APPENDIX QUALITY INFORMATION
Information about the labour force characteristics of the Indigenous population has been collected in the LFS using international standards for defining and measuring employment and unemployment. The distinction between paid and unpaid work is fundamental to these definitions. As a result, the definitions may not always give adequate recognition to the 'unpaid' activities which some people undertake, for example as part of social and cultural obligations. However, the use of standard definitions across all population groups is seen as important for ensuring objectivity and uniformity in measuring Australia's labour force. This section provides information on issues specific to Indigenous estimates from the LFS. Information on the methodology and concepts used in the LFS more generally are presented in the Explanatory Notes section with additional detail published in Labour Statistics: Concepts Sources and Methods (cat. no. 6102.0.55.001).
In the LFS, information is generally collected by specially trained interviewers using a standard LFS questionnaire, from people in dwellings selected in the LFS. The questionnaire used in the survey is published in Information Paper: Questionnaires used in the Labour Force Survey (cat. no. 6232.0).
Special procedures are used to collect information in some Indigenous communities in remote areas to minimise respondent load and achieve maximum cooperation and response rates. In particular, a 'short form', which collects the minimum data required to derive basic labour force characteristics, is available for use when interviewers encounter significant cultural, language or operational difficulties in remote areas.
From April 2001, the monthly LFS has included a question on Indigenous status. This has enabled information on the self-identified Indigenous status of respondents to be collected for each month of the survey, increasing the amount of information available on Indigenous people from the LFS.
The question asked to determine Indigenous status is:
Although the question on Indigenous origin allows people to identify as being of Aboriginal origin only, Torres Strait Islander origin only, or as both Aboriginal and Torres Strait Islander origin, it is not feasible to derive separate labour force statistics for these three groups because of the small numbers of Indigenous people in the LFS sample.
- Is ... of Aboriginal origin, Torres Strait Islander origin or both?
The LFS has a monthly sample of around 64,000 people aged 15 years and over, in urban, rural and remote areas across all states and territories of Australia. This sample size is sufficient to provide reliable monthly estimates of level and change in the major labour force characteristics of the total population for Australia and for each state and territory. However, the survey sample is not specifically designed to provide monthly estimates for small population groups, such as the Indigenous population, or for small geographic areas.
The methodology used to produce the labour force estimates for Indigenous Australians from 2002 onwards is based on combining the LFS samples for each of the twelve months of the year. This pooling has resulted in a sample of between 11,500 and 14,000 records each year. This sample size is adequate for providing broad aggregates of labour force characteristics at the state and territory level. However, the sample size per year is still quite small, resulting in constraints on the level of disaggregated data available from the survey. Note that as the sample is pooled from monthly data throughout the year, and most people are interviewed for the LFS in more than one month (up to eight consecutive months), the number of individuals interviewed each year (e.g. 3,500 in 2006) is substantially lower than the sample used in producing the estimates (12,600 in 2006).
In remote areas of Australia, the LFS sample is more clustered than in non-remote areas, so that the majority of the survey's sample of Indigenous people living in remote areas tends to be concentrated in a relatively small number of clusters. This has significant implications for estimates for Indigenous people in remote areas. There is a high degree of variability in employment opportunities between Indigenous communities, while employment opportunities for people within a particular Indigenous community are likely to be similar. A key factor here is whether the Community Development Employment Projects (CDEP) scheme has been implemented locally. CDEP schemes operate in many remote communities and can be a major source of paid employment for community residents. Conceptually, the participants in a CDEP scheme are considered to be employed in the LFS. However, the LFS does not currently separately identify whether or not an Indigenous person or community is participating in a CDEP scheme.
Not all Indigenous communities in remote areas participate in the CDEP scheme. In a community with a CDEP scheme, the majority of adult members of the community could have an 'employed' labour force status when surveyed in the LFS if they worked as part of CDEP in the reference week. Because of the small number of Indigenous communities selected in the LFS, there is the potential for wide statistical fluctuations in estimates of labour force characteristics of remote Indigenous communities over time if the sample moves from communities participating in CDEP to those which do not (and vice versa).
