Effects of parametric uncertainty and technological change on input-output models / CAC No. 156

A method for restructuring input-output (I-O) models is demonstrated so that real technological options are more accurately reflected. Parameters are reordered to provide information on (1) how to set priorities for acquiring updated data; (2) how to construct more accurate I-O tables for a given cost; and (3) how to identify technologies where a small change will have a large impact on a policy goal. Selective updating of a model's parameters allows uncertainties to be reduced. Only two percent of the parameters were updated in the demonstration for each of four classes of functions (scalar, vector, submatrix, and Leontif inverse), which was sufficient for the scalar and vector importance functions. A reduction in tolerance occurs as the range space increases and wider classes of potential problems are introduced. 23 references. (DCK)