The Journal of Productivity Analysis, 1: 43-78 (1989) o 1989 Kluwer Academic Publishers, Boston. Manufactured in the Netherlands Economic Performance and “Frontier” Efficiency: A Product Life-Cycle Approach* JULIEN VAN DEN BROECK Department of Economics, University of Antwerp (RUCA), Rijksuniversitair Centrum Antwerpen, Mddelheimlaan 1, B-2020 Antwerpen, Belgium FERNAND BROECKX Department of Statistics, University of Antwerp (RUCA), Mddelheimlaan I, B-2020 Anhverpen. Belgium LEONARD KAUFMAN Department of Statistics, Free University of Brussels, Pleinlaan I, B-1050 Brussels, Be&urn Abstract We propose to introduce some economic performance variables of the firms and relate these with the corresponding “frontier” efficiency measures of 13 industries of the Belgian manufacturing sector in 1978. In order to find groups of similar firms by applying a cluster technique, we assume that this relationship is of a product life cycle nature. In the light of the product life cycle scheme itself and variations of it we can consider the possible ex- planatory variables of technical inefficiency in each group of firms as sources which intensify or slow- down the economic performance of the firm under consideration. Some cautious characterization of the top and bottom level clusters is formulated and linked to the efftciency issue. 1. Introduction As we mentioned in a previous paper [van den Broeck 19881 we think it is high time to emphasize the explanation of technical efficiency differences among firms with the aim of increasing total productivity by reducing the inefficient use of pro- duction factors. However, bearing in mind that the real explanation of technical inefficiency probably depends in part on the possibility of quantifying non- economic variables such as motivation, complexity, hierarchy of authority, cen- tralization, formalization and so on, which is only possible by surveying the firms under consideration very carefully, [cf. Armandi 19811, we think it makes sense to broaden our scope concerning efficiency matters first before analyzing the com- plex technical inefficiency problem. The complexity of analyzing technical efficiency makes it necessary to proceed cautiously and consider all quantifiable factors before considering non-quanti- fiable factors. Therefore, we suggest relating the efficiency measures to other *This paper was processed by N. R. Adam