Temporary Agency Work and Precarious Employment: A Review of the Current Situation in Australia...

Introduction

Over the past two decades, there has been extensive growth and internationalisation of the employment agency sector, with organisations such as Kellys, Drake, Randstat, Manpower and Adecco being established within most OECD countries. The spectacular growth of the industry has made it more visible while alerting trade unions and regulators to its potential implications for employment conditions. In Australia and New Zealand, there have been several public inquiries into the sector, in part this has reflected the general ignorance surrounding its dimensions and the implications for public policy (Labour Hire Task Force 2001; Productivity Commission 2005).

Despite its visibility, agency work is an enigma for the analysts - it is part of the romance of self employment, flexibility and freedom, and it is also part of a process that undermines employment conditions, collectivism and employee rights. It is linked to new work and to the new economy, yet it is also linked to traditional areas of 'temping' such as seasonal and replacement work (Burgess/Connell, 2004a). There are clearly many research angles which, in this paper, are restricted to discussion of three questions:

* How is agency work regulated in Australia and New Zealand and is the regulatory approach likely to change?

* Does the relative share of agency work of total employment match its visibility and flexibility and what may limit employer choice?

* Why do people choose agency employment if it is a very precarious form of work?

Regulation, labour flexibility and temp agency work

The cost effectiveness, flexibility and precariousness of agency work is closely linked with regulation and regulatory effectiveness. The confusion and uncertainty surrounding agency work has resulted in some countries having extensive regulations governing both the operation of agencies and the agency employment contract (De Ruyter 2004; Storrie 2002). This is clearly not the case in Australia and New Zealand which have, according to Burgess et al. (2004a), a 'wild west', unregulated approach. While this article explores the regulatory status of agency work, it also raises the question of why agency work is unregulated in Australia and New Zealand. It is a rather paradoxical situation considering the comprehensive work regulations associated with the centurylong tradition of arbitration and conciliation systems and well-established union movements. Although the recent rise in agencies and agency work may have benefited from the dismantling of the arbitration and conciliation systems alongside a dramatic weakening of union influence, the high visibility of agency work make it an obvious target for future regulatory interventions. Thus, this article debates the likelihood of further regulatory constraints being imposed on agencies and agency work.

Discussions of agency work have highlighted the high degree of flexibility that it offers employers. The rise of the temporary agency sector has illustrated how organisations can shed their internal labour costs and shift their recruitment, training and on-costs to the agencies and the agency workers (Storrie 2002). Where organisations retain 'core' employees they can use agency workers as a way of combining numeric and functional flexibility which makes a broad range of skills available to the organization (Burgess/Connell 2004b). Murtough and Waite (2000: 27) suggest that agency workers are more attractive to user firms where recruitment and training costs are low, demand is irregular, output cannot be stored and where the cost of a poor match between employee and employer is high.

However, research has shown that agency work is probably a relatively small proportion of overall employment in Australia and New Zealand. As agency work offers organisations many advantages that are associated with labour flexibility (Hall 2005), why then does it account for a relatively small share of the workforce? In media and agency reports, there is plenty of hype about positive job opportunities and the strong growth of agency work but the ambiguity and uncertainty surrounding it means that it can be very difficult to estimate its extent and effects. Thus, it is evident that there is a lack of any solid, statistical information surrounding the extent of agency work.

Another paradox is the contrasting views of glamour and precariousness associated with agency work. The glamour of self-employment, 'executive leasing' and freedom to choose one's own work-life balance contrast with the uncertainty, 'disposability' and income/benefit risks that are associated with agency work. While it is clear that these images will have different validity for different groups of agency workers, it is unclear how representative the images are across the agency workforce. In this article, the authors attempt to gauge the relative importance of precariousness noting that perceptions of insecurity can change over time. While this article takes a somewhat skeptical approach to the extent and sustainability of 'lifestyle temps', it is important to note that there has been a significant rise in recent research findings which stress these worker preferences.

Agency employment stands out among employment arrangements as being a very ambiguous and uncertain employment arrangement. The agency intermediates between the worker and the hiring organization. A triangular employment relationship is established that incorporates sub-relationships between the agency and the worker, the agency and the hiring organization, and the hiring organization and the worker. The agency intermediates as a broker and is typically paid a fee for service for each placement. By inserting itself into the employment relationship, it becomes debatable who the employer is in relation to any engagement: is it the agency, or is it the hiring organization? Such ambiguity means that the assignment of rights and responsibilities in the employment relationship can generate confusion and create gaps in which employment arrangements and workforce collectivization can be restructured.

The other area of ambiguity is over employment status. Not all agency workers are employees, some may be placed as self-employed contractors. As such, these workers do not receive the protection and benefits associated with employee status. Moreover, since they are not employees, they are located outside the domain of trade union membership.

