Introduction
Adopting active labour market policies (ALMPs) is a strategy used to pursue the political aim of work inclusion across western countries, and works to ensure well-functioning labour markets and prevent long-term unemployment and marginalisation. However, ALMPs have been criticised for targeting the supply-side of the labour market, and not directing enough measures towards the demand-side (Cronert, Reference Cronert2019; Ingold and Stuart, Reference Ingold and Stuart2015; Orton et al. Reference Orton, Green, Atfield and Barnes2019; van Berkel et al., Reference van Berkel, Ingold, McGurk, Boselie and Bredgaard2017). A supply-side orientation is claimed to place too much responsibility on the unemployed. Recent developments are therefore based on the assumption that employment services must include both the supply and the demand side of the labour market (Frøyland et al., Reference Frøyland, Andreassen and Innvær2018). Although such a ‘combined workplace-oriented approach’ is highlighted as a positive element of activation policy, this has received little attention in the literature (ibid. Orton et al., Reference Orton, Green, Atfield and Barnes2019). This paper deals with the street-level organisation (SLO) (Brodkin, Reference Brodkin, Brodkin and Marston2013), which has a crucial role as the final implementers of ALMPs through direct interaction with the public. For the staff in SLOs, a combined workplace-oriented approach consists of efforts to convince firms to hire unemployed clients, organise meetings between clients and potential employers and to sustain job placement (Greer et al., Reference Greer, Breidahl, Knuth and Larsen2017: 25). Thus, pursuing this approach of strengthening employer engagement implies that frontline staff move beyond merely acting as ‘information providers’ to taking the role of intermediary workers (Bonet et al., Reference Bonet, Cappelli and Hamori2013; Ingold and Valizade, Reference Ingold and Valizade2017).
However, the role of the SLO as job brokers has traditionally not been as successful as may be imagined. Employers tend to view public employment services as primarily maintaining unemployed peoples’ income security, and not meeting employers’ needs. It is especially the SLO’s role of assisting the weakest workers that may hinder employer cooperation and job brokering (Larsen and Vesan, Reference Larsen and Vesan2012: 466-7). Therefore, the SLO does not have the easiest starting point when it comes to collaborating with employers.
This paper directs attention to the Norwegian Employment and Welfare Service (NAV), which, despite extensive efforts, has not been able to effectively shift unemployed clients into work. This is arguably due to the lack of liaison with local employers, a criticism that has been highlighted in numerous evaluation and research reports (Fossestøl et al., Reference Fossestøl, Breit and Borg2014; Mandal et al., Reference Mandal, Ofte, Jensen and Ose2015; Proba, 2011; 2012; Office of the Auditor General, 2014; Vågeng Committee, 2015). As a response to this critique, NAV offices improved the organisational structure specialising in youth unemployed, as well as adding to their service portfolio ‘Supported Employment’ – an insourced activation measure that constitutes a combined workplace-oriented approach (Frøyland et al., Reference Frøyland, Andreassen and Innvær2018; Strand et al., Reference Strand, Grønningsæter, Nielsen and Holte2020).
While insourcing employment services in the Norwegian context is an initiative to enhance employment services within a holistic SLO context of post-NPM, the international research literature argues that insourcing is a response to the creaming and parking dilemmas of outsourcing (Greer et al., Reference Greer, Breidahl, Knuth and Larsen2017). Carter and Whitworth (Reference Carter and Whitworth2015: 280) point out that the risk of creaming and parking is ‘not exclusive to outsourced provision’ but specifically related to the diverse needs and heterogeneity of users (Carter and Whitworth, Reference Carter and Whitworth2015: 283). In addition, Greer et al. (Reference Greer, Breidahl, Knuth and Larsen2017: 159) argue that
while it is clear that marketization does not deliver on its promises, and that public authorities have some legitimate reasons for insourcing, it is not clear that insourcing will produce better services. Will these services be susceptible to the old accusation of hierarchical services, i.e. that they are insufficiently focused on outcomes and cost savings, unable to innovate, unresponsive to users, and therefore represent poor value for money? We doubt that the answer is so straightforward.
Thus, the research on organisational design of employment services is still ideologically freighted, warranting closer examination of the types of transaction between public authorities and public and non-public actors, and the dilemmas and trade-offs inherent in these.
