Trade diversification patterns help explain the level of utilization of trade opportunities by countries, mainly the least developed. Empirical analyses show an inverse U relationship between trade diversification and level of development. Trade diversification measures used do not take into account differences in complexity of exports, and complexity indices only consider products with comparative advantages. This study seeks to cover both gaps by analyzing the differences in the determinants of trade diversification, considering the complexity of products exported by 19 Western Hemisphere countries from 1962 to 2017. The results show that after controlling for economic complexity, the inverted U relationship disappears. Development of financial markets positively affects the complexity of trade diversification in the long term, while the terms of trade that have a negative effect on trade diversification does not affect the complexity-corrected indices. In the short term, transaction costs and trade openness appear to have a significant effect.
TopIntroduction
International trade offers countries important opportunities to expand markets and boost economies of scale, with the degree of trade diversification being one of the main determinants of this process (de Piñeres & Ferrantino, 1997; Haddad et al., 2013; Lederman & Maloney, 2003). However, most developing countries have not been able to insert themselves adequately into this dynamic. Taking advantage of these opportunities is for many the great task in public policy design (de Ferranti et al., 2002).
There exists consensus in the literature on the economic effects of diversification but not on its determinants (Jetter & Hassan 2015). They are usually associated with better use of the benefits of trade: on the one hand, diversification achieves greater macroeconomic stability by diversifying risk in an international price instability context (Haddad et al., 2013; Hamid, 2010); on the other hand, it generates positive externalities such as the learning process (learning by exporting) and knowledge spillovers (Hausmann & Klinger, 2007; Hausmann & Rodrik, 2002; Vettas, 2000). This in turn allows other exporters to access existing or potential markets (Vettas, 2000) with less uncertainty and therefore lower costs. This results in greater economic growth (Acemoglu & Zilibotti, 1997; de Ferranti et al., 2002; Hidalgo et al., 2007; Lederman & Maloney, 2003). In addition, the very dynamics of trade induce higher levels of productivity, by boosting both activities of the export sector and related sectors (Melitz, 2003). The extent and depth of these effects depend on diversification patterns, particularly if countries are specialized in certain products that use standardized inputs1 that can serve many other products (transversality), since it reduces the costs of introducing new products (information externalities) (Hausmann & Rodrik, 2002). However, contributions in this area are still scarce.
The study of trade diversification patterns, has been addressed from different approaches at regional and global levels without a clear consensus. Two lines of analysis can be distinguished. The first one investigates the link between economic development (measured through GDP per capita) and diversification. De Benedictis et al. (2009) find a positive relationship between both variables. On the other hand, Imbs and Wacziarg (2003), Klinger and Lederman (2004) and Cadot et al. (2011) show the presence of an inverted “U”, where countries as they develop increase their trade diversification towards a greater number of sectors, but after a certain (high) level of development, they start specializing again (Imbs & Wacziarg, 2003). This would confirm that trade specialization is a phenomenon of both less developed and more developed countries.