Latest release

Overview of family, household and income unit standards

Family, Household and Income Unit Variables
Reference period
2014

Background

The Overview of Family, Household and Income Unit Standards explains the precise concept of the family that is used in the ABS as the basis for standards for statistics on the household and family, and defines the concepts that are used in the different standards. It contains a brief description of the standards themselves, and explains how operational procedures and standard question modules enable the collection of data which conform to the standards. It also contains a summary of the changes made between these standards and 'Standards for statistics on the family' released in 1995, and a glossary of terms containing definitions applicable to all the standards.

The ABS began publishing standards on aspects of the family and the household in Australia in 1993 and many were consolidated in Standards for Statistics on Family Variables, 1995 (cat. no. 1286.0). These standards were reviewed in 2005 and a revised suite was published under the same catalogue number. A minor review of the standards was undertaken in 2013.

The broad purpose of using these standards in ABS collections is to produce an accurate statistical picture of the structures of Australian families. They provide definitions of concepts, methods of data collection, derivation procedures and output formats for use in all relevant ABS and external statistical collections. Debates related to social policy and social support can then draw on consistent evidence about families taken from a range of sources. The standards are underpinned by a consistent definition of the family, which is sufficiently comprehensive to inform debate and sufficiently robust to be useful in a variety of statistical applications.

Key issues that can be examined by applying these standards include the diversity of families and the changing structure of households; relationship breakdown and family formation and dissolution; the extent of caring that takes place in families; the amount of unpaid household work; the combination of paid work and family responsibilities; the way in which men and women share family responsibilities; support from absent or extended family members; the needs of disadvantaged families; and the importance of secure and affordable housing for families.

Terminology used in these standards is defined in the Glossary.

Introduction

Introduction

The standards that form the basic suite of ABS variables on the family and the household are: 

  • Relationship in household.
  • Family composition.
  • Household composition.
  • Relationship between families.
  • Social marital status.
  • Registered marital status.
  • Income unit composition.
     

Summary of major differences between this and Standards for Statistics on Family Variables, 1995 (Cat. No. 1286.0)

Relationship in household

The standard recognised the two separate input classifications for 'Relationship in household', which are applied depending on the collection methodology used, in 2005. The standard emphasises the prime purpose of this variable as an input to family formation rather than as an output variable in its own right.

Family composition

The name of the variable was changed from 'Family type' to 'Family composition' in 2005. The classification was revised to reflect a more logical sequence of decision-making in family formation, however the resultant output remained fully compatible with 'Family type'. Appendix A of the 'Family composition' standard provides a full correspondence table between the two classifications.

Household composition

The name of the variable was changed from 'Household type' to 'Household composition'. The structure of the 'Household composition' classification was changed to allow integration with 'Family composition' as this better suits ABS output requirements. In particular users of data on families and households could now bring data together consistently, as the presence or absence of individuals unrelated to families in a household was no longer a core criterion at the base level of the classification.

Social marital status and Registered marital status

These standards underwent minor changes in 2005 to clarify the application of question modules and to maintain consistency with other standards in this suite.

Income unit composition

This standard was introduced in 2005 for use when a statistical unit is required analyse the levels and distribution of income. Income units are two or more people who pool their income and savings, or an individual with sole control over their income, consumption and savings. They are formed in similar fashion to families within households using 'Relationship in household' data. They have many characteristics in common with families.

Underlying concepts

Concept of family

The concept of the family is central to ABS Family standards. In devising its definition of the family, the ABS has recognised that notions of what constitutes a family vary considerably. Some people consider their family to be the relatives with whom they live while others extend the definition of family to include relatives who live in other dwellings. For some, the notion of family includes people who are unrelated.

In the statistical context, it is necessary to have a clear and comprehensive definition of a family that considers potential analytical uses of the statistics as well as the practicalities of data collection. This inherently requires narrowing the definition of the family unit and restricting who is considered a family member.

