Sectors associations and similarities in input-output systems: An application of dual scaling and fuzzy logic to Canada and the United States

Abstract.

Various authors addressed interindustry linkages in the context of input-output systems, however many focused on the identification of key sectors in the economy. Sonis et al. (1996) offered as a field of influence theory an alternative approach focusing on the analytical importance of elements and combinations of elements. In an attempt to further the understanding of the linkages in an input-output system, the objective of this paper is to examine the cluster structure of sales and purchases profiles when the principle of ‘excluded middle’ is violated by the use of fuzzy sets. The identification of fuzzy clusters is based on an analysis to derive an alternative decomposition of␣interindustry flows (Dridi and Hewings 2002), it is a complementary approach to the so-called ‘Matrioshka principal’ (Sonis and Hewings 1990) and will be presented first; the approach looks into the hierarchy of the statistical dependence between supply and demand in input-output systems. The analysis on the cluster structure and the interindustry flows is conducted using the data analysis technique known as dual scaling (Nishisato 1980, 1994). Results of the analysis will be illustrated using input-output tables of the US and Canada.