Answers are grounded in this paper and cite their sources.
In this paper, we disentangle the total local average treatment effect of retirement on mental health among elderly women in Europe into the indirect effect mediated through grandparental childcare and the direct effect through all other mechanisms. Applying a semiparametric estimation strategy, our estimates point to a positive direct effect of retirement on mental health and a negative indirect effect through grandparental childcare that partially offsets these gains. The results highlight the usefulness of causal mediation studies for policy analysis and illustrate the importance of allowing for full effect heterogeneity in settings with treatment-mediator relationships.
No takes yet. Share an insight, caveat, or question.
Siflinger et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: