Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.
By Jorge Uriarte-Landa and Benoît-Paul Hébert
Full article in PDF
Work-life conflict among older workers remains largely understudied, despite its potential impact on health, productivity and retirement decisions. This article examines work-life balance among older workers, 55 years of age and over, focusing on dissatisfaction with work-life balance, the most commonly reported reasons for dissatisfaction, and key factors and personal characteristics associated with work-life conflict.
While there has been a plethora of work-life balance studies, most of this research has focused on younger workers with children. In Canada, work-life balance of older workers has started to garner attention, but mostly in the context of elder care (Pyper 2006, Habtu and Popovic 2006, and Williams 2005). Gaining a more comprehensive understanding of this issue is important for several reasons. First, work-life conflict has been associated with negative health and productivity outcomes (Duxbury and Higgins 2003). Second, research suggests that work-life balance plays an important role in retirement decisions. In the U.S., work-family conflict among 52- to 54-year-olds has been associated with higher odds of planning to retire within the next ten years (Raymo and Sweeney 2005). In Canada, over 25% of retirees report that they would have continued working if they had been able to work part time or shorter/fewer days, while 6% would have done so if they had suitable caregiving arrangements (Morissette, Schellenberg and Silver 2004).
Within this context, this article sets out to examine work-life balance among workers age 55 years and over, using data from Statistics Canada's 2005 General Social Survey (GSS). The first section introduces some of the potential sources that may contribute to work-life conflict among older workers. The second compares selected socio-demographic, household, and work-related characteristics of older workers with their core-age counterparts (25 to 54). The prevalence of dissatisfaction with work-life balance as well as the most commonly reported reasons for dissatisfaction are then presented. Finally, multivariate analysis is used to measure the impact of various factors on the probability of work-life balance dissatisfaction among older workers (see Data source and definitions).
Several recent studies have pointed to the need to broaden the scope of work-life balance research beyond the context of families with children to include older workers (Yeandle 2005, Hirsch 2003, Gardiner et al. 2007). They argue that as workers get older, many are likely to experience changes in their family situations, health or interests outside of work that may become sources of work-life tension. Examples of such changes include:
In 2005, there were 2.3 million older workers (age 55 and over) in Canada, representing 14% of the total workforce (Table 1). About three-fifths of these workers were men—a higher proportion than among core-age workers (25 to 54). The majority of older workers (84%) were age 55 to 64.
As expected, the GSS data show that older workers were far more likely to have a disability than their younger counterparts (38% vs. 26%).1 Thus, for many older workers, functioning at work and outside of work might be challenging unless appropriate aids, supports and accommodations are provided.2
Older workers were less educated than their younger counterparts. Overall, 54% of them had a post-secondary degree or certificate, compared with 62% of core-age workers. A higher proportion of older workers than core-age workers reported annual personal incomes of $100,000 and over (11% vs. 7%)—likely the result of higher seniority and work experience.
Self-employment and part-time work were quite common among older workers, possibly indicating a conscious transition towards retirement (Table 2).3 They were twice as likely as their core-age counterparts to work less than 30 hours per week (20% vs. 9%). Self-employment was particularly high at 31%, compared with 18% among workers age 25 to 54.
Older workers—perhaps due to their high self-employment rate—had more flexibility than their younger counterparts in terms of when and where they worked. Almost one-half of them (48%) indicated that they were able to choose the start and end times of their work days, compared with 41% of core-age workers. About one in four usually worked some hours from home (excluding overtime), compared with one in five core-age workers.
There was no discernible difference between older and younger workers in terms of work schedule types. Overall, 72% of older workers had a regular daytime schedule (i.e. non-shift work),4 virtually the same proportion as in the core-age group.
The occupational profile of older workers was also similar to that of their younger counterparts (Chart A). As in the core-age group, over one-half of older workers were employed in three broad occupational groups: sales and service (21%); business, finance and administration (21%); and trades, transport and equipment operators (14%). This was followed by occupations in management (9%) and social science, education, government, and religion (8%). One notable difference with respect to younger workers was that older workers were twice as likely to have jobs in primary industries (6% vs. 3%).
Older workers differed from their younger counterparts in terms of household characteristics. Just 3% of older workers were living with children under the age of 15, compared with 40% of core-age workers (Table 3). At the same time, older workers were more likely than those age 25 to 54 to be living with a spouse or common-law partner (77% vs. 72%).
