A COMPARISON OF THE CENSUS AND THE LABOUR FORCE SURVEY
The ABS's Labour Force Survey (LFS) and the Census of Population and Housing both collect information about labour market activity of people aged 15 years and over. The LFS is a household survey which provides Australia's official estimates of employment and unemployment, which are released on a monthly basis. The Census on the other hand is conducted every five years and aims to measure the number of people living in Australia on Census night and the dwellings in which they live. The Census also collects information about a range of characteristics of people, including, but not limited to, information about their labour force status.
The Census provides a count or snapshot of people living in the country on Census night. Being a Census, it is a good source of information for small population groups and areas and allows for the analysis of industry and occupation data at a more detailed level. But it is conducted on a five yearly basis, therefore not an appropriate source of data for monitoring ongoing labour market conditions. The LFS on the other hand is conducted monthly and is designed specifically to measure changes over time in the Australian labour force. Therefore it provides a more accurate estimate of key labour force statistics of the Australian economy as well as a range of more detailed data.
While both collections measure the same concepts surrounding the labour force in Australia, there are a number of differences between the two that should be considered. This article describes the differences between the collections and explains why estimates from the Census and the LFS are not directly comparable.
DIFFERENCES BETWEEN THE CENSUS AND THE LFS
Scope and coverage
While the scope of the Census is similar to that of the LFS, there are differences which can have an impact on the data. The scope of the LFS is limited to the civilian population of Australia, therefore members of permanent defence forces are excluded from the survey. In contrast, the Census includes everyone living in Australia on Census night, including the defence forces. This has the impact of increasing the number and proportion of employed people in the Census relative to the LFS, given that all members of the permanent defence forces are employed.
The Jervis Bay Territory, the Territory of Christmas Island and the Cocos (Keeling) Islands are currently out of scope for ABS household surveys, including the LFS, but are in scope of the Census.
The Census includes only those people who are in Australia on Census night, while the LFS includes usual residents who are out of the country for less than 6 weeks.
Although the Census aims to capture information about every person in Australia on Census night, it is inevitable that in such a large undertaking some people will not be counted. To account for this, a Post Enumeration Survey is undertaken a few weeks after the Census, to estimate the number and characteristics of people who were not counted on Census night.
By contrast, the LFS is a sample survey and weighted to an independent population benchmark based on the Estimated Resident Population (ERP). This ensures that the LFS estimates add up to an independently estimated distribution of the usually resident civilian population aged 15 years and over, regardless of any sample lost due to non-response.
The ABS uses overseas arrivals and departures data to estimate the number of people overseas on Census night in calculating the ERP, but no adjustment is made to the Census data based on these estimates. The LFS estimates for August 2011 are currently benchmarked to the ERP based on the 2006 Census. The rebasing of the LFS benchmarks to the 2011 Census-based ERP estimates will be undertaken in early 2014.
The Census uses a self-enumeration methodology where households are required to complete the Census form themselves and therefore relies entirely on the initial response provided by the respondents. By contrast, the LFS is conducted predominantly by trained ABS interviewers, either face-to-face or over the phone. Post the 2011 Census the ABS commenced the introduction of on-line self-enumeration for the LFS. For more details see the April 2013 issue of Labour Force, Australia (cat. no. 6202.0). Although both enumeration methods can lead to an error if the question is misunderstood, this is less likely to happen with the LFS where the interviewer is able to clarify responses to questions asked. The interviewer is also able to request additional information if answers provided are not sufficiently detailed to ensure accurate classification or coding of industry and occupation of employment.
The Census determines labour force status (i.e. whether a person is employed, unemployed or not in the labour force) from responses to four questions, whereas in the LFS, labour force status is derived from around 30 questions. Many of the additional questions in the LFS relate to the circumstances where a person may be defined as 'employed' even though they were not at work during the survey reference week. The use of the reduced question set in the Census may result in some employed people being incorrectly categorised as either unemployed or not in the labour force, and some unemployed people being categorised as not in the labour force. The additional questions in the LFS therefore results in a more accurate classification and better measure of the employed and unemployed at a point in time.