There are two main contexts in which the term 'family' is likely to be used as a data concept. The first is providing data about the extent to which people may provide support and assistance to their relatives. In this context, the definition of family needs to be restricted only by a specification of the types of relationships which apply. In the second context, which this suite of standards mainly concerns, the label 'family' is given to the key statistical unit used in the analysis of data about the characteristics and circumstances of families. For this purpose a family is constrained to people who live together in a single household. Without this constraint it would not be possible in any practical way to place a boundary around the statistical unit 'family'.

Often the concepts of family and household refer to the same set of people when applied to a particular dwelling. This is because the family is a subset of the household by definition and, in Australian society, a household frequently comprises a single family. The family and the household are however two distinct concepts and do not comprise identical populations. People who live alone, live in group households, or share a household with a family to which they are unrelated are, according to the ABS' statistical definitions, members of households but not members of families. Furthermore, a household may be comprised of two (or more) families.

The ABS defines household as:

  • 'One or more persons, at least one of whom is at least 15 years of age, usually resident in the same private dwelling'.

This definition is similar to the System of National Accounts (SNA) definition which is:

  • 'A small group of persons who share the same living accommodation, who pool some, or all, of their income and wealth and who consume certain types of goods and services collectively, mainly housing and food.'

The ABS definition varies from that of the SNA only in specifically allowing lone person households, and in removing any reference to consumption.

For statistical purposes family is defined as:

  • 'Two or more persons, one of whom is at least 15 years of age, who are related by blood, marriage (registered or de facto), adoption, step or fostering, and who are usually resident in the same household. The basis of a family is formed by identifying the presence of a couple relationship, lone parent-child relationship or other blood relationship. Some households will, therefore, contain more than one family.'

Although the majority of households in Australia are one family households, as the basis of a family is formed by identifying the presence of a couple relationship, lone parent-child relationship or other relationship, some households will contain more than one family. This definition of family has therefore been written to include households which contain more than one family. For more information on how this definition is used to form families see the 'Relationship in household' and 'Family composition' standards.

This definition of family also forms the basis for defining Income units. Income units are formed either by families or by individuals not in couple or parent/dependent child relationships within a household. Income units differ from families in that related, non-dependent individuals will form separate income units rather than being attached to the family nucleus. For more information see the 'Income unit composition' standard.

Because ABS surveys only collect data from dwellings where at least one person aged 15 years or older can be identified, the definition of a family applies an age limit of 15 years and over to at least one member of the family. The definition also restricts the concept of a family to those usually resident in the same household because in most ABS social surveys the household is the unit selected in the survey. A concept of family which extended beyond the household would allow some individuals to be included in more than one family. In addition to leading to double counting of particular individuals in statistical collections, failure to apply an explicit boundary to the concept would make it difficult for the ABS to measure the number and characteristics of families consistently.

Although the definition of a family is constrained to a household for statistical and classificatory reasons, the ABS does produce statistics about wider family network (e.g. in the General Social Survey, and special purpose statistical collections such as the Family Characteristics Survey). Many aspects of family life are not confined to those who live as part of one household. A major emphasis of the Family Characteristics Survey is the ways in which members of family networks, who live in different households, give and receive support. This makes it possible to examine areas of support which are applicable to both the household family and the extended family network.

The household family, as described in this document and related standards, is the standard for all ABS social surveys. Thus, for the purposes of ABS statistics, a person is not considered a member of a particular family if he or she usually lives in another household, or is an unrelated individual over 15 years of age living in the same household (e.g. friend, boarder, housekeeper). However, unrelated individuals under 15 years of age living in the same household are treated as family members.

Non-family members over 15 years of age living in a family household (such as boarders) are classified as part of a family household for the purposes of 'Household composition', but are not classified as part of the family for 'Family composition' coding. For further information see the 'Household composition' standard.

The ABS aims to provide an accurate statistical picture of Australian society to be used as the basis for informed decision making and therefore derives a 'Social marital status' classification of 'Married in a de facto marriage, same-sex couple'. However, counts for this classification are usually too low to be output separately as doing so may breach ABS confidentiality provisions. Such data is therefore output in the 'Married in a de facto marriage' category accompanied by a footnote stating that the category includes partners in same-sex de facto relationships.

