Labour market issues

and Bruce

Table of Contents

Strengths of the 2002 NATSISS
An illustration of new information available from the 2002 NATSISS
A limitation of the survey
Concluding comments

The continuing low employment rates and general labour market disadvantage of Indigenous Australians have been well documented (Altman & Nieuwenhuysen 1979; Daly 1995; Hunter 2004a). However, our understanding of the reasons for this labour market disadvantage is constrained by the limited data available for the Indigenous population. This lack of understanding hampers the development of labour market and related policies to improve labour market outcomes for Indigenous Australians.

Before the collection of the 2002 NATSISS, the main source of data on Indigenous labour force status—and the only sources of data that could be used to reliably measure change—have been the five-yearly censuses from 1971 to 2001. [1] While the census data can provide valuable information on trends in labour force status, working hours, occupation and industry, there is very limited information on other important labour market topics such as the duration of unemployment, difficulties experienced in finding employment, and the identification of discouraged workers. Furthermore, the census has very limited or no data on a range of economic, demographic, social and cultural factors which are likely to be important in explaining labour market outcomes.

The only other nationally representative data on Indigenous Australians is the 1994 NATSIS. [2] Although the 1994 NATSIS provides data on a much wider range of topics than the census, these data are now over a decade old, and the 2002 NATSISS provides a valuable new source of information on labour market issues.

The 2002 NATSISS collects similar information to the 1994 NATSIS on labour market issues, so it represents a valuable and timely addition to data sets with information on Indigenous labour market outcomes. In broad terms, the information on key labour market variables is comparable between the 1994 NATSIS and 2002 NATSISS, allowing for changes over time to be assessed. [3]

The purpose of this chapter is to provide an overview of the labour market information available in the 2002 NATSISS and to describe some of the key strengths and limitations of the data. In order to illustrate the value of the 2002 NATSISS, three examples are offered of highly useful types of data that are available.

While no data set is ideal, we consider in some detail an important limitation of the 2002 NATSISS data. This relates to the omission of key variables from the data set, specifically labour market experience and the length of time spent with the current employer. We use an alternative data set with information on labour market experience to illustrate the potential significance of its omission for statistical analysis of both wages and joblessness. The value of our method is that it can be applied to illustrate the significance or otherwise of the omission of other variables from the 2002 NATSISS.

Other chapters in this volume (Biddle & Hunter; Webster, Rogers & Black) and a number of ABS publications provide a detailed overview of the 2002 NATSISS, including sampling, exclusions issues and non-sampling matters. In this chapter, discussion of these issues is limited to those that are specifically related to the labour market data in NATSISS.

Strengths of the 2002 NATSISS

There are three main areas in which the 2002 NATSISS data has advantages over the census for the analysis of labour market issues.

First, the 2002 NATSISS accurately identifies CDEP scheme employment, which is not the case in the census. [4] This is a major limitation of the census because the CDEP scheme represents a crucial difference between Indigenous and mainstream labour market experiences. Under the scheme, funding is allocated to CDEP organisations for remuneration for participants at a level similar to, or a little higher than, income support payments, with the finances being enhanced with administrative and capital support. It is thus used as a means to provide employment, training and enterprise support to Indigenous participants (see Altman, Gray & Levitus 2005 for a detailed discussion of the CDEP scheme). To illustrate how important the scheme is, we note that in 2002, employment in CDEP accounted for over one-quarter of the total employment of Indigenous Australians, with around 13 per cent of the Indigenous working-age population being employed in the scheme.

The importance of identifying CDEP employment for different areas of Australia is illustrated in Table 10.1, which shows Indigenous labour force status by region using the 2002 NATSISS. [5] In non-remote areas, just 4.7 per cent of the Indigenous working-age population was employed in the CDEP scheme. In these areas, failure to take account of CDEP employment is likely to have a relatively small effect. But in remote and very remote areas, 16.9 per cent and 42.2 per cent respectively of the working age population was employed in the scheme (see Table 10.2).

Table 10.1. Indigenous labour force status by region, 2002




Very remote







 CDEP employed




 Mainstream employed




Total in the labour force




Population (no.)a

196 300

23 100

49 850

a. Table population is Indigenous persons aged 15–64 years.

