4160.0 - Measuring Wellbeing: Frameworks for Australian Social Statistics, 2001
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 12/10/2001
Page tools: Print All

Frameworks

COUNTS, ESTIMATES AND PROJECTIONS

Organising a complete count of all members of a population at a given point in time is an essential step to accurately measuring its size. This is one of the functions of conducting a census. Modern Australian population censuses, conducted at five yearly intervals, also provide a wide range of demographic, social and economic data about the entire population and about populations living in separate areas within Australia down to fine levels of geographic detail. It is from this base, together with information from other sources which measure changes in population size, that the ABS produces official population estimates relating to other points in time and so provides a regular stream of up-to-date statistical information about the overall size, structure and distribution of the population. In terms of geographic coverage these estimates relate to the areas administered by the three levels of government: Commonwealth, State and Local. They include annual (mid-year and end of quarter) estimates of the population of Australia and each of the States and Territories, and annual mid-year (i.e. 30 June) estimates of the population of areas that correspond with Local Government Areas (namely Statistical Local Areas). It is from the basis of these population estimates that projections of the population are produced. Those published include projections for Australia, the States and Territories and capital city/balance of state areas. The ABS also produces, with users, projections for smaller geographical areas.

THE DEMOGRAPHIC MODEL

A simple yet practical stock/flow model (sometimes referred to as a balancing equation) serves to provide the conceptual foundation for much of the work involved in producing official population estimates and projections. The balancing equation, shown below, informs that after an initial estimate of a population has been made, a subsequent estimate of the size of the population can be provided by using information about numbers of all people gained and lost from the population as a result of births, deaths and movements of people, as immigrants and emigrants, to and from other areas.

 P1 = P 0 + B - D + I - E Where P0 = Population at the beginning P1= Population at the end of a period B = number of births during period D = number of deaths during period I = number of immigrants who arrived during period, and E = number of emigrants who left during the period.

The terms immigrants and emigrants usually refer to people who move to and from a country and so are most apt in the context of producing national level population estimates. However, they are also used in the balancing equation in a conceptual sense to refer to people who move to and from any area that is the focus of interest. Thus for a given area within Australia, such as a State or Territory, they include people moving to and from other parts of the country. As such the model takes into account both international and internal migration as components of population growth.

DEMOGRAPHIC TECHNIQUES

Clearly where regular high quality information is available about each of the components of population growth (namely births, deaths and migrations) there is a basis for continuously updating a census count for points in time up until a new set of census data becomes available. Furthermore, by studying the various components of growth, in terms of their rate of change over time (i.e. trends in fertility, mortality, overseas and internal migration), how those trends differ among population sub-groups and the factors that affect those trends, it becomes possible to propose how those components might change into the future and so produce scientifically based population projections. This science, known as demography, provides the logic for producing population estimates and projections for particular geographic areas. It also provides an array of concepts and methods such as cohort analysis, life tables, and hazard models, that support the analysis of demographic trends.

A particular technique that builds on the foundation of the balancing equation but simultaneously uses information about the sex and age of the people involved in each element of the equation, known as the cohort component method of population estimation, is used extensively to produce population estimates and projections. However, various other techniques are also used, especially in instances where the information needed to support the cohort component method is not available or of suitable quality. For example, information on internal migration within States is often not directly available.

COUNTING RULES

There are always various governing rules that determine who should be included or excluded from any population count. These rules typically involve criteria on who belongs to a population in terms of their residency status (whether they usually live within an area or not) but may also include other criteria based on an individual's personal characteristics. For instance, by international convention, foreign diplomats and their families are always excluded from census counts due to their foreign allegiance.

Currently, in Australia, there are no other criteria based on a person's characteristics, such as citizenship or race, to exclude them from census counts or official population estimates. This has not always been the case. Until 1967, section 127 of the Constitution required the exclusion of some Aboriginal people - 'In reckoning the numbers of the people of the Commonwealth or of a State or other part of the Commonwealth, Aboriginal natives shall not be counted'. This was interpreted as requiring the exclusion of people with more than 50% Aboriginal blood. To enable this exclusion, census questions were used to identify 'half-castes' and 'full-bloods' and the census was not undertaken in the remote areas that were only inhabited by Aboriginal people.

PLACE OF ENUMERATION COUNTS

It is for practical reasons that population censuses taken in Australia count people according to their actual place of residence on the night of the census. This maximises the count ensuring that everyone will be counted somewhere. However, it is known that the counts for particular areas can be substantially affected by this approach. Holiday areas, along the coast or around the ski slopes, are examples of areas which can have large seasonal differences in their population associated with short term flocking of visitors, including overseas visitors, to and from those areas.

