Is it because you can’t, or don’t want to? The implementation of frontline sanctions in Norwegian social assistance

ABSTRACT Welfare conditionality, particularly stringent benefit sanctions, has become increasingly prevalent in Western welfare states in recent decades. Although mandatory activation has received great attention in research, the need for further studies on the implementation of frontline sanctions has become evident, mainly where enabling measures are concerned. This article contributes to frontline and deservingness literature, examining how caseworkers cope with non-compliance cases and which deservingness assessments they invoke, in the Norwegian social assistance. Drawing on fieldwork and interviews with caseworkers at a Labour and Welfare Administration (NAV) office, this study shows how caseworkers informally construct a typology of non-compliance cases: ‘unreachable’, ‘incapable’ and ‘unwilling’. The way caseworkers perceive the client’s attitude, level of need and control over neediness seem to be decisive in distinguishing between cases. Caseworkers display a tendency to attribute non-compliance to incapability, which leads them to renegotiate activation requirements. As such, caseworkers’ deservingness assessments seem to shape sanctioning practices. Based on these findings, this article argues for a more nuanced understanding of frontline sanctions.


Introduction
What happens in frontline social assistance services before a benefit sanction is imposed? The decision to sanctioni.e. reduce, suspend or end a welfare benefitis often made by caseworkers at welfare offices, who assess non-compliance situations based on political, organisational and normative frames. Caseworkers are thus the link between citizens and state: implementing and co-producing activation policies in each encounter (Brodkin, 2013). Earlier research shows sanctions can be imposed in an improper, standardised and punitive manner (Lens, 2006(Lens, , 2009. Assuming caseworkers are key actors in activation policies (Frøyland, 2018), this article studies the implementation of frontline sanctions at the Norwegian social assistance (SA). As such, it further contributes to research calls on decision-making in SA (Solvang, 2017) and the implementation of sanctions within enabling measures (Caswell, Larsen, van Berkel, & Kupka, 2017, p. 197).
Activation policies have become common among the OECD countries since the 1990s, making welfare benefits increasingly conditional on behavioural requirements (Clasen & Clegg, 2007). In conditional welfare, the receipt of benefits is attached to mandatory activities and, benefit sanctions apply for non-compliance with these conditions. The research on benefit sanctions has been inconclusive. While some studies show increased transition to employment (Boockman, Thomsen, & Walter, 2014;Røed & Westlie, 2012), others show transition occurs in detriment of stability, qualifications and earnings (Hillman & Hohenleitner, 2012). 1 Accordingly, some studies indicate sanctions cause stagnation, destitution, limited behavioural change and impaired positive mental health (Malmberg-Heimonen & Vuori, 2005;Watts & Fitzpatrick, 2018). Research seems thus divided into two main trains of thought. One points to sanctions as an incentive and motivator, promoting workforce entry, while the second regards it as coercive and problematic particularly for vulnerable clients.
Studies that focus on sanctioning practices, show how these are shaped by the accountability logic in welfare governance, organisational traits and occupational background of caseworkers . Also, caseworkers can be ambivalent about sanctions: they feel the pressure to achieve organisational goals as well as a commitment to the social work ideals (Caswell & Høybye-Mortensen, 2015), questioning thus social work's compatibility with activation work (Hasenfeld, 1999;Van Berkel & van der AA, 2012).
Sanctions are controversial mainly for two reasons. First, the welfare user 2 is placed in a 'throffer' situation that combines an offer with a threat, i.e. sanctions apply if the activation offer is not accepted or fully executed (Molander & Torsvik, 2015), and secondly due to its increased severity in the last decade (Watts & Fitzpatrick, 2018). Sanctions question the legitimacy of activation and, the state's role in regulating the individual's behaviour (Eriksen & Molander, 2018). In this way, welfare conditionality evokes debates on 'who should get what, and why?' (Van Oorschot, 2000), probing notions of fairness, solidarity, reciprocity and deservingness. The legitimacy of activation is reliant thus on the public belief about welfare users. That is, it is justified by the belief that some are unwilling to work (Molander & Torsvik, 2015). In other words, welfare users' characteristics and behaviours trigger different perceptions of deservingness, which in turn define the degree of legitimacy of welfare benefits (Van Oorchot & Roosma, 2017). However, assessing who are (un)willing to work is an intricate task that requires caseworkers to decide accurately (Molander & Torsvik, 2015).
