2940.0 - Census of Population and Housing - Details of Undercount, 2011  
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 21/06/2012   
   Page tools: Print Print Page Print all pages in this productPrint All

ESTIMATES OF NET UNDERCOUNT


INTRODUCTION

Understanding the Census count

Net undercount is the difference between the PES estimate of the number of people who should have been counted in the Census and the actual Census count. The Census count includes persons who have been imputed in non-responding dwellings in the Census, that is, the Census count equals persons counted on Census forms plus imputed persons for non-responding Census dwellings.

Net undercount is therefore a measure of the combined outcome of Census enumeration and data processing. For more information about imputed persons for Census non-responding dwellings and the adjustments made for them in the PES estimates, see Components of net undercount and the Identifying Census late returns Technical Note (in Explanatory Notes).


2011 PES net undercount estimates

In the following tables, net undercount is presented as both level estimates of persons and rates, together with their associated standard errors (SEs). The net undercount rate expresses the net undercount (i.e. undercount minus overcount) as a percentage of the PES estimate of a given population (i.e. as a percentage of the number of people who should have been counted in the Census). All estimates of net undercount based on geography have been calculated on a place of usual residence basis, meaning they are based on the location where a person lived, or intended to live, for six months or more in 2011.


AUSTRALIA

The 2011 Census counted 21,504,721 usual residents of Australia (including imputed persons in non-responding dwellings). This was around 374,540 persons fewer than the estimated population usual residents who were present in Australia on Census night. This equates to a net undercount rate of 1.7%. In other words, 98.3% of the usually resident population were included in 2011 Census counts.

While the net undercount rate decreased from 2.7% in 2006 to 1.7% in 2011, it is important to note that PES estimates of net undercount are not strictly comparable over time due to changes in both Census and PES methodologies. The PES is designed to provide the best measure of Census coverage at a single point in time rather than as a time series, with improvements made to the PES and Census in each cycle. This is particularly true for 2011 with the introduction of Automated Data Linking (ADL) in the PES which has made it difficult to directly compare level estimates and proportions of net undercount from one Census to another. Compositional analysis is therefore much more illustrative.

The ABS has estimated that the introduction of ADL for 2011 has resulted in a net undercount that was 246,985 persons less than if the 2006 methodology had been used. This estimate is subject to sample error. For further details see the Statistical Impact of ADL Technical Note (in Explanatory Notes).

Table 1 shows the net undercount rates and associated standard errors (SEs) for Australia for each Census from 1971 to 2011. The 2011 estimate of 1.7% continues the historical trend of almost complete coverage in Australian Census counts.

1 NET UNDERCOUNT RATE, Australia - 1971-2011

1971
1976
1981
1986
1991
1996
2001
2006(a)
2011(b)
%
%
%
%
%
%
%
%
%

Net Undercount
1.4
2.7
1.9
1.9
1.8
1.6
1.8
2.7
1.7
Standard Error (SE)
0.1
0.04
0.1
0.1
0.1
0.1
0.1
0.2
0.2

(a) Care should be taken when comparing estimates from 2006 onwards with previous years due to changes made to PES estimation and the inclusion of remote areas and discrete Indigenous communities in the PES sample from 2006.
(b) Care should be taken when comparing 2011 estimates with previous years due to changes made in PES linking and matching methodology. For more information see Linking and matching.



AGE AND SEX

The likelihood of counting a person in the Census has traditionally varied according to age and sex. As has been observed in previous Censuses (both in Australia and overseas), young adults are the age group who are most likely to be missed in the Census, with young adult males being more likely to be missed than their female counterparts. In contrast, older adults are much more likely to be counted.

Tables 2 and 3, and graph 4, show that this was also true for the 2011 Australian Census. In particular, males aged 20-24 years again had the highest net undercount rate (7.8%) followed by males aged 25-29 years (7.5%). The net undercount rate for females was also highest for those aged 20-24 years (6.0%). While the undercount rate for 25-29 year old females (4.0%) was higher than most age groups, in contrast to 2006 it was noticeably lower than the rate for females in their early 20s.

The lowest net undercount rate was for people aged 55 years and over (-0.1%).

In general, males had a higher net undercount rate (2.2%) than females (1.2%).

