Kène Henkens

Assessing one’s own life expectancy is becoming increasingly important as pension reforms worldwide shift more pension risks to individuals. A relevant question in retirement planning is therefore: do individuals have an accurate idea of how long they will live? And do they take future life expectancy improvements into account? To examine this question, we conducted in June 2024 a survey among a representative cross-section of the Dutch population and matched the individual respondents’ assessments in terms of (1) the probability of reaching the age of 90 years or older as well as (2) their expected age of death. These subjective expectations were compared with the age- and education-specific projections from expert demographers from Statistics Netherlands. In general, men and women of working age tend to underestimate their chances of living up to the high age of 90. A noteworthy result is that in particular higher-educated women substantially underestimate their chances of reaching 90 years of age. The findings imply that both men and women tend to underestimate their life expectancy and ­tacitly do not take account of the progress in life expectancy that expert demographers expect based on scientific research. The underestimation of life expectancy is of considerable importance in societies where specific groups on the labor market (e.g., self-employed) have to make their own decisions in matters of pension savings.

Policies aimed at extending working lives pose challenges for older workers with high job demands, caregiving duties, and health problems. To address these challenges, policymakers have proposed facilitating work hour reductions in the years before retirement. However, it is unclear whether older employees in demanding situations are relatively more likely to (plan to) reduce their work hours before reaching the state pension age, and whether work hour reduction arrangements in their company enable them to do so. Using the NIDI Pension Panel, this study examines preretirement work hour reduction plans and behaviours of 1704 Dutch employees aged 60–63 at baseline. Ordinal and binary multilevel logistic regression models were used to test the hypotheses. One in four employees in our sample plans to reduce work hours before reaching the state pension age, and one third eventually reduces them. When a work hour reduction arrangement is available, older employees are more likely to reduce their work hours. Older workers who experience high job pressure are more likely to reduce their work hours, particularly when a work hour reduction arrangement is available. Chronic health conditions and caregiving burden were not associated with planned or actual working hour reductions. Overall, the availability of work hour reduction arrangements is key in enabling work hour reductions, facilitating them for some—but not all—employees in demanding situations. We discuss how such arrangements contribute to sustainable ageing, considering the tensions between financial and social sustainability.

This collection is focused on understanding what we know and what we need to know on the topic of work, ageing and retirement. Work, Employment and Society has published 44 articles on this topic over 35 years, from 1989 to 2024. We analyse this article corpus and identify four themes that shed insight into what sociologists of work and ageing know: the macrostructural perspective, the organisational perspective, individual life courses, and retirement as process. Overall, we find a historical pattern whereby studies have shifted from a top-down structural approach utilising mostly quantitative survey data to an individualised and agentic approach utilising mostly qualitative data. Recent scholarship has begun to blend both structure and agency, introducing theoretical and methodological sophistication to derive more holistic understanding of ageing at work. We select 11 articles using various perspectives, theories, and methods that illustrate these trends. We conclude with suggested key directions for future research.

Objectives: Maintaining control over the retirement transition is crucial for post-retirement well-being. Yet, studies show that between 10% and 40% of employees experience involuntary retirement. This study investigates three separate reasons for involuntary retirement: poor health, organizational pressures, and reaching the quasi-mandatory pension age. Guided by a life-course framework of agency and structure, we analyze how collective and individual perceptions of development, accommodation, and post-retirement work climates shape separate reasons for involuntary retirement, as opposed to voluntary retirement. Methods: Using multi-level panel data from the Netherlands, we estimate multi-level logistic regression models on 3,735 respondents nested in 501 organizations. We assess how organizational climates, which vary systematically between organizations, and individual-relative perceptions, which vary systematically within organizations, are associated with different involuntary retirement reasons. Results: Reaching the quasi-mandatory retirement age was the main reason for involuntary retirement. Across organizations, a supportive organizational climate for development reduced involuntary retirement. At the individual level, employees who perceived more supportive accommodation and post-retirement work climates relative to their colleagues were less likely to experience involuntary retirement. Organizational and individual-relative climates revealed distinct associations with different reasons for involuntary retirement. Discussion: Our results suggest that involuntary retirement is a multi-dimensional phenomenon shaped by micro-, meso-, and macro-level constraints. Supportive organizational policies and practices enhance employees’ agency, with some differences depending on the reason for involuntary retirement. Policies encouraging training of older workers may increase both labor force participation and older employees’ choices as they near retirement.

Limited social integration can have adverse effects on both migrants and host communities. For the case of international retirement migrants, previous evidence suggests that migrants struggle to integrate into the local community. Using a novel survey of Dutch retirement migrants based on a probability sample (DRM 2021), we describe and explain Dutch retirement migrants’ social integration with locals, Dutch migrants, and non-Dutch migrants in thirty-five destination countries. Findings show that, on average, Dutch retirement migrants have the most contact with locals, followed by Dutch migrants and the least contact with non-Dutch migrants. According to our multilevel analyses, an important country-level predictor of decreased local contact was linguistic dissimilarity between origin and destination languages. Contact with locals versus migrants was also affected by the relative size of the migrant group in the destination. Significant individual-level sociocultural predictors were previous connections to the destination, which was associated with increased local contact, and stronger national identification, which was associated with having more friendships with Dutch migrants. Overall, the results reveal the effect of preferences and opportunity structures on retirement migrants’ social integration.