Even though few older workers had young children, more than one-quarter (26%) participated in unpaid child care—compared with 48% of core-age workers (Table 4). There were important differences with respect to the core-age group in terms of the location and intensity of child care. While core-age workers most frequently provided child care inside their households, older workers' child care took place predominantly outside their homes—likely reflecting care of someone else's children (e.g. grandchildren). On average, older workers who provided child care devoted substantially less time to this activity than their younger counterparts (12 hours versus 33 hours per week).
Many older workers were elder care providers—20% indicated that they provided care to seniors in need of assistance, compared with 16% of workers age 25 to 54. As in the core-age group, most of this care took place outside the household.
Finally, about 5% of older workers provided some form of care or assistance to non-senior adults.
Older workers were less likely to be dissatisfied with their work-life balance than their core-age counterparts. (Chart B). Overall, 14% of older workers reported being dissatisfied with the balance between their jobs and home lives, compared with 25% of workers age 25 to 54. This 11 percentage point spread with younger workers was the same for men and women.
The proportion of older workers who were dissatisfied with their work-life balance varied significantly by age (Chart C). It hovered around the 15 to 16% range for those age 55 to 64, dropping sharply to 5% for those age 65 and over. This pattern will be revisited later in the paper.
As shown in Chart D, close to one-half of older workers who were dissatisfied with their work-life balance reported spending too much time on the job (45%). This was by far the most commonly reported reason for dissatisfaction within this group, followed by not having enough time for the family (27%), other employment-related reasons (21%), and not having enough time for other activities (13%).
Older workers were very similar to core-age workers in terms of the reasons behind work-life balance dissatisfaction, with one important exception. Older workers were considerably less likely than their younger counterparts to associate their dissatisfaction with not having enough time for the family, the most common reason among the younger group (46% versus 27%).
Although spending too much time on the job was by far the most common source of work-life conflict for older workers, not many were willing to cut back on their work hours at the sacrifice of pay. Indeed, only 27% of those reporting this source of work-life conflict indicated that they would prefer to work fewer hours for less pay (at their current wage rate). Thus, for many of these workers, financial considerations—and not necessarily lack of employer flexibility—appeared to be the key factor making it difficult to cut back on hours.5
Although cross tabulations indicate that the risk of work-life balance dissatisfaction among older workers varies with age, multivariate analysis is required to determine whether this finding persists after controlling for other relevant characteristics. One issue is that older workers are increasingly likely to leave employment for retirement as they age. Research has shown that this is not a random process and people with certain characteristics are more likely to withdraw from the labour market. In particular, workers most likely to experience work-life conflict may also be more likely to retire, thereby self-selecting out of the sample providing information on work-life balance.
Ignoring this self-selection could result in biased estimates (Heckman 1979). This difficulty was addressed by using a probit model with selection following the method proposed by Van de Ven and Van Pragg (1981). This technique provides estimates of the probability of work-life balance dissatisfaction based on a set of explanatory variables while controlling for the selection of older individuals out of employment.6
The probit model with selection was used to investigate the relationship between dissatisfaction with work-life balance and the characteristics of older workers along three dimensions: socio-demographics, care responsibilities and job-related characteristics. The results are reported in terms of marginal effects: the change in the probability of reporting work-life balance dissatisfaction associated with a given characteristic (Table 5).7
In stark contrast to the descriptive results, the multivariate analysis did not point to a decrease in the risk of work-life balance dissatisfaction after age 65. Indeed, after controlling for other factors and accounting for the self-selection of older individuals out of employment, no discernible difference in the risk of work-life conflict by age was found.
Further analysis revealed that correcting for selection mattered mostly for age. While estimates obtained from the probit model with selection were generally close to those produced by a probit model that did not account for selection (results not shown), this was not the case for the age variable. Whereas the model without selection still pointed to a decline in the probability of reporting work-life dissatisfaction with age, the model with selection indicated that age was related to the probability of working, but not to the probability of reporting work-life dissatisfaction.
Previous research suggests that women tend to experience higher levels of work-life conflict than men (Duxbury and Higgins 2008). This holds true among older workers too as women were 11 percentage points more likely than men to report dissatisfaction with work-life balance.
It was hypothesized that gender could be mediating the effects of some of the variables in our model (e.g. caregiver-related variables). However, tests for interactions between gender and these variables did not reveal any significant effects.
These findings are significant in light of the increasing presence of women among older workers. According to Labour Force Survey data, female representation among workers age 55 and over has been increasing steadily for more than three decades.8 If this trend persists into the future, it is likely to put upward pressure on the overall prevalence of work-life balance dissatisfaction among older workers.
Having a disability was associated with a higher chance of experiencing work-life conflict. Indeed, the probability of being dissatisfied with work-life balance was over seven percentage points higher for older workers with disabilities, relative to those without disabilities.