The estimates derived from the LFS are based on information obtained from a sample of dwellings, therefore the results are subject to sampling error. The sample is designed to ensure that sampling error is reduced to a minimum at the national and state/territory levels. However, it can be higher for labour force regions or for detailed breakdowns. The estimates are therefore accompanied by information on the quality of the estimates including relative standard errors. By contrast, the Census is not subject to sampling error because it aims to collect the information from everyone in Australia on Census night.
Treatment of non-response
Another difference between the two collections is the treatment of non-response. For various reasons, a proportion of households that are occupied on Census night either cannot or do not return their Census forms and are referred to as the non-response population. In the 2011 Census, this population was 3.7%. To account for this, persons in non-responding households are imputed, along with some demographic characteristics. In contrast, the LFS does not include non-responding households because only fully responding households contribute to the estimates, with any non-response being compensated for by the weighting process.
|Table 1. Summary of differences between the Census and the LFS|
|Scope and coverage||Includes anybody living in Australia on Census night.||Limited to the civilian population.|
|Includes usual residents who are out of the country for less than 6 weeks.|
Excludes members of the defence forces.
|Frequency||Conducted every five years.||Monthly collection. |
|Reference period||The week prior to the Census night.||The week prior to the week in which the survey is conducted.|
|Collection methodology||Self-enumeration.||Interview - personal visit or telephone.|
|Questions||Labour force status determined from four questions.||Labour force status determined from around 30 questions.|
|Treatment of non-response||Some items are imputed. |
Others included in 'not stated' category.
|Treated as 'not stated' therefore excluded and adjusted through the weighting process.|
The following section compares Census data on four key labour market variables with original data from the August 2011 LFS. The 2011 variables compared are labour force status, hours worked, industry of employment and occupation of employment.
Labour Force Status
Comparing the labour force status between Census and LFS data show that there are differences in the estimated number of persons across the labour force status categories, as well as for the total number of people aged 15 years and over. While the LFS estimates of employed people, unemployed people and those not in the labour force are significantly higher than the Census count, the data produced similar results using proportions.
The Census and LFS also differ in the way they determine full-time and part-time employment. In the Census, a person is considered to be working full-time if they worked 35 hours or more in all jobs during the week prior to Census night or part-time if they worked less than 35 hours. In addition, they have a category for 'employed away from work'.
The LFS designates full-time workers as persons who (a) usually work 35 hours or more per week in all jobs, or (b) although usually working less than 35 hours a week, actually worked 35 hours or more during the survey reference week. In the LFS, part-time workers are those who usually work less than 35 hours per week, and either did so during the reference week, or were not at work in the reference week. People employed but away from work are categorised as either employed full-time or part-time based on actual hours worked.
Table source: ABS 2011 Census of Population and Housing 2011, Labour Force Australia, Detailed (cat. no. 6291.0.55.001).
|Table 2. Labour Force Status|
LFS, August 2011(a)
Number of persons
Number of persons
|Labour force |
|Employed total |
|Employed, worked full-time |
|Employed, worked part-time |
|Employed, away from work (a) |
|Unemployed total |
|Unemployed, looking for full-time work |
|Unemployed, looking for part-time work |
|Not in the labour force |
a. In the LFS people employed but away from work are categorised as either employed full-time or part-time based on usual hours worked.
b. The number of people with labour force status 'not stated' in the Census (975,929) have been removed to enable better comparison with the LFS.
|c. Components may not necessarily add up to the totals due to rounding. |
The hours worked data in the Census refers to hours worked in the reference week and are therefore compared with the LFS data item 'actual hours worked', rather than 'usual hours worked'.
Comparing the actual hours worked using Census and LFS data indicated that they broadly had similar distributions. But the category 40 hours had 18.8% for the Census and 14.6% for the LFS. Also a higher proportion of employed people in the Census indicated that they worked 35-39 or 40 hours, compared with the LFS. This may partly be due to rounding of hours worked by people in the Census (self-enumeration) to their perception of full-time hours (e.g. 38, or 40 hours), while more precise estimates may be provided when responding to LFS interviewers.
