Uncoupling in the third age – the importance of the existential context for late-life divorce

Abstract

Late-life divorce is increasingly common in many Western countries, however, studies on this transition remain scarce. The purpose of this article is to study attributed reasons for late-life divorce, and if any life phase-typical aspects can be identified in these attributions. Qualitative interviews were carried out with Swedish men and women aged 62–82, who after the age of 60 had divorced from a cross-gender marital or non-marital co-habiting union (N = 37). The results, analysed using principles from Grounded Theory, revealed four different types of narratives: (a) incompatible goals for the third age, (b) personality change caused by age-related disease, (c) a last chance for romance, and (d) enough of inequality and abuse. A central insight and an original contribution generated by the study was the importance grey divorcees attributed to the existential conditions of later life in their divorce decisions. The results are discussed in relation to theories of late modern intimacy and the third age.

Introduction

Historically, widowhood was the primary pathway out of marriage, but for half a century it has become common to have divorced earlier in life and, as a consequence, to enter old age unmarried or, increasingly in the last decades, to divorce late in life. The incidence of late-life divorce has rapidly increased in many parts of the Western world (Kingston, 2007). Brown and Lin (2012) have referred to the rapid increase in grey divorce in the United States of America (USA) – a doubling of divorce rates for people aged 50+ between 1990 and 2010 – as a ‘grey divorce revolution’. Sweden is another example of this demographic trend where divorce incidence rates among people aged 60+ have increased rapidly since the millennium (from 1.9‰ in 2000 to 3.3‰ in 2020). Today, one in ten Swedes getting divorced is aged 60+ (Statistics Sweden, own calculations).

Already four decades ago Uhlenberg and Myers (1981) predicted a future increase in late-life divorce. They gave four reasons for this prediction:

1. (1) A growing acceptance of divorce in society. Indeed, since the divorce revolution in the 1960s and 1970s marriage has increasingly been framed as a vehicle for self-fulfilment, and divorce as a solution whenever marriage fails to fulfil this promise (Cherlin, 1992; Giddens, 1992; Hackstaff, 1999; Coontz, 2005).

2. (2) An increase in remarriages that are more prone to end in divorce than first marriages. Indeed, as divorce rates have increased so have the number of remarriages. About 40 per cent of all new marriages in the USA are remarriages for at least one of the spouses (Raley and Sweeney, 2020) and among people aged 50+ the risk of divorce is 2.5 times higher in remarriages than in first marriages (Brown and Lin, 2012).

3. (3) An increase in female labour force participation that would make women economically less dependent on their spouses. Indeed, women's workforce participation in Sweden has increased since the early 1970s and so has women's income relative to men's (from 75.8% in 1975 to 83.2% in 2013; Statistics Sweden).

4. (4) A longer life that lengthens exposure to the risk of divorce. Indeed, in the USA, average life expectancy has increased by more than five years from 1980 to 2020 and, in Sweden, by five years for women and eight years for men. As Brown and Wright (2017: 6) write: ‘Individuals who survive to age 65 can expect to live another 20 years, which could be a long time to spend with someone from whom one has grown apart.’

Despite these early predictions of a future increase in late-life divorce and the above-described demographic change, research on grey divorce remains scarce. The purpose of this article is to study the reasons late-life divorcees attribute to their divorce, and if any life phase-typical aspects can be identified in these attributions. Based on our empirical findings, we argue that the existential conditions that surround the third age constitute an important context for the decision to divorce in later life.

Previous research

Gerontologists have for a long time investigated the consequences of marital dissolution through widowhood, for loneliness, identity, social support, contacts with adult children, etc. (see e.g. Lopata, 1973, 1996; Martin Matthews, 1991; Carr and Utz, 2001). Corresponding research about late-life divorce remains scarce, although like widowhood it is a life-altering transition that may have important consequences for the experience of ageing and later life.

Within the wider field of social science, divorce has attracted more attention and been the object of a large number of journal articles, decade reviews (Amato, 2010; Raley and Sweeney, 2020), research handbooks (Fine and Harvey, 2006) and conferences (the yearly Conference of the European Divorce Research Network). In the widest of terms, research has focused on reasons for divorce (a moot topic in widowhood research) and its consequences, in terms of economy and wellbeing of ex-spouses as well as the development of children and parent–child relationships, etc. However, most of this research has focused on uncoupling in unions of young or mid-life adults with children in the household. Very little attention has been directed towards late-life divorce. In the following, we provide a brief overview of research about predictors and motives for divorce generally before we focus on late-life divorce.

An important epistemological distinction has been made between predictors and motives for divorce (Hopper, 1993; de Graaf and Kalmijn, 2006; Allen and Hawkins, 2017). Predictors are factors that increase the risk of divorce while motives are subjective reasons for divorce. They do not always overlap. Many predictors are structural in nature (e.g. teenage marriage), and individuals are not necessarily aware of or reflect on their divorce decisions in such terms. In a decade review, Amato (2010; see also Lyngstad and Jalovaara, 2010) lists a number of commonly identified demographic predictors for divorce, including e.g. marrying as a teenager, being poor or unemployed, being in an interracial marriage, having stepchildren, having experienced premarital co-habitation, a previous divorce or a parental divorce or having had premarital children.

Motives are subjective reasons and to some extent rhetorical constructions that help individuals understand and legitimise their decision to divorce (Hopper, 1993; Walzer and Oles, 2003). In a Dutch study, de Graaf and Kalmijn (2006) identify and divide such motives into three categories: relational issues (e.g. communication or sexual problems), behavioural problems (e.g. violence, alcohol abuse or excessive spending), and problems related to work and the division of labour (e.g. who takes care of the home). Motives are best known by those who initiate divorce, but both partners can usually imagine and attribute reasons for their divorce. In responses to an open-ended survey question, Amato and Previti (2003) identified attributed reasons for divorce, including infidelity, incompatibility, alcohol or drug problems, growing apart, personality problems, communication problems and abuse.

