Contracting out publicly funded vocational education: A transaction cost critique

Abstract

Contracting out publicly funded vocational education and training (VET) in Australia to private providers has been accompanied by persistent concern at decline in the quality of training. Using transaction cost economics, this outcome is ascribed to the characteristics of publicly funded VET as a commodity and the conditions under which it is privately produced and consumed. The article concludes that, in general, publicly funded VET does not meet the minimum conditions for efficient contracting out. The economic and social consequences of inadequate quality VET provision are potentially severe.

Introduction

Milton Friedman (1955) first outlined the arguments for separating the role of government as a direct producer of educational services from its role in financing education and training. Contracting out to competing private providers would, he argued, better meet the needs of students and employers, reduce cost and improve pedagogical innovation. A student ‘voucher’ system was also advocated as the most efficient means to effect the separation of government as purchaser of educational public goods from their provision. Friedman explicitly included vocational education and training (VET) in this economic analysis.

Since the early 1990s, Australian governments have been incrementally contracting out publicly funded vocational training by creating a ‘training market’ based on competition between public and private VET providers for public training funds (Brown, 2006; Hampson, 2002). In 2008, Australian governments agreed to make all public VET funding open to competition between public and private providers (Council of Australian Governments (COAG), 2008) and, in 2011, a student voucher system was introduced as one means of allocating public training funds. Over the last two decades, the system has moved from one based on a virtual monopoly of public provision, mostly through Technical and Further Education (TAFE) colleges, to a position, where, in 2012, over one-quarter of publicly funded VET enrolments were with a private or other non-government provider (National Centre for Vocational Education Research (NCVER), 2013c: Table 11). These are major changes to the primary system for developing formal intermediate-level skills in Australia.

Following Friedman (1955), the principal rationale for this major shift in government training policy in Australia was based primarily on ‘concepts and language of economics [as] … government had redefined VET as a “product” that was subject to the market forces of “supply” and “demand”, driven respectively by the principles of “competition” and “choice”’ (Anderson, 2005: 13). Despite this rationale, to date evaluations of contracted-out VET have been formulated almost exclusively within an educationist disciplinary framework (Allen Consulting Group, 2013; Anderson, 2005; Ferrier et al., 2008; Mitchell, 2012; Noonan and Allen Consulting Group, 2010; Schofield, 2000), or occasionally from a political science perspective (Ryan, 2002), but not an economic framework. These evaluations find evidence of benefits in contracted-out publicly funded training, but also persistent and widespread poor-quality private provision.

The purpose of this study is to address this research gap by assessing the validity of the economic arguments for contracting out publicly funded VET and to explain why, under this new funding system, the locus of concern over quality is primarily in private provision of publicly funded VET. The study draws on transaction cost economics (TCE) for this analysis. TCE is concerned with identifying the conditions for efficient contracting and the economic effects of deviations from these conditions. The analytical focus of TCE is on the characteristics of the good or service being contracted-out, the conditions under which the good/service is produced and consumed and the incentives created by these characteristics and conditions for principals and agents in a contractual relationship.

The article does not describe in any detail the immensely complex and highly articulated structure that plans, funds and regulates the publicly funded Australian VET system nor the rapidly evolving mechanics for implementing contracting out and student vouchers. Guides to this system are available (NCVER, 2013c).

Scope of private provision of publicly funded VET in Australia

Publicly funded VET plays a highly significant role in initial and continuing post-school skill formation. In 2012, 1.94 million people participated in publicly funded VET or around 12.5% of the total population aged 15–64 years (NCVER, 2013a: Table 2). In 1996 TAFE, the principal public provider, and other government entities, accounted for 83% of total enrolments. Most of the remaining enrolments (15%) were accounted for by Adult and Community Education (ACE), which comprises a diverse collection of decentralised, community owned, not-for-profit organisations (NCVER, 2013b:Table 11). By 2012, the share of TAFE and ACE enrolments had declined to 65% and 6%, respectively. There has been a corresponding increase in the number of students enrolled in ‘Other Registered Training Providers’ (ORTP) from 23,000 in 1996 to 555,000 in 2012. The share of total enrolments in these providers increased from 2% in 1996 to 28%. The ORTP category is an aggregation of different provider types but, importantly ‘Privately Operated Organisations’, that is, privately owned for-profit entities, accounted for 23% of total enrolments in 2012 (NCVER, 2013a). In 2012, ORTP received AUD1.4 billion in public funding, up from AUD0.46 billion in 2008 (NCVER, 2012b: Table 11). (Separate financial data on private providers are not available.)

