Street-level bureaucrat in the introduction programme – client-centred and authority-centred strategies to handle challenging working conditions

ABSTRACT

Departing from the theory of street-level bureaucracy, this article uses a qualitative approach to examine employment officers’ perceptions of working in the Swedish introduction programme for newly arrived migrants. The programme is managed by the Swedish Public Employment Service, and the aim of the study is to illuminate how street-level bureaucrats perceive the prerequisites of implementing integration policy within the introduction programme, and how they respond to these prerequisites. The study shows that the interviewees perceive the working conditions as difficult, characterized by a pronounced tension between organizational demands and migrants’ needs. To manage this dilemma, the street-level bureaucrats apply several coping strategies, and we highlight two broad patterns of practice. Within the client-centred pattern, attempts were made to use discretion to assist the participants in accordance with their needs. Within the authority-centred pattern, the street-level bureaucrats applied a formal and rule-oriented understanding of their assignment, concentrating their efforts on maintaining the functionality of the introduction programme. The most important implication of the study is that it reveals a mismatch between the politically formulated integration policy and the actual needs of the migrants, as perceived by the interviewees. The current integration policy is heavily influenced by a workfare logic, causing the introduction programme to be focussed on providing support connected to labour-market matching. However the programme lacks adequate structures to support its participants to handle e.g. practical, health-related and psychosocial issues that would indirectly facilitate labour-market participation. Thus, this study encourages politicians and policymakers to formulate a more holistic integration policy.

KEYWORDS:

• Street-level bureaucrats
• coping strategies
• integration
• introduction programme
• integration policy

Introduction

Following the conflicts in Afghanistan and Syria, large numbers of forced migrants sought asylum in Sweden during the years 2013 to 2016, with the number peaking in 2015. Of all the EU countries, Sweden received the highest number of asylum seekers per capita in 2015–2016 – a total of 178,100 (Zaun 2017).¹ Evidently, armed conflicts and oppression presents humanitarian crises with severe consequences for the individuals forced to abandon their homes and origins. In Sweden however, the situation was rather interpreted by the government as a threat towards the national order and security, alleged to put an unbearable pressure on the welfare system (Lundberg 2017; Hansen 2017), which led to the introduction of new legislation that made it more difficult for asylum seekers to get a residence permit (Act 2016:752) . Meanwhile, far from the political corridors where asylum policy was being debated, professionals at the street-level continued their endeavours to assist those who had been granted residence permits to manage their situation in Sweden.

Many of the asylum seekers who came to Sweden before 2017 were assessed as needing protection, and were granted permanent or (after 2016) temporary residence permits. These individuals – as all migrants who get a residence permit in Sweden – are offered a two-year introduction programme. Since 2010 this programme has been managed by the Swedish Public Employment Service (SPES), and its aim is that the migrant is financially self-sufficient within two years. When the programme was introduced, it was designed and staffed to include about 20,000 individuals, but due to the changing migration patterns, in 2017 the programme served almost four times that number, resulting in significantly increased caseloads (SPES 2017a, 35). Also, compared to other labour-market programmes at SPES, the introduction programme contained more administrative duties, giving the employment officers less time to work with the programme participants (SNAO 2015, 129).

Using a qualitative approach, the aim of the article is to examine how employment officers perceive the prerequisites of implementing integration policy within the Swedish introduction programme, and how they respond to these prerequisites.² The article’s point of departure is the theory of street-level bureaucracy developed by Lipsky (2010). Central to this theory is the understanding that public street-level bureaucrats continually balance organizational demands and service users’ needs. In every specific institutional context, the nature of – and relation between – political policy expectations, organizational prerequisites and clients’ needs create certain working conditions. Depending on these working conditions, the street-level bureaucrats exercise discretion to apply different ‘coping strategies’ and ‘patterns of practice’ as they try to handle work. The theory helps us understand how policy operates through the practice of the employment officers, which, in turn, will affect the participants opportunities for (labour-market) integration. The study is guided by two interlinked questions:

• How are working conditions perceived by the street-level bureaucrats employed in the introduction programme?

• What coping strategies and patterns of practice are revealed through the interviews with the street-level bureaucrats?

