Retirement behavior of the Swedish notch babies: evidence from the job episodes in the survey of health, ageing and retirement in Europe

Haodong Qi, Lund University

The 1994 Swedish pension reform introduced cohort differentials in benefit accounting. Those born in 1938, the "Swedish Notch Babies", were the first recipients whose benefits were partially computed by the Notional Defined Contribution scheme, while older cohorts remained unaffected. This paper examines the aftermath of the reform by analyzing the differences in retirement behavior between the 1937 cohort and the Notch Babies. Both static and dynamic programming retirement models are implemented using Hierarchical Bayesian Estimation. Retirement propensity is measured by the required rate of replacement (R*). It reflects the level of pension entitlements relative to labor earnings necessary in order for an individual to retire. Large R* implies low retirement propensity, and vice versa. The empirical results are based on the working life history in the Swedish Survey of Health, Ageing and Retirement in Europe (SHARE). The estimated hyper-parameter (R*) in both static and dynamic programming models are nearly identical, 0.76 and 0.73, respectively. This is mainly because the estimated discounting factor in the dynamic model is small, ß=0.31. At the individual-level, R* differs considerably across cohorts. For the 1937 cohort, tertiary education has large and significant effect on R*, while gender and health have no impact, ceteris paribus. However, among the Notch Babies, the positive effect of higher education on R* is reversed, while R* is much higher for men than women, ceteris paribus. Such cohort differences are identical in both static and dynamic models. The implication of the analysis is three-folded. First, future utility flows have little impact on the retirement decision for both the unaffected and notch cohorts. Secondly, the effect of the reform at the population level is negligible. Finally, the reform increases the retirement propensity for those with completed tertiary education, while simultaneously prompting male workers to prolong working life.

  See paper

Presented in Session 10: Pensions and retirement