Core variables describing family structure

The four main variables used to describe family structure within a household are 'Relationship in household', 'Family composition', 'Household composition', and 'Social marital status'. The standards for each of these variables provide comprehensive information on the conceptual and practical issues relating to data collection and processing. Each standard includes a description of the name of the concept, its definitions, the classification criteria, the classification and code structures, a discussion of conceptual issues, standard questionnaire modules, and output classifications for the presentation of data.

Relationship in household

'Relationship in household' is a characteristic of each individual living in a household. It is used to describe the type of familial relationship (if any) an individual has with other individuals in the household. This concept is central to the application of statistical standards to the family and the household because relationships within a household provide the key for identifying the type and number of families that are in the household.

Family composition

'Family composition' differentiates families based on the presence or absence of different relationships, including couple, parent-child, child dependency or other relationships. The 'Family composition' standard includes a hierarchical 'Family composition' classification with four levels. At the highest level of the classification the four family types are:

  • Couple family without children.
  • Couple family with children.
  • One parent family.
  • Other family.

A second level of the classification introduces the concept of dependent children aged under 15 to provide more detail of 'Family composition' within couple and one parent families. A further level distinguishes families with dependent students aged 15-24 years from those with other dependent children. At the most detailed level the presence of non-dependent children is identified in each category of couple and one parent families. The separate identification of opposite-sex and same-sex couples within couple families is also possible when required.

The 'Family composition' classification does not identify family members outside the family nucleus. For example a family may contain a couple and their dependent children, but also the parent of one of the couple (a 'Lone ancestor'). Identification of such people within a family is done using 'Relationship in household' data.

Household composition

'Household composition' is used to identify family households, the number of families in a household, the presence of non-family members in family households, and the type of non-family households. Non-family households comprise lone person and group households. 'Household composition' can be combined with 'Family composition' in statistical output to provide information on families within households. A standard framework for publishing this information is provided in the 'Household composition' and 'Family composition' standards.

Marital Status

There are two distinct standard variables for Marital status: 'Registered marital status' and 'Social marital status', with the latter recommended for most applications. This distinction is in accordance with the recommendations of the United Nations Economic Commission for Europe (UNECE). (Conference of European Statisticians Recommendations for the 2010 Censuses of Population and Housing, United Nations, 2006.)

'Registered marital status' (which categorises people as Never married, Widowed, Divorced, Separated and Married) was once used in most applications. However, the increase in number of de facto marriages has made the concept less useful for identifying couple relationships. As 'Social marital status' provides a category for de facto marriages, and classifies people according to their usual living arrangements rather than their registered marital status, this concept can identify couple relationships within a household by using one of the standard sets of questions that identify relationships.

A person's 'Social marital status' is determined by whether or not they form a couple relationship with another person and the nature of that couple relationship, regardless of their 'Registered marital status'. The classification structure has, at its highest level, the categories married and not married. Within married are separate categories for registered and de facto marriage. Within de facto marriage, at the next level of detail, there is a distinction between opposite-sex and same-sex couples. These are further divided at the next level of detail into male and female same-sex couples. Although the classification contains this detail, it is unlikely that information will be published at this level in many statistical collections.

These core variables are all designed to produce and present information about families and households on a consistent basis. They are therefore based on a common set of underlying concepts and rely on a common approach to data collection. In particular, data on 'Relationship in household', 'Family composition', 'Household composition' and 'Social marital status' are obtained from a common set of relationship questionnaire modules. These modules are presented in full in the 'Relationship in household' standard.

Definitions of common concepts

A number of concepts are common to most of the variables overviewed here. In addition to the family, which is discussed above, these concepts include:

  • Household.
  • Household and family reference persons.
  • Usual residence.
  • Couple relationship.
  • Child.
  • Child dependency.
  • Nominal child and nominal parent.
  • Income unit.

Household

The concept of a household underpins the collection and dissemination of statistics on families and households. In the ABS a household is operationally defined as:

  • 'One or more persons, at least one of whom is at least 15 years of age, usually resident in the same private dwelling.'