In social assistance (SA), where income support is often the last resort, sanctions can have more drastic consequences than in other benefits, undermining the goals of income protection, poverty prevention and social exclusion. European welfare states with strict SA sanctions risk violating the right of minimum of subsistence, mainly due to the lack of mitigation clauses. Mitigation clauses include hardship clauses (possibility to apply for allowances after a sanction e.g. food), reparatory clauses (benefit restoration once compliance resumes) and discretionary clauses (grant caseworkers with powers not to sanction or make exemptions) (Eleveld, 2016). Due to these clauses, Norway the country in study, presents mild sanctions compared with other countries. The discretionary clauses are here particular relevant since caseworkers' perceptions of deservingness seem central in assessing sanctions (De Wilde, 2017). Caseworkers use their discretion to assess cases and classify welfare users in worthiness categories (Fineman, 1991;Maynard-Moody & Musheno, 2003), based on the client's willingness to work (Nothdurfter, 2016).
With this background in mind, this article contributes to a broader understanding of frontline sanctions by studying how caseworkers cope with non-compliance cases and which deservingness assessments they invoke. I draw on the Norwegian SA, where activation is mandatory for people under the age of 30 years, and sanctions established as a discretionary clause. This implies that caseworkers are required to assess the 'welfare user's will or capability ' (para. 20a.2.8, Directive 35 of 2012) to comply with the requirements before sanctioning. As such, this study examines more specifically, how caseworkers differentiate unwilling from incapable cases and, what reasons they express for sanctioning.

Discretion at the frontline
Research has given major insights on how caseworkers' discretionary practices become policy (Lipsky, 2010). The 'elaboration of rules, guidelines, or instructions cannot circumscribe the alternatives' (Lipsky, 2010, p. 15) at the frontline, therefore policy implementation involves inevitably discretion. Caseworkers develop thus coping behaviours such as routines and work simplification to deal with this work complexity.
Discretion can be understood as the degree of flexibility caseworkers have 'to judge, decide and act' (Molander & Grimen, 2010, p. 169). These authors distinguish between structural and epistemic discretion. The 'structural' involves the room within which discretion is exercised, i.e. 'an area where one can choose between permitted alternatives of action' (p. 169). Policy delegates this freedom to caseworkers and specifies the standards for action, such as the Norwegian Social Assistance Act of 2009 (para. 20a) that grants caseworkers with the discretionary powers (not) to sanction. 'Epistemic' discretion concerns reasoning, i.e. the 'cognitive activity carried out by an agent when he or she is making judgments and decisions under conditions of indeterminacy' (Wallander & Molander, 2014, p. 2). Decision-making is thus not random; it is based on certain reasons that justify certain choices. The epistemic discretion is particular relevant in this study in that caseworkers' reasoning is central in differentiating unwilling from incapable cases. When deciding whether to sanction, caseworkers consider a set of information about the client's compliance that allows them to define the case as (un)willing to work. They combine this information, reason about the adequacy of the alternatives and then determine which intervention is the most appropriate to that specific case (Molander, 2016, p. 27). This study provides further accounts of which information sets and deservingness notions shape caseworkers' reasoning when coping with non-compliance.

Deservingness approach
These information sets about the clients are often simplified through general beliefs that allow caseworkers to categorise people. The client's main attributes can signal characteristics from certain deserving categories (Jilke & Tummers, 2018). That is, those who are seen as deserving (of help) are more likely to be considered investment worthy (Maynard-Moody & Musheno, 2003) and prioritised due to their efforts and needs (Jilke & Tummers, 2018).
Deservingness literature has focused on the social legitimacy of welfare benefits mainly through public opinion studies (Van Oorchot & Roosma, 2017). Caseworkers' deservingness perceptions have been less studied (De Wilde, 2017). The underlying idea is that stereotypes about welfare users shape public support and the legitimacy of benefits. SA can be linked to poverty and an inability to be selfsufficient, i.e. it is often laden with normative notions about the clients. Deservingness notions are more evident in SA, since it does not require earlier contribution payments. As such, the support of welfare benefits is greater for more deserving welfare users (Van Oorchot & Roosma, 2017).
