Many existing jobs are prone to automation, but since new technologies also create new jobs it is crucial to understand job transitions. Based on empirical data we construct an occupational mobility network where nodes are occupations and edges represent the likelihood of job transitions. To study the effects of automation we develop a labour market model. At the macro level our model reproduces the Beveridge curve. At the micro level we analyze occupation-specific unemployment in response to an automation-related reallocation of labour demand. The network structure plays an important role: workers in occupations with a similar automation level often face different outcomes, both in the short term and in the long term, due to the fact that some occupations offer little opportunity for transition. Our work underscores the importance of directing retraining schemes towards workers in occupations with limited transition possibilities.
del Rio-Chanona, R.M., Mealy, P., Beguerisse-Diaz, M., Lafond, F. & Farmer, J.D. (2019). 'Automation and occupational mobility: A data-driven network model'. Working Paper. arXiv:1906.04086