Early impacts of the Victorian Training Guarantee on VET enrolments and graduate outcomes

In early 2008, in response to changing labour market demands and concerns over skill development and use in the Australian population, the Council of Australian Governments (COAG) initiated the National Agreement for Skills and Workforce Development. A component of the agreement focused on reforming the training market to be more demand-driven and responsive to the labour market. Victoria was the first state to introduce reforms, with the first round, referred to as the Victorian Training Guarantee (VTG), implemented between July 2009 and January 2011. The Victorian Training Guarantee differs from the national reforms in three main ways: places available are uncapped and based on student demand, although there is an upskilling requirement for those aged 20 years and over; there is full contestability between public and private providers for places; and there is greater flexibility for providers in the setting of course fees.

The Victorian system has come under scrutiny from opposing governments, other jurisdictions, providers, industry, and the public. What is apparent is that there has been a significant increase in vocational education and training (VET) enrolments since the reforms were first implemented. This research finds that, between 2008 (pre-reform) and 2011 (post-reform), the Victorian Training Guarantee was estimated to have led to a 35-percentage-point growth in enrolments, with much of this growth in private providers. This increase is far greater than that which has occurred in other states/territories over the same period. What is not as clear however is the impact of the training guarantee on the outcomes for learners of different ages and those from a non-English speaking background, Indigenous students and students with a disability. This is one focus of this research, which uses data from the NCVER Student Outcomes Survey and the National VET Provider Collection. The research examines only the first round of reforms, implemented between July 2009 and January 2011. Subsequent reforms introduced in Victoria in 2012 are not part of this analysis.

Key messages

• For 15 to 19-year-olds, the Victorian Training Guarantee is estimated to improve the likelihood of being in full-time employment six months after training; this group was also satisfied with their course. The outcomes were not as positive for those aged 20 to 24 years, however, possibly suggesting that the upskilling requirements of the training guarantee are limiting the potential benefits for those looking to change their occupation.
• The increases in enrolments for students from a non-English speaking background or who have a disability were not as great as for those who were not in these equity groups. The Victorian Training Guarantee was estimated to have no effect on Indigenous students’ enrolments.

Rod Camm
Managing Director, NCVER
 

Since July 2009, the Victorian vocational education and training (VET) sector has undergone a number of demand-driven reforms, the objective being for the sector more responsive to current and future skill needs.

The aim of this project is to estimate the short-run effects of the first round of reforms — the Victorian Training Guarantee (VTG) introduced between July 2009 and January 2011 — on student enrolments, their course choices and their outcomes. A particular focus is on how the reforms have affected the enrolments and outcomes for equity group members: students from non-English speaking backgrounds (that is, English as a second language), Indigenous students and students with a disability.

Given that other states have either recently implemented (South Australia), or have plans to introduce similar reforms, the analysis presented in this report provides a timely investigation of the likely impacts of the adoption of demand-driven models of VET provision.

At present, there are insufficient data available to evaluate the effects of subsequent Victorian VET reforms; that is, those introduced in 2012 under the ‘Refocusing Vocational Training in Victoria’ initiative (Department of Education and Early Childhood Development 2012),  including changes in course subsidy levels and the deregulation of the course fee structure.

The key feature of the Victorian Training Guarantee is an entitlement to a publicly funded place in vocational education and training; the training guarantee also gives students the freedom to undertake the course of their choice with the public or private provider of their choice. Prior to the training guarantee, the provision of vocational education and training in Victoria, as in other states, was primarily supply-driven, in the sense that public funding was allocated directly to providers in the form of a block grant, based in part on historical enrolments and skill forecasts. In practice this meant a cap on the overall number of publicly funded places, with the allocation of places across students determined by providers, which was often on a first-come first-served basis.

At the time of the introduction of the Victorian Training Guarantee, its likely impacts on student outcomes would have been somewhat uncertain. On the one hand, removing the cap on subsidised places could be expected to increase enrolments, including for equity group members who, under the old regime, may have missed out on a place. Greater freedom of course choice and competition between providers might also be expected to enhance skill acquisition and lead to better employment outcomes. On the other hand, it could be that student choices may not align with skill demands and course quality under the new system, particularly where information on course quality and outcomes is lacking or not easily accessible.

This report builds on analyses of the impacts of the Victorian Training Guarantee undertaken by Skills Victoria (2012a) and the Essential Services Commission (2011) and also a parallel study by the authors (Leung et al. 2013¹).The main contribution of this study is its examination of student post-training outcomes as well as course choices and course completions (Leung et al. 2013). A further contribution of this present study is the use of detailed information from New South Wales, in combination with multivariate analysis, to construct counterfactual outcomes for Victoria (outcomes that reflect what would have happened in Victoria had the reforms not been implemented) in order to isolate the impacts of the Victorian Training Guarantee from the impacts of other changes, such as the rollout of the national education entitlements and changes in economic trends, which occurred at the same time. The construction of counterfactual outcomes to isolate the effects of the training guarantee sets our analysis apart from that produced by Skills Victoria (2012a), with the estimates from the latter representing only changes in student numbers by comparison with 2008. This should be borne in mind when comparing estimates from this report with estimates from Skills Victoria. Because access to a VET entitlement under the Victorian Training Guarantee varies by age, we conduct separate analyses for those aged 15—19 years, 20—24 years and those aged 25 years and more.

