FEATURE ARTICLE: REBASING OF AUSTRALIA'S POPULATION ESTIMATES USING THE 2016 CENSUS


INTRODUCTION

In this issue, the ABS used the results of the 2016 Census of Population and Housing to undertake two main sets of calculations to produce preliminary rebased population estimates. The new population estimates are said to be based on the 2016 Census because they use the 2016 Census as their main data source. Rebasing is the process of preparing this new population base from the most recent Census. Rebased estimates supersede all previously released estimates.

The first set of calculations constructed a new estimated resident population (ERP) for 30 June 2016 from which quarterly ERP is calculated forward. This ensures that population estimates for the following intercensal period (2016-2021) are as accurate as possible, based on the most recent Census.

The second set of calculations revised the 19 intercensal quarterly estimates preceding 30 June 2016 (i.e. 30 September 2011 to 31 March 2016). This ensures that the ERP time series for the previous intercensal period are comparable with later estimates.

These estimates will be updated again using more up-to-date data to produce final rebased estimates in the December quarter 2017 issue of this publication (released in June 2018). No subsequent revisions are then planned to be made to these estimates up to and including 30 June 2016.


CONSTRUCTING THE PRELIMINARY ERP BASE FOR 30 JUNE 2016

The preliminary rebased 30 June 2016 ERP is constructed from the 2016 Census count in three main steps.

1. Moving people back home

Firstly, people are counted by their place of usual residence rather than their place of enumeration. This accounts for interstate visitors on Census night and removes overseas visitors.

2. Including people who have been missed

The second step addresses people missed or counted more than once in the Census by:

• adjusting for Census undercount and overcount using the results from the 2016 Census Post Enumeration Survey (PES);
• adding in the number of Australian residents temporarily overseas (RTOs) on Census night using data on international travellers obtained from the Department of Immigration and Border Protection; and
• applying demographic adjustments designed to resolve any other anomalies not accounted for in the PES or RTO adjustments.

3. Backdating from Census night to 30 June

The third step backdates the ERP at Census night (9 August 2016) to the ERP at 30 June 2016 by;

• subtracting births;
• adding deaths;
• accounting for net interstate migration; and
• accounting for net overseas migration.

The steps are illustrated in the following diagram:

The table below shows how the ABS arrived at 30 June ERP for Australia, states and territories after applying the adjustments above.

ADJUSTMENT COMPONENTS OF ESTIMATED RESIDENT POPULATION, preliminary - 30 June 2016
Persons ('000)		NSW	Vic.	Qld	SA	WA	Tas.	NT	ACT	Aust.(a)
Components as at 9 August 2016:
	Census count, actual location	7 565.0	5 946.1	4 844.5	1 674.8	2 517.8	504.2	259.1	400.6	23 717.4
	plus - Residents absent interstate	84.6	99.4	56.2	31.9	23.7	15.5	5.5	13.0	330.1
	less - Interstate visitors	73.4	45.7	113.6	15.6	35.9	6.8	26.8	11.4	330.1
	less - Overseas visitors	95.9	73.1	83.9	14.4	31.2	3.0	9.0	4.9	315.5
equals - Census count, place of usual residence		7 480.2	5 926.6	4 703.2	1 676.7	2 474.4	510.0	228.8	397.4	23 401.9
plus - Allowance for under-enumeration(b)(c)		59.2	86.1	60.6	3.0	9.5	0.3	12.0	-4.1	226.5
plus - Demographic adjustment(c)		-3.0	-2.5	-2.0	-0.7	-0.9	-0.2	-0.1	-0.2	-9.6
plus - Residents temporarily overseas(c)		217.2	183.4	94.3	35.2	78.5	7.9	5.1	11.2	632.7
equals - ERP as at 9 August 2016(c)		7 753.5	6 193.6	4 856.1	1 714.1	2 561.5	517.9	245.9	404.3	24 251.5
Backdating components to 30 June 2016:
	less - Births(c)	11.1	8.5	6.6	2.2	3.8	0.7	0.4	0.6	33.8
	plus - Deaths(c)	6.4	4.6	3.6	1.6	1.8	0.5	0.1	0.2	18.9
	less - Net interstate migration(c)	-1.0	1.3	1.5	-0.5	-1.1	-	-0.2	-	. .
	less - Net overseas migration(c)	10.6	9.1	2.7	1.0	1.6	0.2	0.1	0.4	25.8
equals - preliminary ERP 30 June 2016(c)		7 739.3	6 179.2	4 848.9	1 713.1	2 559.0	517.6	245.7	403.5	24 210.8
. . not applicable
- nil or rounded to zero (including null cells)
(a) Includes Other Territories - see paragraph 2 of the Explanatory Notes.
(b) Includes Census undercount and overcount from the 2016 PES and minor adjustments to address additional data coherence and quality matters.
(c) These estimates are preliminary rebased and will change for the compilation of final rebased estimates.