As part of the estimation process, the LFS estimates of Indigenous persons employed, unemployed and not in the labour force are calculated in such a way as to add to independent estimates of the size and structure of the Indigenous population. Independent population estimates (benchmarks) for the Indigenous population aged 15 years and over by sex, broad age groups and remoteness type, for each state and territory of Australia, as at 30 June, were used for each year of estimates. These benchmarks reflect the low series population projections based on the 2001 Census as published in Experimental Estimates and Projections, Aboriginal and Torres Strait Islander Australians, 1991 to 2009 (cat. no. 3238.0). These projections are not predictions or forecasts, but are projections of the latest Census data reflecting the growth in population which would occur if certain assumptions about future levels of fertility, mortality, migration and unexplained growth were to prevail over the projection period. Where information in this publication is presented by geographic area (such as remoteness, or state or territory), this data is based on a respondent's place of enumeration.
Standard LFS definitions and concepts are presented in the Explanatory Notes. However, there are some issues specific to Indigenous labour force estimates which are discussed here.
In the LFS, employment is defined as working for at least one hour, either for pay or profit, or as unpaid work in a family business; or being absent from work but remaining formally attached to a job. Persons who participate in labour market programs are counted as employed, unemployed or not in the labour force according to how they respond to questions in the LFS about their actual activity in the week before interview.
The CDEP scheme is a program provided by the Federal Government for (primarily) Aboriginal and Torres Strait Islander people living in remote, rural and urban areas. It enables an Indigenous community or organisation to pool the unemployment benefit entitlements of individuals into direct wages for those people who choose to participate in local employment in various community development or organisation programs as an alternative to receiving individual income support payments. The relationship between CDEP organisations and the individual participants who are undertaking paid work is treated by ABS as an employer-employee relationship. The individual participants are considered to be in paid employment, even though they are paid for their work from funds originating as unemployment benefits. It is unclear at this stage whether changes to the CDEP scheme resulting from recent government reviews will impact on how such activity is classified by the ABS in the future.
The definition of unemployment is based around the concepts of actively looking for work and being available to start work. Where job opportunities are scarce or may not exist, such as in some remote parts of Australia, there may be little incentive to actively look for work. As a result, there may be relatively little unemployment measured in such regions, with the population who are not employed falling mainly into the 'not in the labour force' category.
RELIABILITY OF ESTIMATES
When reading the following summary of findings, or using the data presented in the tables section, it is important to consider the relative strengths and weaknesses of the LFS as a data source for Indigenous labour force information. It is also important to remember that all LFS estimates are subject to sampling errors. This is particularly relevant to these Indigenous estimates because of the small size of the Indigenous population and the small number of Indigenous persons in the LFS sample. These estimates are also subject to non-sampling errors due to particular collection difficulties in remote areas, in addition to imperfections in reporting, recording or processing of data that can occur in any survey or census. In addition, it is also important to note that the population benchmarks used for compiling Indigenous labour force estimates are not predictions or forecasts, but are projections of the latest Census data reflecting the growth in population which would occur if certain assumptions about future levels of fertility, mortality, migration and unexplained growth were to prevail over the projection period.
The Technical Note provides more information about the reliability of estimates based on a sample survey. However, in general terms, very small estimates are subject to high sampling errors (relative to the size of the estimate). Estimates with a relative standard error (RSE) of 25% or greater would not be suitable for most practical purposes. These estimates are indicated by an asterisk (*) in the tables presented in this publication. The actual standard errors for level estimates are provided in the Technical Note tables L1 to L5.
Comparisons over time of Indigenous labour force characteristics should be interpreted with particular caution. It is important to take into account the high standard errors associated with annual movements. For example, the estimated number of Indigenous people unemployed decreased from 29,000 in 2005 to 26,700 in 2006. However, the standard error on this movement (2,700) means that this decrease is not statistically significant. The standard errors of the year to year movements are provided in the Technical Note tables M1 to M5.