Accountability in personalised Supported Employment-based activation services

Abstract

This article explores the accountability experiences and orientations of frontline workers implementing personalised activation services in the Norwegian Labour and Welfare Administration (NAV). The study draws on observations and interviews (2021), with three teams of employment specialists using Supported Employment as an approach to personalised service provision, in a national programme called Extended Follow-up. Adopting the theoretical lens of accountability as behaviours of account giving, three strategies on how to adopt the accountability regime at the frontline are highlighted: (i) reporting, where one team complied with accountability requirements as the perceived best practices for achieving success in work inclusion; (ii) mitigating, where accountability requirements were fulfilled but combined with attention to how to best meet jobseekers’ needs; and (iii) reframing, where accountability requirements were challenged and redefined. The study highlights how accountability requirements may be interpreted variably, promoting personalised service innovation on the one hand or stagnation on the other.

Introduction

The shift from protecting labour to promoting work (guided by the concept of employment) as the most favoured option for participation in society has become a defining feature of industrialised Western democracies (Bonoli, 2010). Collectively referred to as activation policies, referring to programmes that encourage and compel welfare recipients to find paid employment (Howard, 2012), welfare-to-work programmes are touted as representing ‘a terrain of introducing innovative organisational approaches’ (Bergmark et al., 2017; Van Berkel and Van der Aa, 2012). Combined with the focus on activation policies, there is a call for more personalised and flexible forms of following up for jobseekers that take into consideration the individual’s particular situation (Fuertes and Lindsay, 2016). In this sense, personalisation entails tailoring services to individual client needs, rather than a static set of policy prescriptions (Fuertes and Lindsay, 2016: 526). Howard (2012) notes that personalised service provision in activation calls for an increased focus on case management while at the same time widening the range of options available for frontline professionals to meet client-specific needs. In the case of Extended Follow-up, which is the focus of this study, frontline workers in activation are tasked with finding ‘innovative’ long-term solutions for jobseekers through extended active engagement with the market and other service providers (Bakkeli and Breit, 2021).

However, as scholars of frontline implementation have observed, although personalisation and flexibility are seen as crucial preconditions for labour market integration, especially for people with complex support needs (Rice, 2017), personalised service provision remains rife with contradictions (Andreassen, 2019; Bakkeli and Breit, 2021; Freier and Senghaas, 2021; Fuertes and Lindsay, 2016; Ingold and Stuart, 2015). Key among these contradictions are pressures attached to performance measurement and hierarchical forms of accountability that are intended to control frontline workers’ behaviour (Johnson and Bagatell, 2020; Jordan, 2018; Sinai-Glazer and Krane, 2021).

An extensive body of scholarship focuses on how these forms of hierarchical control, expressed in various forms of formal accountability procedures, continue to create tensions at the frontline of service provision (Brodkin, 2008; Johnson and Bagatell, 2020; McGann et al., 2019; Rice, 2017; Van Berkel and Knies, 2016; Griffith and Smith, 2014). A related body of literature has drawn attention to the challenges faced by frontline workers in personalised activation services while trying to balance user needs and limited resources with performance measurement, resulting in their adopting various forms of coping strategies (Fuertes and Lindsay, 2016; Gjersøe and Strand, 2021; Ingold and Stuart, 2015; Rice, 2017; Yeatman et al., 2009). Nevertheless, this literature has been carried out mainly in the context of contracted-out services that involve pay-for-performance models (Ingold, 2018; Ingold and Stuart, 2015; Jordan, 2018). There is a need for further research to understand how the mechanisms of personalisation and accountability are met at the frontline, when welfare bureaucracies become ‘central players in the new game of flexible and intensive employment services’ (Considine cited in Schaan, 2017: 4).

This study aims to fill this gap and extend this body of scholarship both empirically and theoretically. First, we wish to add to existing knowledge on how accountability mechanisms impact personalised service provision by exploring public-led innovative activation work that often requires challenging and immersive work by frontline workers. Second, we wish to extend the debate on accountability and frontline activation work by highlighting variations in strategies that frontline workers adopt when faced with accountability requirements beyond coping strategies such as parking, (a practice where workers decide to focus on a proportion of their clients), creaming (selecting clients with performance targets that are easy to realise), rule bending, and rule breaking (Tummers et al., 2015; Van Berkel Knies, 2016; Freier and Senghaas, 2021; Gjersøe and Strand, 2021; Raspanti and Saruis, 2021). Our approach is to combine Dubnick’s (2005) concept of account-giving behaviour through reporting, mitigating, and reframing and Brodkin’s (2008) approach to street level accountability that considers how lower-level bureaucrats respond to street level conditions. This approach is based on bounded rationality, and considers policy as ‘uncertain,’ and inductively determines policy’s content based on an analysis of street level practices (322).

