A price-responsive framework for interregional input-output

In interregional analysis, supply models are based typically on defined sets of `representative firms' with given `averaged' technological and cost coefficients, such that, when aggregated as identical units into both sectors and regions, they reproduce the total output of each sector in each region. Then, deterministic models obtain profit-maximising patterns of input demands and corresponding output, prior to aggregating these quantities as if the firms were truly identical and homogeneous. In this paper, the above optimisation models are replaced by a probabilistic framework, which nevertheless guarantees convergence to the above classical deterministic solutions in the limit. The models are estimated on `observed' flows, technological coefficients and prices, and a CES function (optionally) handles substitution between primary factors. These new input-output models become explicitly price-responsive, allowing t×tonnement with spatial models of final demand. <!--ID="" Acknowledgments. Some of the above ideas evolved from collaborative work with BÃrje Johansson. Also, the author is very grateful for constructive advice from Kingsley Haynes, Geoffrey Hewings and Tatsuaki Kuroda. Of course, these colleagues are not responsible for any perceived deficiencies in the above presentation.-->