ArticlePDF Available

Abstract and Figures

In this article, we investigate the effects of economic conditions, families, education, and gender ideologies on the labor force participation rates of women in eleven age groups in 117 countries. We find that participation rates of young and older women are partly explained by sector sizes and the level of economic development. However, to explain the labor force participation rates of women between 25 and 55 years, we need to study families and gender ideologies. We find these women are more likely to participate when paid maternity leave schemes exist, enrollment in pre-primary education is higher, and countries are less religious.
Content may be subject to copyright.
Working Women Worldwide. Age Effects in Female Labor
Force Participation in 117 Countries
JANNA BESAMUSCA, KEA TIJDENS, MAARTEN KEUNE and STEPHANIE STEINMETZ
*
University of Amsterdam, Netherlands
Summary. In this article, we investigate the effects of economic conditions, families, education, and gender ideologies on the labor
force participation rates of women in eleven age groups in 117 countries. We find that participation rates of young and older women
are partly explained by sector sizes and the level of economic development. However, to explain the labor force participation rates of
women between 25 and 55 years, we need to study families and gender ideologies. We find these women are more likely to participate
when paid maternity leave schemes exist, enrollment in pre-primary education is higher, and countries are less religious.
Ó2015 Elsevier Ltd. All rights reserved.
Key words — female labor force participation, motherhood, care, work, global
1. INTRODUCTION
Women are increasingly seen as the motor of sustainable
human development (UNDP, 2013; World Economic
Forum, 2014). Policy makers interpret women’s emancipation
as a proxy for equal opportunities (UNDP, 1995), micro loan
projects invest in women to improve the welfare of entire fam-
ilies (World Bank, 2012), women’s labor market integration
enhances potential for economic growth (World Economic
Forum, 2014), and female labor incomes help reduce poverty
(Buvinic & Gupta, 1997). Women’s work is thus framed as a
major force in shaping countries’ economic and human devel-
opment. In consequence, a growing number of policies are
geared toward improving women’s position and participation
in societies.
However, some debate exists as to the extent to which these
policies are beneficial to women themselves. First of all, as the
World Bank Commission on Growth and Development (2008)
comments in its final report, there are vast and unexplained
differences in countries’ experiences. Secondly, several authors
have warned that in the case of women’s employment quantity
and quality cannot be equated (c.f. Horton, 1999; Norris,
1992). Labor force participation is not always a free choice
(Elson, 1999) and may be limited to low-paid and
labor-intensive sectors and occupations (Kucera & Tejani,
2014; C¸ agatay & O
¨zler, 1995). Low economic activity may
also be desirable for some women, in particular when educa-
tion or retirement is substituted for employment (Clark &
Anker, 1993; Van Klaveren & Tijdens, 2012).
While in the third case education is clearly preferable to
employment, the first two dilemmas certainly include a trade-
off. Jordanian female teachers interviewed by Adely (2009)
describe both how they entered paid work out of necessity
and that it offers them new social networks and a legitimate
space outside the domestic sphere. In a more material consid-
eration, Sassen (1996) argues that even low-paid work
increases women’s autonomy and grants access to the public
domain. Considerable evidence exists that paid work strength-
ens wives’ position in the household (c.f. Anderson &
Eswaran, 2009; Schultz, 1990). Gray, Kittilson, and
Sandholtz (2006) point out that women’s labor incomes offer
an avenue for mending at least part of their general disadvan-
tage and Iversen and Rosenbluth (2008) find it is associated
with a greater presence of women in the public domain. We
do not argue here that female labor force participation equals
gender equality, but will advance that it can be a starting point
of a long process toward emancipation.
Any beneficial effects of female labor force participation,
however, require that policy effectively targets the group of
prime age women that sustain families. In order to do so, we
need to deepen our insights into the way in which female labor
force participation differs across countries and between groups
of women. Yet, our understanding of its dynamics is far from
complete. In their recent overview article on female employment
patterns, Steiber and Haas (2012) point out that the large major-
ity of studies either compare different women in a single country
or women in different countries, but rarely both. In addition,
with a few notable exceptions (c.f. Bloom, Canning, Fink, &
Finlay, 2007; Lincove, 2008), research in the last decade has
been split into studies of industrialized countries on the one
hand, and developing nations on the other.
In order to overcome this binary developing-industrialized
divide, this article contributes to the discussion by evaluating
which country characteristics can explain aggregate female
labor force participation in 117 countries at very different levels
of economic development. Moreover, we distinguish between
age groups, arguing that women in various stages of a life course
are confronted with different encouragements and impediments
to labor force participation. We focus on four domains of coun-
try characteristics that have been established to affect female
labor force participation rates in previous research. We firstly
look at much studied indicators of overall female labor force
participation, including economic conditions ruling the neces-
sity and opportunities to work, as well as education. We then
include two domains affecting mothers in particular: families,
including both family composition and care demands, and gen-
der ideologies that govern the extent to which women are
encouraged or discouraged from working. We argue that previ-
*The authors thank the participants of ILO’s 3rd Regulating Decent
Work conference (3–5 July 2013, Geneva) and the 8th IZA/World Bank
Conference on Employment and Development (22–23 August 2013,
Bonn), as well as two anonymous reviewers for comments on earlier
versions. The authors would like to acknowledge the contribution of the
Labour Rights for Women project. Final revision accepted: April 25, 2015.
World Development Vol. 74, pp. 123–141, 2015
0305-750X/Ó2015 Elsevier Ltd. All rights reserved.
www.elsevier.com/locate/worlddev http://dx.doi.org/10.1016/j.worlddev.2015.04.015
123
ous macro studies have underestimated the effects of those
domains by looking only at the female labor force as a whole,
ignoring that policies may affect women of various ages in differ-
ent or even opposite ways.
In order to disentangle the respective influences of the four
domains and explore their interaction, we study them sepa-
rately and refrain from employing the commonly used com-
posite indices (such as the UN Gender Empowerment
Measure or the Gender Development Index). We draw on a
variety of international data sources to construct a unique
country-level dataset of female labor force participation rates
for eleven age groups in 2010 and selected indicators that can
be attributed to the four domains in 117 countries. In Sec-
tion 2, we review the previous literature and Section 3details
the methodology and dataset. Section 4reports the results and
shows that models that take account of both country and age
differences lead to a fuller understanding of aggregate female
labor force participation. Section 5concludes and explores
avenues for further research.
2. THEORETICAL FRAMEWORK AND HYPOTHESES
We are far from the first researchers to look into the dynam-
ics of aggregate female labor force participation. Many
authors have pointed out that labor can take many different
shapes, both productive and reproductive (for an overview
of the debate, see Tancred, 1995 or Benerı
´a, 1992). In this arti-
cle, however, we aim to shed light on women’s remunerative
work, because we research the conditions under which women
join the labor market. In line with the ILO (1982), we define
labor as paid, productive work, performed outside the purely
familial sphere, but not necessarily in the formal labor mar-
ket.
1
We thus focus on access to the labor market, rather than
women’s position in it.
Since the 1970s, scholars from various fields of social
sciences have researched a range of formal and informal insti-
tutions to explain country differences and similarities in the
extent to which women join the labor force (c.f.; Boserup,
1970; Clark & Anker, 1993; Lincove, 2008; Pampel &
Tanaka, 1986; Semyonov, 1980). Their work shows us that
women’s attachment to paid labor is rarely, if ever, uncondi-
tional. Around the world, women divide their time between
household work, child rearing, home making, family enter-
prises, and the formal or informal labor market (Bardasi &
Gornick, 2000; Barrientos & Kabeer, 2004). Women might
both work or stay at home out of sheer necessity, societal sta-
tus, or beliefs (Haas, Steiber, Hartel, & Wallace, 2006).
Female labor force participation, then, is informed by the
way societies facilitate or impede it (Chang, 2004). In this con-
text, institutions, defined as webs of interrelated rules and
norms that govern social relationships(Nee, 1998, p. 8) are
essential. Various scholars have categorized countries accord-
ing to their ideal-typical institutional settings or gender
regimes, describing the key policy logics of welfare states
in relation to gender(Pascall & Lewis, 2004, p. 373). Some
institutional arrangements, these scholars have convincingly
argued, are more conducive to the labor force participation
of women than others’ equality (c.f. Chang, 2000, 2004;
Korpi, 2000; Whitehouse, 1992).
In addition, we argue that these institutional constraints
weigh differently on women of different ages due to their dis-
tinct position in the life course (for an overview of life course
theories, see Heinz & Kru
¨ger, 2001). Education may keep
school age women out of the labor market, while increasing
the opportunities for graduates. Care tasks are omnipresent
in the lives of mothers and grandmothers, but much less so
for young women. While the timing of life courses is different
in various parts of the world, we argue that, on the aggregate
level, women everywhere go through stages of school-going,
transitions to adulthood, motherhood with care for young
children, care for older children, and grand-motherhood. In
short, we view women’s capabilities to work or not, as revolv-
ing around a balance of economic, educational, family, and
gender ideological influences that affect women in different
manners depending both on the country they live in and their
position along the life course. In the remainder of this section,
we study the relation between each of these four domains and
the aggregate female labor force participation rate (FLPR).
(a) Economic conditions
Economic conditions can provide both the necessity and the
opportunity to work. While in some societies labor force par-
ticipation is required to make ends meet, in others families
may designate a single earner (Steiber & Haas, 2012). Further-
more, the availability of suitable jobs may draw women into
the labor market or keep them out. Economic conditions affect
female labor force participation through economic necessity
and through the sizes of its respective sectors, which determine
the kind of work that is available (Pampel & Tanaka, 1986).
The relation between the level of economic development and
aggregate female labor force participation is generally
observed to be u-shaped (Haghighat, 2002; Lincove, 2008;
Tam, 2011). However, as Semyonov (1980) observes, while
economic development may be indicative of the opportunities
women have in a labor market, the association is created by
social factors. In recent years, Semyonov’s argument has been
confirmed in multi-country studies by Chang (2004) and
Lincove (2008) who both find that there is no u-shaped rela-
tionship when countries at similar levels of development are
compared or when studying countries’ changes in economic
development and in FLPRs concurrently.
The u-shaped relation, then, should be attributed to various
social relations and labor market structures associated with
higher and lower levels of economic development. Economic
necessity is the first of those explanations. Increases in the wel-
fare of workers in periods of industrialization are associated
with the material possibility for wives to withdraw from paid
labor as a sign of affluent family status (see Goldin (2006)
for the USA, Safa (1977) for Latin American or Bhalla and
Kaur (2011) for India). Elson (1999) describes this as the move
from labor force participation for survival to a genuine choice
for (middle class) women to work or not. In a study of Wes-
tern and Eastern European countries, Haas et al. (2006,
p. 767) too, argue that theorizing should take account of
the economic necessity for many women in less prosperous
countries to work full time. Thus, where income from work
is desirable in high-income countries and essential for survival
in low-income countries, withdrawal from the labor market
can be a luxury of sorts in middle-income countries. We expect
that the relation between a country’s level of economic prosper-
ity and the FLPR is u-shaped (H1a).
However, it is questionable whether the abovementioned
economic conditions have the same effect throughout the life
course. Looking specifically at labor force participation of
older women and men in 151 countries, Clark and Anker
(1993) conclude that the FLPR of older women decreases with
economic development and accompanying changes in the
organization of society, such as the availability of old age pen-
sions. The same expectation can reasonably be voiced for
young people, who benefit from the increased educational
124 WORLD DEVELOPMENT
opportunities that are associated with higher levels of eco-
nomic development (Van Klaveren & Tijdens, 2012). School
and pension aged workers are thus exempted from the
economic necessity to work. We expect that a higher level of
economic prosperity is associated with a lower FLPR of women
of school-going age and approaching the retirement age (H1b).
Broadly expecting the same result but following a different
reasoning, scholars have studied the size of different sectors of
economic activity as indicators of the opportunities for women
in a labor market (Mehra & Gammage, 1999; Rendall, 2013;
Schultz, 1990). In one of the first extensive works including 70
countries at various developmental levels, Pampel and Tanaka
(1986) reason that women’s jobs in agriculture disappear in
the process of industrialization, thus leading to a decline in
FLPRs. As economies grow further and a services sector devel-
ops, women are drawn back into paid work due to greater labor
demand and the easier reconciliation of work and family in ser-
vices compared to industry. Studying the evolution of female
labor force participation rates in four middle-income coun-
tries—Brazil, Mexico, India, and Thailand—Rendall (2013)
recently argued that female employment opportunities increase
as the available jobs become more intellectually, as opposed to
physically demanding.
Haghighat (2002) analyzing the effect of economic growth
on the share of female employment in three sectors in 136
countries, notes that the effect of economic development on
the female share of employment is negative in agricultural, flat
for industrial and positive on services employment, thus trac-
ing out the u-shape. Studying 67 Turkish provinces, Tansel
(2001) finds a u-shaped relationship with GDP if agricultural
labor is included, but not when analyzing only the nonagricul-
tural female labor force. Lincove (2008, p. 59), similarly con-
cludes that while service employment increases female
participation...industrial labor [does not reduce it]. This sug-
gests that Rendall’s results might be indicative of a process
where heavy industrial labor is replaced by more service-and
export-oriented industry, a process that is currently prominent
in Central America and South East Asia. Studying the manu-
facturing sector in these regions, Kucera and Tejani (2014)
find that the initially labor-intensive production methods in
export-oriented manufacture are associated with a feminiza-
tion of the labor force. Rendall’s shift from physical to intel-
lectual work or Haghihat’s observed increase in services
employment, should then be interpreted as the upward turn
of the u-shaped relation between economic conditions and
the FLPR. Therefore, we expect that large agricultural and ser-
vices sectors are associated with high FLPRs,whereas large
manufacturing sectors are associated with low FLPRs (H2a).
In agriculture, a disproportionally large effect may be
expected on younger and older women, since many of the jobs
in the sector are associated with early entry and late exit from
the labor market. In its 2010 report, the UN’s Food and
Agriculture Organization points out that 60% of child laborers
work in agriculture.