A consequence of such movements is that the size of a population of a given area can be influenced by the time of year the count is undertaken. However, in order to increase the likelihood that people are counted at their usual pace of residence, recent censuses (conducted on a Tuesday night in August since 1991) have been scheduled to avoid conflict with school holidays and other periods with high levels of population mobility.

COUNTS OF USUAL RESIDENTS

The disadvantages of place of enumeration counts for any area can be overcome by providing counts of people who usually live within an area. It is on this basis that official population estimates are provided and (notwithstanding some differences in operational definitions) on which information from all other ABS social surveys are produced.

Producing such counts from the census, which provides the starting point for producing official population estimates, is enabled by asking people to give their usual place of residence on the census form and then coding them accordingly. The criteria used to determine a usual place of residence in the census is the address at which the person has lived, or intends to live, for a total of six months or more in the census year. If a person is not a usual resident of the place at which they were counted on census night they are deemed to be a visitor, and, if their usual address was in Australia their address information is used to determine the area (i.e. the Statistical Local Area (SLA) and State or Territory) in which they usually live. Information about the population that usually lives within a given area can then be generated. Overseas visitors are excluded from such counts.

The steps just described are those used to produce census based profiles of usual residents living in various geographic areas. However, because the census does not obtain information about usual residents who were temporarily overseas at the time of the census, additional information is needed to obtain complete estimates of usual residents. In practice this data is obtained from international movement statistics and refers to those Australian residents who were overseas on census night and returned in the twelve month period subsequent to the census date.

SERVICE POPULATION ESTIMATES

The balance of usefulness between the counts just described is not entirely in the direction of a usual resident count. A measure of the number of people in an area at a particular time may provide an indication of the pressure on services and infrastructure as well as on the demand for accommodation and related tourism and travel needs. The usefulness of place of enumeration counts from the census depends on the coincidence of the census date with the peak service demand in the area of interest. The census data may not be optimal in this regard as the date of the census is chosen to minimise the effect of population movement. Usual resident counts are, therefore, often taken as a starting point for obtaining service population estimates.

Interest in having population estimates that reflect peak demand for services arises from many different service providers. These include those concerned with tourism, water supply, sport and recreation, traffic volume management, bush fire prevention, policing, and so on. Clearly the specific demands for where and when people should be counted can vary with each service. Thus there can be interest in estimates compiled according to where people spend their time during the day (at work or undertaking other activities) as opposed to where they live, and population estimates that relate to particular days of the week or particular times in the year. The population census collects information about the location of people's places of work which, for the areas in which such information has been coded, has helped to meet some of these needs.

ESTIMATED RESIDENT POPULATIONS (ERPs)

Producing official population estimates involves a number of steps in addition to those already described of converting the five yearly census counts from a place of enumeration to a place of usual residence basis. These include adjusting the census counts for census under-enumeration, and, where the census date falls on a date other than 30 June, making an adjustment so that the estimate applies to 30 June. Post-censal estimates for Australia and the States and Territories are then calculated by the cohort component method, using the 30 June census year estimate and flow data on births, deaths and overseas and interstate migration. The estimates given by these procedures are formally known as 'Estimated Resident Populations' (ERPs).

The procedures for producing post-censal ERPs differ according to geographic level because they depend on the availability of suitable data. For instance, in regard to migration at the national level, only data on international migration is needed while for lower levels of geography additional data on internal migration is also needed. The data on international migration is obtained from records of international movements of persons which distinguish between permanent, long term and short term arrivals and departures to and from each State and Territory. At the State/Territory level, internal migration data is primarily modelled from Medicare records. Again, at the Statistical Local Area (SLA) level there is no direct data available about overseas or internal migration so various estimation techniques are used.

POST CENSAL AUSTRALIA/STATE ESTIMATES

To estimate the Australian population at 30 June one year after a census year, all births occurring during the preceding 12 months are added to the census year ERP, all deaths are deducted; the people leaving the country both long term (12 months or more) and permanently are also deducted; and the people arriving both long term and permanently in the country from overseas are added. Short term movements (arrivals and departures) for less than 12 months are excluded from population estimates. Taking into account interstate migration similar procedures are used at the State/Territory level and the procedures are repeated for each subsequent year.

POST CENSAL SLA ESTIMATES

For post-censal years, the absence of migration data at the SLA level means that it is not possible to estimate SLA populations directly by taking into account natural increase and net migration. Instead, ERPs are calculated using a mathematical model. Local knowledge, including that advised by local governments, may be used to adjust the outcome of the model for a particular SLA. In the mathematical model a relationship is established between changes in population and changes in other indicators over the period between the two most recent censuses.

The choice of indicators varies across the States and Territories, depending on availability, and includes dwelling approvals, electricity connections, Medicare enrolments and drivers licenses. Changes in these indicators are then used to estimate changes in the population of each area since the last census. The choice of indicators also varies across SLAs depending on aspects such as whether the SLA is urban or rural, its population is growing or declining, and whether the area has a high or low proportion of houses or medium and high density dwellings.

