Project: ‘Active Ageing Index (AAI)’


UNECE Grant No: ECE/GC/2012/003






Developing a Conceptual Framework


Part 1

Active aging from a life course perspective:

a review of academic literature


European Centre Vienna

April 2012



Paper prepared for a discussion during the 1st Meeting of the

Expert Group on Conceptualizing and Measuring Active Ageing


May 10 and 11, 2012

Hotel Bloom Brussels




Active aging from a life course perspective:


a review of academic literature



Pieter Vanhuysse

April 2012



This review paper aims to survey recent social scientific advances in the academic literatures, mainly in the fields of economics, sociology, social psychology and public policy, that theoretically inform the debate about active ageing in advanced market democracies today (for a complementary recent survey of evolving conceptual discourses in the international policymaking community, see Walker 2010). The paper is structured as follows. The first section briefly surveys various conceptual approaches to active aging and their limitations, with a special emphasis on the successful aging, productive aging, and WHO approaches. The second section discusses at greater length the rationale for adopting a life course perspective to active aging, and surveys insights from the economics and public policy literatures on human capital formation across the life cycle. The third section reviews recent evidence on the incidence of productive participation in society through volunteering and similar forms of social participation, and on its beneficial health and cognitive effects especially for older individuals. The fourth section surveys recent advances in the literatures in psychology and economics that indicate a set of reasons, based on the subjective well-being and cognitive effects of retirement, for aiming to activate older workers particularly, both by reducing early retirement and by prolonging working lives, and to reactivate already retired individuals who are still willing and able to work. The last section briefly reviews evidence on un-retirement – the re-entry into paid employment by already retired workers.


1. Different concepts of active aging

The evolving policy-oriented literature has typically presented alternative multidimensional views of active aging that emphasize the importance of maintaining and fostering the physical, mental and (sometimes) social well-being of individuals as they age (for longer discussion of the changing discourse among the international community, see e.g. especially Walker 2010; also WHO 2002; Hutchinson et al. 2006; Van Dijk and Turner 2010; Zaidi and Zolyomi 2011; Eurostat 2012). In a recent review of different concepts in this literature to date, Buys and Miller (2012: 104) distinguish between the related concepts of active, healthy, productive, and successful ageing. However, the latter three concepts as described in the literature are presented as related to active aging, and even alternative conceptions of active aging, rather than as full alternatives to it. Active ageing is defined by the WHO (2002) as ‘the process of optimizing opportunities for physical, social and mental well-being throughout the life course, in order to extend healthy life expectancy, productivity and quality of life in older age’. Healthy aging on the other hand can be defined as the ‘ability to continue to function mentally, physically, socially and economically as the body slows down its processes’. Productive aging is in turn defined as ‘any activity by an older individual that produces goods or services, or develops the capacity to produce them, whether they are to be paid for or not’. Lastly, successful aging is defined as a ‘low probability of disease and disease-related disability; high cognitive and physical functioning and active engagement with life.’

These different approaches tend to have in common certain highlights such as fostering aspects of physical, emotional, social and economic well-being, but at the same time they differ in their definition of both the processes and the priorities of the ageing experience (Buys and Miller 2012: 103). For instance, one dominant approach, the successful aging approach most identified with Rowe and Kagan (1998), defines successful aging as ‘adaptive and agentic, wherein individuals are able to “call the shots” in order to attain optimal outcomes’ (Hendricks and Hatch 2006: 308). The 10-year MacArthur Study argued that aging is neither predetermined nor inflexible, and that individuals are capable of exercising personal preferences to modify their environments and lifestyles (Rowe and Kagan 1998).

That is, successful aging is a result of lifestyle choices of individuals that avoid risk factors and lead to better outcomes. The successful aging approach claims rather strongly that such lifestyle choices are more important in determining the experience of older age than hereditary factors or even certain diseases. Not surprisingly therefore, the successful aging approach has been criticized for applying the term ‘successful’ to aging and for focusing on the impact of individual choices on the quality of the ageing experience, thus downplaying the importance of broader societal influences and resources (Buys and Miller 2012: 104). Two such lifestyle choices tend to be singled out as especially crucial in this approach: what people eat (diets) and how people live, that is, (active) exercise, (active) social interaction, (active) mental stimulation, and constructive (or positive) mental outlooks. For instance, maintaining an optimistic outlook may add up to ten years to life (Danner et al. 2001; Hendricks and Hatch 2006).

In a classic study from the US nuns project that fits the successful aging approach, Danner et al (2001) investigated handwritten autobiographies from 180 Catholic nuns, composed much earlier when these nuns had been on average 22 years old. Danner et al (2001) scored these essays for emotional content and relate them to survival during ages 75 to 95. They found a statistically significant and strong inverse association between positive emotional content in these writings and risk of mortality in late life. As the quartile ranking of positive emotion in early life increased, there was a stepwise decrease in risk of mortality, resulting in a 250 percent difference between the lowest and highest quartiles. In other words, Danner et al (2001) conclude that positive emotional content in early-life autobiographies was strongly associated with longevity as much as six decades later.

