Choosing to Retire? A Study of Women’s Patterns for Retiring or Continuing to Work

Women of the baby-boom generation are the first generation of women to present an old age-retirement trajectory similar to the traditional male model. Using a narrative approach, we collected 20 life stories from older women, mostly retired, all born in Canada. We present five patterns that influenced respondents’ decisions to end their career or, inversely, postpone the moment of retirement: 1) “Choosing myself” to fully take advantage of the retirement years; 2) Rational retirement: the deliberate choice to stop paid employment and mourn one’s career; 3) The break: obligation to retire; 4) Retirement as an extension of the retirement of a spouse-breadwinner, and 5) Postponing retirement. Our analysis reveals that patterns vary, depending on socio-economic contexts and state of health, relationship to work, spouse’s retirement status, care work, and the desire for freedom. They also reflect life-long social roles associated with care work, remunerated work, housework, and civic commitments. The Aging-retirement trajectory depends on


Introduction 1
Social and generational changes in second post-war welfare states were transformative for female baby boomers (Arber, Davidson, & Ginn, 2003;Mc Daniel, 2007). While women's work trajectories are not homogeneous (Cau-Bareille, 2011;Nichols, 2019), many women did have the opportunity to pursue post-secondary studies, enter the labor market, have a career, and enjoy the "choice" about whether and when to have children, enter public and political life, etc. Breaking out of the compulsory role of wife-mother-homeworker, the trajectories of women baby-boomers are far more influenced by emancipation, individualization, and social mobility than those of previous generations. Historically, this is the first generation of women have been active in the labor market to reach retirement age in massive numbers (Mc Donald, 2006;Sawyer & James, 2018), thus following a work-retirement trajectory constructed on a masculine archetype (Charles, 2007;Foster & Walker, 2013). Even for men, however, access to an old-age pension in Canada 1 is a relatively recent phenomenon in the history of post-second war welfare state societies. The age of retirement in Canada, all sectors of employment combined, is nearly 64 (Statistics Canada, 2018), while there is no compulsory retirement age. 2 We know that aging-retirement trajectories are gendered (Calasanti, 2010;Calasanti, 1996;Price, 2000;Price & Nesteruk, 2010). Women's experiences in the labor market are different from men's, partly because they are still somewhat less present in the workforce and have lower incomes than men (Gin, Street & Arber, 2001;Women's Philanthropy Institute, 2018). Moreover, socio-economic issues (i.e., poverty; financial insecurity, unequal retirement income) that affect women more particularly, have been well documented (Herd, 2009;Madero-Cabib & Fasang, 2016;Riekhoff & Järnefelt, 2018;Townson, 2009). More women than men are single parents and live alone as they get older (Quéniard et al., 2019;Gambold, 2013;Gazso & McDaniel, 2010;Townson, 2009). Women are also more affected by other issues, such as care work, for which they remain primarily responsible, whether as the primary care-giving parent or as a close caregiver (Ginn, Street & Arber, 2001;Mc Daniel, 2007; Vanier Institute of the family, 2017). These specific realities that circumscribe women's life courses naturally shape their retirement experiences (Gambold, 2013;Newton, Chauhan, Spirling & Stewart, 2018;Prince;2000;Price & Nesteruk, 2010). 3 These interactions between the public and private spheres help maintain the social inequalities in the marketplace and family domain (Tremblay, 2019), which may be even more so in the post-Covid-19 pandemic context, as this crisis in employment seems to hit women more than men. Indeed, if previous economic recessions have mainly affected male sectors such as manufacturing or construction, the COVID crisis has hit more severely service sectors where women are more numerous, and women have thus been harder hit by unemployment and even more precarious positions. Statistics Canada data indicate that the gender gap in employment has increased between men and women, mainly because of women's family responsibilities (Qian and Fuller, 2020). 4 Also, unemployment in Canada went from about 5-6% in January 2020 to a high of 13.7% in May 2020 and 10.9% in July 2020 (Statistics Canada, 2020). Quebec is the province hardest hit by the pandemic, as the unemployment rate has increased from 4.5% in February, to a high of 18.2% in April, returning to 9.2% in July (Observatoire des inégalités, 2020). Some female sectors were more hardly hit; for example, the accommodation and food service industries remain 25% below pre-pandemic level, information, culture, and leisure still 14% below (Renner, 2020) a few months later. On a monthly average, 12% of Canadian workers were laid off in the months from March to June, while in previous recessions, Statistics Canada only reported job losses between 2.5 and 3.5%, indicating the utter strength of the backlash (Statistics Canada, 2020).

