Working after childbirth: a lifecourse transition analysis of Canadian women from the 1970s to the 2000s.
Gaudet, Stephanie ; Cooke, Martin ; Jacob, Joanna 等
A FAMILIAR THEME EMERGING FROM POLICY studies and gender analyses
over the past few decades is the growing number of women in the paid
labor market, with the most notable rise among women with young
children. In 2001, 80 percent of Canadian women aged 25 to 54
participated in the paid labor market, placing Canada near the Nordic
countries in this regard (Jaumotte 2003). Between 1976 and 2004, the
largest increase in employment1 has been among mothers with children
under the age of three (Statistics Canada 2005a). This supports the now
common assumption that a majority of Canadians are living in "full
adult worker" families. The abandonment of the previous "male
breadwinner" model is an important aspect of the new welfare
regime, prevalent in several Organisation for Economic Co-operation and
Development countries, in which policies regarding work and care giving
explicitly or implicitly assume women's full labor market
participation (Lewis and Giullari 2005).
Questions remain concerning how different cohorts of Canadian women
have experienced these labor market transitions after childbirth. These
changes have taken place over time and, although women's employment
participation rates are high, Canadian men and women still have
different work trajectories, mainly because the paid work careers of
many women are interrupted by childbearing and caring work in a way that
men's usually are not (Baker 2005; Kempeneers 1987). Although
interruptions are often temporary and facilitated by maternity and
parental leave policies, some women remain out of the labor force beyond
the time supported by paid leave, and others do not return to work for
years after the birth of their children (Marshall 1999; Zhang 2007).
Widespread employment of mothers has been a fact since the 1970s,
but we know little about the evolution and the variations of this social
phenomenon. It is important to understand how this lifecourse transition
has taken place across cohorts in the Canadian context. For example,
comparing the likelihood of returning to work for women who had their
children in the early 1970s to later childbearing cohorts can provide
insight about how these different cohorts of women will enter older
ages, as lifelong labor force activity affects income later in life.
One reason for the lack of knowledge about these phenomena is a
dearth of longitudinal data in Canada, especially data that extend to
periods before the 1990s (Bernard 2003; Kempeneers 1989). Long-term data
are important in order to understand the specific patterns that have
arisen in the last 30 years as Canadian families have progressed from a
strongly gendered role demarcation model toward one of greater equity.
In this article, we focus on the variation of the transition to paid
work after childbirth among Canadian women who had their first children
between 1970 and 1999. We use retrospective data from the 2001 General
Social Survey (GSS) to examine women's entry to work after the
birth of a first child, and how this has changed since the early 1970s.
We do not directly address the length of women's labor market
withdrawal for childbearing, but rather the probability that a woman
will be engaged in paid work within two years following first birth.
This time period might appear arbitrary; however, this is a common
approach used by researchers in various countries to measure
postchildbirth employment (Hofferth and Curtin 2006; Marshall 1999).
This period is long enough to include most women who planned to return
to work after childbirth, and is longer than most standard maternity
leaves. Our analysis is grounded in a lifecourse perspective, and is
focused on individual transitions and characteristics that influence
trajectories, as well as the social and historical context that
structures individual lives.
MOTHERS IN THE LABOR MARKET
Breaks in employment can have strong implications for women's
individual lifecourses, as well as for gender equity in employment in
aggregate. The length of an absence from paid work after childbirth is
of critical importance to individual women's long-term economic
well-being, their opportunities for human capital development (Kenjoh
2005), their subsequent professional mobility (Felmlee 1995) and
eventual retirement. It also has important effects on families'
well-being. Women's full-time employment before birth and quick
return to work after birth have been found to be strongly protective
against poverty among Canadian families (Juby et al. 2005). It is
therefore important to understand the characteristics of women who are
more likely to remain out of the workforce for some time.
Since the 1990s, the majority of postindustrial countries have been
developing social policies based on the "full adult worker
family" model (Haas 2005; Jenson and Sineau 2001; Lewis and
Giullari 2005). This evolution is rooted in a change in values, but also
in economic pressures and macrosocial change (Ranson 2005). Changes in
family structures, including the increase in lone parent families headed
by women, have also helped to normalize the idea of the financial
independence of women. Equal opportunity, equal citizenship and gender
equity objectives have led governments to promote full employment for
women (Stoltzfus 2003), and Canada is in line with this trend (Baker
2005; Doucet 2010).
Labor market participation by mothers is the new reality facing
contemporary societies (Hochschild 2003) and an issue of growing
significance for public policy in western and nonwestern countries
(Heymann 2006). Several researchers point out that, although women are
approaching levels of labor force participation previously associated
with men, some aspects of social organization are still based on a male
breadwinner model (Lewis and Giullari 2005). For example, most unpaid
work at home is still done by women (Doucet 2010; Fast and Frederick
2004; Kershaw 2005), and women predominate among part-time workers
(Beaujot 2000; Phillips and Phillips 2000). One of the challenges in
understanding this phenomenon is to be aware of the disparity between
the normative policy goals of the full adult worker model and the
empirical reality of mothers (Crompton 2006). The increases in
mothers' employment soon after childbearing since the 1970s may
lead analysts, employers, or policymakers to forget this disparity.
TRANSITIONS FROM A LIFECOURSE PERSPECTIVE
The question of women's return to work after childbirth is
well suited to a lifecourse approach because this key transition is at
the nexus of individual biography and social structure (Reiter 2009).
Lifecourse is a multidisciplinary perspective that reflects the
"intersection of social and historical factors with personal
biography" (Elder 1985:24). Hence, lifecourse patterns and the
timing of "transitions" or changes in status or roles
(MacMillan and Copher 2005) are expected to fluctuate across time,
space, and populations. The lifecourse focuses on how the experience of
transitions in interrelated "domains," including the family,
work, and employment, affect one another (Elder 1985; Kruger and Levy
2001). This approach helps to connect research on the labor market to
that on the family, and provides an especially appropriate framework for
those who wish to investigate the relationships between women's
childbearing and employment transitions.
For individual women, the employment transition following a first
birth is among the key turning points in the lifecourse. The birth event
constitutes a more significant rupture in the life trajectories of women
than it does in those of men, and for some women it remains associated
with an extended absence from the labor market, resulting in a loss of
long-term human and social capital. These work interruptions related to
childcare can lead to lower incomes over the long term, and contribute
to overall gender disparity in income (Phipps, Burton, and Lethbridge
2001), although analyses show that the gap between Canadian mothers and
nonmothers income has decreased over the last decade (Zhang 2007).
Furthermore, the career trajectories of women with young children
may suffer when they have the primary responsibility for rearranging
their work schedules to care for children who are ill, and when they are
the partners performing a greater share of the unpaid work in the home.
From a lifecourse perspective, these women capitalize their resources,
in the sense that they continue to gain human and financial capital by
maintaining their attachment to paid work. However, they face different
challenges from those of men, even if they have similar labor force
attachment, including issues related to workplace equity, accessibility
of daycare services, overtime work culture (Blair-Loy and Jacobs 2003;
Wharton and Blair-Loy 2006), and wage equity (Fortin and Huberman 2002).
At least in the short term, they are also more likely to have health
problems related to stress, particularly as they balance childcare and
paid work (Duxbury and Higgins 2002, 2003; Higgins, Duxbury, and Lyons
2005).
Traditionally, lifecourse analyses have mainly documented the
sequence and the timing of transitions, including how these patterns
have changed over time (George 1993). However, more recent lifecourse
research has turned to understanding social differentiation within key
life transitions (Dewilde 2003; Hynes and Clarkberg 2005; MacMillan
2005). Instead of merely analyzing the sequencing of life events, these
researchers are investigating the types of lifecourse trajectories and
are relating them to socioeconomic characteristics, such as education or
social class (Hogan and Astone 1986).