The productive ageing approach, in turn, focuses less on health and medical issues and instead highlights the fact that the common idea that productivity stops once a person becomes older, is misguided. Productive aging have become an important focus on the agenda of both policy makers and social scientists across the aging world (e.g. Erlinghagen and Hank 2006). Cross-national comparisons of productive activities in middle and later adulthood suggest considerable diversity between countries with regard to older citizens’ engagement in both paid employment and various domains of unpaid work (e.g. Hank 2011). The productive aging approach proposes a primary focus on the economic contributions of older people in terms of paid employment and volunteering. A frequent criticism of this approach is that in addition to excluding health as a dimension, it also implies almost by definition that many older people, namely those who are not or less productive, are therefore of lesser value to society. This productivity concept implicitly or explicitly is the active aging concept underlying most labor market participation and activation oriented studies by economists and labor market policy experts in the literature (see e.g. Remery et al 2003; OECD 2006; Eurofound 2007; Skirbekk 2008; Saint-Paul 2009; Dorn and Souza-Poza 2010; von Below and Toursie 2010; for a double focus on paid employment and unpaid activities, see Zaidi and Zolyomi 2011).

This productive focus can also explain why, as Walker (2010) argues, the concrete policy instruments thus far developed to translated active aging into public policies have been overwhelmingly in the labor market policy realm. Activation policies have acquired a central status within the labor market policy mixes of aging OECD member states in recent years, often based on an implicitly ‚productivist’ concept of active aging, albeit one that has increasingly become life course-oriented, rather than older workers-oriented, in recent years. For instance, Europe 2020, the 10-year strategy proposed by the European Commission in 2010, has as its main aim the revival of the European economy at a time of crisis by means of smart, sustainable, inclusive growth. One of Europe 2020’s five 'headline targets' to boost growth and employment was raising the employment rate of the working-age population to at least 75%. Similarly, among the seven 'flagship initiatives' of the Europe 2020 Strategy, the agenda for new skills and jobs set out to modernize labor markets by facilitating labor mobility and the development of skills throughout the life cycle with a view to increasing labor participation.

Such a productivist approach is in many ways both appropriate and needed. When looking at past trends since the 1960s until today, the direction has generally been strongly upward for most OECD countries as regards life expectancies at 65, yet strongly downwards as regards labor force participation for older workers from ages 55 or 60 until 64 (e.g. OECD 2006; Wise 2010; Marin 2012). It should be noted that labor market mobility is particularly low among older workers everywhere in the EU. Most new hires among older workers tend to be workers who change jobs, rather than previously unemployed or inactive workers (OECD 2006). Inactivity, in particular, is a one-way street for most older workers: once older workers leave the labor market, they are very unlikely to get back into it subsequently. Combined with steadily increasing life expectancy both at birth and at retirement age, these are clearly untenable developments that need to be addressed by the extension of working lives, including the activation of older workers. As many experts have noted, working longer must be seen as a logical concomitant of living longer (Marin 2012, Wise 2010; Hering 2012).  

The World Health Organization, in its Active Ageing Framework paper (2002), proposes a conceptually useful and very comprehensive multidimensional approach to active aging that synthesizes and subsumes the previous approaches (see also Buys and Miller 2012; Walker 2010). It is based on three basic pillars for active aging: health, defined as encompassing ‘all aspects of physical, mental and social wellbeing, as expressed in the WHO definition of health’ (pp. 47-51), participation, (defined as providing ‘education and learning opportunities; recognize and enable participation in formal and informal work; encourage full participation in community life’ (pp. 51-52)) and security, defined as ensuring ‘the protection, safety and dignity of older people by addressing the social, financial and physical security rights and needs of people as they age’ (p. 52).

This comprehensive approach positions active ageing as the preeminent conceptual framework (or umbrella term) for investigating and understanding the impacts of multiple aspects of older adults’ activities on quality of life and their general well-being as they age. But importantly, this approach, like the others reviewed above is mainly conceptual, not empirical. As various reviews of active aging approaches note, this comprehensive WHO nevertheless provides little guidance and even less consensus regarding the specific quantifiable indicators to be used for measuring active aging (e.g. Buys and Miller 2012; Hutchinson et al. 2006; Walker 2010). Note, in addition, that the WHO approach mainly considers (chronologically) old people. That is, even though it pays lip service to the concept, it does not provide a life course perspective.


2. The need for a life course perspective: lifelong learning, but the earlier the better?

The advent of highly competitive, globalised knowledge economies has increased the need for public policies aimed at boosting human capital over the entire life course (Esping-Andersen 1999, 2002a, 2002b, 2009). The economics literature indicates that the knowledge economy has increased both the initial ‘entry ticket’ and the subsequent payoffs throughout the life course of active human capital such as marketable knowledge, non-cognitive traits and skills. That is, knowledge, skills and formal education are both increasingly required for finding good jobs, and increasingly rewarded once these jobs are found. On the requirements side, those with low levels of literacy record much lower levels of labor market participation and much higher risks of unemployment in most OECD economies. On the rewards side, for instance, in the USA the gap in earnings between individuals with an advanced degree and those with a high school degree has increased from around 1.8/1 in 1979 to around 2.6/1 in 1999 (Powell and Snellman, 2004: 213).

Moreover, early life accomplishments and family and educational environments often have significant impacts on individuals deep into adulthood. The reason is that younger children have longer time horizons over which to recoup the benefits of human capital increases over the life course - or, conversely, over which to be scarred by early life deficiencies in their educational, nutritioanl or family environment. Second, there are dynamic complementarities that further increase the importance of focusing on young life cycle stages: early learning makes later learning and cognitive functioning easier and more effective (Heckman 2000; 2007). Fogel (1999, 2003) has shown the empirical relationship between early-life nutrition and adult health. Similarly, regarding educational outcomes, Banks and Mazzona (2012) evaluate the causal effect of a real-world education policy reform (the historical raise in the minimum school leaving age in England from 14 to 15) on cognitive abilities at much older ages – after age 50. They find a large and significant effect of the education policy reform on memory and executive functioning at older ages as measured by cognitive tests. This result is particularly remarkable since the reform had a powerful and immediate effect on about half the population of 14-year-olds. Other studies similarly underscore the positive association between education and old age cognitive abilities. Thus Glymour et al. (2006) find evidence of a causal relationship between education and memory. Exploring geographical variation in US compulsory school reforms they found a positive effect of an additional year of education on memory test scores at older ages.