Motives pose a number of epistemological challenges for researchers. Individuals are not necessarily aware of or able to account clearly for the reasons for their divorce. Divorce is a complex life decision that might not be possible to reduce to a single decision or reason. Divorce decisions can take a long time to mature (Allen and Hawkins, 2017). Reasons to leave are combined in different ways and weighted against reasons to stay, and the balance can change several times before, if at all, they are acted on – a process referred to as divorce ‘ideation’. From the individual's point of view, single motives are in this sense often a simplification of a complex decision. Similarly, from the researcher's point of view, every categorisation of divorce motives is an analytical construction that reduces the motives presented by informants in a specific way. As a consequence, lists of divorce motives tend to vary between studies.

After the historical shift from at-fault to no-fault divorce legislation in many Western countries in the 1970s, divorce motives have increasingly shifted from individual faults to the failure of relationships to satisfy needs for self-realisation (see e.g. Wu and Schimmele, 2007). Basing themselves on Dutch data of divorces occurring from 1949 to 1996, de Graaf and Kalmijn (2006) showed that divorce motives had shifted over time: from ex-partners' behavioural problems (such as violence) in the beginning of the period to motives of emotional dissatisfaction (such as growing apart or not receiving enough attention) at the end of the period. But motives can also shift over the individual lifecourse and be different in later life than earlier life phases. This brings us to research about late-life divorce.

Reasons for late-life divorce

Like general divorce research, studies of grey divorce have investigated both reasons for and consequences of late-life divorce. Research has focused on the consequences of grey divorce for wellbeing (Hammond and Muller, 1992; Bair, 2007; Bowen and Jensen, 2017; Carr et al., 2019; Lin and Brown, 2020), personal economy (Crowley, 2018; Lin and Brown, 2021), and the relationship to adult children and grandchildren (Aquilino, 1994; King, 2003; Shapiro, 2003; Greenwood, 2012; Abetz and Wang, 2017; Crowley, 2018). Studies have also focused on predictors and motives for late-life divorce. Many factors associated with divorce earlier in the lifecourse are also associated with grey divorce (see e.g. Wu and Penning, 1997; Rokach et al., 2004), while other reasons tend to be directly associated with later life, such as the risk in a long-lasting marriage to grow apart or fall out of love, but also with the particular existential conditions that characterise later life (see below).

In a seminal study by the American Association of Retired Persons (AARP) based on survey responses from divorcees aged 40–79, Montenegro (2004) found many self-reported reasons for divorce in the second half of life to be similar to those reported by people in earlier phases of life, while others were more life phase-typical. The three most commonly reported motives for divorce in the study are recognisable from general divorce research: verbal/physical/emotional abuse, different values/lifestyles and cheating. The fourth reason, ‘falling out of love’, reported by one-quarter of the respondents, is related to experiences in the later life phase of having grown apart.

Other studies corroborate how a long relationship can cause spouses to grow apart. In a study from the USA, based on interviews with 310 divorcees, Bair (2007) finds that growing apart, infidelity and abuse were among the most common reasons attributed to grey divorce. In a recent US study of 80 informants who had divorced aged 50+, Crowley (2018) found that the most common reason for divorce reported by men was that the spouses had grown apart. Other reasons were the wife's infidelity, her financial or mental health problems, or problems related to their children. The most common reason reported by women was the husband's infidelity. Other reasons were his pornography/alcohol/drug addiction, mental health problems, verbal/emotional abuse and that the spouses had grown apart. Other studies also stress the importance of growing apart for late-life divorce (see e.g. Amato and Previti, 2003; Canham et al., 2014).

Besides ‘growing apart’, it is reasonable that transitions typical for later life could impact divorce decisions, similar to how transitions earlier in the lifecourse, such as having children, can reveal new aspects in a relationship that may trigger a divorce (see e.g. Doss et al., 2009). Such late-life transitions are retirement, children leaving the household and the onset of illness.

Illness becomes increasingly likely as individuals grow older and may alter the dynamics of a relationship, by changing the ill person's personality and by turning the healthy spouse into an informal care-giver and the ill partner into a care receiver. In a longitudinal study, Wilson and Waddoups (2002) found that for spouses in late mid-life (51+) with high marital satisfaction at baseline, the risk for divorce was higher when the onset of illness caused a health mismatch between the spouses, compared to when the spouses were either both healthy or both in poor health. In another study of late mid-lifers, Karraker and Latham (2015) found that onset of illness in the wife, but not the husband, was associated with an increased divorce risk.

Children (and even grandchildren, see e.g. Brown et al., 2021) often constitute a common project for a couple to unite around. Mutual live-in children can be a reason to avoid or postpone divorce, as an intact marriage is often perceived as beneficial for children and a resource for managing the parents' everyday life. Consequently, children moving out of the household can increase the risk for divorce in mid- or later life (Montenegro, 2004; Bair, 2007; Canham et al., 2014). Retirement and the empty nest also mean that the couple have more time to spend together, something that may improve an already high-quality spousal relationship, but also increase the divorce risk if the spouses find that they no longer have anything in common.

Yet another existential condition for late-life divorce of which there is little research is the finality of life – the growing awareness that life time is slowly running out and that important life decisions cannot be put off much longer.

However, not all studies find a relationship between late-life transitions and divorce risk. Lin et al. (2018) did not find any association between an increased risk for grey divorce, and illness onset in the partner, retirement or children moving out.

This study adds to the scarcely studied field of late-life divorce. The unique contribution is the results that show how the existential conditions of the third age constitute an important context for the decision to divorce in later life.