With respect to this particular lifecourse transition, it is
important to consider not only how women's back-to-work transitions
have changed, but also whether the characteristics that predict this
transition have also changed. In other countries, studies have been
conducted on cohort changes in women's work transitions following
childbirth. Stanfors (2006), using multinomial logistic regression,
compared the odds of part-time paid work, full-time work, or not working
at all after childbirth, for five cohorts of Swedish women born between
1949 and 1969. These models found significant cohort differences, even
after controlling for previous labor force participation, number of
children, parents' employment, religiosity, and self-reported
"career orientation." In one U.K. study, Smeaton (2006)
compared two birth cohorts of women (aged 23 in 1988 and 2000) in terms
of their odds of returning to work one year after childbirth. While
there were expectations of an increasing polarization of the timing of
women's return to work, due partly to the cost of childcare, their
results showed a reduction in the importance of women's social
class in predicting this work transition. However, a gap remains in the
Canadian research regarding the questions of how women's
postchildbirth employment has changed between cohorts.
THE EFFECTS OF STRUCTURAL CHANGE IN THE CANADIAN LABOR MARKET ON
WOMEN's EMPLOYMENT TRANSITIONS (1970s-1990s)
The sociohistoric period in which women take the key lifecourse
decision to work after childbirth is important, as a key element of the
lifecourse perspective is how individual biographies are shaped by
particular social and historical contexts. In this paper, we focus our
analysis on the 30 years from 1970 to 1999--a period in which important
changes in the economy and the labor market took place.
The recessions of the early 1980s and early 1990s may have been
particularly influential on labor market transitions. Some have
identified 1979 as the beginning of the economic change of the period,
as the postwar increase in real wages ceased.
Since the vast majority of Canadian families depend on labour
market earnings as their primary income source, trends in the level of
real wages are key to their economic well-being--and the crucial novelty
of Canadian economic events since 1979 is the fact that the growth of
average real wages largely halted. (Osberg 2007:14)
The labor market participation of Canadian women rose continuously
between the early 1960s and early 1990s. Various factors explain this
trend: economic needs, individual values and attitudes, and social
changes. Among those, high rates of family dissolution, women's
increasing educational attainment, and gender equality claims in the
workplace rerated to the second-wave feminist movement (1960-1990) have
been key influences on women's labor market behavior.
While paid employment had previously been largely closed to
mothers, the 1970s labor market opened up to women with school- and
preschool-aged children. However these changes took place very slowly
throughout the 1970s and 1980s. As of 1976, 60.9 percent of women under
55 years old and without children were working, compared with just 27.6
percent of mothers with children under three. It has only been since the
mid-1980s that a majority of mothers of preschoolers were employed in
the labor market. The year 1985 represents the milestone at which a
majority (52 percent) of mothers with children aged three to five were
employed; 1987 marked a similar point for mothers with children under
three (Statistics Canada 2005b).
As well as providing a reminder that mothers' widespread labor
force participation is a relatively recent phenomenon, these figures
demonstrate how women's trajectories in paid work are linked to the
institutionalization of children's lifecourses. In short,
women's employment is facilitated by the availability of caring
services for children. In 1991, only 22 percent of preschool-aged
Canadian children had a daycare space, and the greatest shortage in
space availability was for infant care (Phillips and Phillips 2000).
The period of expansion in women's labor force participation
ended in the early 1990s, a time of economic crisis, high unemployment
rates, and a dramatically changing labor market. New technology,
just-in-time management, downsizing, and out-sourcing to developing
countries with far lower labor costs were affecting the labor market
demand in Canada. Over the 1990s, many Canadians lost their "good
jobs" in unionized workplaces, especially manufacturing. A new
"flexible" labor market with many "bad jobs,"
including lower pay, short-term contract and self-employment, and
nonstandard working hours began to expand (Canadian Economic Council
1990).
EFFECTS OF INDIVIDUAL CHARACTERISTICS ON MOTHERS' EMPLOYMENT
TRANSITIONS
Given the changes in the patterns of women's labor force
participation and the possibility of the variation of women's
lifecourses in this regard, an important research issue is the
identification of the various factors that may lead to different work
trajectories. Although many scholars would emphasize the importance of
structural class and gender effects on the employment transitions of
mothers (Kempeneers 1987), the majority of empirical studies examining
this transition are rooted in neoclassical economic theories based on
rational choice and preference (Charles et al. 2001). Viewed from the
perspective of the dominant ideology, women are seen as benefiting more
than their husbands from involvement in nonremunerated caring work,
while men's higher incomes reinforce their continuous labor force
attachment within the family strategy (Becker 1991). Women's
decisions to leave work for care giving or to work part-time in order to
combine work and family roles are therefore cast as a manifestation of
women's (and men's) preferences in light of the available
options (Hakim 1995, 1996, 1998). Similar Canadian results have
confirmed the fundamental idea of Hakim's research in the United
Kingdom--that women make rational choices regarding their working time,
considering the utility of childcare and the available wages (Cleveland,
Gunderson, and Hyatt 1996).
This theory contrasts somewhat with theories that explain the
choice women make in terms of their lifecourses, their occupations, and
national policies, as well as gender and class structures (Crompton and
Harris 1998). We share this sociological perspective by using a
lifecourse analysis in which both structure and individual agency are
taken into account. The choices women make remain constrained by
structurally determined employment opportunities and also by the points
in their lifecourses at which they find themselves faced with employment
decisions (Houston and Marks 2003; Lewis and Giullari 2005; Marks and
Houston 2002).
Important inequalities between men and women, and particularly
between mothers and fathers, still prevail in the labor market, and
gender ideology forms a social-structural context for these decisions
(Crompton 2006; Zhang 2007). Studies of women's work intentions can
demonstrate the relevance of structural determination. Crompton, for
example, found that a number of women had not managed to obtain the
employment conditions after childbirth (in terms of full- or part-time)
that they had hoped for while pregnant, suggesting that there can be a
wide gulf between women's "preferences" and their
realities (Crompton 2002). Similarly, Tomlinson (2006) examined British
women's preferences in choosing to work full-time or part-time and
found that they are shaped by a number of factors, including the policy
context and their own care networks. She concluded that women make
"reactive" or "compromised" choices that affect
their subsequent trajectories.
Other factors are often associated with mothers return to work such
as their occupational qualifications and educational attainment
(Elliott, Angela, and Egerton 2001); culture and religious values (Glass
and Leda 2006); the educational attainment of the maternal grandmother
and her employment trajectory (Sanders 1997); age at first birth
(Elliott et al. 2001; Hynes and Clarkberg 2005); the availability of
personal networks for care giving; and the presence of a spouse actively
involved in childcare (Feeney et al. 2003; Helms-Erikson 2001). Support
from the workplace and access to quality childcare services are also key
considerations influencing a mother's decision to return to work
(Barrow 1999; Killien 2005), as well as family policies such as
maternity/parental leave and affordable quality daycare. Owing to
methodological constraints, in this paper we will focus on the effects
of women's education and training, their age at childbearing,
employment before childbearing, and spousal incomes.
Mothers' Education
In research undertaken in several countries, women's education
(defined by credentials and training) is related to the likelihood of
working after childbirth. In many countries, it would appear to be the
most significant factor affecting reentry into the labor market after
birth (Elliott et al. 2001). However, understanding the influence of
mothers' education is complex because education and credentials can
have a mixed impact on transitions, depending on the type of jobs and
workplaces.
In a comparative study of five European countries illustrating
different types of welfare regimes, researchers found that it was only
in Sweden that women with university degrees did not return to work
sooner than those without (Gutierrez-Domenech 2002, 2005). This is
noteworthy because this prototypical social-democratic country has a set
of family policies based on employment equity for men and women and on
well-integrated work-life balance provisions. These data suggest that in
other types of welfare state, university-educated women's
transition to paid employment is quicker than those without a university
degree. Thus, women without university degrees in countries where social
policies are less egalitarian seem to have fewer incentives to return
quickly to work. If we consider education as a proxy for social class,
then this may suggest a growing class differentiation in these
lifecourse transitions in these countries.
