Skip to content

This report was published by the former Department of Families, Community Services
Contents | Downloads as PDF [397KB]

Education, work and fertility: a HILDA survey based analysis

Habtemariam Tesfaghiorghis 1

Research and Data Management Branch, Department of Family and Community Services

back to top

1. Introduction

The issue of work and family balance is high on the Australian policy agenda. The issues of fertility, family support, and balancing family and work responsibilities are being given prominence, as demonstrated by the significant measures announced in the May 2004-05 Federal Budget. Research and information is needed in this important area to inform policy debate and development.

Concomitant with fertility decline since the 1960s, there has been increasing female participation in education and the labour market. Over this period, in Australia there has been strong attachment to the labour market by women of childbearing and rearing ages (McDonald 2001a, p. 17). Increasing levels of female education are associated with increasing labour force participation. Education, ceteris paribus, is expected to lower fertility, as increasing educational duration leads to postponement of age at first marriage and age at first birth, and education may embody fertility behaviour that lowers fertility. If educated women did not compensate delayed births at later ages, then fertility would be lower.

Bratti's (2003, p. 30) study of Italy found that education raises the job attachment of women, highly educated women work in the period surrounding a birth event, and that education causes fertility postponement. According to Bratti (2002, pp. 27-8), early withdrawal of highly educated women from the labour market 'is costly both in terms of current opportunity costs (that is, wages) and future accumulation of human capital'.

The fact that, in general, the proportion of women with children is higher among those that do not work than those who work, and among those who work part-time than full-time, reflect the difficulty that women face in combining work and family responsibilities (Sleebos 2003, Figures 11 and 12). The type of jobs held by mothers is also important for decisions to have children, as part-time jobs generally allow women greater opportunities to combine work and family responsibilities (Sleebos 2003, p. 24). McDonald (2001a, p. 18) observes that:

... while the evidence is somewhat mixed, the balance of evidence is that countries in which there is little opportunity to work part-time have lower fertility rates because the choice between work and family is too stark.

Provision of child care is also important for women with children to have lifetime labour market careers so as to support their career development and on-the-job training (Gustafsson & Stafford 1994, p. 335).

It is thus hypothesised here that both education and labour force participation combine to lower fertility. The Household, Income and Labour Dynamics in Australia (HILDA) survey, provides an excellent source of information on fertility, particularly since the 2001 Census does not contain a question on total number of children a woman ever had. Thus, the HILDA survey, with its rich demographic and socio-economic data, provides a good opportunity to explore the associations of education and work with fertility.

This research-based on primary analysis of the 2001 HILDA survey Wave 1 dataset, Release 2.0-aims to contribute to an understanding of work and family balance issues for working-age women through examination of the following questions and issues:

The analyses in this paper comprise the following:

The main results are presented in '7. Summary'; and key issues, challenges and implications raised in the findings are highlighted in '8. Conclusion'.

back to top

2. Data and methods

Public knowledge and discussion on fertility is based on period/current fertility, which is the fertility experienced by different cohorts of women that gave birth in a particular year/period. It is usually measured by age-specific fertility rates and/or period total fertility rate (TFR). The period TFR, derived by summing the age-specific fertility rates, is a synthetic measure that gives the average number of children a woman would bear in her lifetime. The TFR in 2002 is 1.75 children per woman. Unlike current fertility, cohort fertility is the fertility experienced by a particular cohort of women-birth cohort, age cohort or marriage cohort-during their reproductive life. The completed cohort fertility is measured by completed fertility rate, which is also known as cohort TFR.

Instead, the analysis here is based on a measure of lifetime fertility-that is, the number of children ever born by a woman, which the HILDA survey collected. The measure used is the mean number of children ever born (MCEB)/average parity, which is calculated by dividing the total number of children ever born to women of a given age group by the total number of women in the given age group. For the purpose of comparative fertility studies, MCEB figures are provided by age group and overall for the whole age range. Thus, what is measured is fertility of age cohorts, as given by the mean number of children for that age cohort. The average numbers of children for women with completed fertility-those aged 40 years and over-are equivalent to completed fertility rate.

In addition, adjusted MCEB figures (labelled 'adjusted' in the tables in this paper) are given. Adjusted MCEB is a measure calculated to control for the effect of age composition differences on fertility. This is done by assuming that each category of a variable of interest has the same standard age distribution-in this case, as the age distribution of total women in the HILDA survey. Thus, any fertility effect that remains after controlling for age composition can be attributed to the variable of interest. The sum of the products of MECB, and the corresponding standard population over the age range, provide an adjusted MCEB.

The HILDA survey data on number of children ever born are subject to sampling errors, misreporting and parity not stated (unstated number of children born). The HILDA survey is a nationally representative longitudinal survey of 7682 households, which administered household interviews and person interviews of all persons aged 15 years and over within the selected households. A guide to the calculation of standard errors is given in a HILDA survey technical document (Horn 2004). The results here are based on population-weighted figures and associated standard errors are not given. In the analysis cells, a small number of cases are combined with similar categories to produce meaningful results. The data on number of children ever born usually suffer from parity not stated. This was the case in Australian censuses, except the 1996 Census (McDonald 1998, p. 7).