Research has found that, in Australia and New Zealand, a similar ambiguity and uncertainty surrounds the status, and assignment of rights and responsibilities within the employment relationship. As shown in the next section, this is also related to the number of agency jobs and workers. In Australia, there have been several legal cases and industrial disputes surrounding the use of agency employment arrangements that appeared to erode the conditions of employees and to de-unionize workplaces (Underhill/Kelly 1993; Stewart 2002). It is also problematic that agency work or 'temping' is normally classified as 'casual' or temporary employment with the worker being engaged on a very short-term basis. As a consequence, the worker does not qualify for employment rights (such as notice of employment termination), nor does the worker qualify for employment conditions that are associated with continuity of service, the major ones being holiday and sickness benefits (Campbell/Burgess 2001).

In New Zealand, the employment status of agency workers is very fluid. Predominantly, agency workers are employees of the agency: 'The temp is legally employed by the agency, which invoices the client organization for the hours she has worked, and pays her wages and other associated benefits as agreed.' (Alach/Inkson 2003: 6). However, there are also many agency workers that have independent contractor status or oscillate between the two types of employment roles. The status of temporary or independent contract worker is one selected by the worker and usually follows the type of status they have previously experienced and anticipate in the foreseeable future. (RCSA 2000: 4). Additionally, agency workers can be self-employed or an employee at times when they are not 'working' for an agency (Rasmussen/Deeks 1998).

Agencies are faced with virtually no regulatory constraints. In both Australia and New Zealand, there are no reporting obligations, financial bonds do not have to be posted by the agencies and there are no limitations on the occupations/industries that can be covered through agency employment arrangements. The agencies also have fairly wide leverage in terms of the contract of employment, the ability to re-engage agency workers and over the employment status of agency workers. Finally, agencies have contributed directly and indirectly to the recent sharp decline in union density. Non-unionized agency workers have been a source of friction between trade unions and particular employers in Australia where contracted agency labour has been used in an attempt to replace unionized employees (Campbell/Watson/Buchanan 2004).

This ambiguity and uncertainty lends itself towards exploitation and the development of appropriate and industry-specific regulations. This has already occurred in some countries, where there has been a development of extensive regulations governing both the operation of agencies and the agency employment contract (De Ruyter 2004; Storrie 2002). In countries such as Germany and Italy, as well as at the European Union level, considerable regulations have been introduced concerning agency work. The regulation of agencies and the agency employment contracts can place limits on the length of engagement, restrict re-engagements, limit the proportion of agency workers relative to employees, require agencies to provide financial assurances, clearly assign employee rights and employer responsibilities, and ensure that minimum wage and other employment conditions are not contravened by the agency employment contract (Burgess/Rasmussen/Connell 2004b). Clearly, this has not been the case in Australia and New Zealand which raises the obvious question of why not?

An immediate answer could be that the two countries have just taken a characteristic 'liberal' Anglo-American path with few regulations of agency work. This would fit with general discussions of national differences where diverse Varieties of capitalism' with a range of institutional patterns prompts distinct regulatory paths (Hall/Soskice 2001). The unregulated approach of Australia and New Zealand fits with the 'liberal market' version often assigned to Anglo-American economies. It also fits with Whitley's (1999: 41-44) 'compartmentalized national business system archetypes'. While a discussion of the two national approaches within these general explanations could warrant a journal article in itself, it seems a problematic answer and it does not appear to fit well with the historical and institutional employment relations traditions of Australia and New Zealand.1

In both countries, the long-term establishment of an arbitration and conciliation system provided a strong regulatory model which fostered detailed regulatory interventions. However, the situation regarding temporary agency employment can hardly be considered a 'liberal' approach. While it may have 'compartmentalized' employment relations within the Whitley conceptual framework it did so based on the strong presence of state sponsored institutional intervention. Rather the answer appears to be linked to the focus of regulatory interventions and the timing of the rise in agency employment. Historically, regulatory interventions in Australia and New Zealand focused on collective bargaining and standard employment. While atypical employment was regulated in many awards and collective agreements it is mainly focused towards the protection of standard forms of employment. For example, casual employment was tolerated as long as there was certain minima applied and it did not undermine standard employment conditions.

Furthermore, there has been a sharp separation between employees and the selfemployed. Thus, there have been areas of the economy where collective bargaining and union activity have been fairly rudimentary. Probably more important though is that the rise in agency work has coincided with labour market deregulation and a sharp decline in collective bargaining and union membership. Especially in New Zealand where there was a radical change to its labour market regulations in the 1990s (Decks/Rasmussen 2002). Moreover, both countries have recorded widespread organisational change in both the private and public sectors and a growth in atypical employment. Agency employment has been fueled by these changes and agencies have benefited from the unregulated 'space' of the two countries' labour markets which have grown considerably over the last two decades (Burgess/Rasmussen/Connell 2004b; Spoonley 2004b).