2 NET UNDERCOUNT(a), Sex by age group - 2011

Males
Females
Persons
Age
no.
SE
no.
SE
no.
SE

0-4 yrs
8 488
5 925
8 775
5 114
17 263
8 017
5-9 yrs
13 559
5 914
6 367
5 045
19 926
7 781
10-14 yrs
4 465
5 124
1 449
5 538
5 914
7 664
15-19 yrs
18 005
6 396
17 890
6 577
35 895
9 044
20-24 yrs
62 594
9 392
45 555
7 730
108 148
12 812
25-29 yrs
61 035
10 613
31 602
7 647
92 636
13 444
30-34 yrs
30 419
7 225
14 527
6 528
44 946
10 407
35-39 yrs
15 746
7 330
1 462
5 909
17 207
9 424
40-44 yrs
8 768
6 307
12 255
6 077
21 023
8 856
45-49 yrs
11 367
6 141
-5 876
5 448
5 491
8 063
50-54 yrs
11 848
6 279
1 900
5 428
13 747
8 411
55 yrs and over
-5 498
7 396
-2 161
7 626
-7 659
11 500
Total all ages
240 796
24 534
133 744
20 772
374 540
38 315

(a) A negative value indicates a net overcount.

3 NET UNDERCOUNT RATE(a), Sex by age group - 2011

Males
Females
Persons
Age
%
SE
%
SE
%
SE

0-4 yrs
1.2
0.8
1.3
0.7
1.2
0.6
5-9 yrs
1.9
0.8
1.0
0.8
1.5
0.6
10-14 yrs
0.6
0.7
0.2
0.8
0.4
0.6
15-19 yrs
2.4
0.8
2.6
0.9
2.5
0.6
20-24 yrs
7.8
1.1
6.0
1.0
6.9
0.8
25-29 yrs
7.5
1.2
4.0
0.9
5.8
0.8
30-34 yrs
4.1
0.9
1.9
0.9
3.0
0.7
35-39 yrs
2.1
0.9
0.2
0.8
1.1
0.6
40-44 yrs
1.2
0.8
1.5
0.8
1.3
0.6
45-49 yrs
1.5
0.8
-0.8
0.7
0.4
0.5
50-54 yrs
1.6
0.9
0.3
0.7
0.9
0.6
55 yrs and over
-0.2
0.3
-0.1
0.3
-0.1
0.2
Total all ages
2.2
0.2
1.2
0.2
1.7
0.2

(a) A negative value indicates a net overcount.

4 NET UNDERCOUNT RATE, Sex by age group - 2011
Graph: 4 NET UNDERCOUNT RATE, Sex by age group—2011



STATES AND TERRITORIES

The challenges facing Census enumeration vary between states and territories. Table 5 shows the rates of net undercount for Australian states and territories for Censuses from 1991 to 2011.

As in previous Censuses, in 2011 the Northern Territory recorded the highest net undercount rate of all states and territories (6.9%), while the Australian Capital Territory continued to record the lowest net undercount rate (0.7%).

While the two territories reflected the minimum and maximum net undercount rates, Victoria and South Australia continued to show relatively low rates (both 1.1%). Western Australia had the highest rate for a state (2.5%), emphasising the continued coverage challenges in that state. All states and territories had a lower net undercount rate in 2011 than in 2006, except for Tasmania which was relatively consistent at 2.0%. The greatest decreases were for Queensland (3.7% in 2006 to 1.8% in 2011) and Victoria (2.3% in 2006 to 1.1% in 2011).

It is important to note the effect of the introduction of Automated Data Linking (ADL) when considering the changes from 2006. For more information, see the Statistical Impact of ADL Technical Note (in Explanatory Notes).

5 NET UNDERCOUNT, State/territory of usual residence - 1991-2011

1991
1996
2001
2006(a)
2011(b)
Persons
Rate
Persons
Rate
Persons
Rate
Persons
Rate
Persons
Rate
no.
%
SE
no.
%
SE
no.
%
SE
no.
%
SE
no.
%
SE