The HILDA survey reporting on number of children ever born is complete. The reason is the use of better trained interviewers in the HILDA survey than the self- completion questionnaire used in the censuses. Further, the MCEB figures for comparable age groups from the HILDA survey is consistent with 1996 Census and Australian Bureau of Statistics (ABS) birth registration data (Tesfaghiorghis 2004b).

Thus, the quality of the MCEB data from the HILDA survey are reliable and cell sizes for most tabulation categories are large enough to warrant differential fertility analysis by education and work.

This is exploratory research based on a descriptive analysis. As such, it will attempt to identify the key issues in the work and family balance by a preliminary investigation of the factors involved and their associations. Thus, the question of causal relationship is not addressed here. The methods include analysis of MCEB using cross-tabulations, graphs, and application of standardisation to control for age composition effects on fertility. Analyses of the historical trends in completed fertility and a multivariate analysis of the factors influencing fertility are the subjects of another paper (Tesfaghiorghis 2004b).

back to top

3. Education and fertility

The analyses look at the association between fertility, age left school, highest educational level, and time since completing full-time education.

Age left school and fertility

The more young women stay longer in school at a higher rate, the more it is expected to lead to fertility postponement. A higher proportion of women are staying in school longer. Fifty-one per cent of women left school at ages 16 or younger in the 2001 HILDA survey compared to 71 per cent in the 1986 Census. Fifty-four per cent of the 15-19 year-olds in the 2001 HILDA survey were still at school compared to 44 per cent in the 1986 Census. This is also consistent with increasing school retention rates for young women, as national retention rates of secondary school students doubled from 35 per cent in 1980 to 77 per cent in 1992, with higher rates observed for females than males (AIHW 2003a, p. 272-3).

The analysis of HILDA survey fertility data shows that the age at which a woman leaves school is negatively associated with fertility. For example, among women aged 30-34 years those who left school at 18 years or older had lower mean numbers of children (1.0 child on average), than those who left at 15 years or younger (1.98 children)-table not shown. The corresponding figures for those aged 45-49 years were 1.94 and 2.59 children, respectively. The results show the MCEB by age consistently declines with increasing ages of leaving school, which may be partly to do with an orientation towards work and career rather than motherhood. Overall, after controlling for differences in age composition, women aged 15-64 years and who left school at age 18 or older had 1.60 children compared to 2.25 children for those who left school at 15 years or younger.

Highest education level and fertility

Table 1 presents MCEB according to age group and highest educational level of women. It also includes other descriptive measures (see last four rows of table) such as overall reported MCEB for women aged 15-64 years and adjusted MCEB (adjusted for differences in age composition) by educational level, per cent of total women, and mean age of women according to educational level.

Table 1 shows a number of interesting results:

The mean age figures show that women with postgraduate degrees, followed by those with Year 11 or lower education and those with certificates I & II were the oldest, while those with Year 12 were the youngest. After controlling for the effects of age composition, the adjusted results support the decline of fertility with increasing education. Women with postgraduate degrees had the lowest overall mean number of children (1.1 children), while those with Year 11/below had the highest fertility (2.0 children). The latter accounted for 37.0 per cent of women aged 15-64 years, while the former accounted for 6.5 per cent.

Table 1: Mean number of children ever born by age group and educational level and summary measures by educational level-2001 HILDA survey

Age group

Postgraduate(a)

Bachelor

Diploma(b)

Cert. III & IV

Cert. I & II(c)

Year 12

Year 11(d)

15-24

0

0.02

0.02

0.19

0.25

0.07

0.2

25-29

0.32

0.24

0 .59

0.92

1

0.64

1.57

30-34

0.76

0.82

1.23

1.65

1.69

1.38

1.94

35-39

1.59

1.77

1.85

2.13

2.61

1.8

2.37

40-44

1.59

1.81

2.13

2.12

2.47

2

2.58

45-49

1.68

1.87

2.18

2.25

2.14

2.27

2.61

50-64

1.94

2.28

2.61

2.62

2.55

2.31

2.91

               

MCEB - reported(e)

1.37

1.16

1.6

1.68

1.87

1.02

2

MCEB - adjusted(f)

1.12

1.27

1.51

1.67

1.76

1.46

1.96

% women

6.49

13.42

8.71

9.96

11.44

12.96

37.02

Mean age

41.39

36.67

38.87

37.81

39.02

31.24

39.31

(a) Comprises doctorate, masters, graduate diploma/certificates. Those with doctorates or masters had the lowest fertility and account for about 2 per cent of all women.

(b) Comprises advanced diploma, diploma.

(c) Includes women with certificate not defined.

(d) Year 11/below also includes women with undetermined educational level.

(e) MCEB - reported = MCEB to women of a given educational level aged 15-64 years, calculated by dividing the total number of children born to the total number of women in the given educational level.

(f) MCEB - adjusted = MCEB to women aged 15-64 years in the given educational level, after controlling for differences in age composition.

Source: Primary analysis of 2001 HILDA survey Wave 1 dataset, Release 2.0.