The market approach has held sway in Australia under the Howard government and it appears set to further escalate in the coming years (Economist 2005; Oram 2005). It is unlikely, therefore, that there will soon be any regulatory interventions covering agency work in Australia. Since 1999, New Zealand has moved in a different political direction under the Clark governments. The recent 'Europeanisation' of New Zealand employment relations has witnessed a string of public policy reforms and employee entitlements and protective measures have increased considerably (Haworth 2004; Rasmussen 2004). Given this background, it is remarkable that there has been no attempt to tighten the regulatory requirements of agencies and agency work. What is even more remarkable is that there appears to be no consideration (that is no public policy discussion) of how the precarious aspects of agency work could be regulated. Accordingly, there are no signs of change to the 'wild west' regulatory approach towards agency work within the two countries.

What limits employer use of temp agency work?

It is a fundamental assumption of this article that, contrary to its reportedly strong growth, agency work constitutes a relatively small share of overall employment in Australia and New Zealand. What is still a major research issue, however, is just how widespread is agency work within the two countries. This lack of clarity is also evident in other countries. For example, Storrie (2002) indicates that agency work is not included in the employment typology in many EU countries and that many national authorities responsible for gathering labour market statistics have yet to include agency work in their national surveys. Short-term employment practices associated with labour hire may also be associated with clandestine activities such as the employment of illegal immigrants and income tax avoidance. As such, there is undoubtedly a degree of under reporting of total employment within the temporary work sector.

Overall, there are several problems associated with estimations of agency employment in Australia and New Zealand. First, there are potential differences in employee status as some agency workers may be employed on a self-employed contract basis. Thus, counts of agency employees will understate total agency employment (Campbell/Buchanan/Watson 2004). second, agency assignments may be very shortterm: hours, days or weeks. Since the labour force survey takes stock employment estimates at one point in time, it may also miss the potential short-term assignments associated with agency employment. The problem is that the temporary placements are sometimes for periods that are so short that between the monthly employment counts, many agency jobs have been created and also terminated. This is also influenced by the fact that an individual agency worker can shift between assignments across several agencies, and hold multiple short-term assignments over one week or one month. Since agencies are performing multiple functions linked to the provision of employment services, the placement of workers can cover both labour hire (agency employment) and job placement. These factors leave significant room for estimation errors. For example, the number of job placements by Australian agencies in 2001 has been estimated to be around 3 million, representing about one third of the total employment stock. Yet, it has been estimated that only around three per cent of workers are agency workers (Burgess/Rasmussen/Connell 2004b). Third, the traditional counts of employment that are firm-based may exclude agency workers since workers may be employed or paid by agencies or by hiring organisations. Moreover, agency workers, who may not be regarded as employees of the hiring organization, will not be counted by such measures.

In light of these issues, the Australian Bureau of Statistics (ABS) have asked employees in recent labour force surveys whether they were paid by an agency and this has created a baseline from which to estimate the number of agency workers.

"The 2001 data suggests that 161,800 persons (95,400 males and 66,400 females) can be counted as temporary agency employees. This represents 2.2 per cent of all those classified as employees in the survey (or around 1.8 per cent of the total workforce)." (CampbeU/Buchanan/Watson 2004: 135).

However, as per the previous discussion, the ABS surveys only include employees and many agency workers may be dependent contractors/self-employed. Thus, these survey figures are bound to underestimate the number of agency workers and Campbell et al. (2004: 135) suggest, therefore, that the actual figure could be quite a bit higher "perhaps up to 250,000 persons". On this basis, agency employment could constitute over 2 per cent of the total workforce. Similarly, Burgess and Connell (2004b: 27) estimate that "around 2-5 per cent of employees are agency workers, a minority of whom are paid directly by the agencies". Hall (2005) cites data from the Household, Income and Labour Dynamics in Australia (HILDA) survey which indicated that agency employees constitute 3.7 per cent of all employees. Hall suggests that this estimate would place Australia toward the top end of the proportion of agency workers in OECD workforces (OECD 2002). Overall, it is apparent that there is considerable guesswork involved in estimating the extent of agency work in Australia.

The estimation of agency work is even more problematic in New Zealand where there are no official statistics on the number of agency workers, or indeed, the number of agencies. As there has been growth in New Zealand agencies it is reasonable to assume that there has also been growth in the numbers of agency workers. For example, one industry source has suggested that "the temporary staff placement industry has grown five-fold over the last decade." (Press 2003: 2). The only available figures have been provided by the sector's interest organization, the Recruitment and Consulting Services Association (see RCSA 2000: 5) and, on the basis of these figures, Rasmussen et al. (2004a:167-8) have suggested that agency workers could constitute over 1 per cent of the workforce. As the Recruitment and Consulting Services Association does not include all agencies, however, agency work undoubtedly involves more people than publicised.

