Wednesday 10 September 2003, 2pm, Quiet Room.
Family background, schooling and childlessness in Australia: what HILDA tells us
Nick Parr, Macquarie University.
This paper analyses the extent to which permanent childlessness among Australian women and men varies according to the size and type of family in which they were brought up, and the type of schooling they had. As in most other developed countries, the proportions of Australian women and men who have remained childless until their 40s and early 50s have increased considerably in recent years. Official estimates show the continuation of current first order birth rates would result in 24 per cent of women remaining permanently childless. While the effects on fertility or family formation of some of the educational and family background variables considered by this study have been analysed in other developed countries, such studies appear absent for Australia.
The data used are from wave one of the Household Income and Labour Dynamics in Australia (HILDA) survey, a large-scale, nationwide, longitudinal survey of the household population of Australia conducted in 2001 by the Australian Commonwealth Government's Department of Family and Community Services. A multi-stage, cluster sample design was used, and 13,969 men and women from 7,682 households and 488 census collection districts, which were stratified by state or territory, and metropolitan or non-metropolitan, were successfully interviewed. Data were collected on family formation and background, employment and unemployment history and status, and income.
The analysis is restricted to the 2,051 female and 1,879 male respondents who were aged between 40 and 54 years at the time of interview. Aspects of women's and men's education considered include the sector in which they were educated, and the grade at which they left school. Aspects of family background considered include:
-
the person's number of siblings;
-
whether the parents died, divorced or separated when the person was a child;
-
the countries in which the parents were born; and
-
the parents' occupations when the person was aged 14.
In addition the effects of a person's age, country of birth, and Aboriginal or Torres Strait Islander origin are controlled for. The pathways through which these early lifecourse variables affect whether a person stays childless are also considered. Multilevel logistic models are used in the analysis. Such models incorporating estimates of the between cluster variance of residuals offer improved estimation of the significance of coefficients and related goodness-of-fit statistics.
The results show that women who were educated in non-government schools are significantly more likely than women who were educated in government schools to be childless. Women who remained in school to year 12 are more likely than less educated women to remain childless. Having a smaller number of siblings is shown to be associated with a significantly increased risk of childlessness. Women who at age 14 had a father who was either dead or absent, and women who at age 14 had a father who was employed in a professional occupation, are more likely to remain childless.
The likelihood of a woman being childless has a significant inverse relationship with age between 40 and 54. There are also significant differences in rates of being childless by a woman's country of birth. Women born in southern or eastern Europe, the Middle East, south Asia, Africa, Latin America, or the Pacific Islands are significantly less likely than the Australia-born to be childless.
For men, as is the case for women, having a smaller number of siblings and having stayed in school until year 12 are significantly related to the likelihood of being childless. Men who when aged 14 had a father in a professional or white collar occupation have an increased likelihood of staying childless. The likelihood of a man being childless declines significantly over the 40 to 54 age range.
For most of the educational and family background explanatory variables the recent trend in Australia has been one of increases in the percentages of children in the groups which are more likely to remain childless. The proportion of full-time school students attending non-government schools rose from 28 per cent in 1992 to 32 per cent in 2002. Although rates of retention from year 7/8 to year 12 both for males and for females are slightly lower now than they were during the recession of the early 1990s, the rates for 2002 were more than double those for 1982. Between 1991 and 2001 the percentage of children under the age of 15 who are in lone parent families increased from 14 per cent to 20 per cent.
The proportion of employed males who are in professional occupations rose from just under 13 per cent in 1991 to 16 per cent in 2001. Moreover, trends in numbers of children ever born for women aged 40-44 suggest that although a slight increase since 1996 may have taken place, between 1981 and 1996 the average number of siblings their children had fell substantially.
The trend towards increased levels of childlessness may be viewed (positively) by some as symptomatic of a waning of the stigma that was attached to childlessness, of improvements in the effectiveness of contraception and access to abortion, and of women's enhanced opportunities to pursue education and careers. However, rising rates of childlessness may also be seen as symptomatic of women (and, to a lesser extent, of men) continuing to face work-family conflicts (McDonald 2000). With the leverage on fertility rates which is exerted by the childless component, the effect on the ageing of the population of rising levels of childlessness is of concern to some.
It should be noted that in the Australian context inference from trends in these early lifecourse variables to future levels of childlessness is complicated by the substantial migrant component of the population. In both the female and the male samples 29 per cent of the 40-54 age range had been born overseas. Even so, in view of their relationships to childlessness shown by this study, the trends in the early lifecourse variables may be of concern to those, such as Sex Discrimination Commissioner Pru Goward and economic demographer Peter McDonald, who have argued for the government to adopt a proactive stance to at least arrest or even reverse Australia's falling birth rate.
Email: nparr@efs.mq.edu.au
The decline of fertility in Scotland
Patrick Carroll, Pension and Population Research Institute (PAPRI)
Until the 1980s fertility in Scotland was higher than in England. But fertility in Scotland has since declined to a total fertility rate (TFR) of less than 1.5 whereas it is still over 1.6 in England. There is no official explanation of this difference. But it is assumed by the Government Actuary for purposes of making official population projections that fertility in Scotland will continue to be lower than in England for some time to come.
Relevant factors that might explain this change include social, cultural and economic variables. It is proposed to examine, in so far as available data allow, parallel trends in marriage rates, proportion married, births outside wedlock and male and female unemployment rates.
Less parity progression is observed when there is a higher proportion of extramarital births. The greater increase in male unemployment in Scotland since 1980 is a fundamental influence as is the improved educational and professional career opportunities for women in Scotland. Housing tenure in Scotland is somewhat different and this is considered relevant. Right to buy for tenants in social housing is identified as a contributing factor in fertility decline. It is suggested that cultural differences and a different medical tradition lead to different abortion rates and use of contraceptives in Scotland, and this is also to be considered.
Email: papriresearch@btconnect.com or pstcarroll@btconnect.com
What can we tell from women's fertility intentions?
Steve Smallwood, Office for National Statistics
Since 1979, information about the future childbearing intentions of women has been collected for all women of childbearing age in the General Household Survey. This presentation shows some of the findings based on this data published in Population Trends 112 by Smallwood and Jefferies, and presents further analyses by marital status and education.
Levels of intended fertility are shown to have fallen over time but not at the pace of actual fertility. Current intentions are higher than current and historical fertility rates suggest that women will achieve. The analysis shows that fertility intentions collected during the 1980s have not proved to be good predictors of subsequent fertility, either in level or trend. Alternative assumptions based on the certainty of women's answers are shown to demonstrate how intention levels would look under different scenarios.
When looking at data by partnership status some support is found for the aggregate accuracy of intentions of married women. Analysis of women by education shows lower intentions for women educated to tertiary level.
Email: steve.smallwood@ons.gov.uk