Welfare-to-work, employment provision and the performance system

The implementation of welfare-to-work, workfare and other active labour market policies has been central to processes of welfare reform in OECD countries (Considine, Reference Considine2000; Lødemel and Trickey, Reference Lødemel and Trickey2000; Peck, Reference Peck2001; Lindsay et al., Reference Lindsay, McQuaid and Dutton2007). In the UK, a ‘work-first’ approach has been predominantly adopted in which the emphasis is on placing those unemployed or economically inactive in work as quickly as possible, regardless of the type or quality of the work (Finn, Reference Finn2003; Daguerre, Reference Daguerre2004). This is in spite of the presence of elements of a ‘human capital development’ approach in some welfare-to-work programmes that aims to address skills needs through education and training in order to improve the ‘employability’ of individuals (Lindsay et al., Reference Lindsay, McQuaid and Dutton2007). UK and European Union policy on the future direction of welfare-to-work has underlined the importance of adopting a ‘personalised’ approach that is responsive to the individual needs of those unemployed (Freud, Reference Freud2007; Department for Work and Pensions, 2008; Lindsay and McQuaid, Reference Lindsay and McQuaid2009). The development of more flexible programmes that allow services to be tailored to the different and sometimes multiple needs of more disadvantaged groups is advocated with a view to facilitating transitions to employment (Dean, Reference Dean2003; Lindsay and McQuaid, Reference Lindsay and McQuaid2009).

Reforms to the organisation of publicly funded employment services in the UK and other countries include the contracting of third sector providers with experience of working with more disadvantaged groups, which has been seen as key to addressing the needs of those groups (Considine, Reference Considine2000; Struyven and Steurs, Reference Struyven and Steurs2005; Freud, Reference Freud2007). In the UK, emphasis is given to supporting the role of providers of specialist services for refugees as part of an overall strategy to increase the employment of refugees by providing ‘a responsive quality service that equips them with the skills they need to get jobs’ (Department for Work and Pensions, 2005: 34). There has, however, been limited research that has examined refugees’ experiences and perceptions of the responsiveness of employment providers to their needs, or providers’ experiences and perceptions regarding what influences responsiveness.

Conflicting dynamics are evident in welfare-to-work policy between, on the one hand, an emphasis on the tailoring of provision to the individual needs of those unemployed and, on the other hand, the overarching aim for an ‘employment-first welfare state’ (Finn, Reference Finn2003), by which the focus has been on moving welfare claimants into a job, irrespective of whether it meets the needs and interests of the individual (Considine, Reference Considine2000; Peck, Reference Peck2001; Finn, Reference Finn, Sol and Westerveld2005). The latter aim is reflected in the system for assessing the performance of the public employment service (Jobcentre Plus in the UK) and contracted providers, which is oriented towards the achievement of job outcomes. Job outcome targets provide the incentive structure for Jobcentre Plus advisers to adopt a work-first approach when delivering assistance for the unemployed. These performance measures are transferred to providers contracted to deliver employment programmes through a system of output-related funding by paying a fixed proportion of funding for each programme participant who enters work after completing the programme. The model, therefore, includes financial incentives for providers to achieve these performance measures.

The effects of this performance system point to tensions with the responsiveness of providers to the needs of more disadvantaged groups within the labour market. First, in terms of the type of users whose needs are addressed, research in Australia, the UK and USA has found that job outcome measures encourage ‘creaming’ – the selection of the most ‘job-ready’ individuals who are more likely to enable providers to achieve the job entries by which their performance is assessed and their services funded (Cragg, Reference Cragg1997; Dockery and Stromback, Reference Dockery and Stromback2001; Heckman et al., Reference Heckman, Heinrich and Smith2002; Barnes et al., Reference Barnes, Hudson, Parry, Sahin-Dikmen, Taylor and Wilkinson2005; Casebourne et al., Reference Casebourne, Davis and Page2006). Evaluations of the Jobs and Training Partnership Act (JTPA) programme in the USA found substantial evidence of ‘creaming’ of participants by providers at the enrolment stage of the programme due to financial incentives attached to job outcome measures (Cragg, Reference Cragg1997; Heckman et al., Reference Heckman, Heinrich and Smith2002). Characteristics of those less likely to be enrolled by providers included being African-American, with lower levels of educational attainment and without recent employment experience. Evidence of ‘creaming’ was also found among providers of Intensive Assistance programmes for the long-term unemployed under the Australian Job Network system. Providers were reluctant to attract Indigenous Australian clients due to the perception that placing them in jobs would be difficult and therefore unlikely to produce a funded output (Dockery and Stromback, Reference Dockery and Stromback2001). Likewise, evidence of the ‘parking’ of participants was found, whereby providers make limited effort to find work for those who are difficult to place because the chances of success are low (Struyven and Steurs, Reference Struyven and Steurs2005).

