RESEARCH-AND-DEVELOPMENT, HUMAN-CAPITAL AND TRADE PERFORMANCE IN TECHNOLOGY-INTENSIVE MANUFACTURES - A CROSS-COUNTRY ANALYSIS

This paper comprises an extensive examination of the relationship betw een commonly utilized policy indicators of national technological capa bility and observed trade success in technology-intensive manufactures , across 52 countries, in the 1980s. A central focus is the relative p erformance of policy indicators which directly measure innovative acti vity (such as patent output and expenditure and employed scientists an d engineers in R&D) versus those measuring general efforts at increasi ng national levels of human capital. The study design is guided by the inter-country, intrasectoral approach outlined by Dosi et al. [19] - it emphasizes differences in trade performance based on the notion of international competitiveness (but also compares results for conventio nal revealed comparative advantage indices). The large number of count ries included, and the nature of the policy objectives adopted, have n ecessitated a highly aggregated approach using nation-wide attributes and trade in one broad technology-intensive sector as the measure of d emonstrated international competitiveness. Simple correlation and mult iple regression techniques are used to assess the association of four main sets of national attributes (indicators of direct innovative acti vity, general human capital development, physical capital formation an d natural resource endowment) with technology-intensive trade performa nce.The empirical findings provide further support for the importance of national differences in technological capability as a determinant o f structural, and international, competitiveness within the technology -intensive sector of production. They suggest that policy indicators m easuring actual innovative activity and physical capital formation hav e a much closer link to technology-intensive trade than the general ed ucational effort or output indicators. ''Demand-side'' measures - whic h proxy the actual employment and application of resources for R&D pur poses - demonstrate the strength of association required for reliable policy evaluation of real national differences in technological capabi lity. The poor performance of the general human capital supply variabl es highlights the need for a careful consideration of the optimality o f the allocation of substantial levels of scarce national resources to broad-based educational output. A number of other significant results concerning the link between technology indicators, technology-intensi ve trade and economic performance are discussed.