Saturday, October 15, 2016: 2:35 PM
A standard gravity model in the tradition of Anderson and van Wincoop (AER 2003, Vol. 93 No. 1, pp. 170-192) is applied to migration data and used to construct several regional divisions of a multi-regional economy. This establishes a theoretically and empirically founded description of the economic geography with no use of physical distance data. The methodological approach is exemplified by using US Census data for interstate migration to develop a large set of regional structures of the United States. The results are compared with the official Census Regions and Divisions of the US that have been in use for more than a century. The methodology is also applicable to other regional levels and to other types of data for resource exchange, such as trade in goods or services. Based on a chosen data set, the model can contribute to the understanding of the strength and structure of regional ties that arise from resource exchange and how this develops over time. The discussion of results with migration data makes reference to preliminary results of similar analysis with some European data. The simplicity of the approach, combined with its apparent potential to provide descriptions of the economic geography that are typically but not always intuitive, looks promising for further research. The methodology can also be seen as a bold, simple attempt to stretch the use of economic theory beyond its normal scope and to reflect on issues that often receive more attention from economic geographers and regional planners than economists.
- - -