2
Thus, while the initial negative effect of
declining agriculture on labor force participation is shared by
all age groups, older and younger women are expected to
remain out of the labor force to a greater extent. We expect
the positive effect of large agricultural and services sectors and
the negative effect of large manufacturing sectors to exist for
all women but to be strongest for young and older women
(H2b).
(b) Education systems
Women’s opportunities to work can be further strengthened
by investments in human capital. More training and education
equips women for different occupations and raises their rela-
tive skill levels compared to men (c.f. Abramo & Valenzuela,
2005; Apps & Rees, 2001; Engelhardt & Prskawetz, 2004;
Lincove, 2008). Some authors have also argued that a higher
share of girls in schools leads to more gender egalitarian atti-
tudes as boys and girls study together and are taught the same
skills (c.f. England, Gornick, & Shafer, 2012; Spierings, Smits,
& Verloo, 2010). Education, then, can both enlarge opportuni-
ties in the labor market and can make paid employment more
attractive compared to home making.
Particularly in economics, the relative position of women in
a labor market is considered quintessential to their decision to
participate. New home economics predict that as women’s
wages and education go up, the opportunity cost of leaving
the labor market for the sake of childbirth and care work
increases (c.f. Apps & Rees, 2001; Engelhardt & Prskawetz,
2004). For a sample of 17 high- and middle-income countries,
England et al. (2012) argue that higher educated women are
more often employed than their lower educated counterparts,
because the opportunity–cost effect is stronger than the
income-effect. Instead of withdrawing from the labor market
when a husband’s salary suffices (income effect), highly edu-
cated women refuse to forego the careers that will bring addi-
tional income and the application of their gained skills
(opportunity cost). Tansel (2001) finds that highly educated
Turkish women are more likely to be in the labor force than
their lowly educated peers, as do Bhalla and Kaur (2011) for
India and Aromolaran (2004) for Nigeria.
Studying 18 Latin American countries, Abramo and
Valenzuela (2005) alternatively posit that the FLPR is higher
at higher educational levels and household incomes, because
more highly educated women have more means to outsource
their homework. In turn, these career women create informal
sector jobs in their households for lower educated women,
resulting in a cascading effect. More women are drawn into
the labor market as the vanguard requires other women to
perform domestic tasks left undone. The literature thus sug-
gests that higher levels of education increase the FLPR. It also
suggests that this effect becomes stronger as women reach
higher levels of education, increasing the opportunity cost of
staying at home. Thus, we expect that higher levels of female
enrollment in education are associated with higher FLPRs and
that this association is stronger for higher levels of education
(H3a). We do not expect major age effects of education, aside
from inclusion in educational institutions delaying labor mar-
ket entry. We hypothesize that the effect of female enrollment in
education is negative for women below 20 (H3b).
Several studies indicate that male levels of education do not
weaken but strengthen female labor force participation
(Aromolaran, 2004; England et al., 2012; Ganguli,
Hausmann, & Viarengo, 2013; Spierings et al., 2010). Using
data from the Integrated Public Use Micro Data Series
(IPUMS), Ganguli et al. (2013) find that larger education gaps
are positively associated with gender gaps in labor force par-
ticipation. England et al. (2012) find positive effects of male
education and posit that education inculcates
gender-egalitarian values.Spierings et al. (2010), researching
female labor force participation in North Africa and the Mid-
dle East, find positive effects of a more equal share of boys and
girls in education suggesting enrollment parity creates more
gender egalitarian attitudes. Parity of boys and girls in the
educational system both implies that women and men have
approximately the same skills, as well as creates an environ-
ment in which gender egalitarian values can flourish. We
expect that more gender parity in educational enrollment is asso-
ciated with higher FLPRs (H4).
WORKING WOMEN WORLDWIDE. AGE EFFECTS IN FEMALE LABOR FORCE PARTICIPATION IN 117 COUNTRIES 125
(c) Families
Women’s opportunities to participate in the labor force are
in practice often limited by the division of care work in a soci-
ety. As Barrientos and Kabeer (2004) point out, the burden of
domestic and care work is an impediment for working women
in countries at any level of development. Care burdens, either
by the extent to which they exist or their incompatibility with
paid work, have the potential to hinder female labor force par-
ticipation. Therefore, it is an important factor to be considered
as a source of competing time demands that women face. It
can be expected that the larger the time demand is, the smaller
women’s capabilities will be to participate in the labor force.
The most straightforward way to proxy the care burden is
through the average number of children per woman. While
the relation between fertility and labor force participation
has been called endogenous (c.f. Steiber & Haas, 2012), a point
we will address in Section 3, fertility is one of the most
researched indicators of female labor force participation (c.f.
Ahn & Mira, 2002; Bloom et al., 2007; Engelhardt &
Prskawetz, 2004; Mishra, Nielsen, & Smyth, 2010). A number
of studies have investigated whether fertility rates can effec-
tively cause changes in FLPRs (Agu
¨ero & Marks, 2010;
Cruces & Galiani, 2007; Orbeta, 2005; Angrist & Evans,
1998). Mishra et al. (2010) show that a one percent increase
in fertility leads to a 0.4% drop in FLPRs in the G7 countries.
Orbeta (2005) found that each child below school going age
lowers a Philippine woman’s probability of labor force partic-
ipation by 7.2% while Cruces and Galiani (2007) found a 5%
decline in Argentina and 3.5% in Mexico.
Inherently, the abovementioned effects are focused on moth-
ers and grandmothers, who are caregivers in a way that
younger women are not (for an overview of grandparenting
in industrialized countries, see Arbor & Timonen, 2012). In
one of the few studies including developing countries, Bloom
et al. (2007) report that the effect of one additional child per
woman in a country reduced the labor force participation
of women between 25 and 29 years by ten to 15% and that
of women between 40 and 49 by five to ten percent. Thus,
we expect that higher care burdens are associated with lower
FLPRs of women of childbearing age,and not associated with
FLPRs of women below 20 and above 50 (H5).
While the relation between the total fertility rate and the
FLPR was traditionally found to be negative, since the late
1980s several studies have shown a coincidence of high fertility
rates and high female labor force participation in some
high-income countries and a low–low combination in others
(for a historical overview, see Ahn & Mira, 2002; Engelhardt
& Prskawetz, 2004). Cruces and Galiani (2007) point out that,
while dropping fertility rates can be directly linked to the
increase in female labor force participation in Argentina, it
explains only a small share of the hike in participation rates
of Mexican women. This would suggest that some societies
are more successful in mitigating the incompatibility of work
and motherhood than others (c.f. Engelhardt & Prskawetz,
2004; Gornick, Meyers, & Ross, 1997; Apps & Rees, 2001).
Two of the most frequently employed policies to increase
opportunities for mothers to stay in the labor market are
childcare and maternity leave. Studying 22 industrialized
countries, Mandel and Semyonov (2006) find that the number
of fully paid weeks of maternity leave and the percentage of
pre-school children in publicly funded childcare are both asso-
ciated with a higher FLPR. There is a large body of literature
that links paid maternity leave arrangements in industrialized
countries, providing continued income and guaranteeing the
right to return to one’s old job, to continued labor force par-
ticipation of mothers (c.f. Aisenbrey, Evertsson, & Grunow,
2009; England et al., 2012; Gornick et al., 1997; Steiber &
Haas, 2012). Several authors, however, noted that arrange-
ments allowing extremely long periods of maternity leave
reverse the positive effect and actually decrease labor force
participation (Stryker, Eliason, Tranby, & Hamilton, 2011;
Hummelsheim & Hischle, 2010). We expect that the relation
between the length of maternity leave and the FLPR forms an
inverted u-shape,meaning that both the absence of as well as
the presence of very long maternity leaves are associated with
lower FLPRs of women between 25 and 44,whereas brief mater-
nity leaves are associated with higher FLPRs of women between
25 and 44 (H6a).Furthermore,we expect that higher levels of
wage replacement during maternity leave are associated with a
higher FLPR of women between 25 and 44 (H6b).
Stryker et. al. (2011) argue that widely available and afford-
able childcare both reduces the time incompatibility of work
and motherhood, as well as reshape[s] cognitive expectations
and normative evaluations about the acceptability or desir-
ability of childcare provided outside the home and by someone
other than the mother. For a number of European countries,
they find public childcare facilities increase female labor force
participation. These findings are confirmed by Gornick et al.
(1997) as well as Hummelsheim and Hirschle (2010).We
expect that higher enrollment in childcare is associated with
higher FLPRs of women of 25 and above (H7a).
Hummelsheim and Hirschle (2010) find the effects of child-
care arrangements in Belgium and Germany are largest for
mothers with children under three and decrease or even disap-
pear due to cultural attitudes at later ages. We therefore argue
that care burdens begin to affect female labor force participa-
tion when women become mothers and are reduced as children
grow older and less in need of constant care. We expect that
childcare and maternity leave provisions are not associated with
the FLPR of women below 25 and above 44 (H7b).
(d) Gender ideologies
Next to reducing competing time demands of care tasks and
investments in women’s capabilities, societies may also influ-
ence female labor force participation more indirectly. As
Nee (1998, p. 10) points out, [t]he cultural heritage of a soci-
ety is also important because customs, myths, and ideology
matter in understanding the mental models of actors.
Through formal and informal institutions, countries may
express a preference, objection, or indifference to the inclusion
of women in the labor market. Goldin (1995) argues that if the
societal prejudice against working wives is strong, they will be
much less susceptible to the economic motivations discussed in
Section 2(a). Cultures can mediate the acceptability of a
woman’s choice to work, thus determining the popular image
of appropriate female behavior (Hakim, 2000). By guarantee-
ing equal treatment, countries can provide women with a gen-
uine choice to integrate into the labor force as well as convey a
formal commitment to gender equality. Traditions, reflected
among others in the prevalence of religious beliefs may con-
vene the inappropriateness of working. On the other hand,
the visibility of women in public life, such as in politics or pub-
lic office can signal the de facto acceptance and respectability
of such a choice.
In a sample of 13 high-income countries, Whitehouse (1992)
finds no relation between the adoption of equality legislation
and female labor force participation. However, taking a more
global sample, Chang (2000, 2004) reports that countries that
ratified the ILO gender conventions had higher FLPRs. While
one may argue that such treaties can be implemented in a
126 WORLD DEVELOPMENT
variety of ways and that they may be adopted both in coun-
tries with a long tradition of nondiscrimination legislation as
well as in those that make a policy shift at the moment of rat-
ification, the results are encouraging. The mere presence of
equality legislation may be the first token of a gender equality
commitment and is at present the most suitable measure that is
available for a large number of countries. We expect that the
existence of anti-discrimination laws is associated with a higher
FLPR of women of all ages (H8).
Quite a few studies have looked into the effect of religion on
female labor force participation (c.f. Clark, Ramsbey, & Stier
Adler, 1991; Haghighat, 2002; Hummelsheim & Hirschle,
2010; Lincove, 2008). Contrary to studies into the effect of
the intensity of religious views or practices (c.f. Amin &
Alam, 2008; Chadwick & Garrett, 1995; Heineck, 2004;
Lehrer, 2004), we treat religion rather as a proxy for the mores
in a society as a result of the historical imprint that a denom-
ination has left on a culture. The prominence of religious
beliefs has been associated with lower FLPRs (c.f.
Psacharopoulos & Tzannatos, 1989). Clark et al. (1991), for
instance, explore the labor force participation of women in
Islamic, African, Asian, Marxist, Western, and Latin Ameri-
can world regions. They report that, in comparison to Western
countries, women in Islamic and Latin American countries are
less likely to join the labor force. Studying data from the
World Values Survey from up to 97 countries, Seguino
(2011) links religious institutions not only to more traditional
beliefs with regard to gender roles, but also to more unequal
labor market outcomes.
Islam in particular is often found to be associated with lower
female labor force participation as women are more explicitly
restricted to the private sphere (Clark & Anker, 1993;
Haghighat, 2005; Lincove, 2008). We argue that all major reli-
gions have attributed different roles to women and men. This
applies particularly for mothers, as women’s role in religious
thinking is often associated with motherhood. However, we
argue that the extent to which their teachings have left an
imprint on society and influence behavior will be dependent
on their pervasiveness or dominance in a country. A religion
to which the large majority of a country’s population is affili-
ated, will thus have a stronger influence on behavior than a
smaller religious community. We expect that the pervasiveness
of religions in general and Islam in particular,is negatively asso-
ciated with the FLPR and that this association is strongest for
women between 25 and 44 (H9).
Juxtaposed to the influence of tradition on women’s roles
in a society, is their visibility in public life today. The equal
presence or near absence of women in institutions, politics
and media is related to the self-evidence of a woman’s
choice to work (Chafez, 1990). Haghighat (2005) has argued
that the political empowerment of women can moderate the
effect of religion. Iversen and Rosenbluth (2008, p. 481),
too, interpret the presence of women in politics as a signal
that a society has loosened its attitudes towards appropri-
ate levels of gender specialization and traditional gender
roles. Studying 23 OECD countries, Iversen and
Rosenbluth (2008, p. 481) find that countries with higher
FLPRs have a higher representation of women in Parlia-
ments, as do Stockemer and Byrne (2012) for a sample of
120 countries. Following this line of reasoning, we argue
that the political rights and participation of women reflect
the extent to which the presence of women in the public
domain is accepted in a country. We expect that a greater
presence of women in the politics is associated with higher
FLPRs of women of all ages (H10).