Note: The remote areas in this chapter, in contrast to most other chapters in this monograph, refers to remote areas that are not classified as very remote by ARIA. That is, tables are not derived from ABS (2004c), which generally provides aggregate results for all remote areas.

Source: Customised cross-tabulations from the 2002 NATSISS (derived from Altman, Gray & Levitus 2005: Table 1)

Using the 2002 NATSISS, it is possible to estimate the effects of CDEP employment on a range of important outcomes, such as income and working hours. With the 1994 NATSIS, it also allows for analysis of trends in labour force status (including non-CDEP employment) to be identified with more confidence than has been previously possible using census data combined with administrative data. It is also possible to analyse changes in the determinants of mainstream employment at an individual level (although not for the same individual, which would require longitudinal data).

Further, the 2002 NATSISS can also be used to estimate the associations between CDEP employment and a range of social, health and cultural variables.

The second major advantage of the 2002 NATSISS is that, for the first time, analysis of labour market issues is possible in very remote areas of Australia. The ability to do this is highly valuable because the labour market context of very remote areas (and, to a lesser extent, remote areas) is very different from those in the rest of Australia, for reasons now discussed.

First, Indigenous people in very remote areas are often living in communities in which the majority of the population is Indigenous. Second, these communities are in sparsely populated regions of Australia which are extremely distant from markets, both geographically and culturally. Third, these regions were colonised relatively late, with some parts of Arnhem Land and central Australia as recently as during the last 50 years. This has meant that customary (kin-based) systems and practices remain robust and there is ongoing contestation between mainstream Australian and Indigenous world views.

Furthermore, according to conventional economic and social indicators, there is a growing disparity between Indigenous people living in remote areas and both Indigenous and non-Indigenous Australians living in non-remote areas (ABS 2004a, 2004c). There is evidence that some discrete Indigenous communities in remote Australia are in economic and social crisis.

The different labour market context in respective regions is illustrated clearly by the fact that in non-remote areas the mainstream employment rate is 41.2 per cent, in remote areas 31.7 per cent, and in very remote areas just 14.9 per cent (see Table 10.1). Human capital and demographic characteristics also differ dramatically across regions. For example, education levels are much lower in remote and very remote areas than in non-remote areas, and the proportion of the population speaking an Indigenous language is much higher in remote areas than in non-remote areas. These factors are bound to influence the nature, variance and quality of Indigenous labour market experiences and it is a real bonus that the information is part of the 2002 NATSISS.

A third advantage of the 2002 NATSISS is that it contains information on a wide range of somewhat unusual social, demographic, cultural, and economic variables which are potentially important for understanding labour market outcomes. Examples include health status, speaking an Aboriginal language, having used an employment service, access to transport, and having been arrested. Note that many of these variables are not available from the census.

Table 10.2. Labour market data collected in the 2002 NATSISS

Labour force status

Employment support

 Duration of unemployment

 Whether used employment support services

 Hours usually worked in all jobs

 Whether needed employment support services

 Full-time/part-time status

 Reasons did not use employment support services

 Employment sector


 Precariousness – job security in next 12 months

Level of income

 Whether work allows for cultural responsibilities

 Personal gross weekly income

 Household gross weekly income


Source of income

 Whether CDEP participant

 All sources of personal income

 Duration on CDEP

 Main source of personal income

 Considers CDEP participation to be a job

Government pension/allowance

Barriers to employment

 Type of government pension/allowance (primary)

 Whether had difficulties finding work

 Type of government pension/allowance (auxiliary)

 All difficulties finding work

 Government support

 Main difficulty finding work

 Time on government support in last two years

Discouraged jobseekers


 Whether would like a job


 All reasons not looking for a job


Source: Derived from ABS (2005b)

In general, it appears that questions relating to labour market topics are very similar (virtually identical) in the community and non-community questionnaires (see ABS 2005b and the list of variables in Table 10.2). While there may be some effects generated by differences in the data collection method (CAPI versus paper-based questionnaire), we do not anticipate this will have introduced major biases. While those analysing the data will need to carefully consider the extent to which the remote and non-remote data is comparable for their particular application, our reading of the questionnaires suggests that there is no particular reason for expecting there to be comparability issues.

The questions are also, in large part, standard ABS questions. This allows comparative studies of labour market outcomes for Indigenous and non-Indigenous Australians using the 2002 NATSISS and other data sets, such as the GSS.