The study addresses the following question: how do frontline workers respond to top-down demands for accountability in personalised service provision? The research question will be examined by exploring how teams in three case offices responded to accountability demands. Methodologically, we adopt an approach based on Smith (2006) that examines work as concrete practices and how frontline work is hooked up in institutional relations. This enables us to provide a nuanced understanding of the institutional realities that shape frontline workers’ approaches to back-to work services.

The proceeding section provides an analytical framework for the study, starting from street-level studies, before focusing on accountability and the relevance of the three behaviours of account giving (reporting, mitigating, and reframing) to this area of frontline work. The context of Extended Follow-up is then presented, followed by a description of the research methods used. Using Dubnick (2005), the findings are presented through a comparative framework, highlighting the variations in account-giving behaviour in the three teams. A discussion of the implications of the findings is then presented in lieu of relevant literature.

Accountability and personalisation in frontline work

Most literature highlights accountability as ‘a formal mechanism, an institutional relation, in which an actor can be called to account by a forum’ (Bovens, 2007). Brodkin (2008) notes that this understanding of accountability is based on formal hierarchical mechanisms that regard policy implementation as a linear process. Applied to street-level bureaucracy, this way of framing accountability requires bureaucrats to devise means of putting policy goals into practice, enhancing democratic authority, and avoiding ‘administrative shipwrecks’. Informed by a bureaucratic model of compliance, studies addressing street-level implementation using this logic seek to identify obstacles to the linear progress of policy as it makes its way from legislation to realisation (Brodkin, 2008: 320).

Several studies challenge the compliance model and characterise policy as indeterminate and thus street-level implementation as a continuation of policy by other means. Inspired by Lipsky’s 1980 (cited in Brodkin, 2008) seminal work on street-level bureaucracy, the street-level perspective on accountability assumes bounded rationality and provides ‘a missing link to the behavioural aspects of accountability mechanisms on individual actors; the felt or experienced accountability’ (Lieberherr and Thomann, 2019). Brodkin (2008: 327) notes, ‘…by examining how policy is delivered at the frontlines, a street level approach to accountability has the potential to bring into view those discretionary practices (practices where professionals can consider different options for action when situations are open to interpretation), that systematically shape the policy experience (…) it examines both the conditions of work and the content of practice…to explain the particular form that implementation takes, in specific settings.’

This is relevant, particularly in personalised service delivery, where a web of ‘multiple accountabilities’ (Hill and Hupe, 2007: 296) may produce contradictory action imperatives. Indeed, rather than following pure institutional prescriptions, personalisation requires frontline workers to always be proactive (Lieberherr and Thomann, 2019). Frontline workers rely on multiple accountability relationships (in our study between employment specialists and supervisors, employers, and jobseekers) that may not be aligned, requiring them to constantly determine how to act. In this web of multiple relations, ‘different actors can have different and even conflicting perceptions of accountability’ (Van der Tier et al., 2021: 193). Fuertes and Lindsay (2016: 528) note that accountability systems that target individual caseworkers as a way of minimising costs may lead to more standardised rather than personalised services. They argue, in the UK contracted-out service context, that street-level practice in the face of performance management reflects ‘a rigorous imposition of standardised work-first approaches from the top-down’ (Fuertes and Lindsay, 2016: 529).

Our research focuses on how frontline workers’ behaviour aligns with multiple accountabilities and how this plays out in the context of personalised activation services within a public-sector in-house context. In the case of Extended Follow-up, we argue that standardisation through accountability routines leads to multiple accounting behaviours, reflecting frontline workers’ efforts to ‘reconstruct agency in terms of its own internal logic, rather than the logic of managerial command and control’ (Brodkin, 2008: 328). Dubnick (2005) argues for an understanding of accountability that considers as its core behaviours attached to account giving and the social function of giving accounts. His model is inspired by a sociological understanding of accountability that seeks to connect micro and macro levels of policy implementation (Mulgan, 2000). This understanding combines both the formal and informal conceptualisations of accountability (Lieberherr and Thomann, 2019). This is particularly useful for our understanding of accountability of frontline workers, underlining the multifaceted nature of street-level workers’ ability to ‘apply, adopt or undermine formal policy’ (Fuertes and Lindsay, 2016: 529). Dubnick’s (2005) model suggests three types of account-giving behaviour that may be adopted by individuals called on to justify their actions: reporting, mitigating, and reframing.