For the SLOs who engage with employers, organisational dynamics both within and at their boundaries are critical (Ingold, Reference Ingold2018), yet these remain under-explored (ibid. van Berkel et al., Reference van Berkel, Ingold, McGurk, Boselie and Bredgaard2017). Thus, by directing attention to dynamics at the organisational level, i.e. of the SLO, this paper addresses the lack of knowledge on insourcing a combined workplace-oriented approach by asking: how is the combined approach realised within the organisational context of a welfare bureaucracy? The paper sheds light on how the public agency’s role as an intermediary leads to practices of creaming that are typically linked to private providers. Mitigating the risk of creaming clients is a core challenge of outsourcing (Carter and Whitworth, Reference Carter and Whitworth2015); the paper shows that this challenge remains equally significant in public, insourced provision of employment engagement services.
Creaming
The contracting out of employment services that emphasise employer engagement has appeared promising with regard to efficiency and service innovation but has often disappointed (Wright Reference Wright, van Berkel, de Graaf and Sirovátka2011). A major problem of outsourcing is creaming (Considine Reference Considine2001; Maynard-Moody and Musheno, Reference Maynard-Moody and Musheno2003; Carter and Whitworth Reference Carter and Whitworth2015; Greer, et al., Reference Greer, Breidahl, Knuth and Larsen2017; Sainsbury, Reference Sainsbury, van Berkel, Caswell, Kupka and Larsen2017), which can be defined as:
‘creaming’ refers to provider behaviour that prioritises attention for unemployed claimants with fewer barriers to work and who are therefore felt to be easier, cheaper, and also more likely to move into paid work and release outcome payments. (Carter and Whitworth, Reference Carter and Whitworth2015: 279, emphasis added)
Concretely, creaming happens through ‘sorting job-ready clients from those distant from the labour market’ (Greer et al., Reference Greer, Breidahl, Knuth and Larsen2017: 149). The sorting procedures will usually happen rather informally, through subtle cues and the use of basic information on the users. Creaming implies unequal access of clients to services (Greer et al., Reference Greer, Breidahl, Knuth and Larsen2017: 147). Consequently, the labour market chances of the most disadvantaged are not improved by the service.
Creaming is most often linked to financial risk for the external provider who receives ‘payment by results’ through a contractual relationship with the public agency (Greer et al., Reference Greer, Breidahl, Knuth and Larsen2017: 147). In such cases, creaming happens because it releases outcome payments, as stated in the quote above.
As a reaction to the wave of marketisation and outsourcing services, insourcing employment services is introduced as a response to the dilemmas of marketisation such as creaming. Thus, insourcing means to take the services out of the market (Greer et al., Reference Greer, Breidahl, Knuth and Larsen2017: 157). A ‘make’ instead of ‘buy’ decision is typically pursued in order to ‘improve client outcomes or boost the organisational stability of in-house public-sector provision.’ However, creaming is not only to be considered in relation to financial risks of a contracted provider, but to
reflect the reality that providers – of whatever sector – simply may not have the scope or resources to affect certain characteristics of claimants (e.g. mental or physical health conditions) or their local areas (e.g. a chronic lack of suitable jobs locally) that are relevant to their (un)employment. (Carter and Whitworth, Reference Carter and Whitworth2015: 283).
Two factors can affect the risk of creaming (Greer et al., Reference Greer, Breidahl, Knuth and Larsen2017: 147). First, if the client group is homogenous, there are fewer opportunities to differentiate between the users regarding who will achieve best outcomes (ibid.). Creaming involves a selection of clients according to an assessment of their employability (Greer et al., Reference Greer, Breidahl, Knuth and Larsen2017: 49). If the diversity of users is wide, the scope for sorting for employability increases. Second, the extent to which outcomes can be measured plays an important role. If the services involve training, the outcome is less measurable than if the service is job placement. Thus, because the outcome of employer engagement services is highly measurable (obtained employment or non-obtained employment), such services arguably bear a risk of creaming users. Obtaining results through numerical targets is also evident within public administrations that are equipped with a performance management system. Generally, performance management can make staff sort and prioritise those clients who are easiest to help in order to hit targets (Lipsky, Reference Lipsky2010). Creaming thus happens because it is easier to provide services to clients who are more job-ready.