3. DATA AND METHODOLOGY
(a) Data
For the study of aggregate female labor force participation,
the most comprehensive dataset currently available is the 6th
edition of the ILO Estimates and Projections of the Econom-
ically Active Population (EAPEP). The dataset aggregates and
harmonizes data from selected national labor force and house-
hold surveys that are comparable for different age groups and
include both urban and rural areas. EAPEP is a
cross-sectional time series containing estimates of countries’
population sizes, economically active populations, and labor
participation rates for women and men. Data for 191 countries
for the time period during 1990–2010 are reported for 11 age
categories. Due to the linear interpolation method used to
construct the dataset, we do not perform any longitudinal
analyses and include only the last available year, 2010 (for a
detailed description of the imputation procedure: see
International Labor Organization, 2011). For reasons of reli-
ability 23 countries, mainly African nations and dictatorships,
for which no actual observations are available, were removed
from the sample.
3
Secondly, to avoid statistical complications,
32 countries whose labor force is below one million were left
out of the analysis.
4
To complete the Global Dataset on Women and Work,
described in Appendix II, data for the four explanatory
domains—economic conditions, education, families, and gen-
der ideologies—were gathered from a range of publicly avail-
able international sources. Variables were selected on the basis
of their availability for both industrialized and developing
countries. Because we aimed to maximize the number of coun-
tries in the analysis, we were occasionally forced to leave out
variables that are undeniably important for a woman’s deci-
sion to work or not, but are not easily measured in some parts
of the world. We perform the analyses on the final sample of
117 countries for which data on FLPRs as well as the four
domains are available.
(b) Operationalization
The dependent variable is the female labor force participa-
tion rate, which measures the share of economically active
women to all women in the relevant age group in a country.
The variable is broken down into ten five-year age categories
and one age group for women of 65 years and above, creating
11 observations per country.
To measure the effect of economic conditions, we specified
the level of economic development and the relative size of eco-
nomic sectors. For economic wealth, we use per capita GDP
in 2010.
5
To test the hypothesized u-shaped relationship, we
standardize the variable with a mean of zero and a standard
deviation of one, and calculate a square term.
6
In the absence
of data on the size of sectors relative to total employment, we
measure sector sizes by their value added as a share of GDP.
7
We include predictors for agriculture, manufacturing and ser-
vices, which are three broad but nonoverlapping sectors.
In the domain of education, we use variables for female
enrollment levels and gender parity in education. We take data
from 2010, except in a few cases where this year is not avail-
able and we select the previous or following year.
8
We intro-
duce two variables for the gross female enrollment rate in
primary and secondary education, measuring the share of girls
in education as a percentage of all girls in the relevant age cat-
egory. Secondly, we introduce two variables for gender parity
WORKING WOMEN WORLDWIDE. AGE EFFECTS IN FEMALE LABOR FORCE PARTICIPATION IN 117 COUNTRIES 127
in education, reflecting the share of female-to-male enrollment
in primary and secondary education. Due to nonavailability of
data for many developing countries, we exclude measures for
tertiary education.
For the families domain we formulated expectations regard-
ing the care burden,maternity leave, and childcare. As noted in
Section 2(c), several authors have argued that the choices to
have children and to work are made together rather than inde-
pendently of each other, bringing up questions regarding the
direction of causality as well as whether the two processes
may simultaneously be caused by other exogenous factors
(Browning, 1992; Steiber & Haas, 2012). If such an endoge-
nous relationship exists, only part of the variation captured
by the fertility rate can be attributed to a direct effect of having
children and we overestimate the effect. Therefore, we test the
effect of care burdens both by introducing the total fertility
rate as well as constructing a measure that attempts to address
the alleged endogeneity. We create a scale containing four
items affecting the mean care burden and measuring the con-
sequences of past fertility decisions, rather than the concurrent
fertility and FLPRs. We use the share of the population below
15 to represent dependent family members and the share of the
population above 65 years for the number of adults as a mea-
sure of people without small children with whom both work
and care tasks could be shared. We also introduce the average
age of first marriage of women and the average life expectancy
of women to accommodate country differences with regard to
the relative share of a life span spent nursing children.
9
The
share of the population above 65, age of first marriage and life
expectancy are reverse coded in order for a higher score on the
scale to represent a higher care burden. The scale (Cronbach’s
alpha .79) yields near identical coefficients to a fertility vari-
able in a bivariate regression, but halves the effect sizes of fer-
tility in the multivariate regressions. We interpret this as an
indication of endogeneity and use the scale instead of the total
fertility rate. We measure both the length in days and wage
replacement levels of maternity leave. As we assume an
inverted u-shaped relation between the FLPR and the length
of maternity leave, we also calculate a squared term. World-
wide data on the enrollment of small children in childcare is
not available. Therefore, we use pre-primary enrollment of
both boys and girls as a proxy for childcare.
10
To measure gender ideologies, we look at the existence of
anti-discrimination legislation, the religious background of a
country, as well as the presence of women in politics. For the
existence of legislation protecting women from discrimination
in the labor market, we look at ratification of the ILO treaties
on maternity, nondiscrimination, and equal pay. We create a
scale by adding the three items (Cronbach’s alpha .68). To
measure religious backgrounds, we use data collected by the
Pew Research Center Forum on Religion and Public Life, con-
taining data on adherence to various streams of Christianity,
Judaism, and Islam, as well as Buddhism, Hinduism, Confu-
cianism, a multitude of smaller Asian, syncretic and animist
religions and the share of nonadherents. We construct a mea-
sure of religious dominance, reflecting the size of the biggest
religion in a country as a proxy for the cultural imprint of reli-
gion in general. Secondly, we construct a variable measuring
the influence of Islam. To create the variable we divide the
share of the population adhering to Islam by the size of the
largest religious group, or by the share of nonaffiliated people
if this is the largest group, creating a scale that runs from 0 (no
influence) to 1 (Islam is the largest religion). Finally, to mea-
sure women’s current level of representation in the public
domain, we use a variable measuring women’s right to vote,
run in elections, hold government office, join political parties
and to petition officials. The variable is a four-point scale run-
ning from 0 (no rights) to 3 (rights are guaranteed both by law
and in practice).
(c) Analytical strategy
Taking into account that observations of different groups of
women in a country are not independent from each other, we
model the effects of the specified four domains on the FLPR of
eleven age groups in 117 countries, using two-level hierarchical
models (c.f. De Leeuw & Meijer, 2008; Hox, 2010). We specify
a model with a random intercept for the countries as well as a
random slope for the age variables, where FLPRij is the FLPR
by country and age group.
FLPRij ¼c00 þc10Ageij þc20 Age2
ij þc01Xjþc02 ðXjAgeijÞ
þc03ðXjAge2
ijÞþc04 MLPRij þd0j
þd1jAgeij þd2jAge2
ij þeij
We introduce a random intercept, which consists of the
grand mean intercept for all countries (c00) and a level 2 error
term (d0jÞthat contains the deviation from the mean intercept
for each country. In the null model (model 1) we introduce
only the dependent variable. Model 1 shows that 33% of the
differences in participation levels are differences between coun-
tries (level 2 variance), whereas 67% are found within coun-
tries (level 1 variance). The within-country unexplained
variance is vastly reduced when we include a second-order
orthogonal polynomial term for age (model 2), which intro-
duces a contrast coded first and second-order centralized vari-
able for the 11 age categories.
The 117 countries in our sample differ vastly in terms of the
timing of life events like childbirth, the end of full-time educa-
tion and retirement. Therefore, in model 3 (Table 1) we allow
the age effects to vary between countries and add a random-
ized age slope to account for the differences in the timing of
different stages of women’s life cycle across the countries.
Thus, we include a grand mean effect of age (c10Ageij ) and
age squared (c20Age2
ij), as well as two level 2 error terms
(d1jAgeij;d2jAge2
ijÞ. In our models, we restrict the level 1 errors
assuming homoscedasticity; identical models allowing for
heteroskedastic errors (not shown) were run and yield similar
results.
To measure the effects of the four domains, we use the run-
mlwin package in stata. Firstly, we introduce a vector of
country-level explanatory variables (c01Xj). We interact those
variables with the randomized age and age square variables
creating two cross-level interaction terms (c02ðXjAgeij Þ;
c03ðXjAge2
ijÞ) in order to estimate the heterogeneous effects
by age groups. In order to test the effects of the different
domains, we follow a three-stage approach and examine if
the results hold in all models. For each of the four domains,
we firstly test each hypothesis in one model and then run a
model combining all indicators of the respective domain.
Finally, we run a complete model including all four domains.
We seek to measure only those effects that are specific to
women. By controlling for the male labor force participation
rate (c04MLPRij ), we both isolate those effects that specifically
apply to women, as well as filter out the nongender biased
measurement differences in the national data that the ILO uses
as sources of the EAPEP dataset. We standardize all variables
so as to make coefficients comparable in the multivariate mod-
els. To allow for a more intuitive understanding of the age
effects than afforded by the relatively complicated interaction
128 WORLD DEVELOPMENT
terms with the orthogonal polynomials in the regression mod-
els, we use predicted values to plot the differential effects on
the eleven age groups. In order to test the robustness of the
results, we run the same models on split samples along country
income groups and discuss the results in Section 4(f).
4. RESULTS
At first glance, the same dominant pattern can be observed
in all countries (Figure 1): the FLPR is lowest at the extremes
of the horizontal axis, at young and old age, and higher in
between. On average, 55% of women are in the labor force.
Yet fewer than two in ten women above 65 and less than three
in ten women below 19 are, compared to seven in ten women
between 30 and 49 years. However, both the level of participa-
tion across the life cycle as well as the exact location of the
peak and the rate of decline differ vastly across countries.
Among the 117 countries are those whose FLPR at no point
falls below 50% as well as states where it never reaches that
mark.
When we plot the FLPRs per age group in nine selected
countries we find diverse patterns (Figure 1). For instance,
while low-income Sub-Saharan African countries like Burundi
have high levels of participation across age groups, many Mid-
dle Eastern and North African countries, like Algeria, show
comparable patterns but on a much lower level of participa-
tion. Contrastingly, most Asian, Latin American, and Euro-
pean countries reveal large differences in levels of
participation of the different age groups. Some countries, like
the Czech Republic and, to a lesser extent the USA, show a
double peak, indicative of mothers withdrawing from the
labor market for a couple of years when children are young.
(a) Economic conditions
We hypothesized that the relationship between countries’
level of economic prosperity and the FLPR was u-shaped, as
well as that female labor force participation is higher in coun-
tries with large agricultural and services sectors and lower
where manufacturing sectors are big.
The relation between per capita GDP and female labor force
participation takes the expected u-shaped form (Table 2), but
050 100
050 100
050 100
15 40 65 15 40 65 15 6540
Algeria Burundi Czech Republic
Netherlands Nicarag ua Sweden
Thailand Ukraine United States
Age group
Figure 1. Female labor force participation rate by age for selected countries.
Table 1. Female labor force participation—null models
Model 1 Model 2 Model 3
Constant 55.349
***
(1.564) 55.349
***
(1.564) 55.349
***
(1.564)
Age 4.673
***
(0.310) 4.673
***
(0.449)
Age squared 17.529
***
(0.310) 17.529
***
(0.665)
Level 1 variance 485.826
***
(20.086) 123.792
***
(5.118) 51.256
***
(2.369)
Level 2 variance (cons) 242.061
***
(37.467) 274.973
***
(37.425) 281.568
***
(37.423)
var(age) 18.890
***
(3.087)
var(age^2) 47.051
***
(6.764)
Log Likelihood 5916.0957 5116.2534 4819.6890
AIC 11838.19 10242.51 9685.997
BIC 11853.67 10268.31 9722.118
***
p< 0.01,
**
p< 0.05,
*
p< 0.1.
Source: Global Dataset on Women and Work, sample contains 1,287 age groups in 117 countries, Standard errors in parentheses.
WORKING WOMEN WORLDWIDE. AGE EFFECTS IN FEMALE LABOR FORCE PARTICIPATION IN 117 COUNTRIES 129
is nonsignificant and thus cannot confirm the hypothesis. The
linear term in model 4 shows the association is initially nega-
tive (1.222), forming the downward half of the u-shape and
indicating that a higher per capita GDP is initially associated
with a lower FLPR The positive square term (2.233) models
the point at which the association turns positive, forming the
upward half of the u-shape where GDP and the FLPR rise
together In model 5, agriculture has a strong positive effect
(6.230), as does the services sector (5.437), implying that coun-
tries with larger agricultural and services sectors do indeed
have higher overall FLPRs. We then control for both per cap-
ita GDP and sector sizes (model 6), so as to be able to measure
the impact of sector size on the dependent variable for coun-
ties with the same level of per capita GDP and vice versa.
The effects of agriculture become stronger and those of ser-
vices weaker, but remain significant. The main effect of manu-
facturing is not significant in any model and the effect of per
capita GDP remains nonsignificant but becomes positive.
The significant interaction between age and per capita GDP
confirms that there are differences in the way age groups are
affected. The negative coefficients of the interaction effects
between age and GDP (2.530) and the positive effects of
the interaction with GDP squared (0.730), indicate that the
u-shape is more pronounced for younger and older women.
Figure 2 visualizes the u-shaped relation for each of the age
groups, showing the FLPR at the vertical axis and the stan-
dardized GDP variable on the horizontal axis. Most age
groups experience only a small dip in FLPRs between the low-
est values for per capita GDP toward the mean, and then show
a steady positive effect. The initial negative effect of GDP is
larger for women below 20 and above 60, but the recovery is
also quicker. Figure 2 also shows that the participation rates
of women of different ages are most similar in countries with
very low or high per capita GDP, where the regression lines
for the age groups almost touch. The participation rates of
young and older women diverge most from the middle-age cat-
egories in countries with mid-level per capita GDP, as is
shown by the large distance between the age groups on the ver-
tical axis. Thus, contrary to expectations, the relation between
per capita GDP and the FLPR shows the most pronounced
u-shape in the case of school age and older women.
The effects of sector sizes are also different for women of dif-
ferent ages, as shown by the significant interaction terms and
the diverging regression lines for these age groups (Figure 2).
The largest age effects exist in countries with small agricultural
sectors, large manufacturing sectors, and large services sectors.
