Chapter 2: Training Days Episode 5: Walter Isard’s Legacy

Arriving in London on the first day of September 1970, I was met by Chris who by prior arrangement would be flying back home to the US with me.  I made friends with Chris since 1969 when he was a visiting Fulbright scholar spending a year in Monash undertaking a study of the colonial Pahang civil administrative service for his Yale history Ph.D, and had become acquainted with us amongst the Malaysian student crowd.  He had kindly offered to help me settle in into Penn.   We departed for New York in one of those new Boeing 747, and on arrival he took me home to his parents house in Stamford, Connecticut, to stay a few days before driving me down to Philadelphia over the Labour Day weekend.  That was a welcome introduction for me to American life, that I had only previously known through their television programmes.  In the City of Brotherly Love, Chris helped me find a room to rent on 42nd Street, West Philadelphia, within walking distance of the Penn campus on Locust Walk and 40th Street.

Transit in London, 1970

When the new fall semester opened I registered in the Masters level  courses offered by the Regional Science Department, which required eight taught courses in preparation for the preliminary examinations before one is admitted into the PhD program.  These required courses were necessary to provide a common base for the students undertaking graduate work in regional science who came from diverse undergraduate disciplines including economics, geography, mathematics, engineering and planning.  I was quite pleased about that because that enabled me to double-up on my formal economics courses that I had missed in Monash.  When I told Bruce Allen, the coordinator of the Masters program that I had gone through Samuelson’s textbook, and had audited some economic undergraduate courses in my bachelors degree he was quite pleased and said I should not have any problems with the graduate micro- and macro-economics courses as part of the regional science entry programme.  Fortunately, I was allowed a waiver of the foreign language requirement because of my German course at Monash.  But what pleased me most was the requirement to take up Ron Miller’s mathematics course in the first term, to be followed up in second semester with the methodology course also taught by him with applications in regional science.  For this the Methods in Regional Analysis book that I had discovered  while in Monash was most helpful.  I also took up John Parr’s urban systems course, which was very familiar for me because of my exposure to urban geography in Monash, and similarly Tom Reiner’s regional development course, which was my specialty in my undergraduate work. Walter Isard, professor and departmental head, taught the required theory course for both semesters of the regional science master’s programme. 

Ben Franklin, founder of Penn, in the snow in front of the original College Hall

Walter Isard was of course the legendary founder of regional science as a stand-alone discipline in social science.   In the 1940s he joined other economists searching for a more sophisticated way to measure and understand economic activity at the sub-national regional scale. After graduating in economics from Harvard in 1943, and working with Leontief, the Russian émigré who developed input-output analysis to its most sophisticated level, he devised and taught a course on location theory and regional development. He was influenced by the German location theorists, which resulted in the path breaking Location and Space Economy book I mentioned earlier.  With the help of Lawrence Klein, the economic Nobel Laurette in econometrics, he established the first Regional Science Department in the famous Wharton School at the University of Pennsylvania, which produced the  first Ph.D in regional science, Bill Alonso, in 1960.  The field attracted a wide range of scholars interested in urban and regional economics, economic geography and regional development planning. The core concepts of regional science were developed and propagated through the Regional Science Association (since 1989 the Regional Science Association International) and the Journal of Regional Science, both of which he founded in the 1950s.

Walter Isard, 1919-2010

By the time I arrived at Penn, regional science was more than ten years old and the field of choice for those interested in the regional impacts of national or global processes of economic and social change.  In 1969, Isard and his students had published the General Theory of social, economic, political and regional change, and this became the main text of the theory course conducted by Isard.  I was profoundly influenced by this text, which defined much of my approach to policy analysis in my future work.  The  regional science programme opened my intellectual vision to a wider interdisciplinary scheme of things, that would prepare me for further acquisitons of knowledge, in fact far beyond that specialty itself into science, mathematics, cosmology and even deeper into development economics. Walter Isard left an indelible legacy in the field of regional science, whatever happens later in its fifty year history when its fundamental issues devolved into new economic geography and urban and spatial economics, but even more for me personally a celebration of intellectual exploration and discovery.  Isard died in Philadelphia in 2010 aged 91.