In the same vein, in one key study on family environments, Crosnoe and Elder (2004) use a longitudinal and decade-long dataset of talented children started in 1922 to study whether and how family experiences early in life relate to patterns of adjustment and functioning in later life. They identify altogether four different holistic life course profiles of aging (less adjusted, career-focused but socially disengaged, family focused, and well-rounded). Moreover, they find a clear connection of these aging profiles with experiences much earlier in the life course. That is, they statistically identify three specific family experiences in childhood and adolescence that can predict membership in holistic profiles of aging in the later years: the socio-economic status of the family, early parental divorce, and parent-child attachment. The authors argue that these distal associations from early to later life reflect both ‘mediational pathways’ (that is, adult experiences and current circumstances such as intact marriages, alcoholism, and educational attainments can explain the observed influence of early experiences) and supplemental pathways (the significance of early life experiences on aging profiles is direct, i.e. not filtered through adult and current experiences).

Clearly therefore, given the strong impact of early-life dimensions such as cognitive and non-cognitive educational and behavioral outcomes and life environments, there is a strong case for adopting a life course perspective on active aging. According to Walker (2010: 596), active ageing should be: ‘a comprehensive strategy to maximize participation and well-being as people age. It should operate simultaneously at the individual (lifestyle), organisational (management) and societal (policy) levels and at all stages of the life course’ (emphasis added). As Settersten (2006: 4) puts it, understanding the life course is about describing individual and collective experiences and statuses ‘over long stretches of time and explaining the short- and long-range causes and consequences of these patterns. … And just as the life-course perspective takes a dynamic view of individuals and groups, it takes a dynamic view of environments, probing when and how environments change and probing reciprocal connections between changing individuals and changing environments. For all of these reasons, there is a natural, mutual attraction between scholarship on the life course and scholarship on aging.’

Applied to work and retirement, this means we ought to pay special attention to shifts in the boundaries between the three boxes: education, work, leisure (early retirement but also un-retirement, longer education, increased longevity; lifelong learning due to fast technological change).[i] Yet, as Settersten points out, the three-box structure remains largely intact in policymakers’ minds, although a growing number of scientists are becoming more aware of boundary shifts. In one good example of an life course-sensitive policy analysis, Waldfogel (2006) offers a book-length empirical social science survey of what public policies children need, specifically distinguishing four different stages for policy recommendations: infancy, preschool, school age, adolescence.

A cumulative body of evidence from education economics shows that investment though public policy in early life is particularly effective (and cost-efficient) precisely because it has large and manifold impacts on the subsequent evolution of cognitive abilities and thereby contributes to more active individuals throughout the lifecycle (e.g. Heckman 2000; 2007; Carneiro and Heckman 2003). Even when viewed from a purely economic point of view, early life interventions constitute an efficient use of tax revenues – and a more efficient use of tax revenues than equivalent expenditures on citizens in later stages of life, such as labor market training or older worker activation (Heckman 2000). Compared to randomly assigned controls, participants in a number of early childhood education pilot programs across the USA score systematically better on a wide range of variables measuring educational achievements and high school graduation rates, and, later on in life, employment rates, monthly earnings, house ownership, and so on. Carneiro and Heckman (2003) conclude that such targeted high intensity early life intervention is superior to state interventions aimed at older age groups.             

Policies targeting the early life years can also be goal-effective, because they intervene at a time when most neural and cognitive progress and priming is made. Ability gaps across socioeconomic groups open up at early ages, both for cognitive skills such as educational outcomes and IQ, and for noncognitive skills such as perseverance, motivation, and self-control. Human capital deficits tend to arise mainly from inadequate learning environments in the family. When controlling for parental education and maternal ability, early ability gaps are almost eliminated (Heckman 2000; 2007; Carneiro and Heckman 2003).

This literature indicates that returns to late childhood investment in young adolescents, let alone to investment in adults later in life, are low compared with the returns to early childhood investment, even though investment in the later life course stages can be justified on numerous other grounds, such as equity, or poverty or social exclusion prevention. Equally key here is the role of non-cognitive abilities. Heckman et al. (2010) show that the Perry Preschool Program in the US, which enhanced the subsequent economic and social performance of participants, did little to boost participants’ adult IQ levels, but significantly boosted their non-cognitive abilities.

The argument that government intervention in learning, training and skill acquisition is higher in earlier than later lifecycle stages has also been bolstered at the other end of the career cycle. Kristensen (2012) uses a unique individual employer-employee longitudinal dataset on the entire 1936-1944 birth cohort in Denmark conditional on being active in the labor market in 2001, in which year the selected cohorts were aged 57-65. The dataset provides information on comprehensive government co-sponsored training records spanning three decades and was employed in order to study the effect of these older workers’ participation in formal life-long learning on the decision to retire early. The dataset includes both basic courses targeting individuals with low to medium levels of formal education and focusing on basic literacy and numerical skills as well as language classes, vocational and technical courses targeting all groups of workers and with a short duration of a few days or less than two weeks, and post-secondary courses including general and more specific training from college education and up to university level. Contrary to expectations that more lifelong training will lead to longer working lives, she finds that participation in formal lifelong learning has only a marginal impact in prolonging working life by delaying retirement. On average, one year of training only adds one month to career length. This indicates lifelong training is not enough to substantially prolong careers and activate the older working age population, although it does not necessarily mean that government co-sponsored training programs cannot increase elderly workers’ productivity.


