phase diagram analysis of non-linear differential equations with 2 variables. Lotke?
      7. Convex Sets.
         K.M.S.T., Chapter 5.

_______________________________

MEMORANDUM
January 14, 1965

To: Curriculum Committee
From: Phoebus J. Dhrymes
Subject: Mathematics, Microeconomics, Statistics and Econometrics in the Economics Graduate Training Program

1. Mathematics

It has become quite apparent to me during the course of the last term that our students are woefully equipped to handle instruction involving even very modest and elementary mathematics.

I think it is quite generally accepted that a student specializing in Theory, Econometrics and to a lesser extent International Trade and Industrial Organization would find it increasingly difficult to operate as a professional economist, and indeed seriously handicapped in satisfactorily carrying on a graduate study progress, without adequate mathematical training. With this in mind Albert Ando and I have prepared a tentative list of topics that graduate students ought be minimally familiar with and which has been presented to the Mathematics Committee.

This could form a remedial (and a bit beyond) course to extend over a year and to be taken (by requirement or suggestion) by students intending to specialize in the fields mentioned above during their first year of residence.

2. Microeconomics

It has been my experience in teaching Econ. 620 that one semester is a rather brief period for covering the range of microeconomic theory a graduate student in Pennsylvania ought to be exposed to. As it is the case at both Harvard and MIT, I would propose that the course Econ. 620 be extended to a year course. Roughly speaking, the topics to be covered might be:

1. Theory of Consumer Behavior
   1. the Hicksian version
   2. the von Neumann-Morgenstern version, including the Friedman-Savage paper
2. Demand functions, elasticities, etc.
3. Theory of the firm; output and price determination
   1. Production functions
   2. Cost functions and their relations to i.
   3. Revenue and profit functions and the profit maximizing hypothesis
   4. The perfectly competitive firm and industry, and their equilibrium; comparative statics; supply functions
   5. The monopolistic firm
   6. Monopolistic competition
   7. Duopoly and oligopoly
4. Factor employment equilibrium
   1. Factor demand functions
   2. Factor employment equilibrium under various market institutional arrangements
   3. Some income distribution theory
   4. Factor supply.
5. General Equilibrium Analysis; Input-Output models
6. Welfare Economics (Samuelson; Graaf)
7. Capital Theory (Fisher, Wicksell, recent contributions)
8. (Marginally) Some revealed preference theory; or neoclassical growth models; or alternative theories of the firm (e.g., Cyert and Marsh)

It would be desirable if students were sufficiently well-equipped mathematically to handle these topics at some level intermediate between Friedman’s Price Theory Text and Henderson and Quandt; however, since this is not the case at present some other alternative must be found, such as in the manner in which the propose mathematics course is taught, and the order in which topics above are covered. The split of the subjects could be a) through c) or d) for the first semester and the remainder for the second semester. Clearly, neither the topics proposed nor the split represent my immutable opinion and there is considerable room for discussion.

3. Statistics

At present the statistical training of our students suffers from their inadequate mathematical preparations.

It is my opinion that minimally we should require of our students that they be familiar with the elementary notions of statistical inference, estimation, testing of hypotheses and regression analysis at the level of, say, Hoel, or Mood and Graybill, or any other similar text, (a semester course). For students intending to specialize in Econometrics or other heavily quantitative fields, then it should be highly desirable that a year course be available, say at the level of Mood and Graybill, Graybill, or Fraser, Hogg and Craig, Brunk, etc., with suitable supplementary material. Since, we do have access to a statistics department it might be desirable for our students to take a suitable course there.

Again, due to the problems posed by the mathematics deficiency of incoming students, some accommodation must be reached on this score as well.

4. Econometrics

Econometrics should not be a required subject; rather the requirement—minimal requisite—should be confined to the one semester course indicated under III. It would be desirable to offer a year course to be taken after the statistics sequence and which would cover at the level of, say, Klein, Goldberger, or my readings showing applications and problems connected thereto.

Topics, could start by reviewing the general linear model, Aitken estimators and similar related topics; simultaneous equation and identification problems, k-class estimators, 3SLS, maximum likelihood estimation, full and limited information, Monte Carlo methods.

Also selected topics from Multivariate Analysis; specification analysis, error in variable problems; elements of stochastic processes theory and spectral and cross spectra analysis.

It might be desirable to teach these subjects in the order cited above, although it would appear preferable to have multivariate analysis precede the review of the general linear model.

5. General Comments:

I generally agree with Albert Ando’s memorandum on proposed curriculum revision in so far as they pertain to Mathematics requirements, Macro-economics and Monetary Theory.

I think that at present we require our students to take too many courses. I would favor only the following requirements; the basic Micro and Macro year courses. At least a semester of statistics, as indicated under III, and one semester in either economic history or history of economic thought—although I do not feel too strongly on the latter. I presume, in all of this that students in our program are only those ultimately aiming at specialization in Theory, Econometrics, International Trade, Industrial Organization, and possibly Comparative Systems, or Soviet Economics. It is my understanding that our curriculum will not cover those concentrating in Labor Relations, Regional Science or Economic History.

Thus, through their first year our students would be taking more or less required courses, with the second year essentially left open for their special fields of concentration.

Thus, the course program of a typical first year student will look more or less as shown in Albert Ando’s memorandum, p. 4, although I would be somewhat uneasy about requiring 6 hours of mathematics in the first term and 6 hours of statistics (econometrics) in the second term of the first year. Nonetheless I do not object strongly to this, and indeed in this past term many of the students taking 620 and 621 had in effect taken a six-hour course in Mathematics, 611 as taught by Dorothy Brady and approximately 3 hours as taught by Albert Ando and myself.

Quite clearly the above are merely proposals intended to serve as a basis for discussion an ultimately for guidance of entering students in planning their program of study rather than rigid requirements.

Source: Duke University, David M. Rubenstein Rare Book and Manuscript Library. Economists’ Papers Archive, Lawrence Klein Papers, Box 19, Folder “Curriculum”.

Images: Left, Albert Ando; Right, Phoebus Dhrymes. From the respective obituaries above.