Time since full-time education and fertility

Of all women aged 15-64 years, 15 per cent left full-time education less than five years ago-about 10 per cent each 5-9, 10-14 and 15-19 years ago, and 53 per cent 20 or more years ago. The results on the association between time since completing full-time education and fertility show that the shorter the duration since completing full-time education, the lower the fertility. Women are not only postponing childbearing while pursuing full-time education, but are also postponing childbearing for up to 10 years of completing full-time education, reflecting the time needed to build careers and relationships. Young women aged less than 30 years and who left full-time education less than 10 years ago had zero or negligible fertility (table not shown). This is supported by the ABS trends in age-specific fertility rates, which consistently show declining fertility trends for young women under 30 years and a recent stabilisation or a small increase for those aged over 30 years (ABS 2003, Table 2.8). Women aged 25-34 years and who left full-time education 10-14 years ago had on average borne 1.0 child, compared to 0.4 children for those who left 5-9 years ago.

back to top

4. Work and fertility

This section will examine fertility differences by labour force status, educational level, and labour force status; and occupation of main job, number of hours worked, and actual and preferred work hours. Of all women aged 15-64 years, 66 per cent participated in the labour force (33 per cent worked full-time and 29 per cent part-time, and 4 per cent were unemployed). Twelve per cent had marginal attachment to the labour force and 22 per cent were not in the labour force (see note (a), Table 2).

Labour force status and fertility

Fertility varies by labour force status. At every age, women employed full-time had the lowest MCEB. Full-time workers aged less than 30 years had negligible fertility. This is also true for part-time workers aged less than 25 years. Part-time employment is associated with lower fertility. While part-time employed women had the second lowest MCEB by age group, those not in the labour force had the highest fertility. Women aged 20-44 years who were marginally attached to the labour force had a higher mean number of children than those not marginally attached to the labour force at every age but the 30-34 age group. Young unemployed women had higher MCEB than part-time employees but had a similar fertility level at older ages. The unemployed looking for full-time work, and those looking for part-time work, each comprising about 2 per cent of all women aged 15-64 years, had different fertility levels but were combined together because of small numbers. The former had similar fertility to full-time employees, while the latter had similar fertility to those not in the labour force.

The adjusted results show that the overall MCEB to women aged 15-64 years was 1.23 children for those employed full-time, 1.68 children for part-time workers, and 1.71 for the unemployed. By contrast, it was 2.09 children for those marginally attached to the labour force and 2.03 for those not attached to the labour force.

The results in Table 2 clearly show that labour force participation is associated with lower fertility, particularly full-time employment. The relevance of this finding is that women who participate in the labour force have difficulty in combining family and work responsibilities.

The association between educational level, labour force status and fertility is presented in Table 3. The results show that for women who had the same level of education, their fertility varies according to their labour force status. For example, the MCEB to women with postgraduate degrees increases from 1.1 children for those employed full-time to 1.6 children for those employed part-time, and about 2.0 children for those not in the labour force. Given the same level of education, labour force participation tends to lower fertility. The exception is Year 11 or lower education.

Those employed full-time are better educated; 46 per cent had diploma or higher degrees versus 28 per cent for those employed part-time and 18 per cent for those not in the labour force. Further, a higher proportion of highly educated women were in employment, particularly in full-time work, relative to those with lower educational levels. For example, 54 per cent of all women with postgraduate degrees worked full-time and 30 per cent worked part-time. These were 50 per cent and 28 per cent for those with bachelor degrees, and 18 per cent and 27 per cent for those with Year 11 or lower education. Nevertheless, even after controlling for educational differences, the results show that employment, particularly full-time, is still associated with lower fertility.

Table 2: Mean number of children ever born by age group and labour force status, and summary measures by labour force status-2001 HILDA survey

Age group

Employed

Unemployed

Not in the labour force
 

Full-time

Part-time

  

Marginally
attached(a)

Not marginally
attached(b)

15-19

0

0

0.06

0.04

0.05

20-24

0.03

0.1

0.59

1.11

0.65

25-29

0.16

0.99

1.15

1.96

1.59

30-34

0.64

1.54

1.78

1.95

2.27

35-39

1.53

2.26

1.6

2.74

2.35

40-44

1.77

2.37

2.36

2.84

2.51

45-64

2.22

2.52

2.54

2.63

2.84

           

MCEB - reported

1.2

1.63

1.23

1.83

2.28

MCEB - adjusted

1.23

1.68

1.71

2.09

2.03

% women

33.08

29.02

4.05

11.86

21.99

Mean age

37.31

36.34

30.94

34

43.95

(a) Persons not in the labour force are considered to be marginally attached to the labour force if they want to work and are actively looking for work but not available to start work in the reference week; or want to work and are not actively looking for work but are available to start work within four weeks.

(b) Persons not in the labour force are not marginally attached if they do not want to work; or want to work but are not actively looking for work and are not available to start work within four weeks (Freidin et al. 2004).

Source: Primary analysis of 2001 HILDA survey Wave 1 dataset, Release 2.0.

Table 3: Mean number of children ever born by highest educational level and labour force status-2001 HILDA survey

Educational level

Employed

Not in labour force

 

Full-time

Part-time

Marginally attached

Not attached

Postgraduate

1.06

1.65

1.87

2.01

Bachelor

0.88

1.42

1.58

1.61

Diploma

1.14

1.74

2.02

2.46

Certificates III & IV

1.16

1.87

1.96

2.23

Certificates I & II

1.31

2.09

2.16

2.3

Year 12

0.72

0.9

1.42

1.71

Year 11/below

1.86

1.77

1.85

2.48

Source: Primary analysis of HILDA survey Wave 1 dataset, Release 2.0.