Theoretical perspective

The purpose of this project is to investigate attributed reasons for late-life divorce, and if any life phase-typical aspects can be identified in these attributions. To investigate this issue, we first need to locate grey divorce within the larger normative change surrounding intimate relationships that has taken place in large parts of the Western world since the 1960s. Theorists of late modern intimate relationships have described this as a turn from a socio-historical context in which strong external expectations used to guide the establishment of intimate relationships (e.g. a pre-given sequence for union formation: dating–marriage–sex–common household–children) and their form (e.g. gender roles within the union) towards a context in which union formation and form are elective and intimate relationships are shaped through negotiations between relatively equal partners. Cherlin (1992) has described this change as a deinstitutionalisation of marriage, in which, for example, the roles of family members are increasingly negotiated rather than externally given. Using a different conceptualisation, Beck and Beck-Gernsheim (2002) have argued that contemporary family relationships have become highly individualised, and are increasingly negotiated to fit the individual partners' needs and wants.

Part of this change is an increased acceptance both of union dissolutions and of a multitude of ways of being intimate together. Giddens (1992) has argued that intimate relationships in late modernity are increasingly about individual self-realisation. With the concept ‘pure relationship’ he describes a contemporary relationship ideal where intimate relationships exist not out of a mutual sense of duty, but for the needs of the individual partners, and are highly malleable and easily dissolved whenever a partner feels that it no longer fits their needs or life plans. Similarly, Illouz (2018) describes how we increasingly have come to treat partners, ourselves including, as market commodities that can be chosen and unchosen, and how relationships as a consequence have become more fragile and open to dissolution (or ‘unloving’) when they no longer satisfy our needs. Increasing divorce rates are perhaps the most visible expression of this new view on relationships. Hackstaff (1999) sums up the normative change as a move from a ‘culture of marriage’ in which marriage was the only socially accepted form for intimate relationships, highly regulated and difficult to dissolve, to a ‘culture of divorce’ where marriage is one union form among many, highly negotiable, and easily dissolved.

Today's older people have to a large extent been part of this development, belonging to generations who were young adults or mid-lifers in the 1960s and 1970s and who have lived most of their lives in this new normative landscape. One consequence is that they can be thought to embrace late modern norms on divorce and marriage. Another consequence is that many older people have complex relationship histories and may currently be living in relationships of shorter duration, with fewer common resources binding them together (such as children, property and memories) in times of relationship stress, e.g. in the face of diverging life goals or ill-health in a partner.

Our investigation into life phase-typical reasons for late-life divorce also revealed the need for a theory about the existential conditions that characterise later life. Such theories constitute a core of social gerontology. Most well-known (and criticised) is Cumming and Henry's (1961) disengagement theory which postulated that society and older people both disengage from each other in a functional preparation for the individual's demise. According to the theory, a central driving force for the individual's disengagement was an ‘awareness of finitude’ as the individual approaches death, inciting a withdrawal from social roles and relationships. Both Tornstam (2005) and Carstensen (1992; Carstensen et al., 1999) have developed this idea in their respective theoretical works. In the theory of Gerotranscendence, Tornstam argues that as they age, people tend increasingly to shun superficial human relationships and focus on those that are truly important. In the theory of Socio-emotional Selectivity, Carstensen further clarifies the functional aspect of this withdrawal. Superficial relationships are important earlier in the lifecourse as a way of gaining knowledge about oneself and the world. Towards the end of life, it instead makes sense to be selective and invest in close relationships that foster emotional comfort and stability.

As our results below will show, divorce in later life can partly be understood in terms of an awareness of finality, however, seldom as a withdrawal in preparation of death, but rather as a way of making possible a new future and realising outstanding life projects. The developmental theories presented above are primarily concerned with the final part of life associated with a decline in health – often referred to as the fourth age. This is not the stage of life where late-life divorce is primarily increasing. One reason might be that a divorce is perceived as ‘too late’ and ‘too cumbersome to be worth it’, if it does not hold the promise of a new future.

A complementary perspective that fits our purpose is Laslett's (1991) theory of the third age. Laslett argues that the increase in welfare in large parts of the industrialised world over the last century, with improved health, increased average lifespan and improved individual economy, has created a new life phase in-between the (re)productive second age, characterised by work and raising children, and the declining health, functional incapacity and dependency of the fourth age. This third age (which Laslett refers to as ‘the crown of life’; Laslett, 1991: 78) offers opportunities for realising new or outstanding life goals. Gilleard and Higgs (2000, 2011) connect the third age to the rise of consumer culture after the Second World War and its focus on individual self-fulfilment in a market context. As described in late modern family theory (above), in this project a partner can be an important companion but also an obstacle if life plans diverge or if the partner is no longer fit to participate (see also Brothers and de Jong Gierveld, 2011; Bildtgård and Öberg, 2017). In contemporary European Union societies, this life-period in relatively good health constitutes roughly 10–15 years after retirement (at 65 years) (Eurostat, 2020).

Tying together these two existential aspects of the third age – increased time for self-realisation and an awareness of finitude – and basing themselves on Heidegger's (2008 [1927]) argument that being is time, Bildtgård and Öberg (2017: 159) have argued that the existential structure of the third age can be characterised as a paradox of time. On the one hand, the individual disposes of a large amount of free time after having left working life and their children having moved out – time that can be invested in outstanding or new life projects. On the other hand, remaining life time is running out (as demonstrated by people in the individual's surroundings who attract chronic illnesses or die), which results in an awareness of life's finitude and a heightened sense of urgency with regards to life projects. Within this existential interpretation, divorce can be understood as a way of freeing oneself to be able to realise the potentials of the third age. Below we use this theoretical insight to analyse how people reason about late-life divorce.

