Chapter 12 The changing characteristics of older couples and joint retirement in Canada

The changing characteristics of older workers
Expectations of joint retirement
Conclusions
Bibliography

Increased labour force participation of women has been one of the most important social revolutions of the later half of the twentieth century in Canada and other industrialized countries. Women have become progressively more involved in career-oriented occupations and have accumulated continuous work histories. No longer are they marginal contributors to the family income or economically dependent on their husbands. They have become equal partners in most family decisions regarding economic issues, including those involving retirement (Henkens and Van Solinge 2002).

In this chapter, we examine one of the important decisions that most individuals involved in working couples have to make - whether to retire at the same time as their spouse. The relationship between women's level of economic independence and joint retirement intentions has not received much attention in Canada. What are the correlates of intentions of joint retirement in Canada? In what ways are men and women different (or similar) in deciding to retire at the same time as their spouses? Using the 2002 General Social Survey (GSS) data on 'near-retirees' aged 45 to 59, this study addresses these and other related questions.

We pay particular attention to the relationship between the characteristics associated with women's increased labour force participation - greater contribution to household income, increased pension coverage, and involvement in managerial and professional occupations - and their intentions to retire jointly with their spouse.

The chapter is divided into two sections. In the first section, we document changes in the characteristics of individuals and couples nearing retirement from the early 1980s to the early 2000s. Emphasis is placed on the changing employment characteristics of women and the implications this may have for couples approaching retirement. In the second section, we examine expectations of joint retirement among women and men in dual-earner couples and the factors associated with such expectations.

Top of Page

The changing characteristics of older workers

The employment and income characteristics of older women have changed significantly over the past two decades, altering the circumstances in which individuals and couples make retirement decisions. Women now approaching retirement have far greater attachment to the paid labour force over the course of their lives than did earlier cohorts. Between 1981 and 2004, the labour force participation rate of women aged 45 to 54 increased from 55% to 80%, and of those women who were paid employees¹ , the share with job tenures of 10 years or more increased from 30% to 48% (Table 12.1). Over this period, the share of women aged 45 to 54 working on a full-time basis increased from 73% to 79%, again testifying to greater labour force involvement.

The sectors of the economy in which 'older women' (defined as age 45 to 54) are employed have changed as well. The share of female employees aged 45 to 54 working in the public sector, broadly defined to include health, social services, education and related services, and government and religious organizations, increased from 37% to 45% between 1981 and 2004 (Table 12.1). This is an important consideration given that pension coverage is higher in the public sector than in most other industries (Statistics Canada 2004). Moreover, the share of older female employees belonging to a labour union increased from 33% to 40% over this period, with a large portion of this increase attributable to changes in the industrial profile of these employees (Morissette, Schellenberg and Johnston 2005).

Changes have also been evident in the types of occupations in which older women work. Between 1981 and 1998, the share of older female employees in managerial occupations increased from 7% to 16%, and the share in professional occupations increased from 20% to 27%. While clerical, sales and service occupations still account for a significant share of female employment, the growing importance of managerial and professional jobs is certainly clear. This is consistent with rising levels of educational attainment, as the share of older employees with a university degree increased from 8% to 21% between 1981 and 2004.

Women have also been making gains in earnings. Between 1980 and 2000 the share of female employees (aged 45 to 54) with annual earnings of $40,000 or more increased from 21% to 35% (Table 12.1). More broadly, Heisz, Jackson and Picot (2002) report that the weekly earnings of full-time female workers increased by 13% between 1989 and 2000, while the weekly earnings of comparable male workers remained unchanged. As a result, the gender earnings gap, while still persistent, narrowed over this period (Drolet 2001). Finally, gains have been made in pension coverage, as the share of female tax filers aged 45 to 54 reporting coverage increased from 42% to 46% between 1991 and 2001.

Sustained employment over the life course also means that women now approaching retirement have made larger contributions to private and public pensions than earlier cohorts, and consequently, will be eligible for larger retirement benefits.

Table 12.1
Selected characteristics of paid employees aged 45 to 54, Canada, selected years.

