Import Substitution and Changes
in Structural Interdependence:
A Decomposition Analysis
D. P. PAL
In an input-output framework, import substitution in a sector influences and is influenced by import substitution in other sectors because of the simultaneity that exists between the sectors. The higher the degree of such simultaneity, the higher the extent of indirect repercussion. Estimates of import substitution made in an interindustry framework may thus be called measures of the total (direct plus indirect) import substitution. In this case the functional interdependence of the different sectors is considered in estimation. The estimates obtained may thus be called functional estimates, which are contrasted with the statistical estimates obtained in the isolated framework.
Import substitution is usually defined at two levels. At the absolute level it is the reduction of imports over time. At the relative level it is the reduction of the import coefficient over time. Controversy arises regarding the nature of imports (and hence of import coefficients) to be taken into account. Chenery, Shishido, and Watanabe ( 1962) treated imports as perfect substitutes for domestic goods and lumped them (i.e., imports and domestic outputs) together to define import substitution as the reduction of the actual import/total supply ratio (i.e., the actual import coefficient) over time.
Morley and Smith ( 1970) extended the Chenery-Shishido-Watanabe measure. They defined the concept of "implicit imports," which are obtained through the transformation of the actual imports of different products, into "the domestic production necessary to substitute completely for imports, holding all final demands constant." Import substitution in the sense of Morley and Smith is the reduction of implicit imports to implicit supply (i.e., the implicit import coefficent). The novelty of the Morley-Smith measure is that it counts both direct and indirect production requirements of imports, and thus imputes import substitution to those sectors that have no imports at all.
In the Chenery-Shishido-Watanabe measure (and hence in the Morley-Smith measure) domestic goods and imports are lumped together without distinction in final demand and intermediate demand and hence in the input-output balance equations. Guillaumont ( 1979) made such a distinction and measured the "import content" of the actual import coefficient (here, import/domestic output ratio, mi/xi) of a "given