Rethinking Job Displacement in the Age of Automation: A Conceptual Model of Workforce Transition and Reskilling

Abstract

The debate about automation and jobs has been centred on the question of how many jobs are susceptible to being replaced by machines, and has relied on the task-based estimates of job exposure to machines that range from a factor of ten times between studies. In this paper, the question is challenged as being under-specified. Exposure is an attribute of tasks, whereas displacement is a consequence for workers and both are linked by a transition process that is not modelled by exposure metrics. The fates of workers in similarly exposed jobs vary widely based on where nearby jobs are located, whether their employer redeploys or lets them go, and the institutional and regional infrastructure that surrounds them—workers reskilling is considered a post hoc, undifferentiated supply-side solution. This paper presents the Adaptive Transition Capacity (ATC) model, a multi-level model which reframes displacement as a product of the task exposure on the one hand and the capacity to transition on the other, both at the individual, organisational and institutional-spatial level, in a bounded latency window between onset of exposure and separation. The study is guided by an approach of theory-synthesis that combines the literature on task-based automation, human capital and the relevance of skills, displacement economics and evaluation of active labour market policies. The model results in a typology redeployment/reconversion/stranded proximity/displacement cascade and seven testable propositions, as well as an operationalisation protocol that specifies constructs, indicative measures, and placeholder estimands. The contribution is theoretical - no estimates of the contribution are reported. Employer implications, labour market institutions and design and timing of reskilling interventions are discussed.