Introduction
In the 20 years since the publication of my book Widowhood in Later Life (1991), significant changes have occurred in how widowhood is experienced in Canada, especially at a population level, and in how it is studied and understood. Widowhood has changed both as a status in Canadian society and as a process experienced by individuals. Our understanding of variability in the experience of widowhood – in terms of gender, or socioeconomic or ethnocultural diversity – has also deepened over time.
The status of widowhood is given to an individual who has not remarried following the death of his or her spouse (Martin-Matthews & Davidson, Reference Martin-Matthews, Davidson and Birren2005). Widowhood is also a process of transition, progressing from the illness to the death of the spouse and related events involving burial and mourning, grieving, and reconstruction of one’s social world that follow. This article begins with a focus on how the demographic profile of widowhood has changed over two decades, and how conceptual and methodological approaches have brought new insights to the complexities and variations in the experience of widowhood. It then considers the current state of knowledge relating to the three broad themes that framed the analysis in Widowhood in Later Life: adaptation, social support, and variability.
The discussion of adaptation includes issues of widowhood as a relative life event, forewarning of and socialization to the role of widowed person, and grief and coping mechanisms in bereavement. The brief examination of social support reviews literature on patterns of support, especially familial and financial supports and resources. Widowhood in Later Life dealt with variability between women and men, rural and urban widowed individuals, older and younger widowed persons, and between widowed people with and without children. Marital status comparisons were a focus of the book, but are not in this article. Reflecting recent trends in the literature to focus more on variability within widowhood, rather than compare across marital statuses, here I consider two primary aspects of variability: gender and ethnocultural diversity. Throughout, the findings of Canadian research are emphasized,Footnote 1 but are placed in the context of the international literature.
1. Understanding Widowhood
The Changing Demographic Profile of Widowhood
The number of widowed people in the population is increasing, but the proportion of elderly people who are widowed is decreasing, as is the duration of widowhood in later life. How do we explain these patterns, and their apparent contradiction?
In many countries, the proportion of the population aged 15 and over who are widowed is approximately six per cent. But widowhood is both sex-selective and age-related. Half of all marriages end with the death of the husband, but only one fifth with the death of the wife. Thus, widowhood is primarily associated with women – old women. Almost three quarters of the widows in North America are over the age of 65. The average age of the widowed population is 75 years, for both men and women. In North America, nearly half of women aged 65 and over are widowed, with this proportion increasing to four fifths of those aged 85 and over. In contrast, less than one in five men is widowed at 65, although almost half of men aged 85 and over are widowed.
As the baby boom population (born between 1946 and 1961) crosses the threshold of age 65 over the next 15 years and moves through later life, their sheer numbers dictate that a larger number of widowed people will be in the population. But widowhood as an expectable life event and characteristic of old age (especially for women) has been in decline for quite some time (Martin Matthews, Reference Martin Matthews and Marshall1980, Reference Martin Matthews and Marshall1987a, Reference Martin Matthews1991; Martin-Matthews, Reference Martin-Matthews and Neysmith1999; Martin-Matthews & Davidson, Reference Martin-Matthews, Davidson and Birren2005), due to medical advances which have delayed it, and due to societal trends such as increasing proportions of people entering old age as divorced, or single, people.
Connidis (Reference Connidis2010) provided a thorough analysis of trends and changes in patterns of widowhood for both men and women (and, indeed, for all marital statuses) in later life, examining data from Canada and the US over a 35-year period. Her analysis characterized the long-term decline in the likelihood of widowhood for both men and women, its occurrence later in old age, the increasing proportion of older people who live out longer periods of later life as members of a married couple, and the impact of divorce. These changes have, in the words of Allen and Walker (Reference Allen, Walker, Calastani and Slevin2006), “transformed the intimate landscape for old people” (p. 160).
To complement the detailed analysis presented by Connidis (and which is beyond the scope of this article), Figure 1 provides a graphic illustration of changing cohort patterns over time in the prevalence of widowhood among women in Canada. It traces the trends in widowhood among women born in different decades, and illustrates the substantially decreasing likelihood, over the past 35 years, of widowhood throughout the female life course.
Although the trend of a steady increase in the average age at widowhood is true of many countries worldwide, it is not a global phenomenon. The epidemic of HIV/AIDS in sub-Saharan Africa has meant a reduction in the average age of widowhood for both men and women. However, in these circumstances widowhood is more likely to be of short duration, partly as a result of remarriage patterns among younger generations. Also, when both partners are infected, as is frequently the situation, the period of survivorship is truncated.
Conceptual Approaches to the Study of Widowhood
Over the past 20 years, the conceptual foundations of widowhood research have shifted significantly. Role theory dominated the work of Lopata (Reference Lopata1987a, Reference Lopata1987b, Reference Lopata1996), arguably the most influential and prolific of scholars of widowhood over the past 40 years. This perspective approaches the study of widowhood in terms of role loss and role exit – the cessation of any stable pattern of interaction and shared activities. In her final book, Current Widowhood: Myths and Realities, written two years after her own widowhood, Lopata (Reference Lopata1996) acknowledged the influence of more recent studies of widowhood which
helped pull away the dismal image (with which even I first approached this subject) of the ever-limited, ever-suffering, ever-dependent widow. The picture of widowhood that now emerges is much more complicated and varied than I had been able to grasp in my first study in the late 1960s. (p. xiv)
This more complex, varied, and nuanced understanding of widowhood is in many ways a function of the theoretical lenses and conceptual approaches brought to our understanding of it, moving well beyond role theory. My writing of Widowhood in Later Life was framed within a symbolic interactionist approach that considers individuals as active creators of their social worlds, focusing primarily on how older people interpret and give meaning to events and situations in their lives. From the standpoint of symbolic interactionism, social roles merely set the conditions for individual interaction, but they do not determine action. As I wrote then (Martin Matthews, Reference Martin Matthews1991), “Changes in the self derive from continual adjustments in the person’s notions of how others will respond to her actions, and from the meaning she gives to her own actions based on the earlier responses of others” (p. 10).
