The change in labour skills in Australia over the business cycle.

• Author: Kelly, Ross
• Position: Contribute Article

This paper presents an analysis of skill change for each of several skill dimensions for Australia for the period 1991 to 2006. This period is of particular interest since it covers three phases of the business cycle--recession, full employment and excess demand. The pronounced shedding of low skill workers and increased demand for skilled workers observed in many countries over the last two decades has been attributed to a number of different causes. In this paper the attributes of different occupations are used to obtain measures of distinct skill dimensions--motor, cognitive and interactive, plus education. The results indicate that there were very significant changes in skills mix during the three phases of the business cycle. The mean level declined for motor skills but rose for the other dimensions, particularly interactive. The results have important implications for policy, particularly in relation to employment, unemployment and training.

Introduction

The structure of the Australian economy has changed dramatically over the last two decades, with a sustained shift away from agriculture and manufacturing. The growth industries have been in services, consistent with long term trends in advanced and many developing economies. These trends have had a profound effect on the skills that are in demand in the economy which, in turn, has also altered the structure of skills and average skill levels within industries. Other important factors affecting the skill content of the employed work force include significant institutional changes, growth in part-time and casual work, increased participation in post-compulsory education and a greater participation of women in paid work.

Over the last thirty years there has been observed in many countries a pronounced shedding of low skill workers and increased demand for skilled workers (Gautie 2002). This development has been attributed to a number of different causes, the most commonly cited being skill-biased technological change. In addition, the business cycle has different effects on demand for labour with different skills (AFPC 2009). There have been many studies of the influence of new technologies--in particular, information and communication technologies (ICT)--on skill demand. The overwhelming majority of these studies find a positive link, suggesting a high degree of complementarity between skill and technology (Berman et al 1997; Bound and Johnson 1992; Goldin and Katz 1998; Kelly and Lewis 2003; Maglen and Shah 1999; Pappas 1998; Wolff 1995, Spitz 2004).

It has been argued that ICTs change the composition of skills in the economy in two ways. First, the direct substitution of easily automated labour intensive jobs by computer-based technologies will alter the composition of skills. It can also eventuate from the organisational complementarity that exists between computer based technologies and managerial and professional jobs (Autor et al 1997, 2000; Caroli 1999; Gera and Masse 1996; Mukoyama 2003).

It is well-known that the business cycle does not affect all labour skills equally. The evidence comes mainly from studies in the 1990s following the 1991 recession (Lewis and Seltzer 1996; Borland 1999). For instance, recessions disproportionately impact on the lower skilled and increase the likelihood of lower skilled workers becoming unemployed (AFPC 2009). Also, the long-term unemployed tend to have lower levels of skills, education and experience; and long-term unemployment tends to exacerbate the underlying problem, leading to hysteresis (Lewis and Seltzer 1996). Finally, 'structural mismatch' worsened during the most recent long period of high unemployment, and this mismatch continued even after 17 years of uninterrupted economic growth (Lewis 2008).

In this paper the attributes of different occupations are used to obtain measures of four distinct skill dimensions--motor skills, interactive skills and cognitive skills and education. Motor skills are essentially the ability to do physical tasks. Cognitive skills relate to the possession of and ability to create knowledge. Interactive skills refer to the ability to relate between managers and employees, employees and employees, and employees and customers. Education is the measure typically used in labour economics as a measure of skill. The paper presents an analysis of skill change for each of these aspects of skill for Australia for the period 1991 to 2006--a period containing all stages of the business cycle. The analysis examines the pattern of industry skill demand by analysing skill changes separately for full-time and part-time workers and for the sub-periods 1991 to 1996, 1996 to 2001 and 2001 to 2006.

Employment in the Australian Economy

Australia experienced a deep and protracted recession in the early 1990s. The period of recovery left a large number of the low skilled workforce in long-term unemployment or marginalised employment (Norris and Wooden, 1996). Since the 1991/92 recession, and up to the Global Financial Crisis in 2008, there was a remarkable period of growth, impacting on total employment in Australia (Lewis 2008). This, however, has followed very different trajectories across different industry sectors and occupations (Table 1).

The overall growth in employment has differed between industries over the period from 1991 to 2006 (see Table 1) and the changes have occurred at different stages for a range of industries, suggesting that different drivers are influencing the structural changes. For example, the utilities sector (electricity, gas and water) experienced substantial re-structuring and labour shedding throughout the 1990s as a result of the partial de-regulation and privatisation of the sector. Manufacturing has experienced steady decline over the entire period, due for the most part to reduced protection and the increasing productivity and sophistication of manufacturing in Asia, particularly China. This has enabled a wider range of consumer goods to be cost effectively sourced from overseas.

Growth in each of the sub-sectors of manufacturing in terms of employment since 1991 has been mixed. The standout is a 60 per cent decline in the number of people employed in textile, clothing, footwear and leather (TCFL) manufacturing. Employment in TCFL fell from 10.8 per cent of total employment in manufacturing in August 1991 to 3.9 per cent. Since 2001, the demand for raw materials to feed the rapid industrialisation in China and India has worked its way through the resource rich states of Australia (mainly Queensland and Western Australia). This has been a boon for industries more directly aligned with the fortunes of the mining sector, such as construction, where employment has risen sharply. The impact of the mining boom on the economies of the resource states was so pronounced that they have essentially been running at over-full employment and have recruited heavily from overseas to fill skill shortages (Lewis and Corliss 2009). Construction employment was also significantly boosted by Australia's population boom, fuelled by increased immigration. The changing structure of industry will, in principle, alter the pattern of demand for occupations and, as a consequence, the skill types and skill levels in demand. This needs to be taken into consideration by both industry and training authorities in their forward planning. As shown in Table 2, the mix of occupations in the Australian economy has also changed quite dramatically. Total growth in employment for the 15 years to 2006 was just over 1.8m persons, with the higher skilled occupational groupings growing significantly faster than the traditional blue collar unskilled and trade-qualified occupations. In the occupation grouping, advanced clerical and service workers, total employment actually fell. The scale and duration of the economic expansion in Australia make this reduction quite remarkable.

The...