This phenomenon may be due to a number of factors. University
training is often associated with better salaries as well as with
qualitative factors that might affect work attachment, such as autonomy
and job satisfaction. Armstrong and Armstrong (1994) explain that women
often work in jobs with fewer opportunities to regulate their own hours
of work, less decision-making power, and fewer outlets for
creativity--factors that are usually related to job satisfaction. This
lack of control over work might be a barrier for some women trying to
balance work and family, and women with higher education may be more
likely to have the types of jobs that offer the autonomy to make this
balance, or more ability to change their jobs in favor of better
arrangements. The relationship between these aspects of work and
education is not straightforward, however, and has recently been
changing. Although the working hours of Canadian women have generally
been longer for those with higher education, the gap in the number of
hours worked between college-educated women and those with university
degrees almost disappeared between 1997 and 2006. This was due in large
part to a significant decrease in the proportion of women with
university degrees working more than 41 hours/week (Statistics Canada
2008).
Analyses based on longitudinal data from Britain (Elliott et al.
2001) shed light on important distinctions with respect to women's
education and their return to work after childbirth. Specifically, it
appears that it is not so much education as the link between education
and a given segment of the employment market that influences
women's return to work. British women with occupationally specific
degrees, such as architecture, nursing, and education, return to work
faster than women with nonoccupational degrees, such as degrees in
administration, literature, and so on. Again, job satisfaction or the
requirement of hours of practice to keep a professional license to
practice might influence this transition.
Few Canadian empirical studies exist on the influence of education
on this specific life transition. Using data from the 1993 to 1996
Survey of Labour and Income Dynamics, Marshall (1999) found no
significant and independent effect of education (high school or less
versus higher education) on the likelihood of Canadian women returning
to work "quickly" (within one or two months), or within two
years of childbirth. However, she did find that women who had shorter
tenure in their jobs and who were nonunionized or part-time workers were
more likely to return to their jobs quickly--within two months of
childbirth.
Using U.S. data for women born between 1958 and 1965, Han et al.
(2008) found that, among women who gave birth in 2001, older women and
those with degrees above the bachelor's level were more likely than
younger or less-educated women to have returned to work nine months
after birth. However, at two months, the younger and less-educated women
were more likely to be employed. The authors speculate that women with
more work experience and better education may be more likely to have
savings or maternity leave available, which might make them more likely
to return eventually, but would also allow them to delay this
transition. Moreover, the employability of these women is greater in a
competitive labor market. They have better chances of finding a new
position or of returning to their employment even if they delay their
return. Women with weaker labor force positions may both need to return
in a relatively short period and be less likely to find work if they
extend their leave and miss job opportunities. Obviously the U.S. policy
context is different from that of Canada. More limited maternity leave
programs and private childcare markets in the United States result in
differences in returns to work.
Age at First Birth
Since the 1970s we have seen a significant change in life
trajectories for women, as the average age of entry into a conjugal
relationship and parenthood continues to increase (Ravanera, Rajulton,
and Burch 1998). The lengthening of the transition to independence has
resulted in ever more Canadian women having their first child after the
age of 30. In 1983, only 14 percent of first births were to mothers over
the age of 30, compared with 48 percent in 2003 (Statistics Canada
2005a). These temporal changes in the entry into adulthood are mainly
explained by the longer period young adults, especially young women,
spend in postsecondary education (Cote 2006). These changes appear to
have consequences for women's employment transitions after giving
birth as well, because highly educated young women are more likely to
have children later in life, to remain in stable couple relationships,
and to live in high-income households. All these factors are associated
with a fairly quick return to the labor market after childbirth (Juby et
al. 2005). A number of studies correlate having a child later in life
with a greater likelihood of labor market reentry following childbirth
(Han et al. 2008; Houston and Marks 2003). However, Marshall's
(1999) analysis found no significant independent relationship between
age and returning to work within 24 months of childbirth, among Canadian
women who had given birth in the period between 1993 and 1996.
Employment Status before First Birth
Employment status before childbearing is obviously an important
factor influencing women's subsequent work transitions. This is
true for most lifecourse trajectories in which previous conditions,
statuses, and roles form the context of later choices. In the case of
U.S. women's transitions to work after childbearing, Hynes and
Clarkberg (2005) found that college degree-holding women working in
full-time jobs were more likely to promptly reenter employment than were
part-time workers or women who were outside the labor market before
delivery. Women's career or job satisfaction might also be a strong
factor influencing the length of her absence from the labor market after
childbirth.
Part of the effect of prechildbirth employment on later transitions
may be that having a job immediately before childbearing may allow
access to maternity leave, which not only facilitates temporary
withdrawal from work by replacing income, but also has a "job
holding" function, encouraging return to the same employer. There
is considerable international evidence that access to parental or
maternal leave increases women's rate of return to work after
childbearing (Baker 2005; Barrow 1999; Burgess et al. 2008; Haas 2005;
Hofferth and Curtin 2006; Klerman and Leibowitz 1999; Zhang 2007).
Family Policies
The effects of social policies on this key life transition have
been of interest to many scholars. Three types of policy arrangements
might be of importance in mothers' behavior: maternity and parental
leaves, childcare policies, and early childhood benefits. In the
Canadian context, maternity leave was been extended in all provinces
during the 1980s, although there were important provincial variations.
In 1971, 15 weeks paid maternal leave was introduced in Canada for
female employees, followed by the addition in 1990 of 10 weeks of
parental leave to be taken by either parent. By the end of the 1970s,
all provinces had passed laws ensuring women employees had access to at
least 17 weeks of unpaid maternity leave (Baker 2005). All of these
programs excluded nonstandard workers who did not contribute to
Employment Insurance, however. In 2006, the province of Quebec extended
the program for self-employed mothers and fathers. The most recent
change--one year of combined parental and maternity leave--was
implemented in 2000, standardizing leave provision across the provinces.
(2) In the province of Quebec, a $5 per day childcare policy began in
1997, with the cost increasing to $7 a day in 2000, had an important
impact on women's return to work (Cleveland 2008; Lefebvre and
Merrigan 2008). However, as the present paper is mainly concerned with
the evolution of maternal transitions to work from 1970 to 1999, the
effects of these more recent policy measures cannot be addressed in our
analysis.
Spousal or Family Income
Another factor affecting women's return to work is their
access to the various economic resources that allow them to balance paid
and unpaid work after childbearing (Cleveland et al. 1996). The presence
and salary of a spouse is clearly important in this context, and
previous research on this transition generally finds that spousal income
is especially important for women's return to work, particularly in
the event of an unplanned pregnancy. For those women who have children
in the absence of a spouse or partner, balancing childcare and work is
even more complicated, and clearly has potential implications for a
woman's labor market participation. However, among those with
partners, low spousal income may encourage fast reentry, and high
spousal income may give more choices for women to express their
preferences to either stay home or return to work (Hakim 1998).
Most of the previous research regarding the effects of structural
context and individual characteristics on women's employment
transitions has been done outside the Canadian context. We have little
information on the characteristics that affect Canadian women's
employment or how this has changed over the past few decades. It is
these disparities that we mainly address in the present paper.
RESEARCH QUESTIONS AND HYPOTHESES
Our main research questions were: (1) what are the characteristics
that affect Canadian women's employment and (2) how have
women's employment transitions after the birth of a first child
have changed over time? In particular, we asked whether women have
become more likely to be engaged in paid work fairly shortly after
childbearing, whether the probability of this transition can be
predicted by women's education or spousal income, and whether the
effect of education seems to have changed for more recent childbirth
cohorts. Three hypotheses guided our analysis. First, following
Marshall's (1999) findings, we expected that the probability of
women working two years after childbirth had increased over the decades
of the 1970s to the 1990s. In particular, we expected that the economic
crises of the early 1980s and 1990s would have had seen increases in
women's paid work after childbirth. Second, following the review
above, we also hypothesized that mothers' educational attainment
and availability of family resources, principally the income of a
spouse, would be positively related to their likelihood of working while
their children are still young (preschool-age). Third, we expected that
education had become more important in women's labor force
attachment over the period, and therefore the strength of education
attainment in predicting women's work after childbearing would have
increased.
