Most studies of economic growth concentrate on aggregate data. However, careful examination of intersectoral growth patterns shows a large divergence among sectors. These are driven as much by intersectoral linkages as they are by the separate characteristics of each sector. Arguing, for example, that a certain sector, e.g., machine tools, has a higher growth potential than, e.g., agriculture certainly is right but does not sufficiently take into account that all sectors are interlaced to each other, but rather to a highly diverging extent that then accounts for a diverging dynamic.

We investigate those interdependencies of sectoral dynamics on an empirical basis by using a time series of German input–output (10) tables (58 sectors) of more than a decade (1978 to 1990) and derive the growth pattern of the economy in terms of its sectoral growth. Techniques designed to illuminate the impact of one sector's development on others are applied. Thus we proceed in establishing the analytical tools for this purpose in Sections 1 and 2, showing and analyzing the empirical, i.e., factual, observations in Section 3 and deriving a theoretical hypothesis compatible with those findings in Section 4.

The opportunity to analyze those qualitative differences in sectoral interdependencies – with the option also to recognize dynamic differentials – was opened up by the so-called qualitative input–output analysis (QIOA), which was developed by Czayka (1972), Schnabl and Holub (1979), Holub et al. (1985), as well as Holub and Tappeiner, following those just mentioned. In France, some other applications of graph-theoretical methods for structuring production have been brought forward (see Bellet et al. 1989, Torre 1993).