2.1. General Findings
There is no fixed or absolute cost of a child. The cost of raising a child increases with household income, although the rate of increase in the estimated cost generally declines with income level. Higher income households have greater living standards, which children share. Such households, for example, are more likely to pay for private education, buy brand name clothes and go on expensive holidays.
The cost of the first child is often found to be greater than that for each subsequent child. This is due to economies of scale resulting from hand-me-downs and shared infrastructure (such as bedrooms and furniture). When there is a large number of children in a household, there can be diseconomies in scale due to the extra costs of upgrading to large houses and cars (or second cars).
Estimates of the cost of a child generally tend to increase with the age of the child. The main exception is when all-day childcare is required for preschool age children to enable the parent(s) to undertake employment.
Depending on the sensitivity of the method used, the cost of a particular child varies according to the situations of the household, such as their geographical location and the working arrangements of the parent(s).
2.2. The Studies
During the last twenty years, there have been several studies into the costs of raising children in Australia. The major ones are Lovering (1984), Lee (1988), SPRC (1998), NATSEM (1999), Valenzuela (1999), Henman (2001) and AMP-NATSEM (2002). See AIFS (2000) for a useful overview and summary.
The results of the earlier studies by Lovering (1984) and Lee (1988) were regularly updated using changes in prices and wages data and published by the Australian Institute of Family Studies in their magazine Family Matters. Due to new costs of children research, the Institute ceased updating these costs in 1999.
The recent studies by SPRC (1998) and NATSEM (1999) have recalculated costs of children from first principles using more up-to-date data and more sophisticated methods. As such, they should be regarded as superseding the earlier studies. Henman (2001) further extended and updates the research of SPRC (1998) to address the costs of children more explicitly and enhance the sensitivity of the research to specific cases. The results reported in this report have been derived by updating and extending the results of Henman (2001) to the present.
2.3. The Different Approaches
Broadly, there are two different approaches to measuring costs of children.
- Expenditure or behavioural based estimates. This approach uses survey data on the expenditure of Australian households to create mathematical models that relate household expenditure, household type and living standard levels. It is the approach used by NATSEM (1999), AMP-NATSEM (2002), Valenzuela (1999) and Lee (1988), with the more recent studies using Australian Bureau of Statistics (ABS) survey data from 1993-94, with estimates updated. This approach provides estimates of the cost of children, or the proportion of household expenditure spent on children, for households at different income levels.
The key strength of this approach is that the costs of children estimates are based on what people actually spend. However, this is also a weakness, particularly with respect to low-income households, as it fails to obtain an estimate of what community standards regard as necessary to spend on children. More importantly, due to the nature of the base data, the approach is not particularly sensitive to different household configurations, such as geographical location, labour market status of adults and number and ages of children.
- Basket of goods or normative based estimates. Called the budget standards approach, this involves the identification of a basket of goods and services required by a particular household type to achieve a specified standard of living. The basket of goods and services are costed at current prices, with the cost of durable goods averaged over their lifetime (ie. similar to depreciation). Costs of children estimates using this approach measure what is needed to be spent on children to meet community standards, rather than what can be afforded.
This approach is used by Lovering (1984), albeit in a partial manner, SPRC (1998), Henman (2001) and here in this report. The Henman study is based on SPRC (1998) data, but provides greatly extended coverage of cost of children estimates updated to December 1998. Henman’s research also provides the only cost of children estimates sensitive to geography, providing estimates for each capital city.
A key strength of this approach is that it is sensitive to the circumstances and requirements of different household types, such as geographical location, the number of adults and their labour market status, the age and sex of the children, whether a child has a disability, and housing tenure. Because the estimates are based on a detailed list of goods and services, the assumptions are relatively transparent and therefore more readily open to debate and alteration. As the approach is normative, it also overcomes distortions in measuring the cost of children due to the income constraint in low-income households. A weakness of this approach is that the estimates technically relate to specific household types and are not immediately generalisable to provide an overall average. The approach does to some extent make assumptions about what should be consumed by households, rather than measuring what is actually consumed.
This report adopted the budget standards method for calculating the costs of children.