The association between disability and work-life conflict has also been reported in studies targeting the workforce age 15 and over (e.g. Frederick and Fast 2001).9 However, this finding is of special relevance in the context of older workers, given the sizeable proportion of individuals in this group reporting a disability (38%).
Elder care is frequently identified as a major source of tension in the work-life balance literature. It is often complicated by distance as the care recipients frequently live in different communities from the caregivers. Those providing ‘indirect' care from afar tend to experience feelings of guilt and increased stress. Furthermore, elder care providers typically have had to adjust their priorities, including spending less time with their own families, paying less attention to their own health, and taking fewer vacations (Duxbury and Higgins 2008, Duxbury and Higgins 2005).
The finding that elder care responsibilities place workers at a significantly higher risk of experiencing work-life conflict also applies to older workers. Older workers who provided four or more hours of elder care per week were over 14 percentage points more likely to report dissatisfaction with work-life balance relative to those without any elder care responsibilities. While few older workers spent this amount of time in elder care in 2005 (8%), this share is likely to increase in the future as Canada's population continues to age and the number of seniors who need support increases.
Work-life balance dissatisfaction among older workers varied significantly depending on occupation. Consistent with findings in the broader work-life balance literature (Skinner and Pocock 2008), those in managerial jobs faced the highest risk of experiencing work-life conflict. The probability of being dissatisfied with work-life balance was nine percentage points higher for managers, relative to workers in sales, service and manufacturing occupations. Jobs in social sciences, education, health and the arts were also associated with a higher probability of dissatisfaction.10
The strong positive association between the amount of hours worked and the likelihood of experiencing work-life conflict has long been established in the work-life balance literature (Kanter 1977). Working long hours limits the amount of time workers are physically available for family or other non-work-related activities (Voydanoff 1988). At the same time, high job demands can build up over time and hamper one's ability to function outside of work (Guerts and Demerouti 2003).
Older workers were no exception to this rule. Those working 50 or more hours per week were over 20 percentage points more likely to report dissatisfaction with work-life balance, compared with those working less than 30 hours per week.
The presence of a partner can have a mixed impact on work-life balance. On the one hand, marriage can increase demands outside of work while simultaneously decreasing the amount of control individuals have over their time. On the other hand, a spouse can be a source of emotional and tangible support in times of stress, thereby increasing an individual's sense of control (Duxbury and Higgins 2008).
In the case of older workers, having a spouse or common-law partner decreased the risk of work-life balance dissatisfaction, particularly if that partner was employed.11 Those with an employed partner were between 8 and 12 percentage points less likely to be dissatisfied with their work-life balance than those without a partner.12 In contrast, there was no discernible difference in the likelihood of work-life balance dissatisfaction between older workers with non-working partners and those without partners. These effects did not differ significantly for men and women.
Previous research suggests that enjoying work can reduce stress on time and work-family balance (Frederick and Fast 2001, Williams 2005). This seemed to be the case for older workers too. The probability of being dissatisfied with work-life balance was over 37 percentage points lower for those who were very satisfied with their jobs, relative to those who did not enjoy what they did.
Self-employment also appeared to lower the likelihood of work-life balance dissatisfaction among older workers, perhaps by allowing them to gain better control of their work activities relative to paid employees. The probability of being dissatisfied with work-life balance was six percentage points lower for the self-employed, relative to those who were in paid employment.
This contrasts with results from studies targeting the general workforce. Most notably, a recent Organisation for Economic Co-operation and Development (OECD) study of European workers age 15 to 64 found that being self-employed was significantly associated with increased conflict between work and family life (OECD 2004). Also, Skinner and Pocock (2008) found that paid employees and the self-employed in Australia were equally satisfied with their work-life balance.
Work-life balance studies targeting the younger workforce have consistently found a strong association between child care provision and the risk of work-life conflict. Interestingly, no discernible difference in the risk of work-life balance dissatisfaction was found between older workers who were participating in child care activities and those who were not. One potential explanation is the level of responsibility associated with this type of care. As noted earlier, older workers were likely to be providing care to children who were not their own (e.g. grandchildren), and, thus, presumably did not bear primary responsibility for this type of care in most instances. Consequently, older workers might have much more flexibility than their younger counterparts in terms of the timing and amount of child care they provide, thereby reducing the risk of conflict with their own work demands.
Overall, 14% of Canadian workers age 55 and over reported being dissatisfied with their work-life balance in 2005. Close to one-half of those who were dissatisfied felt they spent too much time on the job, while over one-quarter indicated that they did not have enough time for their families. Financial considerations—and not necessarily lack of employer flexibility—appeared to be a major factor making it difficult to cut back on hours.
