Question:
I am familiar with Karl Fox et al. "functional economic areas" framework of regional modelling. I am also aware of REMI products. These latter models are widely used by state and local governments. But I am not aware of the availability of computer models that would use functional economic areas per se. Can you inform me if there are such models that might be an alternative to REMI? Thank you.Answer:
Any subnational economic modeling includes two challenging components: data and structure. Detailed subnational data for the US is automatically suspect because input-output tables are not collected at any subnational level. Interstate (domestic) trade is also a problem. (A small price to pay for the interstate-commerce clause: no state or municipal-level customs clearance => no data). The Commodity Flow Survey is the only source for domestic trade and there are significant problems with these data. They are not designed to measure source-destination trade as modeled by trade economists. My understanding is that the CFS tracks freight based on waybills, so trade from A to C that stops in B can be counted as A to B then B to C. Effectively inflating the amount of trade. IMPLAN is a for-profit company that constructs sub-national accounts (to the zip code level if you like) and a interface that allows input-output ‘impact’ calculations. Remember the finer the resolution the more suspect is the data because it is based on some scheme for sharing out the aggregate US accounts. The IMPLAN model is based on input-output techniques developed in the 1950s. Input-output studies have not been seen in top general-interest academic journals in some time, although they may still appear in regional-economic journals. REMI seems to provide data and models for policy advocates. The REMI structure is eschewed by academics because of its questionable theoretic foundations and general lack of transparency. There are a number of contemporary general equilibrium models (especially in the area of environmental economics) that utilize the IMPLAN data. These models include price responses and stand up to academic review with varying degrees of success. Another state-level effort has been led by the CoPS group out of Australia (Victoria University) doing consulting for the US government. My understanding of this model is that it computes the US general equilibrium and then shares out the impacts to each state. There is a new open-source push to build transparent, replicable, and extensible subnational accounts (as a substitute for IMPLAN) for contemporary general-equilibrium modeling analysis. This effort is being led by Tom Rutherford at the University of Wisconsin under the WiNDC project (http://windc.wisc.edu/index.html).