RDP 2002-07: An Exploration of Marginal Attachment to the Australian Labour Market Appendix C: Data Definitions

All variables are measured at the wave 1 interview with the exception of the local unemployment rate.

Age is measured at recruitment, and is reported in five-year bands, e.g. 15 to 19 years. We have replaced this with the age at the midpoint of the band.

Male is set to one if the respondent is male, and zero otherwise.

Degree or diploma is set to one if the respondent's highest educational qualification is a higher degree, a post-graduate diploma, bachelor degree, under-graduate diploma or an associate diploma, and zero otherwise.

Vocational qualification is set to one if the respondent's highest educational qualification is a skilled vocational qualification, a basic vocational qualification, or other post-school qualification, and zero otherwise.

Incomplete secondary is set to one if the respondent's highest educational qualification is less than the highest level of secondary school or the respondent is still at school. The small number of respondents still participating in education were included in this category since parameter estimation was unstable if this was included as a separate variable. Exclusion of such individuals would not be desirable since the decision to participate in education is related to perceived labour market opportunities (Lewis and Koshy 1999).

Couple family is set to one if the respondent's family type is a couple family, and zero otherwise.

Dependent children is set to one if the respondent has a dependent child aged under 15 years of age in the family, and zero otherwise.

Has a disability is set to one if the respondent has a disability, and zero otherwise.

ESB migrant is set to one if the respondent is a migrant from an English speaking country, and zero otherwise.

NESB migrant is set to one if the respondent is a migrant from a non-English speaking country, and zero otherwise.

Year of arrival in Australia is the number of years since arrival in Australia. It takes the value of zero for respondents who were born in Australia.

Older is set to one if the respondent is aged 45 years or older, and zero otherwise.

Local unemployment rate is captured by the rate of unemployment in the respondent's postcode at the time of the 1996 census. Given that the only geographic data in SEUP are the variables which indicate the decile of socio-economic status (SES) for each postcode at the time of the 1991 census (i.e., the so-called ‘SEIFA index of relative advantage’), the local unemployment rates are averaged out for each SES decile using standard ABS concordance files and allocated to each SEUP respondent. The SEIFA index is not defined for 49 postcodes throughout Australia. However, the loss of information is not large given this number reduces to only 19 missing postcodes if the focus is on areas with valid data on unemployment rates.

An alternative measure was constructed to examine the variation in unemployment rates over the survey period. This second measure uses quarterly ABS data on labour force status in each Labour Force Region, which is distributed to each postcode using weights constructed from Department of Social Security data on unemployment-related benefits and 1996 census data on working-age population and labour force. Between 1714 and 1996 unemployment-related beneficiaries failed to provide valid postcode data. This is miniscule compared to the number of unemployed throughout Australia (<0.25%). All weights were normalised so that they summed to one in each Labour Force Region. An estimate on the quarterly variation in unemployment (as well as labour force and population) in each postcode is linked to SEUP information on the SES of an area. Compatible data from all sources were only available for seven quarters. One issue is that the LFS started using the 1996 census geography from September 1997 and hence it is not possible to accurately match areas using their 1991 postcode information provided in the SEUP. This exercise revealed that there was little variation in local unemployment rates, at least within each decile of SES, in the period examined (data available from authors on request).