Larger agricultural sectors are associated with higher overall
FLPRs, but Figure 2 shows that the effect is much larger for
women below 20 and above 60 years. Manufacturing has
hardly any effects on most age groups, but negative effects
on women under 25 and of 55 and older. Service sectors have
a positive effect on most age groups, but a negative effect on
the labor force participation of women under 20 and above 60.
We thus find significant main effects of sectors sizes, but not
of GDP; we find significant age effects for both. The younger
or the older women are, the more u-shaped is the relationship
between their labor force participation rate and per capita
GDP. This would suggest that the association between the
level of economic prosperity and female labor force participa-
tion runs primarily through shifts in the timing of entry into
and exit out of the labor market. Women are more likely to
work in countries with larger agricultural and services sectors,
while the predicted negative effect of manufacturing is found
only for younger and older women. When controlling for sec-
tor sizes, the nonsignificant effect of per capita GDP becomes
positive, except for younger and older women.
Table 2. Effects of economic conditions on female labor force participation rates
Model 4 Model 5 Model 6
Constant 51.798
***
(1.985) 53.967
***
(1.358) 54.052
***
(2.012)
Age 4.067
***
(0.434) 3.538
***
(0.315) 4.298
***
(0.446)
Age squared 3.466
***
(0.923) 2.827
***
(0.712) 3.224
***
(0.934)
Per capita GDP 1.224 (2.856) 5.089 (3.614)
Per capita GDP squared 2.233 (1.416) 0.053 (1.531)
Age * GDP 1.636
***
(0.583) 2.530
***
(0.752)
Age * GDP^2 0.467 (0.285) 0.730
**
(0.319)
Age^2 * GDP 1.541 (1.066) 0.852 (1.372)
Age^2 * GDP^2 0.399 (0.527) 0.246 (0.581)
Agriculture 6.230
***
(1.826) 8.042
***
(1.999)
Age * Agriculture 0.416 (0.393) 0.277 (0.419)
Age^2 * Agriculture 1.091 (0.672) 1.304
*
(0.761)
Manufacturing 0.152 (1.419) 0.405 (1.368)
Age * Manufacturing 0.083 (0.300) 0.050 (0.284)
Age^2 * Manufacturing 0.941
*
(0.519) 0.954
*
(0.519)
Services 5.437
***
(1.747) 3.802
***
(1.772)
Age * Services 0.113 (0.369) 0.525 (0.368)
Age^2 * Services 1.989
***
(0.638) 1.848
***
(0.672)
Level 1 variance 24.123
***
(1.115) 24.122
***
(1.115) 24.119
***
(1.115)
Level 2 variance (cons) 219.771
***
(29.020) 206.982
***
(27.348) 191.303
***
(25.298)
var(age) 6.405
***
(1.129) 7.120
***
(1.222) 6.131
***
(1.093)
var(age^2) 28.142
***
(3.967) 25.726
***
(3.651) 25.622
***
(3.638)
Log likelihood 4378.2505 4374.5806 4363.1885
AIC 8788.501 8787.161 8776.377
BIC 8871.062 8885.202 8905.379
***
p< 0.01,
**
p< 0.05,
*
p< 0.1.
Source: Global Dataset on Women and Work, sample contains 1,287 age groups in 117 countries, Standard errors in parentheses, controlled for male
labor force participation rate.
130 WORLD DEVELOPMENT
(b) Education systems
We hypothesized that both enrollment and gender parity in
primary and secondary education are positively associated
with the FLPR. In our models (Table 3), we confirm that
higher enrollment and parity in primary education have posi-
tive effects (3.206 and 3.392). One standard deviation increase
in girls’ enrollment in primary education or gender parity in
primary education is associated with a three percent increase
in the FLPR. Contrary to expectations, however, gender par-
ity in secondary education has a negative effect (4.365) and
female enrollment in secondary education is nonsignificant.
Thus, while higher levels of completed education may effec-
tively improve women’s human capital and their individual
position in the labor market, we cannot confirm here that con-
tinued enrollment in education also improves aggregate female
labor force participation.
We find relatively few significant age effects of education on
female labor force participation. The only significant age
effects are found in the interaction term with girls’ enrollment
in secondary education in model 7, indicating that effects are
more strongly negative for younger and older women. How-
ever, these effects disappear when gender parity in secondary
education is introduced in model 9. We thus confirm the pos-
itive effects of primary education and gender parity in primary
education. However, we reject the hypothesis that these effects
would be stronger for secondary education, where we effec-
tively find a negative relationship.
(c) Families
We hypothesized that FLPRs will be lower in countries
where the care burden for young dependents is larger. We also
expect that labor force participation will be higher where brief
maternity leave arrangements and higher wage replacement
levels exist, and a larger share of children are enrolled in
pre-primary education.
To measure the care burden, we use the dependency scale
described in Section 3(b). Its main effect is not significant
(Table 4). We find no significant effect of pre-primary enroll-
ment and wage replacement during maternity leave, but the
length of maternity leave shows the predicted inverted
u-shape. The strongly positive untransformed term (7.350)
shows that women are more likely to work as they are entitled
to longer periods of maternity leave. The negative squared
term (1.759) indicates that this effect wears off as the length
becomes more extended and women are effectively less likely
to work when maternity leave periods are very long. When
tested together in one model (model 13), the effects are
unchanged.
However, while the grand mean effects do not reveal much
effect of families on FLPRs, the picture changes when we dis-
tinguish between age groups. All indicators, with the exception
of wage replacement during maternity leave, have significant
age effects. The largest age effects are found in countries with
low care burdens, moderately long maternity leave periods
and high enrollment in pre-primary education. As shown in
Figure 3, the effect of a higher care burden is especially nega-
tive for women between 30 and 54, the group most likely to
have small children at home. However, for women under 20
and above 55 the effect is positive, which could indicate that
they substitute mothers in the labor force. Especially women
of 60 and above, whose labor force participation rate is rela-
tively high in countries with a higher care burden, are much
less active in societies with lower care burdens.
We also find significant age effects of the length of maternity
leave. As Figure 3 shows, the effect on some age groups is
actually opposite to the grand mean effect. The length of
maternity leave has the hypothesized inverted u-shaped rela-
tionship to the labor force participation of women between
20 and 59. However, it has u-shaped effects on women of 60
and above or below 20. Variation in participation across age
groups grows as pre-primary enrollment levels increase. For
women between 20 and 59, the effect of pre-primary enroll-
ment is positive, whereas it is decidedly negative for women
under 20, as well as those of 60 years and older.
We thus confirm the hypothesized negative effect of higher
care burdens on the labor force participation of women
between 30 and 54, but also find positive effects for both
younger and older women that hint at possible substitution
020 40 60 80 100
Predi cted FL PR
-1 0 1 2 3
Per capita GDP
020 40 60 80 100
Predi cted FL PR
-1 0 1 2 3 4
Agriculture
020 40 60 80 100
Predi cted FL PR
-2 0 2 4
manufacturing
020 40 60 80 100
Predi cted FL PR
-3 -2 -1 0 1 2
Services
average
15-19 yrs
20-24 yrs
25-29 yrs
30-34 yrs
35-39 yrs
40-44 yrs
45-49 yrs
50-54 yrs
55-59 yrs
60-64 yrs
65+ yrs
Figure 2. Age effects of economic conditions.
WORKING WOMEN WORLDWIDE. AGE EFFECTS IN FEMALE LABOR FORCE PARTICIPATION IN 117 COUNTRIES 131
Table 4. Effects of families on female labor force participation rates
Model 10 Model 11 Model 12 Model 13
Constant 54.015
***
(1.432) 55.653
***
(1.542) 53.952
***
(1.428) 55.881
***
(1.555)
Age 3.606
***
(0.314) 3.301
***
(0.350) 3.528
***
(0.314) 3.329
***
(0.343)
Age squared 3.120
***
(0.716) 3.738
***
(0.722) 2.826
***
(0.698) 4.105
***
(0.722)
Dependency scale 1.019 (1.606) 3.339 (2.654)
Age * Dependency 0.589
*
(0.340) 1.363
**
(0.551)
Age^2 * Dependency 2.059
***
(0.565) 0.793 (0.853)
Wage replacement maternity leave 0.549 (1.362) 0.886 (1.394)
Age * Maternity pay 0.250 (0.281) 0.342 (0.281)
Age^2 * Maternity pay 0.435 (0.449) 0.210 (0.444)
Length maternity leave 7.350
***
(2.473) 8.528
***
(2.796)
Length maternity leave squared 1.759
**
(0.690) 1.880
***
(0.726)
Age * Length maternity 0.574 (0.522) 1.203
**
(0.564)
Age^2 * Length maternity 4.913
***
(0.820) 4.722
***
(0.890)
Age * Length Maternity^2 0.183 (0.143) 0.278
*
(0.146)
Age^2 * Length maternity^2 1.057
***
(0.228) 0.946
***
(0.231)
Enrollment in pre-primary education 1.219 (1.411) 2.168 (2.157)
Age * Pre-primary 0.214 (0.289) 0.488 (0.436)
Age^2 * Pre-primary 1.859
***
(0.491) 1.723
**
(0.687)
Level 1 variance 24.121
***
(1.115) 24.124
***
(1.115) 24.120
***
(1.115) 24.121
***
(1.115)
Level 2 variance (cons) 231.140
***
(30.505) 214.522
***
(28.321) 230.056
***
(30.364) 212.949
***
(28.116)
var(age) 7.015
***
(1.208) 7.026
***
(1.209) 7.190
***
(1.231) 6.537
***
(1.146)
var(age^2) 25.726
***
(3.652) 21.339
***
(3.080) 25.482
***
(3.620) 19.616
***
(2.856)
Log Likelihood 4380.2988 4366.0933 4380.5874 4357.9712
AIC 8786.598 8770.187 8787.175 8765.942
BIC 8853.679 8868.228 8854.256 8894.944
***
p< 0.01,
**
p< 0.05,
*
p< 0.1.
Source: Global Dataset on Women and Work, sample contains 1,287 age groups in 117 countries, Standard errors in parentheses, controlled for male
labor force participation rate.
Table 3. Effects of education on female labor force participation rates
Model 7 Model 8 Model 9
Constant 54.038
***
(1.391) 53.947
***
(1.383) 54.029
***
(1.357)
Age 3.613
***
(0.313) 3.510
***
(0.312) 3.606
***
(0.313)
Age squared 3.133
***
(0.724) 2.741
***
(0.707) 3.107
***
(0.722)
Girls’ enrollment in primary education 3.691
***
(1.373) 3.206
**
(1.501)
Age * Primary 0.272 (0.281) 0.283 (0.315)
Age^2 * Primary 0.279 (0.501) 0.025 (0.558)
Girls’ enrollment in secondary education 1.058 (1.378) 0.194 (1.824)
Age * Secondary 0.534
*
(0.299) 0.446 (0.396)
Age^2 * Secondary 1.344
***
(0.510) 0.955 (0.684)
Gender parity in primary education 4.802
***
(1.766) 3.392
*
(1.957)
Age * Parity Primary 0.084 (0.366) 0.054 (0.409)
Age^2 * Parity Primary 1.099
*
(0.651) 0.882 (0.728)
Gender parity in secondary education 4.676
***
(1.765) 4.365
**
(1.910)
Age * Parity Secondary 0.382 (0.367) 0.182 (0.399)
Age^2 * Parity Secondary 0.275 (0.650) 0.145 (0.711)
Level 1 variance 24.123
***
(1.115) 24.124
***
(1.115) 24.122
***
(1.115)
Level 2 variance (cons) 217.554
***
(28.731) 215.045
***
(28.401) 206.750
***
(27.319)
var(age) 6.938
***
(1.198) 7.110
***
(1.220) 6.923
***
(1.196)
var(age^2) 27.101
***
(3.831) 27.240
***
(3.849) 26.722
***
(3.782)
Log likelihood 4379.1489 4379.8647 4375.3242
AIC 8790.298 8791.729 8794.648
BIC 8872.859 8874.291 8908.17
***
p< 0.01,
**
p< 0.05,
*
p< 0.1.
Source: Global Dataset on Women and Work, sample contains 1,287 age groups in 117 countries, Standard errors in parentheses, controlled for male
labor force participation rate.
132 WORLD DEVELOPMENT
effects. We also find opposite effects across age groups for the
length of maternity leave. The FLPRs of women between 20
and 59 are lowest in the absence of maternity leave and where
leaves are longest, with higher participation at the values in
between. We observe reversed effects for younger and older
women, whose FLPRs are the mirror image of the prime age
groups.
(d) Gender ideologies
We expected a positive effect of nondiscrimination legisla-
tion and women’s rights on female labor force participation,
as well as a negative effect of religion in general, and Islam
in particular. We do not find significant grand mean effects
of anti-discrimination legislation (Table 5). We do find signif-
icant positive effects of women’s political rights. FLPRs are
lower in countries where women hold few political rights
and higher where they are guaranteed. They are also lower
in countries where the biggest religion is larger and where
Islam is dominant. The effect of political rights is weaker when
controlling for all indicators together (model 17), but it
remains significant and positive.
While the main effect of anti-discrimination legislation is
never significant, the interaction terms with age are. The neg-
ative interaction with age squared (1.439) indicates that the
existence of anti-discrimination legislation is associated with
lower labor force participation of younger and older women.
Figure 4 shows the negative effect on women under 25 and
of 60 and over, as well as the nearly flat lines of the nonsignif-
icant positive effect on women between 25 and 54 years. While
some age effects exist for political rights in model 15, there are
no significant age effects when controlling for the other indica-
tors (model 17). Both religion in general and the dominance of
Islam, however, do interact significantly and positively with
age, indicating the negative main effect is smaller for younger
and older women. This implies that, in line with expectations,
the reduction in labor force participation accounted for by
religion is focused on women of childbearing age and mothers.
We thus confirm the positive effect of political rights on
women of all ages and the negative effect of religion on all
women, but especially on mothers.
