Abstract
Export diversification is central to economic development. However, most resource-rich countries have failed to diversify. In understanding the determinants of diversification different strands of literature emerge. One view highlights the role of macroeconomic and trade-related factors linked to the Dutch disease, such as the real exchange rate, type of commodity, and international commodity prices (Agosin et al., 2012; Lederman & Maloney, 2007). Another perspective focuses on path dependence, primarily examining product relatedness measures. This perspective suggests that a nation's current productive capabilities shape its future production possibilities. The latter offers different advantages, such as analysing diversification at the product level instead of export concentration measures, which may be subject to several biases. However, this framework pays little attention to the determinants that shape a country’s productive capabilities, enabling product relatedness. This pa
per introduces an alternative measure of product relatedness, adapting the approach proposed by Nomaler and Verspagen (2022) to encompass a broader set of unobservable characteristics. Our regression framework also integrates macroeconomic factors and relevant controls (i.e., international prices, exchange rate, energy and mineral dependency, GDP per capita) to explain diversification at the product level. We do this in a cross-country setting covering more than 5,000 products between 1995 and 2019; furthermore, we distinguish between different types of products to understand how variables affect diversification in non-extractive sectors vis-à-vis extractive sectors. Results demonstrate that our product relatedness measure is a robust predictor of diversification, especially in extractive sectors, which exhibit greater path dependence. However, macroeconomic factors, such as international prices, level of development, and commodity dependence, play a decisive role in explaining differences in diversification patterns, and excluding them may overestimate the predictive power of product relatedness.
per introduces an alternative measure of product relatedness, adapting the approach proposed by Nomaler and Verspagen (2022) to encompass a broader set of unobservable characteristics. Our regression framework also integrates macroeconomic factors and relevant controls (i.e., international prices, exchange rate, energy and mineral dependency, GDP per capita) to explain diversification at the product level. We do this in a cross-country setting covering more than 5,000 products between 1995 and 2019; furthermore, we distinguish between different types of products to understand how variables affect diversification in non-extractive sectors vis-à-vis extractive sectors. Results demonstrate that our product relatedness measure is a robust predictor of diversification, especially in extractive sectors, which exhibit greater path dependence. However, macroeconomic factors, such as international prices, level of development, and commodity dependence, play a decisive role in explaining differences in diversification patterns, and excluding them may overestimate the predictive power of product relatedness.
Original language | English |
---|---|
Publisher | UNU-MERIT |
Publication status | Published - 19 Sept 2023 |
Publication series
Series | UNU-MERIT Working Papers |
---|---|
Number | 030 |
ISSN | 1871-9872 |
JEL classifications
- o11 - Macroeconomic Analyses of Economic Development
- f14 - Empirical Studies of Trade
- l78 - Industry Studies: Primary Products and Construction: Government Policy
- o13 - "Economic Development: Agriculture; Natural Resources; Energy; Environment; Other Primary Products"
Keywords
- Extractive Sectors
- Export Diversification
- Product Relatedness
- Macroeconomic Factors