Rising inequalities in work autonomy by algorithmic control? An empirical analysis of employees in German large-scale work organizations

Algorithmic technologies designed to control work activities enable management tasks that were previously performed by supervisors, for example, to be increasingly automated. To what extent algorithmic control simply replaces the previously issued work instructions or objectives (e.g., by supervisors or management) or whether it fundamentally changes employees' perception of work autonomy are matters of debate. The aim of this paper is to investigate these matters, specifically whether and to what extent algorithmic control fundamentally changes existing inequality structures where employees of higher occupational status already perceive more work autonomy. Results based on linked employer-employee data from 5,138 employees at 160 large German work organizations showed that only those employees who received daily algorithmic instructions reported having less perceived work autonomy compared with employees who did not receive such instructions. This finding held true for all occupational status groups. However, the employee groups with higher occupational status were less likely to receive daily automatically generated work instructions which corresponds to the inequality generating mechanism of opportunity hoarding. In this comparison of work organizations, the more often algorithmic instructions in a workplace are applied, the greater the inequality in perceived work autonomy among different occupational status