METHODS AND DATA
In order to address these hypotheses, we used the 2001 GSS, Cycle
15 on Family History (GSS 15). (3) This nationally representative survey
collected retrospective data on a variety of family transitions,
including childbirth, union formation and dissolution, work, and
education, as well as information on children, siblings, and spouses,
from 24,310 Canadians (Statistics Canada 2003). Complete work and
childbirth histories in the GSS 15 make it possible to calculate the
length of time in months between the birth of that child and the
beginning of the next spell of employment. We used binary logistic
regression models to investigate how women who began childbearing in the
1970s, 1980s, and 1990s differed in the probability of remaining out of
the paid labor force for at least 24 months after their first birth, and
how this was affected by family and personal characteristics. Consistent
with previous research (Hynes and Clarkberg 2005; Marshall 1999), the
24-month time period was analyzed. Twenty-four months is especially
relevant in Canada because it extends beyond the maximum duration of
paid maternity leave.
As discussed above, the probability of a woman beginning work after
the birth of a child is affected by a number of factors. Unfortunately,
the information on income, household, and family characteristics in the
GSS 15 was collected only for the time of the survey, whereas the timing
of births and job absences was collected retrospectively. Therefore, as
the time since the childbearing and work decisions increases, it becomes
increasingly unlikely that the characteristics in 2001 represent the
characteristics at the time of those decisions.
In order to ameliorate this potential problem with the data, we
estimated two sets of models. In the first, we investigated the effects
of birth cohort on the probability of a woman beginning work within 24
months of the birth of her first child (Table 1). Six five-year
childbirth cohorts were included as main independent variables, along
with women's education and employment status in the year before the
birth of their first children. We also included a number of variables as
controls. These included the mother's age at first birth, with the
expectation that women who had children later in life might be more
likely to be strongly engaged in a career than younger women and also
return to work faster. It was expected that immigrants to Canada may
behave differently with regard to their work after childbearing, so an
indicator of place of birth was included as a control. One important
reason that women may not return to work is the birth of subsequent
children, so we included a control variable indicating whether a second
birth took place within 24 months following the first birth. The GSS 15
does not collect information on characteristics of employment, and we
were therefore not able to include aspects of occupation or industry for
jobs held before or after the first birth.
The second set of models was a separate analysis of the women who
had their first children between 1995 and 1999. By limiting the time
frame, we were able to include some additional family and demographic
characteristics that could not be included in the cohort models because
they were only measured in 2001. It is acknowledged that we cannot be
certain that their values in 2001 reflect women's characteristics
at the time of childbearing, even in this restricted time period.
The GSS 15 data set includes individual income and family income,
both as categorical variables, but does not include a measure of spousal
income. We estimated an ordinal proxy for spousal income, for those
women who were living in families with a spouse or partner, by
subtracting the midpoints of the individual income categories from those
of the family income categories. Although a direct measure of spousal
income would very much be preferred, we can be fairly confident that
those in the lower categories of this variable do have access to fewer
family financial resources than those in the higher categories.
These models also include a control for lone parenthood as well as
indicators for province of residence, which may partially capture
regional differences in maternity leaves as well as some labor market
differences. Again, these are included with the caveat that some of
these women would not have lived in the same province in 2001 as when
they had their first children. The other controls includes in the first
models; age at first birth, being born outside of Canada, having worked
in the 12 months before the first birth, and having a second child
within 24 months, were included in these models as well.
After excluding a small number of women for whom the date of first
birth was unavailable, and approximately 150 for whom the timing of
their prechildbirth employment was unclear, there were 4,889 individuals
left for the cohort analysis. Of the women who had their first births
between 1995 and 1999, there were 592 included in the second model
(Table 1).
RESULTS
Using the population weights provided on the data set, our sample
resulted in an estimate that 41.2 percent of the women who had first
children during this period, and who were alive to be in the Canadian
population in 2001, were working within 24 months of their birth of
their first child. As Table 1 shows, the percentage who reported working
within this time rose across childbirth cohorts, from less than a third
of those in the 1970 to 1974 cohort, to more than 61 percent in the 1995
to 1999 cohort.
As described above, logistic regression models were used to
investigate these cohort changes in the probability of women's
starting work within 24 months of a first child, and these are presented
in Table 2. The effect of cohort remained after introduction of
education and the control variables, although the 1975 to 1979 and 1980
to 1984 cohorts were not significantly different from 1970 to 1974. As
Figure 1 shows, the likelihood of a woman working within 24 months of
childbirth rose so that women who had their first children between 1995
and 1999 were 2.8 times as likely to work within 24 months, as those in
the 1970 to 1974 cohort. The increase was particularly strong after the
1985 to 1989 childbirth cohort, and most of the total increase occurred
over the last 15 years of the period.
[FIGURE 1 OMITTED]
A woman's education had a clear effect on the likelihood that
she was working within two years of her first child. Controlling for the
other variables in the model, including cohort, women with less than a
high school diploma had the lowest odds of beginning work within this
time (OR = .499). The odds of having worked within two years increased
with each education level, although there was no significant difference
between women with university degrees and those with completed trade,
technical, or other non-university qualifications (Figure 2).
In Model 2, interaction terms for women's education and
childbirth cohort were included to assess whether the effects of
women's education on their work transitions had changed with
childbirth cohort (Table 2). This changed the main effects of cohort
somewhat, insofar as the effect of 1975 to 1979 birth cohort became
significant and much stronger. The general pattern of the other cohort
variables remained the same. Overall, the effect of including the
interactions was not great, in terms of the significance of parameters
or fit of the model. Only one interaction term was significant at the
.05 level, although several were just below this level of significance.
Recent Childbirth Cohorts (1995-1999)
Table 3 presents the model results for women who had their first
births between 1995 and 1999. In this cohort, those women with less than
high school education were significantly less likely than
university-educated women to have returned to work, controlling for the
other variables in the model (OR = .295). The estimated income of a
woman's spouse or partner indeed had a significant, although
nonlinear, effect on her likelihood of starting paid work, independent
of a woman's own education and the control variables. As shown in
Figure 3, women with estimated spousal incomes in the middle categories
(between $30,000 and $79,999) in 2001 were significantly less likely to
have been engaged in paid work within two years following the birth of
the first child, roughly a third as likely as women with estimated
spousal income in the highest category (80,000 or more). Women with
estimated spousal incomes in the lowest categories were the most likely
to have worked within 24 months, not significantly different from the
highest category.
[FIGURE 2 OMITTED]
DISCUSSION
It is generally accepted that women's labor force
participation has been increasing since the 1970s, and that we are
approaching a "full adult worker model" with roughly equal
gender representation in paid work. Indeed, with each successive
childbirth cohort since the 1970s, women have been more likely to begin
work within two years of the birth of their first child. However, if we
look to the transitions, we see that nearly 60 percent of women who had
their children over this period were out of the labor force for more
than two years. This may have important economic consequences for women
and their children since, in the long term, it affects women's
retirement savings, their development of human capital, and their career
advancement.
To understand these changes, we can consider aspects of the social
and economic context. One explanation is that there are clearly broad
cultural changes regarding gendered patterns of paid work and the
balance of "earning and caring" (Beaujot 2000). The increasing
level of education attainment by younger cohorts of women is also a key
element explaining changes in women's employment and family
trajectories. These, however, have also taken place alongside important
changes in the labor market, as we discuss below.