The observed fertility differences by labour force status and educational level may reflect the choices and constraints working women face. The issue of choice is that women who have a career orientation, high level occupation and work full-time are less likely to have children or find it hard to pursue a career. They would have lower fertility. However, highly educated Italian women were found to have higher marital fertility because they were more likely to work and able to purchase external private child care with their incomes (Bratti 2003, pp. 15-17). The issue of constraint is that the higher the numbers of children women have, the harder they find it to combine work and fertility.

Occupation and fertility

The association between the occupation of main job and fertility is examined in Table 4. The table sets out the MCEB by age group and occupation of main job, and overall reported and adjusted MCEB by occupation. When account is taken of the significant age differences, the adjusted results show that fertility was high among cleaners, factory workers and other labourers. It was also relatively high for intermediate workers, advanced clerical and service workers, and elementary sales and service workers. Fertility was lowest among the science, business and information professionals. Health and education professionals and associate professionals, and tradespersons had lower fertility. Further, these overall results are supported by completed fertility rates for women aged 45-64 years. Completed fertility rates ranged from a low of 1.85 children for science, business and information professionals to 2.4 children for intermediate workers and elementary sales and service workers, and a high of 2.62 children for cleaners, factory workers and other labourers.

Table 4: Mean number of children ever born according to occupation of main job and age group-2001 HILDA survey

Occupation

MCEB by age

Overall

 

15-24

25-34

35-44

45-64

Reported

Adjusted

Managers & administrators

0.2

0.92

1.71

2.24

1.74

1.4

Science, business & information professionals

0

0.26

1.57

1.85

0.89

1.05

Health, education & other professionals

0.02

0.57

1.9

2.3

1.44

1.35

Tradespersons & related workers

0.05

0.69

1.67

2.26

1.14

1.32

Advanced clerical & service workers

0

0.91

2.47

2.24

1.77

1.53

Intermediate sales/service & production workers

0.05

0.93

2.09

2.4

1.42

1.52

Elementary sales & related workers

0.02

0.82

2.1

2.4

1.02

1.48

Cleaners, factory & other labourers

0.08

1.21

2.13

2.62

1.84

1.73

Source: Primary analysis of 2001 HILDA survey Wave 1 dataset, Release 2.0.

The results appear to indicate that professional occupations are associated with lower fertility, while lower level occupations are associated with higher fertility. These occupational fertility differences could also be due to differences in educational composition of occupations.2 Those in lower level occupations generally have lower education levels, while those in higher level occupations have higher educational qualifications. It is also possible that those with higher levels of fertility choose different occupations to those with lower fertility, rather than occupation influencing fertility.

The observed fertility differences by labour force status and occupation type may arise because couples make labour supply decisions based on their characteristics or circumstances, which are aimed at optimising family outcomes, rather than maximising income. It could also be that women's fertility is constrained by, or facilitated by, the hours and working conditions in their industry or occupations.

Fertility by number of hours worked

Of all employed women aged 15-64 years, 28 per cent worked 20 hours or fewer per week, 19 per cent for 21-34 hours, 32 per cent for 35-40 hours, and 21 per cent for 41 hours or more. Those who worked for 20 hours or fewer were younger, followed by those who worked for 35-40 hours, while those who worked for 21-34 hours were older followed by those who worked 41 hours or more (see last row of Table 5).

As is seen in Table 5, the numbers of hours worked per week are associated with fertility. The MCEB by age was lower for those who worked long hours, particularly 41 hours or more. Of those who worked shorter hours, those who worked for 20 hours or less had higher fertility.

It is clear from Table 5 that working more hours are associated with lower fertility. However, this is not a causal relationship. It could be that women with high fertility chose to work fewer hours to care for their children and family, or women have chosen to work full-time and long hours over having a high number of children. The data appear to suggest that women with a high number of children are working part-time to care for children and other personal/family responsibilities (table not shown).

Table 5: Mean number of children ever born by age group and number of hours worked per week, and summary measures by hours worked-2001 HILDA survey

Age group

Number of hours worked per week

 

20 hrs or less

21-34 hrs

35-40 hrs

41 hrs+

15-19

0

0.03

0

0

20-24

0.06

0.18

0.04

0.01

25-29

1.15

0.74

0.18

0.12

30-34

1.78

1.25

0.7

0.54

35-39

2.39

2.08

1.65

1.4

40-44

2.49

2.23

1.81

1.73

45-64

2.59

2.45

2.3

2.12

MCEB - reported

1.53

1.78

1.17

1.23

MCEB - adjusted

1.77

1.57

1.28

1.16

% women

28.07

18.64

32.14

21.15

Mean age

34.16

39.62

36.3

38.83

Source: Primary analysis of 2001 HILDA survey Wave 1 dataset, Release 2.0.

Fertility by number of actual and preferred weekly work hours

Of all employed women, 27 per cent preferred to work fewer hours, 56 per cent preferred the same hours, and 17 per cent preferred more hours. Those who preferred to work fewer hours were older, while those who preferred to work more hours were younger.