Previous research

The early literature on the ‘purchaser-provider’ model of contracting out government services was concerned primarily with quantifying the direct benefits to the public purse in terms of reduced cost and productivity gains from private provision (Domberger et al., 1986). A later critical literature adopted a more sceptical attitude to the sources of measured cost reduction, where they occur, suggesting that greater work intensity and loss of real income by workers in affected industries may be important factors in the apparent increased efficiency (Quiggin, 2002). A persistent issue in these studies is diminution of service quality, especially when the performance characteristics of outsourced activity are difficult to specify (Hart et al., 1997). Other studies have identified a variety of negative outcomes such as uncertainty over accountability for service delivery between provider and purchaser and loss of residual public service capacity to specify, assess, supervise and manage outsourced projects (Denniss and Toner, 1999; Ranald, 1993).

Despite the large body of previous research on the contracting out of government services, there has to date been no economic evaluation of its rationale or effects in Australia in relation to public funding of VET. Nevertheless, the substantial body of literature which has evaluated contracting out from an educationist perspective can be marshalled to aid economic analysis. This literature is of direct relevance to this study as the evaluations were conducted by VET experts and their key findings provide an independent source of empirical evidence on policy outcomes.

The early literature on ‘User Choice’, as contracting out was initially termed, found that, in general, the policy objectives were being met (KPMG Consulting, 1999; Schofield, 2000). There was a significant increase in the number of private suppliers competing with public providers, and employers expressed satisfaction with their ability to choose a provider and negotiate the timing, place and content of training. Subsequent evaluations confirmed these results (e.g. Ferrier and Selby-Smith, 2003). Anderson (2005) reported that the ‘outcomes of market reform in VET appear to be positive in relation to choice and diversity, responsiveness … flexibility and innovation’ (p. 33). Overall, competition has had a positive effect on the ‘adaptability’ and ‘creativity’ of providers (Ferrier et al., 2008: 44). Contestability has also been associated with reduced cost of training delivery both by public providers (Anderson, 2005: 28; Productivity Commission, 2012: 419) and by other providers (Ferrier et al., 2008: 21).

Nevertheless, a number of problems are also identified in the competitive training market. There are higher transaction costs for users and providers in the ‘contestable’ market arising from higher search, selection and marketing costs, from customisation of training and from higher regulatory compliance costs introduced in response to concerns over the quality of training delivery in contestable markets (Anderson, 2005: 10; Productivity Commission, 2011: 148). Higher transaction costs may offset ‘productive efficiency’ gains flowing from a competitive training market (Anderson, 2005: 27–28). Lower direct unit costs may also, as Quiggin (2002) argues, reflect an income transfer from teachers, in the form of lower wages and conditions, to the owners of private providers and to government. The benefits of greater customisation of training also seem to be more easily gained by larger firms (Anderson, 2005: 28).

Finally, a persistent finding in these evaluations is lower quality of training in many private providers in the contestable market compared with standards in public delivery. A range of practices in some non-public providers are claimed to give rise to lower quality. These include less rigorous student entry requirements and assessment standards (Halliday-Wynes and Misko, 2013: 32; Productivity Commission, 2011: 122; Schofield, 2000: viii); significantly shorter duration of training to complete a given qualification raising concerns about the ‘volume’ of both training delivered and skills acquired by students (Halliday-Wynes and Misko, 2013: 20–21; National Quality Council, 2010; Productivity Commission, 2011: xlv) significantly lower level of qualifications held by teachers (Halliday-Wynes and Misko, 2013: 26; Productivity Commission, 2011: Table C.16) and inadequate teaching resources (Allen Consulting Group, 2013: 10). The studies cited do not quantify the scale of the quality problem but suggest it is both widespread and serious. An indicator of this concern is that major employer associations, including the Australian Industry Group and Australian Chamber of Commerce and Industry (ACCI), who were key architects of the present system, now argue that contracting out has reduced the quality of provision, that it has failed to meet industry skills needs and that reputational damage to the training system has reduced the incentive of employers and workers to acquire VET qualifications (Mitchell, 2012: 34–35, 38–39). ¹

