The Employment and Poverty Paradox: A Causal Puzzle and a Family Affair

While micro-level research shows that employed individuals are less likely to experience poverty, macro-level trends reveal disappointing poverty reduction despite substantial employment growth across European countries – constituting an employment-poverty paradox. This study addresses this paradox by examining the heterogeneous effects of employment on poverty across different family types, while rigorously accounting for selection processes and confounding through leveraging panel data, varying assumption about causal structure, simulations, and formal sensitivity analysis. We apply a multi-group Kitagawa-Blinder-Oaxaca decomposition to examine how employment prevalences and premiums contribute to poverty differences across varying family-types. Our results suggest that differences in employment prevalences contribute minimally to poverty gaps, while premium effects drive most observable effects. Part-time employment shows even weaker poverty reduction effects across all family types. Sensitivity analysis reveals vulnerability to unmeasured confounding, suggesting true causal effects may be smaller still. In sum, employment is not a panacea for poverty reduction across family types.