The changing labour market characteristics of women have altered the characteristics of couples approaching retirement. Of married and common-law couples in which the older partner is aged 45 to 54, the share of couples in which both partners worked a full year² increased from 27% to 46% between 1980 and 2000, and the share in which both partners worked full time and full year increased from 20% to 37% (Table 12.2).

The relative financial contributions of male and female partners have changed as well. Between 1980 and 2000, the proportion of couples (older partner aged 45 to 54) in which the female partner contributed 40% or more of the combined income of both partners more than doubled, rising from 19% to 43%. The share of couples in which the female partner contributed less than 20% declined from 56% to 27%.

Table 12.2
Married or common-law couples in which the older partner is aged 45 to 54, selected characteristics, Canada, 1980 and 2000.

All in all, the past two decades have witnessed substantial changes in the characteristics of individuals and couples approaching retirement. The cohort of individuals now poised to make the transition into retirement is comprised of a larger share of women than any previous cohort. In 1981, women accounted for 35% of employed persons aged 45 to 64, while in 2004 they accounted for 46%.³ Many analysts have shown that women and men tend to experience retirement differently. For example, women typically retire earlier than men.⁴ And among dual-earner couples, the retirement of women has typically been more contingent on the retirement of their husbands, while the retirement of men has been less contingent on that of their wives (Hurd 1990, Smith and Moen 1998, Henkens and Van Solinge 2002, Gustman and Steinmeir 2000). With the changing sex composition of retirees, the effects of differences in the retirement plans and expectations of women and men will be more far-reaching in the years ahead.

Furthermore, changes in the labour force characteristics of older women bode well for their imminent retirement transition. This point can be illustrated by briefly considering the experiences of women who retired between 1992 and 2002. Among this group, positive retirement experiences were most prevalent among those with higher levels of education, employment in managerial and professional jobs, and employment in the public sector (Table 12.3).

Women with these characteristics were more likely to have retired before age 60, were more likely to have retired voluntarily, and were more likely to have retired because they could afford to do so than women with lower levels of education, employment in sales and service jobs, and employment in consumer services. Other things being equal, the share of women positioned to make a favourable transition into retirement has been on the rise.

Table 12.3
Selected characteristics of retirement for female recent retirees, by education, occupation and industry, Canada, 2002

Top of Page

Expectations of joint retirement

A further implication of the changing labour force characteristics of older women pertains to the retirement decisions made by couples. Specifically, many couples must now make two retirement decisions rather than just one, and must balance the preferences and constraints faced by partners who both make substantial contributions to the household income. For women, greater investments in education and training, prolonged attachment to the paid labour force, employment in occupations with career trajectories, and significant earnings and retirement savings may translate into increased autonomy and independence in their retirement decision vis-à-vis their spouses. Thus, in addition to taking into account their husband's retirement plans, women must increasingly weigh the financial and non-financial returns of their own employment when making their retirement decision.

As O'Rand and Farkas (2002) suggest, "women's increased relative economic independence has, in all likelihood, influenced their retirement timing relative to that of their husbands." One way to gain further insights on this issue is to examine the prevalence and correlates of joint retirement. It is to this issue that we now turn.

A growing number of studies indicate that couples prefer to synchronize the timing of their exits from the labour force (Blau 1998, Drobni 2002, Henretta, O'Rand and Chan 1993, O'Rand and Farkas 2002). When 'joint retirement' is not possible, evidence suggests that later-retiring spouses tend to leave the labour force as quickly as possible, with later-retiring wives doing so at a faster rate than later-retiring husbands (Ruhm 1996, Henretta, O'Rand and Chan 1993).

Yet in spite of such preferences, many couples do not synchronize their labour force exits. In Canada, Gower (1998) reports that 33% of spouses in retired couples left the labour force within one year of each other, while 12% did so within the same month. In the U.S., Szinovacz (2002) reports that 42% of couples retire within one year of each other, while 56% retire within two years of each other⁵ . Of course, the incidence of 'joint retirement' depends on how wide a timeframe is used by researchers to capture the labour force exits of both spouses.