The solution to quality concerns offered in these evaluations is the creation and enforcement of more rigorous pedagogical standards (Halliday-Wynes and Misko, 2013: 9–11; Productivity Commission, 2011: xviii; Schofield, 2000: viii; Wheelahan and Moodie, 2010: 56–58). This is also the solution adopted by government. Skills Australia (2011), the predecessor of the current national VET regulator the Australian Skills Quality Authority (ASQA), stated that ‘[t]he safeguards we propose for this more market-oriented approach are the implementation of tougher regulatory expectations and performance incentives for providers’. (p. 2) Indeed, the results of the national VET regulators’ activity provide additional evidence as to the locus and scope of the ‘quality problem’. ²

It will be argued that TCE analysis reveals limits to this regulatory solution to quality problems in the training market, arising from the unusual character of publicly funded VET viewed as a market commodity and from the perverse economic incentives facing many private providers, students and employers in this market. The first step, however, is to explore the economic arguments for contracting out in their own terms.

Economic arguments for contracting out of VET

In Australia, the original case for competition in the allocation of publicly funded VET places was made in the 1990 Deveson Report, an inquiry into the Training Costs of Award Restructuring. The report argued that creating a ‘training market’ would impose a competitive discipline on public and private providers and improve technical efficiency by creating incentives for training providers to minimise costs. It would also improve allocative efficiency by more closely tying training provision to user demand and improve dynamic efficiency by promoting innovation in service delivery (Deveson, 1990: 9). ³

The subsequent Finn Review (1991) recapitulated these arguments and also emphasised that ‘greater choice’, ‘greater flexibility in access’ and cost discipline on providers in a training market were pre-conditions to meet an anticipated greatly increased need for workforce training (p. 112). These training aspirations were later converted into precise training targets when the COAG (2008: 26) resolved ‘to halve the proportion of 20-64 year olds without qualifications at Certificate III level by 2020, and to double the number of higher [VET] qualification completions by 2020’.

The Productivity Commission (2011) provides the most recent and exhaustive restatement of these ideas and proposes an additional reason for contracting out. It asserts that industrial instruments governing TAFE employees restrict managerial prerogative in the deployment of teachers to meet the training needs of industry. Increased competition, it argues, will induce TAFE managers and employees to accept ‘contemporary human resource practices’ defined as greater wage, numerical and functional flexibility (Productivity Commission, 2011: liii). It also argues that two complementary changes are required for the efficient and effective extension of contracting out. These are the use of Community Service Obligations (CSOs) to ensure enhanced competition and competitive neutrality between public and private providers (Productivity Commission, 2011: 71). In addition, moving to decentralised resource allocation through student vouchers will, it claims, also improve allocative efficiency by creating a more demand-led as opposed to supply-led system (Productivity Commission, 2011: 68–69).

Notably absent in all of these studies is a sustained consideration as to the suitability of publicly funded VET for contracting out. It is to this task we now turn.