The main ambition of the article is to make an empirical contribution. In this regard, the empirical findings show that the employment officers perceive the working conditions as considerably difficult, characterized by insufficient resources and a pronounced mismatch between the integration policy (i.e. what support they are expected to deliver) and the perceived need of the participants of the introduction programme (i.e. what support participants require). Theoretically, Lipsky’s concepts of ‘discretion’, ‘coping strategies’ and ‘patterns of practice’ are used to analyse the empirical findings, and in terms of theoretical contribution, we outline a client-centred and an authority-centred pattern of practice in order to understand how street-level bureaucrats handle difficult working conditions. Each of these patterns of practice include several reoccurring and interlinked coping strategies. In the client-centred pattern, the coping strategies applied aim at assisting the participants take their individual circumstances as a starting point, while in the authority-centred pattern, the employment officers concentrate their efforts on maintaining the rules and functionality of the introduction programme. We also contribute theoretically by discussing how these two patterns of practice relate to the citizen agent vs. state agent narratives proposed by Maynard-Moody and Musheno (2000). Here, the study shows that street-level bureaucrats, even if closer to the position of the citizen agent, can orient actively towards ideals characteristic of the state agent narrative as well.

Theory: street-level bureaucrats and coping strategies

Lipsky (2010) terms public services such as schools, the police and the social services ‘street-level bureaucracies’, and the front-line practitioners as ‘street-level bureaucrats’ (cf. Prottas 1979). The work of these street-level bureaucrats is characterized by (at least) two common denominators.

Firstly, their work cannot be fully predicted, scripted or controlled by routines and legislation. By necessity, street-level bureaucrats must exercise an amount of discretion to make decisions in response to the complexity of each specific case (Lipsky 2010; Brodkin 2011). Such professional discretion has been questioned because it is difficult to monitor and account for (e.g. Polsky 1991), yet is often understood as a vital aspect of work that enables individualized treatment (e.g. Handler 1986; Wallander and Molander 2014).

Secondly, their work is characterized by a constant tension between, on the one hand, formal policies and organizational prerequisites, and, on the other hand, the needs of their clients (Lipsky 2010, 162; cf. Borelli and Lindberg 2018). Lipsky formulates this tension as the dilemma of the street-level bureaucrats, where difficult working conditions – such as limited time and resources, and policy directives that are contradictory or perceived as unrealistic to achieve – prevent street-level bureaucrats from conducting their work in direct agreement with their professional judgement (Lipsky 2010, xi–xii; cf. Tummers et al. 2015). Working conditions might differ depending on organizational and national context, creating diverse possibilities for street-level bureaucrats to use discretion and to deliver acceptable services (see e.g. Hupe and Buffat 2014). Thus, while the use of discretion is directed by street-level bureaucrats’ individual attitudes and judgements, the expressions of discretion is also affected by the institutional context (cf. Wallander and Molander 2014).

Coping strategies and patterns of practice

To sustain services under harsh conditions, Lipsky suggests that the street-level bureaucrats develop coping strategies. Lipsky sees heavy caseloads as the main reason for this need to create ‘shortcuts and simplifications to cope with the press of their responsibilities’ (Lipsky 2010, 18), and has identified an array of coping strategies, including: allocating resources unevenly; increasing control of clients; psychological and emotional detachment; formulating personal goals that depart from policy; specialization that de-emphasizes complex tasks; routinization; and denying discretion.

In Lipsky’s original theory, coping strategies are formulated as a way for street-level bureaucrats to ease the burden of their work, typically risking negative outcome for clients. Many have continued to theorize coping strategies (e.g. Winter and Lehmann Nielsen 2008; Brodkin 2011, 2012; Astvik and Melin 2013; Borelli and Lindberg 2018; Jilke and Tummers 2018),³ and more recent scholars have emphasized that coping strategies can also be deliberately applied to support clients. For instance, Evans (2013, 748) shows how street-level bureaucrats bend and break organizational rules to benefit clients, and Dubois (2010) highlights using personal resources to assist clients as a coping strategy. Tummers et al. (2015) classify coping strategies in three distinguished categories: ‘moving towards clients’, ‘moving away from clients’ and ‘moving against clients’. This clarifies that coping strategies can be used to benefit clients, as well as to benefit the street-level bureaucrat or to reprimand clients. For this article, we adopt the definition of coping strategies proposed by Tummers et al. (2015, 1100) as ‘behavioral efforts frontline workers employ when interacting with clients, in order to master, tolerate, or reduce external and internal demands and conflicts they face on an everyday basis’.

