Academic journal article The Economic and Labour Relations Review : ELRR

Outsourcing and the Use of Contractors: Evidence from the AWIRS

Academic journal article The Economic and Labour Relations Review : ELRR

Outsourcing and the Use of Contractors: Evidence from the AWIRS

Article excerpt


While the growth in importance of outsourcing and its implications for labour market arrangements have drawn widespread comment (e.g., Boreham, 1992; Gome, 1998; Quinlan, 1998), there is relatively little information or detailed research concerning the phenomenon in Australia, with most attention from researchers focused on the challenge contract labour poses for established conceptions of labour law (e.g., Brooks, 1988; Stewart, 1992; Underhill and Kelly, 1993; Creighton, 1994). Notable exceptions, however, are studies conducted by the National Institute of Labour Studies (NILS) for the Australian Taxation Office in 1994 (VandenHeuvel and Wooden, 1995; Wooden and VandenHeuvel, 1996) and a 1995 comparative study of Australian and New Zealand businesses reported in Brosnan and Walsh (1998). Each of these studies involved large surveys, either of households or of business establishments, and lead to the conclusion that the use of contractors by Australian businesses is relatively widespread, though still only accounts for a small minority of the Australian workforce. (1)

The first of the NILS studies, for example, involved an interview-based survey of households conducted by the Australian Bureau of Statistics as part of its May 1994 Population Survey Monitor. The sample was obtained using techniques parallel to those used in the monthly labour force survey, and hence the data collected should be representative of the Australian population. (2) The results from this survey, reported in VandenHeuvel and Wooden (1995), suggest that about 7.5 per cent of all Australian workers employed in the non-farm sector might be classified as self-employed contractors, though just over one-in-five of these so-called self-employed contractors actually described themselves as wage and salary earners.

If such estimates are to be believed, then the importance of outsourcing must be large indeed, given this estimate was only intended to cover the self-employed. To these self-employed must be added the many persons employed on a conventional basis by organisations that specialise in providing contract labour services. The second NILS study, reported in Wooden and VandenHeuvel (1996), involved a workplace-based survey that was designed to provide estimates using this wider definition. The results from this survey suggest that together, the various types of outsourced labour represented just over 10 per cent of total labour requirements in the non-farm sector. The sample used, however, was not representative of the total population. Most obviously, the sample was restricted to firms employing at least 100 employees. Further, this study is subject to the usual problems associated with response bias, with completed responses received from just 522 of the initial sample of 1634 businesses.

Brosnan and Walsh (1998) also reported on results from a workplace-based survey, but one which employed a much larger and more representative sample than that used in the Wooden and VandenHeuvel (1996) study. This study, however, is still subject to the criticism that responses may not be random, with useable responses received from 1440 workplaces. Given an initial 'effective' sample numbering 4588, this gives a response rate of 31 per cent (almost identical to the response rate achieved in the NILS workplace-based survey). The results from this survey suggest that contractors and consultants, defined as persons who contract to provide labour services to an organisation, but who are not direct employees of that organisation, represented only 4.2 per cent of the total labour requirements of Australian businesses.

These findings are thus at odds with the estimates reported in Wooden and VandenHeuvel (1996). Which is to be preferred, however, is not obvious. On the one hand, the more representative nature of the sample used in the Brosnan and Walsh (1998) study suggest that its estimate should be more reliable. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.