Supported Employment (SE) – A Combined Workplace-Oriented Approach
For decades, Supported Employment (SE) has been used as a model to support people with significant disabilities to find and retain employment in the ordinary, competitive labour market (Bond, Reference Bond2004). The program was originally developed in a US-welfare context, directed at assisting people with developmental disabilities. Later, the program has spread to an ever widening range of user groups such as people with mental illnesses, physical disabilities, traumatic brain injury and autism (Wehman, Reference Wehman2012), but also to more broadly defined groups such as refugees (Sveinsdottir et al., Reference Sveinsdottir, Bull, Evensen, Reme, Knutzen and Lystad2020a) and young unemployed (Bakkeli et al., Reference Bakkeli, Frøyland, Spjelkavik, Berg, Undem and Vestøl2020; Strand et al., Reference Strand, Grønningsæter, Nielsen and Holte2020).
SE constitutes a ‘place-then-train’ strategy, where the employer serves as an arena for training the client, thus focusing on the clients’ needs in relation to a specific employer and thereby becoming directly integrated into the labour market. SE therefore represents a rejection of the so-called ‘train-then-place’ approach – a supply-oriented model of stepwise labour market integration involving preparing the individual in a sheltered environment before taking steps towards entering the workplace. The combined workplace-oriented approach, such as SE, is thus based on a recognition that supply-oriented and demand-oriented measures are equally important (Frøyland et al., Reference Frøyland, Andreassen and Innvær2018). The SE-method also entails that job coaches interact closely with jobseekers and employers as intermediaries and the job brokering work of the job coaches vis-à-vis the employers is vital (Aksnes, Reference Aksnes2019; Ingold, Reference Ingold2018). Such brokering work consists of building relations and negotiating the value of jobseekers’ labour to the employer. Because SE rests on the notion of assisting people with a disadvantaged position in the labour market, one could expect creaming to be less likely to happen when delivering SE services.
Implementing SE in a SLO
As a concept, the street-level organisation (SLO) brings into analytical focus how changes in governance and management affect the local level ‘through patterns of informal practices’ that contribute to advance political change, capturing the ‘new organizational environment’ (Brodkin, Reference Brodkin, Brodkin and Marston2013: 18-19). It builds on and extends Michael Lipsky’s (2010) concept of street-level bureaucracy. The notion of the SLO draws attention to an arena where policy projects are advanced ‘indirectly through administrative means, including through management reforms that alter the arrangements and conditions of street-level work’ (Brodkin, Reference Brodkin, Brodkin and Marston2013: 23). An important point is that the informal practices of the SLO do not only adapt policies but can also affect citizens’ rights.
Ingold (Reference Ingold2018) emphasises the complex role performed by employment advisors in SLOs, which includes approaching and engaging with employers in a ‘sales’ relationship, as well as engaging with clients in order to find good matches. The SLO advisor’s role is performed both internally and externally and can therefore be characterised as a ‘boundary spanner’ role, similar to that of private recruitment agencies (ibid.).
What distinguishes SLO as an intermediary is the disadvantaged nature of their clients compared to the more ‘ordinary’ unemployed clients of private agencies. As an intermediary, the SLO is required to handle their clients’ insecurities at the same time as managing the risk of hampering employer relations with ‘damaged goods’, and also to engage with clients to convince them to commit to a particular job. Rather than contractual, the relationships they manage are interpersonal, based on trust and characterised by the voluntary nature of the employer’s engagement with the SLO (ibid.). The relationship between the SLO and the employer has also been described as taking the form of a ‘reversed asymmetry’ where SLO frontline workers become dependent on the ‘powerful’ employers who they treat as their clients (Raspanti and Saruis, Reference Raspanti and Saruis2021).
Empirical context: Insourcing Supported Employment in the Norwegian Employment and Welfare Service (NAV)
In 2017, the Norwegian Government issued a policy reform with an ‘intensified youth initiative’ instructing NAV to improve in-house employment services directed at young clients (Ministry of Labour and Social Affairs, 2016). The growth in insourcing of employment services in NAV has taken place in tandem with the continuation of a marketisation trend of employment services. This has resulted in Government supporting a dual strategy, where central policy decisions to insource more employment services into NAV coexist with the continuation of the system of market based employment services (Official Expert Report, 2020). Thus, insourcing of employment services in Norway does not represent a reversal of marketisation such as in other countries, but may instead be seen as a policy trend characterised by a high degree of hybridisation in the form of co-existence of both New Public Management (NPM) as well as post-NPM organisational principles in employment services (Christensen and Lægreid, Reference Christensen and Lægreid2011: 407; Fossestøl et al., Reference Fossestøl, Breit, Andreassen and Klemsdal2015).