Van Oorschot (2000) defined a framework of (perceived) client's characteristics that includes five criteria (Van Oorchot & Roosma, 2017, pp. 13-14). The first is control over neediness, which concerns the responsibility level people have for their own situation, i.e. clients who are considered responsible for their neediness are seen as less deserving. The second concerns the client's attitude. Those who are compliant, make efforts and conform to the dominant norms, are regarded as more deserving. The third is reciprocity, which is associated with the willingness to work or contribute. Welfare users who have or are expected to contribute in the future to a common good are perceived as more deserving. Identity is the fourth characteristic, which entails that needy people who are perceived as 'closer to "us"' are more deserving. The fifth concerns the level of need, i.e. those who are perceived as needier are more deserving. In this study, this perspective enables to understand which norms caseworkers consider important when assessing non-compliance cases and sanctioning.

Social assistance in Norway
SA is the final safety-net for people with insufficient means of subsistence. It is means-tested, temporary 3 and aims at securing income protection and making people self-sufficient. SA is administrated by the Norwegian Labour and Welfare Administration (NAV) at the municipal level 4 and claims and eligibility assessments are ongoing. Granted SA can prompt a meeting with a caseworker, who further assesses needs and activation measures. Caseworkers are responsible for assessing claims, follow-up and sanctioning.
As in other OECD countries, Norway introduced work requirements in SA in the early 1990s (Lødemel & Trickey, 2001), however only in 2017 did activation become mandatory, for people under the age of 30 5 (para. 20a, Social Assistance Act of 2009). SA is conditional on participation in work-oriented activities unless 'extreme reasons speak against it' 6 (para. 20a), and is justified for young adults as an 'incentive that motivate and influence the welfare user to overcome difficult life situations and rapidly enter into activity that can help them emerge from passivity' (para. 20a.1, Directive 35 of 2012).
The same perspective is identified at a range of political parties that support mandatory activation based on human capital and paternalistic concerns (Eriksen & Molander, 2018). Both NAV managers (Sadeghi & Terum, 2020) and caseworkers (Gjersøe, Leseth, & Vilhena, 2019;Terum, Torsvik, & Øverbye, 2017) express the same positive and 'caring' approach. These findings differ from international studies (Watts & Fitzpatrick, 2018, p. 99), suggesting thus the need for a broader understanding of sanctions in welfare policy.
NAV offices offer SA clients a variety of activation measures (Grødem & Vilhena, 2019). One is what I here refer to as activation centresan 'in-house' measure managed by the NAV offices. If young clients fail to participate, the benefit can be reduced. However, the Social Assistance Act of 2009 (para. 20a) offers only general guidance on what trigger sanctions: 'In the event of a breach of conditions, a Decision to reduce the benefit can be made, if reduction information was provided at the original benefit Decision'. Caseworkers are required to assess the reasons for non-compliance before imposing sanctions, based on the client's (in)capability or (un)willingness to comply with the requirements, and the consequences sanctions may cause the client/family (para. 20a.2.8, Directive 35 of 2012). In this way, the Norwegian case gives the opportunity to study in depth caseworkers' assessments on unwillingness and incapability.

Methods and data
This case study draws on approximately 110 h of fieldwork and 12 interviews 7 with caseworkers at one medium-sized NAV office 8 in the south of Norway. I observed team and client meetings, activities at the activation centre and interactions in informal settings. Caseworkers were between 30 and 60 years old and all had higher education in social sciences, mostly social work. Their tenure at NAV ranged from three to 20 years. Most of them were women (three men). For clarity purposes, from this point on, caseworkers exclusively assigned to the activation centre are denominated as social workers.
The office was selected based on its activation repertoire, i.e. having its own activation centre established in recent years. The activation centre was located within walking distance of the office and opened four days a week, from 9.00am to1.00pm. The centre was mainly for young SA clients though others attended when appropriate. 9 Despite a variety of clients, caseworkers described the typical clients as young men of Norwegian background with low education and work experience, and complex life stories.
The purpose of the centre was to support people in finding work or training. Activities there included job counselling, job search, CV writing, job applications, interview training, talks and lectures, and networking. Additionally, social workers provided counselling on health, housing, economics, and family issues besides work. The social workers registered the client's weekly attendance at the centre and sent it to the office caseworker. Based on this information, caseworkers authorised biweekly payments. Following Norway's employment rules, sickness days were granted after two months at the centre; prior to that, a medical certificate was required. Absence due to children illness was accepted without certificate. Other situations such as temporary work or job interview were accepted without proof. Otherwise, clients were sanctioned for no-show and arriving or leaving before 11.00am. If they were late (over 15 min) or left early (before 1.00pm), the day's rate was reduced to half. These rules were defined by this NAV office, i.e. rules can vary between municipalities.
