Economic Complexity and the Emergence of New Ideas

Abstract

Recent research has shown a strong link between the economic complexity of a country and important economic indicators such as GDP per capita and lower inequality. However, the mechanisms underpinning this relationship remain poorly understood. While it is often argued that complex ecosystems provide a fertile ground for novel combinations of ideas leading to the emergence of impactful new technologies and industries, this process is hard to measure with aggregate data structured around industrial codes offering a lagging view of the economy. We seek to overcome this challenge by combining official data about industrial activity in UK Local Authority Districts (LADs) with a novel dataset containing the content and meta-data from almost a million UK business websites. We measure economic complexity using ECI and a Fitness based measure, which respectively capture a location’s specialization in unique, knowledge intensive sectors, and a weighted measure of economic diversity. We use a complex networks approach to topic modelling to detect ‘topics’ in the websites of businesses and analyse their relation to economic complexity and industrial sectors before going on to measure emergence by looking for novel words in the websites of businesses in different locations and sectors. We show that LADs with high ECI scores have a stronger share of companies active in emerging technologies, even after we control for their industrial composition. Further, and contrary to our initial expectations, emergent companies are more industrially diversified in locations with high ECI scores than locations with high Fitness scores. One potential reason is that the innovative sectors that high ECI locations specialize on are lead adopters of emerging technologies that are subsequently diffused into other parts of the local economy.

Keywords: Economic Complexity, Emergent Technologies, Web Mining

JEL Classification: O30, R11