5
As usual, it is mainly young workers, newly hired and the less educated who are hit first, but this time women were more hard-hit than men, like personal services, retail (except for food), and the hospitality sector (hotels, restaurants) were very hard hit and have not yet fully recuperated, especially as restaurants have been closed for at least two periods of time. In Quebec, women went from a 58.6% employment rate to 47.8% in April 2020, and returned only to 54.7% in July, while men went from 65% to 52.6% and back to 61.7% for the same months. The pandemic thus hit women's employment more than men, and furthermore, the risk of increasing the social inequalities in aging is real. In this regard, although the present research was done before the pandemic, the interest for women's different passages from work to retirement, the object of this paper, appears relevant to capture in order to better understand the conditions or dynamics that lead women to retire in the contemporary context. In this perspective, based on qualitative research that studies older women's daily activity, our contribution aims to present, different dynamics that led women to end their paid employment and retire or to stay in the labor market. Before turning to the empirical findings, we will present an overview of the state of knowledge about retirement and especially, knowledge about women's retirement. Next, our methodological framework will be explained. The dynamics for leaving (or not leaving) paid employment and entering retirement will then be described, and, in the last section of the article, discussed.

Retirement: midway between a transition and a passage 7
The concept of retirement has changed many times in the last decades (Byles et al., 2013;Price, 2000;Price & Nesteruk, 2010;Sawyer & James, 2018). In the West, retirement has traditionally referred to a specific event in a person's life, a moment of transition between the world of work and the world of well-deserved leisure and rest. This concept of retirement, characterized by "non-productive" activities, now seems obsolete (Byles et al, 2013). The idea of a retirement cut-off age within a 3-part cycle (study-work-retirement) now seems to be vanishing due to the ever blurrier boundaries between the stages of life (Guillemard, 2010). Retirement is no longer represented as an "abrupt transition" (Byles et al. 2013) and some describe it more as a passage (Boboc & Metzger, 2013). Hooyman et al. (2002, cited in Byles et al., 2013 describe it as a "fluid concept" describing different phases of transition and fluctuations between employment entries and exits in old age. Indeed, the transition to complete retirement can be long and complex; many people alternate between retirement and work, or combine both, while others decide to permanently cease any form of paid work (Canadian Longitudinal Study on Aging, 2018). Thus, the concept of retirement becomes polymorphic, depending on the contexts in which it appears in international studies (Maestas, 2010). For many of the baby boom generation, it represents a passage into a new stage of life, brimming with new possibilities (Byles et al, 2013;Mansour et Tremblay, 2018b). This "new beginning" or "time of revitalization" (Gibson, Ashton-Shaeffer, Green, & Corbin, 2002) was evoked by the activist and positivist discourse and practices associated with the "young-old" in normative theories on successful, productive, and active aging (Kojola & Moen, 2016;van Dyk & Turner, 2010). 3. Women's retirement choices and activities when paid employment comes to an end 8 Retirement is an increasingly diverse phenomenon for women (Byles & al, 2013;Newton & al. 2018;Sawyer & James, 2018;Women's Philanthropy Institute, 2018). Concerning the numerous interrelationships between the family and work spheres, many researchers "stress the necessity of including family life courses in retirement studies along with employment trajectories" (Madero-Cabib & Fasang, 2016: 44). For example, the close link between motherhood, care to work, and paid work may explain why women holding traditionally female jobs leave paid work earlier (Cau-Bareille, 2011). According to Byles et al. (2013: 38) "one of the most common reasons given [by women] for retiring was for personal health problems or physical disabilities and having attained some level of financial security." These contextual factors connected with socio-economic status and health influence the decision of when to retire, as do the family situation, care work, and the work context (Newton et al., 2018). In addition, retirement decisions must be situated in the intertwining of the micro (individual experiences), meso (workplace, organizations), and macro (social policies, types of socio-political regimes) contexts (Kojola & Moen, 2016). They are also influenced by gender (Byles et al. 2013;Newton, et al, 2018) and social class (Kojola & Moen, 2016). 