Both caseworkers and clients gave their informed consent, and all names are pseudonyms. Field notes were rewritten and interviews transcribed. Data were read, coded, and analysed in Nvivo. Data were first coded inductively to grasp the informants' point of view on activation and sanctions. It was initially coded incident-by-incident (Charmaz, 2008) focusing on what caseworkers said about sanctions and what happened at non-compliance events. Secondly, data were read thoroughly and the initial codes were merged into code groups based on repeating themes, similarities and differences. This stage resulted in the following themes: 'unreachable', 'incapable' and 'unwilling'. I drew backand-forth on the analytical concepts of discretionary reasoning (Molander, 2016) and deservingness (Van Oorschot, 2000) to make sense of how caseworkers cope with non-compliance and the moral evaluations used to construct their assessments.
Coping with non-compliance I was told: 'the best way to understand activation is to attend the activation centre'. In an office meeting yesterday, it was clear that most clients are in an activity of some sort. The centre is the main one. It is located in another building. No NAV logos are visible outside, just an A4 sheet with the centre's name: 'Lucky Chance'. Nina, one of the social workers, shows me around, explaining many of the activities there are group based, which is 'good' for clients with social anxiety, she adds. (Field notes) At this NAV office, the activation centre was the locus of SA mandatory activation, i.e. young clients were typically referred there. Caseworkers characterised it as a low-threshold measure with tailored activities to the clients' needs. They described activation with phrasing that viewed it as enabling: 'a way of helping people', 'show respect and care', 'an opportunity' and 'taking people seriously'. They asserted that it would be 'foolish, not expecting people to be active' because they 'make a plan that is best' for the client and focus on 'the content [of activation]'. As Erik explained: We don't think people are incapacitated because they have social anxiety. Having a disorder that sucks, that doesn't make them incapable of existing, right? We believe (…) that it's very dangerous to make people passive.
Generally, caseworkers shared an understanding that activation is in the client's best interest. When addressing sanctions, they tended to refer to mandatory activation, as Elin says: Elin talks about sanctions as consequences, from a helping perspective: some welfare users may not be fully aware of what it means to receive SA, therefore caseworkers ought to help them. Notwithstanding, the centre had a defined monitoring arrangement, as the following field note describes: Nina opened the team meeting: 'I suggest we start with the list', which meant the clients they were responsible for. I saw the list in front of her. On the left were the names and on the right the weekdays. Next to each name was a handwritten code. I spotted an A and an L, between what looked like scratches. I later learned, that A stood for absent and L for late.
This excerpt shows how non-compliance was determined primarily at the centre. At weekly team meetings, social workers discussed the status of the cases based on the attendance list. All type of concerns were discussed, though commonly the client's steps to employment. If non-compliance was an issue, it became clear at these meetings. Non-compliance was described by absence, being late, irregular attendance, low participation or avoidance. It is essential to note that most clients were described as compliant and willing to work, 'doing what they should do', however caseworkers devoted a considerable amount of time to some non-compliance situations, which are examined in the following sections.
'Hard to get hold of'unreachable This situation exemplifies cases where caseworkers could not 'get hold of' clients. Each morning, the social workers called the absent clients. When clients were consistently unreachable, social workers registered absence and sent it to the office caseworker. In such cases, caseworkers considered sanctions to be the only tool available. As Linda explains: Interviewer: How much absence triggers a reduction? Linda: It varies a lot. I mean, two weeks without any attendance at all, we cannot reach you, and there is no medical certificate or anything, then we discharge you. If the caseworker sends you back when you claim SA again, we can take you in, but until then we take a break.
Linda conveys two impressions. The first is that sanctions vary for different situations and the second is that in cases like Jon's, sanctions were considered a pro forma, which did not seem to inspire doubts since there was 'nothing else they could do'. Unreachable clients were described in two ways: 'not being in complete need' or 'having other options' (unwilling to comply); and being 'too ill to come in [NAV]' (incapable to comply). These perceptions seemed to be based on the prior information about the client. Erik asserts: A lot of them [clients] are not in complete need, but we cannot evaluate that because they have a right [to SA] (…), if they participate in activation. They can live with their parents and get food every day and that is all very well and easy, and they sit in their rooms, gaming. They can choose to participate, but maybe it is easier not to, compared to someone who has an apartment with his girlfriend and is used to money. That's the difference between them.
