OVERVIEW
INTRODUCTION
Tuesday, 8 August 2006 was Census night in Australia. Every person present in Australia on Census night, excluding foreign diplomats and their families, should have been included on a Census form at the place where they stayed.
Whenever a Census is undertaken, questions about the completeness and accuracy of the Census count invariably arise. In such a large and complex exercise, it is inevitable that some people will be missed and some will be included more than once (or included when they shouldn't be). In Australia, the Census Post Enumeration Survey (PES) is used to determine how many people were missed in the Census and how many were counted more than once. The PES is a household survey conducted by the Australian Bureau of Statistics (ABS) shortly after the Census. The survey provides an independent check on Census coverage and also identifies the key demographic characteristics of the population that have been missed or over counted in the Census.
WHAT IS UNDERCOUNT?
In the 2006 Census, some people were missed (undercount) and some were counted more than once (overcount). As is usually the case, in 2006 more people were missed than over counted. The difference between the Census count and the number of people who should have been counted is called the 'net undercount' of the Census. (In this context, 'Census count' includes person records imputed for dwellings that were non-responding in the Census.)
Some of the reasons why people are missed in the Census include:
- they were travelling and were difficult to contact
- they mistakenly thought they were counted elsewhere
- there was insufficient space on the Census form in the household they were staying at and they did not obtain additional forms
- the person completing the form thought that young babies, the elderly or visitors (for example) should not be included
- they did not wish to be included due to concerns about confidentiality of information or a more general reluctance to participate
- the dwelling they were in was missed because it was difficult to find (e.g. in a remote or non-residential area)
- difficulty with completion of an eCensus form (Internet)
- the dwelling they were in was mistakenly classed as unoccupied.
Some of the reasons people are counted more than once in the Census include:
- they were included on the Census form at the dwelling where they usually live, even though they stayed, and were counted, elsewhere on Census night
- they were overseas on Census night and so should not have been counted at all, but were included on the Census form at the dwelling where they usually live.
The net undercount for a category of person is obtained by taking the PES estimate of the number of people in the category who should have been counted and subtracting the Census count of the number of people (in the category). Net undercount for a category of person is the net result of the PES estimate of gross undercount, gross overcount, differences in classification between the PES and Census (e.g. age, sex, Indigenous status) and imputation error in the Census.
Rates of net undercount vary significantly for different population groups depending on factors such as age, sex, ethnicity (including Indigenous status) and geographic location.
WHY MEASURE UNDERCOUNT?
There are a number of reasons why measuring undercount in the Census is important:
- to augment the Census count for the purpose of deriving an estimate of the resident population for 30 June of the Census year
- to provide users with an assessment of the completeness of the Census count, allowing them to take this into account when using Census information
- to evaluate the effectiveness of Census collection procedures so that improvements can be made for future Censuses.
Accurate resident population estimates are required for demographic, social and economic studies, as well as for the allocation of seats in the Federal House of Representatives, and Commonwealth payments to states and territories and local government.
CENSUS POST ENUMERATION SURVEY (PES)
The PES interview process determines whether each person in the sample should have been counted in the Census, and the category in which they should have been counted (such as age, sex, Indigenous status, region of usual residence). The match and search process determines how many times each person in the PES sample was actually counted in the Census. PES output processing and estimation combines and weights results from the match and search process to produce an estimate of the number of people who should have been counted in the Census. Net undercount is the difference between this estimate and the actual Census count (including imputed persons for non-responding dwellings).
A number of improvements have been implemented in the 2006 PES. These include:
- expanding the scope of the PES to include remote areas and discrete Indigenous communities
- the use of a Computer Assisted Interviewing instrument to replace the paper questionnaire of previous PESs
- special procedures to limit the impact on the PES of Census extended follow-up
- a more computerised match and search system for processing the PES data, including the recording of more detailed information about the reasons for matching decisions
- an improved estimation method, including a specific adjustment for dwelling-level non-response
- weighting for 'late return' and imputed dwellings designed to make these dwellings representative of this combined group in the Census
- new estimator (dubbed PREG), allowing:
- enhanced model for coverage and response e.g. separate effects for persons sampled in discrete Indigenous communities
- handling of differences in reporting (e.g. age, Indigenous status) between Census and the PES.
Estimation
PES estimation involves assigning a 'weight' to each selected PES dwelling and then to each person for whom a PES response was obtained. The PES only covers people associated with private dwellings and discrete Indigenous community dwellings available for selection during PES enumeration. However, the weights attached to persons allow the PES sample to represent the whole population of interest - that is, all people present in Australia on Census night, including people in non-private dwellings, such as hotels, hospitals and jails, which are not covered by the PES.
Dwelling weighting for the 2006 PES comprises two stages. For private dwellings selected in the PES that were found in the Census, the first stage of weighting adjusts the PES selection weight such that the adjusted weights add up to the Census private dwelling count within categories based on geography and dwelling characteristics. A first-stage weight adjustment is also applied to private dwellings selected in PES that were missed in the Census. For dwellings in discrete Indigenous communities, a similar first-stage weight adjustment is applied based on dwelling counts for communities within each state and territory.
The second stage of dwelling weighting applies a non-response adjustment so that the responding PES dwellings represent other dwellings from which no response was obtained.
The initial stage of person weighting adjusts the dwelling weights to ensure that the PES estimates of people counted in private dwellings and discrete Indigenous community dwellings (other than late return or imputed dwellings) in a set of benchmark categories match the actual Census counts for these categories. The weight adjustment applied to a person does not depend on whether they responded in the Census, but only on characteristics of the person as reported in the PES.
As a final step in weight adjustment, the initial person weights are adjusted so that the PES estimates also represent people in non-private dwellings.
Further information on PES estimation can be found in Appendix 3 of this publication.
Print Page
Print All