‘Planning for uncertainty’: narratives on retirement transition experiences

Abstract

Retirement is a major life transition which is associated in public discourses with reduced economic productivity and a raft of personal vulnerabilities. Consequently, governmental, health and employment sectors have promoted ‘active’ planning of affordable and ‘healthy’ retirements. This study presents a qualitative exploration of retirement transition and preparation experiences among 52 men and women from rural and urban areas of North East England, United Kingdom. The sample was diverse in terms of social class, income level, health status and type of work exit. Health, finance, social relationships and third-age opportunities were required resources for a good transition into retirement, and a degree of planning was required to mobilise these resources. However, the degree of choice and control around the transition to retirement was highly variable and socially structured. The notion of planning was embedded as a normative practice, particularly in relation to finances, but the practice of planning was highly contingent primarily due to personal circumstances (ill-health, bereavement, relationships) and work exit (redundancy, work stress, changes to shift patterns or hours). The findings offer insights into the reasons why many people do not plan and indicate that many of the assumptions associated with retirement planning warrant further consideration both theoretically and practically.

Introduction

It is widely acknowledged that moving into retirement is a major life transition, indeed it is generally conceptualised as a critical life event that can significantly affect the future (Kim and Moen 2002). A central issue within governmental and other social narratives of retirement is an imbalance between the health and social care needs of an ageing population, and support of such services by a proportionally smaller workforce. This concern underpins policy change in the United Kingdom (UK) and other Organization for Economic Co-operation and Development countries, with increasing emphasis on individual responsibility for maintaining or improving health and social relationships, and planning for financial self-sufficiency in retirement.

The once predictable pattern associated with ageing and retirement is changing (Biggs 1993). No longer is retirement a single event marking the end of paid employment (Wang and Shultz 2010), but a ‘process that varies in its timing and duration’ (Marshall, Clarke and Ballantyne 2001: 380). Increasingly, retirement has become a more individualised experience in terms of the age at which it occurs, the form it takes (e.g. gradual/abrupt, voluntary/involuntary, early/late) and the increasing complexities associated with ensuring adequate provision (Phillipson 2004). In policy terms, perhaps the most significant in the European context is the shift from early retirement as a norm, to actively encouraging and incentivising those approaching retirement to remain longer in the workforce (Ebbinghaus 2006; Reday-Mulvey 2000).

In the UK, the state pension age for women has been steadily increasing since 1995 and will reach parity with men in 2018; in 2020 it will increase for men and women from 65 to 66, further institutionalising the extension of working life (Loretto, Lain and Vickerstaff 2013; Sinclair, Moore and Franklin 2014). Despite these changes in policy and encouragement to remain in work, Phillipson (2004) and Vickerstaff and Cox (2005) found that among the majority of current and near future retirees in the UK there is not much desire to work until state pension age, let alone work longer. This may be linked to the ‘promise’ of the third age and narratives regarding retirement as a time of personal development and leisure. This narrative may clash with one that prioritises personal responsibility to the state (through continued work). Nevertheless, in other moments these narratives are complimentary – such as placing emphasis on maintaining or improving health in order to lead an active and enjoyable life, but also to minimise potential burden on services. The degree to which individuals – and couples – are able to achieve these aims is dependent in part on a range of variables that may be beyond their immediate control (Phillipson 2004; Vickerstaff and Cox 2005).

Retirement decision making is relatively underdeveloped theoretically (Jex and Grosch 2013; Wang and Shultz 2010), with the notable exception of Beehr's (1986) theoretical model which emphasises that retirement is a complex psychological process which may occur over a long time period. Sequential decision-making models have been proposed (Feldman and Beehr 2011; Jex and Grosch 2013) and incorporate discrete, but related, decision-making stages. These models emphasise retirement decision making as a rational process, but other more general theories related to ageing are also employed in studies of retirement. For example, continuity theory (Atchley 1999) supports empirical work demonstrating the benefits of maintaining routines and roles before and after retirement (Feldman and Beehr 2011), particularly those that are rewarding and socially productive (Nazroo 2015). Retirement decision-making theories begin with the premise that planning (Jex and Grosch 2013) or ‘imagining the future’ (Feldman and Beehr 2011) is the starting point. Wang and Shultz (2010) suggest that the process begins with informal planning which progresses to more deliberate and overt formal retirement planning. Current evidence, such as it is, indicates that there is a great deal of variability in the degree to which individuals actively plan for retirement (Ekerdt et al. 2001). Numerous factors contribute to this apparent lack of planning, leading us to suggest that the assumptions associated with retirement planning warrant further consideration.