9 Activities pursued in retirement are as varied as the reasons for stopping work. In this perspective, regardless of the types of activities pursued in retirement, care work, manifested in diverse daily practices, remains central to women's lives (Dentinger & Clarkberg, 2002;Marchand, 2018a). More specifically, women are still very much involved in formal volunteering-for instance, in organizational structures (Charpentier et Quéniart, 2009;Lie, Baines, & Wheelock, 2009;Pavelek, 2013;Price & Nesteruk, 2010;Seaman, 2012)-and informal volunteering or commitments such as caregiving within the family and intergenerational networks (Byles & al;Charpentier et Quéniart, 2007;Dentinger and Clarkberg, 2002;Martinez, Crooks, Kim, & Tanner, 2011;Price & Nesteruk, 2010). When women retire, their lives often continue to be heavily centered on the family (Price & Nesteruk, 2010), while others continue fullor part-time paid work after turning 65 (MacEwn 2012; Price & Nesteruk, 2010). Leisure is another aspect of retired women's pathways (Price & Nesteruk, 2010). In short, retired women combine various roles and social, political, and civic commitments (Byles et al., 2013;Charpentier et Quéniart, 2009;Kulik, 2015;Marchand, 2018a). 10 As part of the research continuum investigating "the unique nature of women's retirement" (Prince, 2000: 82), our study, based on a narrative method (life stories), considers the daily life practices of Canadian women aged 60 to 70. The analysis of these stories revealed retirement as an omnipresent theme. We, therefore, focus on the dynamics that led participants to leave the workforce, or, inversely, continue working and delay retirement. To capture the unique experiences of women baby boomers, our article focuses on the important biographical transition of leaving paid employment and entering retirement. We propose to describe the patterns observed in women's retirement experiences, and then, we discuss the gendered aspects of women's experience of aging and finally, the results issues in light of the dominant active aging contexts and policies.

Methodological framework 11
Our theoretical frame is inspired by different perspectives: critical perspectives of aging, social gerontological, and feminist studies. The results presented in this paper emerge from a larger qualitative study that focused on different ways in which older women experiment with citizenship through their daily activities (Marchand, 2016;2018a). However, the study revealed some unexpected results on older women's retirement transitions and experiences. We thus decided to analyze them in a specific article. Using the life story method to understand the everyday reality lived by actors, we collected 20 life narratives from women aged 60 to 70, mostly retired, all born in Québec (Canada). The narrative method permits a rich description of individual histories and explores how actors-subjects structure and construct their life experiences, attributing to them a singular meaning and significance (Chase, 2005;Creswell, 2012). Each participant's narrative considers her sociohistorical context. The narratives thus enabled us to uncover the social reality, events, and the emotional and psychological states experienced by the respondents and relate to them both temporally and spatially. We interviewed 20 women from different backgrounds 2 The sample was selected based on the "maximum variation" (Miles & Huberman, 2014) method, to obtain cases representing a variety of positions and experiences related to the research question. This method generates a "constellation of cases," which are first analyzed vertically and horizontally, and then, comparatively (Pires, 1997). The participants were recruited via social networks and by the snowball method. The criteria for diversity were the following: Age (60 or 70); Being a carer or not; having had children or not; being married, single or divorced; having been employed or stayed at home most of one's life; occupational category; level of education; living in a rural or urban area; socioeconomic status; health condition. All interviewees were francophones and interviews were done in French. Parts selected for the article were then translated into English.