3. Active aging as social participation: ties that bind and benefit

Active and inclusive participation in society all along the life course is not just a value and benefit in and of itself. It also has important health and cognitive side-effects both for older and for working-age citizens (on participation generally, see Gallie et al. 2003; for recent empirical evidence, see for instance NCVO 2011; Zaidi and Zolyomi 2011). Volunteering is important in later life because it is a key dimension of productive and healthy ageing (Warburton and Grassman 2009). Continued participation and involvement in society is recognized by the WHO (2002) as a key component of ageing well. The benefits of volunteering in later life are well documented, with a strong body of evidence linking such activities with positive health and social outcomes for seniors (Hinterlong, Morrow-Howell & Rozario, 2007).

Recently, a number of studies have also shed light on the social determinants of participation. Hank (2011) uses SHARE data to study the social determinants of cross-national variation across eleven European countries in three measures of non-market productive aging - volunteering, helping, and caring. He finds that some of the observed cross-national variation in older citizens’ unpaid productive engagement is due to differences in population composition according to age or education. Second, societal images of aging (as operationalised, however, by older worker employment rates) do not exhibit statistically significant associations with formal volunteering, informal helping, or adult caregiving. On the other hand, older citizens tend to engage significantly more in productive aging activities in those countries that are freer (as measured by the Freedomhouse Index of Civil Liberties) and that have larger welfare states (as measured by total social spending as a share of GDP. Hank argues that longstanding cultural and institutional setups of a society constitute reference frames which older individuals perceive as relevant for their decision whether to engage in volunteering, helping, or caring.

In the same vein, Broese van Groenou and van Tilburg (2012) use Dutch data from the Longitudinal Aging Study Amsterdam on almost three thousand ’young old’ citizens (defined as those between 55 and 69 years old) in order to study how individual and societal developments affect social engagement in later life. In an affort to distinguish the effects of age, aging and cohort, two different cohorts were studied in two different decades in two decades: those aged 55-69 in 1992 and in in 2002, with a six-year follow-up on both cohorts to allowed for cohort-sequential analyses. Broese van Groenou and van Tilburg find that, regardless of age, the 2002 cohort of young olds volunteered more often than the 1992 cohort. Moreover, over six years, volunteering increased for the 55- to 59-year-olds, stabilized among the 60- to 64-year-olds, and declined among the 65- to 69-year-olds, and these age-differential changes were observed in both cohorts. These effects remained significant after adjusting for gender, education, religious involvement, health, employment status, network size, and partner status. A higher education, religious involvement, staying in good health, and maintaining a large network increased the likelihood of volunteering.

As regards older citizens, friendship and social networks are increasingly known to have a positive effect on overall health and longevity later in life. For instance, Giles et al. (2005) study 1477 Australians aged 70 years or more living in residential care facilities and find that even after controlling for a range of demographic, health, and lifestyle variables, greater networks with friends protect significantly better against mortality in a ten-year follow-up period. Older citizens in the highest tertile of friends networks were 22 percent less likely to die during the study period than those in the bottom tertile. Similarly, Orth-Gomer et al. (1993) study a random sample of 736 middle-aged Swedish men for six years, and find that a lack of social support predicts all causes of mortality in population studies. They measure both emotional support from very close persons (’attachment’) and the support provided by the extended network (’social integration’). Both ’attachment’ and ’social integration’ were lower in men who contracted coronary heart disease, with a significant effect for social integration and an almost significant effect for attachment. When controlling for other risk factors in multiple logistic regression analyses, both factors remained as significant predictors of new coronary heart disease cases. Smoking and lack of social support were the two leading risk factors for coronary heart disease in these middle-aged men. Only smoking was as important a risk factor as lack of social support.

Interestingly, adopting a life course perspective on active aging would lead one to expect active labor market policies to also have a positive effect (often unintended) in enhancing the ‘active capabilities’ not just of older citizens, but also of working age citizens. In the wake of liberalized labor markets, growing international competition and increasingly punctuated working careers, working-age citizens in advanced economies, even traditional labor market insiders with highly protected jobs, increasingly demand more activation policy spending rather than standard employment protection, as they view it as an investment in their future re-employability chances (Anderson and Pontusson 2007; Tepe and Vanhuysse 2013). In fact, since the onset of economic crisis in 2008, citizens generally see labor market policies as more important for providing them with a sense of employment security in current crisis times than job protection policies such as employment protection legislation (Chung and van Oorschot 2011).

But beyond economic security, there are some indications that investing in active labor market policies, in addition to upskilling jobholders and re-activating unemployed workers workers (the main intended goals), can have positive side-effects on the social participation levels of working-age citizens. In a cross-national study of the impact of active labor market spending on citizens’ social networks and their perceptions of social exclusion in 17 European countries, Anderson (2009) shows that activation spending tends to significantly boost citizens’ social connectedness and to significantly reduce their sense of social exclusion. In countries that spend more on activation policies, citizens report more frequent social interactions both with family members and with friends and co-workers. Moreover, citizens in such high activation-spending countries tend to be significantly more active in voluntary organizations and to report more often that they participate in social activities. Anderson (2009) argues that this heightened level of social participation and social inclusion is boosted both by the overall investment in active labor market policies and by these social programs’ orientation to enhance people’s skills (but see Breidahl and Clement 2010 on Denmark). Moreover, these beneficial effects on social participation and social inclusions tend to be more pronounced among individuals who occupy a more precarious labor market positions – i.e. labor market outsiders. In other words, activation policies have the potential to affect both the employability and the social lives of those citizens who are most targeted by these policies.[ii]


