2900.0 - Census of Population and Housing: Understanding the Census and Census Data, Australia , 2016

ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 02/05/2018 First Issue

Page tools: Print

Print Page Print all pages in this product

Should I use Area of Enumeration Data or Usual Residence Data?

The Census is collected on a “Place of Enumeration” basis. This means that wherever you are staying on Census night is where you fill in a form and are counted. One of the questions on the Census form also asks “Where does the person usually live?” Responses to this can be the same as the current location (place of enumeration) or at a different address entirely. In effect, this creates two different sets of data. The enumerated dataset is a count of where people actually were on Census night. The usual residence dataset records where people usually live. Both sets of data have different uses and advantages, allowing them to be used in different situations.

What is Area of Enumeration Data?

The place of enumeration is the place at which the person is counted i.e. where they spent Census Night, which may not be where they usually live. It is considered neither a person, family nor dwelling level variable but rather a geography and can be applied to all levels of data, allowing access to data on person, dwelling and household topics. Census counts based on place of enumeration can be provided for individual Mesh Blocks (MBs), and any aggregations of Mesh Blocks, such as postal areas or Statistical Areas Level 2 (SA2s) or above. This means that an Area of Enumeration data set can be as small as a single suburb or as large as Australia as a whole. A detailed description of place of enumeration can be found in the 2016 Census Dictionary.

Area of enumeration data provides a good representation of any given day in time because there will always be people visiting, occupying non-private dwellings, travelling, etc. that are not apparent when usual residence data is used. This can provide insight into atypical situations, such as holiday resort areas; areas like the Gold Coast or snow fields in winter, which may show a larger place of enumeration count compared with their usual residence count.

In 2016, 93.9% of people were enumerated at home (counted on Census Night where they usually live).
In 2011, 95.2% of people were enumerated at home. *
In 2006, 95.3% of people were enumerated at home.

* The 2011 TableBuilder Pro dataset - Counting Families - is incorrectly labelled as Place of Usual Residence. It should be labelled as 2011 Census - Counting Families, Place of Enumeration.

What is Usual Residence Data?
A place of usual residence (PURP) refers to the area where a person usually lives regardless of where they were on Census Night. It is NOT considered a geography; rather it is considered a person level variable and therefore can only be applied to persons not families or dwellings. The responses to questions relating to place of usual residence in the Census form code them to a Mesh Block of usual residence. The PURP variable is hierarchical which means that Census counts based on PURP counts can be also be aggregated to higher levels of geography. When using usual residence data people are not coded to a specific dwelling or family and are instead linked back to the area in which they usually live. Like all person level fields, usual residence can ONLY be used when counting persons. Further information beyond person data and usual geographic location cannot be determined. A detailed description of usual residence data can be found in the 2016 Census Dictionary.

Usual residence data records where people usually live based on Census night. This means you are only dealing the data from the residents of a particular area, not that of visitors. Because of the nature of usual residence data and how it is collected, it is less likely to be impacted by seasonal influences. This is useful for organisations which use the locally resident population, for example the age structure of the people who live in a certain area can be used to determine the allocation of funds for health services to a region.

Comparing Data

In the two tables below we can set the total populations for four areas in Queensland at the SA3 geography level. The Area of Enumeration table contains the total number of persons residing in each SA3 on Census night, including overseas visitors and guests. The Usual Residence table contains the total number of residents that usually live in dwellings located in these specific geographic areas.

Area of Enumeration Total Population	Usual Residence Total Population

As seen in the tables, certain areas contain an increased total population when looking at place of enumeration. Areas such as Surfers Paradise and Brisbane Inner have larger enumerated populations when compared with their usual residence totals, most likely due to an influx of tourists and visitors. This is not always the case, however. Certain places such as Townsville or Toowoomba might have their populations remain static over Census night, with little or no increase in their total populations. In some cases the Enumerated population might even be smaller than the usual residence.

One reason that the Census is typically carried out on an August date is with the aim of minimising the impact of travel, avoiding major public holidays and school holidays. When large numbers of people are away from their usual residence, it increases the difficulty of data collection.

Why can’t I count Families or Dwellings using Place of Usual Residence?

Firstly, a dwelling does not move. It does not have a ‘place of usual residence’. Likewise, all the variables associated with a dwelling belong to that particular dwelling and cannot be moved. It is not logical to count the number of bedrooms in a dwelling by usual residence. The number of bedrooms belongs to a particular dwelling and they do not move, regardless of whether the people within them do.

Secondly, using ‘place of usual residence’ will produce a number that is misleading and likely to be incorrectly interpreted.

For example: Using the illustration below, imagine there are five dwellings with male and female residents represented as outline figures. Now count the number of dwellings by sex:

image: first dwelling two females, second dwelling one male one female, third dwelling, one male, fourth dwelling one male one female, fifth dwelling two males one female

This would produce a table showing the count of dwellings containing females as four and the count of dwellings containing males as four, insinuating a total of eight dwellings. However there are only five dwellings altogether, not eight.

Family variables are only derived for people counted at their usual residence. Temporarily absent persons are used to help classify types of relationships and families existing in a household, but they are not used in the derivation of any other Census characteristics or in other Census output. Family and household structures are based on persons usually resident. If ALL members of a family or household are temporarily absent from their usual residence, no family records are created for them and the family or household is not counted.

Uses for each Dataset

Area of Enumeration	Place of Usual Residence
A point in time (Tuesday night in August) – Provides a ‘snapshot’ of mobile populations.	Address where people usually live or have lived recently (6 months or more) – Less likely to be influenced by seasonal factors such as school holidays and snow seasons.
Available for all person, dwelling and household topics	Available for person datasets only
Useful to gauge temporary population such as tourists and itinerant workers	Useful for all services which use the locally resident population such as planning for local government services
Does not provide the ‘typical’ population of an area	Cannot determine characteristics of absent people
Population may be inflated in some areas, such as tourist locations and CBD (central business district)	Does not capture population mobility
Available data for all (Census) years back to 1991	Available for data from 2006 onwards. Censuses prior to this have provided limited information on usual residence
Example question: How many visitors were staying in Surfers Paradise on Census night?	Example question: What is the age structure of the people who live in a certain area?