12 To be coherent with the storytelling method, non-directed interviews were conducted, where the interviewer acted more as a guide based on a general question asking the women to tell us what they did during a typical week. Then, when respondents stopped talking, the interviewer would ask stimulus questions to go further on the answers. While the focus was on the present, women often referred to past events of their lives (divorce, stopping work, etc.) This made it possible to cover various periods of their lives before retirement. The interviews lasted between 2 and 4 hours. All the interviews were recorded and fully transcribed. In deconstructingreconstructing the corpus (Paillé & Muchielli, 2012), we began with a codification system guided by "different realities" (empirical, discursive, and reflexive reality; cf. Bertaux, 2010). Next, we referred to fact-based segments (empirical reality) to reconstruct the chronology of everyday life in which actions and concrete practices occur. Segments about discursive and reflexive aspects were classified under broad categories (Paillé & Muchielli, 2012), referring to the "social worlds" of the respondents (Bertaux, 2010) (i.e., private/domestic universe, family universe, and social universe) and two categories of meaning (practices themselves and contribute to society). The step of data "condensation" (Miles & Huberman, 2014) provided us with an enriched and thus deeper reading of each story. Beyond a reconstruction of everyday life, we managed to piece together a dialogical analysis of the stories, revealing the messages underlying the discourses and their content, and thus, the points of tension, disruption, consonance/dissonance. Third, because we wanted to conduct a comparative analysis of the stories, we synthesized the data using the broad category themes for the discursive and reflexive segments (Paillé & Muchielli, 2012). To create a crosscutting overview of themes for the entire sample of respondents, we produced comparative tables, drawing on the themes related to each of the previously introduced six categories. This method allowed us to compare our reconstructed corpora and to analyze the "reappearance of the same situations, rationales, and similar actions" (Bertaux, 2010: 95) or, on the contrary, antagonistic processes or mechanisms, as well as the tensions and ambiguities that cross the narratives.
14 Finally, let us mention a few limits of the research and paper. One of the important limits to this study concerns the diversity of women profiles from the point of view of the race (all Caucasian), and social class (middle class for the majority), and language (all Francophones. Because of the small number, we were unable to analyze the effects of gender with the other intersections. In this regard, a more heterogeneous sample would probably contribute to refining the complex realities of leaving careers and entering retirement. Future research should try to include more people of different socio-cultural and economic contexts. Otherwise, we could have extended the theoretical perspectives to include Bridge Employment, which puts the accent on sources of revenue and also symbolic investment in work. While we did address this aspect of symbolic investment in work to a certain extent in the analysis, it is an issue that could be analyzed in more detail in future work; however, this might require a higher number of respondents invested in work after retirement to address more fully the Bridge Employment issue and its impact on retirement (Mansour &Tremblay, 2018).

Results
15 Analysis of the life stories revealed very diverse aging trajectories. The women interviewed belonged to a "pivot generation", that is, "a generation that had split with earlier family models after the introduction of changes far more radical than those of previous generations and the following generation" (Attias-Donfut, 1995: 53, author's translation). In the West, women, and men of this generation underwent social changes instituted by the feminist revolution, which modified socialization trajectories, models of reference, and the identity of a large part of the population. Notwithstanding the uniqueness of individual trajectories, we did note common experiences (e.g., long presence in the labor market for most, divorce, break for children in some cases, and reentering the labor market once children were older, etc.). For all participants in the study, ending paid employment and entering the world of retirement represented a turning point in their lives, and different dynamics informed the decision to stop (or not stop) working in later life. These dynamics are analyzed based on the respondents' different contexts, status, socio-economic situations, and health status; this led to defining five patterns that reflect women's retirement experiences.
5.1. "Choosing myself" to take advantage of my retirement years 16 For one group of respondents, leaving paid employment was a choice, even for those who had not yet turned 65, which is often seen as the "normal age" of retirement in Canada, as even if there is no obligation to retire, it is this age that entitles a person to the full old-age pension benefit paid by the federal government. For this first pattern, two primary factors influenced the decision to leave paid work: health issues and those related to the uncertainties in the aging trajectory.
17 Regarding health, the respondents grouped under this pattern considered themselves "in good health." They nonetheless expressed a strong desire to remain in good health  (60), is like a "gift" we give ourselves when we're still "young" and in good health. When we anticipate uncertain health in old age and become conscious of our potential decline and the emergence of fragility or serious diseases, we consider our possible choices from this perspective: "I worked with older people for 26 years and in the last 5 years, it began to hit me : the walker, the wheelchair, Alzheimer's . . . seeing people becoming more dependent, it was a shock. I said, 'oh . . .that's me in 15 or 20 years. I needed to step back." (Pauline, 64, retiree from the field of administration).