Recent major contributions to our understanding of widowhood in later life have similarly adopted a symbolic interactionist approach. This is most particularly exemplified by the work of Canadian researcher van den Hoonaard, in her studies of the experience of widowhood among older women (van den Hoonaard, Reference van den Hoonaard1999, Reference van den Hoonaard2001), comparisons between men and women (van den Hoonaard, Reference van den Hoonaard2002a, Reference van den Hoonaard2002b, Reference van den Hoonaard2009a), and a focused analysis of older men (van den Hoonaard, Reference van den Hoonaard2010).
U.K. researcher Chambers (Reference Chambers2005) and Australian researcher Foxman-Feldman (Reference Foxman-Feldman2005) have similarly adopted conceptual approaches that focus not only on pathologies and decrements of widowhood, but also on the context of past and current life events and behaviours in shaping the experience of widowhood. Chambers has approached the study of widowhood through more critical gerontology and feminist life course perspectives, moving away from assumptions of the link between observed behaviours and perspectives and of widowhood alone as an explanatory factor.
Chambers challenges the public discourse that problematizes the experience of widowhood and casts it only in terms of negative consequences; she considers a complexity of experience rooted in personal biography and the female life course, dispelling many myths of widowhood in the process (Chambers, Reference Chambers, Bernard, Phillips, Machin and Harding Davies2000). She has argued that, without the conceptual approach of a life course perspective and a “multiple narratives” focus, “it is all too easy … to misconstrue and misinterpret older widows’ current needs and aspirations” (2005, p. 264). Her work challenges prevailing images of widowhood as an experience isolated from the rest of women’s lives; of older widows as a homogeneous group; and of older widows as without agency and self-determination (Chambers, Reference Chambers, Bernard, Phillips, Machin and Harding Davies2000).
Van den Hoonaard (Reference van den Hoonaard2001, Reference van den Hoonaard2010) has similarly moved beyond a “problematizing” approach to widowhood, focusing on the ways in which the experience of widowhood intersects with class, gender, ethnicity, culture, religious affiliation, and environment, and, for men, with understandings of masculinity. As Connidis (Reference Connidis2010) has noted, the more critical perspectives exemplified by the work of Chambers, van den Hoonaard, and others “links the impact of a partner’s death on individuals to socially structured practices regarding marital status, gender and age” (p. 107).
The emergence of “resilience” as a guiding concept is comparatively new in studies of widowhood (O’Rourke, Reference O’Rourke2004). Resilience has been defined as “the ability to maintain a stable [psychological] equilibrium following loss, without long-term consequences” (Bonanno, Wortman, & Nesse, Reference Bonanno, Wortman and Nesse2004). Four models of behaviour have been associated with resilience in widowhood: (a) reorganization, (b) adaptation, (c) finding positive benefit, and (d) compensation (Moore & Stratton, Reference Moore and Stratton2002). Intrigued by the potential utility of the concept, Bennett (Reference Bennett2010a) re-examined 60 of her interviews with widowers to identify those who were resilient and those not, and to consider the processes involved in achieving resilience. She identified resilience among 38 per cent of the men in her studies, with her synthesis of the literature emphasizing the role of resilience in initial responses to loss of spouse (in bereavement) and longer-term reconstruction of life without the spouse (in widowhood).
Although this article focuses on widowhood in later life and not on bereavement per se, the two fields of enquiry are clearly relevant to one another. Understandings of the experience of bereavement and grief have changed as well in recent decades. It is now well recognized that “doing well” after a loss is not necessarily a cause for concern but rather a normal response for many older adults (Boerner, Wortman, & Bonanno, Reference Boerner, Wortman and Bonanno2005). Redefinition of understandings of grief also reflect a “shift in focus from the end of the living bond to the place of the inner representation of the dead or absent person in the inner world of the survivor” (Machin, Reference Machin, Bernard, Phillips, Machin and Harding Davies2000, p. 113). This focus, described by Machin (Reference Machin, Bernard, Phillips, Machin and Harding Davies2000) as more than a revisiting of memories, is rather “a fuller integration of the person of the deceased and the meanings held in the relationship with the deceased in the present and for the future” (p. 117). This perspective is reminiscent of Unruh’s observation that “most important for the study of social worlds in aging lives is the idea that social worlds may be meaningful even though they do not exist in any concrete sense” (Unruh, Reference Unruh1983, as cited in Martin Matthews, Reference Martin Matthews1991, p. 35).
Researching the Widowhood Experience
Two decades ago, I urged that “highest priority … be given to the examination of widowhood as a process of transition, rather than merely as a status held by most older women. There has been relatively little research on the way in which social relationships and support networks influence and are themselves changed by the process of becoming widowed” (Martin Matthews, Reference Martin Matthews1991, p. 113). Particularly in relation to widowhood, I decried the “persistent habit” (characteristic of my own work as well) of “studying role transitions after the transitional event”, and called for prospective and longitudinal studies to enhance understanding of relationships between illness, caregiving, dying, and the transition to widowhood.
There have been a number of noteworthy developments in the literature, with two highlighted here. The first is the increasing importance of prospective research designs and, in particular, of comparisons between the pre- and post-widowhood lives of individuals. Researchers such as Wortman and colleagues (Boerner et al., Reference Boerner, Wortman and Bonanno2005; Bonanno et al., Reference Bonanno, Wortman and Nesse2004; Carnelley, Wortman, Bolger, & Burke, Reference Carnelly, Wortman, Bolger and Burke2006) have written for years using such prospective data sets as the Changing Lives of Older Couples (CLOC), a prospective study of 1,532 married individuals aged 65 years and older, which enables the analysis of the changing life circumstances of study participants as they are widowed. For example, a recent study used CLOC data in examining the extent to which positive and negative support from children prior to and after spousal loss, and changes in support from pre- to post-loss, affect widowed older adults’ depressive symptoms, anxiety, and anger 18 months following widowhood (Ha, Reference Ha2009).