When street-level bureaucrats’ coping strategies reoccur concurrently they can be understood as a ‘pattern of practice’ (Lipsky 2010, 81). Investigating the introduction programme at SPES, Larsson (2015, 194) revealed two patterns of practice. From the ‘social worker position’, the street-level bureaucrats used discretion to transgress the formal assignment and to focus on the totality of the client’s life. From the ‘employment officer position’ the street-level bureaucrats followed the formal duties and focussed narrowly on labour-related issues. These positionings recall the two basic narratives of street-level bureaucrats theorized by Maynard-Moody and Musheno (2000): the ‘state agent’ and the ‘citizen agent’. The state agent narrative is described as the prevailing scholarly narrative, and the ideal-typical state agents implement policy and ‘conform to rules, procedures and orders’ (Maynard-Moody and Musheno 2000, 337; cf. Oberfield 2010). In this narrative, discretion is understood as inevitable, yet mainly problematic, as it alters policy (cf. Polsky 1991) and is said to be guided by street-level workers’ self-interest (Maynard-Moody and Musheno 2000, 339). The citizen agent narrative, on the other hand, is described as the street-level workers’ own perception of their role. The citizen agent does not focus on policy or rules, even if they mostly follow them (cf. Meyers, Glaser, and Donald 1998). Rather, the citizen agent focusses on the relationship with clients, using discretion to respond to the clients’ needs and specific circumstances (Maynard-Moody and Musheno 2000, 348). However, being a citizen agent does not mean always doing everything to assist clients. Citizen agents are pragmatists, rather than idealists, as they are aware of the restraints of sometimes unrealistic policies. But rather than strictly following rules, they exercise moral and normative judgement (of deservingness) to determine when and how to help citizens (Maynard-Moody and Musheno 2000, 351; cf. Borelli and Lindberg 2018; Jilke and Tummers 2018). If the citizen agent deems it appropriate, they may transgress rules to assist clients, while falling back on the rules if they consider the client unworthy.

According to Lipsky (2010, 18, 82), coping strategies are often used without explicit organizational consent, and counter to official policy. Nonetheless, they are often crucial in keeping the agency functioning (Lipsky 2010; cf. Brodkin 2011), indicating that the use of coping strategies should not be interpreted as an implementation failure (Holm Vohnsen 2017). However, when difficult working conditions compel street-level bureaucrats to resort to coping strategies, this inevitably affects the quality of work (Brodkin 2011, 2012; Astvik and Melin 2013). Thus, investigating the coping strategies applied by SPES employment officers is important, as they will affect the services provided and thereby the opportunities for the participants of the introduction programme to achieve (labour-market) integration.

The employment service and the introduction programme

The Swedish Public Employment Service (SPES) is Sweden’s second largest government agency, employing 13,700 individuals in 2018, only outnumbered by the Police Authority.⁴ The mission of SPES is to support the unemployed and to match jobseekers with available work, thereby contributing to a well-functioning labour market. In 2010, the responsibility for the introduction programme was transferred from Sweden’s municipalities to the state and was established as a particular labour-market programme at SPES. The intent was to make the programme uniform, and to increase the focus on labour-market integration (Qvist 2012; Larsson 2015; Brännström et al. 2018) – reflecting the workfare logic and activation policy that characterize most contemporary welfare provision (see e.g. Greer 2016; Floros and Bak Jørgensen 2020). Approximately 15% of the unemployed registered at SPES in 2018 were participants in the introduction programme (SPES 2019a, 8). The other operations of SPES include general support targeting all unemployed, as well as a variety of other particular labour-market programmes intended to assist e.g. young adults, long-term unemployed and disabled people (see SPES 2019b for an overview).

The introduction programme

SPES translates the Swedish name of the introduction programme (‘Etableringsprogrammet’) as the ‘establishment programme’ in English. However, here we use the more theoretically and internationally known term ‘introduction programme’. In Sweden, the field of politics that concerns migrants is still called ‘integration policy’. In practice however, there has been a conscious shift in terminology towards talking about ‘establishment’ rather than ‘integration’, and this shift in terminology is alleged to increase the focus on labour-market participation.⁵ In content, the Swedish ‘establishment programme’ is similar to the equivalent programmes in the other Nordic countries, however, e.g. mandatory language tests have not been part of the Swedish programme (see Fernandes 2015).