19 A second important factor in the decision to stop paid work is embodied in the temporal dimension, namely the reappropriation of all social times. As Chantale (67) explained : "Every hour in the day belongs to me," adding that "now, I am totally in control of my time." The phrases "Taking time, having time" were like a constant tune in these women's stories, most of whom chose to retire after "spending my life on the run between work and kids." Catherine is a good example of this preoccupation, describing herself as an "authentic girl of '68," a product of the 1970s, anti-conformist, freedom-loving, and fleeing the constraints of the work world. As soon as she turned 65, she retired, "not wanting to spend a minute more at my job." 20 Last, ridding oneself of the stress and demanding responsibilities of a profession that requires a high level of education is an important, if not crucial, factor in the decision to retire. This was the case of Christine (63), a former CEO who made a deliberate decision to end her career at 57 : "I've noted a big difference in my stress level. Every decision I made affected a thousand people. This changed from one day to the next. It was huge for me. Just huge…For me, the major difference is my stress level." Like the other respondents grouped under this rationale, she affirms that she has "never regretted her decision" to retire from the workforce when she was still relatively young. This is not the case for all women, however, as we will see with the next rationale.
5.2 Rational retirement : the decision to end a much-loved career 21 Retiring is "mourning a career that I loved." This comment accurately summarizes a pattern in which the departure from paid work is a deliberate choice but nonetheless one that is traversed by tensions and constraints related to family contexts and circumstances. Still very much attached to her job, Marie-Andrée (64) was spending more and more time with her grandchildren. The hospitalization of her husband led her to consider a career end from the perspective of her overall life path : "I said to myself that I had more years behind me than in front of me and how did I want to live the years I had left ?" She added that it was important for her to "take care of her relationship." Her situation resembles that of Claire, who was also carrying out several care tasks. Several years before she finally left her job, she had already cut back her work hours to care for her sick mother. Retiring was a difficult choice for her, she Entering retirement was not something that Marie-Andrée and Claire had planned to do "so soon" in their trajectory, but they made this choice to be able to spend more time with family, especially children and grandchildren. The importance of family and the centrality of the grandmother's role, as well as attention to the couple, played a fundamental role in the decision these women made to retire. In a similar vein of attention to loved ones (caring about), the spouse's activity can also precipitate the decision of married women to leave the workforce. This was the case of Françoise (64), married to a "born builder," a "man of the land" who bought a house in the French countryside as a renovation project. Due to the long stay overseas necessitated by this project, she ended her career, an experience she found difficult : "For sure, it was hard to stop working, there was a period of mourning. I had to find another project, so I threw myself into the renovations." 23 Like Françoise, Michelle also chose the timing of her retirement, mostly to be closer to her family, but experienced entering retirement as a "shock" : "The first year was really hard. As I said to my friend, G. : 'I didn't enter retirement, I fell into it. It's not the same. I guess I just wasn't ready. . . In the North, I was the nurse, everyone knew me up there and then I came here where I'm nobody . . . it was a shock." (Michelle, 67, retiree from the field of health and social services) 24 Retirement, for this woman who worked some 30 years with Indigenous communities in Canada's far North, meant returning to the "city", a letdown of sorts ; a porous zone between two stages of life where she engaged in a process of personal reconstruction in the privacy of her home to "build a new system of references." In short, this pattern for leaving paid work is pervaded by uneasiness : in response to family situations and circumstances, the respondents grouped under this rationale made the reasoned decision to retire, a biographical transition that was not without clashes and ambivalence.

The break : obligation to retire
25 In contrast to the two preceding patterns, the few participants who corresponded to this third pattern were all forced to leave their paid jobs due to health-related incidents or diagnoses. Odette, a woman who had always been "as healthy as a horse," was diagnosed with advanced fibromyalgia and had to leave her job "the very next day." This forced retirement was experienced as a brutal break, generating a series of consequences in different areas of her life (economic, social, psychological/emotional, etc.). This "hurricane," as she describes her illness and its impacts, not only upset her life, but also her sense of belonging in the world : Marie's story evokes a similar turning point because after suffering a stroke, she was paralyzed on her left side, a physical condition that precipitated her into early retirement before she turned 60. For both women, family support was a crucial pillar during the forced transition, which involved a series of psychological and emotional reactions, on top of which came the gradual loss of a social network as a result of the abrupt cessation of employment. These realities are present in the narrative of Lorraine (63), who became blind in her fifties when she was teaching. After a difficult transition period, treatment and rehabilitation, she now considers that she is "living a second life, even though she had never imagined her life as it is today." In short, the aging-retirement rationale characterized by a break refers to a dramatic event that severs the employment trajectory, followed by family and social upheaval. Faced with these biographical splits, the respondents employed diverse resilience strategies : combativeness in the face of illness and defending their rights with the employer (employment insurance claims), proactivity in seeking various resources, and great personal resilience ; they refused to "feel sorry about themselves" for the rest of their lives.