POPULATION PROJECTIONS

Population projections aim to show how the size, structure and distribution of the population may change into the future. The ABS publishes a new set of integrated population projections, relating to Australia, the States and Territories, and capital city/balance of state areas every two to three years. These projections involve a number of scenarios about likely outcomes with each scenario being based on a different combination of assumptions about future changes in fertility, mortality, and overseas and internal migration. The assumptions are based on a detailed analysis of recent past experience at the three geographic levels and takes into account likely trends observed from the experience of other countries and the views of relevant government agencies and experts.

As with population estimates the projections for these areas are produced using the cohort-component method which begins with a base population for each sex by single years of age and advances it year by year by applying assumptions about future mortality and migration. Assumed age-specific fertility rates are applied to the female populations of child-bearing ages to provide the new cohort of births. This procedure is repeated for each year in the projection period for each State and Territory and for Australia. It is also repeated to obtain capital city/balance of State projections for each State and Territory. The resulting population projections for each year for the States and Territories, by sex and single years of age are adjusted to sum to the Australian results.

Of course the number of assumptions that might be made about the future are numerous. Reference to a recent set of ABS projections, those produced on the basis of 1999 population estimates showing population change to the year 2101 for Australia and 2051 for the States and Territories, helps to show how these are managed. At the Australian level, two assumptions about fertility are used (high and low), one about mortality, and three about net overseas migration (high, medium and low). Within each of these six combinations, three further State/Territory assumptions about the volume of internal migration are made (large, medium and small net gains and losses for States/Territories). The combinations of these assumptions give rise to eighteen possible outcomes. While summary information is given for all eighteen, only three outcomes are selected for detailed analysis. These are referred to as being high, medium and low variants. Each of the series is available by sex and age group. Derived ‘indicator’ projections such as the median age of the population, the proportion aged 65 years and over, and the proportion under 15 years of age, are also provided for selected years.

POPULATION FORECASTS

According to their particular administrative and planning needs, various government agencies also produce population projections of their own, sometimes using the ABS to prepare them. Often referred to as forecasts, rather than projections, because they aim to predict populations by only using one set of assumptions and more commonly have a short term outlook, these include national level forecasts of the total population produced by various commonwealth government agencies such as the Treasury and the Department of Immigration and Multicultural Affairs, and local area forecasts produced by various State and Territory government planning agencies. These typically use similiar projection techniques to those described above and are usually based on ABS population estimates.

SMALL AREA POPULATION PROJECTIONS

ABS small area population projections or forecasts (for local government areas , postcode areas and the like) that are produced for users with their invovelment in making assumptions about change, typically draw on knowledge of local area planning decisions (sometimes indicated by data on building approvals or expected releases of land) to help predict likely future changes. State government agencies with access to development plans are clearly best placed to contribute their knowledge about possible future impacts in particular areas and so often produce their own small area projections. The need for local area knowledge reflects the fact that local planning decisions can have a greater influence on population growth than trends in mortality, fertility and migration on which projections for broader geographical areas are based.

DEFINING GEOGRAPHICAL AREAS

The Australian Standard Geographical Classification (ASGC) has been developed by the ABS to identify different geographic areas for which statistical information may be provided to suit many different needs. The classification actually embodies a number of separate, but often inter-related, classification structures all built up from Census Collection Districts (CDs), which are the smallest areas for which data are made available from the census.

The areas have been defined according to a number of criteria which are different for each unit in the classification. These criteria variously include; the official boundaries of local and State and Territory governments; streets, rivers and visible landmarks along with numbers of dwellings within an area (usually about 200 in urban areas) to define CDs; the population density of CDs and their location relative to similar CDs to define urban centres, rural localities, and remaining rural balance areas; road distance from major service centres in order to define remote and less remote areas; and so on. More information on the classification can be obtained from the Australian Standard Geographical Classification (Cat. no. 1216.0) and Statistical Geography Vol. 2: Census Geographic Areas (Cat. no. 2905.0).

An aspect of any geographic classification that is of importance to users is its flexibility for adaptation to a wide range of different purposes. This flexibility is closely related to the size of the smallest available building block; the smaller the building block the more flexible the combinations. Currently the smallest unit, which can be aggregated in various ways to define larger areas of interest, is the CD. However, opportunities are emerging through geocoding and Geographic Information System technologies to identify the precise location of dwellings, or other statistical units, and so offer much greater flexibility in combining information to suit various needs.

DESCRIBING MULTICULTURAL AUSTRALIA

An important dimension of the composition of a population, often taken to belong to the field of population statistics, but which is of interest to each of the other areas of concern described in this book, relates to the diverse ethnic and cultural backgrounds of the population. It is due to its association with immigration, as one of the major components of population growth for much of Australia's modern history, that the link between cultural diversity and the other topics discussed in this chapter arises. Another link relates to the task of providing post-censal estimates and projections for groups of interest (such as the number of people in Australia who were born in different countries) as this also involves the use of demographic techniques similar to those used to provide population estimates and projections for the total population.