[FIGURE 3 OMITTED]
Variation of Lifecourse Transitions across Cohorts
We had expected that the probability of women beginning work within
two years of childbirth had increased over this 30-year period, and also
that the increase would be linked to two important economic crises in
the early 1980s and 1990s, which would have made women's employment
more important for family income. We could therefore have expected a
slight decrease in this probability between these two economic crises.
We did see a substantial overall increase, but this was really only
apparent starting with the 1985 to 1989 childbirth cohort, and increased
strongly thereafter.
Although we expected to see the increase begin with the 1980 to
1984 cohort, we can offer some explanation for why the change is not
apparent until 1985 to 1989. It could be that the effects of the
economic crisis were not directly felt by Canadian households before
this period. Interest rates peaked in 1981, (4) but may have only
affected young families once mortgages needed to be renewed. Similarly,
inflation reached its highest point between 1985 and 1987 (McCallum
1998), years after the economic crisis began. Moreover, men's wages
stagnated between 1980 and 1990 (Phillips and Phillips 2000), possibly
effecting the household economy and leading to a need for the
women's second salaries (Baker 2005). The possibility of a delayed
effect may also explain why women's probability of working after
childbirth did not fall in the period between the two economic crises.
If mothers' increasing work participation was a reaction to
difficult economic circumstances, we might expect it to have fallen
again after the economy had improved by the end of the 1990s.
There have been other important changes that almost certainly play
a role in these dynamics, including changes to gender roles that are
somewhat independent of economic conditions. The increase in
women's educational attainment likely increased labor force
attachment. As Crompton (2002, 2006) proposes, women's choices are
the result of a complex nexus of factors, including education and
available work, cultural values, gender roles, as well as the economic
context, that are not easily taken into account in rational choice
theories.
Education and Income as Predictors of Transitions
One potentially important finding is that, while more Canadian
women are returning to work two years after the birth of a first child,
this phenomenon may be shaped by educational attainment. Although we did
not find evidence that this effect has changed across childbirth
cohorts, women with lower levels of education were reliably more likely
to have remained out of the labor force than those with higher
education. In the 1995 to 1999 cohort, it was only women with the lowest
level of education, less than high school, that were significantly less
likely to have returned to work than were those with a university
degree. As well, the knowledge economy and the changes in the 1990s
labor market have meant that employment prospects for those with less
than a high school diploma are few.
The reality lived by women with a low education is worrisome from
the lifecourse perspective. First, the arrival of a child in the lives
of these women can put them at risk of social isolation, as they
withdraw from their daily work environments, sources of interaction, and
social capital (Furstenberg 2005). The fact that many of these women do
not return to work for at least two years has potential economic
implications, such as the loss of access to extended benefits including
additional medical insurance. One of the tenets of the lifecourse
perspective is that these negative effects can accumulate over the
course of women's lives, and may reduce the resources available to
cope with other negative events. Employment before the first birth and a
quick return to work together represent an important protection factor
for women against poverty (Juby et al. 2005).
Our analysis of the 1995 to 1999 cohort, for which we could include
a proxy measure of spousal income, sheds light on another significant
element for Canadian women's labor market entry post-childbirth.
Women in the highest and lowest categories of estimated spousal income
were more likely than those in moderate categories to have returned,
controlling for education and lone parent status. This leads us to
suspect that women with access to high spousal incomes may be able to
afford childcare, allowing them to return to work more quickly. We might
also speculate that this could be related to the norms of a new working
class characterized by high household earnings, high educational
attainment, and strong work attachment by both men and women (Gershuny
2005). As shown by Blair-Loy and Jacobs (2003), women who are engineers,
lawyers, brokers, or other professionals and are part of this new
working class, may have significant issues with work-life balance, and
this may be compounded by their spouses' high attachment to work.
On the other hand, for women with more moderate family resources, the
decision to return to work may involve weighing the various costs of
childcare against the benefits of paid work. Considering the complexity
of this lifecourse decision, it might also be based on a different set
of work-related values.
Although the data presented here do not allow us to do more than
speculate about these reasons for these patterns, the relationships
found between spousal income and women's work presents concerns.
One is that for these women, as with those with low education, the
decision to remain out of the labor force may have implications for
their later lifecourses, in terms of employability and future income.
CONCLUSIONS
The results presented in this article allow us to better understand
changes in the transition to work after childbirth, across cohorts of
women from the 1970s to the 1990s. Other analyses of transitions have
already revealed changes in the timing of the work transition after a
first child since the 1990s (Zhang 2007). Our intercohort analysis
presents the evolution of this key transition in the lifecourse of
Canadian women across three decades. The 30-year period analyzed shows
one way in which personal biographies may have been influenced by period
effects, such as economic crises. Moreover, the transitions analysis
reveals, among other things, that the full adult-worker family model
remains a goal, rather than a reality. It has taken 30 years of
childbirth cohorts for women's labor force participation after
childbearing to reach the present level. It is important to lift the
veil on this situation because general opinion and political rhetoric
often take this model for granted. Many Canadian women did not return to
work two years after a first birth, even in the most recent cohorts. A
portion of this is explained by differences in women's education
and spousal income, suggesting that there are important socioeconomic
differences between women who return to the labor market fairly quickly
after childbearing, and those who do not.
The data we have used give us reason to be cautious about these
findings. The retrospective GSS data suffer from problems with
respondent recall and respondent mortality that are less prevalent with
prospective longitudinal panel data. Marshall (1999), using longitudinal
data from the Survey of Labour and Income dynamics, found that nearly 90
percent of women who had a birth between 1993 and 1996 returned to work
in 24 months. However, that analysis included only those who worked
before the birth. Our estimates do correspond fairly well to some
estimates of the labor force participation of women with young children
during the same period, from the Labour Force Survey (Statistics Canada
2005b:105). Although it would be preferable to have a source of
long-term panel data, there are no sources of Canadian data that extend
back to the 1970s, and which could be used to examine our main research
questions.
One important aspect of policy that we were not able to directly
address through this analysis is the potential effect of affordable,
high-quality childcare services. Taking into consideration analyses
conducted in Quebec since the establishment of daycare services at $5
per day in 1997, we can confidently predict that women would benefit
from the expansion of such services. Lefebvre and Merrigan (2008) show
that, before the implementation of the program in 1997, labor force
participation of Quebec women with preschool children was lower than
that of their counterparts in other provinces. In 2002, some five years
after implementation, Quebec women are the most active in the labor
market in Canada, their rate of participation having risen by 13
percent. Provisions of these services may be important for reducing the
chance that women with low incomes or low education are further excluded
from the labor market after childbirth.
Finally, this analysis produces two conclusions which merit
attention in the public policy milieux. The first is that a large
proportion of women still appear to withdraw from the labor force for an
extended period following childbearing, suggesting that the full adult
worker model is not as fully realized as is commonly assumed. Second, we
find it is women with lower levels of education, and those with moderate
access to spousal income, who are most likely to remain out of the labor
force. The structuring of this transition by these social and economic
characteristics has implications for the distribution of lifecourse
risks.
References
Armstrong, P. and H. Armstrong. 1994. The Double Ghetto. Toronto:
McClelland and Stewart Inc. Baker, M. 2005. Families. Changing Trends in
Canada. Toronto: McGraw-Hill-Ryerson.
Barrow, L. 1999. "An Analysis of Women's Return
to-Work-Decisions following First Birth." Economic Inquiry
37:432-51.
Beaujot, R. 2000. Earning and Caring. Peterborough: Broadview
Press.
Becker, G.S. 1991. A Treatise on the Family. Cambridge, MA: Harvard
University Press.
Bernard, P. 2003. "Les nouvelles statistiques sociales."
Sociologie et societes 35(1):3-18.
Blair-Loy, M. and J.A. Jacobs. 2003. "Globalization, Work
Hours, and the Care Deficit among Stockbrokers." Gender and Society
17:230-49.