Method

This article is part of a larger project aiming to study the lived experience of late-life divorce. To capture this experience the project utilised qualitative lifecourse interviews. Participants for the project were recruited nation-wide in Sweden using advertisements on Facebook and calls for informants in newspapers, retirement magazines and senior radio. Self-recruitment was used to ensure voluntary participation. Participants were informed about the project in several steps of the recruitment process. Written consent was obtained before the interview. People who volunteered for the study were put on a list and the researchers purposefully selected interviewees to fit a sampling frame based on the following pre-selected inclusion criteria, to allow a broad exploration into the lived experience of late-life divorce:

1. (1) Informants should be divorced aged 60+ from a cross-gender union. The age limit 50+, used in several earlier studies on ‘grey divorce’ (see e.g. Brown and Lin, 2012), was deemed too low to capture transitions typical for later life, such as retirement.

2. (2) Since divorce is a highly gendered experience (see e.g. Kalmijn and Poortman, 2006), both male and female informants were included.

3. (3) Participants were chosen who had divorced between one and eight years before the interview. In order to capture experiences of ‘being’ divorced at least one year should have passed since the divorce. This was also an ethical consideration to avoid informants in acute crisis. In order for memories of the divorce to still be current and meaningful no more than eight years should have passed.

4. (4) Since prior experiences of separation can affect the experience of divorce and singlehood, and since higher-order marriages are increasingly common, the sample was constructed to include both divorcees from first and higher-order marriages. Since first marriages are likely to have lasted longer than remarriages, this criterion also guaranteed a variation in the length of the former union.

5. (5) Since the divorce experience is likely shaped by the individual's role in the divorce process, the sample included both initiators and non-initiators of divorce. However, this distinction is not always clear in the divorce narratives (also shown in previous research, see e.g. Walzer and Oles, 2003; Rokach et al., 2004), so the designation used in Table 1 should be seen as a rough guide based on the informant's interpretation.

6. (6) Since in Sweden non-marital co-habitation is in many respects indistinguishable from marriage (Kiernan, 2002) and widespread in the adult population, especially in higher-order unions (Bildtgård and Öberg, 2015a), the sample included ‘divorcees’ from both marital and non-marital co-habiting unions. Thus, in the text ‘marriages’ refer to both marital and non-marital co-habiting unions, and ‘divorcees’ also include people separated from non-marital unions.

Table 1. Sample descriptives

Notes: IP: Interview Person. Participants were born between 1937 and 1957; current age = 62–82; divorce age = 60–77; union length = 10–51 years; current union: single = 22 (59%), living apart together (LAT) = 8 (22%), co-habitation = 6 (16%), marriage = 1 (3%).

The final sample consisted of 37 participants, 13 men and 24 women, born 1937–1957, aged 62–82 at the time of the interview, and divorced at the age of 60–77 years from unions that had lasted 10–51 years, 19 from first and 18 from higher-order unions. Case-by-case descriptions are shown in Table 1.

Semi-structured qualitative interviews were carried out by the two authors (mean = 1 hour 42 minutes, range = 54–157 minutes). Interviews were organised around four main biographical themes: (a) the history of the previous marital relationship, (b) the divorce process, (c) life after divorce, and (d) expectations of the future. These main themes were followed by questions on more specific topics. The main focus for this article is on the second theme – the divorce process. In this part of the interview, the informants were first asked to tell their divorce story freely. This was followed by questions regarding the reasons for the divorce, the process of decision making, as well as emotions and practical issues surrounding the divorce. Initiators and non-initiators were all asked these questions in order to gain insights from both perspectives. To emphasise that the presented motives are not necessarily one's own or entirely understood, in the article we use the concept ‘attributed reasons’ (or simply ‘reasons’) interchangeably with motives. Even if non-initiators were not always fully aware of their partner's motives for initiating a divorce, they all had an idea what the reasons were and had no problem attributing reasons for their divorce.

All 37 interviews were transcribed verbatim and all identifiable situations and personal characteristics were de-identified. The resulting transcripts were analysed with Atlas.ti using Charmaz's (2006) principles for the construction of grounded theory. To increase the validity of the codes, all parts of the coding process was first done separately by the two authors and then discussed until consensus was reached before they were applied. Initially we read all the interviews and applied codes that closely followed the text to gain insight into the participants' own perspectives on the reasons for their divorce and to counter the impulse to read our own presuppositions into the data. Charmaz (2006) also recommends coding using gerunds that capture what is being done in the text. Using this strategy highlighted how divorce narratives are far from neutral but, for example, assign blame and legitimise decisions.

This initial coding phase identified almost 40 motives for divorce – not accounting for different wordings. Initial coding thus provided a good overview of all the reasons that people attributed to their late-life divorce. At the same time, it illustrated some of the central points raised by Allen and Hawkins (2017): decisions to divorce late in life were not generally reached easily, and the motives given for the divorce in the interviews tended to be complex and changing over time, including prior decisions not to divorce. Some motives described long-term dissatisfaction while others appeared to be more of immediate ‘triggers’. Moreover, how motives were combined and weighted in decisions to divorce differed between individual narratives. Thus, to simply count or categorise discrete divorce motives on an aggregated level would misrepresent the importance they were given in the individual narratives. Instead, we searched for a way to present the data that would preserve the integrity of the cases.

In a second phase of ‘focused’ coding (Charmaz, 2006), we went back to the individual interviews and focused on the parts of the text that concerned the attributed reasons for divorce. We focused on the meaning that the attributed reasons had in the individual divorce narratives as wholes. Were they central to the narrative or did they have more of a supporting role? The same reason could carry very different meanings in different individual narratives (e.g. infidelity could be the unique reason for divorce in a story, or simply one reason among many). In every interview we worked to reconstruct the central divorce narrative so that the most important attributed reason(s) and how they interacted became clear.