On the 2002 General Social Survey, non-retired individuals were asked a battery of questions regarding their plans and expectations regarding retirement. In this section, we focus on non-retired individuals aged 45 to 59 who are married or in a common-law relationship in which both partners are employed⁶ . These respondents were asked:

"Do you intend to retire at the same time as your spouse/partner?"
A logistic regression model was constructed using the yes/no responses to this question as the dependent variable. GSS respondents who said they did not intend to retire⁷ and those who did not state whether or not they intended to retire at the same time as their spouse/partner were excluded from the analysis.⁸ A total of 1,180 women and 1,215 men were included in the analysis. Separate regression models were run for women and men recognizing that labour market experiences, domestic roles and retirement transitions vary by sex.

A series of demographic, labour market and financial characteristics were included as independent variables in the model, the results of which are shown in Table 12.4. We begin by discussing the demographic characteristics associated with expectations of joint retirement, followed by a discussion of labour market and financial correlates.

Overall, expectations of joint retirement are quite prevalent among non-retired Canadians in dual-earner couples, with 48% of women and 46% of men saying they expect to retire at the same time as their spouse or partner. Consistent with other research, expectations of joint retirement are most prevalent among individuals who are the same age as their spouse, and decline as the age difference between spouses increases (Gower 1998, Szinovacz 2002, O'Rand and Farkas 2002). For example, the predicted probability of expecting to retire jointly is 60% among women who are the same age as their husband, while it is 39% for those who are five to nine years younger (Table 12.4). Expectations of joint retirement among men are also significantly correlated with the relative age of their spouse. Joint retirement may not be appealing to younger spouses because they do not think they will be psychologically ready to leave their job or will not be eligible to receive pension benefits at the time their older spouse retires.

Table 12.4
Predicted probability of expecting to retire at the same time as one's spouse, for near-retirees in dual-earner couples, Canada, 2000

For women, expectations of joint retirement are also associated with their current age. Women in their late-fifties are less likely to expect to retire jointly than women in their forties. This may reflect a selection effect, as older women who expected to retire jointly with their spouse have already done so. Alternatively, older women may have more conservative assessments of their prospects for joint retirement given that they are closer to the actual retirement transition than their younger counterparts.

In our analysis, individuals in common-law relationships are distinguished from those who are legally married. Marital dissolution is more prevalent among common-law than married couples (Lebourdais, Neill and Turcotte 2000) and common-law couples are less likely to have children (Statistics Canada 1997). In this context, one might expect that individuals in common-law unions approach retirement more independently vis-à-vis their partner than do individuals in legal marriages. However, the evidence does not support this hypothesis, as expectations of joint retirement are not significantly different between these groups.

In terms of family structure, expectations of joint retirement are less prevalent among women with two or more children at home than they are among women with less than two children at home. Compared with women who have few or no children at home, women with dependent children may face more financial demands and have less latitude in the timing of their retirement. The presence of children was not significantly associated with expectations of joint retirement among men. Joint retirement and health operates in a complex way (O'Rand and Farkas 2002). On the one hand, the poor health of one spouse may be associated with greater likelihood of joint retirement if the other spouse must withdraw from the labour force to provide care. On the other hand, poor health may be associated with reduced likelihood of joint retirement if health problems result in financial need due to lost earnings and/or high out-of-pocket health care costs that compel the other spouse to continue working.

Two health variables were included in our model: the respondents' self-assessment of their current health status and whether or not they expected their spouse's health to be a factor in their own retirement decision. Neither variable was significantly associated with expectations of joint retirement among women or men.

Finally, a question regarding the degree of control which individuals feel they have over decisions affecting their life was included. Many studies have found that sense of mastery is an important determinant of attitudes, behaviours and well-being (Soederberg Miller and Lachman 2000, Mirowsky and Ross 1989) and is a significant determinant of labour-market success (Dunifon and Duncan 1998). Women who say they have control over most decisions affecting their life were less likely to expect to retire jointly with their spouse than women who said they have control over some, few or none of these decisions (predicted probabilities of 60% and 45% respectively). This may reflect traditional gender roles, insofar as women in households where husbands are the primary decision-makers may also be those who are most likely to make retirement decisions based on their husband's health and retirement status.