New South Wales
109 200
1.9
0.1
91 400
1.5
0.2
130 106
2.0
0.2
157 578
2.4
0.4
136 647
1.9
0.4
Victoria
78 800
1.8
0.1
74 000
1.6
0.3
67 254
1.4
0.2
113 596
2.3
0.4
56 906
1.1
0.3
Queensland
52 100
1.8
0.1
57 300
1.7
0.3
68 514
1.9
0.2
148 409
3.7
0.4
77 215
1.8
0.4
South Australia
22 300
1.6
0.1
19 300
1.3
0.3
24 293
1.6
0.2
36 281
2.3
0.4
17 283
1.1
0.4
Western Australia
33 200
2.1
0.2
28 100
1.6
0.3
37 446
2.0
0.3
64 150
3.2
0.6
57 918
2.5
0.5
Tasmania
7 700
1.7
0.2
6 600
1.4
0.4
7 410
1.6
0.3
9 535
2.0
0.6
10 261
2.0
0.6
Northern Territory
4 800
2.9
0.7
5 700
3.1
1.6
7 814
4.0
0.6
15 909
7.6
1.5
15 716
6.9
1.3
Australian Capital Territory
4 100
1.4
0.2
3 400
1.1
0.3
3 282
1.0
0.4
4 027
1.2
1.0
2 595
0.7
0.8
Australia
312 300
1.8
0.1
285 800
1.6
0.1
346 119
1.8
0.1
549 486
2.7
0.2
374 540
1.7
0.2

(a) Care should be taken when comparing estimates from 2006 onwards with previous years due to changes made to PES estimation and the inclusion of remote areas and discrete Indigenous communities in the PES sample from 2006.
(b) Care should be taken when comparing 2011 estimates with previous years due to changes made to PES linking and matching methodology. For more information see Linking and Matching.



GREATER CAPITAL CITY/REST OF STATE REGION

The regional differences in net undercount in the 2011 Census for greater capital cities and the rest of state regions are presented in tables 6 and 7.

Greater capital cities are represented by Greater Capital City Statistical Areas (GCCSAs) and represent a socio-economic definition of each of the eight state and territory capital cities. This means each greater capital city includes people who regularly socialise, shop or work within the city but live in small towns and rural areas surrounding the city. 'Rest of state' regions are the areas within each state or territory not defined as being part of the greater capital city.

Different problems are encountered in enumerating different areas of Australia and these are reflected, to a certain extent, in the net undercount rates. In urban areas, locating dwellings is generally easier but contacting occupants and gaining their cooperation can be more difficult. In contrast, in rural and remote areas where dwellings may be scattered over a wider area, locating the dwellings can cause considerable difficulties. In 2011, New South Wales, Tasmania and the Northern Territory had higher net undercount rates in their rest of state regions compared with their greater capital cities. All other states and territories had a lower net undercount in their rest of state regions (compared to their greater capital cities).

At the broad Australia level, the total net undercount rates in 2011 were slightly lower for the rest of state regions (1.7%) compared to greater capital cities (1.8%). This contrasts with 2006, when the net undercount rate for balance of state/territory (3.0%) was higher compared to the capital cities (2.5%). As was the case in 2006, in 2011 the Northern Territory showed the largest difference in net undercount rate between its greater capital city and rest of state region (3.7% and 10.9% respectively).

It is important to note that for the 2011 PES, the Australian Statistical Geographical Standard (ASGS) replaced the Australian Standard Geography Classification (ASGC) as the framework for PES geography. The move to the new geography will allow for improvements in the quality of small area time series data from the Census. However, the change has resulted in an unavoidable break in series and care should be taken when comparing the 2011 greater capital city/rest of state net undercount estimates to the 2006 capital city/balance of state net undercount estimates.

6 NET UNDERCOUNT, State/territory of usual residence - By greater capital city/rest of state region - 2011

Greater capital city
Rest of state region
Total
no.
SE
no.
SE
no.
SE

New South Wales
79 825
21 972
56 821
17 139
136 647
26 865
Victoria
55 079
16 029
1 827
9 823
56 906
17 864
Queensland
41 898
11 244
35 316
13 997
77 215
17 431
South Australia
14 096
5 584
3 186
3 033
17 283
6 045
Western Australia
51 610
10 690
6 308
6 557
57 918
12 004
Tasmania
1 308
2 141
8 953
3 014
10 261
3 076
Northern Territory
4 637
2 254
11 078
2 157
15 716
3 074
Australian Capital Territory
2 595
3 055
na
na
2 595
3 055
Australia
251 050
31 913
123 490
24 771
374 540
38 315