In this second phase, Charmaz (2006) encourages the use of more interpretative codes and memos to complement the initial codes. We asked how we as researchers understood the attributed reasons given by the interviewees? How could they be described in theoretical terms? Since a central objective of the project was to gain gerontologically relevant knowledge about what is distinctive for late-life divorce, one theoretical angle became apparent: when the participants reasoned about why they had divorced late in life, a range of life-phase factors were mentioned that interacted with other attributed reasons in their divorce explanations – such as viewing their ex-partner as an obstacle to achieving their goals for the third age and having the sense that time was precious and running out. This called for a theory of the existential conditions of later life. Since we were asking about reasons for late-life divorce, life-phase connections were expected but the interviews showed the many concrete ways in which the existential structure of the third age impacted the choice to divorce. These ways are an important part of our presentation of the results and constitute the article's unique contribution to research on late-life divorce.

In this phase, we also looked closer for differences in motives between groups. Two such differences were particularly evident: gendered experiences and differences between the stories of initiators and non-initiators. These differences are emphasised wherever relevant.

Finally, we compared divorce motives between interviews in order to identify similarities and differences. We aimed to create a parsimonious categorisation of narratives of late-life divorce that would include all cases with minimal overlap (e.g. it should be easy to identify to which category each case belonged). The result consists of four empirically generated types of ‘late-life divorce narratives’ that are named to capture the central divorce reason as well as the existential context of the decision: (a) incompatible goals for the third age, (b) personality change caused by age-related disease, (c) a last chance for romance, and (d) enough of inequality and abuse.

Results

Incompatible goals for the third age

Intimate relationships tend to be formed around life projects that are shared by partners (see e.g. Giddens, 1992). For most couples, the big life projects of the second age include creating a home, raising children and working, and their respective associated demands. These projects have united the partners through mutual concern or simple necessity – such as the need for a division of labour. The first type of narrative of late-life divorce that we identified was that the completion of the projects of the second age and the advent of the third age shed light on the quality of the relationship, and raised questions about the partners' future time together. Did they still share goals and visions for their future together? Was the partner that had been right for the second age also the right partner for the third age? Not always. An example was the following interviewee who described how the relationship that he had been happy with for his whole adult life suddenly lost its purpose when the projects of the second age were concluded:

I think I was bored. Our relationship didn't break down, it ended because it had no purpose. The only purpose was to live together … Earlier in life the goal of the marriage was to maintain the home. Maintain working life. Make sure the kids did what they should and got married. We had goals together. But when we had achieved them we ran out of new ideas. (Interview Person (IP) 18, man, aged 69)

The main dissatisfaction in these narratives are incompatible goals for what life should offer in the third age. The completion of the projects of the second age released a lot of time that could be filled by activities and it was not uncommon for the spouses (at least one) to realise that they had grown apart. A very common complaint in the interviews was that the former partner had become either too idle (a tired, home-oriented ‘couch-potato’) or too active (hyperactive). This concern appeared to be gendered. Women often considered themselves more active than their husbands, although this was not an absolute rule. IP24 gave words to the frustration of many women in our data:

If there is no woman around to nag a tired old man, nothing happens. I wanted to meet people and be social and see my kids, but he said: ‘Can't we just stay at home, the two of us?’ (IP24, woman, 62)

Her husband's retirement shed light on the situation and made it unbearable:

I thought after retirement perhaps he would take on smaller jobs, but every offer he got was ‘too far away’, or ‘too cold’ or ‘too hot’. When I got back from work I asked him what he had been doing and it wasn't much. I can't stand idleness and I got increasingly annoyed … I think wasting your life like that is an affront to the creator. (IP24)

A central focus of these narratives was that the partners had grown apart and developed different interests and tempos over time, but that this had previously either been concealed by other commitments during the second age, or simply put off because a separation would have been impractical given the partners' mutual dependence in their everyday lives. Many informants described how they had hesitated to divorce before their children moved out of the household, or had put it off because they were consumed by work. IP19 explained how her goals for later life became clearer as the duties of the second age were peeled off, which raised existential questions about how the remains of her life should be spent:

We had no common interests. He liked watching television (TV). I never watched TV. He read no books. He didn't like going to the theatre. As long as the kids were small and growing up it was fine. Then it got empty … It would have been horrible [to remain married] after retirement. As long as you worked and had kids you had no life plan. The life plan becomes clearer the older you get. You have to try and make the most of what you have left. You didn't think about that before, because you felt forever young … When you realise that you are mortal, you know what you want from the time you have left. (IP19, woman, 69)

IP19 stresses how the existential conditions change from the second age to the third, freeing time to be annoyed over personal differences that had previously been concealed by work and child rearing. While some interviewees had been dissatisfied with their relationships for quite some time before the divorce, for others retirement served as a sudden eye-opener to differences between themselves and their partner. IP26, who had spent most of her adult life participating in her husband's project of refurbishing an old house, suddenly realised that her husband had changed and that the house project would never be finished:

When I retired my husband had already been retired for four years. And not a thing had happened in the house. He sat by a table doing crosswords for four years. [My own] retirement opened my eyes. As long as I worked I didn't notice. I didn't see him sitting there doing his crosswords, but when I saw it and could no longer escape to work, I realised that this is not viable. You have this image of how the future is going to be and that image shattered … I felt a stress. How shall I have time to do everything I want before I die? Your last years should be happy. (IP26, woman, 72)

Although her husband no longer had the energy to complete his life project he did not want to give it up, so she decided to leave him. She did not want to spend the rest of her life caught in his project. She speculates that there are different partners for different life phases, and that her former husband was simply not the right one for the third age:

He was the right man then [in the second age], but the wrong man now [in the third age]. We developed in different directions. I ran past him. Earlier in life he was in front and I came after, but with time I discovered that I'm a rather forceful woman. [For him] it took more and more time. (IP26)

The central theme in this type of narrative has been how the completion of the projects of the second age and the advent of the third age, with an increase in unplanned free time, sheds light on the relationship and raises questions about whether the partner that was right for the second age is the right partner also for the third age.