Reporting

Dubnick (2005) posits that reporting as a mode of account giving involves providing information at a time and in a place and form specified by a superior. Reporting in the context of public-sector organisations is unique because it establishes a distinct power relation between reporter and principal and reflects a system focused on oversight and control (Dubnick, 2005; Johnson and Bagatell, 2020). Moreover, the very processes of accounting require reporters to adopt the logics of which they are part. Reporting as a mode of account-giving behaviour reflects an understanding of street-level practice rooted in rational models of policy implementation (Brodkin, 2008). Street-level workers are expected to act as ‘neutral’ participants in the policy process, implementing policy as received from above. Reporting is rooted in expectations of compliance, seeking bureaucratic allegiance, through hierarchical accountability (Brodkin, 2008). There is a critique of this approach based on the realisation that in ‘converting policy into administrative practices, street level agencies must often choose among conflicting objectives, and specify abstract policy elements’ (Brodkin, 2008: 321).

Mitigating

As a form of account giving, mitigating aims to bridge the gap between action and expectations. It is negotiated action among individuals whose identities and decisions are shaped by their social roles. According to Dubnick, mitigated account giving begins with conditions that assume that the principal is judging the agent for an act that is regarded as wrong or unexpected (Dubnick, 2005: 388). Applied to street-level account-giving practices, mitigating arises in situations where ‘“limited resources and unremitting pressure to meet measured dimensions of performance” lead to ‘reliance on coping strategies’ (Fuertes and Lindsay, 2016, quoting Brodkin, 2011; Gjersøe and Strand, 2021; Høiland, 2018), such as creaming and parking. In the case of Supported Employment approaches in the Norwegian Labour and Welfare Administration (NAV), Schönfelder et al. (2020) show that frontline workers may resort to giving false accounts as a means of meeting accountability pressures.

Reframing

Reframing, Dubnick (2005) argues, converts the account giver into an account maker with a purpose. The agent seeks to ‘control and transform how a problematic situation is defined and perceived, transforming his/her relationship with the principal, and provides justifications for a course of action’ (Dubnick, 2005: 390). Applied to street-level practices, reframed account giving positions frontline workers as active agents negotiating and reimagining policy, or as Brodkin (2008) puts it, as policy makers, rather than neutral conduits for top-down legislative actions.

Consequently, by combining a street-level bureaucracy research approach with a sociological definition of accountability, this study ‘opens up the black box of street level practice’ (Brodkin, 2008: 319) and bridges the gap between macrolevel policy prescriptions and street-level accountability, as our findings will show (Van der Tier et al., 2021).

Context: NAV’s Extended Follow-up programme

NAV was formed in 2006 after the employment service, social insurance, and local social assistance agencies merged to form a one-stop shop for welfare access. NAV is responsible for work inclusion and therefore provides a rich case for examining frontline service workers’ experiences and orientations towards accountability. NAV launched Extended Follow-up in 2017, assuming responsibility over part of the activities that were contracted out to private providers. This was done by recruiting and training teams of ‘employment specialists.’ Relying on the European Union Supported Employment (EUSE, 2010) framework, employment specialists are required to establish close and trusting relations with jobseekers, map their job preferences and qualifications, establish contact with employers to find them suitable jobs, and maintain close relationships with both workers and employers after the work relationship is established to ensure long-term employment (EUSE, 2010; NAV, 2020).

As such, Extended Follow-up represents a novel ‘work-place oriented’ approach (Gjersøe and Strand, 2021), aiming for personalised and holistic support for both jobseekers and employers (Ingold, 2018). This way of organising services contrasts with both contracted-out service provision, based on market logic, and supply and demand approaches focusing solely on jobseekers or employers, which have been critiqued in numerous studies (Ingold and Stuart, 2015; Jordan, 2018; Orton et al., 2019). This approach combines both a work-first logic, aimed at rapid labour market insertion, and a human capital development logic, which focuses on strengthening individual capabilities to maintain long-term employment (Andreassen, 2019).

