|Page tools: Print Page|
On this page:
Image of Question
Applicable to: Employed persons
3. Technicians and Trades Workers
4. Community and Personal Service Workers
5. Clerical and Administrative Workers
6. Sales Workers
7. Machinery Operators and Drivers
0998 Inadequately described
&&&& Not stated
@@@@ Not applicable
VVVV Overseas visitor
Total number of categories:
one digit level 8
two digit level 51
three digit level 134
four digit level 478
More Detailed Description
Quality Statement - Occupation (OCC06P)
There are many aspects which can affect the quality of Census data; the following information should be considered when viewing data on Occupation (OCC06P).
Occupation (OCC06P) is based on the Australian and New Zealand Standard Classification of Occupations (ANZSCO) 2006 (cat. no. 1220.0), and was coded using written responses on occupation (Question 38 on the household form) and on tasks performed (Question 39). Where possible standard procedures are used to obtain an occupation code, however the level of detail provided in response to these questions, and therefore the ease with which they can be coded, varies.
The non-response rate for this variable in 2006 was 0.8% compared with 1.2% in 2001. Unlike some other census variables the non-response rate is not affected by persons who were imputed into dwellings that did not return a Census form, as the occupation variables are only applicable for persons with a labour force status of employed.
A principle of coding occupation data is to allocate responses to the most descriptive and detailed level possible, where justifiable from the information supplied. If a response was not detailed enough to allow coding to the 6-digit occupation level (or 4-, 3-, or even 2- digit levels), a Not Further Defined (nfd) code was allocated. As shown in Table 1, an improvement in detailed coding outcomes has been achieved for 2006 data in total, and across most classification levels. In 2001, 8.4% of applicable occupation records were not able to be coded to 6 digit outcomes. For 2006, 7.7% of occupation records were unable to be coded to the most detailed level.
Table 1: Quality of Occupation response coding, 2001 and 2006 (a)
2006 occupation data was coded to both ANZSCO and the Australian Standard Classification of Occupations (ASCO) 2nd edition (cat. no. 1220.0) - as used in 2001. Table 2 shows how the two classifications relate, at the Major Group level:
Table 2: Occupation: Major groups, 2006 (ASCO2 and ANZSCO)
Just under 0.5% of occupation codes were automatically allocated during the Data Load phase, prior to the commencement of the main data coding process. Standard automated coding (AC) processes were then used to determine codes for a further 62.4% of all occupation responses (including where the respondent supplied no information), up from 57% in 2001. (Table 3 shows the proportion of Major Group occupations that were automatically coded in 2001 and 2006.) More complex responses were coded using clerical procedures and this accounted for the remaining 37.1% of the data. All coding processes were subject to sample checks to ensure an acceptable level of quality. Despite these checks, a clerical transcription error occurred during the early stages of processing - before all vetting systems were fully established. A small number of Program or Project Managers (386 in total) had their occupations coded to the correct ANZSCO code but the incorrect ASCO2 code (as Veterinarians). The mistake was not identified before processing was completed, and the records remain in error in ASCO2 output.
Table 3: Automated coding (AC) of stated responses for Occupation, 2001 and 2006 Census (ASCO2)
In 2006, the ASCO2 Major Groups most frequently coded automatically were 'Advanced Clerical and Service Workers', followed by 'Tradespersons and Related Workers'. Those Major Groups least likely to be coded automatically included 'Managers and Administrators' and 'Labourers and Related Workers'. The Sub-Major Groups recording the greatest improvement in automated coding results were '51: Secretaries and Personal Assistants' (an AC rate of 85%, up 22 percentage points from 2001) and '91: Cleaners' (an AC rate of 65%, up 16 percentage points from 2001).
For those records subject to clerical coding, occupations listed simply as Manager, Supervisor, Coordinator, Technician, Team Leader, Service Person, Customer Service, Installer, Labourer, IT, Analyst, and Clerks were the most difficult to classify at anything other than Major Group level.
Comparisons with other data sources
Census data can be used for the analysis of population characteristics at finer geographic levels and for smaller sub-groups than would be reliably available from household surveys. However, at small area data levels outliers (unusual results) may become more apparent to users. This becomes more probable as other data items are incorporated in the analysis, and users are reminded that almost all census data is as originally reported by the respondents. For some variable combinations, the use of interview-based, correlated survey results at a broader geographic level may therefore be more appropriate.
The table below compares occupation data from the 2006 Census at the broader Major Group with that from the August 2006 Labour Force Survey (LFS). The LFS provides Australia's official estimates of employment and unemployment, and includes quarterly information on occupation. Differences in the scope, coverage, timing, measurement of underlying labour force concepts and collection methodologies of the two collections are the major contributors to the differences in the counts of persons in different types of occupations they produce. Labour Force Survey occupation estimates are generally higher as Census data is unadjusted for underenumeration and only includes those usual residents present in Australia on Census Night. In addition Census counts are also affected by non-response (persons imputed into dwellings that do not return a Census form, as well as persons who are included on a completed form but do not respond to relevant labour force questions). Labour Force Survey estimates only include fully responding questionnaires and are adjusted to account for any non-response. (In the table below, persons for whom OCC06P was not stated have been removed to facilitate comparison). In comparing the two data sources below, the proportions of persons in each occupation category are similar at this broad level.
Table 4: Comparison of 2006 Census and Labour Force August 2006, Major Groups (a) ('000)
(a) Using ANZSCO classification
(b) excludes inadequately described responses, and records where occupation was not stated.
The Census can provide occupation data for small geographic areas or population groups, together with a range of other characteristics. It also produces data at the 6 digit level while the LFS only produces estimates at the 4 digit level (and LFS estimates below the national level can be subject to high sample errors). However, users of Census data at this more detailed level should be mindful of the limitations of collecting information via a census self-completed paper or e-form questionnaire where the responses provided are sometimes not sufficiently detailed to obtain an appropriate occupation code, for example descriptions such as "consultant", "manager" or "clerk". The LFS is conducted via personal interviews (either face-to-face, or over the telephone) which allows interviewers to clarify concepts and questions for respondents.
Additional sources of information regarding occupation can be found in other ABS publications and associated collections, including:
Selected Theme Pages on the ABS Website may also contain links to alternative data sources that may be of relevance to users.
The ABS aims to produce high quality data from the Census. To achieve this, extensive effort is put into Census form design, collection procedures, and processing procedures. More details regarding these efforts can be found in:
All are available from the ABS Website.