Economists in the City #4

A Blogged Conference

Economists in the City #4

Economists in the City #4 870 546 Cléo Chassonnery-Zaïgouche

The Institutionalization of Regional Science

In the Shadow of Economics

by Anthony Rebours

The history of regional science offers an interesting case study, as well as a one of the few examples, of the institutionalization of an entirely new scientific field in the years after 1945.  Its foundation by Walter Isard and a group of social scientists in the 1950s represents the most institutionalized attempt to stimulate the relationship  between economics and geography. The original project of Isard, who was trained as an economist at Harvard, was to promote the study of location and regional problems.

And at the outset, regional science was, in various ways, a success. It attracted many scholars from different disciplines, mostly economics, geography and urban/regional planning, and it quickly became institutionalized formally through the foundation of the Regional Science Association (RSA) in 1954 and establishment of a Regional Science Department at the University of Pennsylvania in 1958. At the same time, the creation of the Papers and Proceedings of The Regional Science Association in 1955 and of the Journal of Regional Science in 1958, offered new publication venues for scholars interested in location analysis, in particular quantitative geographers who found it difficult to publish in traditional geography journals. Within economics, regional science influenced analytical works in urban economics, as, for instance, William Alonso’s thesis, widely recognized as one of the foundational works of urban economics, was written at Penn under the supervision of Isard in 1960.

However, the prevailing processes of knowledge production and evaluation which shaped the emergence of this new field were deeply influenced by economics. Geographers became dissatisfied with Isard’s vision of the hierarchical division between geographers and economists, and the primacy given to economic theorizing and modelling as the core of the new regional science. Thus, the social organization of the field of regional science and its interactions with other disciplines mirrored the particularity of economics, a hierarchical discipline organized around a strong theoretical core and an insularity from the rest of social sciences. In this short article, I discuss the findings of an analysis I have conducted of the contents of the main journal for the field – Journal of Regional Science –and associated archival materials, in order to shed light on the ways in which this field was institutionalized.

The emergence of regional science as a field of study

Regional science emerged in a particularly favourable context. In the US, the impetus for studies about regional development which began during the 1930s, supported by the success of the Tennessee Valley Authority program, persisted after the war. The Second World War and the ensuing Cold War provided new opportunities for the development of scientific research with an unprecedented increase in funding, student enrolment and collaboration between academics and external bodies, such as military institutions. This period confirmed economists’ aspirations to be treated as scientists, and resulted in the increasing prevalence of statistical methods and mathematical modelling, and the accompanying theory of rational and maximizing agents.

In 1942, Walter Isard obtained his doctoral degree under the supervision of Alvin Hansen, who was attached to the National Resources Planning Board, and Abbott Usher, who taught him about the  German tradition of location analysis, at the Economics Department of Harvard. There, and during a graduate fellowship at Chicago, he encountered other leading economists, such as Edward Chamberlin, Joseph Schumpeter, Jacob Viner, Frank Knight and Oscar Lange. At the end of the war, Isard produced a series of research articles in which he offered conventional economic analysis of production costs about the regional implications of the development of the airline industry, the atomic energy industry, and the future location of the iron and steel industry. At the time, these industries were considered as particularly important for national security and economic development.

In the late 1940s, Isard became increasingly concerned about the lack of interest among economists in the location of economic activities. His perception of the subject was not really different to his colleagues, but he wanted to improve the theory they used, which, following the British tradition of the late 19th century, suffered from a lack of spatial dimension. He did not seek to challenge the general equilibrium economic theory that was becoming dominant, but sought instead to integrate a spatial aspect within it.  

In the late 1940s, he started to be more active in the promotion of location analysis but failed to convince the American Economic Association (AEA) to organize sessions on regional topics at the annual conventions. In 1949 Isard was recruited to Harvard by Wassily Leontief to develop an input-output approach to regional development. During the war, input-output analysis received much attention because it enabled the American Air Force to identify the best targets for bombing. As a consequence, Leontief had received large research funds to develop his input-output framework. Drawing on Leontief’s financial resources, Isard was able to organize a series of multi-disciplinary sessions on regional research at meetings of various social science associations between 1950 and 1954. An informal newsletter was also created to disseminate the discussions and papers presented at the meetings.  

In 1949, at Harvard, Leontief also persuaded the faculty to create a new course on location theory at the Economics Department in which Isard would teach. During the course, he promoted the same kind of research he was doing at the Leontief project and that he would continue to conduct and support after having established formerly regional science. In a context where there was a large influx of war veterans returning to Harvard to complete their graduate studies, Isard managed to gather around him a core of young scholars to contribute to this work.  

The institutionalization of regional science

In 1954, after four years of informal meetings and discussions, the Regional Science Association was officially created during a meeting held conjointly with the American Economic Association and the American Social Sciences Association this time. Sixty participants from different disciplines—economics, geography and planning being the largest — as well as organisations like the RAND Corporation and Resources For the Future, were present. The papers presented were published in the first issue of The Papers and Proceedings of the Regional Science Association which was established at the same time. In 1956, Isard opened the first PhD program in regional science at the Penn’s Wharton School, and, in 1958, the first Department of Regional Science. The same year, the Journal of Regional Science, the future leading journal of the field, was founded.