4. Well-being and cognition after retirement: mental switch-off, or slow-down with higher flexibility?             

The literatures in psychology and economics indicate a strong set of rationales gruned in the wellbeing and cognitive effects of retirement, for aiming to activate older workers, both by reducing early retirement and by prolonging working lives, and to reactivate already retired individuals who are still willing and able to work. Early retirement is associated not just with lower income and consequently fewer resources to invest in health, but also with less cognitive and physical activity (e.g. Rohwedder and Willis 2010) and with changes in daily lifestyles that may lead to unhealthy behavior (e.g. Kuhn et al. 2010). Note also that at the individual level, healthy life expectancy is increasing faster than life expectancy, which implies that the average period an individual will need elderly long-term care over his lifetime is actually shrinking. Christensen et al. (2008) use a longitudinal survey of the entire Danish 1905 cohort from 1998 to 2005 to assess the loss of physical and cognitive independence in the age range of 92 to 100 years. Multiple functional outcomes were studied, including independence (defined as being able to perform basic activities of daily living without assistance) and having an MMSE (MiniMental State Examination) score of 23 or higher. In the aggregate, the 1905 cohort had only a modest decline in the proportion of independent individuals at the four assessments that took place between age 92 and 100 years: 39%, 36%, 32%, and 33%, with a difference between the first and the last assessment of 6%. For participants who survived until 2005 (meaning one hundred years old), however, the prevalence of independence was reduced by more than a factor of 2, from 70% in 1998 to 33% in 2005.[iii]

The effects of retirement on health and mortality are increasingly understood in recent years. For instance, Behncke (2012) investigates the effects of retirement on various health outcomes with data from the first three waves of the English Longitudinal Study of Ageing and finds that retirement significantly increases the risk of being diagnosed with a chronic condition. In particular, it raises the risk of a severe cardiovascular disease and cancer. This is also reflected in increased risk factors (e.g. BMI, cholesterol, blood pressure) and increased problems in physical activities. Furthermore, retirement worsens self-assessed health and an underlying health stock.

In the same vein, Kuhn et al. (2010) take advantage of a major change to the unemployment insurance system in Austria, which allowed older workers in eligible regions to retire up to 3.5 years earlier than comparable workers in non-eligible regions, in order to control for the solve the adverse health selection problem (that is, unhealthy workers retire because they are unhealthy) and to make a causal argument about the effect of retirement on mortality. They find that a reduction in the retirement age causes a significant increase in the risk of premature death (defined as death before age 67) for men, but not for women. The effect for men is not only statistically significant, but it is also quantitatively important. Kuhn et al. (2010) estimate that one additional year of early retirement causes an increase in the risk of premature death of 2.4 percentage points; this amounts to a relative increase of about 13.4%; or 1.8 months in terms of years of life lost. Kuhn et al. (2010) make a second contribution by providing strong indications that the reasons for these higher levels of premature death among retired men are linked to changes in men’s health-related behaviors mainly associated with smoking, drinking, an unhealthy diets and less physical exercise. Analyzing cause-of-death statistics, they find that excess mortality is concentrated on three specific causes of deaths (ischemic heart diseases, mostly heart attacks; diseases related to excessive alcohol consumption; and vehicle injuries) that jointly account for no less than 78% of the causal retirement effect (while accounting for only 24% of all deaths in the sample). In fact, Kuhn et al (2010) estimate that 32.4% of the causal retirement effect can be directly attributed to smoking and excessive alcohol consumption alone.

Beyond the extra leisure time and the lower stress levels but higher health and mortality risks often associated with being retired, there are manifold influences of labour market exit on subjective wellbeing levels. On the one hand, a key reason for the prominence of voluntary job loss among older workers may be the higher sense of subjective well-being (or quality of life) associated with having the choice to decide one’s own labour market status. While employed people generally record higher subjective well-being levels than those who are no longer active, among the latter group those persons who had some degree of agency and control over their labour market exit clearly record higher levels of well-being.[iv] By contrast, being forced out of the labor market and into early retirement tends to lead to lower levels of subjective well-being. Analyzing 1167 questionnaires of employed, unemployed and retired British men and women aged 50-74, and distinguishing between those voluntarily lacking employment (such as those who had voluntarily retired early) and those whose lack of employment was imposed (e.g. those forced to retire early), Warr et al. (2004) found that the ability to choose one’s current role is significantly associated with higher subjective well-being. In addition, in a study of over 15000 people in 11 countries, Waginger (2009) found that subjective health and quality of life are generally lower, and depression symptoms higher, for those who left the labor market early, even after controlling for those who left early because of health problems. Waginger also reports that forced early exit is associated with still lower levels of well-being than chosen early exit.

Moreover, another strong rationale for activating older workers is that early retirement also appears to lead to a form of cognitive deterioration that has been labeled ’mental retirement.’ Rohwedder and Willis (2010) report that cross-nationally, retirement is associated with significant cognitive decline, as the likely result of the fact that retirees on the whole tend to live less cognitively stimulating lives in retirement as compared to working lives. Suggesting a causal argument from retirement to cognitive decline at the macro-level, Rohwedder and Willis (2010) show that countries with a more permissive early retirement scheme do not just have fewer older workers working, but also record lower average cognitive scores on word recognition tests for this age group.