Having been nearly five years away from home, I had become a man in a hurry to complete my graduate studies at Penn.  The urge to return home to start my career was a powerful motivator.  Rose arrived safely in the early winter of 1971, and we moved into a rented apartment just a block down on 42nd Street from the house I had rented a room in the previous fall term.  

As I was settling in into the first term at Penn, I received news from Geoff that our Kuala Lumpur trend surface analysis paper was accepted by Regional Studies journal.  Now as a published author, I was much encouraged to pursue the new course of studies in regional science.  I completed the eight required courses and sat for and passed the preliminary exams by the end of spring for me to begin the PhD programme.  This required a further 12 courses before you can take the PhD qualifying exam, and passing that to start working on your PhD thesis.  At that point you are classified as ABD (all but the dissertation) status.

I completed those requirements by the end of 1972 by taking the maximum load per semester, including courses during the summer session.  I used this opportunity to take up Klein’s econometrics course, Julius Margolis public economics where I was exposed for the first time to social choice theory, and two other economics courses in advanced macroeconomics and public finance.  The rest were courses in regional science including Tony Smith’s advanced mathematics who focused on measure theory, and Miller brought out his newly-published Mathematical Methods book as the text for the advanced methodology course.  I took this time also to produce two required papers in lieu of thesis for fulfillment of the Masters degree requirement, one for Tom Reiner on regional development policy and a paper for Julian Wolpert on a choice-theoretic model of locational decisions using Isard’s incremental bid-curves framework.  Wolpert had a team of research students working on an NSF-funded study of conflicts in locational decisions, such as highway routing or location of nuclear power plants.  After I graduated in 1973, Wolpert would leave Penn to take up the chair of geography in Princeton University.  With these two papers I duly received the MA in regional science degree in 1972.

Locust Walk, Penn campus in the fall, the famous Wharton School on the left.

Because of my sense of urgency, I did not have much of a social life in graduate school, compared to my undergraduate days in Melbourne.  My circle comprised mainly my fellow enrollees in regional science, which included the Americans Don Elmer and Pete Jones, the Brazilian Rude Valle, a Canadian Richard Zucker and Yamashita a Japanese scholar.  A few seniors were also part of the McNiell Building crowd, namely Kawashima, Fujita and Hong a South Korean. To this can be added Rose’s workmates Rudy and Betty.  Rose took a summer job as a proof reader with TV Guide in Radnor, a  Philadelphia Main Line suburb, and kept that job until the  end of 1972.   She decided to take a rest from continuing her studies.  She would commute every working day from the 30th Street station to work, and take the tram back to our house.  Just in front of the house was O’Malley’s corner store which provided most of our daily needs.  

To this campus community were added in 1971 the few Malaysian students such as Farouk Abdullah and Raja Lope, accompanied by their wives.  Both were doing the Wharton MBA, the former a senior civil servant who would go on to head the Malaysian Natural Rubber Board, while Raja Lope continued to his PhD under Alan Walters at London University, and returned home to become a senior banker.  Across the city to the Temple University campus in North Philly were two UKM scholars, Ismail Hamid and Faisal Othman, studying comparative religion under Ismail Faruqi.   Later in the fall of 1971, Zeti Ungku Aziz arrived with her husband John (Muslim name, Aiman, who also did a DBA at Penn) to undertake her doctorate in monetary economics at Penn.  Zeti would later return to Bank Negara and eventually become Governor of the Malaysian central bank.  I had given her most of my economics notes to assist in her studies.  We only travelled in the States in the two summers we were there, one Greyhound trip to New Orleans and the other to Montreal during the 1972 Olympics.  Other shorter trips were to New York and Washington DC during the Thanksgiving weekends.

I passed the PhD qualifying exams, a four-hour open book affair, at the end of fall 1972, and started working on my PhD thesis.  At the same time I was appointed a lecturer in the department, and asked by Walter to take the undergraduate introduction to regional science course in the spring term, with about twenty students.   This course became the basis later of my introductory course in geographic theory in USM.