(e) Full model
We run a full model, combining the models of economic
conditions, education, families, and gender ideologies to
examine which effects remain significant predictors of the
FLPR (Table 6). Of the economic variables, the negative effect
of manufacturing is now significant. There are also significant
age effects of GDP, indicating the u-shaped relationship with
the labor force participation of younger and older women
remains relevant. Gender parity in primary education contin-
ues to have a positive effect on the FLPR, whereas parity in
secondary education has a reduced negative effect. The educa-
tional variables now include significant age effects, which are
most notable for gender parity in education. Both parity indi-
cators now show that while younger and older women are
among the most likely to work in countries with low scores
on gender parity, they are least likely to work in countries with
high parity.
While the effect of the dependency scale is nonsignificant,
care arrangements retain their effects. The effect of the length
of maternity leave is fully replicated in the total model. Enroll-
ment in pre-primary education now has a negative effect and a
higher wage replacement during maternity leave has a negative
effect on younger and older women. Anti-discrimination legis-
lation continues to be associated with a lower participation of
young and older women, whereas political rights have a posi-
tive effect across the board. The results of religion and Islam
are replicated as they were in the gender ideologies model.
Thus, also controlling for economic development, families
and education, gender ideologies continue to affect FLPRs.
(f) Robustness checks
As indicated in Section 3(c), to ascertain the validity of the
reported results for all levels of development, we repeat all
analyses on a split sample of low and lower middle, upper
middle and high-income countries (findings are not reported
020 40 60 80
Predi cted FL PR
-2 -1 0 1 2
Dependency scale
020 40 60 80
Predi cted FL PR
-2 0 2 4 6
Maternity leave
020 40 60 80
Predi cted FL PR
-2 -1 0 1 2
Pre-primary edu
average
15-19 yrs
20-24 yrs
25-29 yrs
30-34 yrs
35-39 yrs
40-44 yrs
45-49 yrs
50-54 yrs
55-59 yrs
60-64 yrs
65+ yrs
Figure 3. Age effects of families.
WORKING WOMEN WORLDWIDE. AGE EFFECTS IN FEMALE LABOR FORCE PARTICIPATION IN 117 COUNTRIES 133
here but are available on request). The findings show, that
analyses per income level replicate the u-shaped relationship
with GDP, indicating that GDP starts having a positive effect
already in lower middle-income countries. The effects of sec-
tor sizes hold when performing the analysis on the income
groups separately. The results indicate that agriculture is
the most important driver of FLPRs in developing countries,
while service sectors pull women of childbearing age and with
small children into the labor force in industrialized countries.
In upper middle-income countries, the positive effect of agri-
culture exists exclusively for young and older women, while
manufacturing has a negative effect on those same age
groups. We also find that in high-income countries the effects
of education are the opposite of the findings reported in
Section 4(d), which is likely the effect of a lack of variation
on these variables.
020 40 60 80
Predi cted FL PR
-4 -3 -2 -1 0 1
Anti-discrimination
020 40 60 80
Predi cted FL PR
-3 -2 -1 0 1 2
Political rights
020 40 60 80
Predi cted FL PR
-3 -2 -1 0 1
Religion
020 40 60 80
Predi cted FL PR
-1 0 1 2
Islam
average
15-19 yrs
20-24 yrs
25-29 yrs
30-34 yrs
35-39 yrs
40-44 yrs
45-49 yrs
50-54 yrs
55-59 yrs
60-64 yrs
65+ yrs
Figure 4. Age effects of gender ideologies.
Table 5. Effects of gender ideologies on female labor force participation rates
Model 14 Model 15 Model 16 Model 17
Constant 53.923
***
(1.432) 53.937
***
(1.315) 53.926
***
(1.154) 53.960
***
(1.106)
Age 3.489
***
(0.314) 3.478
***
(0.312) 3.466
***
(0.306) 3.504
***
(0.307)
Age squared 2.658
***
(0.699) 2.606
***
(0.699) 2.541
***
(0.624) 2.729
***
(0.613)
Nondiscrimination law 0.445 (1.414) 0.182 (1.103)
Age * nondiscrimination 0.012 (0.287) 0.111 (0.285)
Age^2 * nondiscrimination 1.621
***
(0.495) 1.439
***
(0.386)
Women’s political rights 6.126
***
(1.295) 3.770
***
(1.144)
Age * Political rights 0.154 (0.283) 0.122 (0.293)
Age^2 * Political rights 1.397
***
(0.499) 0.303 (0.399)
Size biggest religion 4.146
***
(1.133) 4.065
***
(1.097)
Age * Religion 0.656
**
(0.277) 0.668
**
(0.281)
Age^2 * Religion 0.872
**
(0.401) 1.085
***
(0.383)
Dominance of Islam 7.916
***
(1.133) 6.685
***
(1.151)
Age * Islam 0.040 (0.277) 0.010 (0.295)
Age^2 * Islam 3.342
***
(0.401) 3.079
***
(0.402)
Level 1 variance 24.127
***
(1.115) 24.134
***
(1.116) 24.139
***
(1.116) 24.126
***
(1.115)
Level 2 variance (cons) 231.129
***
(30.504) 193.735
***
(25.615) 147.184
***
(19.527) 134.672
***
(17.893)
var(age) 7.213
***
(1.234) 7.180
***
(1.229) 6.760
***
(1.175) 6.760
***
(1.175)
var(age^2) 26.263
***
(3.722) 26.904
***
(3.804) 16.544
***
(2.452) 14.489
***
(2.183)
Log likelihood 4382.7754 4373.7813 4329.5908 4317.4199
AIC 8791.551 8773.563 8691.182 8678.84
BIC 8858.632 8840.643 8773.743 8792.361
***
p< 0.01,
**
p< 0.05,
*
p< 0.1.
Source: Global Dataset on Women and Work, sample contains 1,287 age groups in 117 countries, Standard errors in parentheses, controlled for male
labor force participation rate.
134 WORLD DEVELOPMENT
The effect of care burdens is not the same across income
groups. In low and lower middle-income countries, higher care
burdens are associated with higher FLPRs, whereas the nega-
tive effect in the whole sample is replicated in upper middle
and high-income countries. The effects of gender ideologies
are reproduced in all income groups, except for a nonsignifi-
cant effect of the variable for religious dominance in
high-income countries (the effect of Islam is still significant),
which confirms the fading impact of religion on individual
actions in those countries. The effect of political rights is con-
siderably larger in low and lower middle-income countries
than in upper middle and high-income countries.
5. CONCLUSION AND DISCUSSION
In this article, we have studied the country-level effects of
economic conditions, education, families, and gender ideolo-
gies on FLPRs in 117 countries. We paid particular attention
to the inclusion of developing countries in the sample and the
distinctive effects of the predictors on women of different ages.
Our study showed that the FLPR in all 117 countries firstly
increases as women mature and then start decreasing until
retirement. These patterns differ between countries as to their
exact timing, but they occur universally. We see that in most
instances, women below 20 and above 60 cluster together
and behave markedly different from the rest of the population.
Women between 20 and 24 and between 55 and 59 often form
a second group that alternately behaves like younger and older
women or like the prime age group. Some effects we found
applied exclusively to one cluster of age groups or indeed
had opposite effects on various age groups. Such findings,
we believe, signal the value of disaggregating the female pop-
ulation into age groups in macro-level studies.
While we do not confirm a significant u-shaped relationship
between the level of economic development and the FLPR for
all women, we find that a u-shaped relationship does exist for
young and older women, even when we control for educa-
tional expansion in the final model. Thus, our hypothesis that
economic prosperity would drive youth and older women out
of the labor market can only be confirmed for the lower levels
of GDP, with the opposite effect taking place at higher
levels of development. The effects of sector sizes too, which
are confirmed as hypothesized, appear to be driven by the
Table 6. Effects of economic conditions, families, education, and gender ideologies on female labor force participation rates
Constant 53.772
***
(1.518) Age^2
*
Length maternity^2 0.705
***
(0.189)
Age 4.065
***
(0.453) Enrollment in pre-primary education 2.626
*
(1.524)
Age squared 3.234
***
(0.780) Age
*
Pre-primary 0.804
*
(0.434)
Per capita GDP 0.791 (3.019) Age^2
*
Pre-primary 0.110 (0.568)
Per capita GDP squared 1.270 (1.122) Girls’ enrollment in primary education 0.074 (1.131)
Age * GDP 2.981
***
(0.857) Age
*
Primary 0.371 (0.321)
Age * GDP^2 0.697
**
(0.319) Age^2
*
Primary 1.150
***
(0.422)
Age^2 * GDP 0.993 (1.125) Girls’ enrollment in secondary education 3.694 (2.406)
Age^2 * GDP^2 0.580 (0.418) Age
*
Secondary 0.255 (0.684)
Agriculture 1.818 (1.807) Age^2
*
Secondary 0.949 (0.896)
Age * Agriculture 0.312 (0.513) Gender parity in primary education 3.442
**
(1.377)
Age^2 * Agriculture 0.420 (0.675) Age
*
Parity Primary 0.037 (0.391)
Manufacturing 2.166
**
(1.072) Age^2
*
Parity Primary 0.861
*
(0.513)
Age * Manufacturing 0.097 (0.304) Gender parity in secondary education 2.234 (1.523)
Age^2 * Manufacturing 0.066 (0.399) Age
*
Parity Secondary 0.757
*
(0.433)
Services 1.183 (1.371) Age^2
*
Parity Secondary 0.377 (0.568)
Age * Services 0.444 (0.390) Anti-discrimination legislation 0.651 (1.151)
Age^2 * Services 0.218 (0.511) Age
*
Anti-discrimination 0.312 (0.327)
Dependency scale 2.185 (3.259) Age^2
*
Anti-discrimination 0.957
**
(0.429)
Age * Dependency 0.372 (0.930) Women’s political rights 3.297
***
(1.016)
Age^2 * Dependency 0.329 (1.219) Age
*
Political rights 0.215 (0.289)
Wage replacement maternity leave 0.117 (1.054) Age^2
*
Political rights 0.372 (0.379)
Age * Maternity pay 0.622
**
(0.299) Size biggest religion 3.565
***
(0.967)
Age^2 * Maternity pay 0.348 (0.393) Age
*
Religion 0.695
**
(0.275)
Length maternity leave 5.509
***
(2.097) Age^2
*
Religion 0.749
**
(0.360)
Length maternity leave squared 1.015
**
(0.506) Dominance of Islam 9.026
***
(1.281)
Age * Length maternity 0.832 (0.596) Age
*
Islam 0.211 (0.366)
Age^2 * Length maternity 3.427
***
(0.782) Age^2
*
Islam 3.336
***
(0.478)
Age * Length Maternity^2 0.220 (0.144)
Level 1 variance 24.120
***
(1.115)
Level 2 variance (cons) 85.404
***
(11.453)
var(age) 4.874
***
(0.929)
var(age^2) 9.973
***
(1.594)
Log likelihood 4258.8716
AIC 8645.743
BIC 8975.988
***
p< 0.01,
**
p< 0.05,
*
p< 0.1.
Source: Global Dataset on Women and Work, sample contains 1,287 age groups in 117 countries, Standard errors in parentheses, controlled for male
labor force participation rate.
WORKING WOMEN WORLDWIDE. AGE EFFECTS IN FEMALE LABOR FORCE PARTICIPATION IN 117 COUNTRIES 135
youngest and oldest age groups. We conclude that economic
conditions do affect the timing of initial labor market entry
and eventual exit, but do little to explain the level of participa-
tion of women between 25 and 55 years of age.
We find a positive association between girls’ enrollment and
gender parity in primary education, but cannot confirm the
hypothesized positive effects for secondary education. Based
on human capital theory, we had expected effects of higher
levels of education to be stronger than those for lower levels.
Yet, contrary to micro-level studies we find quite the opposite.
It may be that mechanisms that distinguish between women on
the micro level, do not directly translate to the macro level—
that is to say, education may serve as a stratifying factor for
women inside countries rather than between them. It may also
be that our indicator of total enrollment and parity fails to
capture the extent to which educational institutions are hori-
zontally segregated along gender lines. Future research based
on micro data from low- and middle-income countries or at
least disaggregated by educational field, may be able to answer
this question.
As regards the labor force participation rates of mothers, we
suggest the effects of families and gender ideologies are most
promising. In particular, we find evidence of substitution
effects in maternity and care arrangements. While we observe
women of childbearing age being more likely to remain in the
labor force as the length of paid maternity leave increases and
to start to drop out again as the leave period is extended more,
we see an opposite movement for women under 25 and above
55. Similarly, while women between 25 and 59 have higher
FLPRs at higher levels of enrollment in pre-primary educa-
tion, women above and below that age are less likely to work.
While we find few significant effects of anti-discrimination
legislation and positive effects of political rights for women
of all ages, the effects of religion are largest for women of
prime age. We find that a greater religious heritage in a coun-
try is associated with lower FLPRs of women between 20 and
59 years of age, but much less so with the participation of
women who are younger or older than that. These results sug-
gest that the gender stereotypes encapsulated in religious prac-
tices affect female labor force participation primarily through
strong beliefs regarding the role of mothers.
As Barrientos and Kabeer (2004) have pointed out, care
tasks are gendered in places around the world. While policies
on care may differ greatly between low-income and
high-income countries, the issues are universal. We believe
that researchers who are particularly interested in the labor
force participation of women of prime age, may do well to
focus on families and gender ideologies. What is more, our
research shows that for this age group predictors regarding
families and gender ideologies have greater explanatory power
than economic conditions or education do. Policy makers call-
ing for job creation in female-dominated sectors to help
women increase family incomes (e.g., World Bank, 2011),
therefore, should consider they will primarily be drawing
young and older women into the labor force, unless their
efforts are accompanied by investments in care arrangements
and reducing the stigma on work.