Turning to labour force characteristics, expectations of joint retirement among women vary across occupational groups. The predicted probability of expecting to retire jointly with one's husband is 44% among women in managerial, professional and technical occupations, compared with 56% among women in sales and services occupations. In short, women in occupations typically characterized by career trajectories and higher earnings are more likely to view their retirement transition as independent of their spouse. This may reflect differences in career orientations, career trajectories or job satisfaction among women in these groups.

Expectations of joint retirement among women varied markedly between paid employees and self-employed workers. The predicted probability of expecting to retire jointly is 68% among self-employed women compared with 44% among paid employees. Self-employed women may have greater latitude over how and when they retire and hence may be better able to time their retirement to coincide with that of their spouse. Participation in a family-run business in which both spouses work and from which both plan to retire may be another explanation. Evidence from the Census provides some clues in this regard. In 2001, there were just over 170,000 couples in which the older spouse/partner was aged 45 to 54 and in which the female spouse/partner was self-employed. In 53% of these couples (90,700), the male spouse/partner was also self-employed, while in 45% of these couples the male spouse/partner was a paid employee.⁹ The key point is that a substantial share of older women who are self-employed are married to self-employed husbands. ¹⁰ Although we do not know if these couples are involved in the same business or enterprise, it does appear that for self-employed women retirement is a family affair with labour force exits synchronized with their spouses'.

Turning to income characteristics, pension coverage is negatively correlated with expectations of joint retirement among women. Those with coverage have a predicted probability of joint retirement of 44% compared with 53% among those without pension coverage. It appears that pension coverage constrains the latitude that women have for joint retirement. Wives are often two to three years younger than their husband and significant financial costs would be incurred by them if joint retirement meant they would have to leave the labour force before becoming eligible for a retirement benefit, or meant they would have to continue working after becoming eligible for a retirement benefit. Interestingly, the negative correlation between joint retirement and pension coverage is not evident among men, suggesting that women and men have different assessments of the constraints that pension coverage will have on the timing of their retirement relative to their spouse.

The total income received by individuals was not significantly associated with expectations of joint retirement. However, the proportion of household income contributed by individuals was significant. Women who contributed 60% or more of the household income were less likely to expect to retire jointly with their spouse than women who contributed less than 40% (predicted probabilities of 53% and 42% respectively). It would appear that shouldering responsibility for the household income diminishes the scope that married women feel they have to retire at the same time as their spouses. Again, however, this relationship is not evident among men. Finally, women's perceptions that adequate financial preparations for retirement were being made were positively correlated with expectations of joint retirement.

Overall, many of the characteristics that one might associate with greater economic independence among women are negatively correlated with expectations of joint retirement. Women in managerial and professional occupations, those with pension coverage, and those who contribute most of the household income are more likely than others to view the timing of their retirement in an autonomous manner vis-à-vis their spouses. Furthermore, many of the characteristics associated with expectations of joint retirement among women are not significantly correlated with such expectations among men. This is consistent with the view that women and men approach retirement in different ways (Smith and Moen 1998, Henretta, O'Rand and Chan 1993).

Top of Page

Conclusions

The past two decades have witnessed significant changes in the labour market and in the financial characteristics of women in their forties and fifties. This has been evident in terms of women's attachment to the paid workforce over the course of their lives, the characteristics of the jobs they hold, and their earnings. Broadly speaking, these trends have positive implications for the financial well-being of the next generation of seniors in Canada.

The changing employment characteristics of women have also reshaped the profile of couples approaching retirement. A cohort of well-educated, dual-earner couples at the top of the earnings distribution is now poised to make the transition into retirement (Morissette and Johnson 2004). Increasingly, couples must make two retirement decisions rather than just one. In this respect, the 'baby boomers' are different from previous cohorts.

It is interesting to note that while slightly less than half of married individuals now approaching retirement say they intend to retire at the same time as their spouses, evidence from the mid-1990s shows that only one-third of married Canadians actually retired within one year of their spouse (Gower 1998). It appears that joint retirement intentions are more prevalent than joint retirement transitions. One reason may be that some couples are unable to achieve their preference for joint retirement because of financial constraints (such as pension eligibility criteria or financial imperatives to continue working) or because of non-financial constraints (such as career goals or job (dis)satisfaction).