Economic criticisms of contracting out publicly funded VET

This section briefly sets out key principles of TCE analysis, drawn from the work of Oliver Williamson (1985, 1998, 1999, 2000), focussing on the determinants of risk in market transactions. It also applies these principles to the outsourcing of publicly funded VET. ⁴

Determinants of risk in contracting

Williamson analyses risks arising in market transactions by relaxing a number of the key assumptions required for perfect competition. The foremost of these is recognition of ‘bounded rationality’ on the part of consumers and producers, characterised by limited information and processing capacity and systematic biases in agents’ assessments of risk and reward. Compounding this risk is that typically a contractor (agent) knows more than a contractee (principal) about the characteristics of a good or service she or he is offering and about the conditions under which it is produced. This leads to ‘information asymmetry’ between the principal and agent, which can result in inefficient contracting even if all parties act in good faith. ‘Opportunism’ also occurs in market transactions as some producers or consumers cheat by failing to honour contracts, shirk effort or otherwise do their best to ensure they get the best out of any bargain. TCE argues that a key, if not sole, factor determining the extent to which an agent actually engages in opportunism is economic incentives confronting the agent. Ethical considerations are also obviously a factor.

For both principals and agents, the extent of risk attached to contracting under conditions of bounded rationality, information asymmetry and opportunism depends on a range of factors governing the nature of the commodity to be contracted-out and the conditions under which it is produced and consumed.

The factors include first, the importance of the contracted activity to the performance of a principal’s organisation. The more important an activity to the survival, profitability or quality of an organisation’s output, the higher the risk in contracting an activity out. It is not just direct costs and rewards that enter into agents’ decision-making regarding ‘important’ market transactions; TCE argues that externalities also need to enter into an agent’s calculation. ⁵ The risk of acquiring an ‘important’ good or service in the external market is increased when such transactions are undertaken only once or very infrequently. This limits the principal’s scope for learning from market transactions and for improving outcomes from such exchanges.

Second, some goods or services are complex in that it is difficult and/or impossible to specify precisely in a contract their features, performance characteristics or attributes of inputs required for their production. Contracts subject to such ambiguity are described as ‘incomplete’. The scope for opportunism on the part of agents is increased when contracts are incomplete.

Finally, differing risks to principals and agents emerge when the production of a good or service is marked by either high or low barriers to entry. The focus here is on low barriers to entry defined as minimal investment by a producer in human and physical capital and low sunk costs. For an agent, low barriers to entry minimise the risk of adverse action by a principal if the agent is detected acting opportunistically.

TCE reveals that attempting to shift risk between contracting parties arising from these multiple sources itself generates costs and risks. A principal can seek to insulate itself from such risks by the use of short-term contracts. This allows the principal to avoid long-term exposure to opportunistic agents and litigation and other transaction costs in terminating a contract. But short-term contracts will likely raise the hurdle rate of return sought by the agent investing in high-cost assets, especially in the case of assets designed to meet specialised needs of a principal. Risks to both principal and agent in this situation of ‘bilateral dependence’ can be addressed by the use of longer term supply contracts, but these may reduce competitive pressure and the incentive for an agent to lower costs and innovate. Agents may also retain rents arising from efficiency gains that lower production costs. In subjecting an activity that was produced in-house to a long-term supply contract, a principal also risks losing knowledge about the commodity and associated production processes. This may disadvantage the principal in future contract negotiations.

Risk and contracting out VET

Importance of the contracted-out commodity

Risks arising from ‘important’ transactions apply with particular force to outsourced publicly funded VET. The fundamental objectives of publicly funded VET are to improve economic efficiency by redressing market impediments to private investment in VET and to improve social equity by lifting participation by disadvantaged groups in VET (Productivity Commission, 2011: 295). The important role of the VET qualified workforce and VET system in national innovation systems is also increasingly recognised (Toner, 2011). There are large direct costs and externalities for individuals, government and the wider economy if VET provision does not adequately meet these objectives. Persistent concern at the quality of contracted provision implies these risks are realised in practice.