Loneliness is relatively common among older adults and variation in loneliness can partially be explained by the occurrence of various life events. One such event is retirement. However, research on the impact of retirement on loneliness has yielded conflicting results. We address this by studying the impact of retirement on within-person change on two dimensions of loneliness: social and emotional. We argue that retirement might have opposing implications for these two dimensions of loneliness, which can thus serve as an explanation for the contradictory results reported in the literature. Furthermore, we include partner status and partner’s retirement status as factors that may determine the size and direction of the effect of retirement. We propose five hypotheses based on detailed theoretical reasoning, which we investigate using 3-wave large scale (n = 5,238) panel data from the Netherlands, collected in 2015, 2018, and 2023. The results show that retirement only affects emotional loneliness, not social loneliness. Among people with a partner, emotional loneliness increases following retirement, while it decreases among unpartnered retirees. We interpret this finding as retirees being quite capable of substituting the socially integrative function of work, and that retirement can provoke conflict and expose pre-existing issues within romantic relationships.

Insecurities and risks related to ageing in a foreign country could fuel return migration among international retirement migrants. The few studies examining retirement migrants' return have been small in scale and focused mainly on return intentions rather than return behaviour. In this paper, we examine the prevalence and predictors of return migration among retirement migrants and the discrepancy between return intentions and return behaviour. We collected survey data on a representative sample of 5,065 Dutch retirement migrants in 40 destinations and combined them with administrative data on return migration. Three years after data collection, almost 9 per cent had returned to the Netherlands, whereas less than 5 per cent had intended to return during this period. Our findings show how age-related changes and transnational ties to the country of origin increase the likelihood of return. Our results also suggest that retirement migrants may underestimate the long-term implications and social embeddedness of the return migration decision.

Objectives: Most studies on retirement and social network dynamics focus on the closer social network, leaving the role of more peripheral contacts largely overlooked. This article studies how retirement affects contact frequency with a wider range of social ties. We formulate and test differential hypotheses for each category of ties, and additionally examine gender and partner status differences. Methods: We analyze 3 waves of panel data of the NIDI Pension Panel Study, collected in the Netherlands between 2015 and 2023 (n = 5,238). We use 2-way fixed-effects regression models to study within-person change in contact frequency with the different categories of social ties after retirement. To test the differential hypotheses, we conduct Wald tests comparing coefficients across models. Results: The results indicate that retirement stimulates contact primarily with neighbors and friends, to a lesser extent with siblings and (grand)children, and does not affect contact frequency with parents. Contact with ex-coworkers initially increases but then decreases over time. For women, the positive association between retirement and contact with ex-coworkers, friends, and children is stronger. Not having a partner reduces the association between retirement and contact with friends and ex-coworkers. Discussion: Our findings suggest that retirees attempt to replace lost workplace interactions by engaging with their former colleagues outside of work and increasing contact with neighbors. Simultaneously, the results suggest substantial continuity in contact with all ties. We suggest that larger changes might take place outside of the ties studied here, with new contacts, and provide several suggestions for future research.

The number of retired migrants has vastly increased in Western societies and is projected to grow further in the decades ahead. Many of those migrants are more likely than nonmigrants to encounter severe financial hardships during their retirement years, raising the following question: What explains this “migrant pension penalty?” Based on a systematic literature review of 86 articles, this article presents a comprehensive model for understanding the migrant pension penalty. This model is inspired by the life-course framework and shows the most important explanations clustered into four “life course constraints and opportunities.” First, there are institutional constraints and opportunities such as pension policies. Second, the economic constraints and opportunities such as incomes throughout the life-trajectory play a role. Third, sociocultural constraints and opportunities such as language skills and cultural norms for saving are an important explanation for the migrant pension penalty. The connections with others such as families and neighbors—known as linked lives—constitute the fourth life-course constraint and opportunity. Based on this comprehensive model, we summarize the scientific knowledge on the migrant pension penalty, identify knowledge gaps, formulate a research agenda, and provide policy implications for reducing the migrant pension penalty.

Objective: This paper examines variations in how international retirement migrants maintain intergenerational relationships with adult children residing in the country of origin. Background: Despite a greater need for support from family members in the later life course, many retirement migrants move away from their adult children. Prior studies have described retirement migrants’ family ties, but often using non-representative samples and without examining factors that could explain variations. Method: We collected data on a stratified representative random sample of Dutch nationals aged 66-90 who migrated after age fifty to one of 40 different destination countries (N = 2,849). Ordered logistic regression models are performed on three dimensions of retirement migrants’ relationship with adult children. Results: Retirement migrants are emotionally close to their adult children despite little face-to-face contact and not particularly frequent digital contact. Female and married retirement migrants have more contact and closeness with adult children than males and divorced or widowed people. Fewer opportunities for contact, such as a greater geographical distance and a lower socioeconomic status, mainly predict lower contact frequencies but not variation in emotional closeness. Evidence for the need for contact is mixed and the role of cultural norms is limited. Conclusion: The maintenance of family ties over longer distances challenges the view that frequency of contact is a prerequisite for emotional closeness and shows how parents adapt to a new family context.