by Dirk Ehnts
No, this is not about the New York Cosmos. Instead, I have recently co-authored a paper named From New Trade Theory to New Economic Geography: A Space Odyssey (download). It is on the history of economic thought and, as it often happens, not all ideas fitted into the paper. The "troubles" started after reading this part of Carl Sagan's 1980 book Cosmos:
"Most of the world's great cities have grown haphazardly, little by little, in response to the needs of the moment; very rarely is a city planned for the remote future. The evolution of a city is like the evolution of the brain: it develops from a small center and slowly grows and changes, leaving many old parts still functioning. There is no way to rip out the ancient interior of the brain because of its imperfections and replace it with something of more modern manufacture. The brain must function during the renovation. That is why our brainstem is surrounded by the R complex, then the limbic system and finally the cerebral cortex. The old parts are in charge of too many fundamental functions for them to be replaced altogether. So they wheeze along, out-of-date and sometimes counterproductive, but a necessary consequence of our evolution."
This is quite interesting. In the New Economic Geography (NEG), what is deemed to be science is the creation of general equilibrium models. This basically means that the whole system goes from zero to equilibrium in under 3 secs. Wow! In the core-periphery model, history does play a role since there is non-linearity in the model. This is quite realistic, but for me the question becomes whether this model is empirically testable. And until now, the answer has been no. Well, almost. There have been some papers on different aspects of one or the other NEG model, but these mostly tested certain properties.
The Heckscher/Ohlin model of intersectoral trade was nice because you could use it to explain actual trade, and it more or less fitted. With the NEG, I am not aware of anyone who fit the whole thing to a real world case, except for me and my co-authors. You find our results in the Baltic Journal of Economics. The question then is one of methodology. If history plays such a large role, how much sense does it make to model full equilibrium? Would not partial equilibrium be more proper, since we are more interested in marginal changes? For example, there are many economists who believe that the core-periphery structure of Europe will be more pronounced in the future because of agglomeration forces like spillovers and linkages. However, the core-periphery structure might only be sustainable as long as capital flows from the core to the periphery to balance their current account deficits. A rebalancing would probably lead to a growth of industry in the periphery at the cost of shrinking industry size in the core.
What sense does it make to model in general equilibrium? A lot of sense if the assumptions are fulfilled. However, they are not, and this takes away a lot of the models explanatory power. Current accounts have not and will not be balanced. Capital flows matter for industry location. Omitting the monetary side can render the models useless. However, it is inherently difficult to model a monetary side to such a restricted model as the core-periphery model. Nevertheless, today's interesting questions call for such a model.
Dr. Dirk Ehnts is a research assistant at the Carl-von-Ossietzky University of Oldenburg (Germany). His focus is on economic integration and economic geography, covering trade, macro and development. He is working at the chair for international economics since 2006 and has recently co-authored a book on Innovation and International Economic Relations (in German). Ehnts has written at his own blog since 2007: Econblog 101. Curriculum Vitae.