Paper Title:

Abstract:
Economic geography has, for a long time, had a split personality.  The part that leans toward economics takes the long view, looking for how forces for change will ultimately work themselves out in new geographical patterns of development (city systems, specialization, income levels, etc).  The part that leans toward human geography is more interested in what happens in particular places, and hence in differences in outcomes, and in how they are experienced.  Each of these is useful, in and of itself, but they leave economic geography in the position of merely measuring geographical outcomes of economic forces.   Instead, I propose that the two interact in mutually shaping ways, and understanding how they do so is the comparative advantage of economic geography.  Thus,  behaviors that agents use to "do globalization" are strongly structured from the top down (technology, capital flows, market integration), but then they take often surprising forms due to geographical forces (scale, mobility, interaction, and bundling).  There is a group of these forces which can be labeled « context, » and I suggest how territorial context and global forces shape one another.  This evolutionary approach draws on the metaphor of « Coleman's Bathtub »  to suggest how these interactions can be captured analytically. In other words, territories (regions, cities, nation-states), rather than mere objects of empirical explanation, are contexts in which wider economic processes are not just expressed, but created and transformed.  They then, in turn,  generate geographical outcomes.

Discussants:

Gilles Duranton - University of Toronto

Anders Malmberg - Uppsala University