Personality change caused by age-related disease

The second type of narrative concerns how divorce is triggered by one partner's ill-health. This type of narrative is connected to the first type of narrative above, in as much as it relates to the third age as a space for self-realisation. It differs in that the reason for this change is clearly identified as caused by a severe illness. A partner's ill-health (or one's own) can become an obstacle to achieving late-life projects. Our informants primarily problematised ill-health in terms of personality change. Rather than slowly growing apart, the problem of ill-health is about one partner becoming a ‘different person’ in a short space of time, and that the relationship suffers as a consequence. Three personality changes were often mentioned: becoming less energetic, becoming more self-centred or becoming mean-spirited.

Leaving a partner in ill-health was difficult and often put off for a long time. In several of our cases the informants' spouses had been ill for some time before the actual divorce took place. Other life-phase factors typically influenced the decision. If the partner was already ill, divorce was often initiated after retirement, when the partners spent more time together and it became evident how the illness affected the promise of the third age. It was common to refer to the insight that life is finite and remaining time too precious to be spent with a partner who was no longer the person with whom they once fell in love.

A common personality change was that of a partner becoming less energetic due to a disease and unable to participate in the projects of the third age. There were many stories in our interviews of heart attacks, strokes, cancer, late-life burnouts, etc., that changed one of the partners who no longer had the energy to participate in their other half's plans for later life. One female informant described how her ex-spouse was changed by a heart attack. He became cautious and withdrew from physical activities such as sex and excursions:

After the heart attack he withdrew … Our sex life basically ceased … It became increasingly difficult to get him to participate [in activities]. He became the king of couch-potatoes …. He didn't notice me anymore. If I cooked a good dinner and set the table nicely he would crawl out of the sofa and eat, only to go back out and watch sports again. (IP5, woman, 64)

The informant, who was active and in good health, felt that her husband's condition constrained her life and eventually opted for divorce. Another disease-driven personality change was introversion. Some informants described how their partner became increasingly introvert after a disease process, stopped noticing them and stopped being a source of emotional support. IP2 described how her husband went into a severe depression a few years before retiring:

He went into a bubble and remained there for three years. I tried to comfort him but he pushed me away and kept more and more to himself … He stopped noticing me. Although I washed his clothes and took care of the home and cooked food … We never recovered. (IP2, woman, 65)

She hoped the situation would get better after they retired. Instead it got worse. The husband spent his time in the garage while she took care of the home. They hardly met and ‘love died’. She finally decided that the situation could not be allowed to continue. Awareness of life's finality played an important role in her decision to divorce:

He was miserable and I was miserable. I woke up sad and frustrated every day. I was 64 years old and I suddenly realised that the average age of a Swedish woman is 84 years. Do I really want to spend another 20 years like this? No! (IP2)

Another interviewee describes how she left her ex-husband after he relapsed into mano-depressive disorder. Similar to IP2, she argued that time was too precious to spend taking care of a sick husband:

You only have this one life and only I can decide about it. You realise you don't have so many years left. It is now or never … You know that things can happen in the blink of an eye. Not that I should die tomorrow, but you are aware of the possibility. (IP33, woman, 82)

A few interviewees also described how their spouse had become mean-spirited as a consequence of a disease process. For some, it was an exacerbation of a previous bad-tempered personality, while for others it was a total personality transformation. The most extreme example was related by an older male informant who had lived in a happy marriage for 48 years, only to find his wife growing increasingly mean-spirited and violent as part of the onset of dementia:

When I went to bed she had poured salt between the sheets. Another evening when I walked up the stairs she poured water over me … Doors were slammed shut and glasses thrown. I tried not to care and see that it was a disease that had taken hold of her, but I was afraid that she would kill me. You never know what could happen when you are asleep. (IP3, man, 81)

After one violent outbreak the police intervened and encouraged the man to divorce and move out. Shortly after this event he filed for divorce.

Central to the ill-health narrative is that the explanation for the divorce lies in a disease-driven personality change in one of the spouses, which makes that person an obstacle for living the promises of the third age. Also, an awareness that life time is finite and that only a few good years remained was important in these decisions.

A last chance for romance

The third identified type of narrative focuses on how the existential insight that life is running out can prompt a person to leave their union for a new intimate relationship. In our data this was a highly gendered phenomenon – it was mainly men who had divorced their partner for somebody else (although many female informants found a new partner after being divorced). For the men who left their wives, a new intimate relationship was often narrated as a ‘last chance’ for romance. For the women who were left by their men for somebody else, the same event was instead described as prompted by a male ‘age crisis’. In both cases divorce was understood in terms of finality – the insight that time was running out prompted a drive for change.

For some of the men who had divorced their partner, the concern was that the partner did not share their ambitions regarding intimacy in later life. They felt their ex-spouses were not loving enough and they feared being trapped in a loveless relationship. The third age offered a final opportunity for finding love. IP32 was dissatisfied with his wife's harsh attitude and unwillingness to talk about emotions, and saw a chance for a loving relationship when a close female friend's husband passed away. The growing awareness that life is not eternal and that the remaining good years are numbered made him willing to take a chance on a relationship with the widow, and he filed for a divorce soon thereafter:

[I wondered] how long will I live and what should I do with all the time that I have, now that I'm retired? I wanted a warm relationship. It was a mature decision. There was more time when you no longer had to think about work. When you had kids at home and worked, life just went on. And when you are 80 I don't think you have the energy for a divorce, so it was now or never. (IP32, man, 76)

In all these stories, retirement was an important transition, which could make long-term unhappiness unbearable when more time was spent together, or that symbolised a ‘last chance’ for achieving change. IP13 described how he grew increasingly tired of what he experienced as a lack of emotional commitment in his marriage:

She wasn't very cuddly. I thought the relationship lacked in conversation and emotional openness. It wasn't good and our intimate life wasn't much to brag about. She could be very foul-mouthed and tell raunchy stories but she wasn't very interested in sex. (IP13, man, 74)

In preparation for retirement he actively looked for a more romantically oriented partner with whom to spend his later years. Eventually he met a new partner and immediately decided to divorce his wife.