Absence of objective measures of VET inputs, outputs or quality

Efficient contracting and pricing in markets relies on complete contracts that specify accurate information with respect to both the quality and quantity of goods and services demanded and supplied. VET is unsuitable for contracting out, owing to the multiple and overlapping objectives expected from it and the inability to closely align VET inputs to the achievement of these objectives. Objectives of the VET system include

building human capital by inspiring, stimulating and enriching learners from all segments of the community; assisting the workforce to acquire the skills needed by the economy and contributing to social inclusion and civic participation. (Productivity Commission, 2011: 295)

It is difficult to specify even approximately what VET courses, activities or qualifications produce these objectives and in what proportion each objective is being produced. Developing metrics for even a single objective is problematic. For example, ‘providing skills for the economy’ may appear to be relatively straightforward, but there is only an indirect connection between the VET qualifications people acquire and the jobs they get. ‘In 2008, only 30 per cent of recent VET graduates reported that they were employed in an occupation group that was related to their training course’ (Productivity Commission, 2011: 120). As the Productivity Commission (2011: 120–121) appears to acknowledge, it is in fact difficult to establish a valid and reliable metric of the extent to which the VET system matches or mismatches labour market needs.

Because of these multiple objectives and difficulties in controlling for quality change over time in VET inputs and outputs, the Productivity Commission (2011: Appendix D) acknowledges that, despite intensive efforts over many decades, no country in the OECD has yet developed widely accepted measures of productivity, efficiency and effectiveness for either education or VET. It also notes similar conceptual and empirical difficulties with the measurement of quality as ‘there are no unequivocal indicators of teaching quality in VET’ (Productivity Commission, 2011: xliii). Similar definitional problems pertain to the application of CSOs in VET.

Compounding the difficulty of constructing robust metrics of performance and quality in education and training are two conflicting design principles, ‘standardisation’ and ‘flexibility’, that have underpinned the detailed apparatus of the Australian VET system over the last three decades. Resolution of this conflict was achieved by giving priority to flexibility. On the one hand, the current training and funding system developed over the course of the last three decades has been premised on the creation of a standardised, industry-led, national system to replace the previous state-based system in which each jurisdiction applied their own curricula, assessment methods, qualifications and occupational licensing (Karmel, 2012: 7). The content and assessment of training, broadly specified by bipartite Industry Training Councils, are based on the concept of Competency-Based Training (CBT) and embodied in National Industry Training Packages. On the other hand, and concurrently, the principle of flexibility is applied to virtually every aspect of training to ensure training is tailored to the needs of the individual workplace and student.

As part of the priority given to flexibility and customisation in training delivery and assessment, Industry Skills Councils do not generally produce textbooks, teaching materials or assessment instruments. Because teaching and assessment is tailored to the needs of the individual workplace, the resulting ‘lack of standardised national assessments means that there is no standard to ensure that a particular set of skills has in fact been acquired’ (Guthrie, 2009: 13).

External reviews of the VET system commissioned by the federal government have reached similar conclusions. Allen Consulting Group (2013) found that there is a ‘strong general view that the Standards for the Regulation of VET are in need of fundamental revision reflecting concerns about aspects of VET quality’ (p. 9). It found ‘inadequate standards for delivery and assessment’ covering virtually the entire system from ‘specific trainer requirements’ to the ‘volume of learning’ (Allen Consulting Group, 2013: vii–viii). Ambiguity over the volume of training arises in part from the fact that payments to providers in the competitive training market are set according to the delivery of ‘nominal hours’ or ‘the anticipated hours of supervised learning and/or training deemed necessary to conduct training/learning and assessment activities associated with the program of study’ (NCVER, 2014: 113). However, providers are not required to deliver a fixed duration of ‘nominal hours’: under ‘competency based progression’, the training period can be reduced if students are deemed by the provider to be learning at a faster rate than average or in recognition of students’ prior on-the-job learning. Determining the latter is also at the discretion of the provider.

Inadequate standards and resulting ‘incomplete’ contracts create considerable latitude for some providers to cut costs by diminishing the quality and quantity of training. They also create considerable difficulties for VET regulators in enforcing imprecise standards. Because ‘Training Package requirements’ and ‘Standards for RTOs [Registered Training Organisations]’ are

… very broad, there can be significant differences between RTOs in the nature and quality of both learning and assessment resources, and in the actual training and assessment process. These differences can create difficulties in the registration and audit process of RTOs in terms of consistent interpretations by regulators. (Allen Consulting Group, 2013: 10–11)

These are persistent findings in the literature (Schofield, 2000: viii).

