Economics research seminar: Michelle Rendall, Monash University

Presenter

Michelle Rendall, Monash University

Topic

Robots and Technological Unemployment (with Fenicia Cossu and Alessio Moro)

abstract

The speed of automation adoption is faster than in the past, and many experts believe robots will quickly take on tasks previously performed by humans, creating technological unemployment. However, the extent of technological unemployment at the economy-wide level depends crucially on two variables: i) the average time needed for the representative worker to acquire her skill set, which we document is continuously increasing over time; and ii) the impact robots have on the labor market depends on the share of tasks they can perform in each occupation. We provide empirical evidence suggesting that for most occupations, this share is less than 10%, and only in a few cases does it exceed 20%. To quantify the extent of technological unemployment due to robots, we present a multiple-occupation model in which each occupation is defined by a set of tasks, and technological change occurs as an increase in tasks that can be performed by robots. Transitioning across occupations is costly and uncertain, as agents have to go through unemployment to do so, and the time to match to a different occupation is random. We calibrate the model and use it as a laboratory to study the transition of workers across occupations after a technology shock, and the resulting technological unemployment.