Women who had been divorced by their partner for somebody else were rather less romantic about the reasons, but still employed a narrative of finality, only expressed in terms of their ex-partner's ‘age crisis’. The age crisis is an established narrative about late-life separation, in our data told only by women to explain why their male partners left them. In this narrative, an age-related event such as retirement or disease triggered an awareness of life's finitude in the partner, followed by existential questions regarding the life lived and how remaining life time should be spent. Finally, this crisis drove the husband to a new woman who unlocked a new future for him. IP8 described how her husband went into a crisis after retirement:

He stopped working, and I remember he said: ‘I've lived my life all wrong. How could I allow work to come first? Nobody calls me anymore.’ He experienced a crisis … I was concerned but I could never have imagined that he would meet another woman … This all happened after he stopped working. I don't think that he really needed a new woman, he just needed change. A new meaning for his life. (IP8, woman, 73)

In these stories, the age crisis typically led the husband into the arms of a younger woman, who gave him a chance to start over and postpone the feeling of growing old. Some of the men started working again. IP35 described how her husband loathed growing old and how he divorced her after meeting a much younger woman:

He felt desperation that he was running out of time … He felt like getting old was a personal violation … It was horrible when he stopped working. It was a big life crisis … Our friends and neighbours talked about how much he hated, almost dreaded, getting old. And logically, the woman he met was 30 years younger than him. Younger than our daughter … He wanted an infatuation before he laid himself in his grave. (IP35, woman, 74)

In the age crisis narrative, switching partner becomes a strategy for halting the ageing process, at least postponing it. In this narrative the female informants become obstacles for their husbands' goals of rejuvenation.

The core of this type of narrative has been the existential insight that lifetime is limited, and how this insight can prompt a person (often a man) to leave their partner for a new intimate relationship that promises a new romantic future and a new life. The last type of narrative focuses inequality and abuse as a reason for late-life divorce.

Enough of inequality and abuse

The fourth type of narrative explains divorce as emancipation from an unequal relationship dominated by one partner. A common thread in these stories is that one of the spouses had one-sidedly defined relationship goals and the level of personal latitude for each partner – who was allowed to do what. A majority (but not all) of the informants telling this type of story are women who have emancipated themselves from a dominant and self-centred male partner. In difference with the first type of narrative, this type focuses on what the informants were divorcing from, rather than to. These stories also tend to include high levels of conflicts in the former marriage.

The mostly female informants who told these narratives complained that their former partners were self-centred and did not notice them or take note of their interests. In difference with our first type of narrative, this had often been a long-term frustration, but had become more bothersome over time, either because the informants had grown more aware of their partners' dominating behaviour or because the behaviour had got worse. IP20 described living with a partner that had no interest in her life goals, only his own:

He got more and more selfish [over the years]. He wanted to play golf so I started to play golf. He wanted to hunt, so I got a licence for hunting. He bought a boat and I participated, even though it wasn't my interest … When he wanted me to start fly fishing I said NO … He got cancer, twice, and became even more egocentric after that. There was so much he had to do in life – a bucket list. He never even tried to share any of my interests … I thought, can't he ever come with me to the theatre or to a concert? Go to the cinema? Never! He was uninterested in me as a person and in what I wanted. (IP20, woman, 68)

After her retirement, the interviewee decided to divorce her husband. The feeling of not really being seen as an equal partner that had to be considered in decisions was common among these female divorcees. Like IP20 above, IP23 had been frustrated with her husband's disrespectful attitude towards her for many years. She recalls an episode that was symptomatic of her husband's egoism and which led to her decision to file for divorce:

They planned a trip for my 60th birthday. I said: ‘I'm sorry but I can't go because I'm unemployed, we'll have to wait.’ He [the husband] responded: ‘You don't have to come with us [on your birthday trip].’ The next day I said: ‘I've figured out what I want for my 60th birthday.’ ‘What?’, he said. ‘Divorce’, I responded. I thought, ‘perhaps I'll live 10–15 more years’. You start to panic when the years add up. If I shall have time to do something I have to act now. I saw 60 as a crossroads. (IP23, woman, 68)

For all the informants above and for most of the women telling this type of narrative, thoughts of divorce had been present for a long time, but had been postponed due to co-residing children. The informants often argued that they did not want to be the cause of a ‘broken home’ for their children. It was only when the children had moved out that they felt free to divorce.

Many women also described how the power balance between the partners had changed over time during a long marriage. Life phase-typical transitions played an important role in this change. Female informants described how diminished physical and mental resources on their ex-husbands' part, sometimes coupled with a weaker connection to the labour market, meant that their partners had increasingly lost control over them. In return, the women gained independence as their children grew up and moved out. This transition allowed the women to be more active outside the household, working more and/or increasing their social activities. In some cases, the men reacted by lashing out more than before.

