A comment on: "Increasing the Retirement Age: Policy Effects and Underlying Mechanisms"

Rabat'e et al. (2024) investigate the employment effects of a reform to the statutory retirement age (SRA) in the Netherlands. They use a regression discontinuity design and administrative data from the entire Dutch population (CBS Microdata). They find that the employment rate increases by 21 percentage points following an SRA increase. A three-month increase in the SRA increases the average retirement age by 0.61 months. The labor market state prior to the SRA-reform is highly predictive of the treatment effect. The authors argue that employers' effects are likely driving the employment impacts. We replicate the original results and perform various robustness tests. We find that their results are largely replicable and robust against using an alternative estimation strategy (difference-in-difference). However, the code for the mechanical model used to predict the treatment effect is not available in the replication package and therefore cannot be replicated. Additionally, we find that the mechanism claim on employers' effects is not robust and should be interpreted with caution.