RDP 2010-04: Employment Composition: A Study of Australian Employment Growth, 2002–2006 1. Introduction

It has been well established that the distribution of employment changes over the business cycle (for example, see Okun 1973, Borland 2000 and Bils, Chang and Kim 2007). Time series studies show that during upswings, those with lower skills and a lower propensity to supply labour tend to experience proportionately larger increases in employment rates than other segments of the population (and vice versa during downturns). This paper takes a somewhat different approach, using longitudinal data on individuals to examine how the relationships between personal characteristics and labour market status changed in Australia over a period of strong employment growth (Figure 1).

The paper uses data from the Household, Income and Labour Dynamics in Australia (HILDA) Survey, which has been conducted annually since 2001. The survey asks participants for information on a range of personal characteristics and employment status over time. Because some of the key data for our purposes (such as wealth) are only available for 2002 and 2006, we restrict our analysis to a comparison of these two years. Even so, over this period the rate of employment increased steadily and substantially, making it relevant to examining the effect of strong employment growth on people with varying degrees of attachment to the labour market (that is, the unemployed, the marginally attached and those more strictly defined as being ‘not in the labour force’).

The paper is structured as follows. In Section 2, we review existing studies of the relationship between labour force status and personal characteristics, and how this changes as the aggregate employment rate increases. In Section 3, we introduce the data and provide graphical evidence that the employment rates across different characteristics became more evenly distributed between 2002 and 2006. In Section 4, we estimate an econometric model that relates the probability of being employed (versus not employed) to various personal characteristics, and test for a change in these relationships over time. We then extend this analysis in Section 5 by focusing on the relationships between personal characteristics and the probability of being unemployed, marginally attached or not in the labour force (NILF). Again, we test whether these relationships have changed over time. We conclude in Section 6.