Burgess, S., P. Gregg, C. Propper and E. Washbrook. 2008.
"Maternity Rights and Mothers' Return to Work." Labour
Economics 15(2): 168-201.
Canadian Economic Council. 1990. Good Jobs, Bad Jobs: Employment in
the Service Economy. Hull: Government of Canada Publication.
Charles, M., M. Buchmann, S. Halebsky, J. Powers and M. Smith.
2001. "The Context of Women's Market Careers." Work and
Occupations 28:371-96.
Cleveland, G. 2008. '"Benefices et cofits des centres de
la petite enfance du Quebec' in Les services de garde educatifs h
la petite enfance du Quebec." Recherches, reflexions et pratiques,
PUQ. Pp. 27-39.
Cleveland, G., M. Gunderson and D. Hyatt. 1996. "Child Care
Costs and the Employment Decision of Women: Canadian Evidence." The
Canadian Journal of Economics/Revue canadienne d'Economique
29(1):132-51.
Cote, J. 2006. "Emerging Adulthood as an Institutionalized
Moratorium: Risks and Benefits to Identify Formation." In Emerging
Adults in America: Coming of Age in the 21st Century, edited by J.
Arnett and J. Tanner. American Psychological Association.
Crompton, R. 2002. "Employment, Flexible Working and the
Family." The British Journal of Sociology 53(4):537-58.
Crompton, R. 2006. Employment and the Family. The Reconfiguration
of Work and Family Life in Contemporary Societies. Cambridge, UK:
Cambridge University Press, 252 pp.
Crompton, R. and F. Harris. 1998. "A Reply to Hakim." The
British Journal of Sociology 49(1):144-49.
Dewilde, C. 2003. "A Life-Course Perspective on Social
Exclusion and Poverty." British Journal of Sociology 54(1):109-28.
Doucet, A. 2010. "Paid and Unpaid Work: Linking Households,
Workplaces, the State, and Communities."
In New Canadian Families, 2d ed, edited by D. Cheal. Oxford: Oxford
University Press.
Duxbury, L. and C. Higgins. 2002. "Enquete nationale sur le
conflit entre le travail et la vie personnelle." Retrieved October
4, 2004 (http://www.phac-aspc.gc.ca/publicat/work-travail/
report1/index-fra.php).
Duxbury, L. and C. Higgins. 2003. "Le conflit entre le travail
et la vie personnelle au Canada durant le nouveau millenaire."
Retrieved October 4, 2004 (http://www.phac-aspc.gc.ca/
publicat/work-travail/report2/index-fra.php).
Elder, G.H. Jr. 1985. Life Course Dynamics. Ithaca, NY: Cornell
University Press.
Elliott, J., D. Angela and M. Egerton. 2001. "Returning to
Work after Childbirth: A Longitudinal Analysis of the Role of
Qualifications in Mothers' Return to Paid Employment." Pp.
505-24 in Restructuring Work and the Life Course, edited by V.W.
Marshall, W.R. Heinz, H. Kruger and A. Verma. Toronto: University of
Toronto Press.
Fast, J. and J. Frederick. 2004. The Time of Our Lives: Juggling
Work and Leisure Over the Life Cycle (Catalogue No. 89-584). Ottawa:
Statistique Canada, 39 pp. Retrieved October 4, 2004
(http://www.hecol.ualberta.ca/RAPP/documents/
The%20Time%2020Our%20Lives%20203.pdf.
Feeney, J., R. Alexander, P. Noller and L. Hohaus. 2003.
"Attachment Insecurity, Depression, and the Transition to
Parenthood." Personal Relationships 10(4):475-93.
Felmlee, D.H. 1995. "Causes and Consequences of Women's
Employment Discontinuity, 1967-1973." Work and Occupations
22:167-88.
Fortin, N.M. and M. Huberman. 2002. "Gender Occupational
Women's Wages Segregation in Canada: Historical Perspective."
Canadian Public Policy 28:s11-39.
Furstenberg, F.F. 2005. "Banking on Families: How Families
Generate and Distribute Social Capital." Journal of Marriage and
Family 67:809-21. doi: 10.1111/j.1741-3737.2005. 00177.x.
George, Linda. 1993. "Sociological Perspectives on Life
Transitions." Annual Review of Sociology 19:353-73.
Gershuny, J. 2005. "Business as the Badge of Honor for the New
Superordinate Working Class." Social Research 72(2):287-314.
Glass, J. and N.E. Leda. 2006. "Religious Conservatism and
Women's Market Behavior following Marriage and Childbirth."
Journal of Marriage and Family 68(3):611-29.
Gutierrez-Domenech, M. 2002. "Penalty after Motherhood in
Spain. Centre for Economic Performance, LSE." Retrieved March 17,
2007 (http://cep.lse.ac.uk/pubs/download/ DP0546.pdf).
Gutierrez-Domenech, M. 2005. "Employment after Motherhood: A
European Comparison." Labour Economics 12:99-123.
Haas, B. 2005. "The Work-Care Balance: Is it Possible to
Identify Typologies for Cross-National Comparisons?" Current
Sociology 53(3):487-508.
Hakim, C. 1995. "Five Feminist Myths about Women's
Employment." The British Journal of Sociology 46:429-55.
Hakim, C. 1996. "Labour Mobility and Employment Stability:
Rhetoric and Reality on the Sex Differential in Labour-Market
Behaviour." European Sociological Review 12:1-31.
Hakim, C. 1998. "Developing a Sociology for the Twenty-First
Century: Preference Theory." The British Journal of Sociology
49:137-43.
Han, W.-J., C.J. Ruhm, J. Waidfogel and E. Washbrook. 2008.
"The Timing of Mothers' Employment after Childbirth."
Monthly Labour Review (June): 15-27.
Helms-Erikson, H. 2001. "The Transition to Parenthood:
Implications of the Timing of Parenthood and the Division of
Housework." Journal of Marriage and Family 63(4): 1099-110.
Heymann, J. 2006. Forgotten Families. Oxford: Oxford University
Press, 306 pp.
Higgins, C., L. Duxbury and S. Lyons. 2005. "Reducing
Work-Life Conflict: What Works? What Doesn't?" Health Canada.
Retrieved February 19, 2008 (http://www.hc-sc.gc.ca/ewh-semt/
pubs/occup-travail/balancing-equilibre/index_e.html).
Hochschild, A.R. 2003. The Second Shift. New York: Penguin Books.
Hofferth, S.L. and S.C. Curtin. 2006. "Parental Leave Statutes
and Maternal Return to Work after Childbirth in the United States."
Work and Occupations 33(1):73-105.
Hogan, D.P. and N.M. Astone. 1986. "The Transition to
Adulthood." Annual Review of Sociology 12:109-30.
Houston, D.M. and G. Marks. 2003. "The Role of Planning and
Workplace Support in Returning to Work after Maternity Leave."
British Journal of Industrial Relations 41(2):197-214.
Hynes, K. and M. Clarkberg. 2005. "Women's Employment
Patterns during Early Parenthood: A Group-Based Trajectory
Analysis." Journal of Marriage and Family 67:222-39.
Jaumotte, F. 2003. "Labour Force Participation of Women:
Empirical Evidence on the Role of Policy and Other Determinants in OECD
Countries." OECD Economic Studies, No. 37. Retrieved March 8, 2008
(http://www.oecd.org/dataoecd/12/ 39/34562935.pdf).
Jenson, J. and M. Sineau. 2001. "Citizenship in the Era of
Welfare State Redesign." Pp. 240-59 in Who Cares ? Women "s
Work, Childcare, and Welfare State Redesign, edited by J. Jenson and M.
Sineau. Toronto: University of Toronto Press.
Juby, H., C. Le Bourdais, N. Marcil-Grattan and L.P. Rivest. 2005.
