Sustainable performance in people-centric-operations: Empirical Essays

Performance measurement plays an important role in managerial decision-making and has long been a central topic in economics, operations research, and management science. Traditional approaches generally focus on the relationship between inputs and outputs and on how efficiently organizations transform resources into productive outcomes. In people-centric operations, however, such measures may provide an incomplete picture of performance because they do not necessarily capture the conditions under which output is produced or the demands placed on the workers involved in the production process. This PhD therefore examines how analytical tools can be developed to identify potentially unsustainable work configurations in people-centric operations, assess their consequences, and provide adequate operational decision support. Methodologically, the dissertation builds on nonparametric production analysis, econometric methods, and operational data analytics. It first examines how atypical observations can be identified and interpreted in nonparametric performance analysis. It then studies how workload shocks propagate through a workforce and affect subsequent sickness absence, and develops multidimensional measures of employee taskload that account for task composition and worker-job characteristics. Finally, it proposes a probabilistic nonparametric framework for identifying episodes of potential mental overload from operational log data. The empirical applications draw on detailed administrative and operational data from home nursing and railway traffic control. Together, the studies show how performance analysis can be extended beyond observed output and resource use to incorporate the work configurations under which performance is generated and their implications for worker availability and operational reliability.