LAUSR

We investigate the consequences of retirement from work for the overall well-being of individuals aged 50 and above. Well-being is captured by two different concepts: life satisfaction and agency-freedom, i.e. the evaluation of a person’s ability to do the things s/he wants to do and be who s/he wants to be. We use three observation periods of the Survey of Health, Ageing and Retirement in Europe and include nine European countries. The sample counts 62,082 observations (38,344 individuals) of at least 50 years old. A fixed-effects estimation approach is used so that differences in (time-invariant) individual characteristics are taken into account. We control for changes in financial and health situations of the individual and the situations of the partner. When retiring, people do not immediately report (on average) a different level of life satisfaction, but after 2 years, life satisfaction decreases compared to the beginning of the retirement [identified as Atchley’s honeymoon effect (The sociology or retirement, Wiley, New York, 1976)]. If well-being is expressed as agency-freedom, well-being is immediately positively affected, and this effect does not change after 2 years of retirement. This paper also investigates several forms of heterogeneities in the transition from work to retirement. We consider partial, early and joint retirement, part-time employment and self-employment, and job quality. We find that there is no difference in overall well-being between being partially and fully retired, between being retired before or after the normal retirement age or between those who retire simultaneously with their partner and those who do not. However, for some older workers, such as those employed with a low-quality job, retirement can be a relief from their current employment status. In summary, a policy for longer working careers is, on average, not detrimental to well-being, but some specific groups need special treatment.