In the UK, evaluations of Jobcentre Plus programmes have also found evidence of providers selecting out more disadvantaged groups. In the case of the Ethnic Minority Outreach programme (targeted at unemployed minority ethnic groups), contracted providers began to prioritise working with those who had recent employment experience and were, therefore, considered to be more ‘job-ready’, as a result of increasing emphasis placed by Jobcentre Plus on achieving short-term job outcomes as the programme progressed (Barnes et al., Reference Barnes, Hudson, Parry, Sahin-Dikmen, Taylor and Wilkinson2005). Around half of the programme participants had recent employment experience (having left their last job in the previous six months), while only 16 per cent had been out of work for two or more years. In the Action Teams for Jobs programme (targeted at areas with high levels of unemployment), job outcome measures were also found to encourage contracted providers to work with ‘easier to help’ clients (Casebourne et al., Reference Casebourne, Davis and Page2006). Providers considered there to be little incentive to work with clients with multiple barriers to employment as no payment was made for helping those with greater needs make progress in accessing employment.

Second, these performance measures have been found to encourage providers to orient their services to finding the ‘fastest way into work’ for service users (Struyven, Reference Struyven2004), with implications for the type of services that are delivered. An emphasis on job outcomes in employment-related training programmes funded by the former Training and Enterprise Councils (TEC) in the UK led to the provision of types of training oriented towards a limited range of jobs in which these outcomes could more easily be achieved (Felstead, Reference Felstead1998; Gray, Reference Gray2000). In the case of TEC programmes targeted at minority ethnic groups specifically, this incentive structure encouraged providers to focus on placing minority ethnic participants in work in the short term, rather than assisting participants to acquire skills and qualifications, which might have greater impact on improving employment outcomes over the long term (Ogbonna, Reference Ogbonna1998). Job outcome measures attached to funding within the Jobcentre Plus Action Teams for Jobs programme and the New Deal for Young People (targeted at unemployed 18 to 24 year olds) were also considered by providers to encourage them to place participants in any job, regardless of whether or not it was the most suitable job in the medium to long term for the individual, or something in which the participant was interested (Finn, Reference Finn2003; Casebourne et al., Reference Casebourne, Davis and Page2006).

These findings raise questions concerning the extent to which a job outcome-oriented system may inhibit responsiveness if providers are encouraged to place those out of work in any job, irrespective of the needs and interests of the individual. Moreover, they raise questions regarding the extent to which labour market inequalities experienced by minority ethnic groups may be reinforced by diverting resources away from types of assistance that might better address long-term employment outcomes. While short-term job outcomes are used as an indicator of impact (in part because they provide immediate data to providers and funders to measure the ‘success’ of particular interventions), they have indeed been found to be inaccurate as an indicator of employment outcomes over time. In the context of the JTPA programme in the USA, short-term job outcome indicators (based on the employment and earnings of participants on completion of the programme) had either a negative or zero correlation with the employment and earnings of participants over the long term (18 and 30 months following completion) (Heckman et al., Reference Heckman, Heinrich and Smith2002). There is, therefore, the danger that financial incentives attached to short-term performance measures may focus providers on a set of criteria unrelated to longer-term labour market outcomes (Heckman et al., Reference Heckman, Heinrich and Smith2002).

With these issues in mind, the following sections of this article discuss the findings of research in the UK on the responsiveness of employment providers to the needs of refugees.

Responsive to whom? The targeting of assistance

Regarding the type of users to whom providers are responsive, the findings draw attention both to the potential ‘parking’ of refugees (where limited assistance is provided to those considered hardest to help and therefore more difficult to place in work) and to processes of selection of refugee service users (where assistance is targeted at those considered to be more ‘job-ready’).

Refugees’ perceptions and experiences of Jobcentre Plus as unresponsive to their needs partly centred on the limited advice and guidance that they had received from Jobcentre Plus advisers in the context of ‘Work Focused Interviews’. This lack of assistance may have partly reflected the limited time allocated to these interviews and the work load pressures of Jobcentre Plus advisers – indeed, refugees referred to Jobcentre Plus advisers as being ‘very busy’ with ‘many clients to see’. However, it may also have reflected the ‘parking’ of refugees considered by advisers to be among those hardest to help, given the barriers to work faced by refugees, such as limited English language proficiency. While some refugee respondents indicated that they had received information on job vacancies and where to look for vacancies, this minimal assistance was particularly inadequate for those with needs for English language provision.