Together with their employment officer, each participant of the introduction programme sets up an ‘introduction plan’ filled with activities corresponding to full-time occupation (40 hours per week). The activities consist primarily of language training and support aimed at labour-market participation,⁶ but also contain 60 hours of civic orientation (see SPES 2016). The programme is voluntary, but full-time participation is mandatory to receive the introduction benefit of approximately 6,000 Swedish crowns (approx. EUR 570) per month. The programme runs for two years, and the goal is that the individual will be self-sufficient upon completion. Far from all participants reach this goal, but the share of participants leaving the programme for employment or education has increased from below 30% in 2014 to above 43% in September 2018 (SPES 2018, 18).⁷

Between 2010 and 2017 – thus, when the study was conducted – the introduction programme was regulated by a specific law (Act 2010:197) granting the participants a relatively strong right to support (e.g. financial benefits and the possibility to decide what kind of activities to partake in), compared to the rights of participants of other labour-market programmes run by SPES. This act was terminated as of 1 January 2018, and the new legislation (Act 2017:584; Regulation 2017:819, 2017:820 and Regulation 2017:819) reduces the rights of the participants in the introduction programme. The new regulation harmonizes with the legislation guiding the other labour-market programmes at SPES, resulting in extended possibilities to assign participants with mandatory activities and increased possibilities to sanction participants who fail to follow their introduction plan by reducing their introduction benefit (SAPM 2019, 14). Also, the new legislation aims to decrease the administrative duties of the employment officers working in the programme (SAPM 2019, 11). However, the actual content of the programme has remained much the same.

The street-level bureaucrats

All professionals interviewed for this study are employed as ‘employment officers’ (in Swedish ‘arbetsförmedlare’) in the introduction programme. Hence, they share the same title and principal assignment. Among the traditional street-level professions (see Tummers et al. 2015, 1107), the tasks of the employment officers most closely resemble those of a social worker, including assessment of needs and providing assistance to improve the lives of the individuals they meet (cf. Dominelli 2004). However, there is no requirement of social work training to be appointed as an employment officer, and the informants represent a variety of professions, including political scientists, social workers, a psychologist and even a teacher. Thus, the employment officers are a heterogenous group, representing several street-level professions. Presumably, this diversity could affect how they manage their work. Tentatively, due to their training, a social worker by profession might orient more towards the client’s perspectives, while a political scientist might orient more towards the bureaucratic aspects of the job. However, our material is too small to draw any conclusions as to whether diverse coping strategies are linked to certain professions, and Tummers et al. (2015, 1107) show that street-level workers within the same formal profession adopt various coping strategies as well. Thus, for this article we will analyse the informants as a group – as the street-level bureaucrats of the introduction programme. In the following text, we will refer to the informants as employment officers or interviewees when discussing their specific practice or statements, and use the term street-level bureaucrats when making analytical remarks.

Compared to other labour-market programmes at SPES, the duties of the employment officers in the introduction programme include a higher proportion of administrative duties, such as handling payment of financial benefits. The working tasks of the employment officers in the introduction programme are also more varied than those of other programmes as they, besides managing activities oriented directly towards labour-market activation, also managed activities linked to e.g. language training and civic orientation.

Few studies have investigated employment officers’ perceptions of working in this introduction programme. However, in a doctoral thesis, Larsson (2015) interviewed employment officers in the introduction programme. Larsson reveals a work situation that is characterized by limited resources and heaped with administrative tasks (Larsson 2015, 189), where managerial and political demands contradict the needs of the participants (Larsson 2015, 173ff; cf. Neergaard 2004). In an earlier study, Schierenbeck (2003) compared how street-level bureaucrats that work with migrants in Sweden and Israel handle their tasks, arguing that – compared to the more strict rule-following in the Israeli case – the pronounced discretion and focus on individual judgement in the Swedish case might actually hamper labour-market integration, as it enables a practice based on stereotypical perceptions.⁸

The participants

The introduction programme is open to all individuals between 20 and 65 years who have been granted a residence permit, independent of the reason for migration. SPES calls this group ‘newly arrived migrants’. In policy terms, ‘newly arrived’ refers to the first two years after being granted the residence permit, but due to individual circumstances – such as childbirth or illness – some participants in the programme have held a residence permit for longer. Moreover, the term ‘newly arrived’ can be misleading, as many individuals reside in Sweden for quite some time before their permit is finally decided upon. Thus, in this text, we use the terms ‘migrant’ and ‘participant’ for the individuals participating in the programme.