Thus all private dwellings in Australia by definition contain only one household. This definition of household aligns with current ABS practice and makes no significant difference to existing time series data.

Household and family reference persons

The concepts of household and family facilitate the identification of a household reference person and family reference person(s) who are used as the basis for determining the relationships between the usual residents of the household.

The household reference person is a household member whose relationship with all other members of the household may be described in terms which identify the composition of the household and are meaningful to family formation. The relationship between each individual and the household reference person can identify relationships between other members of the household and further clarify household and family composition.

A family reference person is a household member who forms a family with other members of the household and they are used to determine the relationships between other family members. As such, the person who is identified as the family reference person needs to be the person to whom relationships with all other family members can best be described. In households which contain more than one family, a family reference person must be identified for each family. This allows each family living in the household to be treated as a separate entity.

As a family reference person must be identified for each family in a multifamily household, a separate variable must be associated with each person in the household indicating to which family that person belongs. This variable is referred to as the family number. See 'The family number' under 'Coding Procedures' in the 'Relationship in household' standard.

Although the concepts of household and family reference persons are primarily used to code and process data, they can also form output variables. This occurs when characteristics of the reference person are used as indicators of characteristics of the household. For example, the occupation or qualifications of the reference person may be used as an indicator of the socio-economic status of the family or household. Accordingly, it is necessary to have a set of criteria for determining the reference person who is most appropriate for such purposes.

In a single family household, the household reference person and the family reference person are the same person. A standard hierarchical set of criteria is used to determine the most appropriate household member to be the household reference person. These criteria are explained in the 'Relationship in household' standard, 'Appendix A - Identifying household and family reference persons'.

In multifamily households, where family coding is less straightforward, a more detailed procedure is needed to determine the appropriate reference persons. Once the household members have been allocated to families, the responses provided on the collection form are further processed to determine a reference person for each family. The rules for determining families in multifamily households are explained in the 'Relationship in household' standard, 'Appendix A - Identifying household and family reference persons'.

Usual residence

The concept of usual residence is used to constrain the description of relationships, family type, household type and social marital status to people who usually reside with a particular household. People who are temporarily residing with a household during the survey period and are usual residents of another household are considered to be visitors.

The concept of usual residence is based on the fact that each person has a basic attachment to a particular dwelling. However, this concept embodies two forms of attachment. The first is attachment to the dwelling in which a person lives the majority of the time, known as 'Usual residence in a dwelling'. The second is attachment to the dwelling which a person considers to be their family home, known as 'Usual residence in a household or family home'.

The first concept is used for producing estimates of the usual resident population by geographic area, which the Australian Statistician is under a legislative obligation to compile. These estimates are used as the basis for allocating resources. In particular, population estimates based on the Census are used to determine the number of seats allocated to each State and Territory in the House of Representatives and also for the allocation of Financial Assistance Grants by the Australian Government to the States and Territories.

The second concept is used in the Monthly Population Survey (MPS) and most other ABS household surveys. It is used for the collection and output of data for units of analysis such as households, families and income units. In operational terms, it is the dwelling a person perceives to be their home.

Consequently the ABS has two definitions of usual residence, as follows:

  • The dwelling (address) at which a person has lived or intends to live for a total of six months or more in a calendar year.
  • The dwelling (address) that the person regards as their 'home'.

A more detailed explanation of the concept of usual residence and the definitions and applications of the concept can be found in Usual Residence Concepts Sources and Methods Paper, Jan 2004 (cat. no. 1389.0).

Couple relationship

The concept of couple relationship is common to all family related variables. It exists when:

  • two people are usually resident in the same household
  • a social, economic and emotional bonding, usually associated with marriage, exists between the partners and
  • the partners consider their relationship to be a marriage or marriage-like union.

In most statistical collections it is not practical to ask questions which will determine whether bonding usually associated with marriage exists between two people. Accordingly, two individuals are regarded as a couple if both usually live in the same household in either a de facto or registered marriage. The formalisation of these living arrangements through a ceremony is not necessary.

