THE CHOICE OF AN INPUT-OUTPUT TABLE EMBEDDED IN REGIONAL ECONOMETRIC INPUT-OUTPUT MODELS

In this paper, we investigate the role of input-output data sources in regional econometric input-output models. While there has been a grea t deal of experimentation focused on the accuracy of alternative metho ds for estimating regional input-output coefficients, little attention has been directed to the role of accuracy when the input-output syste m is nested within a broader accounting framework. The issues of accur acy were considered in two contexts, forecasting and impact analysis f ocusing on a model developed for the Chicago region. We experimented w ith three input-output data sources: observed regional data, national input-output, and randomly generated input-output coefficients. The ef fects of different sources of input-output data on regional econometri c input-output models revealed that there are significant differences in results obtained in both forecast and impact analyses. The adjustme nt processes inherent in the econometric input-output system did not m ask the differences imbedded in input-output tables derived from diffe rent data sources. Since applications of these types of models involve both impact and forecasting exercises, there should be strong motivat ion for basing the system on the most accurate set of input-output acc ounts.