Patterns of world trade: an application of network complexity analysis
Abstract
This thesis is a contribution to network complexity analysis,
which has shown a high degree of explanatory power in visualising
the process of economic diversification, both across countries
and over time. The thesis comprises a stage-setting introduction,
a primer on the network complexity analysis and measures of
economic complexity, three core chapters and a concluding chapter
that summarises the key findings and makes suggestions for
further research.
The first core chapter (Chapter 3) aims to place network
complexity analysis, which has been criticised as a purely data
driven analytical tool, within a theoretical economic context. By
building a bridge between the mainstream theories of comparative
advantage and the new analytical tool, the chapter demonstrates
that the predictions of the theory of comparative advantage are
borne out in the network analysis.
The next two chapters are novel empirical applications of network
complexity analysis, using a structural construct called the
product space. Chapter 4 deals with the impact of trade
liberalisation on export performance, covering 96 liberalisation
episodes in 129 countries over 1962–2012. The first stage of
the analysis involves constructing measures of product emergence.
These are then used in econometric analysis to examine the
average within-country impact of trade liberalisation on product
emergence and diffusion. The findings indicate that trade
liberalisation is associated with an increase in the rate of
product emergence and a moderate reorientation toward products
that are more dissimilar to those in the pre-reform export
composition.
Chapter 5 explores the implications of global production sharing
on network complexity analysis. The approach taken is based on a
systematic separation of production-sharing based trade (PSB
trade) from the standard (reported) trade data. PSB trade is
further disaggregated into parts and components and final
assembly. It is found that failure to distinguish between PSB
trade and standard horizontal trade tends to upwardly bias the
complexity rankings of exports of some developing countries. This
is because some developing countries are engaged in assembling
and testing some parts and components while others engage in
final assembly. It is also found that parts and components are
clustered in a densely connected core of the product space
alongside relevant final products and have a higher level of
average within-group proximity relative to final products,
indicating higher average co-export potential. In the case of the
machinery and transport equipment sector, parts and component
exports tend to emerge prior to the export of final products,
suggesting that global production sharing could act as a
mechanism in the diffusion process from the periphery to the core
of the product space.
Description
Citation
Collections
Source
Type
Book Title
Entity type
Access Statement
License Rights
Restricted until
Downloads
File
Description