Evidence from the 2002 GSS suggests that the characteristics one might associate with economic independence, such as employment in managerial and professional occupations, pension coverage, and a majority contribution to the household income, are positively associated with women viewing retirement in an autonomous manner vis-à-vis their spouses. This is consistent with the view that as women make gains in the labour force and contribute more to the family income, their own labour market and financial circumstances will weigh more heavily in their retirement decisions.

Our analysis underscores the differences in the way that women and men view the retirement transition. While several demographic, labour market and financial characteristics are significantly associated with joint retirement expectations among women, very few of these characteristics are significantly associated with joint retirement among men. As often noted in the literature, models of retirement that fail to take differences between women and men into account are likely to offer only partial accounts of this transition.

Other researchers have documented some of the implications of joint retirement within the household. Moen, Jungmeen and Hofmeister (2001) report that the transition from a career job into retirement is associated with a decline in marital quality, and that such a decline is most often reported by men and women who retire while their spouses remain employed. Their analysis underscores the correlation between marital conflict and incongruence in the work/retirement circumstances of husbands and wives. Szinovacz and Davey (2005) report that husbands find it difficult staying home alone after retirement when their wives are still working. And Smith and Moen (2004) report that satisfaction with retirement is less prevalent among retirees whose decision to leave the labour force was greatly influenced by their spouse.

Joint retirement has broader implications as well. With the imminent retirement of the baby boom generation, increasing emphasis is being placed on the role that public policies and employer strategies might have in encouraging older workers to remain on the job (OECD 2002). Older women ought to be central to such strategies given that they account for a sizeable pool of well-educated, experienced workers. This is particularly the case in the public sector where relatively large shares of workers will become eligible for retirement in the years ahead (Henry 2004).

Top of Page

Bibliography

Blau, David. 1998. "Labor force dynamics of older married couples." Journal of Labor Economics. 16: 595 to 629.

Drobni, Sonia. 2002. "Retirement timing in Germany." International Journal of Sociology. 32: 75 to 102.

Drolet, Marie. 2001. The Persistent Gap: New Evidence on the Canadian GenderWage Gap. Catalogue no. 11F0019MIE2001157. Ottawa. Statistics Canada.

Dunifon, Rachel and Greg J. Duncan. 1998. "Long-run effects of motivation on labor market success." Social Psychology Quarterly. 61: 33 to 48.

Gower, David. 1998. "Retirement patterns of working couples." Perspectives on Labour and Income. 10, 3: 27 to 30. Statistics Canada Catalogue no. 75-001-XPE.

Gustman, Alan L. and Thomas L. Steinmeir. 2000. "Retirement in dual-career families: A structural model." Journal of Labor Economics. 18: 503 to 545.

Heisz, Andrew, Andrew Jackson and Garnett Picot. 2002. Winners and Losers in the Labour Market of the 1990s. Catalogue no. 11F0019MIE2002184. Ottawa. Statistics Canada.

Henkens, Kène and Hanna Van Solinge. 2002. "Spousal influence on decision to retire." International Journal of Sociology. 32: 55 to 74.

Henretta, John C., Angela M. O'Rand and Christopher G. Chan. 1993. "Joint role investments and synchronization of retirement: A sequential approach to couples' retirement timing." Social Forces. 71: 981 to 1000.

Henry, Pold. 2004. "The near-retirement rate." Perspectives on Labour and Income. 5, 2, February: 18 to 22. Statistics Canada Catalogue no. 75-001-XWE2004102.

Hurd, Michael D. 1990. "The joint retirement decision of husbands and wives." Pp. 231 to 258 in Issues in the Economics of Aging. David A Wise (ed.). University of Chicago Press.

LeBourdais, Céline, Ghyslaine Neill and Pierre Turcotte. 2000. "The changing face of conjugal relationships." Canadian Social Trends. 56, Spring: 14 to 17. Statistics Canada Catalogue no. 11-008-XIE.

Mirowsky, J. and C.E. Ross. 1989. Social Causes of Psychological Distress. Hawthorne, NY. Aldine.

