The design of the Work Programme in international context

This report compares the design and procurement of the Work Programme (WP) with pertinent experience in Australia, the USA and the Netherlands. It considers the risks in the implementation of performance based contracts, their implications for the WP, and their address in the other countries. The review finds that contract and procurement systems in the comparator countries have been in flux as policy makers have sought to secure the advantages of contracting out whilst minimising attendant risks and delivery problems. Critics view such change and re-regulation as indicative of the inherent instability of contracting systems whilst proponents prefer the term ‘continuing improvement’. There are a number of key risks with performance and outcome based contracting, notably the practices of ‘creaming and parking’ and ‘gaming’. ‘Cream-skimming’ occurs when providers select the more job ready or more motivated participants to help meet targets or gain outcomes instead of selecting those with the potential to gain more in the longer term. ‘Cream-skimming’ may become a reality in the wake of the initial surge of mandated participants into WP. With its uncapped budget and volunteer eligibility groups, the WP design may inadvertently induce this practice. ‘Parking’ occurs after recruitment when providers may put less effort and investment into working with harder to place participants. Parking was a particular problem in the first two Australian Job Network contract periods. Evaluations found that although participants were allocated to providers for a year most contact occurred in the first few months and many users received minimal services and saw their provider infrequently. It is possible to envisage a similar outcome in the WP should providers concentrate intensive service provision in the initial stages and maintain different levels of contact with participants thereafter. DWP maintains that the design of the WP will militate against this but critics have warned that ‘parking’ may occur amongst the harder to assist participants within the WP’s differential payment bands. ‘Gaming’ occurs when providers seek to exploit weaknesses in programme design by undertaking activities which allow better performance on contractual criteria but do not improve employment outcomes. At its most extreme ‘gaming’ may involve fraudulent activity. The occurrence of ‘gaming’ in each of the comparator countries reinforces the importance of balancing WP flexibilities with safeguards, and robust risk assurance and audit processes. A further set of risks relate to ‘market failure’. The most obvious is that a severe ‘economic shock’ might undermine provider viability. The Department suggests that WP contracts give providers scope to absorb most labour market fluctuations and there is a ‘change control’ mechanism that enables contractual variations in the event of unforeseen adverse developments. Other, less dramatic, forms of market failure require that the DWP has sufficient safeguards to sanction large prime contractors for poor performance, remove contracts and/or replace an underperforming provider when contracts are due for renewal. It remains unclear how much risk the DWP has effectively transferred to prime providers and the Department’s capacity to tackle under performance and to resist attempts to renegotiate contractual terms is, as yet, untested. A potential long term consequence of the concentration and duration of WP contracts, however, is the accrued advantages of the incumbent primes and the lessening of the risk of potential loss of contract. In future, the DWP also may find it harder to attract new entrants to the market. It is important, then, that the continuing interest and potential competitive threat posed by other providers who have qualified for the ‘Framework’ is sustained rather than assumed. The explicit connection between up-front investment by WP providers, job outcome and long term sustainment payments, and funding from future benefit savings is distinctive to the British system. In Wisconsin welfare to work providers received some of the savings accrued from the reduced benefit expenditure they helped generate. The contract was poorly designed as providers gained income from simply reducing the benefit caseload. The outcome and evidence requirements for WP providers are more rigorous, but it remains important to monitor the number and destinations of WP referrals and participants who leave benefit but do not enter employment. In the comparator countries, there have been challenges around the definition, measurement and authentication of job outcomes. Evidential requirements should be robust but proportionate, seeking to limit the administrative burden that has characterised paper-based audit systems. The definition of a WP job outcome will need to be revised over the period of the contract to better align provider incentives with the Universal Credit which will encourage claimants to take ‘mini jobs’. WP prime contractors have specified the minimum services they intend to provide. They also are obliged to inform service users about the services they will make available. There are, however, few safeguards to ensure the delivery of services and the DWP will have only limited insight into the ‘black box’ of front line delivery. It is important that the safeguards are robust if Ministers and DWP are to resist possible future pressures to introduce the type of prescribed service standards that re-bureaucratised the Australian Job Network. Service user journeys across mixed public and private provision can be complicated, especially for the most disadvantaged. In Australia, the Netherlands and the US problems have arisen with ‘failures to attend’, incorrect assessments, and the imposition of sanctions. Such risks may be heightened with the WP because of its duration, subcontractor delivery chains, and the requirement for participants to maintain regular contact with a private provider whilst continuing to ‘sign on’. Service users will need clear and timely information to avoid ‘mixed messages’ and it will be important to monitor trends in sanctions and the interactions on conditionality between Jobcentre Plus (JCP) and WP providers. Other problems may arise from poor interactions between health assessments and job search and programme activity requirements. Lengthy waiting times and inaccurate assessments, with consequential appeals, have proved difficult to manage both in Australia and New York. The connections between JCP, Atos and WP providers should be tracked closely. Issues of concern may include the timeliness, reporting and accuracy of work capacity assessments, appeals processes, and the management of changing health status of individuals during WP participation. Comparative evidence confirms the importance of ensuring that robust systems are in place to respond to complaints of unfair treatment and poor service delivery. In Australia the Department has committed to clear minimum ‘Service Guarantees’ that apply to all providers. In the USA providers and welfare agencies must have complaints procedures in place to respond should services be unfairly denied to clients or of poor quality. The WP approach seems opaque in this regard, with individual prime contractors having their own minimum service standards, which may vary, and freedom in how they communicate their disputes and resolution procedures to participants. Prime providers must produce a document outlining their minimum services and the ‘first steps’ of their complaints procedures but it is unclear how the full range of safeguards are communicated to service users referred to the WP. In addition, there does not seem to be any mechanism for collating complaints information between JCP, prime contractors and DWP. Concerns have been expressed about prime providers’ management of their subcontractors and the contractual terms and risks passed on to smaller organisations. In response the DWP’s Commissioning Strategy established a ‘code of conduct’ and sponsored the ‘Merlin’ standard to ensure a role for voluntary sector and specialist organisations in WP delivery. The DWP, however, eschews any direct role in supply chain commercial relationships although these forces are most likely to drive change and possible consolidation. Over time it may be important to monitor the impact of WP contracting on the composition of the non-profit organisations involved as well as the quantity of referrals and terms and value of any contracts with primes. The accountability of WP providers is more limited than that of public sector organisations and whilst prime provider contracts have been published, information that was considered ‘commercially sensitive or confidential’ has been redacted and there is little public information available on subcontractor contracts. Contracting out of employment services poses challenges to political accountability. Ministers remain politically responsible for service outcomes and the effective use of public funds even though in practice they will have less control over the actions of WP contractors. There is much scope for ‘blame shifting’ and the responsibility for poor performance is less obvious. In each of the comparator countries audit and oversight bodies have played a significant role in holding policy makers and providers to account, prompting improvements in service delivery, and exposing poor practices. In this context the role of such bodies in Britain will be of particular importance, especially as Ministers, senior officials and providers will have a clear interest in promoting the success of the WP.