Retirement as an extension of the retirement of a spouse-breadwinner
27 It must first be kept in mind that paid work does not figure in the lives of all older women. As Charles (2007 : 5) mentions, until the 1960s and 1970s, most women declared themselves "inactive" because "women's essential activity took place outside the labor market." This is the case of a minority of the respondents who did not engage in paid employment. For them, "retirement," and its associated social rights can be understood "in extension" to the retirement of a husband-breadwinner.
28 At the discursive level, the term "retirement" was rarely used by respondents to describe their own aging trajectory. It is rather the partner's retirement that represents a moment of transition in these respondents' older years. As such, their pathway is conditioned, not by their own pattern for stopping work and commencing retirement, but by those of their spouse. For example, Rita's (66) story is strongly shaped by her role as a caregiver for her husband, who retired due to deteriorating health : "He [her spouse] is very ill ; it takes almost all I've got to take care of everything. I have to do everything, our finances, debts, etc. He can't do a thing, so it's hard. It's tough for the morale too." 29 Carole's situation is altogether different. Still living in the village where she was born, she remembers that when women married and had children in that community, they stayed at home to dedicate themselves to their children's education. Having volunteered in her parish all of her life, it was when her husband retired, healthy and with a good retirement plan, that she was able to enjoy new sports and cultural activities and travel.

Postponing retirement
"For me, retirement equals boredom, stopping everything, and ceasing to think or be involved." These remarks from Suzanne (61) echo the notion of "retirement as a social death" conceptualized by Guillemard in 1972, illustrating the social exclusion experienced by retirees at that time. Although representations and experiences connected with aging have changed, for some respondents retirement represents a setback ; a period of life to be deferred "as long as possible." Respondents grouped under this pattern indicated that their degree of satisfaction with work remained high ; they said they "love" what they do, in their work as an artist, child care educator, university professor, or street newspaper seller. The prospect of their eventual retirement raised questions about their postemployment trajectory and activities since, for them, work was an important place of social integration and recognition. Furthermore, their job or career played a significant role in their personal and professional identity, remaining intrinsically linked to their values and sense of meaning. Hélène exemplifies this : "At the moment, it's hard for me to imagine how I would find meaning or value in my life once I have retired. I don't actually think about it much right now . . . but I'm aware of it and tell myself that probably I won't have the choice in a few years, I'll have to start planning it . . . I mean, the psychological aspect of retirement . . ." (Hélène, 61, artist-painter).
30 Moreover, for these respondents, retirement increased the sensation of "feeling and looking old" tenfold. "When you're healthy and fit, you don't need to stop working," said Suzanne, echoing Hélène, who also stated that she remained at work because "people don't see me as dead wood," adding that in our aging societies, she "can't imagine that we will work for only 35 years." Moreover, as a teacher, her job gives meaning to her life : "My sense of meaning comes from my job, not my hobbies."

Synthesis 31
In sum, with the first two patterns, the cessation of paid employment is preceded by serious thoughts about personal aspirations for the later years of life, health issues, and the importance of taking care of oneself (self-care) and loved ones. What most distinguishes the first from the second pattern is the quest for freedom that motivated the decision to leave paid work, to "choose myself." The third pattern as opposed to the first two : the pattern for entering retirement is constrained and brutal ; it causes a definitive break with productive work, thus staging an abrupt transition, or even a split in the life course between "having always been active" and becoming, suddenly, "inactive". The last two patterns express a form of continuity with life as it has been up until now : postponing retirement enables one to continue paid work and put off issues associated with the biographical transition of retirement.
32 In the case of women for whom retirement is an extension of their breadwinner husband's retirement, there is a sort of continuity with life as they have lived it because the practices of daily life were never associated with paid employment. Nevertheless, the context of a breadwinning spouse's retirement shapes their own relationships to aging-retirement and the practices of their later years. Finally, for those who postpone retirement, they can be seen as implicitly referring to the notion of "retirement as a social death" conceptualized by Guillemard (1972), and retirement is something that would make them 'feel and look older.