Currently, this type of analytical capability in Canada is restricted only to specific types of analyses (such as Statistics Canada’s Longitudinal Administrative Databank from 1993 onward, which allows for comparisons of the dynamics of income during widowhood among women and men aged 65 and over) (Bernard & Li, Reference Bernard and Li2006). However, the launch in 2009 of the Canadian Longitudinal Study on Aging (CLSA), which will track 50,000 Canadians between the ages of 45 and 85 for a 20-year period, promises such opportunity here as well (Martin-Matthews & Mealing, Reference Martin-Matthews and Mealing2009; Raina et al., Reference Raina, Wolfson, Kirkland, Griffith, Oremus and Patterson2009).
The second trend has been noted by Connidis (Reference Connidis2010) in her observation that: “Much of our understanding of becoming and being widowed comes from qualitative research in which in-depth interviews and narratives reveal the process and dynamics of this major life transition” (p. 108). Especially relevant is the emergence not just of more qualitative research, but of research that has focused on penetrating aspects of the experience of widowhood to examine issues of diversity and heterogeneity that were previously unexplored. For example, Canadian research by van den Hoonaard has enhanced our understanding of variability in men’s experiences of widowhood in later life. Her focus on broader issues of masculinity among older widowed men frames their experience of widowhood and old age in ways that extend well beyond widowhood exclusively. Van den Hoonaard’s approach is inductive, drawing exclusively on the ways in which men speak of widowerhood and the social meaning of the experience as reflected in the ways in which they talk about it (van den Hoonaard, Reference van den Hoonaard2010, p. 1786). This approach was also used in Foxman-Feldman’s (Reference Foxman-Feldman2005) focus groups with widowed women, Chambers’ (2005) research with widowed women, and Bennett’s (Reference Bennett2010a, Reference Bennett2010b) and Davidson’s (Reference Davidson1998, Reference Davidson2001, Reference Davidson2002) studies of widows and widowers.
Van den Hoonaard (Reference van den Hoonaard2009b) also provides an interesting methodological insight in her writing about an opportunity she had to observe the participants in her study
when they were not talking to me or to each other about being widows. No one who met them on that day would have thought about them as grieving, struggling widows. … Transcending a single aspect of their lives gave me a much fuller picture than … if I had only seen them when they were talking about their lives as widows, in an interview situation. (van den Hoonaard, Reference van den Hoonaard2009b, p. 1786)
At a time when the application of research findings to the “real world” is being increasingly promoted by research funding agencies (through initiatives such as “research to action” and “knowledge translation”), one qualitative study has been quite innovative in “giving voice” to widows’ accounts of their experiences. Foxman-Feldman’s in-depth qualitative study of widows in Australia was adapted by Alan Hopgood into a highly successful stage production “Wicked Widows”. The theatrical adaptation uses the words of three widows of different ages and with varied experience to convey “a positive message, appreciation and acceptance of their loss and a wicked comment on widowhood” (Health Play, n.d., p. 1). New methodologies are therefore not only being used by researchers to enhance our understanding of widowhood in later life, but also to present that knowledge to the public in ways that are accessible and impactful. Indeed, Feldman, Radermacher, and Lorains (in press) have taken this a step further, subsequently examining the use of a theatre performance about widowhood as a medium for facilitating older people’s engagement with key life events and countering and challenging negative stereotypes.
Thus, the profile of widowhood and the landscape of widowhood research have both changed substantially in 20 years: widowhood itself (largely) of later onset and shorter duration, less “expectable” a characteristic of much of old age; research that is framed by critical, feminist, and life course perspectives that recognize the widowhood experience itself as but one of multiple “narratives” shaping the experience of this stage of life, both for women and for men; and methodologies that range from in-depth qualitative studies of widowed individuals in a wide variety of contexts (and that penetrate the very nature of that variability) to prospective studies with large or nationally representative populations; studies that follow populations of widowed persons for longer periods of time after the loss of spouse; that re-evaluate data in light of new contexts and approaches (Bennett, Reference Bennett2010a); employ multi-method approaches (Feldman, Byles, & Beaumont, Reference Feldman, Byles and Beaumont2000); new data collection strategies, such as the Internet (O’Rourke, Reference O’Rourke2004); and longitudinal research designs intended to capture a broad spectrum of transitions and trajectories.
2. Becoming Widowed: Adaptation
Adaptation to widowhood is shaped by numerous factors, which include its relative expectability, in terms of forewarning and whether on-time or off-time, with many implications for the reorganization of one’s social world and coping processes.
Widowhood as a Relative Life Event
Widowhood has traditionally been considered one of the most stressful of life events and role transitions. However, there are inconsistent findings as to the impact of the timing of widowhood in the life course (age at widowhood) and the duration of the spouse’s final illness on the experience of bereavement. Widowhood as “on time” or “off time” affects individuals in terms of their psychological preparedness, opportunities for anticipatory socialization, and the societal resources and supports available to them. The age at which a person experiences the death of the spouse is a very important feature of the experience because of the way in which the person’s life is embedded in other social roles at that point in time. Similarly, the duration of a spouse’s final illness also influences opportunities for anticipatory socialization to the role of widowed person. However, this relationship appears to be stronger for the case of early, off-time widowhood than for widowhood in later life.
Despite the view of widowhood as a particularly stressful life event, most individuals adapt well over time, although the process of reconstruction of a new life and social order is for some a long and painful one. In a recent study designed “to chart the time course of grief” by focusing on continued involvement with the deceased spouse, emotional resolution, meaning finding, and feelings of personal growth, Carnelly, Wortman, Bolger, and Burke (Reference Carnelly, Wortman, Bolger and Burke2006) found that widowed persons continue to talk, think, and feel emotions about their lost spouse for many years (sometimes decades) later.
Nevertheless, many individuals fare well in widowhood, underscoring the points made earlier that widowhood need not be pathologized or problematized. Comparisons across bereavement studies have revealed three basic patterns of outcome: common or time-limited disruptions in functioning (e.g., elevated depression, cognitive disorganization, health problems) lasting from several months to one to two years, chronic disruptions in functioning lasting several years or longer, and the relative absence of depression and other disruptions in functioning (Bonanno et al., Reference Bonanno, Wortman and Nesse2004). In a large prospective study, Bonanno and colleagues found that almost half of the widowed people studied showed little or no depression following the death of their spouse. Indeed, 10 per cent of those widowed showed improved mental health following the loss. While Bonanno et al. recognized the understanding of widowhood as one of life’s most stressful experiences, their findings emphasize “the importance of maintaining a healthy skepticism toward traditional assumptions in the field and lend credence to the view that we still have much to learn about the variety of ways people cope with loss” (p. 269).