ADJUSTING FOR NET UNDERCOUNT

Adding the net undercount (undercount minus overcount) of people back into the population is a crucial step in arriving at the most accurate ERP possible. Net undercount is the difference between the actual Census count and the estimate of the number of people who should have been counted in the Census. It is the result of Census undercount, overcount, differences in classification between the PES and Census, and imputation error.

The 2016 PES estimate of net undercount for Australia in the 2016 Census was 226,400 people (including additional minor coherence and quality adjustments). For more information on the PES see Census of Population and Housing: Details of Overcount and Undercount, Australia (cat. no. 2940.0) and Information Paper: Measuring Overcount and Undercount in the 2016 Population Census (cat. no. 2940.0.55.002).


RESIDENTS TEMPORARILY OVERSEAS

Residents temporarily overseas (RTOs) during the Census are those who were outside Australia for less than 12 months over a 16-month period, and therefore are added to the ERP. A departing resident is excluded from Australia's resident population only when they leave Australia for 12 months or more over a 16-month period. The preliminary estimate of RTOs on 2016 Census night was 632,700 people. The preliminary estimate models the duration of overseas stay of departing residents. The final estimate of RTOs will be calculated when data for actual arrivals and departures up to 16 months after Census are available, and will be included in the final rebased ERP.


DEMOGRAPHIC ADJUSTMENTS

Demographic adjustments are a range of small adjustments made to the population at the national level. These adjustments ensure coherency with previous population estimates and address known issues, such as age misreporting in the centenarian population. Detail on these demographic adjustments will be provided in the forthcoming Information Paper: Ensuring the Quality of Rebased Population Estimates, 2016 (cat. no. 3250.0); to be released 28 July 2017.


INTERCENSAL DIFFERENCE

Intercensal difference is the difference between new Census-based ERP for 30 June of the Census year and the ERP for the same point in time carried forward from the previous Census base using births, deaths and migration data.

Intercensal difference can be attributed to one or more of three sources;

• the rebased estimate based on the previous Census;
• the rebased estimate based on the current Census; and /or
• the components of population change (births, deaths and migration) between the two Censuses.

The table below shows the preliminary intercensal difference by number and percentage of total population for 2011 to 2016 for Australia, states and territories.

Intercensal difference 2011 - 2016(a)
	Intercensal difference
	'000	%
New South Wales	10.0	0.1
Victoria	108.7	1.8
Queensland	6.9	0.1
South Australia	4.7	0.3
Western Australia	-57.4	-2.2
Tasmania	-1.4	-0.3
Northern Territory	0.5	0.2
Australian Capital Territory	7.1	1.8
Australia(b)	78.7	0.3
(a) A positive number indicates that rebased ERP as at 30 June 2016 was higher than unrebased ERP. A negative number indicates it was lower than unrebased ERP.
(b) Includes Other Territories - see paragraph 2 of the Explanatory Notes.

Information collected in the 2016 Census will also allow the ABS to estimate how much of the preliminary intercensal difference is due to inaccuracies in estimates of interstate migration. To do this, the ABS will analyse data from the Census questions about place of usual residence one and five years ago. These results will be released in the December 2017 issue of this publication (released in June 2018).

After the preliminary intercensal difference is adjusted for revisions to the components of population change (births, deaths and migration), the remaining (un-attributable) portion is referred to as the final intercensal difference. The intercensal difference for each state and territory, birth cohort and sex are spread evenly across the intercensal quarters. Thus the intercensal difference acts as a balancing item, that when combined with births, deaths and migration equals the difference between the two Census-based population estimates. The final intercensal difference cannot be attributed to a particular source.


FURTHER DATA RELEASES

Final population estimates based on the 2016 Census for Australia, states and territories, for 30 September 2011 to 30 June 2016, will be published in the December quarter 2017 issue of Australian Demographic Statistics (cat. no. 3101.0) to be released in June 2018. No subsequent revisions are planned to be made to the final rebased estimates.

Preliminary rebased estimates for SA2s and LGAs will be published in a special issue of Regional Population Growth, Australia (cat. no. 3218.0) to be released on 28 July 2017. Final rebased sub-state estimates will be released by August 2018.

Preliminary rebased estimates of the Australian Aboriginal and Torres Strait Islander population by age, sex and state/territory at 30 June 2016 will be published in the March quarter 2017 issue of Australian Demographic Statistics to be released on 28 September 2017. Further disaggregation of these estimates will be released in August 2018.

Further detail on the 2016 Census rebasing of Australia's population estimates will be published in Information Paper: Ensuring the Quality of Rebased Population Estimates, 2016 (cat. no. 3250.0) to be released on 28 July 2017.