Algorithmic Management and Wage Dynamics: an Individual-level Survey Evidence

Using a novel European-wide individual-level survey on “Impacts of Artificial Intelligence and Algorithmic Management in the Workplace”, in this paper we study the relationship between the use of algorithmic management practices and working conditions. By drawing on the richness of the available data, we are able to examine more closely the relationship between exposure to different intensities of algorithmic management and the wage structure, while accounting for workers' socio-demographic characteristics, occupation, sector, and country of employment. We use a Mincerian framework as our benchmark and employ a quantile specification to better account for the distribution of the wage structure. To remove confounding factors, we apply a propensity score matching procedure. Intermediate levels of algorithmic management (AM) are positively correlated with wages in Eastern Europe, whereas in Western countries positive coefficients are observed at low levels of algorithmic management, while at higher intensities of AM the relationship between the two variables turns negative. When examining this dynamic along the employment structure (white vis-à-vis blue-collar workers) and across manufacturing and service sectors, we grasp more subtle nuances in the nature of this relationship. Thus, we detect that blue-collar workers exposed to higher levels of algorithmic management intensity experience a wage premium, whereas white-collar workers face a wage penalty associated with greater degrees of algorithmic management. This pattern is observed in both Eastern and Western Europe.