Who is the principal?

Ambiguity even extends to who precisely is the ‘principal’ in this competitive training market. A foundation principle of the current VET training system is that it be ‘industry-led’ (Jenkins and Curry, 2013). The definition of ‘industry’ was never exact but it broadly encompassed the training needs of firms, employer associations, unions and, to some extent, government (Hampson, 2002; Ryan, 2002). Under the industry-led ‘centralised’ system, funding is allocated to meet skill needs identified by industry and government.

A direct assault on the notion of a training system principal occurred with the introduction in 2011 of student vouchers as one means of allocating public funding. A voucher ‘allows individuals to choose the education and training that best suits them, within certain boundaries’ (Karmel, 2012: 7). Employer associations in particular have been critical of the training choices made by voucher holders for failing to address the skill needs of industry (Hart, 2010; Mitchell, 2012). A voucher system also makes accountability and developing metrics of training provider performance especially problematic since, by definition, each individual student is a ‘principal’ having his or her own reasons for undertaking training and criteria to assess trainers. Accountability for public funds is, therefore, even more diffuse than under an industry-led system, and the scope for opportunism on the part of both student and provider is increased.

The rational VET consumer

The TCE assumption of ‘bounded rationality’ on the part of market participants arguably applies to many VET students, a large proportion of whom have low initial educational attainment, with limited literacy, numeracy and problem-solving capacity. ⁶ This is not unexpected given that increased participation of disadvantaged groups is an explicit objective of publicly funded VET. The limited literacy and numeracy of many VET students may call into question the efficacy of moves by the VET regulator to publish the results of its provider audits ‘to assist VET consumers in making decisions about enrolling with a provider’ (ASQA, 2013: 34).

Compounding the problem of rational decision-making is that selection by students of courses and providers typifies those transactions that are both important to principals but conducted only once or very infrequently. Such transactions pose a particular risk as they ‘are among the least likely to meet the conditions for effective learning. A young person making a life cycle plan gets no practice and therefore no feedback’ (Conlisk, 1996: 683–684). Furthermore, under some circumstances, there can be incentives for providers to engage in ‘obfuscation’ (Ellison and Wolitzky, 2009). For VET providers, this can include offering misleading advice regarding the quality, status and career benefits for students undertaking their courses. Moves by regulators to ensure the veracity of marketing claims made by RTOs with respect to matters such as employment and earnings of graduates, international recognition of their qualifications and credit transfers to higher VET or university opens up a potentially new, difficult and expensive field of VET regulation. ⁷ The national VET regulator recently completed a sample audit of provider advertising and found

… potential breaches ranged from relatively minor concerns (which could be rectified quickly and easily) to more serious breaches (which could involve major sanctions being applied, including the cancellation of a provider’s registration. (ASQA, 2013: 45)

The difficulty of obtaining reliable information as to the quality of multiple providers also imposes higher inefficient ‘search’ costs on potential students and firms.

Low demand for and supply of quality training

There are particular market conditions creating perverse incentives for students and employers not to demand quality training and for providers to supply this low-quality market. ⁸ From a TCE perspective, all parties act opportunistically by minimising their respective costs in meeting contractual obligations to government for publicly funded VET.

There are a number of conditions giving rise to indifference on the demand and supply side to the quality of training. One condition occurs when a student and employer have limited intrinsic motivation to undertake or support training, respectively, but participation of both in the training system is mandated by government. Second, training providers can gain a pecuniary benefit by lowering teaching delivery costs through the employment of under-qualified teachers and the provision of minimal teaching resources. Third, given that a substantial proportion of the public funds to training providers is dependent on student course completions, provider income can be maximised by reducing both the duration of training and assessment standards. Finally, facilitating each of these conditions is the low risk of adverse action by regulators, owing to the characteristics of publicly funded VET outlined earlier. In addition, in a market where all parties collude, there is a low risk that a participant will inform the regulator about their collective opportunism.

