Demographic change, secular stagnation and inequality: automation as a blessing?

We construct and calibrate an overlapping generations model incorporating demographic change and the possibility to automate the production process to test the hypothesis put forward by Acemoglu and Restrepo (2017). In line with their hypothesis, we find that ageing is a powerful force stimulating the adoption of automation technologies in OECD economies. Ageing-induced automation is found to soften the negative effects of labour scarcity and rising old-age dependency rates on per capita growth, but the compensation is incomplete. One important reason is that automated tasks are far from perfect substitutes for tasks executed by human labour. A second reason is that ageing-induced automation reduces the intensity of positive behavioural reactions to ageing in the form of retiring later and investing more in human capital. Moreover, the partial compensation comes at the price of rising wage and welfare inequality between individuals of different innate ability level and a fall in the net labour share of income. Compared to existing literature, we pay special attention to the theoretical and empirical foundations of the modelling of automation. Theoretically, our work is the first one testing this hypothesis that relates the approach to automation rigorously to the state-of the-art conception by Acemoglu and Restrepo (2018a; 2018b). Empirically, we tested and largely confirmed the validity of our approach and calibration by comparing model predictions of (changes in) automation density to actual data on robotization in a cross-country fashion.