As well as receiving attention from psychologists (Wang 2013), the planning associated with retirement has been considered by economists (Adams and Rau 2011; Banks and Smith 2006; Donaldson, Earl and Muratore 2010) and gerontologists (Ekerdt et al. 2001). Work from this diverse body of literature, primarily from cross-sectional and retrospective research in the United States of America (USA), suggests that both financial and psycho-social planning are important for wellbeing in retirement. Longitudinal research based on the US Health and Retirement Survey shows that financial and psycho-social planning are the two largest contributors to satisfaction and health in retirement, after taking into account health status, reason for retirement and income (Noone, Stephens and Alpass 2009). Adams and Rau (2011) conclude that retirement preparation is critically important to quality of life in retirement. A meta-analysis of the antecedents and consequences of retirement planning and decision making found that a lack of retirement planning is a predictor of subsequent illness, but that retirement planning itself does not predict retirement satisfaction (Topa et al. 2009). Combined with work by Donaldson, Earl and Muratore (2010), these findings present a somewhat contradictory picture of the impact of retirement planning on wellbeing in retirement. This may be due to methodological limitations, e.g. relying on cross-sectional and/or retrospective data. There are also well-documented cultural differences in retirement practices (Hershey, Henkens and Van Dalen 2007) as well as different institutional arrangements regarding retirement norms and resources available for retirees (Fasang 2010; Maltby et al. 2006).

Shultz and Wang (2011) determine that disadvantaged groups are likely to be constrained in their capacity to plan actively for retirement. Phillipson (2004: 160) highlights the ‘significant polarisation [that] has emerged between affluent early retirees and disadvantaged older workers’, a trend that looks set to continue (Brewer et al. 2007). Moen, Kim and Hofmeister (2001: 55) demonstrated how gender has become a ‘key source of heterogeneity in the nature and effects of the retirement process’ due to the gendered nature of career trajectories and the workplace, most notably women having interrupted careers, lower wages and lower pensions.

Individuals have increasing personal responsibility for their retirement and many aspects of planning have become highly complex (Adams and Rau 2011). Yet, despite the evidence as well as the ‘intuitive assumption’ that ‘planning for retirement will lead to positive outcomes’ (Noone, Stephens and Alpass 2009: 296), US data indicate that a considerable proportion of older workers do not engage in any retirement planning (Adams and Rau 2011; Ekerdt et al. 2001). This calls into question the notion that, with the exception of disadvantaged groups, retirement planning is normative. Attention to actual experiences of retirement transitions and associated planning is required, particularly as there is a paucity of qualitative research on experiences of retirement planning (Jex and Grosch 2013).

The focus of this paper is on retirement preparation and transition experiences of people aged over 50. Our research rests within the paradigm of retirement as a longitudinal process requiring multiple adjustments (Wang 2007). We therefore explore the variability of transitions, examining the extent to which planning was incorporated into the various retirement trajectories of our participants. We thus offer insights into why people may not engage in deliberative planning for retirement and the implications this has for policy and practice.

Methods and data

The research took place in North East England, an area whose traditional economic base of ship building, steel making and mining has all but disappeared. The region has one of the most ethnically homogenous populations within England with 93.6 per cent defining themselves as white British compared with 80.5 per cent across England and Wales (Office for National Statistics (ONS) 2012).