NOTES
1. We interpret informal labor is more than a legalistic distinction
between declared and undeclared work. It makes up a substantial part of
labor markets in developing countries, including many own-account
workers, subsistence or family workers, homeworkers and industrial
outworkers, and casual wage workers. Informal work exists in parts of the
economy that have not yet been integrated in the global economy, in
family-owned businesses, in trade and commerce to provide cheaper
products for working-class families (street vendors), in registered busi-
nesses seeking to avoid taxes, in subcontract production of global value
chains, and so on (c.f. Chen, 2001).
2. FOA also stresses that disaggregated statistics of participation in
agricultural activities are only very sparsely available and require further
research.
3. Countries without actual observations are Afghanistan, Angola,
Central African Republic, Channel Islands, Comoros, Equatorial Guinea,
Eritrea, Gambia, Guam, Guinea, Guinea-Bissau, Democratic People’s
Republic of Korea, Libyan Arab Jamahiriya, Mauritania, Myanmar,
Senegal, Solomon Islands, Somalia, Swaziland, Turkmenistan, United
States Virgin Islands, Uzbekistan, and Western Sahara.
4. Countries with less than one million inhabitants over 15 years old are:
Bahamas, Bahrain, Barbados, Belize, Bhutan, Brunei Darussalam, Cape
Verde, Cyprus, Djibouti, East Timor, Fiji, French Guiana, French
Polynesia, Gabon, Guyana, Iceland, Luxembourg, Macau, Maldives,
Malta, Martinique, Netherlands Antilles, New Caledonia, Qatar, Re
´-
union, Saint Lucia, Saint Vincent and the Grenadines, Samoa, Sao Tome
and Principe, Suriname, Tonga, and Vanuatu.
5. Data on per capita GDP for Iran were taken from 2009.
6. The process of standardization does change the distribution of the
variable and as such affects the second order polynomial, however, when
the regression output is converted to predicted values this change is
undone, so that results are unbiased.
7. In cases where data for sector sizes in 2010 was not available, data
were taken from adjacent years. Other years that were included were: 2006
(New Zealand, Nigeria), 2007 (Cameroon, Iran), 2008 (Canada, Chad,
Greece), 2009 (Algeria, Belgium, France, Ireland, Lithuania, Madagascar,
Mali, Spain).
8. Primary and secondary enrollment and gender parity for different
years than 2010 are included. For 2005 (Brazil, Malaysia); for 2006
(United Arab Emirates); for 2007 (Iraq, Namibia, Togo); for 2008
(Bolivia, Botswana, Georgia, Kuwait, Trinidad and Tobago); for 2009
(Canada, Ghana, Madagascar, Oman, Philippines, Russia, South Africa,
Sudan (pre-cession), Thailand); for 2011 (Benin, Ivory Coast, Liberia)
Data from Albania, Ivory Coast, Mauritius were taken from the World
Economic Forum Global Gender Gap report 2011.
9. Marriage data for different years than 2010 are included. For 2005
(Congo, Georgia, Honduras, Republic of Korea, Kuwait, Lao, Moldova,
United Arab Emirates); for 2006 (Australia, Benin, Canada, Haiti, Hong
Kong, Lesotho, Mali, Namibia, New Zealand, Niger, Papua New
Guinea); for 2007 (Democratic Republic of the Congo, Dominican
Republic, El Salvador, Iraq, Lebanon, Mozambique, Nicaragua, Pak-
istan, Peru, Philippines, Saudi Arabia, Sri Lanka, Ukraine, West Bank &
Gaza, Zambia); for 2008 (Bolivia, Egypt, Ghana, Kenya, Liberia,
Madagascar, Nigeria, Sierra Leone, Sudan, Turkey); for 2009 (Azerbaijan,
Belarus, Belgium, France, Israel, Jordan, Kazakhstan, Kyrgyzstan,
Switzerland, United Kingdom, USA, Vietnam); for 2011 (Albania,
Austria, Bangladesh, Bulgaria, Cameroon, Chile, Costa Rica, Czech
136 WORLD DEVELOPMENT
Republic, Denmark, Ethiopia, Germany, Iran, Ireland, Latvia, Lithuania,
Nepal, Netherlands, Romania, Slovenia, South Africa, Uganda and
Uruguay).
10. Pre-primary enrollment for different years than 2010 are included.
For 2005 (Brazil, Malaysia, Pakistan); for 2006 (Namibia, United Arab
Emirates, Nepal); for 2007 (Hong Kong, Iraq, Trinidad and Tobago); for
2008 (Bolivia, Botswana, Georgia, Kuwait, Liberia); for 2009 (Canada,
Ghana, Madagascar, Oman, Philippines, Russia, South Africa, Sudan
(pre-secession), Thailand, Ivory Coast); for 2011 (Benin, Kazakhstan,
Liberia). Data from Mozambique are retrieved from the ministry of
education. Data from Malawi are for 2009 and retrieved from the Ministry
for Gender, Children and Community Development. Data from Tunisia
are for 2009 from the Observatory for Information, Training, Documen-
tation and Studies on the Rights of the Child.
REFERENCES
Abramo, L., & Valenzuela, M. E. (2005). Women’s labour force
participation rates in Latin America. International Labour Review,
144(4), 369–399.
Adely, F. (2009). Educating women for development. The Arab human
development report 2005 and the problem with women’s choices.
International Journal for Middle East Studies, 41, 105–122.
Agu
¨ero, J., & Marks, M. (2010). Motherhood and female labor supply in
the developing world. Evidence from infertility shocks. The Journal of
Human Resources, 46(4), 800–826.
Ahn, N., & Mira, P. (2002). A note on the changing relationship between
fertility and female employment rates in developed countries. Journal
of Population Economics, 15(4), 667–682.
Aisenbrey, S., Evertsson, M., & Grunow, D. (2009). Is there a career
penalty for mothers’ time out? A comparison of Germany, Sweden and
the United States. Social Forces, 88(2), 573–605.
Amin, S., & Alam, I. (2008). Women’s employment decisions in Malaysia:
Does religion matter?. The Journal of Socio-Economics, 37(6),
2368–2379.
Anderson, S., & Eswaran, M. (2009). What determines female autonomy?
Evidence from Bangladesh. Journal of Development Economics, 90,
179–191.
Angrist, J., & Evans, W. (1998). Children and the Parents’ Labor Supply:
Evidence from Exogenous Variation in Family Size. American
Economic Review, 88(3), 450–477.
Apps, P., & Rees, R. (2001). Fertility, female labor supply and public
policy. IZA Discussion Papers, 409.
Arber, S., & Timonen, V. (Eds.) (2012). Contemporary grandparenting:
Changing family relationships in global contexts. Bristol, UK: Policy
Press.
Aromolaran, A. (2004). Female schooling, non-market productivity, and
labor market participation in Nigeria. Economic Growth Center
Discussion Paper Series, 879.
Bardasi, E. & Gornick, J. (2000). Women and Part-Time Employment:
Workers’ ‘Choices’ and Wage Penalties in Five Industrialized Coun-
tries. Luxembourg Income Study Working Paper, No. 223.
Barrientos, S., & Kabeer, N. (2004). Enhancing female employment in
global production: Policy implications. Global Social Policy, 4(2),
153–169.
Benerı
´a, L. (1992). Accounting for women’s work: The progress of two
decades. World Development, 20(11), 1547–1560.
Bhalla, S., & Kaur, R. (2011). Labour force participation of women in
India: Some facts, some queries. Asia Research Centre Working Paper
40.http://eprints.lse.ac.uk/38367/1/ARCWP40-BhallaKaur.pdf.
Bloom, D., Canning, D., Fink, G., & Finlay, J. (2007). Fertility, female
labor force participation, and the demographic dividend. NBER
Working Paper Series, 13583.
Boserup, E. (1970). Women’s role in economic development. New York: St.
Martin’s Press.
Browning, M. (1992). Children and household economic behavior. Journal
of Economic Literature, 30(3), 1434–1475.
Buvinic, M., & Gupta, G. R. (1997). Female-headed households and
female-maintained families: Are they worth targeting to reduce poverty
in developing countries?. Economic Development and Cultural Change,
45(2), 259–280.
C¸ agatay, N., & O
¨zler, S. (1995). Feminization of the labor force: The
effects of long-term development and structural adjustment. World
Development, 23(11), 1883–1894.
Chadwick, B., & Garrett, H. D. (1995). Women’s religiosity and
employment: The LDS experience. Review of Religious Research,
36(3), 277–293.
Chafetz, J. (1990). Gender equity: An integrated theory of stability and
change. Newbury Park, CA: Sage.
Chang, M. (2000). The evolution of sex segregation regimes. American
Journal of Sociology, 105(6), 1658–1701.
Chang, M. (2004). Growing pains: Cross-national variation in sex
segregation in sixteen developing countries. American Sociological
Review, 69(1), 114–137.
Chen, M. (2001). Women and informality: A global picture, the global
movement. SAIS Review, 21(1), 71–82.
Clark, R. L., & Anker, R. (1993). Cross-national analysis of labor force
participation of older men and women. Economic Development and
Cultural Change, 41(3), 489–512.
Clark, R., Ramsbey, T., & Stier Adler, E. (1991). Culture, gender, and
labor force participation: A cross-national study. Gender & Society,
5(1), 47–66.
Commission on Growth and Development (CGD) (2008). The growth
report: Strategies for sustained growth and inclusive development. CGD/
World Bank: Washington, D.C..
Cruces, G., & Galiani, S. (2007). Fertility and female labor supply in Latin
America: New causal evidence. Labour Economics, 14(3), 565–573.
De Leeuw, J., & Meijer, E. (Eds.) (2008). Handbook of multilevel analysis.
New York: Springer.
Elson, D. (1999). Labor markets as gendered institutions: Equality,
efficiency and empowerment issues. World Development, 27(3),
611–627.
Engelhardt, H., & Prskawetz, A. (2004). On the changing correlation
between fertility and female employment over space and time.
European Journal of Population/Revue Europeenne De Demographie,
20(1), 35–62.
England, P., Gornick, J., & Shafer, E. F. (2012). Women’s employment,
education, and the gender gap in 17 countries. Monthly Labor Review,
135(4), 3–12.
Food and Agriculture Organization (2011). The state of food and
agriculture: Women in agriculture: Closing the gender gap for develop-
ment. Rome: FOA.
Ganguli, I., Hausmann, R., & Viarengo, M. (2013). Closing the gender
gap in education: What is the state of gaps in labor force participation
for women, wives and mothers?. International Labour Review.
Goldin, C. (1995). The U-shaped female labor force function in economic
development and economic history. In T. P. Schultz (Ed.), Investment
in women’s human capital and economic development (pp. 61–90).
Chicago: University of Chicago Press.
Goldin, C. (2006). The quiet revolution that transformed women’s
employment, education and family. American Economic Review,
96(2), 1–21.
Gornick, Janet C., Marcia K. Meyers., & Katherin E. Ross (1997).
Supporting the Employment of Mothers: Policy Variation Across
Fourteen Welfare States. Journal of European Social Policy, 7(1),
45–70.
Gray, M., Kittilson, C., & Sandholtz, W. (2006). Women and Globaliza-
tion: A Study of 180 Countries, 1975–2000. International Organization,
60, 293–333.
Haas, B., Steiber, N., Hartel, M., & Wallace, C. (2006). Household
employment patterns in an enlarged european union. Work, Employ-
ment & Society, 20(4), 751–771.
WORKING WOMEN WORLDWIDE. AGE EFFECTS IN FEMALE LABOR FORCE PARTICIPATION IN 117 COUNTRIES 137
Haghighat, E. (2002). Culture, development and female labor force
participation: Disaggregating different sectors. International Review of
Sociology, 12(3), 343–362.
Haghighat, E. (2005). Neopatriarchy, Islam and female labour force
participation: A reconsideration. International Journal of Sociology and
Social Policy, 25(10), 84–105.
Hakim, C. (2000). Work-lifestyle choices in the 21st century: Preference
theory. Oxford: Oxford University Press.
Heineck, G. (2004). Does religion influence the labor supply of married
women in Germany?. The Journal of Socio-Economics, 33(3), 307–328.
Heinz, W. R., & Kru
¨ger, H. (2001). Life course: Innovations and
challenges for social research. Current Sociology, 49(2), 29–45.
Horton, S. (1999). Marginalization revisited: Women’s market work and
pay, and economic development. World Development, 27(3), 571–582.
Hox, J. (2010). Multilevel analysis: Techniques and applications (2nd ed.).
New York: Routledge.
Hummelsheim, D., & Hirschle, J. (2010). Mother’s employment: Cultural
imprint or institutional governance?. European Societies, 12(3),
339–366.
Thirteenth International Conference of Labour Statisticians (1982).
Resolution concerning statistics of the economically active population,
employment, unemployment and underemployment. In: ILO (2000).
Current international recommendations on labour statistics, 2000 edition.
Geneva: ILO.
International Labor Organization (2011). ILO estimates and projections
of the economically active population: 1990–2020. In Methodological
description (6th ed.). Geneva: ILO, Available: http://laborsta.ilo.org/
applv8/data/EAPEP/v6/ILO_EAPEP_methodology_2011.pdf.
Iversen, T., & Rosenbluth, F. (2008). Work and power: The connection
between female labor force participation and female political repre-
sentation. Annual Review of Political Science, 11, 479–495.
Korpi, W. (2000). Faces of inequality: Gender, class, and patterns of
inequalities in different types of welfare states. Social Politics:
International Studies in Gender, State & Society, 7(2), 127–191.
Kucera, D., & Tejani, S. (2014). Feminization, defeminization, and
structural change in manufacturing. World Development, 64, 569–582.
Lehrer, E. (2004). Religion as a determinant of economic and demo-
graphic behavior in the United States. Population and Development
Review, 30(4), 707–726.
Lincove, J. A. (2008). Growth, girls’ education, and female labor: A
longitudinal analysis. The Journal of Developing Areas, 41(2), 45–68.
Mandel, H., & Semyonov, M. (2006). A welfare state paradox: State
interventions and Women’s employment opportunities in 22 countries.