In practice, a couple relationship exists when the relationship between two people usually resident in the same household is reported as: husband, wife, spouse, partner, de facto, common law husband/wife/spouse, lover, boyfriend. Any relationship label which indicates that a couple relationship exists should be accepted, unless one of two circumstances applies. One, if the relationship is further qualified by the respondent to indicate that they do not consider that a couple relationship has been formed. For example, if the respondent indicates that a person is their boyfriend and then goes on to say ''but, we aren't de factos'', the individuals are treated as unrelated. Two, a relationship between two females described as girlfriend is not regarded as evidence of a couple relationship unless one is specifically indicated by the respondent. The word friend on its own is not taken to be sufficient evidence that a couple relationship exists. Thus, the definition of a couple relationship is:

  • A couple relationship is defined as two people usually residing in the same household who share a social, economic and emotional bond usually associated with marriage and who consider their relationship to be a marriage or marriage-like union. This relationship is identified by the presence of a registered marriage or de facto marriage.

The restriction of couple relationships to usual residents in a household is a necessary practical consideration for conducting household-based surveys. Therefore, a marriage or partnership between a person usually resident in a household and a person who is a usual resident of another household (i.e. a visitor) is not recognised in the 'Relationship in household', 'Family composition', 'Household composition' or 'Social marital status' standards. However, a registered marriage between a person who is usually resident in a household and a person who is a usual resident of another household is recognised in the 'Registered marital status' standard.

Child

In these standards the definition of a child is primarily a function of an individual's relationship to other household members, regardless of age. However, all people under 15 years of age are defined as children. The identification and classification of children into different types is based on reported relationships and the presence of dependency relationships within the family. It is through the dependency relationships that the definition of a child is broadened beyond natural, adopted, step or foster son or daughter to include otherwise related and unrelated individuals aged under 15.

The definition of a child is:

  • A child is a person of any age who is a natural, adopted, step, or foster son or daughter of a couple or lone parent, usually resident in the same household. A child is also any individual under 15, usually resident in the household, who forms a parent-child relationship with another member of the household. This includes otherwise related children under 15 and unrelated children under 15.

To be classified as a child, the person can have no partner or child of their own usually resident in the household.

Three types of child are identified in the 'Relationship in household' classification:

  • Dependent child under 15.
  • Dependent student.
  • Non-dependent child.

The differentiation of children into these three types is based upon the dependency criterion and is designed to identify families with different structures and needs. Dependency as used in these standards refers to economic dependency and is applied only to the population of people who could be described as 'children'. It is therefore not intended to measure an aged or disabled person's dependency.

Dependent child

Dependency is assumed to exist when an individual living in a family household is likely to be unable to support themselves financially and is therefore reliant on another usually resident individual(s) for the provision of their financial needs (meals, accommodation, other expenses). Two barriers to full-time employment are used as indicators of economic dependency: age and student status.

The dependency criterion is applied to children under 15 and dependent students in slightly different ways. The reason for applying it to children under the age of 15 is that people of this age are not legally able to work full-time and thus cannot support themselves. This criterion applies not only to natural children but also to nephews, nieces, cousins and unrelated children who live with a family.

The concept of dependency is also applied to another group of people: those 15-24 years old and studying full-time (called 'Dependent students'). They are also assumed to be unable to support themselves financially as, by virtue of their student status, they are generally unable to work enough hours to do so. In this instance, the dependency criterion is not as widely applied as it is to children under 15: unrelated people who live with a family, and otherwise related family members such as nieces, nephews, cousins etc. who are studying full-time are assumed to receive support from outside the household. Therefore only natural, adopted, step, and foster children are defined as dependent students.

The ABS has extensively reviewed the concept of child dependency to determine the most appropriate and operationally feasible method of identifying dependency. As the concept uses economic rather than social, legal, psychological, physical or any other form of dependency, it was acknowledged that an actual measure of economic activity would be more accurate than the current student status indicator. However, practical collection considerations and the necessity to retain continuity in time series led the ABS to conclude that child dependency should continue to be measured by age and student status.