However, there are some indications that this cognitive effect of labor market exit into retirement may not be uniform. Studying the link between retirement and cognitive development, De Grip et al. (2012) find that retired persons face greater declines in information processing speed than persons who remain employed. But, interestingly, retirees’ cognitive flexibility declines less, and this effect persists for six years after retirement. These two opposite effects of retirement on cognitive development are comparable to those of a five to six-year age difference. De Grip et al. (2012) suggest that these two opposite effects can be explained neither by a relief effect (after being employed in low-skilled jobs), nor by changes in lifestyle. And when controlling in addition for changes in blood pressure (which are negatively related to cognitive flexibility), the authors still find lower declines in cognitive flexibility for retired persons. De Grip et al. argue that since the decline in information processing speed after retirement holds particularly for low educated persons, re-activating these persons after retirement could lower the social costs of an aging society. This clearly provides an extra rationale for keeping these older workers in the labor market under cognitively stimulating working conditions, but also for promoting various active forms of being retired, such as social participation to voluntary work. More radically, it also provides a rationale for promoting a second form of active aging: un-retirement.


5. Very active aging: silver workers and un-retirement

Un-retirement is a little studied phenomenon that regards retired workers who go back to the labor market after having already retired. As such un-retirement is a distinct concept from partial retirement (which regards currently working workers leaving the labor market in part). In a study of older workers in the USA, Hardy (2006) finds that only half of all older workers experience retirement as a once-and-for-all withdrawal from full-time employment. The other half phases into full retirement via partial retirement (fewer hours worked), bridge jobs (lower paid jobs, different from previous career jobs) or un-retirement. Hardy finds that the latter phenomenon, un-retirement, is most common among retirees in their early and mid-50s, and it tends to happen mainly within the first two years of retirement. On average, American un-retirees stay un-retired for four years and mainly involve bridge-type jobs paying lower wages and requiring fewer working hours than the previously held jobs, most often in self-employment.

In a first-ever study on European data, Pettersson (2011) studies un-retirement (similarly defined as the extent of re-entry into the labor force) in Sweden for already fully retired persons aged 56 and above between 1994 and 2007. He finds un-retirement to be significant, though smaller in scale than appears to be the case in the USA (the only other country where un-retirement was studied): up to 14 percent of the stock of fully retired Swedes actually go back to work at least once during their spell of retirement. And these un-retirees, most of whom are men, then tend to work, on average, for almost 3 years. Pettersson (2011) finds that the likelihood to unretire is negatively related to the states of being: female; older among the retired; welfare-dependent; born outside Sweden; and having low education; higher pension income; being married to a spouse that has retired. Petterson concludes that un-retirement appears to be a lifestyle decision rather than a decision driven by financial need: once retired, re-entry to the labor market tends to happen as a result of voluntary choice rather than from economic necessity.



Anderson, C. (2009), ‘The Private Consequences of Public Policies: Active Labor Market Policies and Social Ties in Europe,’ European Political Science Review 1(3): 341-373.

Anderson, C., Pontusson, J. (2007), ‘Workers, Worries and Welfare States: Social Protection and Job Insecurity in 15 OECD Countries,’ European Journal of Political Research 46:211-235.

Banks, J., Mazzonna, F. (2012)
’The Effect of Education on Old Age Cognitive Abilities: Evidence from a Regression Discontinuity DesignEconomic Journal, forthcoming.

Behncke, S. (2012), ,Does Retirement Trigger ill Health?’ 
Health Economics, 282-300.

Breidahl, K.N., Clement, S.L. (2010), ‘Does ALMP have an Impact in Social Marginalization?’ Social Policy & Administration 44(7): 845-864.

Broese van Groenou, M., van Tilburg, T. (2012), ‚Six-year Follow-up on Volunteering in Later Life: A Cohort Comparison in the Netherlands,’ European Sociological Review 28(1): 1-11.

Buys, L., Miller, E. (2012), ‚Active Ageing: Developing a Quantitative Multidimensional Measure,’ in G.M. Boulton-Lewis and M. Tam (eds.), Active Ageing, Active Learning: Issues and Challenges, Education in the Asia-Pacific Region: Issues, Concerns and Prospects 15, pp. 103-118.

Carneiro, P. & J. Heckman (2003), 'Human Capital Policy,' in J. Heckman & A. Krueger, Eds, Inequality in America: What Role for Human Capital Policy. MIT Press. Cambridge, MA.

Christensen, K., McGue, M., Petersen, I., Jeune, B. and Vaupel, J.W. (2008), ‘Exceptional longevity does not result in excessive levels of disability,’ Proceedings of the National Academy of Science of the United States of America 105 (36): 13274–13279.

Chung, H, van Oorschot, W. (2011), ‘Institutions versus market forces: Explaining the employment insecurity of European individuals during (the beginning of) the financial crisis,’ Journal of European Social Policy 21(3): 287-301.

Crosnoe, R., Elder, G.H. (2004), ‘From Childhood to Later Years: Pathways of Human Development,’ Research on Aging 24(6):623-654.

Danner, D. D., Snowdon, D.A. (2001). ‘Positive emotions in early life and longevity: Findings from the nun study.’ Journal of Personality and Social Psychology 80(5): 804-813.

de  Grip,  A.,  Dupuy, A.,  Jolles, J.,  van  Boxtel, M. (2012), ‘ Retirement  and  cognitive   development:  are  the  retired   really  inactive?’ Maastricht University: METEOR working paper RM/12/009.

Dorn, D. Sousa-Poza, A. (2010), ‘’Voluntary’ and ‘Involuntary’ Early retirement: An International Analysis,’ Applied Economics 42(4): 427-438.