The Hours worked item was not stated for 2.2% (218,484) of employed persons in the Census and their exclusion may impact on the distribution.
Table source: Source: ABS 2011 Census of Population and Housing 2011, Labour Force Australia, Detailed (cat. no. 6291.0.55.001)
|Table 3. Hours worked |
LFS, August 2011
|0 hours |
|41 and above|
Industry of employment
a. The number of people whose hours worked was 'not stated' in the Census (218,484) has been removed from the Census Total of 10,057,484 to aid comparison.
|b. Components may not necessarily add up to the totals due to rounding.|
Industry of employment in both Census and LFS data produced similar distributions. The industry division 'Retail trade' had the highest percentage of 10.8% and 10.7% for Census and LFS respectively, whilst the division 'Electricity, gas, water and waste services ' had the smallest percentage of 1.2% for both Census and LFS.
The industry of employment was not determined for 2.3% (233,886) of employed people in the Census as it was either ‘not stated’ or ‘inadequately described'.
Table source: Source: ABS 2011 Census of Population and Housing 2011, Labour Force Australia, Detailed, Quarterly (cat. no. 6291.0.55.003).
|Table 4. Industry of employment |
LFS, August 2011
|Industry (a) |
|Agriculture, forestry and fishing |
|Electricity, gas, water and waste services |
|Wholesale trade |
|Retail trade |
|Accommodation and food services |
|Transport, postal and warehousing |
|Information media and telecommunications |
|Financial and insurance services |
|Rental, hiring and real estate services |
|Professional, scientific and technical services |
|Administrative and support services |
|Public administration and safety |
|Education and training |
|Health care and social assistance |
|Arts and recreation services |
|Other services |
Occupation of employment
a. Classified according to the Australian and New Zealand Standard Industrial Classification (ANZSIC) 2006 (cat no. 1292.0).
b. The number of people whose industry of employment was either 'not stated' (109,131) or 'inadequately described' (124,755) in the Census have been removed from the Census total of 10,058,325 to aid comparison.
|c. Some components may not sum up to the totals due to rounding.|
Similar to industry, the distribution across occupation of employment was similar for both the Census and the LFS. The occupation category 'Professionals' had the highest percentage of 21.7% and 21.6% respectively for Census and the LFS. The occupation category 'Machinery operators and drivers' had the smallest percentage of 6.7% and 6.8% respectively.
The occupation of employment could not be determined for 1.9% (189,013) in the Census as it was either ‘not stated’ or ‘inadequately described'.
Table source: ABS 2011 Census of Population and Housing 2011, Labour Force Australia, Detailed, Quarterly (cat. no. 6291.0.55.003).
Table 5. Occupation of employment
LFS, August 2011
|Occupation (a) |
|Technicians and trades workers |
|Community and personal service workers |
|Clerical and administrative workers |
|Sales workers |
|Machinery operators and drivers |
a. Classified according to ANZSCO - Australian and New Zealand Standard Classification of Occupations, First Edition, 2006 (cat no. 1220.0).
b. The number of people whose occupation was either 'not stated' (79,861) or 'inadequately described' (109,152) in the Census has been removed from the Census total of 10,058,325 to aid comparison.
|c. Some components may not sum up to the totals due to rounding.|
Comparisons of the Census and LFS outputs are difficult because their purpose, scope, frequency and mode of collection vary. Understanding these differences helps explain why observed estimates from the LFS are different from the Census. However, this article illustrates that, while the level estimates are different the distributions of data from both sources are generally similar.
For the most authoritative and recent estimates of labour market information including employment and unemployment, labour force statistics published monthly by the ABS in Labour Force, Australia (cat no. 6202.0 ) are the most appropriate, The Census is valuable for detailed data and analysis on the economically active population and for regional and small population analysis.