The results in Table 6 show that the mean number of children born by actual hours worked and preferred hours were lower for those women who preferred to work more hours and higher for those who preferred to work fewer hours. However, for those who worked 41 hours or more, those who preferred the same number of working hours had higher fertility than those who preferred fewer hours.

However, an examination of mean number of children born both by age group and preferred work hours shows that older women who preferred to work more hours had higher fertility than those who preferred the same or fewer work hours. When the differences in age composition were controlled, those who preferred to work more hours had higher fertility (1.62 children), and those who preferred fewer hours had lower fertility (1.29 children)-see 'adjusted' in Table 6. This is because older women who preferred to work more hours have grown up children that do not constrain their labour force participation.

The distribution of women by actual and preferred hours (second panel of Table 6) shows that a relatively high proportion of those who worked shorter hours preferred to work more, while those who worked longer hours preferred to work fewer or about the same hours. For example, of those who worked 20 hours or less, 59 per cent preferred to work the same hours and 36 per cent preferred to work more. By contrast, 58 per cent of those who worked 41 hours or more preferred fewer hours and 39 per cent preferred the same hours.

Table 6: Mean number of children ever born by actual and preferred work hours, and distribution of women according to actual and preferred work hours-2001 HILDA survey

Hours worked

Prefer to work

Per week

Fewer hours

About the same

More hours

20 hrs or less

1.65

1.62

1.36

21-34 hrs

2.08

1.89

1.34

35-40 hrs

1.53

1.03

0.83

41 hrs or more

1.2

1.33

0.68

MCEB - reported

1.45

1.42

1.26

MCEB - adjusted

1.31

1.48

1.65

Mean age

39.73

36.8

32.42

% distribution(a)

      

Hours worked

Fewer hours

About the same

More hours

20 hrs or less

5.5

58.9

35.6

21-34 hrs

15.5

61.8

22.7

35-40 hrs

31.9

60.2

7.9

41 hrs or more

58

39.4

2.6

Total

27

55.7

17.3

(a) Row '% distribution' for categories of preferred work hours adds up to 100 per cent.

Source: Primary analysis of 2001 HILDA survey Wave 1 dataset, Release 2.0.

back to top

5. Age and number of own resident children and labour force status

This section will examine the association between age and number of young resident children in the household and women's labour force participation. Gustafsson and Stafford (1995, pp. 163-4) showed that the majority of mothers in the United States, Sweden and the Netherlands did not work, 58 per cent, 59 per cent and 74 per cent respectively, when the child is less than one year. In the United States, the majority of mothers returned to work, mainly full-time, after the child's first birthday, with the employment rate increasing by child's age. In the Netherlands, the majority of mothers stayed at home to care for their preschool children. It is only when their children reached four years that about 40 per cent of them worked, mainly short hours-that is, less than 20 hours per week. In Sweden, maternal employment rate doubled after the child's first birth with a larger proportion working 20 to 34 hours a week.

Gustafsson and Stafford's (1994, pp. 352-4) modelling of married women's labour force participation in the United States, the Netherlands and Sweden showed that:

... the impact of children under age three is to significantly lower labour force participation, this effect is most pronounced in Sweden, where the generous parental leave program has its impact. Children aged 3-5 have a far smaller impact on participation.

Using the 1986, 1991 and 1996 Australian censuses, McDonald (2001b, pp. 17-18) found that the maternal employment rate remained low between the censuses when the youngest child is a baby (24-8 per cent), and that maternal employment increases substantially as the youngest child reached one or two years, and three or five years. Using the 1986 and 1996 Australian censuses, Gray et al's (2003, pp. 10-12) labour force status modelling found that:

... both the age and number of children has a strong and statistically significant impact upon the labour force status of both couple and lone mothers. The predicted probability of being full- time and part-time employed increases as the age of the youngest child increases.

The HILDA survey will contribute to these findings by providing an insight into the association between the age and number of own resident children and detailed labour force status. The presence of own resident young children, particularly 0-4 year-olds, appears to reduce women's labour force participation. Figure 1 displays the relationship of number of own resident children aged 0-4 years to labour supply choices women make. The results are:

Compared to own resident children aged 0-4 years, a higher proportion of those employed full-time had own resident children aged 0-14 years (20 per cent). This figure is even higher for those employed part-time, 40 per cent (Figure 2). Further, Figure 2 shows that 51 per cent of those marginally attached to the labour force (NILF-MA) had own resident children, with about 30 per cent having at least two children.

Another issue examined here is how labour force status varies according to number of own resident children aged 0-4 years. Figure 3 shows that if women have young children in the household, particularly two or more children, only a small proportion worked full-time. However, the proportion working part-time remained stable at a high level, irrespective of whether women have no children, one child or more than one child in the household-29 per cent, 32 per cent and 28 per cent, respectively.

Figure 1: Percentage of women by number of children aged 0-4 years and labour force status

Figure 1: Percentage of women by number of children aged 0-4 years and labour force status

Source: HILDA 2001.

Figure 2: Percentage of women by number of children aged 0-14 years and labour force status

Figure 2: Percentage of women by number of children aged 0-14 years and labour force status

Source: HILDA 2001.