Erlinghagen, M., Hank, K. (2006), ‘The participation of older Europeans in volunteer work,’ Ageing & Society, 26, 567–584.

Esping-Andersen, G. (1999) New Foundations of Postindustrial Economies. Oxford Polity Press.

Esping-Andersen, G. (2002a), 'Towards the Good Society, Once Again?,' in Esping-Andersen, G., with Gallie, D., Hemerijck,, A., Myles, J., Why We Need a New Welfare State. Oxford: Oxford University Press.

Esping-Andersen, G. (2002b) A Child-Centered Social Investment Strategy,' in Esping-Andersen, G., with Gallie, D., Hemerijck,, A., Myles, J., Why We Need a New Welfare State. Oxford: Oxford University Press, pp. 26-67.

Esping-Andersen, G. (2009) The Unfinished Revolution. Cambridge: Polity Press.

Eurofound (2007), Employment and Labour Market Policies for an Ageing Workforce and Initiatives at the Workplace: National Overview Report Sweden. Dublin: European Foundation for the Improvement of Living and Working Conditions.

Eurostat (2012), Active Ageing and Solidarity between Generations: A statistical portrait of the European Union 2012.

Fogel, R.W. (1999), ‘Catching up with the Economy,’ American Economic Review, 89(1): 1-21.

Fogel RW (2003) The Escape from Hunger and Premature Death, 1700-2100: Europe, America and the Third World, Cambridge University Press, Cambridge.

Gallie, D., Paugam, S., Jacobs, S. (2003), ‘Unemployment, Poverty and Social Isolation - Is there a vicious circle of social exclusion? European Societies 5(1): 1-32.

Giles, L., Glonek, G., Luszcz, M., Andrews, G. (2005), ‘Effect of social networks on 10 year survival in very old Australians: the Australian longitudinal study of aging,’ Journal of Epidemiology, Community, and Health 59: 574-579.

Glymour, M.M., Kawachi, I., Jencks, C. and Berkman, L. (2008). ‘Does childhood schooling affect old age memory or mental status? Using state schooling laws as natural experiments’, Journal of Epidemiology and Community Health, 62(6), pp. 532–37.

Hank, K. (2011), ‘Societal Determinants of Productive Aging: A Multilevel Analysis across 11 European Countries,’ European Sociological Review 27(4): 526-541.

Hardy, M. (2006), ‘Older Workers,’ in R.H. Binstock and L.K. George, editors, Handbook of Aging and the Social Sciences, chapter 12, pages 201– 218. Academic Press, Elsevier.

Heckman, J.J. (2000), 'Policies to Foster Human Capital,' Research in Economics, 54, 3-56.

Heckman, J.J. (2007), ‘The Economics, Technology, and Neuroscience of Human Capability Formation,’ PNAS 104 (233): 13250-13255.

Heckman, J.J., Moon, S.H., Pinto, R., Savelyev, P.A. and Yavitz, A.Q. (2010). ‘The rate of return to the high scope Perry preschool program’, Journal of Public Economics, 94(1-2), pp. 114–28.

Hendricks, J, Hatch, L.R. (2006), ‚Lifestyle and Aging,’ in R.H. Binstock and L.K. George, editors, Handbook of Aging and the Social Sciences, chapter 17, pp. 303– 320. Academic Press, Elsevier.

Hering, M. (2012), ‚Live Longer, Work Longer? Intergenerational Fairness in Retirement Age Reforms in Germany and the UK,’ in Pieter Vanhuysse and Achim Goerres (eds), Aging Populations in Postindustrial Democracies: Comparative Studies of Policies and Politics. Abingdon: Routledge/ECPR European Political Science Series, pp. 79-105.

Herzog, R., James S., House, J.S., Morgan, J. (1991), ‘Relation of Work and Retirement to Health and Well-Being in Older Age,’ Psychology and Aging 6(2): 202-211.

Hinterlong J., Morrow-Howell N., Rozario P. (2007), ‘Productive engagement and late life physical and mental health: Findings from a nationally representative panel study,’ Research on Aging 29(4): 348–370.

Hutchinson, T., Morrison, P., Mikhailovich, K. (2006), A Review of the Literature on Active Aging, document prepared for the Australian Government Department of Health and Ageing, Healthpact Research Centre for Health Promotion and Wellbeing.

OECD (2004), Ageing and Employment Policies: Finland, OECD, Paris.

OECD (2006), Live Longer, Work Longer. Paris: OECD

OECD/Human Resource Development Canada (OECD/HRDC). (2000). Literacy in the information age: Final Report of the International Adult Literacy Survey. Paris: Organisation for Economic Co-operation and Development

Powell, W.W., Snellman K. (2004), ‚The Knowledge Economy’, Annual Review of Sociology 30: 199-220.

Kristensen, N. (2012), ‚Training and Retirement,’ Bonn: IZA working paper 6301.

Kuhn, A., Wuellrich, J.P., Zweimüller, J. (2010), ‘Fatal Attraction? Access to Early Retirement and Mortality,’ IZA Discussion Paper No. 5160.

Marin, B. (2012) Welfare in an Idle Society? Reinventing Retirement, Work, Wealth, Health and Welfare. Vienna: European Centre, Forthcoming.

NCVO (2011), Participation: Trends, Facts and Figures. London: National Council for Voluntary Organisations.

Orth-Gomer, K., Rosengren, A., Wilhelmsen, L. (1993), ‘Lack of social support and incidence of coronary heart disease in middle-aged Swedish men,’ Psychosomatic Medicine, 55(1): 37-43.