In implementing the ‘intensified youth initiative’ reform, NAV had to demonstrate an improved ‘ability to shift young into work’, where enhanced employer contact in NAV is considered an important factor. The follow-up of young users is organised in specialised youth teams, consisting of dedicated frontline workers who only handle users under 30 years of age (Strand et al., Reference Strand, Bråthen and Grønningsæter2015; Gjersøe, Reference Gjersøe2021; Strand et al., Reference Strand, Grønningsæter, Nielsen and Holte2020).
In terms of meeting the requirement for enhanced employer contact, the NAV offices have established an in-house, SE-based follow-up service with trained job coaches specialising in the five-step SE method. This five-step model has been implemented in NAV by developing a quality guide based on the IPS (Individual Placement and Support) fidelity scale. The quality guide includes specific performance targets for the job coaches, such as the number of employers that should be contacted per week, percentage time spent with employers, a maximum number of users per job coach and targets in the form of successful job matches (Bakkeli et al., Reference Bakkeli, Frøyland, Spjelkavik, Berg, Undem and Vestøl2020). The first step of the model comprises the job coach meeting the user and making a cooperation agreement defining roles and responsibilities. In the second step, a mapping exercise and plan for the employment is made based on the users’ own wishes, interests, competence, and skills. Steps three and four involve that the job coach search for, visit and recruit relevant employers, with step four being carried out in the form of visits by the job coach to the employers. In step five the job coach links the unemployed client and the employer in a job match. In addition, the job coach carries out follow-up services to both parties, by visiting the previous unemployed at the new workplace.
By implementing SE in-house, NAV is arguably an innovator within activation work, as combined workplace-oriented approaches are not considered ‘part of mainstream activation policies’ (van Berkel, Reference van Berkel2020: 3). SE represents evidence-based knowledge – a knowledge type placed high on the knowledge hierarchy of the NAV Directorate (Breit et al., Reference Breit, Fossestøl and Pedersen2018). The documented positive effects are particularly strong for the IPS variant of SE (Individual Placement and Support), where numerous RCT studies demonstrate higher employability and better work retention for people with mental illnesses (Bond et al., Reference Bond, Drake and Becker2008; Bonfils et al., Reference Bonfils, Hansen, Dalum and Eplov2017; Brinchmann et al., Reference Brinchmann, Widding-Havneraas, Modini, Rinaldi, Moe, McDaid and Mykletun2020; Frederick & VanderWeele Reference Frederick and VanderWeele2019; Sveinsdottir et al., Reference Sveinsdottir, Lie, Bond, Eriksen, Tveito, Grasdal and Reme2020b). Other variants of SE, like the five-step model, lack the same evidence base (Bakkeli et al., Reference Bakkeli, Frøyland, Spjelkavik, Berg, Undem and Vestøl2020), and a recent Norwegian study concluded that the five step SE applied in NAV has the same employment outcomes as ordinary, market based activation measures (Berg et al., Reference Berg, Hauge, Markussen and Zhang2021).
Methods and Research design
In order to shed light on how a combined workplace-oriented approach is realised within the organisational context of a welfare bureaucracy, the paper reports on qualitative data collected for a study of the implementation of the intensified youth initiative reform (Strand et al., Reference Strand, Grønningsæter, Nielsen and Holte2020). Qualitative data are suitable for specifying organisational features of street-level work because such data provide rich contextualised insights to how caseworkers are ‘nested within the organisational context’ (Gofen et al., Reference Gofen, Sella, Gassner and Hupe2019: 343; Kekez, Reference Kekez and Hupe2019).
22 qualitative interviews were conducted with caseworkers in youth teams and job coaches who worked according to the five-step SE model on activating young, as well as office leaders, in five different local NAV offices. In three of the offices, four interviews were conducted, and in two offices, five interviews were conducted. The SE job coaches were not part of the youth team but liaised closely with the caseworkers to select young clients that the job coaches could provide with follow-up after the SE steps. The offices selected were in three different regional parts of Norway and included two larger offices situated in cities, two offices in larger, urban municipalities and one smaller office in a semi-rural municipality. The selection of offices in different regional areas, of different sizes and degree of urbanity secured a variety of information to the empirical study, constituting a strategic selection of offices (Flyvbjerg, Reference Flyvbjerg2006). The strategic selection accommodated variations in offices as well as contextual factors. The different regional areas of the selected offices involved a variation of management through different county office administrations. The different areas also involve specific regional labour markets and unemployment rates. Thus, the offices selected for this study include variations on several dimensions.