Rather than being a theoretical curiosum, a number of documented examples attest to the operation of these market conditions on an apparently large scale. Two examples are provided. Traineeships, like apprenticeships, combine work and formal vocational training in a specific occupation. Unlike apprenticeships, the great majority of traineeships are in low-paid and mostly low-skill service occupations. These include sales assistant, fast food operative, cleaning, labouring, machinery operation, hospitality, security guarding, labouring, road transport, aged care and child care and office administration. The number of traineeships grew rapidly from the late 1990s as a result of government policies which greatly increased their attractiveness to employers (Cully, 2008). Most trainees are paid a mandated low ‘training wage’ and employers receive generous employment and training subsidies. Around 250,000 trainees are now employed at any point in time, around three times larger than traditional apprentices (NCVER, 2012a). However, there have been persistent concerns at the low quality of training provided under traineeships (Ferrier and Selby-Smith, 2003; Schofield, 2000). One study found that ‘[w]ithout improvements in quality across all traineeships their value may continue to be dismissed’ (Smith et al., 2009: 11).

Many traineeships have a number of features which are not conducive to demand for or supply of quality training. The prime reason trainees report for entering a traineeship is to get a job (Cully, 2008). In contrast to traditional apprentices, Cully and Curtain (2001b: ix) found that many trainees had, at best, only a tenuous connection to the training aspect of their traineeship. For these people, undertaking a traineeship offers

a path out of unemployment into the secondary labour market or an artificial barrier into a higher paying job, while for others it offers better prospects with their current employer or a different employer, some of whom do not require completion of the qualification to be persuaded of the person’s competence. (Cully and Curtain, 2001a: 212–213)

The key point is that entry into what previously were unregulated low-skill jobs, mostly for those disadvantaged in the labour market, now requires participation, by both the trainee and employer, in a formal system of regulated training. Karmel (2012: 11) found ‘there is a range of traineeships for which there is apparently little skills acquisition during the traineeship’ and questioned ‘why public money is going to traineeships in some occupations’.

The New South Wales (NSW) Independent Commission Against Corruption (ICAC) identified a related set of conditions giving rise to collusion between students, employers and private trainers to diminish quality. The ICAC investigated training and certification in a number of licensed occupations, including safety induction and heavy plant operation in the construction industry (ICAC, 2004), building contractor and building certifier for residential building (ICAC, 2005) and the licensing of security guards (ICAC, 2009). Briefly, the ICAC found students were motivated by the financial returns to these licensed occupations, but many lacked the prior educational attainment or English language proficiency to legitimately attain the licences. In some cases, students had worked or were working in the industry before licensing requirements were introduced, leading them to believe they had to ‘jump through bureaucratic hoops’ to remain in or re-enter the industry. Second, employers were indifferent to quality of training because of high labour demand for the licensed occupations. For example, ‘demand created by the building boom’, associated with the Sydney Olympic Games, led employers to pressure assessors to pass their employees to ensure adequate labour supply (ICAC, 2004: 72). Alternatively, some licences allowed for self-employment, a situation permitting an absence of employer monitoring of graduate quality. Third, compliant private training providers could increase their income by reducing the quality and duration of training and accelerate student completions. ⁹ The ICAC concluded the scale of fraudulent training provision threatened both public safety and survival of legitimate training providers for these licensed occupations.

Low barriers to entry

As noted above, many VET courses, such as a broad range of traineeships, require minimal investment by providers in specific physical and human assets, and there are inadequate standards relating to teacher qualifications and teaching resources, which can be exploited by providers. The funding system also reduces the incentive of providers to invest in these assets, owing to the relatively limited duration of some government training contracts, which can be as short as one semester, and uncertainty over continuity of funding in a competitive market (Allen Consulting Group, 2011: 10). Government could redress these disincentives by increasing the duration of contracts, but this would contradict an objective of contracting out which is to maximise both competition between providers and flexibility in meeting fluctuations in demand for training.

The significance of low barriers to entry for many courses is that minimal financial and other barriers to entry and exit shorten the investment horizon of the provider and reduce the potential cost of regulatory penalties for providers acting opportunistically.

