On the units of geographical economics

The new `geographical economics' introduced by Paul Krugman and others has been dismissed by geographers for a number of reasons, ranging from (mis)treatment of externalities to reliance on formal modelling. These criticisms may be appropriate, but they miss some underlying unit assumptions that cause geographical economics to deal poorly with issues of scale and space. The theories produced by Krugman and others rely on the economic units of the firm, industry and economy for analysis. In contrast, geographers often use spatial units which not only reveal different processes, but can also delimit the economic units. Acknowledgement of the difference in the assumed units of analysis between geographical economics and economic geography is necessary before any useful discourse between the two fields can occur.