Appendix 1: Data sources and variables used

This study uses extracts of data taken from the quarterly Household Labour Force Survey (HLFS) that cover the period March 1991 to September 2005, (the last quarter that the Department of Labour had detailed extracts available for use at the time of this study). This source was chosen because it offers a long-term data source and can reveal the turning points in what is often a very gradual change. There are limitations with this data source. In particular the classification of industry/ occupation data has changed over time and there is evidence that the reliability of occupational data in the HLFS is affected to some degree by non-sampling (e.g. coding) error (Spier 2006).[12]

The HLFS variables considered in this study included the following:

• ethnicity
• gender
• full-time/part-time employment status
• hours usually worked
• industry classification
• occupation classification
• post-school qualification. (Yes or No).

Family code was also looked at but did not appear to be reliable. This dimension would be useful in order to better understand the family dynamics that may influence participation in an older aged population (especially females).