The New Data Revolution in Regional Science: A Tribute to the Late Stan Czamanski

Introduction

The methodology of regional science is traditionally based on a sound mix of conceptual framing and operational (or applied) analysis. In recent years, we have witnessed an avalanche of new empirical data used in regional science research (e.g., social media, “big data,” sensor data, App data, etc.). Such new data emerge in particular as a result of digital technology applications and interactive media. But also in traditional statistical domains we observe a rapid expansion of the data base for regional science research, to the extent that a gradual shift from basic thinking toward applied research is taking place.

One of the pioneers in regional science who has devoted a significant part of his scientific work to applied research has been the late Stan Czamanski (1918–2012). He was fascinated by the question of how to introduce space in a measurable way as a key variable or a structural component in socio-economic research at the sub-national level. Based on a solid conceptual framework of the space-economy, he was able to design operational macroeconomic regional accounts and regional input-output tables, which he used as practical instruments to map out complex spatial-economic phenomena. He grew up in the quantitative research tradition pioneered by Walter Isard, Wassilly Leontief and Lawrence Klein among others; they all tried to develop measurement models for studying industrial structures and (spatial-)economic linkages, inter alia by adjusting (inter-)national data to regional economies.

The lack of high-volume computer capacity necessitated the first generation of quantitative (regional) economists to employ innovative means to measure and understand regional activities and spatial patterns of development. Czamanski’s seminal work on regional product accounts, spatial input-output tables, regional econometric modeling and industrial complex analysis is indeed noteworthy. He was inspired by the motto: “if you cannot measure it, you cannot say much about it that is meaningful.”

Since the first data revolution in regional science (the period 1960–1980) during which many attempts were made to quantify social science phenomena in space—a trend that can also be observed in geography, sociology or psychology—, quantitative measurement constituted a formidable step forward in analytical research in the spatial sciences. With the rise in computer capacity, a new trend emerged in the 1990s with the advent of spatial econometrics, characterized by a particular emphasis on spatially dependent or interacting phenomena leading to adjusted model specifications and test mechanisms related to space-time autocorrelation in complex spatial systems. In more recent years, we observe a new trend, viz., the third data revolution in regional science, where large volumes of data are becoming available as a result of the digital revolution, the wide-spread use of social media (e.g. Facebook, Twitter) and the use of digital information in spatial marketing or public administration.

This (third) data revolution prompts many unprecedented challenges to current regional science research in terms of new conceptual framing, in terms of data storage and in terms of data handling (e.g., privacy requirements, new statistical data mining techniques, spatial visualization methods, quantitative tools for digitized qualitative information). It is foreseeable that in the next decade(s) regional research will have to face the challenge of combining traditional statistical data with contemporaneous digital information sources to address critical global issues and to tackle urgent urban- and regional-oriented issues, e.g. climate change, health, well-being, viruses, quality of life.

At the age of the advancement of sophisticated sets of digital tools for a wide range of disciplines and research areas, the potential for measuring well-being, health and the perception of urban happiness is growing in a data-rich urban environment (Kourtit 2021; Kourtit, Elmlund, and Nijkamp 2020). The permanent pressure to improve the current digital information systems and data-analytics approaches is a clear manifestation of the relevance of advanced conceptual and methodological thinking on the principles and assumptions of (strategic) regional and urban planning. This is often done by focusing on advanced qualitative, quantitative and mixed analytics methods of critical input and output possibilities of the complex policy issues and urban multi-level quality of life (embodied in the XXQ-principle; see Nijkamp 2008). This further emphasizes the importance of a comprehensive and systematic hierarchical process of fit-for-purpose data decomposition (based on an information “cascade principle”) at all relevant geographical scale and time levels (see Mumford 1961; Hall 1998; Taylor 2004; Bettencourt et al. 2007) in the urban space-economy in combination with “intelligent transformation” (see Dick 2005; Beaverstock, Smith, and Taylor 2000; Pumain 2006; Nijkamp 2008; Kourtit and Nijkamp 2018; Kourtit 2021) to build intelligent and strategic policy handles.

A prominent question will also be whether the unprecedented rise in spatially relevant data volumes leads to better insights into the complexity of the space-economy, leave aside whether this may lead to better policy decisions. To address this question, a two-day international Advanced Brainstorm Carrefour (ABC) workshop (in April 2019) was held in Tel- Aviv and Haifa in honor of the late Stan Czamanski. This meeting was organized by The Regional Science Academy (TRSA) and supported by the Tel- Aviv University and the Technion—Israel Institute of Technology. The ABC workshop brought together scholars from all continents to reflect on the new data-analytics challenges of the post-Czamanski age. Both conceptual and applied studies addressing the above-mentioned challenges were welcomed at this fruitful meeting. We thank both universities for their hospitality in hosting this international expert meeting. Finally, we also thank the Ax: son Johnson Foundation (Stockholm, Sweden) for supporting new activities regarding a better understanding of urban agglomerations and statistics in the New Urban World.

The present special issue offers a selection of various contributions and findings from this ABC workshop. The first part, devoted to methodological aspects and precedents on measurement contains three contributions. The first contribution in this part is offered by Geoffrey Hewings and Peter Batey; the authors focus on progress in achieving greater integration between demographic and economic components in system-wide modeling. It underlines the importance of Czamanski’s Baltimore model as well as other influential research based on the two-sector economic base model, before proceeding to review more complex, spatially-disaggregated economy-wide models. Next is a paper by Kieran Donaghy that provides an overview of the “Rebuilding Macroeconomic Theory Project” (RMTP), which serves to footmark the development of macroeconomics and to outline responses by prominent macroeconomists to a set of questions posed by the organizers of the RMTP project. And finally, in this first part, Dani Broitman, Itzhak Benenson and Daniel Czamanski offer a comprehensive agent-based model of a closed system of cities, focusing on the relative influence of globalization and the localized entrepreneurial ecology on innovation.

The second part of this special issue addresses issues related to data analytics in regional science; it contains also three contributions. The first paper in this paper is written by Peter Nijkamp and Waldemar Ratajczak. They authors offer a concise discussion of the history, the complementary side-roads and the vast application potential of the gravity model in regional science and geography, and in spatial sciences in general (including trade theory). The second paper is by Marie Hårsman Wahlström; Karima Kourtit, Bart Neuts and Peter Nijkamp, the authors conceptualize the nature and composition of urban characteristics of place attachment and appreciation (“city love”) in terms of a “city soul” and “city body” analytical approach. Finally, Daniel Felsenstein and Michael Beenstock propose an experimental design that may serve as a methodological prototype for further tests of “common correlated effects” (CCE) as a solution to the absent spatial data problem.