Family diversification in Australia--the increasing share of blended and step families.

Author:Forster-Jones, Joanna

Non-traditional couple families with children appear to be becoming more widespread in Australia. These include families with step children and families that consist of both step children and the couple's natural children (blended families). Measuring the prevalence of these non-traditional families is difficult. However a new family-blending variable in the 2006 census gives us a more accurate picture than before.


Australian families have undergone extensive change over the last century. Families are becoming smaller and family forms are becoming more diverse. Many factors have influenced changing household forms. Within families changing gender roles, social expectations and relationship patterns have resulted in major changes. We are experiencing what has been described as a 'mainstreaming of family forms that were once regarded as marginal'. (1)

The number of sole parent families has risen from 14 per cent of families with children under 15 in 1986 to 88 to 22 per cent in 2004 to 06. (2) The majority of these families are formed through relationship breakdown. Half of the 26,000 divorces in 2006 involved dependent children. (3) Breakdown of de facto partnerships, also an increasingly common family form, (4) adds considerably to the number of children affected.

Being a sole parent is often a transitional phase prior to forming a new relationship. As people re-partner following relationship breakdown, more complex family forms emerge. These may include children from one partner's or both partners' previous relationship(s), children born within the new relationship or a combination of child types. Families with natural children of both partners only are commonly referred to as 'intact families'. Those with step children only are 'step families', while those with both step and natural children are 'blended families'.

Estimates of the number of step and blended families have been available from previous censuses and from surveys. However for the 2006 census of population and housing additional coding was undertaken which was designed to better identify such families. This paper provides some background on previous estimates then outlines the way in which these family types were identified in the 2006 Census.

The paper examines the new family-blending variable available in the 2006 census. The count of step and blended families is disaggregated to provide previously unpublished figures of family-blending type by the sex of the step parent. The accuracy of both census counts and previous estimates is explored, highlighting strengths and weaknesses and exploring additional research opportunities available with the introduction of the new variable.


When analysing complex family relationships, surveys have the advantage of being able to include interrelated questions that fully explore the nature of connections within a family. They have the scope to include both current circumstances and previous relationship history, and are able to focus in depth on an area of interest--in this case, family form. In addition, as relatively few people are included, many surveys are conducted by an interviewer either by phone or face-to-face. This ensures greater consistency in the way responses are collected and coded.

However, few large surveys have included data on family blending. Most recent research has used data from the 1997 and 2003 surveys of family characteristics or waves of the Household Income and Labour Dynamics Australia (HILDA) survey.

The survey of family characteristics is conducted by the Australian Bureau of Statistics (ABS). It is a household-based survey, conducted as a supplement to the monthly population survey (MPS), which includes around 30,000 private dwellings. The survey of family characteristics is completed by households in the sample with at least one child under the age of 18. (5)

HILDA is a longitudinal panel survey, conducted in annual waves by the Melbourne Institute of Applied Economic and Social Research, under contract from the Australian Government Department of Family, Community and Indigenous Services (FaCSIA). The initial wave in 2001 included 7,682 households, re-interviewed in subsequent waves. (6)

To produce estimates for Australia as a whole from a survey sample, a complex weighting process is undertaken. Total population counts known as benchmarks are used to calculate weights for particular groups. The weights are then applied to the survey data to produce national-level estimates. Finally, standard errors are calculated to account for the degree to which the sample selected in the survey may differ from the full population.

For the 2003 survey of family characteristics, the ABS used counts at the person and household level, by state, part-of-state and household composition as benchmarks. (7) In this case, benchmarks at the family level were not available. The estimated resident population (ERP)--itself an estimate, as the name indicates--was used to provide benchmark totals. At the time of survey processing, the ERP calculated from the 2001 census was not yet available, so a 2003 ERP projected forward from 1996 census data was used. Household composition types used for benchmarking did not include a family-blending type breakdown.

Clearly, the estimation process will affect the accuracy of estimates produced for intact, step and blended families. In the case of the 2003 survey, the estimation process is far from ideal for producing accurate estimates of family-blending type.

Estimates from the survey of family characteristics are most commonly cited in Australian research into diverse family forms, and the 2003 estimates are shown in Table 1. The majority of couple families with children are intact (1,775,500 or 90 per cent). Step families comprise five per cent of all couple families with children, while blended families comprise four per cent. The remaining one per cent are couple families with children who do not have a parent-child relationship with either member of the couple. While step and blended families accounted for nine per cent of couple families with children under 18, they were home to over 382,000 children under 18, or over ten per cent of all children in couple families.

Many people are aware of the large number of sole-parent families in Australia--estimated at just over one in five families with children under 18 in the 2003 survey of family characteristics.

However, it is less commonly known that the same survey estimated that a further eight per cent of all families with...

To continue reading