For many people, widowhood represents a crisis not in the sense of a threat of catastrophe but rather as a turning point, a crucial period of increased vulnerability and yet heightened potential. This is evident in Bennett’s (Reference Bennett2010a) study of older widowers, and in numerous autobiographical accounts of widowhood (Didion, Reference Didion2007; Oates, Reference Oates2011). Many individuals realize that potential, although as they advance in age and grow increasingly frail, the threat of becoming dependent looms large in their lives. One of the real advances in the understanding of widowhood in recent decades has been the recognition of the important role of health, and especially of health declines in later life, in shaping how widowhood is experienced (Chambers, Reference Chambers2005). Davidson (Reference Davidson2001) found that extrinsic factors of older age and poor health are highly salient for elderly widowers’ choices and constraints concerning re-partnering.
Although there is evidence that widowhood is associated with an elevated risk of morbidity and mortality, especially among those who are widowed more than once (Zhang, Reference Zhang2006), research on the relationship between widowhood and morbidity and mortality is fraught with contradictory findings. Short-term decreases in health status and perceived health following widowhood have been observed, but long-term health appears largely unaffected. Despite early research findings of high mortality rates among widowers, more current epidemiological studies (Jin & Christakis, Reference Jin and Christakis2009) have found no significant relationship between bereavement and mortality.
Socialization to the Role of Widowed Person
As Lopata (Reference Lopata1996, p. 212) observed, and subsequent qualitative studies attest, there exists no institutionalized role of widowhood, only a pervasive identity of widow that alters relationships. What, then, are the factors that influence the transition to what used to be considered this “roleless role”? As Connidis (Reference Connidis2010) has noted, the caregiving experience that often precedes the death of a spouse has implications for the transition to widowhood. Caring for a spouse during his or her final illness allows some widowed persons to anticipate their loss and to feel some relief in the death (Hansson & Stroebe, Reference Hansson and Stroebe2007). In accounts of their experiences of widowerhood, many elderly men acknowledge how periods of caring for their ill wives provide opportunities to acquire domestic skills that ease their transition to widowhood (van den Hoonaard, Reference van den Hoonaard2010), which often does not happen when a spouse dies suddenly (Moore & Stratton, Reference Moore and Stratton2002).
The anticipation of widowhood, as when a spouse is ill, facilitates adaptation (Martin Matthews, Reference Martin Matthews1991). This was found in studies 20 years ago, and remains true today. Among women in particular, lengthy illness enables conversation about the dying partner’s death and the future life of the surviving spouse (van den Hoonaard, Reference van den Hoonaard2001). There is far less evidence that men do this (Moore & Stratton, Reference Moore and Stratton2002; van den Hoonaard, Reference van den Hoonaard2010). Connidis (Reference Connidis2010) noted that “following widowhood, these exchanges are enmeshed with the experience of being widowed more than of being married; a husband’s expressed views of her future are part of a widow’s present” (p. 110). Bennett (Reference Bennett2010b) extended this even further, in her examination of how identities of older widowed women remain strongly associated with their husbands and with their identity as a wife: “At the heart of their experiences is a sense that they do not become only a widow; rather they reconstruct an augmented identity, that of wife/widow” (p. 214).
Grief and Reorganization
Twenty years ago, it was well understood that many individuals take two to four years to adjust to the loss of their spouse, and that the early period of bereavement is typically associated with “profound psychological disorganization and feelings of status loss” (Martin Matthews, Reference Martin Matthews1991). As I have indicated here, new research methodologies have enabled a more nuanced understanding of variability in this process, and a call for the re-evaluation of many of the assumptions that have guided interventions with bereaved persons. Prospective studies of absent grief, for example, suggest that these are not cases of under acknowledged problems related to loss, but rather evidence of doing well among resilient widows and widowers (Bonanno et al., Reference Bonanno, Wortman and Nesse2004, Reference Bonanno, Moskowitz, Papa and Folkman2005).
Widowhood clearly provides opportunities for personal growth and independence, and, for some, release from an unhappy union. In my earlier studies, I noted that the focus of many widows on surviving and prevailing as they take charge of their lives and become more independent over time suggests that adapting to widowhood represents, for some, a transformation rather than a recovery (Martin-Matthews, Reference Martin-Matthews and Neysmith1999). Qualitative studies and narrative accounts of widowhood (Bennett, Reference Bennett2010a, Reference Bennett2010b; Bennett & Vidal-Hall, Reference Bennett and Vidal-Hall2000; Chambers, Reference Chambers2005; Davidson, Reference Davidson1998, Reference Davidson2001; Feldman, Byles & Beaumont, Reference Feldman, Byles and Beaumont2000; Foxman-Feldman, Reference Foxman-Feldman2005; Moore & Stratton, Reference Moore and Stratton2002; Root, Reference Root1998; Scott, Reference Scott1998; van den Hoonaard, Reference van den Hoonaard2001, Reference van den Hoonaard2010) affirm that widowhood fundamentally involves a renegotiation of aspects of daily life and, as reported for women in particular, a reconfiguration of one’s sense of self and relationship to others.
3. Being Widowed: Social Supports and Resources
Social support has been extensively examined in the gerontological literature, with strong evidence of a positive relationship to health. However, the types and sources of support beneficial at one point in the process of adaptation to widowhood may not necessarily be appropriate at a later point in the transition. Researchers have also questioned the assumption of the inherent value of social support. Loose, low-density networks, rather than tightly bound, all-embracing networks, may be most appropriate in enabling widowed men and women to develop new social roles consistent with their new status (Martin-Matthews, Reference Martin-Matthews and Neysmith1999).
