Mapping industrial patterns and structural change in exports
Charlotte Guillard
#2020-005
This paper proposes a new methodology for identifying patterns in the
organisation of industries and their evolution over time, based on the
temporal network structure of the product space. To do this, I apply a
community detection algorithm on 5-year snapshots of the product space
from 1975 to 2000. This exercise enables us to identify different
clusters of related products and to follow their evolution over time. I
find that the product space is highly modular, that is it contains well
delimited clusters of products. The community structure and its
evolution show that the factors explaining industrial patterns and
structural change are more complex than the traditional divide between
low, medium and high-tech industries. Several common drivers can be
identified to explain the emergence and evolution of different
communities including the experience in a technological domain, factor
abundance, scale economies as well as global value chains and vertical
integration. Moreover, I find that technological domains and boundaries
between industries are not always clear-cut and can evolve over time.
Keywords: Structural change, Capabilities, Economic Complexity,
Networks, Community Structure, Exports
JEL Classification: O11, O14, O33, O25, P40, E14