In operationalising the five-step model of Supported Employment, NAV has developed a quality guide based on the Individual Placement and Support (IPS) fidelity scale (Bakkeli et al., 2020; Gjersøe and Strand, 2021). This stipulates specific performance indicators at the national and local levels to which employment specialists must adhere. These guidelines are connected to time spent out of the office (the requirement at the time of our study was 40% of the work week out of the office), contacts with employers (ranging from 4-6 per week), caseload (12-20 jobseekers per employment specialist), transition into work (65% over the course of the year), and team composition (maximum 10 per team) (NAV, 2020). In addition, local team leaders have varying reporting requirements for employment specialists regarding how many people they have in training, in work, or on wage subsidies. This contrasts with the mainstream practice in NAV, where although caseworkers have high (100 or even more) caseloads (Fossestøl et al., 2020: 9), they are not necessarily subjected to individual performance measurements.

Methods and data

This study is part of a PhD project registered with the Norwegian Centre for Research Data (NSD) as meeting ethical research guidelines. The data include interviews with and observations of three teams of employment specialists and their leaders at three NAV offices conducted between January and May 2021. We call them Team Peri-Urban, Team City, and Team Rural, after their respective geographical locations.

We purposively selected the teams, considering those that had been in operation for more than three years, to gather the views of employment specialists with both long and short experience in the job. We also selected teams with differing leadership models and sizes to increase the possibility of variation in how employment specialists experience their work. The caseload per employment specialist ranged from 12 to 16, and the jobseekers were selected from NAV’s classifications of follow-up groups that need more than standard support. The client groups in all the teams were reported to have mental and physical health challenges or to be recovering from drug-related abuse (we did not have prior knowledge of their work approaches or performance achievements prior to case selection).

Prior to conducting the personal interviews, the first author attended 21 hours of online video meetings of employment specialists, discussing their daily routines, challenging cases, and strategies. She also attended two half-day seminars in which employment specialists discussed two of the documents used to guide their work: the national professional guidelines and the quality guidelines. During each observation session, the researcher took notes on the main discussion, forms of reporting, recurring issues of contention, references to texts and forms, and descriptions of the virtual room setting. The author also used this participation to recruit respondents for interviews.

The interview data consist of 14 interviews with employment specialists and 4 interviews with leaders of employment specialist teams. All the informants provided verbal and written consent, and all the interviews were conducted and transcribed by the first author using codes such as JSTR (for job specialist Team Rural) and TLTR (for team leader Team Rural) to protect the informants’ anonymity. All the respondents had at least a bachelor’s degree, and they had diverse disciplinary backgrounds, ranging from education to social work, organisational studies, psychology, engineering, and health sciences. The interviews were conversational, and the aim was to elicit experiences related to daily activities, reporting, and documentation connected to their work (Smith, 2006; Campbell and Gregor, 2004). Since documentation and reporting practices were constantly discussed in team meetings, the researcher followed this thread, asking questions about perceptions of reporting and performance management, views on supported employment, and challenges related to their work. Smith (2006) argues for institutional ethnography as a methodological approach to investigate ‘how things actually happen’ and how the daily lives of various people come to be organised and coordinated across different spaces.

Thus, to counteract answers couched in ‘institutional language and discourse’ (Smith, 2006), the researcher asked the respondents to describe what they did from the time they came to work to when they left for the day, with particular emphasis on accountability practices, such as what they reported, how, to whom, how often, and what their experiences related to these reporting routines were. Due to COVID-19 restrictions, the interviews were carried out through video calls (Teams or Zoom).