The interviews were carried out between June and September 2018. A semi-structured interview guide was used and all interviews were audio recorded and transcribed in full. The research design was approved by the Norwegian Centre for Research Data. The interviewees were informed about the purpose of the research project before the interview started and consented to participate in the research on a voluntary basis. All interviewees were informed that their consent to participate could be withdrawn at any time. None of the interviewees withdrew their consent upon participation. All interviewees are anonymised in this article.
In the first stage of the analysis, the transcripts were read several times and coded inductively for themes. In the next phase, themes were merged and patterns identified in light of theoretical perspectives from SLO theory and previous research. In particular, attention was directed to how governance signals of the youth reform affected the organisational dynamics within and at the boundaries of the NAV office. A central theme emerging from the analysis of data was that SE implied particular dependencies on employers and jobseekers. Concerning the dependence on employers, the job coaches were concerned with building relations, which in turn had consequences for their selection criteria of young users. These selection criteria had implications for the internal dynamic between the staff in the NAV office.
Empirical findings
Building relations
SE and the five-step model represent a resource-demanding service. The job coaches had fewer users than the ‘ordinary’ caseworkers in the youth team and were thus allowed time to provide close follow-up of both the client and employers. One job coach said the following about the resource allocation:
I now have a portfolio of 11-12 youths that I will follow up closely towards work following the Supported Employment method. It is an enormous shift. If I am going to put it into perspective comparing today with where we were 10 years ago, then we are on a totally different planet now.
Although the job coaches had fewer users to follow up (generally between 12 and 20), there were strong expectations placed upon them to deliver results in terms of successful transitions to work for their young clients as well as the number of employers they contacted and formed a relationship with. The performance targets were linked to each of the steps of the SE method, which shaped their work in terms of how much time they spent outside the office and how many new contacts they made with employers. Adhering to the performance targets thus became a way for the job coaches to demonstrate that they were providing high quality activation services.
IPS [Individual Placement and Support – a method within SE] has a gold standard which is 40 percent time spent outside the office and six company contacts, or six cups of tea that are not linked to follow-up, that is six new employer contacts each week. In terms of time outside the office, I am at approximately 20 percent, but in terms of cups or new company contacts, I am on at least six new each week that are not follow-ups of those who are in work. That is six new contacts, or follow-ups of new employers. Some are cups one, two and three. We count cups.
Having ‘cups of tea’ was a strategy for establishing contact with employers. Meeting performance targets, as outlined in the quality guide, became the evidence of success. The informants emphasised that the ideal was for clients to obtain paid work and to receive follow-up at the workplace. The idea to lean more towards the demand-side of the labour market was clearly expressed by one NAV office leader:
And equally important as the close follow-up of the user – and you can quote me on this – is the follow-up of the employers. If we do not have them on board then we do not have an arena for inclusion.
Building close and personal relations with local employers was interpreted by the informants as spending time with them and taking an interest in their recruitment needs. Thus, the demand-side of the activation process was implemented as a type of labour market intervention through relational work. The activation work was thus hinged on the ability of the job coaches to establish personal contact and to build relations with local employers. Consequently, other types of demand-side activation measures were not promoted, such as wage subsidies. The emphasis was rather on developing a one-to-one relationship with each employer, built through the personal investment of each job coach. One of the job coaches explained how this worked:
We need to be close. We will not achieve anything if we cannot build a relationship with an employer. So, we work following a method called three cups of tea, you form a relationship with an employer, talk a little bit about what they do in the company, and then we can match that with what the youth has an interest in, and what they know.
‘Cups of tea’ was about establishing a relationship between the job coach and the employer (as part of steps three and four in the SE method referred to above). When clients started to experience problems at the workplace, the solution pointed to was relational work. Hence, job coaches pointed to themselves, and their ability to build relations with employers and young clients, as the means to making activation work more successful. In other words, the job coaches put themselves centre stage in the activation process. The way that job coaches organised their work could be seen as close to the ideal of a ‘good’ activation worker according to both NAV-leaders, youth team caseworkers and job coaches alike. The ideal activation worker was considered someone who is not solely inside the office but instead is ‘out there’, is more ‘business like’ and who possesses the sought-after ‘labour market competence’. This fits to the job description of the job coaches. However, stimulating employers through relational work in order to promote the labour market participation of vulnerable youth was a task that brought forth a new set of dilemmas for the street-level organisation, which we present in the remaining findings.