Discussion
33 What most interested us in considering these five patterns was the analysis of the contexts, individual and collective experiences, and social relations underlying these decisional or event patterns that lead to retirement. As various authors have noted, health is unquestionably a recurring reason behind women's choice or obligation to stop working (Byles et al. 2013 ;Cau-Bareille 2011 ;Kojola & Moens, 2016 ;Price & Nesteruk, 2010). Beyond the objective health situation, the rationales reflect a subjective relationship to health in the course of aging : the link between work and health, which may become increasingly difficult as we age ; the uncertainty of health status in the coming years ; the desire to be healthy to enjoy new freedom, not to mention the emotional states caused by the state of health, one's own or that of a sick spouse, which forces women to retire. Even the decision to continue working "while I'm still healthy" requires a subjective evaluation of one's state of health. As Cau-Bareille (2011 : S25) remarked, "Health is a state of balance -in fact, a state of fragility -and one's personal assessment of this balance changes over time." 34 Family trajectory and caregiving responsibilities also influence women's career exit and retirement patterns (Byles et al, 2013 ;Dentinger and Clarkberg, 2002 ;Price & Nesteruk, 2010). In our study, women who have evolved according to the traditional male breadwinner-female caretaker model, even after having held paid employment for a large part of their lives, are more likely to make a rational choice to retire, because of the predominance of care work, particularly with the arrival of grandchildren. It is interesting to note that while this generation of women has been active in the labor market, almost as much as men, retirement often comes along as an important break and while some chose to postpone it, women's intertwined relation with family often imposes retirement. It must be noted that "even if women of this generation have manifested the desire to exist outside the family sphere, intergenerational mutual aid still has the effect of steering them into domestic care duties" [Translation] (Pennec, 2009 : 149). In the case of the caregiver, it is important to question the structural conditions in which this choice is made, to avoid to essentialize care work, which would fall under the self-sacrifice relating to the "feminine nature" (Marchand et Firbank, 2016b). 35 However, the care work performed by many older women contributes to the traditional gendered division of labor once they have retired, but also has a differing influence on women's and men's perceptions of employment (Byles et al., 2013 ;Dentinger and Clarkberg, 2002 ;Prince & Nesteruk, 2010 ;Farges & Tremblay, 2016). Similarly, the "retirement of the spouse-breadwinner" pattern, like the "retirement as Husband's retirement" pathway described by Gibson and others (2002), reproduces, in retirement, the traditional male breadwinner-female caretaker model. The retired male breadwinner model, which perpetuates the gendered division of labor, results in women's continued economic insecurity in their old age (Calasanti, 2000 ;Prince, 2000). 36 Also, the desire to postpone retirement to keep working as long as possible was not associated with financial considerations for the respondents we interviewed. Ass in other studies, the reasons given were more about contributing to society, continuing to be useful and active, and sharing their knowledge (Gibson et al. 2002 ;Kojola & Moens, 2016 ;Mansour et Tremblay, 2018a ;b ;Newton et al. 2018). Similarly, for professional women, work gives life meaning. The central place of work in the lives of women of this generation is also evoked in the findings of Kojola & Moens (2016) and Newton et al. (2018). Moreover, we hypothesize that a strong professional identity is linked to the desire to extend one's career. In contrast to this pattern, participants who chose to retire as soon as possible reflect the quest for freedom, expressing a deep-seated desire to take advantage of this "new period of life" by liberating themselves from the constraints of work. "Freedom from time constraints enabled them to develop a more holistic lifestyle (Gibson et al., 2002 : 261). Last, as with the data of Newton et al. (2018), which shows that women who have the luxury to choose the timing of retirement exhibit a higher degree of satisfaction than women who do not, our findings also reveal the influence, in everyday life, of having, not having, or having only a little power to determine the timing of retirement.
