Abstract:  Population aging is pushing governments to extend working lives, yet workers in demanding occupations often leave their jobs well before any retirement age binds. This paper examines whether reducing workload can keep them in their jobs longer. I study a Dutch program that subsidizes reduced teaching loads for older secondary-school teachers, exploiting a 2014 reform that restricted eligibility by date of birth in a regression discontinuity design. Using administrative records on the universe of Dutch secondary-school teachers, I find that eligibility raises take-up but has no significant effect on retention. This null conceals opposing responses by contract size. Full-time teachers are 6.1 percentage points more likely to continue teaching, and the teaching they provide by staying offsets the reduction in their teaching load. Part-time teachers, who already work reduced schedules, take up the subsidized reduction without staying longer, resulting in a net loss of teaching hours. Eligibility also appears to improve physical health, with linked health-insurance claims showing a reduction in pain-medication prescriptions. Workload reduction can thus ease strain but retains only workers for whom reducing hours is otherwise costly.

Abstract: Higher education is strongly associated with better health, yet the extent to which this relationship reflects a causal effect rather than differences in pre-determined characteristics remains unclear. Using population-wide Dutch administrative data, we estimate the effect of higher education on chronic conditions and healthcare expenditures. We employ a fully nonparametric partial-identification framework that exploits monotone instrumental variables derived from parental wealth and parental health. We find that a substantial part of the observed health gap between graduates and non-graduates is attributable to selection. Nevertheless, higher education causally reduces chronic morbidity, particularly lifestyle-related diseases, and lowers healthcare expenditures among graduates.