One is left with the impression that there is a considerable gap between the visibility of agency work and its actual extent. The above estimates put New Zealand on a comparable basis with a number of European countries while Australia appears to be on par with the United Kingdom. However, there is a considerable jump in the estimated 6 per cent temp agency density in the Netherlands (Storrie 2002). Australian and New Zealand estimates also indicate that there has been considerable growth in agency work. This conclusion is supported by the influx of international agencies as well as a growth in the number of agencies overall. Still, the actual level of agency work is debatable as the available figures appear to contain several flaws which could account for significant under reporting.

Why are employers interested in utilizing temporary agency workers?

Temporary agency employment possesses many of the characteristics associated with 'flexible labour' and for this reason it has obvious attractions to employing organisations. The rationale for hiring temporary workers tends to be the same as many other workplace initiatives - labour cost savings associated with downsizing, increased global competition, the introduction of new technology and the need to respond quickly to an ever-changing marketplace. The rise of the temporary agency sector can be interpreted as one manifestation of increased flexibility in which the restructuring of internal labour markets and lower internal labour costs ('headcount costs') are associated with organizations externally shifting recruitment, training and on-costs to the temporary agencies and temporary workers. Agency employment not only offers flexibility and cost saving possibilities, it also potentially removes responsibility for compliance with many employment regulations such as unfair dismissal, employment insurance, employment benefits and pension responsibilities onto the agency worker. In this context, 'temping' allows for a shifting of the responsibilities and risks associated with direct employer responsibilities.

The ability to avoid legislative requirements, thereby increasing so-called procedural flexibility, is considerable with agency employment. Through agency arrangement employer responsibility can be ambiguous, especially if sub-contracting and multiple hires are involved (Rubery et al. 2000). Trade union influence can also be mitigated since the employment engagement may be short-term or involve a contractor or even a sub-contractor. In an analysis of the increase in agency employment in Australia, the Productivity Commission (2005) suggested that changes in the industrial relations and the competitive business environment were the major reasons promoting increased agency employment over the period 1990 to 1995. Considerable legislative change to industrial relations in Australia has increased the ability of managers to utilise agency workers and to extend managerial prerogative (Burgess/Macdonald 2003). In the past, such practices were moderated and limited by collective agreements or closed shop arrangements with trade unions but it has now become a generic response from workplace managers concerning their rationale for increasing usage of agency workers.

Nonetheless, it would be a mistake to suggest that it is only flexibility advantages that are driving agency employment in Australia and New Zealand. An industry survey of 150 user organisations in Australia reported by Hall (2005) suggested that the main reasons for agency work were: to fill vacancies; reduce administrative costs, provide access to skills and 'labour flexibility'. Interestingly, the use of agency workers was regarded as complementary, rather than as a substitute for, ongoing and internal staffing arrangements. Similarly, the Labour Hire Taskforce (2001) suggested that traditional reasons were largely behind the decision to utilise agency employment arrangements. These reasons included covering absences, meeting unexpected and short-term labour needs and providing specialist skills. In the recent tight labour market, agency work has been strongly connected to ongoing skill shortages in construction, nursing and IT (all relatively heavy 'temp' users) as well as the ongoing restructuring of businesses in both the private and public sectors (Burgess/Rasmussen/Connell 2004b).

In New Zealand, the growth of agencies and agency workers has been driven, to some degree, by demand. The radical and comprehensive reforms in the post-1984 period created larger markets in the central government sector, the newly corporatised and privatised firms, local government and in the health sector. The business cycle especially the major downturns after the 1987 share market crash and after the 1996 Asian crisis - prompted private sector restructuring and further growth in agency work and agencies.

Across the various agency markets in New Zealand there appears to have been a steady growth pattern regardless of business cycles. This growth has been fuelled by organisational restructuring, the emphasis on flexibility and lean staffing, as well as the small size of many New Zealand organisations. Agencies have themselves tried to develop their markets through the promotion of agency work as a flexible and interesting option. There have also been continuous attempts by agencies to create closer relationships with employers (through obtaining preferred supplier status) and becoming 'strategic partners' with their client companies. This has happened to some degree in banking, local government, law firms and hospitals where the 'preferred supplier status' has gone a long way in some cases and in some smaller firms where agencies provide extensive consultative and advisory services (particular when there are no dedicated human resource management staff). This is clearly an area where agencies see further growth opportunities and where they work hard to increase their ability to recruit and develop high quality staff in professional services (such as accounting, nursing, secretarial and administrative work).

