Return to work (RTW) in later life is often studied through the lens of health limitations, institutional incentives, or individual motivation in isolation. Yet re-entry into the labour market at older ages is inherently multidimensional, shaped by the interplay of personal, psychosocial, and institutional factors. Advanced multivariate methods, particularly Structural Equation Modelling (SEM), can capture these complex interdependencies, including latent constructs, mediation, and moderation.

This talk first introduces the core logic of Structural Equation Modelling (SEM), emphasizing its advantages for ageing research, including the modelling of latent constructs, mediation, and moderation. It then presents a bibliometric mapping on modelling RTW in later life using SEM based on a Web of Science dataset. The analysis highlights the research directions in the field, identifies dominant thematic clusters, and demonstrates a clear methodological gap: despite the complexity of RTW processes, the use of SEM remains highly fragmented.

Building on these findings, the talk proposes a SEM model for RTW in late working life. The talk concludes by outlining empirical implementation strategies using survey and register data and identifies opportunities for verifying the proposed model.