Trade indices and trade patterns

The formation of trade patterns along the development process has recently been the object of considerable scrutiny. How export diversification evolves, empirically, along the path of economic development was, up to now, a relatively little-explored question. Opening up new export markets is an entrepreneurial gamble which, if successful, is quickly imitated. The inability of export entrepreneurs to keep the benefits of their activity private leads to a classic public-good problem. Klinger and Lederman (2004, 2005) show that poor institutions appear empirically to compound the problem, lending support to the Hausmann-Rodrik (2003) view.

This project analyses indices of trade specialisation and investigates the determinants underlying the development of selected trading nations’ specialisation patterns over time. It contributes to the existing literature by adopting new approaches to the analysis of countries’ transitions across patterns of specialisation – or the lack thereof – combined with refined methods of classification of specialisation patterns across industries and products at the highest level possible of disaggregation. With particular emphasis on trade policy, it then seeks to identify the determinants underlying transitions in specialisation patterns on the basis of detailed country studies and across countries and time.

The empirical methods applied are manifold. Inter alia, they include Markov transition matrices for the analysis of changing specialisation patterns over time; system-GMM estimators, suitable for the analysis of large panels with few periods; conditional nonlinear quantile regression techniques to examine the scale and shape of the entire response distribution and to deal with excessive heterogeneity in the industry data.