S3.1 Gender and social inequalities in the late career: Empirical evidence and critical perspectives
Chair: Loretta Platts
Discussant: Martin Hyde
European policy discourses on lengthening life expectancies and the sustainability of pensions systems have focused on incentivizing citizens to prolong their working lives. However, older people are a highly heterogeneous group in terms of their health and financial situations and have highly varied possibilities for participation in the labour market. These presentations draw together findings from Germany, Russia, Sweden and the United Kingdom to explore social and health inequalities in later life related to life course working patterns, retirement processes and pensions systems. During this symposium, a range of new empirical findings will be presented, accompanied by critical reflections from the discussant Martin Hyde to place them into the wider policy and political picture. The presenters will explore a variety of questions: How might lifelong working patterns contribute to the gender pension gap? Are returns to employment after retirement really so rare? How do gender and social position affect older people’s situations in retirement? What are the implications for health and healthcare use of extended working lives and retirement? Might prolonged working lives generate new sites of social and health inequalities? What might be the broader implications of an expansion of part-time, temporary jobs that merely supplement pension income?
S3.1.1 The role of mid-career income for gender pension gaps in Sweden
Stefanie König, Boo Johansson, Marie Kivi
Gothenburg University, Sweden
Background: This study investigates the importance of mid-career earnings for the gender gap in pensions and potential scarring effects of low income earlier in life. Methods: Swedish income register data from 1990 to 2015 was linked to the ‘Health, Ageing and Retirement Transitions in Sweden” survey from 2015. We investigated the gender pension gap of 960 retirees and worries about pension income of 2518 older workers between the age of 60 and 66 years. Blinder-Oaxaca decompositions were applied to analyse the gender gap in pensions and linear regressions were used for the analysis of financial worries. Results: Results show that gender differences in mid-career income play a stronger role for the gender gap in pensions than late career earnings. Mid-career earnings are furthermore related to higher subjective worries about pension income and explain gender differences for this variable. Results are related to the context of the current Swedish pension system and the role of the preceding pension reforms is discussed. Conclusions: It can be concluded that different aspects of the reformed pension system potentially contributed to higher gender gaps in pensions.
S3.1.2 Predictors of returns to paid work following retirement: A prospective analysis of
“unretirement” in Germany, Russia and the United Kingdom
Loretta G. Platts1, Karen Glaser2
1 Stress Research Institute, Stockholm University, Sweden, 2 King’s College London, UK
Background: Individuals may return to paid work following retirement, a phenomenon described as “unretirement”. Previous work has focused on single countries, which precludes international comparison. Therefore this study harmonizes data from Germany, Russia and the United Kingdom (UK) to examine intercountry variations in the rate and predictors of unretirement. Methods: Data were drawn from the German Socio-Economic Panel Study (1991–2013), Russian Longitudinal Monitoring Survey (1994–2013) and, for the UK, British Household Panel Survey (1991–2008) and Understanding Society (2010–2014), harmonized ex post. Unretirement was examined using Cox regression in relation to demographic covariates, education, health, and financial adequacy. Results: Unretirement was common; cumulative proportions of 0.42 Russian, 0.25 British and 0.17 German retirees unretired. In each country, retirees with more education, higher income, and better health more often unretired, but not those concerned about their finances. Conclusions: While recently retired people may represent a pool of potential labour, there was little indication that those in financial need were unretiring.
S3.1.3 Prolongation of working life and physical functioning in old age in Sweden
Harpa Sif Eyjólfsdóttir1, Isabel Baumann2, Neda Agahi1, Johan Fritzell1, Carin Lennartsson1
1 Aging Research Center, Karolinska Institutet, Sweden, 2 Zürich University of Applied Sciences, Switzerland
Background: Previous research examining the impact of retirement on physical health have shown mixed results, which may be due to socioeconomic differences. Individuals with high socioeconomic status (SES) who extend their working lives may have better working conditions, while those with poor with low SES may extend their working life for economic reasons despite health issues, resulting in poor physical function in old-age. This study examines whether the association between retirement age and late life physical functioning differs by SES. Methods: Data are used from two interrelated nationally representative longitudinal studies covering a 46-year period: The Swedish Level of Living Survey (LNU) and the Swedish Panel Study of Living conditions of the Oldest Old (SWEOLD), linked to annual tax register data. Propensity score matching was used to further investigate the effects of working longer than age 65 on physical functioning in old age. Results: Preliminary results showed that for musculoskeletal pain, self-rated health and ADL limitations, no significant differences were observed between those who retired later than age 65 compared to those who retired earlier in any socioeconomic group. For mobility limitations, there were significant differences for unskilled manual workers only, for whom prolonging working life was associated with fewer mobility limitations. Conclusions: Prolonging working life was generally not associated with physical functioning in later life in both high SES and low SES groups.
S3.1.4 Consequences of retirement on healthcare use: Applying an inequality-sensitive
approach using Swedish register data
Martin Wetzel1, Stefanie König2, Susanne Kelfve3
1 University of Cologne, Germany, 2 Gothenburg University, Sweden, 3 Linköping University, Sweden
Background: Healthcare costs are a major part of welfare contributions in
Sweden. These costs tend to increase over the life course and may be affected
by life events such as retirement. Targeting potential high-risk groups at an
earlier stage may help to prevent increases. Therefore this study examines whether
retirement is associated with intensified healthcare use overall and whether
there are gender and educational differences in healthcare use trends. Methods:
Based on Swedish register data, the study examines all Swedes who were still in
paid work in 2009 and had retired in 2010. Their use of healthcare (specialist
visits and nights spent in hospital) was followed from three years prior to
five years following retirement using panel regression models. Results: Preliminary findings showed that
retirement affected healthcare use only marginally (specialist visits) or not
at all (hospital nights). While we found gender differences in the amount of
healthcare use, no differences in changes at retirement were found. Similarly,
no systematic differences by education level were found in developments of
healthcare use at retirement. Conclusions: We conclude that the healthcare
system in Sweden does not need to target recent retirees, whether male or
female and higher or lower education level. Other risk groups (i.e., disability
pension recipients) might however need special attention.