Does Promoting High Tech Products Spur Development?

Abstract

In the past decade, contributors to the endogenous growth literature have identified a variety of ways that trade policy might affect long run growth. Among these, Lucas (1993) has argued that the mechanisms emphasized by Krugman (1987), Stokey (1988 and 1991) and Young (1991) provide an especially appealing characterization of developing countries: growth is accomplished by concentrating resources in those goods whose production processes induce learning and knowledge spillovers. Hence trade policy, by influencing the mix of production, can affect long run growth rates.

Despite its appeal, the Lucas/Krugman/Stokey/Young (hereafter LKSY) view remains largely untested. To distinguish it convincingly from other theories that relate trade to growth requires information on product-specific market shares and their evolution, as well as the technological sophistication and productivity growth rates associated with each product. Comprehensive product-level data of this kind are rarely available, and they are certainly missing in the relatively aggregated data sets that the empirical growth literature has focused upon. Nonetheless, by exploiting plant-level panel data, it may be possible to get much closer to testing the LSKY view than the existing empirical growth literature has done. If particular products may be associated with particular plants, and if technological sophistication may be associated with the plant-specific engineer- and technician-intensity of production, these data should provide a reasonable basis for inference. This paper begins from the premise that they do.

After reviewing the theoretical models of interest (section II), we devote considerable time to rendering the concept of a "learning industry" empirically meaningful (section III). We then use plant- and industry-level data from Colombia and Morocco to characterize rates of movement up the continuum of products from low-end (little learning potential) to high-end goods (section IV). Finally, we look for evidence that relatively rapid productivity gains accompany relatively rapid movement up the goods continuum.

JEL Classification: O31, O32, O38