The analysis of enrolments presented here draws on detailed administrative data on all enrolments in publicly funded VET courses over the period of interest taken from the National VET Provider Collection, managed by the National Centre for Vocational Education Research (NCVER). The analysis of post-study outcomes, however, is limited by the availability of suitable survey data (taken from the Student Outcomes Survey). Survey data on post-study outcomes are only just beginning to become available for those enrolling after the reforms. Here we limit the sample of analysis to course graduates who enrolled in January or February 2010 and completed by the end of 2011. This has two important implications. First, the available sample is so small that some of our estimates of the training guarantee impacts on post-study outcomes are imprecise and are not as robust as we would like. Second, those in the 25 years and older age group who enrolled in 2010 enrolled prior to the extension of the entitlement from diploma-level courses to any course that is higher than their existing qualifications held in 2011. Therefore, the outcome results for the 25 years and older cohort who enrolled in 2010 are unlikely to give a clear picture of the impacts of the full introduction of the training guarantee for this age group; neither are they likely to give a clear picture of the impacts of the diploma-level entitlement in place in 2010, given that the extension to all higher-level courses had already been pre-announced. Nonetheless, we present the preliminary impacts for this group here for completeness.

As more data become available, obtaining a clearer picture of the effects of the training guarantee using the methods applied in this report should be possible. Even so, future analyses will still need to rely on the Students Outcomes Survey, and will therefore be restricted to analysing outcomes six months after course completion.

Results

Overall, we estimate that the Victorian Training Guarantee has substantially increased new enrolments in vocational education. In 2011, new enrolments in New South Wales were 6% higher than they were in 2008. In Victoria they were 41% higher. Our estimate of the impact of the training guarantee on new enrolments in 2011 is therefore that it led to an additional 35 percentage points of growth in enrolments. The training guarantee also increased enrolments for two key equity groups (disabled students and students from non-English speaking backgrounds), although to a lesser extent than the increase for non-equity group students. The Victorian Training Guarantee is estimated to have had no discernible impact on Indigenous enrolments. In total, the estimated impacts of the training guarantee on new enrolments are generally consistent with the changes in student numbers reported by Skills Victoria between 2008 and 2011 (2012a), but with some noticeable differences for specific cohorts; namely, Indigenous people and people with a disability. In these cases, changes in student numbers may not merely reflect the impacts of the training guarantee, but also changes in other circumstances at the same time.

Much of the increase in enrolments has been realised as increased enrolments in private institutions. Between 2008 and 2010, the Victorian Training Guarantee was associated with a 60-percentage-point higher growth in enrolments with private providers; between 2008 and 2011 this figure was 300 percentage points. Behind this growth in private provision is a 36-percentage-point higher growth in the number of private providers between 2008 and 2010 and, between 2008 and 2011, a 48-percentage-point higher growth. TAFE (technical and further education) enrolments on the other hand were relatively unaffected, with the Victorian Training Guarantee associated with a seven-percentage-point lower growth than otherwise would have been expected between 2008 and 2011. The suggestion is that private providers have done better than TAFE in responding in the short run to increased demand for publicly subsidised places under the training guarantee.

Taking the limitations of data on post-study outcomes into account, we draw three main conclusions from the multivariate analysis. First, for those aged 15 to 19 — the age group with an open entitlement to a publicly funded course of their choice and with their provider of choice — the evidence presented in this study suggests that the Victorian Training Guarantee has generally exerted positive impacts on outcomes. We find that for course graduates the training guarantee is associated with a statistically significant five-percentage-point improvement in the chances of being full-time employed six months after study and a statistically significant four-percentage-point improvement in the chances of being satisfied with their course. These positive effects may work through a number of different channels, including greater access to training related to skill shortage areas and improved quality of training resulting from greater competition.

Second, the Victorian Training Guarantee appears to have had less positive effects on those aged 20 to 24 years who completed a higher qualification compared with the 15 to 19-year-olds. By and large, the magnitude and direction of the estimated impacts, although statistically insignificant, are less positive than for the 15 to 19-year-olds.²  Further analyses of the 20 to 24 years age group suggests that these impacts are largely driven by less favourable effects for those who have already attained a certificate level III and above. We cannot rule out the possibility that this result is due to unobserved differences in the characteristics between those who do and do not hold at least a certificate level III that affect the returns from further study. However, the most likely explanation is that the upskilling requirements are limiting the potential benefits from the Victorian Training Guarantee for 20 to 24-year-old course completers: for those with prior qualifications, it makes retraining in areas outside the current expertise more difficult relative to skill deepening in the current field. In cases where an individual’s current expertise is not in high demand, attaining a higher qualification in the same area may do little to improve their outcomes. Because the estimates in this study are for course completers only, this effect may be exacerbated if relatively few of those who retrain at a higher level complete their qualification.

