Modular structure in labour networks reveals skill basins

Labour networks, where industries are connected based on worker transitions, have been previously deployed to study the evolution of industrial structure ('related diversification') across cities and regions. Beyond estimating skill-overlap between industry pairs, such networks characterise the structure of inter-industry labour mobility and knowledge diffusion in an economy. Here we investigate the structure of the network of inter-industry worker flows in the Irish economy, seeking to identify groups of industries exhibiting high internal mobility and skill-overlap. We argue that these industry clusters represent skill basins in which skilled labour circulate and diffuse knowledge, and delineate the size of the skilled labour force available to an industry. Deploying a multi-scale community detection algorithm, we uncover a hierarchical modular structure composed of clusters of industries at different scales. At one end of the scale, we observe a macro division of the economy into services and manufacturing. At the other end of the scale, we detect a fine-grained partition of industries into tightly knit groupings. In particular, we find workers from finance, computing, and the public sector rarely transition into the extended economy. Hence, these industries form isolated clusters which are disconnected from the broader economy, posing a range of risks to both workers and firms. Finally, we develop a methodology based on industry growth patterns to reveal the optimal scale at which labour pooling operates in terms of skill-sharing and skill-seeking within industry clusters.