Findings of the predominant role of family ties in informal helping networks dominate the literature on widowhood. Adult children provide much of the assistance received by older widowed parents. But patterns of contact between widowed individuals and their adult children are not necessarily uniform, either among children or in terms of the duration of widowhood. Most contact is with daughters, and although the support of adult children is crucial during acute grief, it may become less salient over time as friends increase in importance.
Sibling relations hold a unique place in support networks in terms of their longer duration in comparison with other family ties and their essentially egalitarian nature. Research suggests generally strong patterns of emotional support between widowed persons and their siblings, especially sisters, although frequency of contact and exchange of aid may be comparatively low.
However, as many as one in five widowed persons report not having a single living relative to whom they feel particularly close. Indeed, there is substantial evidence for the importance of friendship ties, even though friendships change substantially with widowhood. Because they are not distinguished by the feelings of obligation that characterize family relations, friendships – especially those with members of a married couple – may not survive widowhood. The ability to make new friendships is, for many, crucial to longer-term adaptation to widowhood.
However, there are many sources of variability in the experience of widowhood, all of which will influence patterns of access to available social supports and their perceived benefit. Increasingly we recognize that widowed persons are not passive recipients of social support, but rather people who “… have built, and continue to occupy, full social life spaces as they age”(Lopata, Reference Lopata1996, p. 215). The images of inherent neediness that characterized earlier studies of widowhood are balanced by the recognition of “a definite decrease of dependency on the part of widowed women” (p. 217). To a large extent, in North America at least, this reflects now several decades of married women being less economically dependent on a husband, and “more likely to be involved in a multidimensional life, much of which is also social and emotionally independent of that partner” (p. 217).
Analyses of social support often consider self-help and mutual support groups that represent particular kinds of peer relationships. Although typically utilized by only a minority of widowed people, such groups have been demonstrated to effectively reduce the distress of widowed people in intervention groups. Formal organizational support groups are also consistently underutilized by widows and are rarely used by widowers.
There are very few studies of interventions with widowed populations in Canada. Following a five-week support intervention with widows, Stewart, Craig, MacPherson, and Alexander (Reference Stewart, Craig, MacPherson and Alexander2008) used pre-test, post-test, and delayed post-test measures and found statistically significant impacts including enhanced support satisfaction, diminished support needs, and increased positive affect.
In the research literature over the past 20 years, there is increasing reference also to the self as resource. Recent research has documented the experience among some widowed people of what are called “identifying moments”, the widows’ sudden conscious awareness of their new status and the fact that their identity has changed. In any study of changes in patterns of social support in widowhood, the possibility of change in widowed individuals themselves as an influencing factor is an important consideration.
4. Variability in the Experience of Widowhood
Attention to issues of variability or diversity in the experience of widowhood in later life has increased substantially in the past 20 years. Of all the topics covered in Widowhood in Later Life, the two which have received the greatest attention in the succeeding two decades are analyses of gender (although less likely today to be framed solely or even primarily in terms of “gender comparisons”) and of ethnicity and culture. These topics are addressed here, with some reference to variability related to financial resources in widowhood, and the role of health, which was not a topic of discussion in my previous writings on later life widowhood.
The Gendered Experience of Widowhood
The focus on the female experience of widowhood has been so predominant that only comparatively recently has the word “widowerhood” been included in some, but not all, English dictionaries. As shown earlier, in all societies, women are more likely to be widowed than are men, and the differential increases substantially with age. Reflecting this reality, I observed a decade ago that “the knowledge we have of ‘widowhood’ and ‘widowed elderly persons’ is gendered. What we know of widowhood in later life is of women’s experiences of widowhood in later life” (Martin-Matthews, Reference Martin-Matthews and Neysmith1999, p. 41). But here, fortunately, the research landscape has changed substantially over the past 10 years.
Gender variation in the likelihood of being widowed, and average age when experiencing widowhood, are explained by biological, cultural, and social factors. On average, men die at younger ages than do women (although the gap in life expectancy is decreasing in many societies). Second, in most cultures, women marry men who are several years older than they are. These two factors combined mean that the likelihood of widowhood increases with old age and being female. However, the third explanation for the imbalance in widowhood statistics by gender, relates to remarriage. Men are far less likely than women to become widowed and then to remain widowed. While close to one in five widowers remarry, only one in 10 widows does so.
Many factors distinguish the experience of widowhood for men and for women. The greater “expectability” of widowhood as a life course event for women frequently leads to a mental rehearsal or anticipation of circumstances associated with being “unattached” in later life. Considerable research has debated whether the loss of a spouse is more difficult for men than for women and about “who suffers more”, widowers or widows. Overall, there are striking similarities in the reported well-being of elderly widowed people. Resources such as income, education, and freedom from limiting disability seem to advantage widowers, while widows benefit from the support of close female friends and neighbours as well as adult children and having more “emotionally intimate” female friends (Connidis, Reference Connidis2010).
Loneliness is described as a major issue for both men and women. Financial resources are much more typically a concern for widows (Bernard & Li, Reference Bernard and Li2006; C. Li, Reference Bernard and Li2005), as they are far more likely than widowers to be poor in old age. Longitudinal Canadian data indicate that not only do widows’ incomes decline, but also more of them fall below the low-income threshold following widowhood. And while many widowers fare well financially in widowhood, older widowers in the lowest income groups appear to be more affected economically (Bernard & Li, Reference Bernard and Li2006).
Overall, widowers report smaller support networks and less involvement with friends and family than widows do. In negotiating their social worlds, widowers rely almost exclusively on themselves or their children. Many widows, by contrast, have much more extensive and diverse social networks. Overall, then, widowers appear comparatively disadvantaged in their ties to family and in their domestic skills, but both these patterns may be cohort specific.
However, an emerging and very important line of research moves away from this “gender-comparison” approach to considering gender variability in the experience of widowhood in later life, and instead examines issues of variability within gender. When applied to widowerhood, for example, this approach anchors the analysis of men’s experiences of widowerhood in their own words, and without explicit reference to women’s experiences (van den Hoonaard, Reference van den Hoonaard2010). Using this approach, van den Hoonaard was able to identify how issues of masculinity framed much of men’s narratives about widowhood. She reported how feelings of competency in the completion of household tasks and issues involving sexuality are major areas of concern for widowers (perhaps, again, reflecting cohort effects for older men in general).