It is often suggested by industry sources that agencies are just mediating new forms of labour demands and supply patterns. While there are distinct demand and supply factors (as discussed below), it is important to acknowledge the active role of agencies as they have pursued market opportunities and, through alerting employers and employees to the 'benefits' of agency work. Hence, they have become 'market makers' to some extent (see Peck/Theodore 2004). The positive image of agency work is often based on information from agency sources where cost effectiveness, productivity considerations and 'cutting out the employment hassles' for organizations are continuously repeated (see Grey 1998, Jayne 2004; Press 2003). The growing dominance of international agencies at the top end of the Australian and New Zealand markets is one indication of this.

While agencies are not just passive actors, there has clearly been a growing demand for their services. As employers have reacted to global competition, the introduction of new technology and changing markets through organizational restructuring and work re-organization, it has often been regarded as a necessity to seek further labour flexibility and labour cost savings through the utilization of temp agency workers.

Can employers' limited use of temporary agency workers be explained?

If agency employment offers considerable flexibility advantages to the hiring organisation, and if agency employment can be used as a device for restructuring employment, reducing workforce collectivisation and disciplining employees, then why is the incidence of agency employment relatively low? This is even more puzzling in the case of Australia and New Zealand, where the regulatory environment for agency work remains very light, especially when compared to Europe (de Ruyter 2004). With respect to this apparent paradox, we offer several possible reasons.

The first and major reason is that there are other forms of direct employment available to employers that provide considerable labour flexibility. Although agencies limit recruitment and selection costs for employers they also have a considerable 'mark-up' in terms of commissions. Why would employers, except in the short-term, pay an on-going premium for employing staff? This appears to be a major reason why agency work in Australia and New Zealand only constitutes a minor proportion of the workforce. The rise in agency work has been part of a much larger increase in other forms of atypical employment such as part-time employment, self-employment and casual employment. For example, the incidence of part-time employment in Australia and New Zealand is relatively high by OECD standards (Burgess/Gleisner/Rasmussen 1996) and there are no constraints on the deployment of part-time workers. This is partly related to a rise in tertiary education and to limited childcare provision in both countries, which means that part-time work is one way that work and study or caring arrangements can be combined (Pocock 2003; Rasmussen et al. 2004b). Thus, there is a pool of workers (particularly young females) who are available to fill flexible employment arrangements since they have commitments relating to study or family care.

Also, agency work can now be considered as a viable, alternative option as part of the subcontracting and self-employment arrangements that are becoming entrenched in many industries and organizations. According to recent Australian and New Zealand research, this has been the case in construction and building (Hall 2004), TV and film (de Bruin/Dupuis 2004), health (Lumley/Stanton/Bartram 2004), stevedoring (Dabscheck 1998) and call centres (Hannif/Lamm 2005). In this context, agency work represents another form of subcontracting and risk spreading. That is, agency employment has to be seen as one means, among many, towards business re-organization, and with it the restructuring of employment. Agency employment can achieve these economies and flexibilities but there may be more direct means available that do not require the labour brokerage costs of employment agencies (Rubery et al. 2000).

A second reason is that traditional motivations for employing agency workers are still important. Many organisations require On call' agency workers to meet unexpected shortages in labour or urgent skill needs. These requirements are regarded as appropriate for an agency assignment as long as they are short-term - thereby making agency assignments cost efficient. In some cases, organizational limits on the hiring of permanent staff - a so-called 'headcount' approach - are driving agency employment. This is the case for many larger companies where company earnings are related to the number of permanent staff or in public sector organisations where budgetary considerations limit staff numbers (Lumley et al. 2004). Again, the motivation is specific, normally short-term and complementary to long-term and on-going employment. Thus, there are significant constraints on the overall numbers of agency workers employed.

However, the 'headcount' approach has been a major factor in the rise in agency employment and this has coincided with widespread skill shortages in Australia and New Zealand. Where skills are in high demand or are quite specific - for example, as in health or IT - agency staff can provide a temporary solution. In the longer term, this becomes more problematic since it will be very expensive for hiring organisations to provide continuity in the provision of skills through agency workers. This continuity may be provided through a contractual shift of the agency worker and it has now become more common for agencies to allow - for a fee - that agency workers accept a permanent position. For example, the Australian industry study by Brennan et al. (2003: 55) found that around one fifth of agency placements ultimately become direct employees of the client. Hence, the analogy from Gash (2004) of temporary work being either a 'bridge or a trap'.

'Trying before buying' is built into the activities of most agencies since agency work may be used by client organisations as both a recruitment device and a form of screening potential employees. The placement and intermediary functions of agencies mean that both functions can be conflated so that agencies can simultaneously fulfill both functions. This has often been the case for the unemployed and constitutes a new, often officially-sponsored market opportunity for agencies (Rasmussen/Lind /Visser 2005).2 For those seeking permanent full time employment, temporary work does provide a 'foot in the door' as an employer is given a low risk (and thereby cost efficient) opportunity to 'try before buying'. It also gives the worker an opportunity to see whether they would like to work permanently with a particular company. Again, agency work becomes a process for arranging the recruitment of permanent and waged employees and this curtails the relative employment share of agency employment.