Participants were initially recruited from large employers through which we accessed people in urban and rural locations pre- and post-retirement (N = 22). Further recruitment took place via: a national charity representing older people (N = 3); a physical activity class in a low socio-economic status (SES) neighbourhood (N = 5); a local older people's forum (N = 8); and community-level advertising in low-SES and rural locations (N = 10). Leaflets were also posted in urban neighbourhoods with proportionally high numbers of Black and Minority Ethnic residents. All participants were offered a £10 (€12.31, $16.75) shopping voucher and travel expenses. The present or former occupation of each participant was categorised using the ONS nine main occupational typologies (ONS 2010), with all nine categories represented in the sample (Table 1). Data were collected initially through focus groups (N = 6) and then interviews with either individuals (N = 13) or couples (N = 4). We observed that the details of individual experience throughout retirement transitions (which were often complex) were not always accessible in group format more suited to discussing a shared experience. Consequently, after an initial period of data collection (during which five focus groups were conducted), subsequent data were collected by semi-structured interview with individuals or couples. Data were audio recorded and transcribed verbatim by a transcription company, 10 per cent of which were checked for accuracy. Transcripts were anonymised and pseudonyms used throughout.

Table 1. Participant demographics

Notes: ND: no data. ONS: Office for National Statistics.

Data collection and analysis occurred iteratively, informed by principles of the constant comparative method (Maykut and Morehouse 2002). The authors read and coded a selection of the transcripts and made notes regarding descriptive coding categories, before agreeing an initial thematic coding framework, including brief ‘statements’ or propositions that defined the categories as they developed. Analysis was conducted through the writing of analytic memos, and drawing together arguments based on the data within and across codes. Data analysis workshops, where analytic memos were compared and discussed, shaped the emerging analysis.

Based on our analysis of focus group and interview data from 38 women and 14 men (aged 51–82, mean age 60), we structure our results by (a) considering the diverse pathways into retirement, and the consensus around an ‘ideal’ transition; (b) exploring actual practices of retirement planning and associated contingencies; and (c) examining the multiple ways in which the transition was experienced as ‘unplannable’. As shown by Table 1, our findings are derived from people from all occupational social classes, with and without long-term illness, with and without occupational pensions, as well as people fully retired, working and working reduced hours in anticipation of retirement.

Results

Pathways into retirement

Experiences could broadly be divided into those who anticipated and to varying degrees planned their work exit; those whose retirement was ‘forced’; and the one individual who, following a period of retirement, re-entered the workforce. Among those who anticipated and planned their transition out of work, most individuals left work directly without any phased reduction in hours, but others were able, through negotiation with their employers, to phase their exit over a period of time. Ill-health, redundancy and changing work circumstances were reasons given by those who felt ‘forced’ into retirement. Notwithstanding this range of personal experiences, there was broad consensus on three elements that constituted a ‘model’ or ‘ideal’ retirement transition: choice and control over the timing and mode of exit from work; employer's attitude to retirement, including gradual and flexible options; and having access to resources to prepare for the event. This account from couple Alice and Joe exemplifies all three elements:

Alice: I did go part time for a couple of years [before retiring]. I did four days then three days … Because I knew I was coming up to the age of retirement and I felt the need to get used to being retired because I've worked all my life and I knew it was going to be more difficult than maybe I thought it was going to be just to stop altogether and they [employers] were happy for me to do that…

Joe: …after watching [Alice] go through the scale down, I was happy to do that. I thought that's going to be the best way rather than an absolute stop … I've seen an awful lot [of other retirees] who came to a dead stop and have difficulty coping with it so I knew that to taper down was a good idea.