While chalking up experience in teaching, most of my time was however devoted to research towards my doctoral dissertation.  It was almost natural that I would choose regional development as my subject of focus.  Even more than that, I was attracted to the intersection of theory, analysis and policy that a regional science approach to development calls for.  Thus, I decided to frame my research around developing a policy framework for regional-national development planning, and chose Miller, Reiner and Parr as my dissertation committee, the latter serving as chief supervisor.   In choosing this topic I was very much influenced by Jan Tinbergen’s book on A Theory of Economic Policy, and subtitled my thesis as a contribution to a theory of regional policy.  Tinbergen, a Dutch economist was together with Ragnar Frisch, the first winners of the Nobel Prize in economics, the year before Samuelson got his in 1972.   Tinbergen’s work focussed on econometric models in formulating macroeconomic policies and devising instruments to implement them; this formulation was taken up almost thirty years later by economists in the sub-discipline of mechanisms theory.

My approach was to model the regional policy formulation process as a mechanism to resolve goal conflicts in a multi-level development planning framework.  For this I developed a two-level mathematical optimization model along the lines of Mesarovic’s maximal-infimal framework to arrive at a constrained solution to an aggregated national-regional policy objective.  In real life there would be expected to be such goal conflicts due to differences in regional factor endowments, social and political conditions, and economic and fiscal dictates, which would be difficult to aggregate consistently in the policy planning process.  In the mathematical model I constructed, I framed the goal-conflicts as a two-level objective function which are aggregated (provided the goals are linear and additive, eg between national GDP and regional GDP objectives) and resolved the conflicts between the upper and lower level as maximizing their goal-achievement ( so that should there be a shortfall in goal-achievement due to their respective and cross-level constraints) by minimizing the shortfall in the aggregate national-regional situation.  Such objective functions maybe non-linear, but using the goal-programming method these can be linearized to enable the use of linear programing, a mathematical optimization tool,  to achieve a solution.  To demonstrate this mathematically, I employed Dantzig’s Simplex Theorem that a unique solution exists in which all interactive (between national and regional) terms were cancelled, to achieve the maximal objectives. When I presented this result in a paper to the inaugural Pacific Peace Science conference in Tokyo in November of 1973 after my return to Malaysia, Fujita, a mathematician and regional science graduate supervised by Tony Smith who graduated together with Kawashima and Hong a year ahead of me, and who was the designated commentator to my paper, could not find fault with my method, only that it was highly intuitive.  That was validation enough for me.   Fujita would become a famous regional scientist publishing two seminal works, one on Spatial Economy with the economic Nobel Laurette PauL Krugman and Venables, and the other an authoritative two-volume collection presenting a selection of seminal articles between 1949-2002 investigating the spatial aspect of economic processes and development.  A version of my goals conflict paper would later be published in the inaugural International Regional Science Review in 1974. 

The rest of the thesis was built around this central feature, including the dynamic multi-year version of the model using recursive programming.  The theory part was an expansion of the MA paper I submitted to Reiner on regional policy, incorporating also the choice-theoretic locational model I did for Wolpert, all elaborated further along the lines of my favourite development economist then (famous for his unbalanced growth strategy for developing countries) A.O Hirschman’s dictum to “treat the nation as a (sub-national) region, and the region as a nation” in devising development policy.   The draft version of the chapters were typed by Rose evenings in between her Radnor job, on electric typewriter in the days when there was no desktop word processors yet.  Finally, after stenciling in the mathematical formulas and consultation with my supervisors of the draft chapters as they were finished in sections, I submitted the 300- odd paged manuscript for final rendering by a professional typist.

On a sunny day in June 1973, I took two unbound copies of the type-written manuscript to the Regional Science Office to formally submit them as partial fulfillment of the PhD requirement.  Ron Miller officially accepted the thesis submission, joking after flipping through the manuscript that I had modelled my thesis after an Italian car (he must have in mind the Mesarati!), just as Walter Isard was coming out of his office.  After a brief exchange (“has he completed it? Ok?) with Miller, Walter asked the secretary for the notification form to the Dean’s Office and duly signed it.  They both proceeded to congratulate me for having passed the PhD requirement, just like that! Having completed my training days, and holidaying in Europe in August 1973 en route home, I was conferred in absentia the doctoral degree in regional science at  Penn commencement ceremonies later the same month. Years later an undergraduate student of mine at USM, Rashidah Shuib, undertaking her postgraduate studies in the US would present me with a printed copy of my PhD dissertation from University Microfilms.  That was a nice touch to seal my training days.


Read Chapter 3: An Academic Firmament; Episode 1: Coming Home Here