7 NET UNDERCOUNT RATE, State/territory of usual residence - By greater capital city/rest of state region - 2011

Greater capital city
Rest of state region
Total
%
SE
%
SE
%
SE

New South Wales
1.8
0.5
2.2
0.7
1.9
0.4
Victoria
1.4
0.4
0.1
0.7
1.1
0.3
Queensland
2.0
0.5
1.5
0.6
1.8
0.4
South Australia
1.1
0.4
0.9
0.8
1.1
0.4
Western Australia
2.9
0.6
1.2
1.3
2.5
0.5
Tasmania
0.6
1.0
3.1
1.0
2.0
0.6
Northern Territory
3.7
1.7
10.9
1.9
6.9
1.3
Australian Capital Territory
0.7
0.8
na
na
0.7
0.8
Australia
1.8
0.2
1.7
0.3
1.7
0.2




REGISTERED MARITAL STATUS

Table 8 shows net undercount estimates and rates by registered marital status by sex. The net undercount rates were highest for people identified as never married (3.7%) and lowest for people widowed, divorced or separated (a net overcount of 0.8%). It is important to consider the strong relationship with age when interpreting net undercount estimates by registered marital status.

8 NET UNDERCOUNT(a), Registered marital status by sex - 2011

Males
Females
Persons
Persons
Rate
Persons
Rate
Persons
Rate
no.
SE
%
SE
no.
SE
%
SE
no.
SE
%
SE

Never married(b)
234 038
20 842
4.2
0.4
151 557
16 596
3.1
0.3
385 595
29 990
3.7
0.3
Widowed, divorced or separated
-8 837
10 274
-0.8
1.0
-15 434
9 878
-0.8
0.5
-24 271
14 630
-0.8
0.5
Married
15 594
10 970
0.4
0.3
-2 378
9 966
-0.1
0.2
13 216
18 737
0.2
0.2
Total persons
240 796
24 534
2.2
0.2
133 744
20 772
1.2
0.2
374 540
38 315
1.7
0.2

(a) A negative value indicates a net overcount.
(b) Includes those who are living with a de facto partner and have never been in a registered marriage.



INDIGENOUS STATUS

Special procedures are used in the Census to support the enumeration of the Aboriginal and Torres Strait Islander population, as counting this population continues to present a number of challenges.

The 2011 Census counted 548,147 persons who had been identified as being of Aboriginal and Torres Strait Islander origin, which was 21% more than the 454,799 persons in 2006. A summary of Aboriginal and Torres Strait Islander Census counts is presented in Census of Population and Housing - Counts of Aboriginal and Torres Strait Islander Australians (cat. no. 2075.0), which was also released on 21 June.

The 2011 PES estimated that 662,335 Aboriginal and Torres Strait Islander persons should have been counted in the Census, compared with 513,977 persons in 2006.

Table 9 shows net undercount estimates by Indigenous status. In 2011, the net undercount rate was 17.2%, compared with 11.5% in 2006. The net undercount for 2011 was estimated to be 114,188 persons, which was almost double the 2006 estimate of 59,178 persons.

9 NET UNDERCOUNT, Indigenous status - 2011

Persons
Rate
no.
SE
%
SE

Indigenous  
114 188
14 274
17.2
1.8
Non-Indigenous
1 318 799
37 272
6.2
0.2



It is important to note that these measures refer to the undercount of persons according to their Indigenous status, regardless of whether or not they were actually counted in the Census. In other words, persons who were counted in the Census and had a 'not-stated' Indigenous status will not be included in the Census counts of either Aboriginal and Torres Strait Islander or non-Indigenous persons, but are, instead, a separate category for this classification. They will, however, be included in Census counts for other key categories, such as Age and Sex.

In order to understand the differences between the 2006 and 2011 PES results, the ABS undertook an extensive quality assurance process. The results of this process are summarised in the Improvement in collection of Indigenous status Technical Note (in Explanatory Notes).

This quality assurance process has led the ABS to advise caution when comparing net undercount for Aboriginal and Torres Strait Islander persons between 2006 and 2011. Analysis of data indicates that the main contributing factor for the difference between the 2006 and 2011 estimates was improved PES methodology and procedures, which resulted in better identification of Aboriginal and Torres Strait Islander people in the 2011 PES.