Moen, Phyllis, Jungmeen Kim and Heather Hofmeister. 2001. "Couples' work/retirement transitions, gender, and marital quality." Social Psychology Quarterly. 64, 1: 55-71.

Morissette, René and Anick Johnston. 2004. The Earnings of Couples with High and Low Levels of Educational Attainment, 1980 to 2000. Analytical Studies Branch Research Paper Series. Catalogue no. 11F0019MIE2004230. Ottawa. Statistics Canada.

Morissette, René, Grant Schellenberg and Anick Johnston. 2005. "Diverging trends in unionization." Perspectives on Labour and Income. 17, 2, Summer: 1 to 8. Statistics Canada Catalogue no. 75-001-XPE.

Organisation for Economic Co-operation and Development (OECD). 2002. "Editorial." Employment Outlook. 7 to 12.

O'Rand, Angela and Janice Farkas. 2002. "Couples' retirement timing in the United States in the 1990s." International Journal of Sociology. 32, 2, Summer: 11 to 29.

Ruhm, Christopher J. 1996. "Gender differences in employment behaviour during late middle age." Journal of Gerontology: Social Sciences. 51B: S11 to S17.

Schellenberg, Grant. 2004. The Retirement Plans and Expectations of Non-Retired Canadians Aged 45 to 59. Analytical Studies Branch Research Paper Series, no. 223. Catalogue no. 11F0019MIE2004223. Ottawa. Statistics Canada.

Smith, Deborah and Phyllis Moen. 2004. "Retirement satisfaction for retirees and their spouses." Journal of Family Issues. 25, 2, March: 262 to 285.

Soederberg Miller, L.M.S and M.E. Lachman. 2000. "Cognitive performance and the role of control beliefs in midlife." Aging, Neuropsychology, and Cognition. 7: 69 to 85.

Statistics Canada. 2004. Pension Plans in Canada, January 1, 2003. Catalogue no. 13F0026MIE2004001. Ottawa.

Statistics Canada. 1997. "Common-law unions in Canada at the end of the 20th century." Report on the demographic situation in Canada,1996. Catalogue no. 91-209-XPE. Ottawa

Szinovacz, Maximiliane E. 2002. "Couple retirement patterns and retirement age: A comparison of Austria and the United States." International Journal of Sociology. 2, 2, Summer: 30 to -54.

Szinovacz, Maximiliane E. and Adam Davey. 2005., "Retirement and marital decision making: Effects on retirement satisfaction." Journal of Marriage and the Family. 67: 387 to 398.

Notes

1. In 1981 and 2004, the shares of employed women aged 45 to 54 who were self-employed were 12.4% and 13.6% respectively.
2. Full-year employment is defined as 49 weeks or more.
3. The figures are similar if we consider employed persons with a university degree. In 2004, women accounted for 44% of employed persons aged 45 to 64 who had a university degree, up from 32% in 1990.
4. In an Ordinary Least Squares Regression model, women retired about two years earlier than men after taking into account other factors such as occupation, industry, pension coverage and marital status.
5. In the data used by Szinovacz (2002), month of retirement was not available, so the possible range for the retirement date of each spouse was actually 0 to 35 months.
6. 'Employed individuals' were identified as those who said their main acitivity during the previous 12 months was 'Working at a paid job or business'. Also included were individuals who said their main activity was something other than paid work, but who had at least 26 weeks of employment during the year. To be included in the analysis, individiuals also had to declare that their spouse's main acitivity during the previous 12 months was 'Working at a paid job or business".
7. A discussion of the characteristics of this group can be found in Schellenberg (2004).
8. A total of 176 (or 7%) respondents who were potentially in-scope for the analysis said they did not know if they intended to retire at the same time as their spouse or refused to answer the question.
9. In two percent of these couples, the male spouse/partner was not in the labour force.
10. Similarly, in 2001 there were 307,000 couples in which the older spouse/partner was aged 45 to 54 and in which the male spouse/partner was self-employed. In 30% of these couples (90,700) the female spouse/partner was self-employed, in 60% the female spouse/partner was a paid employee, and in 10% the female spouse/partner was not in the labour force.