A third reason why the use of temp agency workers may be limited relates to a range of potentially adverse human resource implications associated with employing agency workers. Where an organization is attempting to enhance commitment, teams and high work performance, the use of agency workers can compromise these goals (Hall 2005). Temps may, for example, lack commitment and be in search of better assignments or a permanent job elsewhere. Thus, the provision of training and skills upgrades could be counter productive if workers have minimal attachment. Also, the agency workers are often partitioned from the culture of the waged workforce, and, in some cases, they can be seen as a threat to waged workers leading to disruption within the workplace (Hall 2004). Generally, the implications of utilizing temporary workers over long periods are not known in terms of organizational effectiveness. While outsourcing has become very popular, there are also reports of a return to 'insourcing' as better competitive advantages can be achieved through internal (standard and atypical) employment arrangements. There have also been more particular risks, especially in the area of occupational health and safety whereby temp agency workers have placed themselves and their co-workers at risk, because they were not familiar with the protocols and policies specific to an organization's occupational health and safety program (Labour Hire Taskforce 2001).

A fourth reason for employers not using agency temps is that agency employment remains, as discussed below, a preferred option for only a minority of agency workers. Lifestyle choice, independence and a variety of experiences, while promoted by the agency industry, are not the drivers for many workers taking up the option of agency employment (Hall 2005). There will be workers who desire short-term assignments, but in the main, there will be a large pool of agency workers who have taken agency assignments as a temporary or second-best choice. These people would prefer regular and secure employment if it were available or could fit their current circumstances. The whole question of whether people are 'pulled or pushed' into self-employment and/or agency work is a rather problematic one, with the various 'pull and push factors' impacting differently on individuals (Bururu/Irwin/Melville 1998; Firkin 2003; Perera 2003).

Individual choices are often influenced by contextual factors and individual preferences. Agency work can be a risky or even unpleasant existence: besides the highpaid, sought after 'leased executive', specialist or temp, there is a low-paid, continuously job changing, highly stressful side to the lesser skilled temp jobs (Rasmussen et al. 1996; Alach/Inkson 2003). The state of the business cycle will also influence preferences. When the labour market is buoyant, and skill shortages are emerging (as they are currently doing in Australia and New Zealand), it becomes less risky to be an agency worker. However, it is more difficult for temp employment agencies to find suitable recruits, if there are comparable waged and long-term employment positions available. If the state of the labour market offers more options for job seekers then the shortage of agency workers will indicate that agency employment remains a second best option for many job seekers.

Temp agency work: Precarious work or employee opportunity?

Labour flexibility agendas are normally promoted by employers or governments, endeavoring to promote growth, contain inflation or improve international competitiveness (Campbell 1993; Standing 1999). In contrast, employment precariousness often examines the conditions of employment from the worker perspective. But what makes a job precarious? A considerable literature has developed around employment precariousness, in some cases associating it with non-standard and contingent employment arrangements (Capelli et al. 2001; Rodgers 1989). Other factors relating to the precarious nature of work include:

1. the objective conditions of work including hours, rights and protection, pay and conditions of work.

2. the perceptions that the worker holds about the job. Is it a job they derive satisfaction from, feel secure in and would be happy to continue in?

3. the quality of the job, bringing together objective and subjective employment conditions such as the national and international job quality indicators that aggregate and measure job quality (de Ruyter/Burgess 2003; Considine/Callus 2001).

Standing (2002) used the concept of labour security to highlight the duality of labour flexibility, whereby as jobs become more flexible they also tend to become more insecure. This insecurity covers a range of dimensions such as pay, tenure, representation, skills, hours and control. The objective conditions associated with agency employment are such that they are likely to generate work experiences that are insecure and precarious. On balance there is no doubt that, relative to other employment arrangements, temp agency employment has the potential to be precarious for the following reasons:

1. Labour is utilised on demand and usually for short periods.

2. The employment relationship is not ongoing but for a limited duration.

3. For the agency worker both the term of employment and pay will be indeterminate, becoming dependent on the type of assignment (part-time versus full-time), the duration of the assignment and the likelihood of achieving subsequent engagements.

Job insecurity is compounded where the assignment is in the form of a contractual arrangement as opposed to placement as an employee. In Australia and New Zealand with its very minimalist regulatory regime over agencies and the agency employment contract, the potential for insecurity and precariousness is increased. The identification of the employer is uncertain, agencies have limited reporting and financial security obligations (Burgess and Connell 2004b) and agency employment can complicate legal responsibilities with respect to taxation, workers compensation and safety regulations (Labour Hire Taskforce 2001).