Historically, the ABS has a program of continuous improvement in its survey methodologies. Improvements to the PES in 2006 and 2011 are summarised in Survey Enumeration, Linking and Matching and Estimation. While the individual impacts of all improvements made in 2011 cannot be measured, they have resulted in a change in the Indigenous status classification of Aboriginal and Torres Strait Islander persons in the PES and the Census between 2006 and 2011. This has in turn resulted in a noticeably different net category change for Indigenous status in net undercount estimates, and accounted for most of the change in the estimate of net undercount for Aboriginal and Torres Strait Islander persons between 2006 and 2011.

It is also important to note that Indigenous status, as collected in both the Census and PES, is based on responses to a question related to information that some people will consider personal and sensitive. Respondents can choose to indicate in the Census that they are Indigenous or non-Indigenous, or they can choose to not answer the question at all. If no answer is provided, the Census does not impute for this missing response (which is also the case for imputed persons). The Census count is therefore a count of those who were identified by a respondent as Indigenous (i.e. those without a response are excluded).


COUNTRY OF BIRTH

As Census forms are generally completed by one or more persons in a household, those who have come to Australia from other countries and whose first language is not English may find completing a Census form more difficult than other Australians. For several Censuses, special strategies have been employed to promote an understanding of the Census among migrants, in particular that the Government is not using their information for anything other than statistical purposes and to provide assistance in a range of languages.

Tables 10 and 11 show the undercount estimates and rates by country of birth. The countries displayed were the 10 highest ranked (in terms of population residing in Australia) according to the 2011 Census. There were 1,195,432 people (5.6% of the Census count) whose country of birth was not stated in the Census. Since Census does not impute a Country of birth for these people, the PES estimates of net undercount are not adjusted to take account of any imputed persons. As with Indigenous status, these people, while counted in the Census, do not contribute to the Census counts for these categories but do count to PES estimates of their population. For further information about Census not-stated responses and their impact on estimates of net undercount see Components of net undercount.

Of those countries listed, persons born in China had the highest net undercount (55,965 persons) followed by New Zealand (46,536 persons). China also had the highest net undercount rate (14.9%) followed by India (9.7%). Persons born in Scotland had the lowest net undercount rate (1.2%) followed by those born in England and Italy (both 4.6%).

10 NET UNDERCOUNT(a)(b), Country of birth by sex - 2011

Males
Females
Persons
Country of Birth
no.
SE
no.
SE
no.
SE

Australia
584 461
19 579
477 937
16 850
1 062 398
30 156
England
26 257
4 870
17 938
4 549
44 195
7 049
New Zealand
30 504
4 323
16 032
3 770
46 536
6 344
China
28 239
5 375
27 726
5 033
55 965
8 934
India
22 309
4 371
9 386
2 856
31 694
6 075
Italy
4 880
1 631
4 004
1 654
8 884
2 513
Vietnam
8 020
2 736
8 075
2 326
16 095
4 313
Philippines
6 441
1 952
10 665
2 417
17 106
3 603
South Africa
7 099
2 138
4 613
1 791
11 712
3 383
Scotland
-1 348
1 511
2 929
1 674
1 582
2 208
Other overseas
151 369
11 502
122 435
8 990
273 804
15 894

(a) Net undercount is based on Census counts for a category. In the Census, Country of birth was set to not-stated where the response was blank and where imputed person records were created for non-responding dwellings. Hence net undercount estimates for Country of birth do not sum to the Australia total.
(b) A negative value indicates a net overcount.

11 NET UNDERCOUNT RATE(a)(b), Country of birth by sex - 2011

Males
Females
Persons
Country of Birth
%
SE
%
SE
%
SE

Australia
7.3
0.2
5.9
0.2
6.6
0.2
England
5.4
1.0
3.8
0.9
4.6
0.7
New Zealand
11.1
1.4
6.3
1.4
8.8
1.1
China
16.6
2.6
13.5
2.1
14.9
2.0
India
12.0
2.1
6.7
1.9
9.7
1.7
Italy
4.9
1.6
4.2
1.7
4.6
1.2
Vietnam
8.6
2.7
7.5
2.0
8.0
2.0
Philippines
9.1
2.5
9.1
1.9
9.1
1.7
South Africa
9.0
2.5
5.9
2.2
7.4
2.0
Scotland
-2.1
2.4
4.2
2.3
1.2
1.6
Other overseas
11.2
0.8
8.8
0.6
10.0
0.5

(a) Net undercount is based on Census counts for a category. In the Census, Country of birth was set to not-stated where the response was blank and where imputed person records were created for non-responding dwellings.
(b) A negative value indicates a net overcount.