"Pauvrete des familles monoparentales et parcours professionnels
des meres apres la rupture." Recherches sociographiques
XLVI:217-43.
Kempeneers, M. 1987. "Questions sur les femmes et le travail:
une lecture de crise." Sociologie et societes 19(1):57-72.
Kempeneers, M. 1989. "Discontinuite professionnelle et charges
familiales: regards sur les donnees canadiennes." Cahiers quebecois
de demographie 18(1):63-85.
Kenjoh, E. 2005. "New Mothers' Employment and Public
Policy in the UK, Germany, the Netherlands, Sweden, and Japan."
Labour 19(1):5-49.
Kershaw, P.W. 2005. Carefair: Rethinking the Responsibilities and
Rights of Citizenship. Vancouver: UBC Press.
Killien, M.G. 2005. "The Role of Social Support in
Facilitating Postpartum Women's Return to Employment." Journal
of Obstetric, Gynecologic, & Neonatal Nursing 34:639-46.
Klerman, J.A. and A. Leibowitz. 1999. "Job Continuity among
New Mothers." Demography 36(2):145-55.
Kruger, H. and R. Levy. 2001. "Linking Life Courses, Work, and
the Family: Theorizing a not so Visible Nexus between Women and
Men." Canadian Journal of Sociology 26:145-66.
Lefebvre, P. and P. Merrigan. 2008. "Child-Care Policy and the
Labor Supply of Mothers with Young Children: A Natural Experiment from
Canada." Journal of Labor Economics 26(3):519-48.
Lewis, J. and S. Giullari. 2005. "The Adult Worker Model
Family, Gender Equality and Care: The Search for New Policy Principals
and the Possibilities and Problems of Capabilities Approach."
Economy and Society 34(1):76-104.
MacMillan, R. 2005. "The Structure of the Life Course: Classic
Issues and Current Controversies." Advances in Life Course Research
9:3-26.
MacMillan, R. and R. Copher. 2005. "Families in the Life
Course: Interdependency of Roles, Role Configurations, and
Pathways." Journal of Marriage and Family 67:858-59.
Marks, G. and D. Houston. 2002. "Attitudes towards Work and
Motherhood Held by Working and Non-Working Mothers." Work,
Employment and Society 16:523-36.
Marshall, K. 1999. "Employment after Childbirth."
Perspectives on Labour and Income 11(3): 18-25.
McCallum, J. 1998. "Les taux de chomage canadien et americain
dans les annees 1980: un test de trois hypotheses." Revue
L'actualite economique 64(4):494-508.
Osberg, L. 2007. "A Quarter Century of Economic Inequality in
Canada." Research Paper, Department of Economics, Dalhousie
University.
Phillips, P. and E. Phillips. 2000. Women and Work. Inequality in
the Canadian Labour Market. Toronto: James Lorimer and Company Ltd.
Phipps, S., P. Burton and L. Lethbridge. 2001. "In and Out of
the Labour Market: Long-Term Income Consequences of Child-Related
Interruptions to Women's Paid Work." Canadian Journal of
Economics/Revue Canadienne d'Economique 34:411-29.
Ranson, G. 2005. "Paid and Unpaid Work: How Do Families Divide
Their Labour?" Pp. 99-120 in Families. Changing Trends in Canada,
edited by M. Baker. Toronto: McGraw-Hill-Ryerson.
Ravanera, Z.R., F. Rajulton and T.K. Burch. 1998. "Early Life
Transitions of Canadian Women: A Cohort Analysis of Timing, Sequences,
and Variations." European Journal of Population 14:179-204.
Reiter, H. 2009. "Beyond the Equation Model of Society--the
Postponement of Motherhood in Post-State Socialism in an
Interdisciplinary Life Course Perspective." The European Journal of
Social Science Research 22(2):233-46.
Sanders, K. 1997. "Mothers and Daughters in the Netherlands:
The Influence of the Mother's Social Background on Daugthers'
Labour Market Participation after They Have Children." European
Journal of Women "s Studies 4:165-81.
Smeaton, D. 2006. "Work Return Rates after Childbirth in the
UK--Trends, Determinants and Implications: A Comparison of Cohorts Born
in 1958 and 1970." Work, Employment and Society 20(1):5-25.
Stanfors, M.A. 2006. "Labor Force Transitions after Childbirth
among Five Birth Cohorts in Sweden." Journal of Family and Economic
Issues 27(2):287-309.
Statistics Canada. 2003. "General Social Survey Cycle 15:
Family History Public Use Microdata File." Retrieved February 19,
2008 (http://www.statcan.ca/ bsolc/english/bsolc?catno= 12M0015XCB).
Statistics Canada. 2005a. "Naissances. Le Quotidien (July
12)." Retrieved October 4, 2007 (http://www.statcan.ca).
Statistics Canada. 2005b. Women in Canada: A Gender-Based
Statistical Report. Cat. No. 89-503-XIE, 5th ed. Ottawa: Statistics
Canada.
Statistics Canada. 2008. "Table 2 Usual Hours by Education and
Sex (Age 25 and Over). Labour Force Survey." Retrieved May 18, 2010
(http://www.statcan.gc.ca/ pub/75-001-x/2008103/
tables/10534/5004211-eng.htm).
Stoltzfus, E. 2003. Citizen, Mother, Worker: Debating Public
Responsibility for Child Care after the Second World War. Chapel Hill,
NC: The University of North Carolina Press.
Tomlinson, J. 2006. "Women's Work-Life Balance
Trajectories in the UK: Reformulating Choice and Constraint in
Transitions through Part-Time Work across the Life-Course." British
Journal of Guidance and Counselling 34(3):365-82.
Wharton, A.S. and M. Blair-Loy. 2006. "Long Work Hours and a
Cross-National Study of Employees' Concerns." Journal of
Family Issues 27(3):415-36.
Zhang, X. 2007. "Returning to the Job after Childbirth."
Perspectives on Labour and Income 8(12):18-24.
STEPHANIE GAUDET
Universite d'Ottawa
MARTIN COOKE AND JOANNA JACOB
University of Waterloo
This research was funded by a standard SSHRC grant (410-2005-1323).
The title of the project is: << La participation sociale des
Canadiens a travers l'analyse des parcours de vie >>.
Stephanie Gaudet is the principal investigator and the co-investigators
are Paul Bernard, Susan McDaniel and Martin Cooke. Stephanie Gaudet,
Sociologie et anthropologie, Universite d'Ottawa, 55 Rue Laurie E.,
Ottawa, ON, Canada K1N 6N5. E-mail: sgaudet@uottawa.ca
(1.) There was a 40 percentage point increase among mothers'
employment rate during this period (Statistics Canada 2005a).
(2.) In January 2006, Quebec took control over family leave
policies in that province, expanding the program to self-employed
parents and introducing flexibility in the length of the leave.
(3.) This analysis used confidential data made available through
the Research Data Centres Programme. The research and analysis are based
on data from Statistics Canada (2003). The opinions expressed do not
represent the views of Statistics Canada.
(4) The Canadian Treasury Bill rate peaked at 20.85 percent in
August 1981 (Osberg 2007).