As will be discussed in the following section, Jobcentre Plus advisers themselves referred to the difficulties of assisting refugees with limited English language proficiency to find work.

Unlike Jobcentre Plus, which is required to assist all claimants of unemployment-related benefits, the third sector providers, who competed for public and other sources of funding for their services, had greater powers to select whom they assisted and to select out those who were difficult to place in work. Two types of processes of selection of refugee service users were evident among some of these providers. One type of selection – also found by research on other Jobcentre Plus-contracted provision for minority ethnic groups in the UK (Barnes et al., Reference Barnes, Hudson, Parry, Sahin-Dikmen, Taylor and Wilkinson2005) and by research on employment provision in Australia and the USA (Dockery and Stromback, Reference Dockery and Stromback2001; Heckman et al., Reference Heckman, Heinrich and Smith2002; Struyven and Steurs, Reference Struyven and Steurs2005) – occurred at the stage of enrolment of refugee users onto programmes funded by Jobcentre Plus. The pressure to achieve job outcomes in order to receive the associated funding was considered by staff to encourage providers to restrict enrolment to those refugees who were more likely to enter work. In the case of EMPLOY, a third sector provider that delivered an English language and job search assistance programme for refugees funded by Jobcentre Plus, staff emphasised that it was very difficult for some refugee participants to access employment within the timeframe of the programme (26 weeks) and the subsequent post-programme period (13 weeks) in which job outcomes had to be achieved, given the relatively low levels of English language proficiency of some participants, particularly those who were illiterate in their first language. The provider's strategy was, therefore, to avoid taking referrals from Jobcentre Plus of refugees seen as ‘more difficult cases’, including those who were illiterate or older in age.

Given that the provider was largely dependent on Jobcentre Plus as the main source of funding for the programmes that it delivered, staff considered themselves to be under pressure not only to achieve the required job outcomes but to exceed them. This pressure related to concerns to out-perform other local contracted providers in order to maintain ongoing contracts with Jobcentre Plus. While this process of selection may have met the needs of the provider to compete successfully for public funding, it conflicted with being responsive to the needs of those refugees facing greater barriers to employment.

A second type of selection found in the case of another provider, CASA, occurred at the stage of applying for Jobcentre Plus funding. The provider delivered a programme aimed at assisting refugees with experience in health care work to re-enter employment in this field. Although it had not previously applied for Jobcentre Plus funding for this programme because of the financial risk associated with funding that was partly dependent on the attainment of job outcomes, it was, at the time of interview, trying to negotiate a strategy for managing this risk. Staff highlighted the tensions between guaranteeing the job outcomes required by Jobcentre Plus and not compromising the extent to which the programme was responsive to refugees with English language needs who faced the greatest difficulties in entering work.

As a means of balancing the demands of Jobcentre Plus for short-term job outcomes and the needs of refugees, it had devised a system for funding initial stages of service provision for users perceived as being less ‘job-ready’ through alternative sources, while targeting Jobcentre Plus funding towards more ‘job-ready’ refugees. Although the provider therefore intended to continue to support refugees facing multiple barriers to employment, this was dependent on whether the provider could successfully access other sources of funding. Moreover, it meant that Jobcentre Plus funding was directed away from the needs of potentially the most disadvantaged among its service users.

Responsive to whose needs? Type of assistance provided

Another dimension of the influence of a work-first and job outcome-oriented approach concerns the type of employment-related assistance provided for refugees. Pressures to achieve short-term job outcomes encouraged a focus among some of the providers on placing refugees in ‘easy to access’, low-skilled and low-paid jobs, which conflicted with supporting refugees to enter a range of types of work related to their skills and interests.

This conflict was reflected in refugees’ experiences of being assisted if not obliged by Jobcentre Plus advisers to look for a narrow range of low-skilled jobs, such as cleaning and factory work, irrespective of their skills and interests. Refugee respondents who had higher-level qualifications and managerial, administrative or professional employment experience before coming to the UK emphasised that they had been unable to access advice and guidance from Jobcentre Plus advisers that was appropriate to their experience and to the types of work in which they were interested. By contrast, in some cases they had been required by advisers to look for lower skilled jobs, such as cleaning and factory work.