Figure 3: Women's labour force status according to number of resident children 0-4 years

Figure 3: Women's labour force status according to number of resident children 0-4 years

Source: HILDA 2001.

A further insight into the employment constraints faced by mothers of young children is gained by considering the analysis by age of youngest child. The results are presented in Table 7 and Figure 4. The HILDA survey results show that the employment rate of mothers with infants is very low (22.5 per cent), compared to 41.4 per cent when the child reached age one year and 42.6 per cent when the child is aged two years. Thereafter, mothers' employment rates rise slowly with increasing age of their children-from 44.1 per cent when the child turns three years to about 50 per cent when the child is aged four or more years.

Table 7: Comparisons of mothers' labour force status (%) by age of youngest child between 2001 HILDA survey and 2001 Census

Age of child

HILDA survey 2001

Census 2001(a)

 

EMP-FT

EMP-PT

Unemployed

NILF

Employed(b)

Unemployed

NILF

0

6.5

16

1.7

75.8

25.3

2.6

64.5

1

9.9

31.5

5.6

53.1

42.8

3.7

51.4

2

11.8

30.8

4.1

53.3

46.6

4.1

47.4

3

10.8

33.3

3.8

52.2

49.3

4.3

44.4

4

13

35.9

2.5

48.7

51.4

4.5

42.1

5

18.5

35.7

5.8

40

54.9

5.2

38.1

6

14.5

36.3

3.9

45.2

58.9

5.2

34.2

(a) The row percentages for Census 2001 do not add up to 100 per cent because mothers working '0' hours, but did not work during the reference week, were excluded.

(b) Employed includes those who worked one hour or more.

Sources: Primary analysis of 2001 HILDA survey Wave 1 dataset, Release 2.0; and, for 2001 Census, AIHW 2003b, Table A6.3.

Figure 4: Mothers' labour force status by age of children

Figure 4: Mothers' labour force status by age of children

Source: HILDA 2001.

A salient result is that like Sweden, mothers' return to employment is dominated by part-time employment, which rose from 16 per cent when the child is aged less than one year to 32 per cent when the child is aged one year, and stabilised at about 36 per cent when children are aged 4-6 years. By contrast, the full-time employment rate remains low, rising gradually from 6.5 per cent when the child is under one year to 10-12 per cent by ages 1-3 years, and 13 per cent by age four.

The comparisons of HILDA survey results with the 2001 Census show that the pattern and magnitude of maternal employment by age of child are generally consistent. The HILDA survey results are also consistent with McDonald's (2001b, pp. 17-18) finding, though his employment figures are higher than these results.

It is not clear from these associations whether the number of young children is influencing women's labour supply decisions or whether women's labour supply decisions are influencing their fertility decisions. At this juncture, it suffices to say that there is a clear association between the number and age of children and labour force participation. However, the consistent finding that maternal employment is low when the child is less than one year, and increases with the age of child after the child's first birthday (mainly through part-time employment), is relevant to targeting family assistance by age of children and when to support mothers as they return to work.

back to top

6. Fertility and labour force participation

The preceding analysis showed that the employment rates of mothers of 0-4 year-olds is low when they have infants and steadily increases with children's age. Besides the age of young children, it is expected that a high number of children would lower women's labour force participation. Del Boca et al. (2003, p. 12, and Table 10) found that the number of children already present in the family has a significant negative effect on women's participation in the labour market. The association between the number of children ever born and women's labour force status is presented in Table 8.

Table 8: Distribution of women by number of children ever born and labour force status according to age group of women
 

Employed

Unemployed

Not in the labour force

% Total women

Age group/
number of
children
ever born

Full-time

Part-time

  

Marginally
attached

Not
attached

  

15-24

            

Zero parity

29.6

35.9

7.9

13.7

12.9

90.5

1+ child

6.1

11.4

11.8

43.4

27.2

9.5

Total

27.3

33.6

8.3

16.6

14.2

1 346 632

25-34

            

Zero parity

72.1

16.2

3.3

3.4

5.1

45.4

1 child

21.6

36

1.8

18.8

21.8

20.5

2 children

13.5

34.6

3.5

21

27.3

19.5

3+ children

10.9

26.2

5.2

26.7

31

14.7

Total

41.4

25.3

3.3

13.4

16.6

1 436 269

35-44

            

Zero parity

62.5

18.1

1.8

2.6

14.9

15

1 child

42.6

30.6

5.4

8.1

13.4

12.9

2 children

29.4

39

3.2

11

17.4

36.1

3+ children

22.9

37.8

2.1

17.9

19.2

36.1

Total

33.7

34.4

2.9

11.9

17.2

1 476 423

45-64

            

Zero parity

45.8

17.9

2.7

6.9

26.9

8.4

1 child

32.6

22.9

4.4

7.4

32.8

9.8

2 children

32.6

26.1

2.5

8.4

30.5

36.3

3+ children

26.1

26.1

2.5

8

37.3

45.5

Total

30.8

25.1

2.7

8

33.5

2 198 643

Source: Primary analysis of 2001 HILDA survey Wave 1 dataset, Release 2.0.

