Actualización de la Inversa de Leontief

Abstract

Among the many matrix updating techniques available, basic RAS and its extensions are the most popular. RAS is a biproportional adjustment method, consisting in reiterative multiplication of row and column elements of a base matrix by correction coefficients. This technique is almost invariably used on the technical coefficients matrix (or on the intermediate consumption matrix), but it is adaptable to other cases. The most usual formulation of the technique represents it as an optimization program minimizing a distance among matrixes subject to several constraints. This paper shows instead a scale algorithm representation corresponding to the direct application of RAS on the Leontief inverse matrix. The requisite for obtaining a coherent solution in this case is to work simultaneously with the demand model and the price model.