(Alice, aged 67, gradual retirement from private-sector job, occupational social class 6; Joe, aged 65, gradual retirement from self-employment, occupational social class 7)

Alice had a high degree of self-awareness concerning the potential impact of stopping work, partly due to her work with older adults which had given her considerable exposure to retirement transition experiences prior to her own retirement. She had clearly thought about it and, with the support of her employer, was enabled to plan a gradual transition into retirement, exercising a high degree of choice and control, in a way that suited her need to ‘get used’ to being retired. Her partner, influenced by Alice's pro-active and agentic transition, also adopted this gradual approach to exiting work. This can be sharply contrasted with Sam's transition experience, which was sudden as a result of a health problem and followed by a deep depression:

[After a series of health problems] I decided on the spur of the moment to call it a day which was, regrettably, the worst thing I ever did … If I was to do it again, I would do a gradual process. (Sam, aged 65, immediate retirement from private-sector job, occupational social class 6)

Throughout the interviews, participants continually referred to the resources required to achieve a ‘model transition’: finance, health, social relationships, control over timing and mode of exit, sense of purpose and planning for retirement, factors known to influence retirement transitions and subsequent wellbeing in retirement (Heaven et al. 2015; Wang 2013).

Planning for retirement

Narratives on planning focused primarily around two aspects, financial planning and planning a post-working life, i.e. activities, roles and relationships in retirement. The idea of planning was firmly embedded as crucial to a good transition and a good retirement. The potential consequences of not doing so was illustrated by the experience of one focus group participant who suffered serious mental health problems on retiring and reflected that this was due to a sudden and unplanned retirement. The conversation then turned to the importance of planning:

Marilyn: I think preparation is the key word. So many people don't even think that suddenly this whole world is opened up to them … but if you have made no preparation … I have heard of many people suffering how you suffered and going into quite a bad decline … I can't think of anything worse than retiring and then having nothing to do…

Lindsay: I think it's an important point to think about not the day you retire but a couple of years previous.

Peter: As long as you've planned for it.

Marilyn: Planning, that's the key.

(Focus Group 4, post-retirement, aged 58–60, private-sector company, occupational social classes 5 and 7)

Structural constraints, most apparent in relation to financial planning, were repeatedly discussed:

Jen: I know you're supposed to look forward to [retirement] and save up to do it, but you don't always.

Carol: Not everybody's able to do that either, no.

Zoe: Not everybody's able to.

All: No.

Carol: And then over the last few years, they're making you work longer and longer and your pension's going down and down, so it's frightening really, I think.

Zoe: Well, it does take a lot of planning … you really have to plan for your finance, don't you.

(Focus Group 2, pre-retirement, aged 57–60, private-sector jobs, occupational social classes 4 and 7)

The group narrative emphasises the problems of saving for retirement when incomes do not permit this. The now widespread and frequent changes to pension conditions that result in reduced pension payouts provoked further insecurity and anxiety, as well as the requirement to work longer due to the increase in the age of the UK state pension. Whilst the onus is increasingly on individuals to manage their financial affairs successfully to provide for retirement, the continually shifting landscape of pensions, annuities and interest rates makes financial planning challenging. Increasingly, knowledge of complex financial systems is not itself adequate, and individuals are required and expected to plan their retirement finances in an uncertain and turbulent economic climate.

Our female participants in particular mentioned the inadequacy of their retirement finances, reflecting gendered work experiences of lower pay and interrupted careers due to caring responsibilities. Many partnered women were like Jennifer, whose ‘pension doesn't pay for any standard of living’, but who were financially adequate when their partner's pension was taken into account. Marital status was a significant factor in the social structuring of retirement resources, as one of our focus group participants observed:

Single people … have to prepare [for retirement] with even more care … it's just my experience of dealing with hundreds of retired staff, they who were in stable marriages and there was two incomes, two pensions, were all moderately comfortably off, but the single people … they struggled. (Megan, aged 60, immediate retirement from private-sector job, occupational social class 4)

Financial security, preceded by financial planning, was viewed as a cornerstone for a good retirement, yet the constraints in doing so were persistent themes. However, planning for a post-working life was also viewed as crucial:

I think once you've left work, if you haven't planned and given enough thought to how you will fill your time, let's assume one is healthy and they have a little bit of money to do things … if they haven't planned … [you] think, well what do I do now? (Focus Group 3, pre-retirement, occupational social classes 2 and 3)

Mabel: I think it's a good idea to have something in place before you retire, some idea of what you want to do…

James: I would say it's dangerous to have a lack of structure for too long.