The salient points of Table 8 are summarised as follows:

back to top

7. Summary

There were substantial fertility differences by education and labour force status. The younger the age of leaving school, the higher the fertility, irrespective of the current age of women. The research found that time since full-time education is an important influence on fertility, where substantial childbearing occurs on average only after 10 years of leaving full-time education. Increasing educational level is associated with lower fertility. For women that completed their fertility, it is found that education lowers fertility. For younger women, education postpones fertility but may not lower their actual fertility, as they have incomplete fertility.

Labour force participation is associated with low fertility, particularly full-time employment. Compared to those employed full-time, women employed part-time have higher fertility, particularly those who work part-time to care for children or for other personal/ family responsibilities. Those who worked full-time, particularly 41 hours or more per week, had the lowest fertility.

This analysis also found that higher education and full-time employment are associated with lower fertility. It is likely that the opportunity cost to these women will be higher in terms of lost earnings and taking time off to have children. The analysis found that women employed full-time had higher education levels than those employed part-time or in other labour force statuses.

How does labour force participation influence fertility? The analysis of the relationship between labour force status and age and number of own resident children found that most women who worked full-time or looked for full-time work had no resident children aged 0-4 years in the household. By contrast, a significant proportion of those employed part-time and not in the labour force had children aged 0-4 years. Thus, it appears that labour force decisions women make are influencing fertility.

How does fertility relate to labour force participation? It is found that women with 0-4 year-olds have lower employment rates, particularly full-time (Figure 3). The full-time employment rate is small if they have two or more children aged 0-4 years. When the employment rate of mothers with 0-4 year-olds is considered by age of child, it is found that maternal employment rate is low, when the child is less than one year. The maternal employment rate increases substantially with child's age after the child's first birthday. Mothers return to employment primarily through part-time employment. It is also found that a high number of children ever born, particularly for prime working-age women, are associated with lower labour force participation.

It is not clear from these associations whether the age and number of young children women have is influencing labour supply decisions or whether women's labour supply decisions are influencing their fertility decisions. It is very likely that there is reciprocal causality between fertility and labour supply decisions.

back to top

8. Conclusion

There are several conclusions from this research that raise many issues and challenges. First, it appears that women delay childbearing after completing full- time education for, on average, up to 10 years. This may indicate that young women are delaying their fertility until such time as they build their relationships and/or careers, or because they find it difficult to combine work with childbearing.

Second, those who work full-time have lower fertility, while part-time workers and those out of the labour force have higher fertility. Is this because those employed full-time are giving priority to their jobs over childbearing or because of the difficulties they face in combining work and childbearing? Women who are not in the labour force have the highest fertility, and it is likely that many of the mothers currently not in the labour force are doing so to care for their children. A significant proportion of employed women are working part-time so that they can care for their children, and these mothers have the highest fertility among part-time workers.

Third, most women who currently work full-time have no children aged 0-4 years. Of those mothers with own resident children (younger than five years), the majority were either out of the labour force or in part-time employment. Only a modest proportion of mothers with 0-4 year-olds were in full-time employment. The finding of increasing maternal labour force participation by age of children aged 0-4 years is relevant to targeting assistance to support maternal fertility and work.

McDonald (2001b, pp. 18-19) argues that policy should take into account mothers' sharp increase in labour force participation as the youngest child ages from age zero to age four. He argues that all mothers of young children, irrespective of their employment status, need a combination of income support and child care support so that the transition from home to work can be undertaken without experiencing substantial loss of benefits (McDonald 2001b, pp. 18-19).

How are these mothers of young children combining childrearing with employment? The issue of child care is not analysed here, but elsewhere (Tesfaghiorghis 2004a) it has been shown that 84 per cent of mothers both in full- time and part-time employment with children aged 0-4 years used child care. This indicates that child care is supporting mothers to participate in paid employment.

This research has explored the associations between education, work and fertility, where both education and labour force participation combine to lower fertility. It also found that fertility and the number and age of own resident young children are associated with lower labour force participation, particularly full-time employment.

This research has not explored the issues of what can be done to support mothers to combine childbearing and rearing with work responsibilities. So what does the literature say on this?

McDonald (2000, p. 9) states that:

... obviously, among all family services, provision of childcare is highly beneficial to the employment of women and hence to a higher level of gender equity. ... Again, low expenditure on family services matches low fertility.

Gustafsson and Stafford (1994, pp. 343-4) state that the parental leave program extensions in the 1980s combined with the day care system are regarded as having created a fertility boom in Sweden in the early 1990s.

The differing design of family-friendly policies followed by countries may have different effects on family life, fertility and labour force participation (Gilbert & Voorhis 2003, pp. 55-6). For example, the 1998 policy initiative in Norway to pay cash benefits to all families with children up to three years-old if the child was not enrolled in a state subsidised day care centre resulted in higher marriage rates and fertility but lowered labour force participation (TFR = 1.85 in 2000), compared to Sweden (TFR = 1.54 in 2000) with its 'extensive provision of public day care services, which only benefit those who work' (Gilbert & Voorhis 2003, pp. 55-6).

Gilbert and Voorhis (2003, p. 56) observe that:

It is difficult to imagine a reversal of the heightened labour force participation of women. The contemporary issue for family policy is not whether to assist the transfer of labour from household to market, but how to help parents manage this shift in a way that lends adequate consideration to various desires for work and children at different stages of family life.

