Stated Preference Analysis of Full and Partial Retirement in the United States

Gradual Retirement, Financial Incentives, and Labour Supply of Older Workers: Evidence from a Stated Preference Analysis

“…The results of our ranking analysis are thus not in line with our expectation that the introduction of gradual retirement opportunities would lead to substantially different retirement patterns ( Reday-Mulvey and Delsen 1996; Reday-Mulvey 20 0 0 ). However, our results are consistent with other studies which showed that many workers are more likely to choose full retirement over partial retirement (Van Soest et al 2006;Kantarci and Van Soest 2013 ). 19 , 20…”

“…Because the stated preferences experiment presents a hypothetical situation in which workers work part-time, we checked to what extent workers with part-time and full-time jobs respond differently to the vignettes. TableA4in Appendix A shows that our main results are robust to the part-time status of the respondents in our estimation sample.16 See alsoVan Soest et al (2007) andKantarci and Van Soest (2013) . 17 This result is robust to different definitions of early and late retirement.…”

“…A stated preferences approach solves this problem by capturing a broader array of preference-driven behaviours, and can therefore be used to estimate the causal relation between gradual retirement and a worker's age of retirement and labour supply. Concerns may arise about the external validity of retirement options in stated preference studies, especially if these retirement options do not yet exist in practice ( Kantarci and Van Soest, 2013 ). 21 However, the literature shows that stated preference estimates and estimates based on data on actual behaviour are usually quite close ( Louviere et al,20 0 0 ).…”

“…Second, this paper builds on two related studies by Van Soest et al (2007) and Kantarci and Van Soest (2013) . These studies, also using stated preferences experiments, show that many workers can be enticed to work part-time if they are given fair financial compensation, but also that many workers are more likely to choose full-time retirement over partial retirement.…”

“…4 A further advantage of our study is that the base scheme in our stated preference experiment is similar to the actual pension scheme of public sector workers in the Netherlands. While Van Soest et al (2007) and Kantarci and Van Soest (2013) used highly stylized regimes that are not related to actual pension schemes in the Netherlands and do not allow for different durations of gradual retirement in their scenarios, our vignette regimes provide many more choice options, allowing for more realistic estimates of the effects of gradual retirement opportunities on the labour supply. Finally, we go beyond previous studies by showing that introduction of gradual retirement schemes in the Netherlands may have heterogeneous effects on the retirement age and labour supply of workers with health problems and physically demanding jobs.…”