Convergence in Global High Technology? A Decomposition and Specialisation Analysis for Advanced Countries.

This paper investigates whether patterns of technological specialisation of advanced countries change over a period of time in which globalisation in technology occurs. The theoretical analysis of this question is based on models of new growth and trade theory-especially a model by Grossman/Helpman (1991)-as well as on clues from evolutionary economics. The empirical analysis is divided into three parts. First, based on the patent applications in 42 R&D- intensive product groups of ten advanced countries, structural decomposition analysis is used to investigate the effects that determine growth in inventive activity. Secondly, the concepts of: beta- and sigma-despecialisation (or specialisation), which are very similar to the concepts of beta- and sigma-convergence (or divergence), are applied to the same data to investigate whether the countries are converging in technology Thirdly, for a longer but less disaggregated time series, the impact of the adaptation of technical knowledge on foreign product trade is examined by a multivariate autoregressive model (transfer function model). Taking all results together, evidence points to rather stable specialisations whereby the structural effects are not predominant. However, results indicating beta-despecialisation also show that the majority of national innovation systems are now moving towards an average specialisation. Yet, country-specific economic structures seem to remain more important as global changes in technology when their appropriation for international trade is concerned.