Data analysis

The first author recorded and transcribed all the interviews verbatim. First, the research team members individually read each transcript to understand the informants’ embodied experience of what they do (McCoy, 2006). Second, we summarised each interview while listening carefully for recurring themes and references to account-giving practices. Third, we interrogated these informant accounts for traces of how the institutional context of NAV and the labour market organised the informants’ work strategies and the social relations that extended beyond the boundaries of informants’ experiences (Campbell and Gregor, 2004). The research team specifically investigated how accountability manifested as a ruling relation (Smith defines ruling relations as institutional complexes that coordinate the everyday work of administration and the lives of those subject to administrative regimes) and how it organises the flexibility of the work of employment specialists at the local level. During this third round of analysis, clear team differences in how employment specialists responded to system-regulated accountability requirements became apparent. We therefore moved the analysis to the team level to explore team particularities. Analysis at the team level revealed differences in approaches that ranged from complying to coping and contesting accountability. This led us to adopt Dubnick’s (2005) framing of accountability as ‘practices of account giving’ related to the social function of giving accounts. Borrowing from Dubnick, we framed these responses as connected to reporting, mitigating, and reframing. This framework informed the analytical categories through which the findings will now be presented.

Findings

In the following section, we present the ways in which the three teams experienced and responded to accountability requirements in their day-to-day practice.

Team Peri-Urban: Account giving as reporting

Employment specialists at Team Peri-Urban expressed strong faith in the Extended Follow-up initiative and the corresponding accounting requirements. There were constant expressions such as ‘We have to use our method’ and ‘As long as we stick to the method’, with the idea that everything would then work out. The employment specialists took pride in fulfilling weekly accountability requirements related to transition to work, time spent outside the office, and number of weekly new employer contacts. Their discussions on Supported Employment as an approach to work inclusion were often connected to discussions of the accountability requirements, and there was a strong sense that fulfilling the reporting requirements meant that the employment specialists were ‘following the method’ and that this would lead to good results.

In the weekly team meetings, the employment specialists took turns reporting how many employers they had visited in the past week, how many jobseekers they currently had in work and with which supportive measures, and their plans for the next week. Several employment specialists expressed anxiety when they had little to report, starting out with an apology, such as ‘I am sorry things are going slowly this week; the employer cancelled on me’ or ‘My jobseeker didn’t turn up, and I don’t know what to do’. These often elicited replies such as ‘I don’t understand how employers do not give our people a chance’ with respect to employers who ‘ghosted employment specialists’ and ‘My people are sleeping too; I don’t know how to wake them up’ with respect to inactive jobseekers.

The employment specialists on this team described their job as ‘full and stressful’ and said, ‘There are not enough hours in the workday’. Many commented that the workday consisted of ‘fighting fires’. Moreover, for the employment specialists, flexibility and availability to both the employer and jobseeker formed the core of how they organised their workday. One explained,

‘We must be available. Things change very quickly. If an employer calls now and says, “This person here has a challenge; can you come now?”, then I must be able to say, “Yes, I can come in the next half hour”, because I have promised that I will be available (…). I have given both my (private) phone number and email to the employer and jobseeker so that they do not have to call the big NAV. That is what our job is: being available.’

This perceived requirement to be flexible and available at all times seemed to create stress when combined with the reporting regime that the employment specialists in this team had to follow:

‘The fact that you should always be available to everyone you meet, it’s tiring. It is a service profession on steroids, where you should be gentle, sociable, have relationships with people you do not know (…). And then it is not only employers, but also jobseekers. That combo role, and then you should (also) be good at documenting, writing reports…. It is a complex role with a great many things you must be good at…’

Indeed, although several employment specialists in this office viewed the accountability regime as an inevitable part of their role, many shared that they ended up becoming stressed and feeling incompetent. One, for instance, pointed out that if by Thursday she had not been outside the office, she felt strong pressure to leave the office and find something to do, no matter what her actual work needs were, to avoid reporting no time out of the office when Friday arrived. Because reporting no time out of the office did not feel good,

‘(…) I notice that I spend a lot of time talking to the jobseekers at the office. And then I sit there on Friday without having been out, yet that is the most important part of the job…. That can be demanding!’

Team Peri-Urban thus adopted reporting as a mode of account giving aimed at fulfilling system-regulated accountability requirements. This response was closely tied to discourses of compliance, represented in arguments such as ‘As long as we stick to the method, things will work out (in the end, if not now)’. It is telling that the employment specialists on this team, compared to those in the two other offices investigated in this study, experienced their job as stressful and felt not good enough or even incompetent.

Team City: Account giving as mitigating

For Team City, the top-down accountability requirements connected to the practice of Supported Employment were met with ambivalence. The team leader and employment specialists had long experience working as both caseworkers and employment specialists in NAV, and most of the team members expressed a recognition of ‘the need to follow the system’: ‘Let us not be naive; we work in an organisation that is politically controlled. We must provide these numbers, to defend ourselves, because it is what the politicians look at’.