Creaming by motivation
According to the interviewees, young users constituted a mixed group but were also similar in that they shared some key background characteristics. Typically, the interviewees reported that almost none of the young users had completed upper secondary education and they had no or little previous work experience, which means they had little to show in terms of non-formal qualifications to compensate for their lack of formal education. In addition, many of the young clients had some form of mental health problem (either self-reported or documented), and some experienced social problems and struggled with their motivation to work. A main idea of SE is that clients should perform real jobs in real workplaces for real wages. The local NAV offices operated within local labour markets. The NAV offices thus had a strong incentive to maintain good relations with local employers. This included reducing the risk of selling employers ‘damaged goods’. The job coaches had a clear perception of the need to provide companies with job-ready candidates. Providing job-ready candidates was not, however, only a matter of matching the jobseeker’s competence with the employer’s needs; an important precondition set by the job coaches was that of the users’ personal motivation for work.
What is important is that they are motivated for work, ‘engine on’, as we say, that they have that. That is the criterion for receiving follow-up services from me.
Personal motivation was set as an essential criterion for admission to SE. The frontline offices were organised in such a way that, prior to the provision of SE via job coaches, the youth teams handled the inflow of all young users to the local office and played an important role in making the young clients ‘job ready’ before transferring them to a job coach. Thus, despite the similarities between the young users, the caseworkers in the youth teams managed to distinguish between those whom they perceived as motivated to work or not. The practice of referring a young person to a job coach after first getting the young client ready implies that the job coach was responsible for the last of several activation measures.
[The NAV office] has a youth team where two frontline employees work specifically with welfare clients under the age of 30. They send bookings to us as job coaches. We then have a meeting and discuss specific cases, and if it is a person we think we can probably find a job for, then we admit that person to a job coach. If there is a lot of difficult diagnosis and health treatment, then we do not think that they should come to us. The idea is that a client should be in a job within six months.
The youth team acted as a buffer for the job coaches by withholding users who were not ready for their activation. In practice, the ordinary caseworkers in the youth teams prepared the young client and assessed their job-readiness and personal motivation before they were referred to a job coach. However, the job coaches also problematised the importance of motivation:
Motivation should really be there, that is the idea. But perhaps that is what we are all struggling with. One of the most important principles in Supported Employment is self-determination, making informed choices. You should receive sufficient information in order to make a choice: I wish to see a job coach or I do not wish to see a job coach. But in NAV it’s like, how real is your choice really? What are the consequences if I say no? OK, then I lose my money. Or if I say yes, then I will get my money. That is how the NAV system works, so real choices when it comes to job measures do not exist, because the consequences for you [the user] will be too large. That makes motivation difficult.
For the young users, there was a threat of benefit sanctions in case of non-compliance. Using motivation as an admission criterion to SE-based activation services was therefore considered problematic for the uncovering of true motivation. The risk of sanctioning left a question mark as to under what conditions users had become motivated, and how robust this motivation would be when the user had to meet the demands of a real workplace.
In addition, the job coaches’ admission criteria created some friction among the SLO staff. Although motivation may put the emphasis on the client, it also serves as a mechanism for omitting the unmotivated, assumed harder-to-place users from the SE service. One of the regular youth caseworkers described having to assess and perhaps generate the users’ motivation before the job coach even met the user:
I am thinking that I am the one who should produce the motivation in my users, and when they are ready and motivated, then I can get some help from the job coaches. But then a lot of the work is also already done.
The performance targets justified notably unequal workloads between the different categories of frontline workers in the youth teams. Job coaches seemingly had a more ‘favourable’ position, with fewer clients, less administrative work, and a portfolio of motivated users. Informants seemed to justify the special status of job coaches within the NAV office by stressing the stricter performance targets placed on job coaches compared to other caseworkers, as well as their specific labour market competence. Anything that reduced the possibility of hitting performance targets would represent a serious challenge for the job coaches and the NAV office. The strict performance targets were thus a necessary precondition to maintaining unequal workloads in the youth team. Consequently, it was also in the interest of the NAV office to maintain a practice that supported creaming in order for the SE-based services to be given the optimal organisational structure to be able to succeed in terms of activating young persons though the use of SE based methods.