In discussing 'need', Erik distinguishes behaviours and justifies sanctions. Unreachable clients can be considered in control over their neediness and as such less deserving of help. On the other hand, Martin says: 'It's health related, but they can't even go to the doctor. It's difficult to get documentation or an explanation'. As such, caseworkers considered that sanctions could help them reestablishing contact with the client. Depriving clients of support would compel them to re-contact, proving the client's level of need and attitude. Hence, caseworkers seem to legitimate sanctions in this situation mainly as an opportunity to provide help, if the client needed, rather than a (moral) dilemma.
'I see they struggle'incapable This situation illustrates non-compliance cases in which caseworkers describe clients that 'struggle'. That is, caseworkers define non-compliant behaviours as a result of incapability, vulnerability or a difficult life situation. In Kari's case, Nina defined the reasons for non-compliance as health-related and interpreted it as inhibiting. This assessment was based on prior information: Kari showed signs of poor mental health and her behaviour changed. In such cases, where the client was considered incapable of complying, sanctions were generally regarded as unreasonable since clients were not seen responsible for the situation. Nina discloses in a conversation: 'many of them [clients] are unwellsick, I mean […]'. Caseworkers explained incapability situations mostly with mental health problems such as (social) anxiety and depression, and medically unexplained symptoms. The social workers would still register absence and inform the office caseworker. However, they would together discuss the consequences of sanctions, often resulting in adjustments of the requirements to the client's current needs. Lea explains 'it's best to adjust [the requirements] so that they're achievable' while Ida elaborates on the subject when asked about clients being late: We had one [client] that was always late. We tried everything but he still came late. And it finally emerged that he experienced so much anxiety that it was difficult for him to wait [outside before opening hours] with the others (…) Without figuring that out, we could have pushed him through as many work trainings as we wished. But we made a deal: 'you can come at 9:15. If making small talk is what's difficult'. We deal with it, and talk to him about how to do it at the workplace. Because if you share the problem, we can talk about solutions.
Even if late coming initially triggered reductions in this case, the excerpt shows how caseworkers respond to non-compliance situations by trying to understand what might be causing them. For these reason, creating a 'good relationship' was crucial. Once clients opened up about their life situation, troubles and feelings caseworkers were able to use this new information to establish another norm of action for the intervention. When the requirement changes, a sanction is no longer triggered. Simultaneously, these renegotiations implied making exceptions, something the social worker Olivia was concerned about: We [caseworkers] make the rules of the game. We all have to be on the same page. They [clients] know everyone is different, but how do I (…) justify he's leaving early but she can't? I think (…) it's problematic. It's instable and disrupts the group.
Olivia points that adjustments might cause uncertainty and confusion among clients. When requirements are renegotiated, the attendance rules can seem less consequent than intended.
Sometimes, readjusting requirements was insufficient to meet the client's needs. In these cases, clients would be referred to other benefits, as Linda describes: We have two girls, who are very sick. They try as hard as they can, but they shake like a leaf, or sit and cry, or run and throw up. They try, right? But we have to say: 'okay, how can we make this easier for you?' Sometimes it's worth trying [another benefit].
Cases like this depict the focus on tailoring requirements to the client's needs, rather than automatically sanctioning. In this case, the clients show will to comply by 'keep trying', which seems to contribute to caseworkers' assessment of the situation. Though, some cases became 'unreachable', even after adjustments. Martin recounted that occasionally when he renegotiated requirements, clients 'still disappear for a while and seek help later. Then we agree to try again, but they don't show up, again. That happens. Then I just reduce [SA], again. And so it goes'.
