ABSTRACT
The schematic uniformity of standard input-output computations accounts for certain practical advantages of that approach as well as for some of its peculiar limitations. In current statistical practice, the solution of the difficulties described above is sought in aggregation. The lack of perfect correspondence between the sectoral headings of two input-output tables might, however, frequently reflect the presence in one of the two economies of some goods or services that are neither produced nor consumed in the other. The method of double inversion described below permits us to reduce to a common denominator two input-output matrices that contain some comparable and also some incomparable sectors. In contrast to conventional aggregation, such analytical reduction is achieved without distortion of any of the basic structural relationships. The comparability of input-output tables attained through double inversion is limited in the sense that their respective structures are described only in terms of input-output relationships between goods and services of directly comparable kinds.
