This paper develops a general theory relating technology change and skill demand, capable of rationalizing the labor impacts of various technology changes since the 19thcentury. Performers (human or machine) face stochastic issues that must be solved in a given time to complete tasks. Firms choose how production tasks are divided into steps, the rate at which steps need to be completed, and the type of performer assigned to a step. Performers differ in the breadth of issues they can solve (generality) and in their tolerance for working at higher rates. Hu-man performers tend to be generalists with low rate-tolerance. Machine performers tend to be specialists less sensitive to rate. Central to the theory are the cost of fragmenting tasks into smaller steps, the cost of allocating performers to multiple steps, and the negative relationship between step complexity and the rate of completing that step. We derive the cost-minimizing division of tasks and level of automation of production and the demand for workers of different skills that those conditions create. Our theory predicts that the division of tasks under increased complexity is skill polarizing; automation is skill polarizing at lower production volumes and upskilling at higher volumes; and that parts consolidation increases the demand for mid-level skills. We find counterparts to the theory across a range of industrial contexts and time periods, including the Hand-Machine Labor Study covering mechanization and process improvement at the end of the 19thcentury, automotive body assembly, and emerging techno-logical changes in optoelectronic semiconductors used for communications.
Read full abstract