Table 1
Sample Description
All childbirth cohorts
(1970-1999)
Sample % Began work
(%) within 24 months
(N=4,889) (N=2,047)
Independent variables
Childbirth cohort
1970-1974 13.5 28.5
1975-1979 16.3 33.2
1980-1984 20.0 35.2
1985-1989 18.1 42.6
1990-1994 17.5 51.2
1995-1999 14.6 61.3
Total 100 41.2
Education
Less than high school 13.7 25.0
High school diploma 21.7 37.8
Some university, 12.4 42.4
technical, or CEGEP (a)
Completed trade or 32.0 46.2
technical
University degree 20.3 50.6
Total 100 --
Estimated spousal income
Less than $20,000 -- --
$20,000-29,999 -- --
$30,000-49,999 -- --
$50,000-79,999 -- --
$80,000 or more -- --
Total
Control variables
Born outside of Canada 20.5 39.6
Born in Canada (ref) 79.5 42.5
Second child within 24 46.2 40.2
months
No second child (ref) 53.8 43.2
Worked 12 months 19.5 57.9
before first birth (ref)
Did not work for pay 80.5 38.0
12 months before
Lone parent, no spouse/ -- --
partner
With spouse/partner
(ref)
Region
Atlantic -- --
Quebec -- --
Ontario -- --
Prairies -- --
British Columbia -- --
Total -- --
Age at first birth 26.081 (4.209) (c)
Childbirth cohort
(1995-1999)
Sample % Began work
(%) within 24 months
(N=592) (N=2,047)
Independent variables
Childbirth cohort
1970-1974 -- --
1975-1979 -- --
1980-1984 -- --
1985-1989 -- --
1990-1994 -- --
1995-1999 -- --
Total 100.0 61.3
Education
Less than high school 5.9 35.4
High school diploma 14.7 60.2
Some university, 12.9 70.3
technical, or CEGEP (a)
Completed trade or 35.8 64.2
technical
University degree 30.6 67.5
Total 99.9 (b) --
Estimated spousal income
Less than $20,000 8.4 47.2
$20,000-29,999 7.5 49.4
$30,000-49,999 24.6 55.4
$50,000-79,999 30.7 61.1
$80,000 or more 28.9 71.3
Total 101.9
Control variables
Born outside of Canada 21.8 57.8
Born in Canada (ref) 78.2 65.4
Second child within 24 45.8 62.1
months
No second child (ref) 54.2 64.9
Worked 12 months 73.2 77.3
before first birth (ref)
Did not work for pay 26.8 58.6
12 months before
Lone parent, no spouse/ 16.2 53.0
partner
With spouse/partner 83.8 54.6
(ref)
Region
Atlantic 6.6 62.2
Quebec 25.2 66.9
Ontario 38.2 64.1
Prairies 16.8 60.2
British Columbia 13.2 62.0
Total 100.0 --
Age at first birth 27.920 (4.063)
Source: Statistics Canada General Social Survey Cycle 15 Master
File, Weighted data (2003).
(a) CEGEP is a postsecondary program in the province of Quebec.
(b) Totals may not sum to 100 due to rounding.
(c) Mean (standard deviation).
Table 2
Logistic Regression Models of Beginning Paid Work within 24 Months
of First Birth, 1970-1999 Childbirth Cohorts
Model 1
Exp(B) B SE(B)
Intercept .159 -1.8383 .2108 ***
Childbirth cohort
1970-1974 (ref) 1.000
1975-1979 1.194 .177 .118
1980-1984 1.199 .182 .112
1985-1989 1.584 .460 .114 ***
1990-1994 2.127 .755 .114 ***
1995-1999 2.864 1.052 .121 ***
Education
Less than high school .488 -.717 .117 ***
High school diploma .732 -.311 .096 **
Some university, technical, or CEGEP .792 -.233 .110 *
Completed trade or technical .944 -.057 .087
University degree (ref) 1.000
Control variables
Born outside of Canada .860 -.150 .077
Second child within 24 months .861 -.150 .062 **
Worked for pay 12 months before 2.017 .702 .077 ***
first birth
Age at first birth 1.050 .049 .007 ***
Education x cohort interactions
1975-1979 x less than high school -- -- --
1980-1984 x less than high school -- -- --
1985-1989 x less than high school -- -- --
1990-1994 x less than high school -- -- --
1995-1999 x less than high school -- -- --
1975-1979 x high school diploma -- -- --
1980-1984 x high school diploma -- -- --
1985-1989 x high school diploma -- -- --
1990-1994 x high school diploma -- -- --
1995-1999 x high school diploma -- -- --
1975-1979 x some postsecondary -- -- --
1980-1984 x some postsecondary -- -- --
1985-1989 x some postsecondary -- -- --
1990-1994 x some postsecondary -- -- --
1995-1999 x some postsecondary -- -- --
1975-1979 x completed trade or tech -- -- --
1980-1984 x completed trade or tech -- -- --
1985-1989 x completed trade or tech -- -- --
1990-1994 x completed trade or tech -- -- --
1995-1999 x completed trade or tech -- -- --
N = 4,889
df = 13
AIC = 6,186.463
-2LL = 6,158.463
Model 2
Exp(B) B SE(B)
Intercept .135 -1.9991 .263 **
Childbirth cohort
1970-1974 (ref) 1
1975-1979 1.974 .680 .263 **
1980-1984 1.483 .394 .245
1985-1989 1.833 .606 .250 **
1990-1994 2.194 .786 .245 **
1995-1999 3.287 1.190 .247 ***
Education
Less than high school .689 -.372 .280
High school diploma .803 -.219 .273
Some university, technical, or CEGEP 1.440 .365 .313
Completed trade or technical .982 -.018 .259
University degree (ref) 1.000
Control variables
Born outside of Canada .87 -.134 .078
Second child within 24 months .86 -.155 .062 *
Worked for pay 12 months before 2.03 .710 .077 ***
first birth
Age at first birth 1.05 .048 .007 ***
Education x cohort interactions
1975-1979 x less than high school .622 -.476 .386
1980-1984 x less than high school .665 -.407 .378
1985-1989 x less than high school .560 -.579 .399
1990-1994 x less than high school .775 -.255 .400
1995-1999 x less than high school .551 -.596 .435
1975-1979 x high school diploma .541 -.614 .367
1980-1984 x high school diploma .827 -.190 .341
1985-1989 x high school diploma .992 -.008 .347
1990-1994 x high school diploma 1.157 .146 .348
1995-1999 x high school diploma 1.115 .109 .371
1975-1979 x some postsecondary .414 -.881 .420 *
1980-1984 x some postsecondary .534 -.628 .410
1985-1989 x some postsecondary .472 -.751 .398
1990-1994 x some postsecondary .462 -.773 .395
1995-1999 x some postsecondary .721 -.327 .408
1975-1979 x completed trade or tech .570 -.563 .344
1980-1984 x completed trade or tech .929 -.074 .323
1985-1989 x completed trade or tech .914 -.090 .326
1990-1994 x completed trade or tech .778 -.251 .321
1995-1999 x completed trade or tech .903 -.102 .327
N = 4,889
df = 33
AIC = 6,200.979
-2LL = 6,132.979
*** p<.0001.
** P<.01.
* p <.05.
Table 3
Logistic Regression Models of Beginning Paid Work within
24 Months of First Birth, 1995-1999 Childbirth Cohort
Exp (B) B SE(B)
Intercept 3.787 1.332 .677 *
Education
Less than high school .295 -1.220 .428 **
High school diploma 1.317 .275 .330
Some university, technical, or CEGEP .881 -.127 .317
Completed trade or technical .849 -.164 .246
University degree (ref) 1.000 -- --
Estimated spousal income
Less than $20,000 1.371 .316 .305
$20,000-29,999 .775 -.255 .400
$30,000-39,999 .307 -1.182 .388 **
$40,000-49,999 .368 -1.000 .359 **
$50,000-59,999 .558 -.583 .458
$60,000-79,999 .340 -1.079 .363 **
$80,000 or more (ref) 1.000 -- --
Lone parent/no spouse .401 -.914 .327 **
Control Variables
Born outside of Canada .516 -.663 .251 **
Second child within 24 months .726 -.320 .193
Worked for pay 12 months before 2.012 .699 .226
first birth
Age at first birth 1.003 .003 .020
Region
Atlantic 1.720 .542 .367
Quebec .702 -.353 .254
Ontario (ref) 1.000 -- --
Prairies .507 -.679 .271 *
British Columbia .768 -.264 .333
N-585
df = 22
-2LL = 696.210
** p <.01.
* p <.05.