Discussion
In order for the SLO to obtain a good reputation among employers in its local area, it is vital to invest in building relations with local employers (Larsen and Vesan, Reference Larsen and Vesan2012; Liechti Reference Liechti2020; Raspanti and Saruis, Reference Raspanti and Saruis2021). The job coaches in the present study appeared to be rather cautious vis-à-vis local employers because this may be a relationship that they will be relying on for a long time. The main effort of the job coaches of building relations with employers, getting to know their business, and then persuading employers to take on young people means that job coaches’ collaboration with employers takes the form of ‘soft’ interventions. This finding is supported by previous research suggesting that job agents were careful to avoid imposing moral arguments of social responsibility on employers (Aksnes, Reference Aksnes2019). Recruitment through informal networks reduces problems associated with asymmetric information between employers and the unemployed (Larsen & Vesan, Reference Larsen and Vesan2012).
Moreover, Larsen and Vesan (Reference Larsen and Vesan2012: 468) argue that the PES has an incentive to hide the barriers of jobless clients from employers. A key challenge for the staff in NAV was to balance the time intensive task of making young clients ‘job ready’ and avoiding the risk of placing job seekers who were unable to carry out work tasks with local employers. Consequently, the youth teams performed a central and critical role in identifying the young users most likely to be able to find work. This practice can be characterised as creaming within the public SLO, as the users with fewer barriers to work were prioritised.
Reducing the risk of employers turning against the public SLO because of bad experiences serves as an important reason for creaming users by motivation. For the job coaches, working with young people who were considered motivated was felt to be easier for maintaining the relationship with employers. By creaming the users internally, i.e. from the youth team to the job coaches, the risk of ‘agitating’ employers with ill-fitting young unemployed users was reduced. In other words, ‘creaming by motivation’ serves the purpose of maintaining high quality interactions with local employers, without risking the collaboration that successful employer engagement relies on. Ingold (Reference Ingold2018: 715-6) points to a similar tension experienced within contracted SLO providers in the UK, where there was an ‘interdependence of the intra-organisational relationships with both client advisors and clients within their SLOs and their inter-organisational relationships with employers at the boundaries of their SLOs.’ While the contracted SLO in the UK was ‘constrained by the individuals on their caseloads’ (Ingold, Reference Ingold2018: 718), SE job coaches in NAV could filter the most motivated clients into their caseloads.
Although payment by results is not a relevant issue when insourcing employment services to the public agency, other factors contribute to the same unintended consequence of creaming. First, because the outcome of employer engagement services is highly measurable in the sense of obtained employment or non-obtained employment, such services bear a risk of creaming users irrespective of the type of provider (Carter and Whitworth, Reference Carter and Whitworth2015; Greer et al., Reference Greer, Breidahl, Knuth and Larsen2017). Second, the target group of young people in NAV constitutes a rather homogenous group in which the great majority of the users had short educational records and empty CVs. Therefore, the young clients’ own motivation was just about the only asset left for the job coaches to sell to an employer. Thus, although a central mechanism behind creaming is regarded to be heterogeneity in the user group (Greer et al., Reference Greer, Breidahl, Knuth and Larsen2017) the findings in the present study suggest that SLOs can also cream users from a homogenous group.
Sorting users according to their job-motivation is a type of psychologisation of activation work, where a person’s inner self is seen as decisive for being successful in finding a job. This has been observed in other studies of activation work, frequently as a strategy used by frontline staff to place the responsibility for activation on the welfare client (Hagelund Reference Hagelund2016; Born and Jensen Reference Born and Jensen2010). Given that the frontline staff in the youth teams hold on to and prepare the users before referring them to a job coach, it can be argued that frontline workers are still quite focused on supply-side activation. van der Aa and van Berkel (Reference van der Aa and van Berkel2014: 25) argue rightfully that ‘to serve employers requires supply-side policies that are able to deliver suitable and motivated candidates’. At the same time, to deliver unmotivated candidates to employers would imply to go against the will of the jobseeker. A point to be made beyond that of personal responsibility of the unemployed and the need for motivation is that personal motivation is an asset that, in the present study, was applied as a strategy by the public SLO to sort users into a specific service. Thus, motivation as a criteria was used strategically to strengthen the SLO’s output. By sorting clients according to their motivation, NAV effectively proposes a select group of clients through SE, a method which was developed to assist people with disadvantages. The motivation criterion requires a strong degree of discretion. As the ‘unmotivated’ clients receive more traditional supply-orientated measures, the SLO’s practice becomes a two-sided policy, in which clients are treated differently according to the discretion-based criterion of personal motivation. In this way, through the introduction of SE with performance targets, the SLO becomes a site for advancing policy projects that can, through informal practices, affect ‘possibilities for making claims on the state’ and ‘asserting rights’ (Brodkin, Reference Brodkin, Brodkin and Marston2013: 17).