Work-life balance dissatisfaction among these workers was associated with having a disability, providing elder care, working long hours, occupying a managerial position and being a woman. At the same time, having an employed partner, being self-employed and enjoying one's job reduced the probability of work-life conflict. When the self-selection of older individuals out of employment and other confounding factors were taken into account, the risk of work-life conflict did not vary with age.
The strong association between disability and work-life balance dissatisfaction, combined with the high prevalence of disability among older workers (38%), make disability a major piece of the work-life balance puzzle for this population. Further research on the particular barriers faced by older workers with disabilities and more information on how these vary with the nature of their disabilities would shed light on this major source of work-life conflict.
The higher risk of work-life conflict associated with the provision of four hours or more of elder care per week is also worth noting. While only 8% of older workers spent this amount of time in elder care in 2005, this share is likely to increase in the future as a result of population aging.
Extrapolating from ongoing trends, further increases in the share of older workers who are women combined with potential increases in the proportion providing elder care could make work-life conflict more prevalent among older workers in the coming years.
Data are from the 2005 General Social Survey (GSS) on time use (over a 24-hour period on a diary day). The GSS interviews Canadians age 15 and over in the 10 provinces on a wide range of social issues. In 2005, the sample size was 19,600. The target population of this study included all respondents age 55 and over who were working at the time of the survey—resulting in a sample of 1,832.
Work-life balance is based on self reports. The 2005 GSS determined satisfaction with work-life balance by asking “Are you satisfied or dissatisfied with the balance between your job and home life?” Respondents who indicated that they were ‘dissatisfied' were, then, asked eight questions regarding the reasons for their dissatisfaction.
Job satisfaction is measured in the GSS with a scale ranging from 1 (dislike the activity) to 5 (enjoy the activity). These ratings are combined into three categories: “unsatisfied with job” (a rating of 1 or 2), “relatively satisfied” (a rating of 3), and “very satisfied” (a rating of 4 or 5).
Child care activities stemmed from the following GSS questions: “Last week, how many hours did you spend looking after one or more of the children in your household, without pay?” and “Last week, how many hours did you spend looking after one or more children outside of your household, without pay?” Children are defined as being 14 years or younger.
Elder care activities were determined using the following GSS questions: “Last week, how many hours did you spend providing unpaid care or assistance to one or more seniors who live in your household?” and “Last week, how many hours did you spend providing unpaid care or assistance to one or more seniors who live outside your household?” Seniors are defined as being 65 years or older.
The non-senior adult care variable was constructed from the two elder care questions as well as nine other GSS variables. These variables indicated the time spent providing help or assistance to other adults in terms of personal care, medical care, housework, house maintenance, travel, correspondence and other care, as well as time spent caring for a disabled or ill person. Respondents who spent 30 minutes or more during the diary day in these activities and who did not report any elder care activity in the elder care questions were defined as “non-senior adult care providers.”
People with disabilities are those who reported that they had difficulty hearing, seeing, communicating, walking, climbing stairs, bending, learning or doing any similar activities; or who had a physical condition, mental condition or health problem that reduced the amount or kind of activity that they could do at home, at work, at school, or in other activities (like leisure or transportation). The 2005 GSS does not contain any information on the type, duration or severity of disability.
Probit regression estimates the probability of an outcome based on a set of explanatory variables. This technique allows the relationship between each explanatory variable and the outcome to be examined, while holding all other specified variables constant. This article uses a probit model with selection, allowing the estimation of the probability of work-life balance dissatisfaction controlling for selection out of employment (based on the method proposed by Van de Ven and Van Pragg 1981). Results are reported in terms of marginal effects—that is, the change in the predicted probability of being dissatisfied with work-life balance associated with a change in a given variable, controlling for all other explanatory variables in the model. Bootstrap weights are used to estimate the standard errors to account for the complex sample design of the GSS.
Canadian Council on Social Development. 2005. Employment and Persons with Disabilities in Canada. CCSD's Disability Information Sheet: Number 18, 2005. (accessed October 1, 2009).
Duxbury, Linda and Chris Higgins. 2008. Work-Life Balance in Australia in the New Millennium: Rhetoric Versus Reality. South Yarra, Australia. Beaton Consulting. 178 p. (accessed October 1, 2009).
Duxbury, Linda and Chris Higgins. 2005. Who is at Risk? Predictors of Work-Life Conflict. Report Four. Report prepared for the Public Health Agency of Canada. Ottawa. 124 p. (accessed October 1, 2009).