It was common among the female informants to have been in abusive relationships and some had experienced physical violence, verbal threats and harassment. IP25 described how her ex-husband repeatedly abused her for many years of their marriage, sometimes in front of their children:

He held a bottle over my head and pulled my hair so that tufts of hair fell off. My daughter was only 13 at the time, she should not have had to experience that. … A couple of years later he got furious with my son. My daughter reacted and shouted at him. He pushed her on to the floor and held a lit candle from the dining table over her head as a threat. The kids ran up to a neighbour and called the police. (IP25, woman, 73)

She eventually managed to wiggle herself out of the relationship when the husband left to work abroad. However, it was only many years later when their children had grown up and she had established herself on the labour market that she felt free to file for divorce. The story from IP25 demonstrates how entering a new life phase can be important to allow emancipation from a dominant partner. Retirement could also be a trigger for divorce, by making an unbearable home situation even worse:

I've been thinking about divorce for a long time. But when the kids were young I couldn't imagine leaving the kids with their dad … It got worse with both verbal and material violence when I retired, perhaps because I was more at home … He could come in [to the bedroom] and shout at me when I was sleeping … I never knew what was going to happen at home. (IP12, woman, 68)

A few months into her retirement the informant fled her home to a safe house with the help of a social assistant. Later she filed for divorce.

This fourth type of narrative focused on inequality and abuse as explanations for divorce in later life. The female informants who told these stories emphasised the importance of children leaving home to make it possible for them to become independent and break up from their marital relationship. Retirement was another significant transition that could make an already difficult situation unbearable. Finally, the realisation that life time was running out and that if they were going to divorce it was now or never was expressed by some of these interviewees, although this existential condition was less pronounced than in the other three narrative types.

Discussion

The purpose of this study has been to inquire into the reasons late-life divorcees attribute to their divorce, and if any life phase-typical aspects can be identified in these attributions. We have shown how the existential conditions of the third age constitute an important context for the decision to divorce in later life. Most of the motives for divorce that have been identified in prior studies of divorce generally (see e.g. Amato, 2010; Lyngstad and Jalovaara, 2010) and of late-life divorce (see e.g. Montenegro, 2004; Bair, 2007; Crowley, 2018) were present also in the stories of our interviewees. Traditional ‘at fault’ reasons for divorce, such as abuse, addiction or unfaithfulness, are common in the interviews, but it is more common for the interviewees to attribute their divorces to no-fault motives. They still often blame their partners for their divorce, but primarily for not matching them in energy or interests and thus no longer contributing to their life plans. This is line with prior research that has shown that ‘growing apart’ is a common reason for late-life divorce. It also matches the predictions made by theoreticians such as Giddens, Hackstaff and Illouz, that the basis for contemporary intimate relationships tend to be self-realisation, and that relationships that no longer contribute to the individual's life plans can be legitimately ended. This is clearly the case also for late-life divorce.

The most important difference between our results and those of previous research on grey divorce is the weight that our interviewees gave the existential context of the third age for their divorces. As Bildtgård and Öberg (2015b) argue, in a Heideggerian perspective being can be understood as time and the third age as a historical institution characterised by the ‘paradox of time’ – of having lots of available free time after the (re)productive duties of employment and raising children, but also of having a limited amount of (healthy) life time left. This combination means that the third age offers an opportunity for the individual to realise outstanding or new life goals, but also a sense of urgency in doing so, springing from a growing awareness of finitude. For our interviewees, this existential context clearly confers meaning and urgency on to their other reasons for divorce (reasons often mentioned also in earlier scholarship). Retirement and the empty nest raised questions about the future of the marital relationship. While for couples who stay married more time together after retirement and the ‘empty nest’ might strengthen the relationship (Bildtgård and Öberg, 2015b), for many of the older interviewees who had divorced it instead illustrated diverging aspirations for their time together, and increased existing grievances that undermined the relationship. The end of the second age projects around children and work also increased the autonomy of the partners and facilitated divorce decisions, as shown in many of the stories above. The onset of illness could change a spouse's personality, so that he or she becomes an obstacle rather than a resource in achieving the promises of the third age. Finally, a growing awareness of the finitude of life can confer a sense of urgency to the choice of divorce – it is now or never.

The importance of the third age as a coherent existential context for late-life divorce has not been given much attention in prior research, although some of the aspects that we find to be important have been stressed by e.g. Bair (2007) and investigated by e.g. Lin et al. (2018). Considering the importance that our interviewees give to the life-phase context in their divorce attributions, it is notable that Lin et al. (2018), in their longitudinal survey of antecedents of grey divorce, found no significant evidence for the role of the empty nest, retirement or illness onset. It is possible that the difference is due to some methodological problem in the respective studies. However, a more plausible explanation is of an epistemological nature.

Motives and predictors for divorce do not always overlap. What an individual perceives to be important for their decision to divorce is not necessarily directly translatable into a measurable predictor for divorce. As pointed out by Hopper (1993), motives are rhetorical in nature and help to legitimise and give weight to decisions, both in relation to one-self and others. The importance of the existential context of the third age may not always be that it directly triggers divorce, but rather that it confers meaning and urgency to that decision. This would translate as a system of interactions that would be difficult to capture with quantitative data.

Finally, the absence of an expected correlation does not necessarily imply absence of causality. The late-life transitions that were found important in our study (retirement, illness onset, empty nest, increased awareness of finality) may work in two opposite ways: it may strengthen high-quality relationships (lowering divorce risk), while simultaneously acting as a trigger for uncoupling in low-quality relationships (increasing divorce risk). In a quantitative analysis they could take each other out. In any case, it is impossible to disregard from the importance that our older divorcees attribute to the existential context in their explanations for their divorce.

Limitations

A common weakness of studies based on self-recruitment, like this study, is that the educated middle class tend to be more prone to volunteer. Although we have used a sampling framework to guarantee participation from both genders and from both initiators and non-initiators, generalisations should be made with caution. The identified narratives are arguably shared by many grey divorcees also outside the sample, but it is possible that some narratives are not represented in the data, because some groups have not volunteered for the study, perhaps because they are not interested in participating in research or because their stories are not socially acceptable. For example, we have no stories of divorce because of marital abuse from the perpetrator's perspective.