Third, we find no strong evidence to suggest that the employment effects from the Victorian Training Guarantee are significantly different for graduates who have a disability or who are from a non-English speaking background, relative to those not in these equity groups. Results for Indigenous students are inconclusive due to insufficient data.

Implications

The results presented in this study show that demand-driven reforms can improve access for key equity groups, although not to the same extent as for those not part of an equity group, which raises issues of equity of access. The reason for the lower growth in enrolments among equity group members is not answered in this study, but there are several possible explanations. One is that equity groups may have been given priority access prior to the reforms; hence, the entitlement had less of an effect on access to a publicly subsidised place in training. Another is that the lower enrolment response represents short-term barriers to enrolment, either because new providers are yet to develop the capacity to cater for disadvantaged learners, or because equity groups were slow to access information on the training guarantee. Finally, it is possible that in a competitive market, uncertainty surrounding who pays for the cost of student support services may have deterred some providers from enrolling students with special learning needs. In 2010 and 2011, funding arrangements for these services were limited to select groups — Indigenous, those in correction and early school leavers younger than 20 years — and it was unclear whether the hourly fee premium paid to public providers for delivering a ‘full range of services’ included the cost of services to other equity group members. Understanding the reasons underpinning the lower response among equity group members should be a priority. Also, to ensure equality of access under demand-driven reforms, governments should make support for disability services transparent and available on an equal basis for both public and private providers under community service obligations.

The positive post-study outcomes for 15 to 19-year-olds is consistent with results from a parallel study by the authors (Leung et al. 2013), which shows that the Victorian Training Guarantee has led, on average, to increased enrolments in areas of skill demand (measured by the proportion of enrolments in state and national skill shortage areas) and increased course completion rates. These results provide timely support for the introduction of demand-driven VET reforms in other states. However, the poorer outcomes for 20 to 24-year-olds suggest that upskilling requirements aiming to encourage skill deepening may not necessarily lead to better outcomes because they may limit the opportunities for students to reskill in areas better aligned with industry needs.

The positive outcomes from the implementation of the Victorian Training Guarantee for 15 to 19-year-olds do not necessarily mean that broad-based voucher schemes are an efficient use of government funding. The main justification for government funding of VET courses is that there are positive externalities or spillovers that accrue to the community from training. The Productivity Commission (2011) identifies two broad groups of public benefits: accelerated innovation and diffusion of new ideas; and civic benefits, including improved health, well-being and social cohesion. Because of the public benefits from an individual’s training, governments encourage greater participation in training by subsidising training costs. However, given that the public benefits from training vary from course to course, governments should vary their subsidies accordingly.³

While this is intuitive in theory, in practice it is difficult to value the public benefit from any extra enrolments associated with a course subsidy. When faced with such difficulties, governments should instead choose key criteria for varying subsidies and justify how these criteria are related to the public benefit of the extra enrolments from the subsidies. Subsidy levels that taper off with increasing course levels, as introduced in Victoria (Department of Education and Early Childhood Development 2012) and South Australia (Department of Further Education, Employment, Science and Technology 2011), provide a good example of a criterion that may be clearly linked to the public benefits of subsidies. Higher-level courses are well known to have higher private benefits (future wages) than lower-level courses, but lower-level courses may have higher public benefits because they provide an entry point to enable people to develop minimum skills for workforce and community participation. Hence, because the ratio of private to public benefits increases with the level of course, higher-level courses are likely to attract large numbers of enrolments without a subsidy and, hence, assuming diminishing public benefits from extra enrolments, require a lower subsidy to reach the socially optimal level of VET enrolments.

Criteria linked to skill demand, such as ‘industry needs’, as used to justify the variation of subsidies across courses under the Refocusing Vocational Training in Victoria reforms (Department of Education and Early Childhood Development 2012) may be more difficult to link with the public benefits from subsidising vocational education and training. It could be argued that the main public benefit from VET is in preparing young people for work and the social benefits that flow, so that courses associated with jobs that are in high demand are also courses associated with high public benefits. However, barring failures with the labour market, courses linked to higher industry needs will also have higher private benefits, which will attract large numbers of enrolments without the need for higher subsidies.

One reason why high private benefits may not attract large numbers of students is because of a lack of information about the private benefits from various VET courses. MySkills is a first step for students in meeting the need for better information on outcomes from VET study. At present, information on outcomes from MySkills is limited to broad field of study, which may not be indicative of the outcomes from specific courses. To provide course-level information on outcomes, the sample from the Student Outcomes Survey of VET graduates, which is used in the construction of MySkills, must be considerably expanded. At present the sample is not large enough to support the measurement of course-level outcomes. Ideally, the sample will also contain longitudinal information on post-training outcomes, possibly by linking the survey to individual tax record or census information.

1   Available from the Melbourne Institute website: <http://melbourneinstitute.com/labour/publications/reports.html>.
2   The only statistically significant result is a reduced chance of working in a higher-skilled job, which is likely to merely reflect greater skill deepening in an existing occupation instead of retraining for another job.
3   This was the motivation for the Refocusing VET reforms, which, among other things, included a revamping of the course subsidy levels from July 2012 (Department of Education and Early Childhood Development 2012).