In sum, the meaning and consequences of bereavement and widowhood for men and women under varying sets of circumstances are only beginning to be understood. Gender intersects with widowhood in complex ways that also reflect the status of women in society, and the societal value and meaning (and rewards) of marriage. Much work remains to be done in exploring these issues.
Ethnocultural Diversity in the Experience of Widowhood
One of the most notable changes in the field of widowhood research in the past two decades has been the examination of ethnocultural diversity in the experience of widowhood. Twenty years ago, little was known of experiences of widowhood in cross-cultural context beyond what Helena Znaniecka Lopata had identified in her 1987 volume, Widows: The Middle East, Asia and the Pacific, a pioneering effort to advance the understanding of the role of widow outside Western culture (Martin Matthews, Reference Martin Matthews1991).
Today, there is an emergent, rigorous research literature on the experiences of widowhood in international context. These studies, typically focused on women’s experiences of widowhood (although not necessarily in later life), highlight the relationship between the characteristics of widowhood and the status of women in society. As Lopata (Reference Lopata1996) observed, “the location of a widow in the social structure in terms of social class and minority community is a major influence on these resources … [which] include her knowledge of what is available and how to reach her goals, plus – and this is an important plus – her self-concept” (p. 220). These studies also illustrate how variations from one society to another, and within societies, often reflect social and cultural norms surrounding death, mourning, and remarriage. For example, widowhood in India is often described as a definitive and tragic moment in a woman’s life – one in which her identity is stripped away with the death of her husband (Mastey, Reference Mastey2009). Recent research has contributed to our understanding of international comparisons of women’s economic status following widowhood (Burkhauser, Giles, Lillard, & Schwarze, Reference Burkhauser, Giles, Lillard and Schwarze2004), and of the experiences of widows in the Maltese Islands (Cutajar, Reference Cutajar2000); of frailty among African-Caribbean widows (Meneley, Reference Meneley1999); of widowhood in rural India (Chen, Reference Chen2001); of issues of hope in bereavement for Chinese people in Hong Kong (Chow, Reference Chow2010); of self-rated health (Krochalk, Li, & Chi, Reference Krochalk, Li and Chi2007) and depressive symptoms (L. Li, Reference Li2005) among widows in China; and of “mastery of bereavement and widowhood” among Korean women (Yoon, Reference Yoon1996).
Of particular relevance to the Canadian context, however, is the emergence of research on ethnocultural variability in experiences of widowhood in this country. This enquiry reflects substantial changes in the ethnocultural diversity of Canada over several decades. An estimated 30 percent of the Canadian population over the age of 65 was born outside Canada (although the majority have “aged in place” here). More particularly, between 1981 and 2001, the proportion of Canadian residents over the age of 65 who belong to a visible minority increased from two to seven per cent, with the largest number (39%) being Chinese (Turcotte & Schellenberg, Reference Turcotte and Schellenberg2006). In all, Chinese elders comprise about 10 per cent of the elderly population in Canada. Among persons 25 to 54 years of age, the proportion that belongs to a visible minority increased from five to 14 per cent, suggesting that issues of ethnocultural diversity will be salient in Canada for years to come.
When Widowhood in Later Life was written, it was generally assumed that widowed people in ethno-cultural communities benefitted from more extended support networks and remained involved in those networks throughout their widowhood. This myth has largely now been disproven and is less the focus today. However, this body of research in Canada is modest, using small convenience samples, and is very much of a preliminary and pilot-testing nature. Interestingly, much of it is found in the more fugitive or “grey” literature of graduate theses in Canadian universities, as represented by the research on Hindu-Canadian widows (McArthur Lamb, Reference McArthur Lamb1999; Reinarz, Reference Reinarz1999); Greek Canadian widows (Rosenbaum,Reference Rosenbaum1990); elderly Chinese immigrant women (Gee, Reference Gee2000; Wang, Reference Wang1997); and elderly Korean immigrant women (Yoon, Reference Yoon2008).
My own research, as well, has recently focused on issues of ethnocultural diversity in widowhood. Rosenthal, Martin-Matthews, McDonald, and Chiu (Reference Rosenthal, Martin-Matthews, McDonald and Chiu2008) completed an exploratory study of Chinese widows. The project included a secondary, quantitative analysis of data from the Survey of Labour and Income Dynamics (SLID) and a qualitative study of 10 older Chinese widows to explore their widowhood experiences. These women were recruited through a Chinese Community Services organization in Toronto; all were born in China, were community-dwelling, and had been widowed at least two years. The women in the study ranged in age from 70 to 88 years (median age, 81 years), and they had resided in Canada an average of 15 years.
This pilot study reported findings similar to those reported by Chan almost 30 years ago (Chan, Reference Chan1983), with these women “immersing themselves in their culture but reducing their dependency on family… (p. 84)” (as cited in Martin Matthews, Reference Martin Matthews1991). Although four of the 10 women lived with a child, their preference was to live alone. An “ethic of independence” was evident in such comments as: “My daughter asked me to move to their place. I don’t want to move. I am used to living alone. I cook. I feel free” (Interview #7). Widowhood was also seen to bring some benefits, such as independence. When asked whether there were any positive changes in her life since being widowed, one woman said: “Freedom. There is no one to restrict you” (Interview #2).
This small group of widows vigorously expressed their wish to be independent and not be a burden to their children, and to avoid asking for help from their children. Nevertheless, adult children played an important role at the onset of widowhood and thereafter. Several widows co-resided with children for a period of time after their husband’s death, before moving into an independent residence.
Loneliness was the most commonly described problem of widowhood, even for those who co-resided with family members: “You always feel like you have no one to talk to … my son and daughter-in-law … have to work. I feel very lonely because they are all very busy” (Interview #5).
Most widows reported having friends, although ties seemed somewhat superficial. There are a number of barriers to friendship and interaction with friends: older age (with its attendant losses of friends), immigration in later life (leaving lifelong friends in the country of origin), distance from friends, and lack of transportation.
