Working Papers

This paper proposes the use of synthetic training data generated by large language models
to improve machine learning SDG classifiers. It shows that supplementing existing training data with
synthetic data produced by the ChatGPT tool improves the performance of the SDGClassy classifier.
This addition of synthetic data is especially useful in building SDG classifiers given the limited availability
of properly labeled data and the complex, interconnected nature of the SDGs. Synthetic data thus enables
more effective machine-learning applications in this context.

This paper investigates the role of trade costs in exporter dynamics in Africa. In comparison to exporters from other regions, African exporting firms are fewer, smaller and relatively less diversified in terms of products and destinations. African countries also display the highest rates of entry, exit and turnover of exporting firms, exporting products and export destinations. This suggests that Africa’s exporting activity is volatile and subject to a lot of experimentation, with exporters having difficulties in maintaining trade relationships. The analysis also confirms that trade costs are a crucial factor in explaining exporter performance in Africa vis-à-vis other regions, but also…

Productive and technological capabilities matter. The more conventional strands of the literature have emphasized them as major engines of export, growth and development. But how they matter is less clear, and many open questions remain on how capabilities influence export dynamics at microeconomic level. This paper empirically investigates their role on export dynamics in 40 developing countries between 2002 and 2012. In doing so, the paper exploits a country-sector-year database containing exporter-level statistical information. The empirical analysis shows that, within sectors, countries with higher productive capacities have more exporters, and the exporters are larger and charge…

This paper discusses Uzbekistan’s recent experience with structural shifts and industrial policy and the larger implications for existing theories of industrial policy. The paper has a particular focus on various industry policy instruments.

Two major hypotheses are discussed: (1) the hypothesis of Haussmann, Hwang and Rodrik (the more technologically sophisticated the export structure, the better for growth) and (2) the hypothesis of Justin Yifu Lin (export specialization should build on existing comparative advantages and not jump over the necessary technological stages).