These events were key to the institutionalization of the field, and reflected the thinking of Isard and his colleagues about the main focus and boundaries of regional science. This reflected an amalgam of diverse approaches to the study of regional and spatial issues, drawing on different disciplines, in particular economics and geography, with a strong emphasis on the same kind of analytical and statistical methods he learned at Harvard and from his work with Leontief.

In what follows, I look more closely at the constituent parts of the new discipline.

Figure 1. Journal co-citation network 1958-1967 (2 or more co-citation links), the size of the nodes and links being proportional to the number of links. Source Web of Science.

The centrality of economics for regional science is clearly visible in Figure 1, which maps the network of co-citations for articles published in the Journal of Regional Science (JRS) between 1958 and 1967. The co-citation technique allows us to measure the conceptual proximity of different journals cited in different papers of JRS. This technique is complemented by the use of a community detection algorithm in order to identify coherent sub-groups that have stronger links with each other than with the rest of the journals. The most striking feature of this mapping is the distance and the net distinction between economics journals and geography journals. While some leading journals such as the Geographical Review and the Annals of the Association of American Geographers, were among the most cited in the network (respectively the fourth and fifth most cited), geography journals occupied a peripheral position, along with journals of other disciplines like sociology (green). More surprising is the relative prominence of psychology journals (blue) which were more strongly associated with economics journals (purple) and represented the second most cited discipline of the period. However, in the next decade the situation completely changed, and psychology journals received less than 1% of the total citations in 1968–1977 (Table 1). For the whole period, 1958 to 1977, the geography is the second most cited discipline but with only 13,4% of the total citations of the period, far behind economics with 55,2% of the citations.

Table 1. Disciplines cited in the Journal of Regional Science (two percent of citations or more). Source Web of Science, using the National Science Foundation disciplinary categorization of journals.

This first result is consistent with the idea, expressed by Isard in Location and Space-Economy (1956), of a hierarchical division between economists, who provided the analytical foundations of regional science, and the geographers, who provided the empirical facts and testing. Another way to confirm this asymmetrical relationship between economics and geography is to compare the most cited disciplines (Table 1) with the disciplines that cited the most JRS (Table 2). While geography journals were by far the ones cited most in the JRS, with 44,1% of the total citations, they only received 10,7% of the citations within it. At the same time, economics journals, which represented only 12,6% of the total citations to the JRS, were the most cited,  41,5%. Regional science, thus, was more important for geographers than economists, while the reverse was not true as economics was more important for regional scientists than geography. This result is also consistent with the idea that the quantitative turn in geography and the emergence of regional science were closely associated.

Table 2. Disciplines citing the Journal of Regional Science (two percent of citations or more). Source Web of Science, using the National Science Foundation disciplinary categorization of journals.

These trends persisted in the next period (1968–1977). The size of the network in Figure 2 is representative of the increase of publications of the JRS during the period, with an increase in issues per years after 1969. The separation between the cluster of economics journals and geography journals persisted. Moreover, and despite an increase of the proportion of citations to geography journals, economics became even more important in the network and accentuated the difference of size with geography (Table 1). On the other hand, regional science became even more important for geography as the discipline represented 50% of the citations to JRS in the period, while it also received more citations from economics journals in this period than in the preceding one (Table 2). More generally, the data show that economics and geography were the two most important disciplines for the authors who published in the JRS between 1958 and 1977, a trend that has continued after 1967.

Figure 2. Journal co-citation network 1968-1977 (4 or more co-citation links), the size of the nodes and links being proportional to the number of links. Source Web of Science.

The fact that the JRS was much less quoted by economics journals doesn’t mean that it was completely ignored by economists as, in fact, the JRS was among the most cited economic journals in 1970. However, it shows that regional science was more discussed by scholars publishing in geography journals than economics. As already indicated, this situation is certainly related to the dynamics of both disciplines at the time. While, the identity of economics was legitimated and reinforced by its success during the war, in geography, there was an increasing dissatisfaction with the regional geography approach that dominated the field in the1950s. The Cold War context facilitated the promotion of a new generation of quantitative geographers looking for more scientific methods. Most of them were early members of the Regional Science Association, and as Brian Berry, were interested in the potential of regional science to transform geography. On the other hand, the stronger identity of economists meant that when they associated with other scholars, they were inclined to retain their own frameworks and methods, as Walter Isard did for regional science. However, by the mid-1970s, regional science experienced a progressive decline when geographers started to distance themselves from the analytical methods that were promoted by Isard. But even after the Regional Science Department at Penn closed its doors in 1993, regional science journals remained a going concern and continued to promote studies of spatial issues notably from urban economics and, after 1991, New Economic Geography.

Anthony Rebours is currently a graduate student at University Paris 8 and a young fellow of the Center for the History of Political Economy (CHOPE) at Duke University. His dissertation deals with the relationships between economics and neighbouring disciplines such as geography and regional science. It combines archival work and sociological methods for quantitative history.

Other posts from the blogged conference:

Cities and Space: Towards a History of ‘Urban Economics’, by Beatrice Cherrier & Anthony Rebours

Economists in the City: Reconsidering the History of Urban Policy Expertise: An Introduction, by Mike Kenny & Cléo Chassonnery-Zaïgouche