Cultural diversity can be recognised through a large number of attributes. These include the country of birth of the person, their year of arrival in Australia, the country in which their mother or father was born, whether of Indigenous origin or not, or more generally their ancestry. Other dimensions of cultural diversity include differences in religious affiliation and in the use of a range of local and foreign languages. As well as being able to describe the sizes of the populations that make up various groups and how through further immigration their numbers may be changing, there is interest in the wellbeing of groups with different backgrounds. One area of interest relates to the fact that some groups experience disadvantage when seeking to obtain access to life opportunities including employment, education and those related to government and community programs.

It is in response to a widely recognised need for a nationally consistent framework for the collection and dissemination of data on cultural and language diversity that the ABS has developed standards for statistics relating to many of these attributes. The 'Standards for Statistics on Cultural and Language Diversity', as they are formally known, often built up from longstanding procedures, were endorsed by the Council of Ministers of Immigration and Multicultural Affairs (COMIMA) in April 1999. These include recommended questions, classifications, coding structures and output categories for use in interview-based and self-enumerated data collections.

INDIGENOUS POPULATION

The size and geographic distribution of the Aboriginal and Torres Strait Islander population are significant determinants of the distribution of government resources and the provision of services. Population estimates and projections are also essential components to the calculation of key Indigenous social statistics needed to monitor progress in addressing changes in health status or social and economic disadvantage over time. It is necessary, therefore, to identify this population in a wide range of data collections, particularly administrative systems. Classifying people according to their origin can be a sensitive issue. People can have mixed parentage or mixed ancestries so there can be ambiguities (as seen by the person themselves or by others) associated with whether a person belongs with one group or another.

According to the most widely adopted definition of Aboriginal or Torres Strait Islander people (the Commonwealth working definition):

'An Aboriginal or Torres Strait Islander is a person of Aboriginal or Torres Strait Islander descent who identifies as an Aboriginal or Torres Strait Islander and is accepted as such by the community in which he or she lives. (DAA, 1981)'.4

This definition, widely accepted by Commonwealth and other government agencies has three elements, descent, self identification and community acceptance. The choice of which one or which combination of these elements to use in collecting information about Indigenous people can affect the resulting statistics. In practice, it is not generally feasible to collect information on the community acceptance part of this definition and, therefore, questions on Indigenous Status relate to descent and identification only. The ABS standard question is as follows.

 Is the person of Aboriginal or Torres Strait Islander origin? For persons of both Aboriginal and Torres Strait Islander origin mark both 'Yes' boxes No Yes, Aboriginal Yes, Torres Strait Islander

Full details of the standard concepts, questions and classification used to produce statistics on indigenous status are provided in Standards for Statistics on Cultural and Language Diversity (Cat. no. 1289.0). These standard concepts are being used, or actively promoted for use, in a wide range of statistical and administrative data collections. This includes those relating to births and deaths to help support the production of Indigenous birth and death rates as well as population estimates and projections by adopting some of the methods used for producing estimates and projections for the total population of Australia, as described above. To date, largely because of the poor quality of available data, the ABS has used other means to produce population estimates and projections of the Indigenous population which are referred to as being 'experimental'. The experimental estimates for 1991 to 1996 used as their starting point the place of usual residence census count as at 6 August 1996, adjusted for non-response to the Aboriginal/Torres Strait Islander origin question in the census, net census undercount and registered Indigenous births. These 6 August 1996 Indigenous estimates were then 'survived' back to 30 June 1996 and then back to 30 June 1991 using life tables. This method is known as the reverse survival method and assumed zero net internal migration and zero overseas migration.

While the ASGC, described above, provides a basis for obtaining information about Indigenous people according to various geographical areas, the area classifications available in the ASGC have not always been ideal for obtaining area based statistics relating to Indigenous people. This is partly because the geographical distribution of Indigenous people is quite different from the total population on which the construction of CDs are based: many CDs have no, or very few, Indigenous people while others have large numbers. To help maximise the use of available small area data in a way that also accords with commonly referred to localities and administrative regions relating to the Indigenous population, the ABS has developed a separate geographical classification for Indigenous statistics that was first used to disseminate data from the 1996 Census. Like the ASGC, the Indigenous area classification is hierarchical and built on CDs. It consists of three levels: Indigenous Locations (which almost always have a population of at least 80 Indigenous people), Indigenous Areas and Aboriginal and Torres Strait Islander Commission regions. As with the areas presented in the ASGC these areas can also be aggregated to provide statistics according to user requirements.

 Previous Page Next Page