If the experiences of agency workers are reviewed, it is clear that many have an insecure and precarious existence, and would prefer alternative employment arrangements. The Labour Hire Taskforce (2001) reported that 95 per cent of agency workers in Australia were 'casuals' and, as such, they suffer from all the forms of benefit and protection exclusion attached to casual employment (Campbell/Burgess 2001). In addition, the union submissions to the NSW Task force suggested a pattern of persistent pay below the minimum wage. This is possible where agency workers are employed as contractors, not employees. Their pay rates were often below those of permanent workers who performed the same job (Hall 2000). Without a clear designation of employer responsibility, agency workers are not covered for holidays, sickness or severance. Furthermore, there is an imperative for many agency workers to maintain continuous assignments without a break for sickness or holidays, since their non-availability may mean they are relegated in the queue of available agency workers when assignments become available (Connell/Burgess 2001). In between employment assignments, agency workers are effectively unemployed. In Australia and New Zealand, unlike some EU countries, the agency workers are not paid by the agency between engagements (Storrie 2002). Moreover, in the case of work related injuries, it is of concern that ambiguous employment arrangements can result in an absence of occupational health and safety coverage and no insurance cover for rehabilitation and a return to work (Hall 2000).

From the agency perspective they tend to be very explicit about the fact that they do not offer permanent employment as job insecurity is part-and-parcel of agency employment. However, many agencies try to overcome the insecurity aspect through promising interesting assignments, learning on the job, and 'taking the hassles out of job hunting'. Importantly, agencies provide preferential job assignments for their 'regular' agency workers and these workers will often have a continuous employment pattern as long as their particular labour market is relatively buoyant. Still, the OECD (2002) found that lower pay and 'lower' jobs prevailed in the agency sector. A New Zealand survey, reported by Hardy and Walker (2003: 146), found similar substandard outcomes: "...the majority of temps, almost 70 per cent of whom are women, earn below the national average wage." This is a disturbing finding, considering that 45 per cent of the agency workers surveyed had a tertiary qualification and 50 per cent had had a permanent job before temping.

Hall (2005) found similar negative outcomes based on two Australian surveys. The Brennan et al. (2003) survey found that there was a high degree of compulsion attached to the choice of agency assignments (around one third of agency workers) and two thirds of the surveyed agency workers reported that they would prefer direct employment relationships with the client organisation. Hall (2005) also cited evidence from the HILDA employment survey which clearly indicated low job satisfaction and outcomes amongst agency workers:

"Agency workers reported less satisfaction with work and pay, less freedom, lower levels of empowerment, less use of current skills, less need for new skills and lower levels of job complexity. This is a powerful finding and confirms that agency work is associated with lower satisfaction, autonomy, empowerment and skill utilisation for the same occupational status" (Hall 2005).

The factors associated with agency employment can reflect local labour market conditions together with the conflation of the job search and agency functions of the placement agencies. For the unemployed, agency placements may be the only means to access a job as job seekers have the opportunity to obtain work experience through agency employment and to progress to more secure employment.

Job insecurity is a major threat for agency workers and they are also the first to feel an employment downturn through the non-renewal of assignments. For example, in New Zealand, there was a downturn in the Auckland market for personnel and human resource management staff following the 11 September 2001 event in New York. Anecdotal evidence has indicated that this subdued demand lasted for at least 3 years. In such a situation, agency workers will either have to shift into other types of jobs or venture overseas. In short, this is a situation that can primarily be countered through a return of employer demand and lower unemployment. In the last two decades, this has seldom been the case for broader job categories in Australia and New Zealand but it has existed in, for example, IT, engineering and nursing. However, the low level of unemployment in Australia and New Zealand, as in several other smaller OECD countries, may herald another repositioning of agencies (similar to the USA in the late 1990s, see Peck/Theodore 2004).

Knowledge, skills and experience constitute a crucial asset for agency workers and it can be a struggle to keep these market-relevant (Rasmussen/Deeks 1998). Although some agencies provide training for their clerical 'temps' this is the exception rather than the rule. Where employers offer training for temps this tends to primarily focus on compulsory health and safety training and on-the-job training required to familiarize temps with their environment (Connell/Burgess 2002). Agency employment for professional and skilled workers is predominantly based on skills and education gained elsewhere as they are expected to be 'work-ready' and come with the necessary skills (Connell/Burgess 2002). This is one of the reasons why agency workers appreciate long-term assignments (Hardy/Walker 2003) which allow them to gather industry or occupation-specific experience and knowledge.