Overall, this small pilot study found some differences between Chinese widows and what the research literature describes for widows in general. They have smaller informal support networks, siblings are less prominent as sources of support than was found in other Canadian studies (Martin Matthews, Reference Martin Matthews1991), and they seem to be less involved with friends. At this juncture, we cannot assess the extent to which these patterns reflect their status as late-life immigrants to Canada (rather than ethnic Chinese cultural identity in widowhood per se). The study findings suggest that sharing a household with children is not a bulwark against loneliness, and that even shared households can be lonely places. The preference for independent living is also found in the wider population of older adults and in that sense is not surprising. It does, however, run counter to the popular stereotype of the “enviable” multigenerational household that is more commonly found in immigrant groups than in the overall population. Finally, the lives of the widows in this study were generally family-centered, with the women taking great pride in their children’s and grandchildren’s accomplishments and finding meaning in life through their families (Rosenthal et al., Reference Rosenthal, Martin-Matthews, McDonald and Chiu2008).
While studies of ethnocultural diversity among Canadian widowed men and women as yet remain small and quite exploratory, they nevertheless reflect progress over 20 years ago when I lamented the lack of “systematic studies of widowed members of minority ethnic groups” (Martin Matthews, Reference Martin Matthews1991, p. 109).
Other Sources of Variability
It is important to recognize that many of these “sources of variability” intersect and interact and must not be viewed in isolation one from the other. Many issues already discussed both relate to and are influenced by such societal trends as birth rates, societal and cultural norms regarding employment opportunities and labour force participation of women, and policies related to public and private pensions. These in turn impact socioeconomic status and financial resources.
Even in countries of the “developed” world, life expectancy and average age at widowhood can vary widely across gender, ethnicity, and socioeconomic statuses. For men, but less so for women, there is a strong correlation between wealth and low morbidity (illness) and mortality rates. Men with higher incomes and high educational attainment live longer, healthier lives than low-income, poorly educated men. As a consequence, the experience of a surviving spouse, whether male or female, can be vastly different in widowhood, especially in relation to financial security (Burkhauser et al., Reference Burkhauser, Giles, Lillard and Schwarze2004; Connidis, Reference Connidis2010).
Another source of variability may be the housing environment, an issue examined in detail in Widowhood in Later Life. An increasing number of widowed persons live alone, many of them residing in the same community and the same housing unit for many years (Bess, Reference Bess1999). Long residential tenure, often for reasons of sentimental attachment to home as well as inertia at the prospect of relocation, may confine widowed men and women to a home inappropriate for their needs and to a neighbourhood no longer providing a supportive environment. When relocation finally comes for those widowed in later life, it may be into an institutional setting. Although comparatively few elderly people are institutionalized, widowed individuals are proportionally overrepresented among those who seek institutional care in later life (Noël-Miller, Reference Noël-Miller2010).
While it is beyond the scope of this article to consider a broad spectrum of factors that contribute to the variability of widowhood, two developments are important in the Canadian context and warrant mention here: the enactment of the Civil Marriage Act, and the Widowed Spousal Allowance.
On July 20, 2005, Canada became the fourth country in the world and the first country in North America to legalize same-sex marriage nationwide with the enactment of the Civil Marriage Act. By then, an estimated 3,000 same-sex couples had already married, as a result of provincial court decisions that had legalized such marriages in jurisdictions across the country, starting in 2003. Most legal benefits commonly associated with marriage had been extended to cohabiting same-sex couples since 1999. Although our understandings of widowhood following long-term same-sex unions and marriages remain somewhat limited (Bonanno, Moskowitz, Papa, & Folkman, Reference Bonanno, Moskowitz, Papa and Folkman2005; Whipple, Reference Whipple2006), there is a fairly large literature on gay male bereavement, especially as a result of HIV/AIDS. Numerous studies indicate that same-sex partners (whether married or not), and opposite-sex partners who were not married, often experience what has been termed “disenfranchised grief”, wherein “the death of a partner is complicated by the general denial by society of their right to grieve” (Springer & Lease, Reference Springer and Lease2000, p. 300; see also Connidis, Reference Connidis2010). Green and Grant (Reference Green and Grant2008) further argued that sexuality per se is not a direct predictor of disenfranchised grief, but rather that “marginalization and stigmatization owing to sexuality, poverty of social support and ritual are significant” (p. 294).
In 1985, Canada changed the provisions of its Old Age Security Act to allow for benefits for widowed persons aged 60 to 64 years with low incomes, enabling them to “bridge the gap” until these people become eligible for the OAS pension. Since 2000, changes to the legislation have made same-sex common-law partners eligible for the Widowed Spousal Allowance. The fact that widows or widowers are eligible, but single, separated, or divorced persons with the same low income are not eligible, has been claimed as discrimination based on marital status (Martin Matthews, Reference Martin Matthews1991; http://www.servicecanada.gc.ca/eng/isp/pub/oas/allowancesurv3.shtml). This legislation, however, has been upheld following legal challenge under the provisions of Canada’s Charter of Rights and Freedoms. It is noteworthy that for the Widows Allowance, “fewer than half of those eligible are in receipt. Every year thousands of Canadians lose benefits by not applying” (www.shillington.ca/seniors/federal_programs.htm).
5. Summary and Conclusions
I have here considered continuities and changes in the experience of widowhood as well as the conceptual and methodological advances that have shaped widowhood research over the past two decades.
Although widowhood is still an expectable life event (for, by late old age, most elderly people, and women in particular, will be widowed), the length of the period of widowhood is considerably shorter than it was when I described “Widowhood as an Expectable Life Event” (1987a). Its prominence as a status that defines much of old age is far less than it used to be. But if indeed “many current myths concerning widows are old age myths” (Lopata, Reference Lopata1996, p. 212), this may well increase rather than decrease in the future.
