Analysis of the impact of artificial intelligence and robotics on labor productivity and the structure of employment in the digital economy

Abstract

The diffusion of machine learning systems and industrial robotics is changing the ratio between labor inputs and output faster than a consensus on the direction of this shift is being formed. The divergence of estimates remains fundamental: some studies record the displacement of routine labor and the contraction of employment, while others point to productivity growth with a stable headcount and a redistribution of functions within the enterprise. The aim of the work is to assess which components generate the gain in labor productivity when intelligent and robotic solutions are introduced, and how the qualification structure of employment is reshaped in the process. The calculation is performed on an author's model of a manufacturing enterprise for 2021–2026: the dynamics of output per worker are constructed, a factor decomposition of output growth is carried out using the chain substitution method, and a scenario modelling of the personnel structure by qualification groups is performed with a horizon up to 2030. The macroeconomic background (robot density, employment dynamics, the spread of intelligent technologies) is tied to data from Rosstat, the ISSEK HSE and the International Federation of Robotics. On the model data, the cumulative growth of output per worker over the period amounted to about 53,4% with a reduction in the average headcount of 12,0%; the factor decomposition attributes 721,1 million rubles of output growth to rising productivity and only –163,4 million rubles to the change in headcount. The share of routine operational personnel in the model structure falls from 38.4 to 21,5%, while the group of data and intelligent-system support specialists expands from 3,2 to 14,2%. The decomposition obtained shows that the observed effect of adoption is determined primarily by the intensive factor – the growth of output per worker rather than the mere release of jobs; consequently, the impact of intelligent technologies on employment is realized through a shift in the qualification profile rather than through a unidirectional contraction of labor demand.