Neighborhood clusters and citywide technological diversification

The clustering of economic activities is widely considered to be a key driver of economic competitiveness. However, little is known how technological activities are clustered within cities and how such neighborhood concentration is associated with citywide technological diversification, as the existing literature tends to treat cities as homogenous units, neglecting their neighborhood dynamics and heterogeneity. To fill this gap, we develop a novel measurement framework to identify neighborhood-level technology clusters within 260 Chinese cities, using 16 million geo-coded patents from 2003 to 2018. We then link this neighborhood-level clustering measure with the rate, direction, and complexity of citywide diversification in patenting activities. We measure technological diversification as the entry of a new revealed comparative advantage in a city's patenting portfolio and find that neighborhood-level clusters (1) contribute to an increasing rate of diversification at the city level; (2) channel diversification toward technologies related to existing specializations; (3) are associated with more complex diversification. Our results suggest that zooming in on the micro-dynamics of tech clusters provides a more nuanced view of their advantages than is typically assumed.