(Focus Group 1, post-retirement, occupational social classes 7)

These accounts point to a general consensus about the importance of planning for retirement, both in terms of finance and developing a post-working identity. These reinforce findings from the quantitative literature on retirement planning. The narratives also emphasise some of the factors which constrain planning, including finance, but also availability of post-work opportunities.

‘Planning’ for uncertainty

Personal circumstances and mode of work exit were the two main factors that contributed to the unpredictable nature of the retirement transition, and made planning difficult, impossible or undesirable.

Personal experiences – ill-health, bereavement and relationships

Narratives were replete with examples of how ill-health had impacted on retirement plans. For some, ill-health (physical and mental) forced individuals out of work earlier than anticipated, with consequences for finances in retirement as well as activities and past-times that could then be undertaken. For others, it was the ill-health of family or close friends that impacted on their lives, often obliging them to take on a caring role. In the case of Carl, it was the uncertainty surrounding his wife's health that made him circumspect about planning altogether:

My wife's had two bouts of cancer and that kind of focuses your mind on how life might have been … so in my head I had no plans [for retirement]. Other people you hear, ‘Oh we're going to do this and we're going to do the other’. I have nothing in my head that is planned. (Carl, aged 56, self-employed, working, occupational social class 3)

Accounts of how bereavement impacted on retirement plans are illustrated by Joan's account of her partner's sudden death prior to retirement. Joan's retirement had been thought through, discussed at length and planned with her partner yet:

I've now got to sort out [retirement] which is solo effectively, so I'm still trying to get there really … my expectation was that he and I would retire, we would travel because that was what we wanted to do and do other things as well … of course that's all been completely turned on its head … there's a reinforcement of what I've always believed which is the complete randomness of life. (Joan, aged 60, immediate retirement, occupational social class 2)

Another form of loss was the impact of divorce. As well as the personal impact, John's narrative demonstrates how the rift had a considerable effect on his financial circumstances, causing him to change his original retirement plans and work longer:

A divorce kicks in so that [lump sum] has just been decimated. The pension's decimated and you end up with a lot less than you were really geared up to deal with and this in itself is a bit of a slap in the face, so you've got to re-adjust because of that … I can get another couple of years work out of this [job] and it will be the sweetener to losing all that money … and that's why I decided to stay on at work. (John, aged 60, reduced work from full time to part time, occupational social class 3)

John's account is a further demonstration of the close association between resources in retirement and relationship status.

Work exit

Vickerstaff (2006) demonstrated from a UK study that many people do not choose when to retire. In a similar vein, many of our narratives revealed that the lack of choice over the timing of retirement was closely bound up with lack of choice over the manner in which people left or re-entered the workforce, hampering the ability to plan or causing havoc with pre-existing plans. The most dramatic and unpleasant work exit was involuntary redundancy. Redundancy is a major setback to financial planning with immediate consequences and considerable long-term implications. As Kate explains below, her experience of redundancy, not untypical, not only threw her plans off course, but also had a profound impact on her sense of self-worth. Following a period of depression, Kate returned to education part time and at the time of the interview was seeking work:

…it's a real blow to the ego to be made redundant and the fact that I'd thought about finances and think I would have five years more salary and five years more pension contributions at work … it took me oh probably a year to really come to terms and think okay and I still have my moments when I get really angry with work and if I feel a bit down or depressed. I think it's like they're saying to you we don't want you any more, what you did in the past wasn't particularly good … it is a real blow to you as a person. (Kate, aged 61, made redundant from full-time job aged 59, occupational social class 3)

While not classed as redundancy, a number of our participants recounted work situations in which they felt coerced into retirement earlier than they had planned; moving into shift work, changing hours, work patterns or roles were common work-related reasons. Often this was coupled with ill-health, thus pressing people into retirement. As well as the financial consequences, this mode of labour force exit was deeply felt as an unpleasant way to end a lifetime of work and was often associated with high levels of stress and anxiety, not usually conducive to a smooth transition into retirement.

