[Access article in PDF]
Historical GIS as a Foundation for the Analysis of Regional Economic Growth
Theoretical, Methodological, and Practical Issues
The argument of this article starts with three key theoretical postulates: First, regional-scale explanation differs in important respects from either aggregate macroeconomic studies of development at the national level or microeconomic and business historical studies of individual firms. Second, neoclassical economic theory is unsuitable as a basis for the analysis of regional economic dynamics, and a different framework, called adjustment theory, should be used in its place. Third, utilization of adjustment theory in the historical [End Page 575] geographical domain leads directly to the deployment of both the concepts and methods of geographical information science as the foundation for empirical inquiry.
While the first two assertions would undoubtedly merit one or more articles in their own right, space only permits us to develop an outline justification of each. However, the theoretical questions raised in this justification provide the necessary background for the detailed investigation of the role of historical GIS in the analysis of regional economic growth, the main focus of our discussion.
The first section of the article therefore addresses these theoretical issues, then examines a number of methodological issues and design principles that arise from the use of GIS in historical regional analysis. The practical implications of these issues and principles are illustrated by means of two individual case studies of GIS creation, both of which form part of a larger substantive investigation into the spatiotemporal dynamics of regional industrial development in the northeastern United States, during the second half of the nineteenth century. The first case study concerns the development of the railroad network of the Middle Atlantic states prior to 1900 and is thus at quite a broad geographical scale. The second is a more localized study of mining investment in the anthracite coalfields of Pennsylvania during the same time period. The concluding section of the article will summarize key implications and benefits of adopting a GIS approach to the analysis of regional development over time.
Theoretical Issues in Regional Analysis
While national-level studies of economic development abound, those at the regional level, however broadly defined, are comparatively few and far between, although there are a number of notable exceptions (Baer 1981; Baker 1949; Conzen 1977; Cronon 1991; Earle 1992; Engerman 1971; Knowles 1997; Meyer 1983). Also, Douglas North’s (1961) classic study of economic growth is distinguished by a strong regional orientation in the organization of the work. Otherwise, regional contributions to, or impacts of, wider economic growth have to be inferred from industry studies, including studies of railroads (Chandler 1977; Fishlow 1965; Fogel 1964; Vance 1995) and iron and [End Page 576] steel (Walsh 1975; Warren 1973), or detailed historical treatments of individual companies (e.g., Delaware and Hudson Company 1925; Dilts 1993).
Regional analysis differs from national-level studies in at least three key areas:
- The nature of the dynamics of interest;
- The effect of the chosen scale on appropriate objects of study; and
- The interaction between economic decision making and geographical constraints and the resulting pattern of built forms (e.g., industrial plants, transport infrastructure) in specific locations.
The choice of underlying theoretical framework has a very important bearing on the approach to these topics, which will now be examined in turn.
As Walter Isard (1960) has noted, changes in the national economy are a weighted sum of changes in regional rates of growth or decline. In one sense this statement is a statistical truism, but it can also imply that there may be important economic dynamics operating at the regional level that are damped out, or even concealed, in the averaging process that leads to the production of national-level figures for output, employment, and growth. For example, changes in some regions may lead or lag those in others, and the resulting averaged national picture may not actually reflect the fortunes...