Universal Database for Economic Complexity

TL;DR

This paper introduces a comprehensive 'Universal Database' for economic analysis by integrating trade data through machine learning, enabling better assessment of countries' economic development and industry interactions.

Contribution

It presents a novel integrated database combining trade of services and goods, and applies machine learning to enhance economic complexity analysis.

Findings

Emergence of development pathways and industry clusters in the network.

Improved assessment of country competitiveness using the new database.

Identification of high-tech industry interactions.

Abstract

We present an integrated database suitable for the investigations of the Economic development of countries by using the Economic Fitness and Complexity framework. Firstly, we implement machine learning techniques to reconstruct the database of Trade of Services and we integrate it with the database of the Trade of the physical Goods, generating a complete view of the International Trade and denoted the Universal database. Using this data, we derive a statistically significant network of interaction of the Economic activities, where preferred paths of development and clusters of High-Tech industries naturally emerge. Finally, we compute the Economic Fitness, an algorithmic assessment of the competitiveness of countries, removing the unexpected misbehaviour of Economies under-represented by the sole consideration of the Trade of the physical Goods.