American Journal of Sociology, 111(6), 1910–1949.
Mehra, R., & Gammage, S. (1999). Trends, countertrends, and gaps in
women’s employment. World Development, 27(3), 533–550.
Mishra, V., Nielsen, I., & Smyth, R. (2010). On the relationship between
female labour force participation and fertility in G7 countries:
Evidence from panel cointegration and granger causality. Empirical
Economics, 38(2), 361–372.
Nee, V. (1998). Sources of the new institutionalism. In M. Brinton, & V.
Nee (Eds.), The new institutionalism in sociology (pp. 1–16). New York:
Russell Sage Foundation.
Norris, M. (1992). The impact of development on women. A specific-
factor analysis. Journal of Development Economics, 38, 183–201.
Orbeta, A. (2005). Children and the labor force participation and earnings
of parents in the Philippines. Philippine Journal of Development, 32(1),
19–52.
Pampel, F. C., & Tanaka, K. (1986). Economic development and female
labor force participation: A reconsideration. Social Forces, 64(3),
599–619.
Pascall, G., & Lewis, J. (2004). Emerging gender regimes and policies for
gender equality in a Wider Europe. Journal of Social Policy, 33(03),
373–394.
Psacharopoulos, G., & Tzannatos, Z. (1989). Female labor force partic-
ipation: An international perspective. The World Bank Research
Observer, 4(2), 187–201.
Rendall, M. (2013). Structural change in developing countries: Has it
decreased gender inequality?. World Development, 45, 1–16.
Safa, H. (1977). The changing class composition of the female labor force
in Latin America. Latin American Perspectives, 4(4), 126–136.
Sassen, S. (1996). Towards a feminist analytics of the global economy.
Indiana Journal of Global Legal Studies, 4(1), 7–42.
Schultz, T. P. (1990). Women’s changing participation in the labor force:
A world perspective. Economic Development and Cultural Change,
38(3), 457–488.
Seguino, S. (2011). Help or hindrance? Religion’s impact on gender
inequality in attitudes and outcomes. World Development, 39(8),
1308–1321.
Semyonov, M. (1980). The social context of women’s labor force
participation: A comparative analysis. American Journal of Sociology,
86(3), 534–550.
Spierings, N., Smits, J., & Verloo, M. (2010). Micro- and macrolevel
determinants of women’s employment in six Arab countries. Journal of
Marriage and Family, 72(5), 1391–1407.
Steiber, B., & Haas, N. (2012). State of the art. Advances in explaining
women’s employment patterns. Socio-Economic Review, 10(2),
343–367.
Stockemer, D., & Byrne, M. (2012). Women’s representation around the
world: The importance of women’s participation in the workforce.
Parliamentary Affairs, 65, 802–821.
Stryker, R., Eliason, S. R., Tranby, E., & Hamilton, W. (2011). Family
policies, education and female labor market participation in advanced
capitalist democracies. In G. B. Cohen, B. W. Ansell, R. H. Cox, & J.
Gingrich (Eds.). Social policy in the smaller European Union states
(Vol. 9). Berghahn Books.
Tam, H. (2011). U-shaped female labor participation with economic
development: Some panel data evidence. Economics Letters, 110,
140–142.
Tanced, P. (1995). Women’s work: A challenge to the sociology of work.
Gender, Work and Organization, 2(1), 11–20.
Tansel, A. (2001). Economic development and female labor force
participation in Turkey: Time-series evidence and cross-province
estimates. Economic research forum, Working Paper no.200124.
UNDP. (1995). Human development report 1995.Gender and human
development. Available: http://hdr.undp.org/sites/default/files/reports/
256/hdr_1995_en_complete_nostats.pdf.
UNDP. (2013). Human development report 2013.The rise of the south:
Human progress in a diverse world. Available: http://hdr.undp.org/
sites/default/files/reports/14/hdr2013_en_complete.pdf.
Van Klaveren, M., & Tijdens, K. (2012). Empowering women in work in
developing countries. Basingstoke: Palgrave Macmillan.
Whitehouse, G. (1992). Legislation and labour market gender inequality:
An analysis of OECD countries. Work, Employment & Society, 6(1),
65–86.
World Bank. (2011) World development report 2012: Gender equality and
development. Available: http://siteresources.worldbank.org/
INTWDR2012/Resources/7778105-1299699968583/7786210-
1315936222006/Complete-Report.pdf.
World Bank. (2012) World development report 2013: Jobs. Available:
http://siteresources.worldbank.org/INTWDR2012/Resources/
7778105-1299699968583/7786210-1315936222006/Complete-Report.
pdf.
World Economic Forum (2014). The global gender gap report 2014. World
Economic Forum. Available: http://www3.weforum.org/docs/
GGGR14/GGGR_CompleteReport_2014.pdf.
138 WORLD DEVELOPMENT
APPENDIX I. LIST OF COUNTRIES
Table 7. Countries in the analysis by world region and income level
Europe and Central Asia
Albania Lower Middle-Income Country
Armenia Lower Middle-Income Country
Austria High-Income Country
Azerbaijan Higher Middle-Income Country
Belarus Higher Middle-Income Country
Belgium High-Income Country
Bosnia and Herzegovina Higher Middle-Income Country
Bulgaria Higher Middle-Income Country
Croatia High-Income Country
Czech Republic High-Income Country
Denmark High-Income Country
Estonia High-Income Country
Finland High-Income Country
France High-Income Country
Georgia Lower Middle-Income Country
Germany High-Income Country
Greece High-Income Country
Hungary High-Income Country
Ireland High-Income Country
Italy High-Income Country
Kazakhstan Higher Middle-Income Country
Kyrgyzstan Low-Income Country
Latvia Higher Middle-Income Country
Lithuania Higher Middle-Income Country
Netherlands High-Income Country
Norway High-Income Country
Poland High-Income Country
Portugal High-Income Country
Republic of Moldova Lower Middle-Income Country
Romania Higher Middle-Income Country
Russian Federation Higher Middle-Income Country
Slovakia High-Income Country
Slovenia High-Income Country
Spain High-Income Country
Sweden High-Income Country
Switzerland High-Income Country
Tajikistan Low Income Country
FYRO Macedonia Higher Middle-Income Country
Turkey Higher Middle-Income Country
Ukraine Lower Middle-Income Country
United Kingdom High-Income Country
North America
Canada High-Income Country
United States High-Income Country
Latin America and Caribbean
Argentina Higher Middle-Income Country
Bolivia Lower Middle-Income Country
Brazil Higher Middle-Income Country
Chile Higher Middle-Income Country
Colombia Higher Middle-Income Country
Costa Rica Higher Middle-Income Country
Cuba Higher Middle-Income Country
Dominican Republic Higher Middle-Income Country
Ecuador Higher Middle-Income Country
El Salvador Lower Middle-Income Country
Guatemala Lower Middle-Income Country
Honduras Lower Middle-Income Country
Jamaica Higher Middle-Income Country
Mexico Higher Middle-Income Country
Nicaragua Lower Middle-Income Country
Panama Higher Middle-Income Country
Paraguay Lower Middle-Income Country
Peru Higher Middle-Income Country
Trinidad and Tobago High-Income Country
Uruguay Higher Middle-Income Country
Venezuela Higher Middle-Income Country
Algeria Higher Middle-Income Country
Egypt Lower Middle-Income Country
Iran, Islamic Republic of Higher Middle-Income Country
Jordan Higher Middle-Income Country
Lebanon Higher Middle-Income Country
Morocco Lower Middle-Income Country
Saudi Arabia High-Income Country
Tunisia Higher Middle-Income Country
United Arab Emirates High-Income Country
Yemen Lower Middle-Income Country
South East Asia
Australia High-Income Country
Bangladesh Low-Income Country
Cambodia Low-Income Country
China Higher Middle-Income Country
India Lower Middle-Income Country
Indonesia Lower Middle-Income Country
Japan High-Income Country
Korea, Republic of High-Income Country
Lao People’s Democratic Republic Lower Middle-Income Country
Malaysia Higher Middle-Income Country
Mongolia Lower Middle-Income Country
Nepal Low-Income Country
New Zealand High-Income Country
Pakistan Lower Middle-Income Country
Philippines Lower Middle-Income Country
Sri Lanka Lower Middle-Income Country
Thailand Higher Middle-Income Country
Viet Nam Lower Middle-Income Country
Sub-Saharan Africa
Benin Low-Income Country
Botswana Higher Middle-Income Country
Burkina Faso Low-Income Country
Burundi Low-Income Country
Cameroon Lower Middle-Income Country
Chad Low-Income Country
Co
ˆte d’Ivoire Lower Middle-Income Country
Ethiopia Low-Income Country
Ghana Lower Middle-Income Country
Kenya Low-Income Country
Lesotho Lower Middle-Income Country
Liberia Low-Income Country
Madagascar Low-Income Country
Malawi Low-Income Country
Mali Low-Income Country
Mauritius Higher Middle-Income Country
Mozambique Low-Income Country
Namibia Higher Middle-Income Country
Nigeria Lower Middle-Income Country
Rwanda Low-Income Country
South Africa Higher Middle-Income Country
Sudan Lower Middle-Income Country
Tanzania, United Republic of Low-Income Country
Togo Low-Income Country
Uganda Low-Income Country
WORKING WOMEN WORLDWIDE. AGE EFFECTS IN FEMALE LABOR FORCE PARTICIPATION IN 117 COUNTRIES 139
Table 8. Global Dataset on Women and Work, description of dataset.
Variable name Variable label Measurement Source
Country Country For a list of countries, see Appendix I EAPEP 6th edition
Age_group Age group Ten five year age intervals (15–19, ...,
60–64) and one age group for 65
years and above
EAPEP 6th edition
INCOMEGRP Income group Countries classified according to four
income groups
World Bank
CULT_REGION World region Countries classified according to six
geographical regions
World Bank
MPR Male labor force participation rate Economically active men as a share
of all men (Ø75.9, r24.5)
EAPEP 6th edition
FPR Female labor force participation rate Economically active women as a
share of all women (Ø55.1, r26.2)
EAPEP 6th edition
MFPR Total labor force participation rate Economically active population as a
share of all women and men (Ø65.5,
r23.9)
EAPEP 6th edition
LPRGAP Gender participation gap Male activity rate minus female
activity rate (Ø20.8, r17.7)
EAPEP 6th edition, own calculation
Economic structure
ECON_GDPPC Per capita GDP Per capita GDP in constant 2000 US
Dollars (Ø7300, r10,060)
World Bank Development Indicators
ECON_AGRI Size agricultural sector Agriculture, value added as a share of
GDP (Ø12.5, r12)
World Bank Development Indicators
ECON_MANU Size manufacturing sector Manufacturing, value added as a
share of GDP (Ø14.5, r6.2)
World Bank Development Indicators
ECON_INDU Size industrial sector Industry, value added as a share of
GDP (Ø30.1, r10.1)
World Bank Development Indicators
ECON_SERV Size services sector Services, value added as a share of
GDP (Ø57.6, r12.7)
World Bank Development Indicators
Families
POP_FERTILITY Fertility rate Mean number of births per woman
(Ø2.6, r1.3)
World Bank Development Indicators
POP_OLD Elderly as a share of the population Share of the population aged 65 or
over (Ø13.3, r8)
EAPEP 6th edition
POP_YOUNG Children as a share of the population Share of the population aged 15 or
under (Ø43.7, r21.5)
EAPEP 6th edition
POP_MARRIAGE Mean age of marriage Mean age of first marriage for women
(Ø24.9, r3.7)
UN World Marriage Dataset
POP_LIFEEXP Mean life expectancy Mean life expectancy for women in
years (Ø73, r10)
World Bank Development Indicators
SOC_MATERNITY_LENGTH Length of maternity leave Length of statutory maternity leave
period in days (Ø111, r58)
ILO TRAVAIL
APPENDIX II. DESCRIPTION OF GLOBAL DATASET ON WOMEN AND WORK
140 WORLD DEVELOPMENT
SOC_MATERNITY_PAY Wage replacement during maternity leave Level of wage replacement (%) during
maternity leave (Ø89.3, r19.8)
ILO TRAVAIL
SOC_PREPRIMARY Enrollment rate pre-primary education Gross enrollment rate (%) in pre-
primary education (Ø60.8, r34.4)
UNESCO Institute for Statistics
Education
EDU_FEM1 Girls’ enrollment in primary education Gross female enrollment rate (%) in
primary education (Ø103.4, r14.6)
UNESCO Institute for Statistics
EDU_FEM2 Girls’ enrollment in secondary education Gross female enrollment rate (%) in
secondary education (Ø79.3, r28)
UNESCO Institute for Statistics
EDU_PAR1 Gender parity in primary education Ratio of female to male primary
enrollment (Ø0.79, r0.05)
UNESCO Institute for Statistics
EDU_PAR2 Gender parity in secondary education Ratio of female to male secondary
enrollment (Ø0.97, r0.13)
UNESCO Institute for Statistics
Gender ideologies
SOC_EQUALPAY_LEX Existence of equal pay legislation Dummy for the ratification of ILO
convention C100
ILO
SOC_NONDISCR_LEX Existence of nondiscrimination legislation Dummy for the ratification of ILO
conventions 3, 103 or 183
ILO
SOC_MATERNITY_LEX Existence of maternity legislation Dummy for the ratification of ILO
convention C111
ILO
CULT_POLRIGHTS Women’s political rights Four-point scale 0 (no rights)...3
(full rights by law and in practice)
(Ø2, r0.44)
CIRI Human Rights Database
RELI_BIG Size of the largest religious group Ratio of the largest religious
affiliation to the total population
(Ø0.78, r0.18)
Pew Research Center Forum on
Religion and Public Life
RELI_CHRD Christian heritage Ratio of Christians to the largest
(non)religious affiliation (Ø0.72,
r0.41)
Pew Research Center Forum on
Religion and Public Life
RELI_MUSD Muslim heritage Ratio of Muslims relative to the
largest (non)religious affiliation
(Ø0.28, r0.4)
Pew Research Center Forum on
Religion and Public Life
ScienceDirect
Available online at www.sciencedirect.com
WORKING WOMEN WORLDWIDE. AGE EFFECTS IN FEMALE LABOR FORCE PARTICIPATION IN 117 COUNTRIES 141
... Z denote the control variables, including age and squared age of the wife (Coen-Pirani et al., 2010), the education level of the wife (Besamusca et al., 2015;Coen-Pirani et al., 2010;Hare, 2016;Keats, 2018;Lincove, 2008), the education level of the husband (Aisenbrey, et al., 2009;Besamusca et al., 2015;Spierings et al., 2010), the husband's income (Angrist and Evans, 1998;Chiappori, 1992;Kleven et al., 2009;Malathy, 1994), and living with other adults who can assist in taking care of the children (such as grandparents or helpers) (Besamusca et al., 2015;Guo et al., 2018;Hare, 2016;Heath, 2017). ...