Research on the implications of old age for the experience of widowhood will therefore continue to be essential. This article has indicated the advances made over several decades in focusing the lens of widowhood research specifically on elderly widowed persons (through the work of Bennett, Bonanno, Chambers, Davidson, Feldman, Lund, van den Hoonaard, and Wortman, among others). Conceptual approaches that frame research questions in terms of the life course have substantially advanced our understanding of widowhood in later life. Combined with in-depth qualitative analyses, studies have focused on the narrative accounts of older widowed women and challenged the dominant public narrative of misery and decline, pointing instead to a complexity of experience that is rooted in personal biography and in female and male life course experiences, rather than in later-life widowhood itself. The concept of resilience is contributing to a new understanding of bereavement outcomes.
In the past, most studies of widowhood did not include comparison groups of those who have not experienced the loss of their spouse, or of those who had experienced other kinds of loss. As a result, conclusions that any observed patterns were, in fact, due exclusively to widowhood were questionable. Prospective research designs now enable us to address this limitation.
Likely reflecting these conceptual and methodological advances, widowhood research has expanded our understanding of adaptation, social support, of the gendered nature of widowhood and issues of ethnocultural variability of widowhood as both status and process. A burgeoning literature examining the gendered nature of widowhood and its ethnocultural framing (and points of intersection between the two) exists today, whereas 20 years ago little was known beyond select comparisons.
There is much potential in these lines of enquiry. As Connidis (Reference Connidis2010, citing Carr, Reference Carr2004) has noted,
gender variations in the response to widowhood reflect socially structured gender relations that are reproduced in traditional marriage. Apparently individual responses to being widowed are linked to the nature of the lost relationship and the broader institutional and structural arrangements in which it was embedded. (p. 108).
A next requisite step in widowhood research is a focused examination of the links between the insights gained from narrative accounts and these broader societal, institutional and structural arrangements. Chambers (Reference Chambers2005) provided a solid beginning to such framing and analysis.
Since biblical times, accounts of the plight of the widowed have evoked powerful images of women at their most vulnerable and in greatest need. Widowhood has become an especially seminal concept, reflecting and influencing our thinking about old people in general, and old women in particular (Martin-Matthews, Reference Martin-Matthews and Neysmith1999). Recent prominent autobiographical accounts (Didion, Reference Didion2007; Oates, Reference Oates2011) and reviews of them in the popular media (Botsford Fraser, 2011; McLaren, 2011) reflect this seminality. As well, literature in the humanities suggests new directions for understanding connections between gender and widowhood. For example, Clark (Reference Clark2009) argued that understandings of the social identity of the widow and of widowhood as a life cycle category facilitate “a historiography of widowhood that offers not only an investigation of the social and economic activities of individual widows, but also of the cultural meaning of widowhood itself” (p. 190).
Into the Future
That foreseeable changes in Canadian society and in health and social forces will impact the experience of widowhood in later life is undeniable. With the aging of the baby boom generation, Canada – like many nations worldwide – is predicted to experience a “rising tide” of Alzheimer’s disease and related dementias (Alzheimer Society of Canada, 2010). Connidis (Reference Connidis2010) has noted that seeing someone through dementia also alters the quality of the marital relationship so that a partner’s death may feel like a relief. Rosenthal and Dawson (Reference Rosenthal and Dawson1991) have developed the concept of “quasi-widowhood” to describe the experiences of women and men separated from their spouses through institutionalization. One can but speculate about how the experience of widowhood will be framed if sizeable numbers of people are widowed following years of dementia in their spouse (as chronicled in Bayley’s [Reference Bayley1999] searing account of the illness and death of his wife, Iris Murdoch), or separated by institutionalization.
Although, as noted earlier, there have been advances in understanding how health decrements influence the experience of widowhood, there is more to be done in this area. As Lund (Reference Lund1989) has noted, many old people experience the death of their spouse against a backdrop of other age-related changes such as chronic illness, disability, and diminished physical and cognitive capacities. Chambers (Reference Chambers2005) has alerted us to the pitfalls of confounding understanding of widowhood with understandings of physical changes and frailty in later life.
One of the constant themes in research on old age for those in ill health is the question of “who will care for them”. It is worth noting that this, too, is an area where the structural potential of kin availability will change in Canada over time. At the start of this decade, more than a third of elderly persons aged 65 to 74 years of age (35%) had four or more children (still living) – they, of course, being the parents of the baby boom generation. This compares to only 11 per cent of younger adults aged 45 to 54 years having four or more living children (Turcotte & Schellenberg, Reference Turcotte and Schellenberg2006). These figures have implications for the availability of adult children to provide care to an older widowed parent (even if geographically proximate and able to do so).
International and comparative understandings of widowhood will become increasing priorities both as the population worldwide ages, and as international migration results in large diasporas of elderly people aging and widowed in foreign lands. As Lopata (Reference Lopata1996) has noted, in many societies “the worst thing that can happen to a wife in widowhood is a drop in status and being controlled by often indifferent or hostile others (p. 217)”. Ethnic and cultural differences mediate the experience of bereavement, not only in terms of age and pension poverty. But, Lopata added,
in America, the strongest effect appears to be a personal one. It is personal not only in the loss and modification of personally built resources in marriage but also in the emotional loss of what was probably the most significant other in her life. (p. 217)
We will need to understand the nature of widowhood when these two contrasting understandings come to co-exist, through migration and cultural assimilation – as they inevitably have and will.
It is appropriate to conclude with the insights of Helena Znaniecka Lopata. Earlier, I noted how, in her final writing on a topic that had been a dominant focus of her career, Lopata observed how much more “complicated and varied” was the picture of widowhood than that understood or recognized when she first started writing on this topic 40 years ago. This is certainly my perception also as I look back over the 20 years since the publication of Widowhood in Later Life. The meaning of widowhood as a defining characteristic of most women’s later years will change again in the next 20 years. As a category defining access to benefits and services in Canada (as it now does), will widowhood continue to have utility?
This analysis has established that the field of widowhood research is changing in terms of the research questions, theories, methods, and relevant societal and policy issues driving our enquiries. And it must continue to do so. For surely, today’s widowed elderly people have characteristics – in terms of social roles, norms, values and expectations; availability of family members; labour force participation and history; education; access to health and social services; health, wealth and financial resources – that are quite different from the attributes of the elderly widows and widowers of tomorrow.