For some participants, an abrupt work exit precipitated problems adjusting to retirement. From Peter's account, we can see that he had planned his work exit in terms of being able to afford to retire, but had not ‘thought through’ or planned his post-working life, largely because he was so busy with his full-time work prior to retiring:

I found, I must admit, having worked so many hours and I loved work and just all of a sudden cutting off … I couldn't really plan anything [due to working away from home and working very long hours] … the only thing I wanted to plan was financial … I didn't have a structure and there wasn't a purpose to get out of bed in the morning. There wasn't a need to have a shave and it's – I just … I got quite depressed really. (Peter, aged 60, retired immediately at 58, occupational social class 7)

Peter's account indicates that financial planning alone was insufficient for a smooth shift into retirement. Many of our, mainly male, participants referred to their own retirement transition in negative terms. The widespread use of expressions such as ‘over the cliff’, ‘all or nothing’, ‘shock to the system’, ‘drop dead point’ and ‘go downhill’, implied a downward trajectory that reflected their own retirement transitions and the ensuing negative emotions. In contrast to Alice's description of a gradual and highly planned transition, many accounts were of the, ‘planned’ but not ‘thought out’ variety. This emphasises Alice's fairly unusual experience which reflects the disposition of her employer as well as the nature of her job in which she developed a high level of awareness of retirement transition issues. In general, most participants did not perceive there was an option for an alternative to an ‘all or nothing’ transition into retirement, as this focus group discusses:

James: Obviously retirement is a one-off decision. You don't get a chance to go back and have another go but there's a lot of people that I feel that they've looked forward to retirement, they've talked up retirement so much … that within a few months of retiring if they got another chance would they actually retire or would they actually do what you're doing which is carry on working on reduced hours? I think there's a lot of people make so much of this big day, the day that I retire, I can't wait for that day…

Mabel: It's a shock.

James: …I think they would say, could I work two or three days, could I just do something but unfortunately it tends to be a one-off chance.

Mabel: Unless you get another job.

James: Unless you get another job, yeah, unless you start afresh somewhere else.

Karen: Yes but who wants to employ anyone of 60, 65?

Scott: B&Q (laughter) [large UK retail chain well-known for employing older workers]

(Focus Group 1, post-retirement, private company, occupational social classes 7 and 9)

Alternatives to a complete ‘all or nothing’ break from paid employment were not perceived to be widespread, nor commonly experienced by our participants. Re-entering the labour force was also unusual. However, Gill recounted her return to part-time work as she viewed her retirement aged 58 as a disaster:

I'd lost my identity … I missed the money, the people, the chat … I done a return to practice [after five years of retirement]… and I went back … what a lovely three years I had … so I worked till I was 67 … I had an accident … I couldn't go to work and I thought that was God saying now it's about time you stopped really. I looked at it that way and we just slowly got it [retirement] right, but it took a while. (Gill, aged 68, retired initially aged 58, re-entered workforce part time aged 63, retired fully aged 67, occupational social class 3)

Gill's original motive for retiring aged 58 was to be with her retired partner who had been made redundant and could not find more work despite a concerted effort. Following a very unhappy period in retirement, Gill returned to paid employment part time. Among participants there was both a perception and a reality that those with higher levels of education and skills were in a stronger labour market position and therefore able to adapt work to suit their needs in their sixties. Among those who wished to work beyond state pension age, none of our participants wished to work full time, anti-social hours or shift work.

Discussion

This study highlights an inherent contradiction in the notion of ‘planning for retirement’, largely as a consequence of the circumstances surrounding the retirement transition period and the lack of choice and control in work exits. The narratives demonstrate that the notion of planning is culturally embedded, but that the practice is constrained. Consequently, as planning for retirement takes place in the face of considerable uncertainty, plans went awry, and some desisted from planning in the first place. Yet, the narratives also demonstrated the pitfalls of not, at the very least, ‘thinking through’ retirement transitions, and there were several accounts of the disasters resulting from ‘unplanned’ transitions. Whilst participants’ experiences of the retirement transition were diverse, there was broad agreement about the ‘resources’ required for a good retirement: health, finance, relationships, social connectedness and third-age opportunities.