Nevertheless, the leader recognised the need for the numbers to mean something and commented that it is not always in the best interests of the jobseeker to focus on reporting. Team City employment specialists pointed out that Supported Employment as a method was ill suited to NAV’s strict regulations and rigorous accountability practices, and their challenge was always trying to fit this ‘round peg into a square hole’. An employment specialist noted this dilemma,

‘Supported employment as a method, is not meant for NAV. If we are to follow the core values, we should have unlimited follow-up; we should not be grading people according to their needs (…), but then you put (supported employment) into this monster that is NAV, with rules, counting, scorecards and everything. Then you start saying, 6 months for limits for such cases, and three years for these; you are trying to trim Supported Employment to fit NAV….’

The informant here referred to the practice in which, in addition to maintaining accountability requirements, NAV set a limit on how long jobseekers could receive Extended Follow-up. At the time of the study, the requirement was between 6 months and a maximum of 3 years. This time limitation was presented as conflicting with the value system of supported employment, which argues for unlimited follow-up.

Our informants also expressed concern about the mismatch between the core assumptions and accountability practices tied to supported employment and Norwegian work life, especially in terms of employment and labour protection:

‘This method has several advantages, but most of the practices tied to it are ill suited to work life in Norway, which has strong job protection. You are asking an employer to take such a big risk; they cannot just fire this person the next day, as they do in the US….’

The respondents were adamant about the need to consider employer perspectives and insisted on using time to build long-term relationships to minimise risk for employers instead of focusing on short-term countable outputs. Thus, key responses to accountability requirements for Team City involved trying to meet accountability demands while attempting to preserve ‘good service’ for both employers and jobseekers. This combination caused indignation and frustration among some employment specialists. Participants, for example, pointed out that choosing to visit employers in whom they were not interested was a way of fulfilling the requirements for supervised employer visits, even though they had no direct interest in visiting these employers. A participant noted this balancing work,

‘After the first field supervision (by the team leader), it became clear to me that in the future, I should choose companies that are not interesting to me (…). I chose companies that I knew I had no interest in cooperating with, because then it does nothing; then, I can be true to the method, because that was what was most important; it did not matter because I knew that I would never talk to that company again.’

This balancing work led to experiences of curtailed flexibility connected to documentation and process results within NAV:

‘I experience very little freedom in my work (…) because you have this with the quantifiable… the detail management, it suffocates you, deprives you of the joy that is most stimulating, when you get a relationship with the employer where you can bring in candidates for years, a generous partner…. The micromanagement suffocates it.’

The requirement for employment specialists to account for how they use their time to the smallest detail is at the centre of the contestations that are inherent in employment specialist work. Although most respondents recognised the importance of being out in the market and building employer relationships, they were sceptical about the counting and reporting regimes attached to these requirements and doubted that being monitored and controlled made them better employment specialists.

Team City respondents seemed trapped between trying to meet the requirements of the system and managing to do their jobs satisfactorily. Several participants presented these two aims as largely incompatible, adding that monitoring and controlling their work practices in fact ignored the core requirements of employment specialist work, such as flexible and long-term trust-building approaches to jobseekers and employers. To mitigate these shortcomings, some employment specialists presented dummy accounts. These included visiting employers that they did not intend to use only to meet the requirement of a good ‘employer contacts’ score while protecting the employers that they wished to contact later: ‘You give them (the Directorate of Labour) what they want, and they back off, so you can do your job’. This double role placed the employment specialists in a situation in which they had to constantly assess how fulfilling a certain reporting practice would affect the actual practice of their work in the job market.

Team Rural: Account giving as reframing

Team Rural’s response to the top-down regulated accountability requirements manifested in the form of reframing what the employment specialists viewed as their core activity: moving people into permanent employment. This team expressed a strong bias against taking the detailed accountability procedures literally, disregarding some micro practices that they deemed cumbersome and unproductive in relation to the mission of moving people into work. The informants frequently commented, ‘We will not do that just to please the system; we will do what we know works and helps us get people into work’. Furthermore,

‘We cannot just be running around hunting for these numbers. I always say that we work with a method, and this method is very qualitative (…). It is not the numbers that say who we are (…). If we have done a good job and the candidate is happy, I do not waste my energy on numbers.’