There is, however, another growing literature on the positive side of agency work. Specifically, there are significant supply-side influences at play as agency work can provide lifestyle choices, flexibility and/or increased income. This has included a tendency to idealize agency work through notions of super-efficient, totally flexible, highly paid 'leased personnel' where 'career resilience' can only be built through being entrepreneurial, adaptable and skilled (see Inkson et al. 2001). Agency work can also be seen as an expression of 'alternative value aspirations' whereby women can build new identities that are less economic and labour market determined (see Casey/Alach 2004). In New Zealand, there has been an increased focus on individual choice amongst agency workers; mainly associated with Alach's rather limited empirical research (see Alach/Inkson 2002, 2003, 2004; casey/Alach 2004). In this type of research, agency work is seen in a more positive light where the focus is on the subjective experience of agency work and how opportunity and choice can work for the individual.

There is nothing new in this positive interpretation of agency work since it has already been mentioned by other New Zealand researchers such as Callister (1997) and Grey (1998) who suggested that temping as a lifestyle or career option was on the rise. Nevertheless, these positive interpretations of temp work raise two questions. First, what percentage of workers constitutes these 'career' or 'life style' agency workers? As mentioned, the evidence indicates that this is a small minority as agency work is seldom the first choice for agency workers (see also Hardy/Walker 2003). Second, it is necessary to stress that this could be a dangerous employment route to take (assuming that agency workers have other choices and that is quite often not the case). For example, Alach and Inkson (2004: 48) found that adverse labour market trends resulted in assignments and incomes being terminated. This is clearly a precarious situation, even though it happened in a period of relatively low unemployment:

The vast majority of agency workers tend to be located in the unskilled and semi-skilled occupations associated with clerical work, construction work, manufacturing and business services. Research evidence suggests that highly paid and highly skilled temp agency workers are a small component of the industry, and while their situation can be idealised, their experience is not representative of the average experience across agency employment (Hall 2005). Precariousness is built into the construction of agency assignments, and this is further compounded by the very light regulatory regime that applies to agency employment in Australia and New Zealand. Public policy issues have been raised with respect to the potential for agency employment to subvert income taxation, worker's compensation, immigration and safety regulations (NSW Labour Hire Taskforce 2001), but even if these issues were addressed, employment arrangements would still remain precarious for the majority of agency workers.

Conclusion

This article has presented a comprehensive overview of recent Australian and New Zealand research relating to agency employment and has discussed three key issues associated with temp agency work. Discussion has shown that:

1. there is a considerable lack of solid information about agency work,

2. there is a variety of opinions about agency work and

3. it seems to be bypassed in the public policy debate.

In addition, agency work appears to have a disproportional impact on trends in employment and employment conditions and, constituting a 'thin wedge' undermining standard employment conditions, it is probably necessary to consider how regulatory standards can be implemented efficiently.

In Australia and New Zealand, agency employment is very lightly regulated. Despite the high visibility of agency work and its ambiguous and precarious tendencies, there has been a remarkable 'wild west' approach to regulation, public accountability and worker protection. This can partly be explained through timing, as agency work has expanded at a time when the Australian and New Zealand labour markets became more deregulated. It can also be partly aligned to a focus on standard employment patterns which has bypassed, to some degree, casual employment and self-employment. This has provided agencies with ample scope to grow. This deregulatory 'wild west' approach can be expected to continue in Australia under the Howard governments' liberalisation of labour market regulation. It is remarkable, however, that the recent 'Europeanisation' of New Zealand employment relations has not prompted any public policy debate in relation to how agencies and agency employment could be regulated in New Zealand.

Media and agency promotion of agency work has given it more visibility and many employers are now aware of its ability to increase labour flexibility. In that light, it is surprising that agency employment constitutes a relatively small part of overall employment (though, as discussed, some under reporting probably occur). A key constraint relating to the utilization of temps is agency costs which often make other forms of labour flexibility more cost effective. There is also the impact of 'trying before buying' where employers use agency work as a screening device for future employees. This dovetails well with the double service role of agencies: recruitment and temporary placements. While these factors enhance the importance of agencies they limit the overall employment share of agency work. Finally, it is important to point to the adverse human resource implications associated with employing agency workers which cut across the fashionable tendency to create a highly committed and loyal workforce.

It has been recognised for some time that temp agency work, as one of the most flexible employment forms, constitutes a rather precarious form of work. Recently, there has been more emphasis in Australian and New Zealand research on individual choices and preferences. This is in line with the more glamorous image projected by the agencies themselves. There are undoubtedly some 'lifestyle temps' but, in common with the other temp statistics, it is unclear how many there really are. We estimate that these more glamorous assignments and 'lifestyle temps' represent a rather limited section of agency employment. Nearly all of the broadly-based surveys and statistics emphasise the aspects of temp agency work relating to choices with constraints, as well as the risks associated with agency work and the often substandard employment outcomes that can result for temp employees and employers.