Theoretical approaches which emphasise a lifecourse perspective, connecting individual lives and the historical and socio-economic context in which these unfold (Giele and Elder 1998), are helpful in contextualising our findings; more specifically, Dannefer's (2003) cumulative advantage and disadvantage model which suggests social processes, such as those dictated by social class and gender, influence and become more pronounced during the process of ageing. We saw this clearly in relation to finances, and choice and control over work exit. So, whilst the notion of financial planning for retirement was embedded as a norm, the possibilities of doing so were structured by occupational social class and gender. A similar observation has been made by Vickerstaff and Cox (2005), based on accounts of white- and blue-collar employed and retired men and women in 2002–2003. The accrual of lifetime resources for retirement as well as the levels of disruption arising from individual circumstances, and the impact of employment practices and changing social policies, was highly socially structured; those on lower incomes and therefore lower pensions had less choice in when and how to retire than those in more advantaged positions. Vickerstaff and Cox (2005) described a fragmentation of retirement transition experiences and increased risks for individuals which they termed ‘structured individualisation’. This process constrained individuals’ capacity to exercise choice or control over their transition. In the UK, growing concentrations of advantage and disadvantage in wealth (Bardasi, Jenkins and Rigg 2002) and health (Marmot 2010) profoundly constrain the exercise of choice and control in planning for retirement. Our findings show that unplanned events could happen to anyone, although those more disadvantaged in terms of health and wealth had fewer resources with which to respond.

Our findings resonate with theoretical perspectives concerning the third age, and ‘cultures’ of ageing (Gilleard and Higgs 2000) arising from the diverse nature of identity in later life (Jones et al. 2008). The need for a social identity in the third age, assert Gilleard and Higgs (2000: 38), ‘demands deliberation and planning – the active construction of a “post-work” identity’. Laslett (1989) indicates that the acquisition of a third age identity depends on decision making, arising from the exercise of individual choice. Individual agency is therefore key to the achievement of a third-age identity. Our data show that whilst individuals varied considerably in their deliberations and planning for this phase of the lifecourse, most recognised the need to consider, if not plan, this post-work period. This was generally couched in terms of imposing a structure. However, the core of these narratives were about the creation of meaning, purpose and identity, congruent with theoretical approaches highlighting the importance of continuity (Atchley 1999) and the establishment of socially productive and meaningful roles in retirement (Moen 1996). Little research has addressed the issue of planning directed towards social roles and identity in retirement and this is an area which requires greater attention (Taylor and Schaffer 2013).

Connecting theories of the third age with cumulative advantage and disadvantage (Dannefer 2003) sheds light on how structural factors, such as social class and gender, play out across the lifecourse to shape the construction of post-working identities. Whilst this was evident in our data, it remains the case that the retirement transition experiences of working-class groups are relatively undocumented. An exception to this is a working-class cohort of 87 men and 10 women born in the UK between 1943 and 1947 (Goodwin and O'Connor 2014). Most interviewees in Goodwin and O'Connor's study recognised that their retirement provision might be insufficient, certainly to afford the leisured life of the retiree promoted by media and marketing; uncertainty and lack of preparation characterised most of the narratives, which related to finances, changing dependencies between family members, and the desire for a structure or role. The authors argue that the reason for vague and unclear retirement plans was primarily because ‘the future as a retiree was an imagined one’ (Goodwin and O'Connor 2014: 585). As a result they conclude that

…for many of this largely working-class sample, the implied advantages of the ‘baby boom’ label are far from the reality, with their experiences being similar to those who have gone before with an adjustment to retirement marked by uncertainty and anxiety. (Goodwin and O'Connor 2014: 586–7)

A reduced material base profoundly influences retirement planning, although it does not mean, as our data indicated, that individuals do not consider their retirement. Contrary to the prominent idea of the ‘planful, deliberative pre-retiree’, prevalent in theories of retirement decision making (Feldman and Beehr 2011; Jex and Grosch 2013), the ‘pre-retiree with uncertain plans is an authentic type of older worker … uncertainty about prospective retirement is a circumstance that should not be ignored’ (Ekerdt et al. 2001: S169).