Ownership and incentives

Up to this point, the argument has considered the appropriateness of private delivery of publicly funded VET. The case for public provision needs to be considered.

The case for public provision reduces to two propositions. First, public sector workers have no claim to an increase in the value of assets they manage (King and Pitchford, 1998: 318). As they are unable to appropriate any financial surplus generated by the public enterprise, ‘no individual can capture the returns from … quality shading’ (Jensen and Stonecash, 2005: 780). In contrast, private providers are subject to ‘high powered incentives’, whereby owners of private RTOs gain a pecuniary benefit by increasing revenue and minimising costs. Employees can share in bonus schemes for improving financial performance, and their continued employment may depend on meeting a variety of financial performance targets. Such incentives, combined with the supply of services that do not meet conditions for efficient contracting, create incentives for opportunism. Public servants are subject to ‘low-powered’ incentives oriented to achieving the often diffuse goals of the public entity set by government (Jensen and Stonecash, 2005: 780).

Second, organisational procedures within public entities erect high practical barriers to public sector workers acting opportunistically. Large process-bound bureaucracies with multiple controls and reporting arrangements make it difficult for systemic quality reduction involving many agents to occur. To take an elementary example, in TAFE the duration of training for a given course is fixed, as courses are based on the full ‘nominal hours’ of funding for the given course suggested in Training Packages. There is, therefore, limited opportunity for teachers or administrators to exploit, for their own self-interest, flexibility provisions that allow providers to deem students ‘competent’ early. ¹⁰

There are, of course, disadvantages to low-powered incentives and process-bound bureaucracy, such as potentially lower incentives to adopt technical and other innovations. Against this must be set persistent concerns regarding the quality of private VET provision. As Williamson (1998: 46) observed,

the common practice of condemning public bureaus because they have lower-powered incentives, more rules and regulations, and greater job security than are associated with a counterpart private bureau completely misses the point. Those features have been deliberately crafted into the public bureau, thereby giving it the desired governance result. ¹¹

Conclusion

The purpose of this article was to assess the validity of the economic arguments for contracting out publicly funded VET and to explain why the locus of concern over quality is primarily in its private provision. The principal conclusions are that, first, from a TCE perspective, in general, the minimum conditions for efficient contracting out of publicly funded VET do not apply. This is due to the characteristics of publicly funded VET viewed as a commodity and the conditions under which it is privately produced and consumed. The conditions for efficient contracting out do not hold either for a large segment of private providers or for many employers and individuals seeking publicly subsidised training. Second, in general, direct public provision of publicly funded training is less subject to pressures to diminish quality due to the objectives, organisational practices and pecuniary incentives facing public providers.

There are a number of public policy implications from this study. Funding needs to be re-balanced to the public VET provider to assign the task of delivering this critical service to the organisation which has within it the least incentive and opportunity to diminish quality. However, the study does not support monopoly public provision of publicly funded VET. Previous evaluations of the training market find some benefits in private provision in terms of greater flexibility with respect to time and location of delivery and pedagogical innovation. Thus, there are some disadvantages in returning wholly to the status quo ante for the delivery of publicly funded training. Against this needs to be set the arguably greater harm caused by reduced quality of VET provision. To identify the appropriate scope for private provision of publicly funded VET, further research is needed to identify those incentives and characteristics within private providers conducive to high-quality provision.

Public policy has responded to rising quality concern by imposing more extensive and intensive regulation on private and public providers (Allen Consulting Group, 2013; ASQA, 2013). This is a problematic solution and highlights a degree of public policy incoherency. Tighter regulation, of necessity, requires much greater standardisation in delivery and assessment, but this conflicts with the priority given to ‘flexibility’ in delivery and assessment. Depending on the scale and scope of new regulatory imposed standardisation, the result could, ironically, resemble a return to the status quo ante.

Decline in the quality of VET provision matters. It diminishes the capacity of the VET system to meet its diverse objectives including to satisfy industry skill needs, contribute to the national innovation system and redress socio-economic disadvantage. ¹²