... Z denote the control variables, including age and squared age of the wife (Coen-Pirani et al., 2010), the education level of the wife (Besamusca et al., 2015;Coen-Pirani et al., 2010;Hare, 2016;Keats, 2018;Lincove, 2008), the education level of the husband (Aisenbrey, et al., 2009;Besamusca et al., 2015;Spierings et al., 2010), the husband's income (Angrist and Evans, 1998;Chiappori, 1992;Kleven et al., 2009;Malathy, 1994), and living with other adults who can assist in taking care of the children (such as grandparents or helpers) (Besamusca et al., 2015;Guo et al., 2018;Hare, 2016;Heath, 2017). ...
... Z denote the control variables, including age and squared age of the wife (Coen-Pirani et al., 2010), the education level of the wife (Besamusca et al., 2015;Coen-Pirani et al., 2010;Hare, 2016;Keats, 2018;Lincove, 2008), the education level of the husband (Aisenbrey, et al., 2009;Besamusca et al., 2015;Spierings et al., 2010), the husband's income (Angrist and Evans, 1998;Chiappori, 1992;Kleven et al., 2009;Malathy, 1994), and living with other adults who can assist in taking care of the children (such as grandparents or helpers) (Besamusca et al., 2015;Guo et al., 2018;Hare, 2016;Heath, 2017). ...
... Las complicaciones de la lactancia y su compatibilidad con el trabajo se originan a partir del incremento de la participación la mujer en el mercado laboral, esto es, a partir de la etapa de la posguerra en la década de 1950. Ello alcanzó un gran auge en la década de 1980 y una continua expansión hasta antes de la pandemia por Covid-19 cuando hubo despidos masivos por las políticas de confinamiento de la pandemia (Fullerton, 1999;Mosisa y Hipple, 2006;Toossi, 2012;Besamusca, Keune y Steinmetz, 2015) En principio, la inclusión de las mujeres en el mercado laboral fue visto como un avance positivo en materia de equidad e igualdad, sin embargo, tal inclusión supuso retos en materia de políticas y regulaciones, así como cambios en las estructuras sociales que hasta ese entonces dominaban. Dentro de los cambios estructurales más intensos que los países debieron enfrentar con la inserción laboral de las mujeres fue el del cuidado de los hijos, pues típicamente esta actividad estaba asociada a las féminas. ...
... Sin embargo, la participación de las mujeres en el mercado laboral ha sido uno de los factores que ha afectado su práctica (Cárdenas, Valle y Fernández, 2020). Históricamente, el papel de la mujer ha sido asociado con el cuidado de los hijos y la lactancia materna (Besamusca, Keune y Steinmetz, 2015). Sin embargo, con la entrada de las mujeres al mercado laboral en el siglo XX, se ha dado una reducción en la duración de la lactancia materna exclusiva, ya que muchas mujeres han tenido que abandonarla para volver al trabajo. ...
Article
Full-text available
Objetivo de este trabajo es indagar sobre los principales factores que condicionan y vulneran la Lactancia Materna Exclusiva (LME) en una región con vocación productiva orientada hacia el turismo: la Riviera Maya de Quintana Roo, México. Metodología: estudio exploratorio y descriptivo sobre el abandono de la lactancia materna exclusiva en mujeres que laboran en el sector turístico de la Riviera Maya, México. Se aplicó un cuestionario de recolección continua para conocer las condiciones y contextos que les impiden amamantar exclusivamente a sus hijos en el trabajo. Resultados: entre los principales hallazgos se encontraron tres factores que vulneran la lactancia materna: la flexibilidad del trabajo impide a las mujeres acceder a la licencia por maternidad; la nula existencia de trabajadoras de lactancia en los centros de trabajo; la distancia entre los centros urbanos de la Riviera Maya y los centros turísticos es bastante considerable en términos de tiempo y dinero. Limitaciones: el estudio solo considera una porción limitada de la población del estado de Quintana Roo. Conclusiones: es necesario considerar estos contextos para adecuar la legislación mexicana sobre LME en el trabajo.
... In an extensive study of 117 countries, Besamusca et al. (2015) wanted to discover how economic conditions, families, education, and gender ideologies determine the women's labour force participation rates in eleven age cohorts. Using data from the International Labour Organization between 1990 and 2010, the logarithmic regression confirmed the Ushape hypothesis of the relationship between the level of economic development and the FLFP for younger and older women but not for women between 25 and 55 years of age (Besamusca et al., 2015). ...
... In an extensive study of 117 countries, Besamusca et al. (2015) wanted to discover how economic conditions, families, education, and gender ideologies determine the women's labour force participation rates in eleven age cohorts. Using data from the International Labour Organization between 1990 and 2010, the logarithmic regression confirmed the Ushape hypothesis of the relationship between the level of economic development and the FLFP for younger and older women but not for women between 25 and 55 years of age (Besamusca et al., 2015). Even though the authors had expected a positive relationship between higher levels of education and the FLFP rate, it was found a positive relationship between enrolment in primary education of women between 25 and 55 years of age and their participation in the labour force. ...
... Scholars have conducted extensive research from various angles. Research based on the human capital theory primarily examines the roles of individual characteristics, such as age, household registration, education level, marital status, work experience, and health status, in influencing female employment [15,16]. Studies grounded in the family economic theory focus on the impact of characteristics like husband's income, number of children, parental care, allocation of household time, and family economic conditions [17,18]. ...
Article
Full-text available
Based on China Household Finance Survey (CHFS) data from 2019, this paper explores the impact of the residential pattern of coresidence with parents on the labor market performance of women in married families with minor children. The study finds that coresidence with parents significantly increases the possibility of female labor market participation and positively impacts women’s employment income. To overcome the potential endogeneity problem of residential patterns, this paper uses the Heckman two-step method and the conditional mixed process estimation method (CMP method) for regression, and the conclusions remain robust. The mechanism analysis shows that coresidence with parents has both grandchild care and elderly care factors, which have a spillover effect and a crowding-out effect on female labor market performance, respectively. Since the spillover effect is more significant than the crowding-out effect, coresidence with parents positively impacts women’s labor market performance. The heterogeneity analysis shows that in terms of labor force participation rate, coresidence with parents has a more significant impact on women in families with children aged 0–6, women in families without boys, and women in families with employed husbands. In terms of income, coresidence with parents has a more significant impact on women in families with employed husbands. This study provides a new perspective for promoting female labor market performance and can serve as a reference for future policy formulation.
... The barriers to closing gender wage gaps are multiple and differ by context, among them being social norms that preclude women from the workforce, participation in work that is unpaid, part-time, or informal, work in lower-paying economic sectors that are predominantly female, and the costs of child-bearing (Jayachandran, 2020;Kleven & Landais, 2017;Ortiz-Ospina & Roser, 2018;Terada-Hagiwara et al., 2018). An analysis of 117 countries found that women are more likely to participate in the labour market if there are paid maternity-leave schemes, there are higher pre-primary enrolments and less religiosity (Besamusca et al., 2015). ...
Article
Full-text available
Motivation Investing in girls’ schooling in low‐ and middle‐income countries (LMICs) is seen as central to improving gender equity. It is argued that interventions to promote girls’ enrolment are appropriate as girls face gendered barriers to school enrolment and completion and investing in girls’ schooling has high economic and human‐development returns. But is this fair to boys and enough for girls? Purpose We ask how appropriate it is to direct development assistance towards improving girls’ school enrolment, compared to prioritizing schooling for both girls and boys, and addressing barriers to gender equality throughout the life‐course. Methods and approach We frame the enquiry through a human development framework with three distinct but interdependent domains: protection of human development potential; realization of human development potential; and use of human development potential. Using publicly available data, we identify indicators likely to be correlated with the degree to which human development potential is protected, realized, and utilized in LMICs. We compare male and female outcomes on each of these indicators to assess gender parity at different life stages. Findings In most regions, girls are ahead of boys in both school enrolment and completion. Girls have better outcomes than boys in several other indicators in early life and childhood. In adolescence and adulthood girls and women fall behind boys and men. This is especially apparent in workforce participation, in unemployment, in pay, and in share of unpaid care work and political participation where women have less favourable outcomes than men. The bias against women is most marked in South Asia and sub‐Saharan Africa. Policy implications A focus on girls’ schooling should be tempered by ensuring quality pre‐primary, primary, and secondary schooling for both boys and girls. Simultaneously we must address causes of gender inequality, including labour market discrimination and social norms which justify the exclusion and exploitation of women and girls.
... Besamusca et al. [31] were the first to compile data at the national level on women's labor force involvement across age groups and individual factors. The ages of the participants influence the conclusions they reach. ...
Article
Full-text available
The fundamental objective of this research is to learn how trade liberalization, male employment, urbanization, and foreign direct investment (FDI) affect women’s participation in the labor force. To continue, this study aims to determine the effects of trade and other factors on women’s employment in three distinct sectors (i.e., agriculture, industry, and service). From 1991 to 2021, we analyzed data from eight SAARC countries. The study’s theoretical foundation was the Cobb–Douglas production function. To better understand the connections between trade liberalization and the SAARC labor market, this paper used panel quantile regression (QR) and generalized method of moments (GMM) to empirically explore the key determinants of female employment in total and three sub-sectors. The QR method was used in the study because it looks at how variables affect each other beyond the data mean. Additionally, our data set does not follow a normal distribution, and the connection between the explained and explanatory factors is non-linear. Trade openness has a beneficial effect on total female employment throughout system GMM and all quartiles. Total female employment also benefits from an increase in GDP and FDI. However, women’s access to the workforce is hampered by urbanization. Many strategies for increasing women’s participation in the workforce across three sectors are addressed in this article. The major finding of this study is the rate of change in female employment across three industries. Women’s participation in the service and manufacturing sectors increases, whereas their participation in agriculture decreases, as a result of increased trade openness. Although these studies can assist policymakers in choosing the best feasible trade adjustments, they will also add to diverse academic and policy discussions on trade liberalization and its gender consequences. Since trade has become more accessible, more and more women are entering the workforce. Therefore, workers should acquire industrial and service-sector-related competencies.
... For example, one study of 117 countries found that women ages twenty-five to fifty-five are more likely to participate in the labor force when paid maternity leave of moderate length is available. 9 Further, job-protected paid leave makes it more likely that women will return to the same workplace. For example, a 1999 analysis found that women's access to paid maternity leave in Britain and paid maternity and/or parental leave in Japan made it more likely that women returned to the same employer; 10 in this way, women's access to leave not only supports their individual employment outcomes but also reduces employers' turnover costs. ...
Chapter
Well into the twenty-first century, achieving gender equality in the economy remains unfinished business. Worldwide, women’s employment, income, and leadership opportunities lag men’s. Building and using a one-of-a-kind database that covers 193 countries, this book systematically analyzes how far we’ve come and how far we have to go in adopting evidence-based solutions to close the gaps. Spanning topics including girls’ education, employment discrimination of all kinds, sexual harassment, and caregiving needs across the life course, the authors bring the findings to life through global maps, stories of laws’ impact in courts and beyond, and case studies of making change. A powerful call to action, Equality within Our Lifetimes reveals how gender equality is both feasible and urgently needed to address some of the greatest challenges of our generation.
... For example, one study of 117 countries found that women ages twenty-five to fifty-five were more likely to participate in the labor force when their countries provided moderate-length paid maternity leave. 19 In Spain, the introduction of thirteen days of paternity leave increased mothers' probability of reemployment following childbirth by 11 percent. 20 And in California, two studies found that the introduction of an individual entitlement to paid parental leave, which was equally available to men and women, was associated with greater wages and working hours for mothers with children under age three. ...
Chapter
Well into the twenty-first century, achieving gender equality in the economy remains unfinished business. Worldwide, women’s employment, income, and leadership opportunities lag men’s. Building and using a one-of-a-kind database that covers 193 countries, this book systematically analyzes how far we’ve come and how far we have to go in adopting evidence-based solutions to close the gaps. Spanning topics including girls’ education, employment discrimination of all kinds, sexual harassment, and caregiving needs across the life course, the authors bring the findings to life through global maps, stories of laws’ impact in courts and beyond, and case studies of making change. A powerful call to action, Equality within Our Lifetimes reveals how gender equality is both feasible and urgently needed to address some of the greatest challenges of our generation.
... The fact observed in the results of this research, that the probability of women's employment increases with age, and then starts decreasing after the breaking point, is supported by the results of several studies. One of them is the study conducted by Besamusca et al. (2015) who, on a sample of 117 countries, found that FLFP increases after the completion of the mandatory education process, and decreases as the retirement phase approaches. In addition, Xin et al. (2021) identified that the turning point for the influence of age on women's employment is between the ages of 35 and 40. ...
Article
Full-text available
During time, women's participation in the labour markets all over the world has increased. However, gender differences in the level of employment still exist, especially in underdeveloped and developing countries. Since empirical evidence shows that a higher degree of women's employment has many positive effects on the economy, it is of great importance for such countries' governments to create mechanisms for increasing it. The