7. Conclusion : the implication for policies and future studies 37 In the present pandemic context, where women are harder hit than men by job losses, this lack of power is surely an important issue to consider. Also, while job losses may lead to the retirement of older women, it is not yet clear that this may have had an impact beyond a few sectors where women of all ages have lost their job ; in Canada, this was the case for hospitality, hotels, restaurants and some retail sectors, but it is not yet clear if these will transform into retirement. Some US researchers indicate this is not the case (Munnell and Chen, 2021), but it might be different for Canada, given the different retirement but also health plans (Tremblay, 2014), as well as strategies that firms can put forward to retain older workers (Mansour & Tremblay, 2019). While all stories here are related to health issues, the sudden break is something which can also be characteristic of persons who suffer job termination or cuts, as is often the case in the present pandemic context, but also in all times of recessions. As well, for many other subjects, the pandemic will surely have important impacts on work-retirement transitions and our five types can be useful in the future to analyze the transitions of women who will have lost their jobs during the pandemic and not been able to return to employment, as many female sectors were particularly hard hit (the hospitality, restaurants and travel industry in particular, but other service sectors as well). Some new categories might well appear following the pandemic which exacerbated, and highlighted, issues related to gender, race, and social class. The analysis of these different power relations, within their specific context, would probably influence entry into retirement for some senior workers. As mentioned previously, not having the 'luxury' to choose the timing of their retirement will surely have important socioeconomic impacts on many, as paid employment was an important part of women's identity in the baby boom generations. 38 The patterns underlying a decision or the obligation to retire vary based on various contexts and processes : state of health that can influence or force retirement, identity relationship to work and career as well as the feeling of remaining in "active" life, husband's retirement status for women who have lived according to the traditional male breadwinner-female caretaker model, family caregiving responsibilities that concern most respondents and, finally, desire for freedom, which is expressed by the possibility of choosing retirement activities and leaving behind the constraints of the working world. Yet, despite the changes that influenced the trajectories of women of this generation, for most of the respondents, the patterns underlying their decisions to stop or continue paid work reveal a form of continuity of the social roles adopted over their life course, whether care work (formal and informal), paid work, household work, or civic involvement. These dynamics, which influence careers and retirement need to be taken into account in economic, retirement, and aging policies, including so-called active-aging policies which are now frequent in many industrialized nations. First, active aging policies are not gender-neutral (Marchand, 2018b), nor are they without political objectives, as the male trajectories still dominate the views on work and retirement. As mentioned by Le Feuvre et al., (2014, p. 14), women are in the 'blind angle' of policies to increase labor market activity, and thus also of policies for active aging, which tend to neglect reproductive and care work assumed by women. As research has shown in recent years, care work, defined as practices and attitudes of caring for others (Brugère, 2010) is embedded in women's trajectories, in employment, a career as well as in aging and retirement.
40 From this perspective, the retirement transition must be understood, in social and economic policies as well as in future research, as occurring at the intersection of social contexts, individual life courses, and gender (Newton et al, 2018 : 3). In this respect, we find the feminist life course perspective very useful for understanding retirement pathways and considering the interdependence of work-related and private or family imperatives (Dentinger and Clarkberg, 2002). To capture the diversity and complexity of women's retirement experiences, in the present and future generations, therefore, it is important, to not only concentrate on the dominant pathways of retirement but also to investigate fully the influences related to the intersection of social relations, such as those associated with age, sexual orientation, ethnic/racial background, etc. in the aging-retirement trajectory. These all impact women's life courses and retirement dynamics. The interviews were conducted in 2014-2015, as part of a doctoral thesis (Marchand, 2016).

ABSTRACT
Women of the baby-boom generation are the first generation of women to present an old ageretirement trajectory similar to the traditional male model. Using a narrative approach, we collected 20 life stories from older women, mostly retired, all born in Canada. We present five patterns that influenced respondents' decisions to end their career or, inversely, postpone the moment of retirement: 1) "Choosing myself" to fully take advantage of the retirement years; 2) Rational retirement: the deliberate choice to stop paid employment and mourn one's career; 3) The break: obligation to retire; 4) Retirement as an extension of the retirement of a spousebreadwinner, and 5) Postponing retirement. Our analysis reveals that patterns vary, depending on socio-economic contexts and state of health, relationship to work, spouse's retirement status, care work, and the desire for freedom. They also reflect life-long social roles associated with care work, remunerated work, housework, and civic commitments. The Aging-retirement trajectory depends on women's collective and individual histories and the impact of gender on other intersecting social relations.