When asked about the 40% requirement for time spent outside the office, a respondent noted,

‘We must use our time sensibly; for me, success is not a top score on the scale. Success for me is a happy jobseeker, so if we have 20% instead of 40, so what? If 40% is what is important, then I choose to score red….’

Team Rural noted that the 40% out-of-office requirement was unsuitable in geographical contexts such as theirs, which required driving long distances between employers, because it led to a false account of how time spent outside related to the results of moving people into work. They therefore consciously chose not to report time spent driving to meet employers. Sometimes, they also chose not to drive to a certain employer located far away but rather followed up with a video call:

‘I do not have to drive three hours to talk to an employer for 30 minutes, just to please the system with scores. I would rather talk to them on a video call and use my time constructively for both me and the jobseeker.’

The members of Team Rural in this sense tried to redefine and modify the value of accountability parameters to their daily work. The team’s reframing was an attempt to adapt to the central aims of personalisation over micro practices of reporting what happened during the day. The team leader recognised that they needed the numbers to defend their practices in the system but chose not to require weekly reports from the employment specialists; rather, they were to mark time spent outside the office in their calendars, and based on this, the leader would calculate an average for purposes of external evaluation.

For this team, reframing created a sieve through which the employment specialists sifted what they defined as their core role, allowing them room to score low on certain scales while delivering on the core of the Extended Follow-up programme, which was to move 65% of their jobseekers into long-term employment annually. Hence, this team scrutinised the formal accountability requirements for their usefulness (or lack thereof) in helping the employment specialists attain this core goal. This scrutiny was anchored in a preference for strong professional discretion regarding sensible time use for the employers, the jobseekers, and the employment specialists themselves over the team’s score on the accountability scale.

While reframing elicited ‘so what’ responses with respect to noncompliance with NAV accountability practices, it also presented a critical but constructive way through which the frontline workers wrote their own narrative of work inclusion within a system of strict control and regulation. In short, they talked back to the system while at the same time producing overarching results that were beneficial to that system. Reframing created a strong sense of ownership and teamwork in Team Rural, which the respondents attributed to having determined that:

‘(…) some of those details (e.g. 4–6 new employer contacts) make sense maybe to people in the city, but here, we must do things in a way that works for us. We know what our job is; we are grownups who have long experience with the labour market, so some of these things that come from the top, we just laugh and set aside, and do our job.’

Conversely, the informants described their work as ‘full of fun’ and commented that the main challenge that they faced was the detailed control systems from ‘above’, which tried to ‘count things for the sake of counting’.

An analysis of account-giving variations

The empirical analysis reveals that ruling relations (Smith, 2006) are connected to accountability in all three of our case study offices (See Table 1) but are most prominent in Team City and Team Peri-Urban. In Team Rural, we see strong tendencies towards local adaptation (Lundberg and Sataøen, 2019) that aim to contest NAV accountability requirements.

Table 1. Experiences and orientations to accountability in Extended Follow-up among Team Rural, Team Peri-Urban and Team City

A possible explanation for these different orientations towards accountability may lie in team composition, location, and leadership. In Team Rural, for example, both the leadership and team members were recruited outside the NAV system, and respondents noted that many in NAV were asking questions such as ‘What has an engineer got to do with working in NAV? Having no prior experience in NAV or social work education, however, seemed to give this team a more proactive attitude towards what to count and what to disregard. Moreover, Team Rural had fewer potential employers to choose from than the other teams, and counting employer contacts per week seemed unproductive. They therefore focused on deeply engaging with the employers in their locality and building strong bonds that would give consistent long-term access to employment opportunities for their job seekers.

In Team City, there were tensions between quality of service and fulfilling the numbers. These arose because the employment specialists had good knowledge of supported employment, but the leadership was intent on fulfilling the reporting requirements. To mitigate this, employment specialists sometimes used dummy accounts. Such strategies were used as a compromise, allowing them to maintain reporting requirements and quality of service simultaneously. This, however, caused stress for employment specialists, and many respondents felt like they were caught between a rock and a hard place.