'He is one of those'unwilling Hans, who was 'in and out of NAV' was next on the list. 'We've tried a lot of things', Linda said, 'but he comes in irregularly and nothing works. He sabotages every opportunity'. Olivia retorted: 'he is one of those, he doesn't want to [work]'. Linda observed, 'He needs individual follow-up as we discussed last week'. The others agreed. Linda said that she would contact the [office] caseworker and suggest [another benefit]. (Field notes) The caseworkers interpreted this type of non-compliance as indicative of unwillingness to comply. When Hans was described by Linda as someone who was 'in and out of NAV', it meant that he engaged in work/work training that did not result in employment and then claimed benefits anew. The caseworkers tended to perceive such behaviours as intentional due to other characteristics that contributed to the assessment. When I asked Elin why she sanctioned a client, she described such characteristics: Often he didn't come to the centre. Then he started work training and came late (…) called the employer saying he had overslept on his third day (…) and the employer didn't allow him to continue. But we knew something like that would happen. He learned that if he got a job there he would get 160kr/hour, and thought it was low. Okay, it's not much, but 'when you don't have income is that ok that you receive benefits of less than 160kr/hour?' That was okay for him. Then, it was like 'sorry, but that's not how it works'. You should always try to make it on your own (…) don't say no to the opportunities you getit doesn't work like that.
In this case, the benefit was reduced. The client wanted to return to the centre, but Elin steered him onto another measure with closer individual follow-up. Elin's description not only reveals the reasons for sanctioning, but also which behaviours are expected. In not meeting expectations, the client is regarded as 'unwilling to comply'. Accordingly, the client seems to be considered responsible for his situation of need since he shows low efforts towards employment. Such non-compliant behaviours were usually defined as sabotage (missing appointments, sneaking away, low participation, irregular attendance, refusing offers) and shirking (inconsistent stories, giving excuses, avoiding, low engagement and effort, not being open about problems). When asked about unwilling cases, Linda responds: Shirkers? Yes, very few, but then we have to be strict. We had one whom we had job after job for. She never wanted it, and always had another contact she was going to talk to. Never did, and finally 'you know what? This is foolish. You don't want, no you don't (…)' Interviewer: What do you mean by be strict? Linda: No breaks! While the others who work and make an effortif they want to have a break to smoke, I don't nag them. One of the boys worked voluntarily for [an activity] from early in the morning until late at night; if he had wanted to go home early the next day, it would have been fine. Because I know I can recommend him to an employer, he is reliable. But when they're not (…) come in late, leave early, take breaks, disappear, don't contribute to the group, don't do the taskswe have to go after them a bit (…).
Linda associates compliant behaviours with 'being reliable', which relates to the client's attitudes and notions of reciprocity. The contrary, being unreliable seemed to be associated with less deserving clients. In situations involving recurrent sabotage, caseworkers expressed frustration, suspicion or mistrust, saying things like, 'I get frustrated, he's an adult'; 'I think he's testing us'; 'I really don't get her'. In these cases, the caseworkers confronted the clients, 'called them out', and tightened monitoring activities, e.g. asked for documentation, proof of activities or restricted privileges. Sometimes, clients were referred to work training with cautious recommendations, or other benefits. Others could become unreachable. Making these assessments typically involved uncertainty. Unwilling behaviours could sometimes be hiding health issues. Thus, caseworkers relied 'on other senses'. Nina described decision-making in this way: What Nina calls 'gut feeling' can be interpreted as a combination of knowledge, experience and observed behaviours that supported caseworkers in decision-making. Simultaneously, she indicates that decision-making is not clear-cut. Caseworkers often argued that sanctioning an unwilling client that was ill, could do 'more harm than good'. They seemed to encounter a dilemma: even when determining behaviours as unwilling, they could not be sure whether the client was 'really sick' relying thus on 'gut feeling'. Sanctioning could augment the client's vulnerability, however not sanctioning could perpetuate benefit claims and 'send the wrong message to the others [clients]'.

Discussion and conclusion
This article addresses the implementation of sanctions at the Norwegian SA. It studies how caseworkers cope with non-compliance situations and which deservingness assessments they invoke. Before imposing a sanction, caseworkers assess the client's reasons for non-compliance with the activation requirements. This assessment is based on a vague guideline about the client's will or capability to comply (para. 20a.2.8; Directive 35 of 2012). However, assessing will and capability is a demanding matter that involves normative appraisals about the client's deservingness. Caseworkers interact with clients who experience problems wider that unemployment, as such they assess non-compliance out from the client's context. To cope with the complexity of this work, caseworkers informally construct a typology of non-compliance situations, based on the client's behaviours and characteristics: unreachable; incapable; and unwilling. This categorisation seems thus to determine the subsequent interactions and interventions. In this way, the caseworker's deservingness assessments seem to have implications for sanctioning practices, i.e. how policy is implemented at the frontline. Caseworkers tend to assess non-compliance situations due to incapability, meaning health related issues. Considering that clients already were assessed capable and willing to work at SA claim, there is at first glance no particular reason for high incapability cases. However, as the last safety-net, SA covers different types of need including cases that are not eligible for other (health) benefits. As such, illness was considered out of the client's control spherebeing ill was not a choicetherefore sanctioning would be unfair and unreasonable. The idea that bad personal choices led to illness was non-existent in caseworkers' assessments. Thus, the degree to which the client is considered responsible for his/her situation and the level of efforts made to overcome it, seem decisive in distinguishing unwilling from incapable cases. Sabotage and shrinking were seen as unwillingness signs while health related signs were ascribed incapability. Hence, the extent to which clients were perceived as reliable and open about problems implied trustworthiness and ultimately deservingness. In contrast, withhold of information created suspicion and made activation demanding.
