RDP 1999-12: Unemployment and Skills in Australia 5. Conclusions

This paper asks two questions. Firstly, what has happened to the market for skilled and unskilled labour in Australia as the aggregate unemployment rate has risen? Secondly, why is the unskilled unemployment rate substantially higher than the skilled unemployment rate, not just in Australia but also in virtually all OECD nations?

In response to the first question, demand for skilled labour has increased markedly in Australia in recent years. This is true whether skill is measured by educational attainment or by occupational skill level. However, wage relativities and unemployment relativities between skill groups have not changed substantially, because the supply of skilled labour has thus far kept pace with the shift in demand. There has been a large increase in the aggregate unemployment rate in Australia since the 1960s; however, this has not been disproportionately focused on the less skilled, and appears to be due to aggregate labour market factors.

In response to the second question, the high unemployment rate of less-skilled workers is associated with two factors: (i) a less-educated worker has a greater probability of exiting employment each period (a high ‘separation rate’); and (ii) once not employed, a less-educated worker has a smaller probability of finding employment (a low ‘matching rate’). The high separation rate for low-skilled workers is consistent with the explanation that less-skilled workers have less firm-specific human capital than skilled workers. However, this does not explain the low matching rate. Three explanations for this low matching rate are considered: that skilled workers are able to search for both skilled and unskilled jobs, that unskilled workers have higher replacement ratios than skilled workers, and that institutional factors or equity considerations have led to a compressed wage distribution. It is not clear which of these explanations is most important, although all three are consistent with various features of the Australian data.

A final comment. One approach sometimes advocated for reducing the natural unemployment rate is that aggregate wage restraint can be achieved by reducing wages at the lower end of the wage distribution – this in turn would place downward pressure on the distribution as a whole (Lowe 1998). The evidence presented in this paper suggests that in the past, wage pressures for less-skilled workers have been in line with aggregate wage pressure. However, this does not necessarily suggest that future attempts to reduce unemployment should disregard solutions focusing on the wages for one group or another. Whether such ‘relative wage restraint’ soutions are the most appropriate way of reducing aggregate unemployment is a question left for future research.