When the Algorithm Sets the Pace: Digital Management, Worker Voice and Job Quality

Abstract Software that allocates tasks, sets the pace of work, monitors behaviour and rates performance has moved from delivery platforms into ordinary workplaces, yet comparative evidence on how widely it is used, who is shielded from it and what it does to job quality remains thin. This study reanalyses the published results of the European Agency for Safety and Health at Work (EU-OSHA) OSH Pulse survey of about 27,250 workers in 29 countries (spring 2022) and sets them against the 2024 establishment survey (ESENER) and the 2025 OSH Pulse wave. A country-by-sector panel of the organisational model index shows that 63 percent of its variation lies between countries and 12 percent between sectors; sector rankings agree across countries (Kendall’s W = 0.28), with commerce, transport and food, ICT and finance, and administrative services highest. Countries with statutory agreement or co-determination rights over new technology do not show lower adoption (difference -0.11 index points; exact permutation p = 0.44). Published regressions link each additional digital management practice to 0.21 more psychosocial risks and 0.17 more health complaints; worker consultation is associated with reductions equal to 0.51 (95% interval 0.32 to 0.70) and 0.78 (0.50 to 1.06) of one practice. Raising consultation to all workers would offset 15 percent of the average psychosocial burden and 22 percent of the health burden, and a package of effective measures 53 and 74 percent. The findings are associations from self-reports, and the panel tests describe variation among cell means rather than sampling error. Purpose - This study measures the diffusion of digital worker management across European countries and sectors, asks whether worker voice institutions are associated with less adoption, and quantifies how far worker consultation offsets the job quality costs of digital management. Design/methodology/approach - The paper reanalyses published OSH Pulse 2022 results: a two-way variance decomposition and Tukey non-additivity test on a 29-country by 8-sector panel, Kendall’s W, exact permutation and bootstrap comparisons of voice regimes, and delta-method and Monte Carlo accounting built on published regression coefficients. Findings - Country differences dominate sector differences in adoption; sector ranks are stable; voice regimes are not associated with lower adoption; and consultation is associated with sizeable but partial offsets of psychosocial and health burdens. Research limitations/implications - The analysis uses published aggregates and coefficients, cross-sectional self-reports and a coarse regime classification, and cannot establish causal effects or test whether consultation moderates the effect of digital management. Practical implications - Employers and regulators should treat consultation on digital systems as one component of a package that includes information, awareness and transparency, and should extend algorithmic management safeguards beyond platform work. Social implications - Because most workers already experience some form of digital management, protection that applies only to platform workers will miss most of those affected. Originality/value - The paper brings comparative panel evidence, institutional comparison and transparent accounting together, and shows how the current regulatory and survey evidence on digital management fits.