Similar tensions were also evident in refugees’ experiences of one of the third sector providers, EMPLOY, that delivered an English language and job search assistance programme funded by Jobcentre Plus. In addition to English language provision, some refugee service users referred to their need to access training and voluntary work or work placements in order to be able to apply for a range of jobs related to their interests. However, they felt that the employment advisers of this provider were, instead, pressuring them to apply for cleaning and factory jobs.

From the perspective of both Jobcentre Plus advisers and staff of EMPLOY, encouraging refugees to look for ‘easy to access’ low-skilled jobs, such as cleaning, irrespective of their previous experience, was the only means of placing those with limited English language proficiency in work in the short term.

Although Jobcentre Plus advisers were able to refer refugees who were claiming Jobseekers Allowance to full-time English language and job search assistance programmes for a period of 26 weeks, the duration of these programmes was considered by advisers to be inadequate for refugees with more limited literacy and English language proficiency to develop a sufficient level of English language skills to be able to enter work by the end of the programme. Re-referral back to further full-time English language or other training programmes was, moreover, restricted for a period of six months. Therefore, the capacity of Jobcentre Plus advisers to be responsive to the skills-related needs of refugees was limited by restrictions on the access of claimants of unemployment-related benefits to education and training provision.

Pressures to achieve job outcomes not only encouraged Jobcentre Plus advisers and staff of EMPLOY to focus their efforts on placing refugees in a narrow range of ‘easy to access’, low-skilled and low-paid jobs, but also in jobs within refugees’ ethnic communities, which were even easier to access as English language proficiency was not required. Indeed, some Jobcentre Plus advisers referred to being dependent on these sources of employment as a means of placing refugees in work.

Similarly, staff of EMPLOY referred to their reliance on more accessible jobs within refugees’ ethnic communities, as it was considered very difficult to place refugees with lower levels of English language proficiency in other types of employment.

While EMPLOY was able to exceed the job outcomes it was required to achieve, staff emphasised that this was primarily a result of some of its programme participants being able to access catering and shop work within their ethnic community.

Although the performance of this provider was successful according to the job outcomes by which it was assessed, these outcomes reflected the social networks of its refugee service users and less the extent to which the provider was responsive to the English language needs of its users. Moreover, by encouraging an orientation towards placing refugees in a limited range of ethnically segmented, low-skilled and low-paid jobs, job outcome measures potentially served to reproduce patterns of labour market inequalities experienced by refugees.

Differences between refugees’ experiences of assistance from EMPLOY and the other third sector providers partly reflected the extent to which the other providers were less dependent on Jobcentre Plus as a funder of their services and, therefore, under less pressure to adopt a job outcome-oriented approach. Types of assistance among the other providers that refugees perceived as responsive to their needs (including respondents who had recently entered or were about to start employment) involved specialist job search programmes targeted at particular professions, such as engineering and health care work. They also involved advice and guidance on, and referrals to, education and training programmes, such as programmes in accounting and business management, in which respondents who had recently entered administrative work had participated. In particular, the arrangement of work placements was perceived favourably, both by respondents with higher and with lower skilled backgrounds, as a means of gaining UK-based work experience in types of work relevant to the skills and interests of individuals. Likewise, work placements were perceived as a means of access to potential employers and as providing a work environment in which to develop English language skills. These perceptions concerned the experiences of users of the provider WORKS that delivered a work placement programme through funding that was not attached to job outcomes. Among these users were, for example, one respondent who had participated in a work placement with a local authority and another who had participated in a work placement with a financial organisation, where they had subsequently been employed. Central to these users’ perceptions of responsiveness was their experiences of the extent to which they were offered ‘choice’ in trying to find work rather than being obliged to enter any job.

The other providers, however, highlighted the difficulties of complying with short-term job outcomes in order to access Jobcentre Plus funding while maintaining a focus on the particular needs of refugees. Those needs might require long-term trajectories of assistance, for example, for those wishing to re-enter higher skilled professions such as engineering.

Not competing for additional Jobcentre Plus funding provided a means of financial risk avoidance, as was the case for REFUGEE NETWORK, which had decided not to re-apply for funding from Jobcentre Plus because of the financial losses it had experienced with a previous programme, having been unable to achieve the required job outcomes within the programme period. However, it also provided a means of maintaining a focus on the needs of refugees. Indeed, WORKS emphasised that it did not wish to compete for additional mainstream sources of funding from Jobcentre Plus as the emphasis on job outcomes would require it to change the orientation of its work placement programme away from assisting refugees to enter types of employment related to their skills and interests towards placing them in any job.