Pettersson, ‘Instead of Bowling Alone? Un-retirement of Old-age Pensioners,’ University of Uppsala Economics Department working paper 2011-14.

Stanton, D. (2008), ‘Sustainable Ageing Societies – Indicators for Effective Policy-Making: A Framework,’ in Marin, B., Zaidi, A., eds. Mainstreaming Aging: Indicators to Monitor Sustainable Policies, Alderschot: Asghate, pp. 507-534.

Remery, C., Henkens, K., Schippers, J., & Ekamper, P. (2003), ‘Managing an aging workforce and a tight labor market: Views held by Dutch employers,’ Population Research and Policy Review, 22, pp. 21−40.

Rohwedder, S. Willig, R.J. (2010), ‘Mental Retirement,’ Journal of Economic Perspectives 24(1): 119-138.

Rowe, J., Kahn, R. (1998), Successful Aging. New York: Random House.

Saint-Paul, G. (2009), ‘Does the Welfare State Make Older Workers Unemployable?’ Bonn IZA Discussion paper no 4440.

Settersten, R.A. (2006), ‚Aging and the Life Course,’ in R.H. Binstock and L.K. George, editors, Handbook of Aging and the Social Sciences, chapter 1, pp. 1– 16. Academic Press, Elsevier.

Skirbekk, V. (2008) ‘Age and Productivity Capacity: Descriptions, Causes, and Policy’ Ageing Horizons 8, pp. 4-12.

Tepe, M. & Vanhuysse, P. (2013), ‘Parties, Unions, and Activation Strategies: The Context-Dependent Politics of Active Labor Market Policy Spending,’ forthcoming, Political Studies, Vol. 61.

Van Dijk, S, Turner, M (2010), ‘Active, Productive and Healthy Aging in Germany and the United States: Some insights and critical remarks from a comparative perspective

Von Below, D. and Thoursie P.S. (2010), ‚Last-in, First Out? Estimating the Effect of Seniority Rules in Sweden,’ Labor Economics 17(6): 987-997.

Waginger, U. (2009), Early Retirement and Well-being in Europe. London: King’s College London: MSc thesis.

Waldfogel, J. (2006), What Children Need. Cambridge, MA: Harvard University Press.

Walker, A. (2010), ‚The Emergence and Application of Active Aging in Europe,’ in Naegele, G., ed., Soziale Lebenslaufpolitik. Wiesbaden: VS Verlag fuer Sozialwissenschaften, pp. 585-601.

Warburton, J, Grassman, E. (2011), ‘Variations in older people’s social and productive ageing activities across different social welfare regimes,’ International Journal of Social Welfare 20  (2): 180-191.

Warr, P, Butcher, V., Robertson, I., Callinan, M. (2004), ‚Older People’s well-being as a Functiom of Employment, Retirement, Environmental Characteristics and Role Preference,’ British Journal of Psychology 95: 297-324.

WHO (2002), Active Ageing: A Policy Framework. Madrid: Second United Nations World Assembly on Ageing, April 2002.

Wise, D. (2010), ‚Facilitating Longer Working Lives: International Evidence on Why and How,’ Demography 47: S131-S149.

Zaidi, A., Zólyomi, E. (2011/12) ‘Active Ageing: What Differential Experiences across EU Countries?’, European Papers on the New Welfare, Paper No. 17: Counter-ageing Policies.




[i] Settersten (2006) goes on to point out that such a life course perspective should include (a) aging along multiple dimensions (physical, cognitive, psychological,…) and social spheres (family, work, education, leisure), (b) aging in multiple directions: decline and growth, (c)  processes and mechanisms how experiences in old age are shaped by earlier periods near and far away in time, (d) how aging-related experiences are shaped by specific characteristics of social environment: proximal settings of everyday life (family, social networks, work,…) and distal settings (state, policies, macro-demographic and macro-economic parameters,…).

[ii] Moreover, Anderson (2009) notes that labor market insiders and outsiders both benefit because they do not have exclusive social ties; when outsiders interact socially, insiders are often parties to the interaction. On the other hand, in a study on Denmark alone, Breidahl and Clement (2010) fail to find an effect of participation in such activation programs on three different indicators of social marginalization relating to individuals’ social networks, their stigmatization and their self-esteem.

[iii] Similar results were obtained for the other functional outcomes. Christensen et al. (2008) argue that the discrepancy between the population trajectory and the individual trajectory is caused by increased mortality among dependent individuals. For the individual, long life brings an increasing risk of loss of independence. For society, mortality reductions are not expected to result in exceptional levels of disability in cohorts of the very old. But at the macro level, due to the developing shape of the population pyramid as a result of overall population aging, the share of the population in need of elderly long-term care is increasing while lower average family sizes reduce the share of elderly people who can rely on family care.

[iv] For instance, investigating the relationship between the choice of workforce participation or non-participation and physical and psychological well-being among 1339 participants over 55 in the Americans’ Changing Lives Survey (which oversampled in the 60+ group), Herzog et al. (1991) found that it was not the amount of work that affected the well-being of retirees, but the choice of current role (i.e. being employed or retired). Those people whose current employment (respectively retirement) status was personally preferred showed higher levels of well-being than those whose current role was imposed. The ability to retire voluntarily was associated with higher life satisfaction and less cognitive decline than exiting the workforce involuntarily. Even if the latter findings were statistically significant for the over-65s only, the results showed consistency in so far as the feeling of control over one’s choice was related to higher health and well-being (see also Waginger 2009).