Duxbury, Linda and Chris Higgins. 2003. Work-Life Conflict in Canada in the New Millennium. A Status Report. Final Report. October. Report prepared for Health Canada. Ottawa. 154 p. (accessed October 1, 2009).
Frederick, Judith A. and Janet E. Fast. 2001. “Enjoying work: An effective strategy in the struggle to juggle?” Canadian Social Trends. No. 61. Summer. Statistics Canada Catalogue no. 11-008-XIE. p. 8-11. (accessed October 1, 2009).
Gardiner, Jean, Mark Stuart, Chris Forde, Ian Greenwood, Robert MacKenzie and Rob Perrett. 2007. “Work-life balance and older workers: Employees” perspectives on retirement transitions following redundancy.” The International Journal of Human Resource Management. Vol. 18, no. 3. March. p. 476-489.
Guerts, Sabine A.E. and Evangelia Demerouti. 2003. “Work/non-work interface: A review of theories and findings.” The Handbook of Work and Health Psychology. Second Edition. Marc J. Schabracq, Jacques A.M. Winnubst and Cary L. Cooper (eds.). Chapter 14. Chichester, England. John Wiley and Sons. p. 279-312.
Habtu, Roman and Andrija. Popovic. 2006. “Informal caregivers. Balancing work and life responsibilities.” Horizons. Vol. 8, no. 3. April. Policy Research Initiative. p. 27-34. (accessed October 1, 2009).
Hirsch, Donald. 2003. Crossroads After 50: Improving Choices in Work and Retirement. Joseph Rowntree Foundation. 58 p. (accessed October 1, 2009).
Kanter, Rosabeth Moss. 1977. Work and Family in the United States: A Critical Review and Agenda for Research and Policy. New York. Russell Sage Foundation.
Marshall, Katherine and Vincent Ferrao. 2007. “Participation of older workers.” Perspectives on Labour and Income. Vol. 8, no. 8. August. Statistics Canada Catalogue no. 75-001-XIE. p. 5-11. (accessed October 1, 2009).
Morissette, René, Grant Schellenberg and Cynthia Silver. 2004. “Retaining older workers.” Perspectives on Labour and Income. Vol. 5, no. 10. October. Statistics Canada Catalogue no. 75-001-XIE. p. 15-20. (accessed October 1, 2009).
Organisation for Economic Co-operation and Development (OECD). 2004. “Recent labour market development and prospects: Working hours and work-life balance.” OECD Employment Outlook 2004. Chapter 1. Paris. p. 46-52. (accessed October 1, 2009).
Raymo, James M. and Megan M. Sweeney. 2005. Work-Family Conflict and Retirement Preferences. On-line Working Paper Series. CCPR-035-05. California Center for Population Research. University of California - Los Angeles. 30 p. (accessed October 1, 2009).
Skinner, Natalie and Barbara Pocock. 2008. Work, Life and Workplace Culture: The Australian Work and Life Index 2008. Centre for Work + Life. Hawke Research Institute for Sustainable Societies. University of South Australia. 67 p. (accessed October 1, 2009).
Statistics Canada. 2008. Participation and Activity Limitation Survey 2006: A Profile of Assistive Technology for People with Disabilities. Statistics Canada Catalogue no. 89-628-XIE - No. 005. Ottawa. 27 p. (accessed October 1, 2009).
Statistics Canada. 2003. Participation and Activity Limitation Survey, 2001: Disability Supports in Canada, 2001. Statistics Canada Catalogue no. 89-580-XIE. Ottawa. 19 p. (accessed October 1, 2009).
Van de Ven, Wynand and Bernard Van Pragg. 1981. “The demand for deductibles in private health insurance: A probit model with sample selection.” Journal of Econometrics. Vol. 17, no. 2. November. p. 229-52.
Voydanoff, Patricia. 1988. “Work role characteristics, family structure demands, and work/family conflict.” Journal of Marriage and the Family. Vol. 50, no. 3. August. p. 749-61. (accessed October 1, 2009).
The authors would like to thank Maud Rivard, Christopher Poole, David Coish, Alex Grey and John Rietschlin for their comments and suggestions.
The views expressed in this article are those of the authors and do not necessarily reflect the views of Human Resources and Skills Development Canada.
Jorge Uriarte-Landa and Benoît-Paul Hébert are with the Policy Research Directorate, Human Resources and Skills Development Canada. They can be reached at 819-994-1757 and 819-934-5730 respectively. Both authors can be reached at firstname.lastname@example.org.