Beyond work related measures, family support programs that could have the potential to reverse declining trends in marriage and fertility rates by strengthening family life, could be tried from a mix of policies. These include paid childrearing leave as an alternative to day care, pension credits towards retirement for parents who stay home to care for young children, and the normalisation of part-time employment with measures that promote choices and that allow different parents to balance work and family life according to their varied preferences (Gilbert & Voorhis 2003, p. 56).

This research has only attempted to identify the key issues in work and family balance by preliminary investigation of the factors involved and their associations. The answers to the problems raised are partial at best. The next phase of the research is to undertake a multivariate analysis of factors that influence fertility and labour force participation, so as to establish the independent and joint effects of the key variables identified in this research.

back to top

Endnotes

1. The author wishes to thank Heather Evert, Helen Moyle and the two anonymous referees for their valuable comments.

2. The following educational profile of selected occupational groups demonstrate the point:

back to top

References

Allen, Susan F 2003, 'Working parents with young children: cross-national comparisons of policies and programmes in three countries', in International Journal of Social Welfare, vol. 12, pp. 261-73.

Australian Bureau of Statistics (ABS) 2003, 2002, Births: Australia, cat. no. 3301.0.

Australian Institute of Health and Welfare (AIHW) 2003a, Australia's young people: their health and wellbeing, AIHW cat. no. PHE 50, Canberra.

- 2003b, Australia's welfare 2003, AIHW cat. no. AUS 41, Canberra.

Barnes, Allison 2001, Low fertility: a discussion paper, Department of Family and Community Services (FaCS) Occasional Paper no. 2, FaCS, Canberra.

Bratti, Massimiliano 2003, 'Labour force participation and marital fertility of Italian women: the role of education', Journal of Population Economics, vol. 16, pp. 525-54.

Del Boca, Daniela, Pasqua, Silvia & Pronzato, Chara 2003, 'Analyzing women's employment and fertility rates in Europe: differences and similarities in Northern and Southern Europe', Centre for Household, Income, Labour and Demographic Economics, University of Torino, Italy.

Fisher, Kathleen & Charnock, David 2003, 'Partnering and fertility patterns: analysis of the HILDA Survey, Wave 1', paper presented at the HILDA Conference, University of Melbourne, Melbourne, 13 March 2003.

Freidin, S, Watson, N, Warren, D & Wooden, M 2004, HILDA Survey Waves 1 and 2: users' guide, downloaded with HILDA Wave 2 documentation from the website http://www.melbourneinstitute.com/HILDA.

Gilbert, Neil & Van Voorhis, Rebecca A 2003, 'The paradox of family policy', Social Science and Public Policy, September/October, pp. 51-6.

Gray, M, Qu, Lixia, Renda, Jennifer & de Vaus, David 2003, 'Changes in the labour force status of lone and couple Australian mothers, 1983-2002', Research paper no. 34, Australian Institute of Family Studies, Melbourne.

Gustafsson, S and Stafford, Frank P 1995, 'Links between early childhood programs and maternal employment in three countries', The future of children: long-term outcomes of early childhood programs, vol. 5, no. 3, pp. 161-74.

- 1994, 'Three regimes of child care: the United States, the Netherlands, and Sweden', in R Blank (ed.), Social protection versus economic flexibility: is there a trade-off?, University of Chicago Press, Chicago and London, pp. 333-61.

Horn, Stephen 2004, 'Guide to standard errors for cross section estimates', HILDA Project Technical Paper Series, no. 2, Melbourne Institute of Applied Economic and Social Research, University of Melbourne.

Martin, J 2002, 'Fertility in Australia: can we afford to keep the Government out of the bedroom?' Australian Social Policy 2001-02, FaCS, Canberra.

McDonald, Peter, 2001a, 'Work-family policies are the right approach to the prevention of very low fertility', People and Place, vol. 9, no. 3, pp. 17-27.

- 2001b, 'Family support policy in Australia: the need for a paradigm shift', People and Place, vol. 9, no. 2, pp. 14-20.

- 2000, 'Gender equity, social institutions and the future of fertility', Journal of Population Research, vol. 17, no 1, pp. 1-16.

- 1998, 'Contemporary fertility patterns in Australia: first data from the 1996 Census', People and Place, vol. 6, no. 1, pp. 1-17.

Sardon JP, 2002, 'The demographic situation of Europe and the developed countries overseas: an annual report', Population: English edition, vol. 57, no. 1, pp. 111-56.

Sleebos, JE 2003, 'Low fertility rates in OECD countries: facts and policy responses', OECD Social, Employment and Migration Working Papers, no. 15, OECD, Paris.

Gilbert, Neil & Van Voorhis, Rebecca A 2003, 'The paradox of family policy', Social Science and Public Policy, September/October, pp. 51-6.

Tesfaghiorghis, Habtemariam 2004a, 'Education, work, fertility and childcare in Australia: a HILDA based analysis', paper presented at the Australian Population Association 12th biennial conference on Population and society: issues, research and policy, 15-17 September 2004, Australian National University, Canberra.

- 2004b, 'Australia's fertility: a HILDA based analysis', unpublished manuscript.

Return to top

Book reviews

Addressing mental health problems as a strategy to promote employment: an overview of interventions and approaches