Team Peri-Urban’s leadership had limited knowledge of Supported Employment but long experience from leadership in other departments in NAV. Most of the employment specialists on this team also had NAV experience. The team leadership had a tight association of supported employment with accountability parameters, which seemed to be the major driving force for compliance. Over time, this created waves on the team between leaders and employment specialists. The leaders felt that the employment specialists demanded too much autonomy, while the employment specialists felt that their leaders’ limited knowledge of supported employment limited the room that they needed to deliver personalised service.

Concluding discussion

We set out to explore how frontline workers experience and respond to top-down demands for accountability in personalised service provision. Our findings reveal that accountability measures largely limit the room for flexibility needed to implement personalised activation.

Our results show that while employment specialists in Extended Follow-up are seen as having conditions that enable flexibility and personalisation, given that they have much lower caseloads than other caseworkers in NAV, in line with extant literature, the imposition of performance measurements and other forms of accountability limits the room for flexibility, which is the essence of personalised service (Andreassen, 2019; Johnson and Bagatell, 2020; Jordan, 2018; Van Berkel and Knies, 2016; Sinai-Glazer and Krane, 2021).

These accountability requirements coordinated teams across sites, producing various account-giving responses from employment specialists in the form of reporting and mitigating (Dubnick, 2005). This resonates with arguments in a range of studies of frontline implementation in personalised services (Gjersøe and Strand, 2021; Howard, 2012; Ingold, 2018; Tummers et al., 2015) and the literature on accountability in frontline work (Brodkin, 2008; Van Berkel and Knies, 2016; Van der Tier et al., 2021), which point out various coping strategies that frontline workers may adopt when faced with competing demands in service provision. At the same time, our analysis illustrates that when subjected to accountability demands, frontline workers mostly respond as ‘active agents’ (Bakkeli, 2022: 2), interpreting policy mandates by drawing on contextual and professional knowledge. This points to how multiple accountability relationships, in this case, between professional and administrative accountability, may require more than top-down regulation, calling on frontline workers to be more proactive in determining how activation looks in practice (Hill and Hupe, 2007: 296; Lieberherr and Thomann, 2019).

Consequently, our study goes beyond the inevitability of accountability as a linear ruling relation and shows how one of the teams responded to accountability requirements by ‘working around’ them through different accountability practices. While two of the employment specialist teams continued to be largely encumbered by accountability regimes, through reporting, and mitigating, one harnessed resources through reframing, allowing for the flexibility that is required to perform their work.

Through reframing, Team Rural opened space for a more qualitative focus on jobseekers and employer relationships, rather than simply meeting targets set by the system. By not allowing themselves to be limited by procedures and checklists, Team Rural turned into an ‘account maker with a purpose’ (Dubnick, 2005) by choosing to receive low scores on some items, such as time out of the office and weekly new employer contacts, in favour of meeting locally acknowledged needs in their daily work. Also interesting is that employment specialists in Team Peri-Urban, who had a strong focus on meeting the accountability requirements, reported substantially more dissatisfaction with their job and a struggle to move jobseekers into work.

Relating to the larger topic of accountability and personalisation in activation, three main insights can be drawn from our findings. First, in the case of NAV’s Extended Follow-up, our findings suggest that the ability to subvert these accountability requirements, instead of accepting them as inevitable, has the potential to empower employment specialists to deliver personalised labour inclusion for both jobseekers and employers. Second, activation work remains ‘an unfinished domain’ (Sinai-Glazer and Krane 2021), shaped and understood differently at a plurality of sites and creating new subject positions for welfare workers as active, enterprising bureaucrats. Consequently, frontline workers involved in innovative and immersive work situations may find it unproductive to account for the multiple relations that cannot be put into numbers. As Brodkin (cited in Andreassen, 2019: 671) clearly states, ‘performance measures are too rudimentary to capture qualitative aspects of practice…with negative consequences for accountability’. This implies that policy makers should reconsider how parameters of accountability may cause a disconnect between frontline workers’ experiences and their day-to-day practices.

Third, returning to the need for personalised activation practices, this study shows that employment specialists’ practice-generated knowledge should be considered beyond schemes highlighting control and reporting. Within the complex of accountability regimes that govern activation work, ‘what works’ (Bakkeli and Breit, 2021) is a moving target that may be applicable only within a particular context. Therefore, further research should examine how innovative and personalised activation work such as in Extended Follow-up can best be implemented and accounted for through locally adapted practice-generated knowledge, rather than the uniform administration measures that are typical of modern activation schemes.