As children over the age of 15 who are not studying full-time can be employed full-time, they are classified as non-dependent children.

Nominal child and nominal parent

The composition of many Australian households and the relationships between their residents are more diverse than those generally regarded as being traditional 'nuclear' families. Often the relationship between adults in the household and dependent children (people aged under 15 or full-time students aged under 25) would not be captured by basic coding methods. This is explained in the 'Relationship in household' standard. To better reflect the nature of those relationships where to all intents and purposes a 'parent-child' relationship exists, and thereby facilitate a more accurate and analytically useful picture of the 'family-type arrangements' that may exist in those households, the ABS uses the concepts of the 'nominal child' and 'nominal parent'.

When a child under 15 is a usual resident of a dwelling where their parent is not a usual resident, that child must be allocated a 'nominal parent'. They thereby become a 'nominal child' who is subsequently treated, particularly for output purposes, as identical to a natural, adopted, step or foster child. Full-time students aged 15-24 years are never allocated nominal parents. Consequently if that student is a usual resident of a dwelling where their parent is not a usual resident, the student cannot be defined as 'dependent'. The student becomes a related (or unrelated) person in that household.

Income unit

The economic wellbeing of a person is not always readily apparent from their individual income e.g. a dependent child. However, the family as defined in these standards will in some cases be too broad to use as a basic statistical unit for analysis of income levels and distribution. Therefore the ABS assumes that there are groups of related people who, by virtue of their relationships, systematically pool all their income and savings and share equitably in the benefits derived from access to these financial resources. Such groups of related people are known as the income unit.

At the basis of the income unit is a relationship between people which is identified by their 'Relationship in household'. Unrelated people do not form income units together but are considered to be single people with sole control over their income, consumption and savings. In a given household an income unit may therefore be an individual, every person in a family or household, or a group somewhere in between. Income units consist in the first instance of people in a couple relationship and any dependent children. Lone parent income units are then identified. Any other person aged 15 or over who is not identified as a dependent student will form a separate one person income unit. Finally unattached children under 15 are allocated nominal parents.

Collection methods

Procedures for processing family variables

The following steps need to take place to process data about families: 

  • Identify the usual residents of the household.
  • Describe the relationships between the usual residents using an initial household reference person.
  • Apply the rules for forming families (see the rules for identifying and coding families in the 'Family composition' standard). At this stage it may also be necessary to reallocate the household and family reference person(s) depending on the complexity of the household (for further details see Appendix A of the 'Relationship in household' standard).
  • Code 'Relationship in household', 'Family composition' and other variables as required.

Collection methods

While information on families can be collected using different methodologies it always follows the same sequence.

Initially, a 'Household reference person' is identified. This person is the first named person entered on the collection form and is called 'Person 1'. They are the reference point to which the 'Relationship in household' is anchored. 'Relationship in household' information provides information on 'Social marital status'. Once relationships within the household have been established and coded, reference is made to the ages of children and further questions about student status are asked to determine whether children aged 15-24 years are dependent or not. Next, rules are applied to the relationship and dependency information to establish the 'Family composition'. 'Family composition' data can then be used to determine the 'Household composition'. If required, additional questions can be asked to determine 'Registered marital status'.

In practice, identification of the 'Household reference person' is either a correction of the identification made in the field, which occurs later during the coding process, or more usually a result of the appropriate person being selected automatically. Automatic selection of 'Person 1' can occur in two ways, depending on the method of collection:

  • In self-completed collections such as the Census, the instruction on the form is to "Enter the householder or any adult household member as 'Person 1', and if present, the spouse or partner as 'Person 2'.
  • In interviewer-based collections, the interviewer instruction is to ask "What are the names of all the people who usually live here?". As the respondent will usually list themselves first, they become 'Person 1' and thus the household reference person. Note that the ABS Interviewers Manual allows the respondent to nominate a different household member as Person 1 and the interviewer is advised to accept this provided that person is a suitable household reference person. For example, a person aged under 15 years is not suitable.

Standard question module

For the 'Relationship in household' question modules see the 'Standard question modules' section in the 'Relationship in household' standard.