We suggest that models of retirement decision making, which generally view decision makers as highly rational, could incorporate qualitative findings which highlight the lived experience and the ‘tradeoffs that people may make in the process of making retirement-related decisions’ (Jex and Grosch 2013: 278). Our data add to a growing body of work which emphasises the heterogeneity of retirement expectations, how retirement is socially structured and, importantly, the need for uncertainty to be embedded into retirement preparation policies and practices, since ‘so far, no theoretical framework has incorporated this notion of uncertainty’ (Wang and Shultz 2010: 186). Among our participants who actively decided not to plan or those whose plans had been ‘blown off course’ (Vickerstaff and Cox 2005), there was still engagement with what might be termed ‘thinking through’ retirement. Whilst this might be regarded as a much looser practice than retirement planning per se, it is evidence of a practical and entirely logical response to the inherent uncertainties in the retirement transition period and what lies beyond. Active acknowledgement that uncertainty is an inherent part of this life phase and elements of retirement cannot, in fact, be planned for, posits an experientially embedded picture that individuals may be more willing to engage with. In practical terms, this may assist individuals to ‘think through’ and thus prepare for various scenarios, taking into account the ‘often messy and disrupted process’ that constitutes retirement (Loretto and Vickerstaff 2013: 65).

The policy and practical implications of this study suggest that retirement preparation programmes need to take account of personal, household and work exit factors, recognising the uncertainties and contingencies that can and do affect retirement preparations and plans. Whilst a focus on financial issues is vital, due emphasis on post-working life is also an essential element (Taylor and Schaffer 2013). Individuals (and couples) may benefit from assistance to reflect on ‘core resources’ within their control (e.g. maintaining family relationships) rather than planning specific details of retirement that might be highly dependent on factors outside their control. Incorporating these factors in a meaningful way would be a considerable achievement. Rigorous evaluation, currently lacking in the pre-retirement programme arena (Adams and Rau 2011; Jex and Grosch 2013), would be a positive next step.

Caution must always be exercised when considering the conclusions of an entirely qualitative study. The strength of this work lies in the depth of the accounts, however, a question remains about the extent to which similar findings would be obtained in other settings or contexts. Methodologically, we have followed accepted practice in fieldwork, analysis and interpretation (Barbour 2003). Our sample included more women than men, a common finding, and there is evidence that older men and women respond differently to recruitment to health promotion research (Carter et al. 1991). Earlier we pointed out that there are cultural differences in retirement practices, as well as differences in resources available for retirees among industrialised nations with established welfare state provision. This might suggest that our findings may be more applicable to a UK context. However, the overarching concepts which we have derived from the data concerning uncertainty appear pertinent to both the UK (Goodwin and O'Connor 2014) and the USA (Ekerdt et al. 2001). Future cross-national research is required to establish the extent to which this concept emerges as significant in retirement transition and planning narratives in other industrialised countries.

Conclusions

Drawing on the lived experiences of retirees and pre-retirees, we have shown how retirement transitions are no longer discrete events, but are diverse, disrupted and socially structured. Narratives revealed how notions of planning for this life-stage were embedded as a normative practice, particularly in relation to finance and less so for post-work roles and activities. However, when it came to the practice of planning, a range of contingencies, conceptualised by our participants as ‘uncertainties’, elicited a much more nuanced and realistic outlook towards planning for retirement. At the same time, interviewees recognised the importance of ‘thinking through’ and preparing for retirement. The very notion of ‘planning for retirement’ is therefore paradoxical. Bearing in mind that many people, for a range of reasons, are not at all prepared for retirement, an ‘imagined future’ might be more realistically prepared for if discourses on retirement transitions were couched more in terms of ‘thinking through’, thus removing some of the ‘certainty’ associated with ‘planning’. Acknowledging this is an important step towards supporting people through their transition into retirement.