Activation requirements seem to become more lenient or stringent based on how caseworkers defined non-compliance. In incapable cases, caseworkers were inclined to re-evaluate whether the requirements were reasonably defined in the first place, i.e. rather than standardly imposing sanctions caseworkers adjusted the requirements. In a way, caseworkers appeared more lenient towards clients with complex stories. In unwilling cases, besides sanctioning, caseworkers were prone to tighten monitoring activities and steer clients onto other measures or benefits.
Considering caseworker's assessments have implications for sanctioning, reductions for late coming can be challenging. Such reductions might be practiced more leniently than desirable. On the one hand, by re-negotiating requirements caseworkers risk extending the clients' employment (re)integration, undermining the purpose of activation. On the other, enforcing requirements and sanctions standardly can be counterproductive, intensify vulnerability, neglect client's needs and ultimately the SA goals. Discretionary sanctioning practices can be perceived as inconsistent by the clients. For that reason, misinformation and misunderstandings can arise. At the same time, clients can take advantage of caseworker's leniency, to hamper activation and extend benefit periods. However, how clients experience sanctions requires further research.
Regarding the reasons for sanctioning, findings indicate caseworkers invoke notions of fairness and reasonability. That is, in some situations, non-compliance could be due to a combination of incapability and unwillingness, as so sanctions could aggravate vulnerable and impoverished situations ('do more harm than good'). In others, sanctions would serve as a 'message to the others'. Additionally, caseworkers justified sanctions as a tool to re-establish contact with clients that 'really need help', supporting thus other research (Dorsett, 2008). Such justification was used mainly for unreachable clients. In these cases, sanctions were imposed in a more standardised way. The difference between unreachable and unwilling clients is that the first make less efforts (if any) to comply with the requirements, while the second complies in some degree but do not meet the dominant expectations. These distinctions are just possible through face-to-face interactions. In this way, caseworker-client interaction is central in activation work as it allows caseworkers to define unclear situations and implement tailored interventions.
The clients' characteristics related to control over neediness, attitude and level of need appear relevant in caseworkers' assessments. The other deserving criteria: reciprocity and identity were less noticeable in this data. This means that welfare users who were not considered responsible for their situation, presented need more candidly and made efforts to comply were regarded as more deserving, even if in practice they did not do 'what they were supposed to'. These findings provide a more nuanced picture of non-compliance at frontline, showing a tangled phenomenon more far-reaching than compliant vs. non-compliant.
This study contrasts with others that show mechanical sanctioning and discretionary practices as punitive and arbitrary (Lens, 2006(Lens, , 2009. A way of understanding the findings is by looking at caseworkers perceptions on welfare conditionality. They are generally favourable to conditionality in SA, associating it with an enabling perspectivea 'way of helping people progress'. Sanctions can also be understood within this mindseta consequence of non-compliance rather than a punishing mechanism. This helping perspective may serve as an explanation for caseworkers' tendency to interpret non-compliance as incapability cases. Notwithstanding, this perspective reflects a wider context, one where social work traditionally has been engaged in SA in Norway (Røysum, 2017) and where activation (Lødemel & Trickey, 2001) and mild sanctions (Eleveld, 2016) have differed from other countries. Moreover, caseworkers' perceptions on conditionality seem wrapped in the wider positive and paternalistic political discourse in Norway (Eriksen & Molander, 2018). In this line of thought, this study might suggest a welfare conditionality form where sanctioning conciliates with an enabling approacha mild conditionality. Whether a mild conditionality increases transition into employment requires though further research.