The application of SE in NAV is implemented with a high degree of loyalty to the prescribed five-step model, associating quality in the activation work and success in job activation with high scores on the quantifiable quality indicators of the SE manual. Previous research suggests that this may make the method highly manual based, less flexible and risk that attention is being drawn in the activation work to achieving a high score on quality scales, and consequently paying less attention to individual needs and to the needs of fitting the activation work to different local contexts (Bakkeli et al., Reference Bakkeli, Frøyland, Spjelkavik, Berg, Undem and Vestøl2020). Certainly, creating a new job role in NAV in the sense of job coaches who have the defined task of developing relationships with employers has very much been NAV’s answer to the need identified for improving their links to the labour market. However, the combined workplace-oriented approach becomes a rather exclusive service, i.e. an activation measure reserved for only a few users. Moreover, a basic principle of SE is that candidate preparation should not be necessary, as the employer should act as the training arena. The practice of creaming therefore contradicts the main ideas of the place-train logic inherent in SE. This creaming can be seen to result from an unclear notion of who is the main client that the SLO should serve – the client or the employer – a confusion highlighted in previous research (Ingold, Reference Ingold2018; Raspanti and Saruis, Reference Raspanti and Saruis2021).
Carter and Whitworth (Reference Carter and Whitworth2015: 280) argue that
[i]t is important therefore for programme design to ensure that the incentives of the economically rational provider to cream or park claimants, depending on type, is mitigated such that meaningful and equitable employment support is provided for all, irrespective of the level, complexity and cost of their support needs.
This paper demonstrates the complex realities and trade-offs within PES under the combined workplace-oriented approach. Although the sorting of employability into the SE-based service was a practice that both job coaches, caseworkers and managers adapted to, the creaming of young users into SE was not an explicit strategy as such. Rather, the goal of the SLO was to find efficient activation strategies to succeed better with work inclusion of young people. However, PES needs to provide services to all young people registering as unemployed. Since the motivated unemployed may more easily be shifted to work without damaging the relationship to local employers, this means that it was ‘common sense’ to have an organisational structure that seemingly supported ‘optimal’ activation in the sense of SE.
The quest to generate successful job matches implies that the employer’s need for profit and reliable employees is emphasised by the frontline workers, which may risk overriding vulnerable youth’s need for employment. Employers’ need for profit and skilled workers may be difficult for job coaches to overcome only by improving relational work with employers and young clients. Although relational work directed at both employers and young clients may be held as the very advantage of a combined activation approach, relational work alone may prove to be a rather flimsy foundation for activation work, unless supported by some of the more structural approaches available in the demand-side toolbox. This may take the form of agreements between the social partners or systematic work directed at changing the Human Resource Management (HRM) policies of larger corporations or public employers who, for instance, commit themselves to take on a certain share of jobseekers with disadvantages, as in the case of the Dutch Job Agreement (van Berkel, Reference van Berkel2020). van Berkel’s study (ibid.: 12) concluded that ‘a range of policy instruments and employer services is needed to be responsive to what might trigger organisations to become engaged’. Motivation is crucial also for the employers, and, moreover, because hiring disadvantaged jobseekers may pose HRM challenges, it is vital to ‘bridge the gap between motivation and participation’ (van Berkel, Reference van Berkel2020: 13). SE in NAV should constitute an answer to this need to be responsive to employers’ support needs. Based on the findings in the present paper, however, the efforts of the public SLO job coaches and their ability to make local employers open their doors to disadvantaged jobseekers should be backed up by formal agreements. The combination of informal and formal networks could strengthen the SLO’s efforts of implementing the combined workplace-oriented approach, and contribute to making the approach a less exclusive service.
Competing interests
The authors declare none.
