Outcome based commissioning: lessons from contracting out employment and skills programmes in Australia and the USA

Despite the relative effectiveness of recent reforms to the employment and skills system the Leitch Review (1996)identified some key problems. Welfare to work programmes had few incentives to focus on skills, job retention and progression; the skills system had little focus on employment outcomes; and the targets and incentives in both systems did not encourage a joined-up service for individuals or employers. The UK Commission for Employment and Skills (UKCES) subsequently proposed the integration of employment and skills provision and the use of outcome based commissioning. The Government is now seeking to implement such changes and to ensure that providers’ success is defined and measured in terms of their responsiveness to labour market needs and the outcomes from their provision. Debates about outcome based commissioning and contracting, and their application to employment and skills provision, have frequently referred to the experience of other countries, especially the USA and Australia. The USA was the first to introduce explicit performance and outcome standards in its employment and skills system and to connect these with financial incentives and penalties. Australia has sought to create quasi markets in its vocational education and training and is the only OECD country to fully privatise its employment assistance system where a major part of provider income is dependent on securing job outcomes. This report assessed the commissioning and contracting processes through which employment and skills provision is procured in both these countries and sought to identify design and implementation issues that may have implications for the reforms currently being introduced within the British system. The findings reveal that in the USA and Australia respective federal Governments increasingly have set explicit performance and outcome targets within wider reform strategies designed to increase employment rates, reduce welfare dependency and increase skills attainment and utilisation. Performance management systems have been redesigned with public sector employment and skills delivery agencies subject to greater scrutiny and accountability and, where service delivery has been made contestable, challenged by the entry of other providers, including those from the private sector. There are multiple variations in the performance and outcome standards that apply to training and employment services but in the USA they typically include job placement rates, earnings, retention in employment and, for training programmes, skills and qualifications obtained. Studies in both countries show that delivery of employment and skills programmes is subcontracted to a wide range of public and private agencies with performance standards and outcome measures reflected in service delivery agreements and/or contracts with providers. The terms of such contracts differ widely, with varying amounts of provider income dependent on securing agreed outcomes or performance standards. Only in some welfare to work and employment programmes is a major part of provider income dependent on securing sustained job outcomes. It is important to note that these contracts are not exclusively outcome based and typically have other performance and process requirements embedded within them. Views on the success of such contracting systems are mixed and even where the results may appear impressive, it is difficult to disaggregate the impact of the outcome based contracting system from that of other policy changes. Moreover, whilst the cost and efficiency gains claimed for US and Australian models appear significant relatively little is known about how far these gains have been offset by high transaction costs or reduced service quality, especially for the most disadvantaged. There is evidence that minimising cream-skimming, creaming and parking are significant challenges in both public and private sector incentive and target driven delivery systems. Such risks may be reduced through contract design and oversight. The inclusion of measures related to job retention, wages and benefits, and earnings gains, for example, all help diminish any incentive to place participants into poor quality jobs. Measures indicating completion of assessments and activities and regular surveys of participant and employer experience help limit the ability of providers to service clients differently. The challenge is to design such process and outcome measures in ways that do not create unnecessary administrative burdens and allow providers flexibility in how they secure outcomes. The findings show that the implementation of performance and outcome based commissioning and contracting has been dynamic and that government agencies and ‘purchasers’ frequently have had to revise performance standards and contractual terms as problems have arisen and conditions have altered. In both countries there has been much ‘learning by doing’ and constant adaptation as officials have sought to establish performance management and payment structures that now aim to increase the duration of job outcomes, reduce creaming, integrate skills provision, improve service quality, and control any potential for perverse incentives or ‘gaming’ of systems. In this process much knowledge has been gained about different contracting models, the relationships between service delivery and performance incentives, and how to define relevant outcomes. This knowledge, and that developed within the UK, must inform policy makers’ efforts as they seek to operationalise employment and skills outcomes, such as those proposed by UKCES, into more effective performance management and contracting systems. The review findings show that in both the USA and Australia there remain legitimate differences in the outcomes sought from ‘work first’ employment and welfare to work programmes and those sought from vocational education and training systems. Efforts to improve the connections and coordination between these ‘work first’ and skills development programmes are, however, hampered by these differences and by the distinct funding streams and bureaucratic mechanisms through which such outcomes are sought. Such problems are shared in the UK employment and skills system where variation in Government targets and incentives makes coordination and integration of services more difficult. This is particularly evident in the contrasts between the sustained job outcomes that DWP requires contracted employment providers to secure against the short term employment entries that enable Jobcentre Plus to meet its employment targets. It is also clear in the primacy given to qualification attainments in the skills system with less regard given to subsequent employment or whether the skills acquired are actually valued and utilised in the workplace. In addition to more general findings from this study there are three particular proposals that, when tested, might augment current efforts to better coordinate and integrate employment and skills provision in the UK. The first proposal concerns how to better integrate skills provision within DWP employment programmes for the long term unemployed. It concerns adapting the redesign of job focused outcome payments within the Job Services Australia system for use with FND and other contracted out DWP programmes. It would involve giving incentives to providers to broker training places with employers and rewarding them when participants they have helped access training are placed in jobs that make use of the skills developed. The second proposal concerns enhancements in the assessment of the employment outcomes from skills programmes and of the qualifications gained. This should involve the use of enrolment and destinations data, collected administratively or through leavers’ surveys, to establish the employment and wage rates of participants. Such data could be used to establish whether participants improved their employment position as a result of their training and the extent to which they utilise any skills gained in their current employment. Such data could be combined into a measure of workforce quality as suggested in Australia. Finally there would be value in swiftly reviewing the many contrasting outcome requirements that exist within the British employment and skills system, and in how they are measured, with a view to developing ‘common performance’ or ‘return on investment’ measures, similar to those being developed in the USA. Such agreed common measures would help minimise different performance, outcome and reporting requirements and facilitate co-commissioning and the alignment of skills and employment funding. They also would help facilitate greater coherence in the performance and outcome standards that providers have to meet. At the same time such a review could critically audit the varying contracting and financial practices of the different public agencies involved in procuring skills and employment provision. The aim should be how to simplify such requirements and, if budgets cannot be integrated, to consider the extent to which ‘master contracts’ might reduce complexity in the system and assess whether employment and skills purchasing might be better managed by the recently proposed single professional procurement agency.