At the time of the study (2016–2017), the vast majority of the participants in the introduction programme were former asylum seekers who had been granted protection – primarily from Syria and Afghanistan (SPES 2017a, 11–12) – or their relatives, being granted residence permits through family connection (SPES 2017a, 13). Even though a new legislation (Act 2016:752) normalized temporary residence permits in 2016, due to the long processing time of residence permit applications, almost all participants of the introduction programme at the time held permanent residence permits.⁹ Thus, in the context of our study, the typical participant in the introduction programme held a permanent residence permit and had experiences of forced migration, presumably indicating needs that differ from individuals migrating willingly.

According to SPES, the participants in the introduction programme at the time of the study had a lower level of education than prior cohorts of migrants and Swedish citizens in general. Approximately half of the participants had not completed the equivalent of Swedish upper secondary school (SPES 2017b, 14f). At the same time, a third of the participants had completed higher education. Similarly, the participants of the programme also constituted a diverse group when it comes to age and prior occupation (see SPES 2018). Concerning gender, in 2016–2017 there was an overrepresentation of male participants: 57% (SPES 2018, 12). In sum, the participants of the programme represented a considerably diverse group. This likely posed a challenge to SPES in providing adequate activities, as it requires support that varies in extent and content (see Qvist and Tovatt 2014, 55ff; SNAO 2015).

Method and empirical material

This article is part of the study ‘The professional and the introduction programme’ (Johansson 2018) and is based on three focus group interviews with professionals hired as employment officers in the introduction programme at three local SPES offices. The groups consisted of 5 to 7 individuals, and a total of 17 informants participated. The study was carried out in consent with SPES, and the local offices were selected in dialogue with officials at the SPES head office. The aim was to reach local SPES offices in municipalities of different size, and where the introduction programme was an established and active part of the operation of SPES. The three offices selected were geographically separated, but all situated in the central part of Sweden, located in a small, medium and large municipality respectively. Information about the study was spread through local managers to all employment officers working in the introduction programme at the SPES offices in question, and the officers interested in participating contacted us. Thus, the selection was based on self-selection. The interviews were led by one of the authors and conducted between March and May 2017, coinciding with the peak of participants in the programme (see SPES 2018, 12). This gave us an opportunity to investigate the perceptions of working at an authority that was under exceptional pressure. At the same time, the timing of the study may be reflected in the results, as working conditions might have been perceived as particularly difficult.

The informants from each office partook in the same focus group, and all interviews were conducted under informed consent at the participants’ workplace. Thus, the participants of each focus group presumably knew, or knew of, each other. In general, the participants were active and the discussions relatively self-propelled. However, even if everyone was involved in the discussions, some informants were more inclined than others to speak, especially when it came to raising critical comments about the operation of the local SPES office. The interviews lasted between 60 and 120 minutes and were recorded in full. Two interviews have been transcribed in their entirety. The third interview was repeatedly interrupted by non-informants entering the room, as well as by informants entering, leaving and re-entering. Thus, this interview was fragmentized and has only been transcribed partially. Consequently, most excerpts come from two of the three interviews. Nevertheless, the analysis is consistent with what was expressed in the third interview. When presented in the article, all names of places and individuals, as well as other signifying characteristics, have been removed or anonymized.

The purpose of the analysis has been to find recurring patterns in the empirical material at a detailed level. Through repeated readings of the material, ‘patterns of recognition’ emerged, which have been categorized and organized into main themes and sub-themes (cf. Braun and Clarke 2006). For this analysis, a theoretical model was determined in advance. We used Lipsky’s (2010) theoretical framework to approach the material, looking for sub-themes in four theoretically implied categories: ‘political and organizational expectations’, ‘participants’ needs’, ‘coping strategies’ and ‘patterns of practice’. These four categories were applied because they harmonize with the empirical aim of the study: to investigate how employment officers perceive the prerequisites of implementing integration policy within the Swedish introduction programme, and how they respond to these prerequisites. In theoretical terms, the nature of – and relation between – political and organizational expectations and clients’ needs constitutes the working conditions (i.e. the perceived prerequisites) under which the employment officers carry out their work (i.e. implement integration policy) (cf. Lipsky 2010, xi–xii,162; Tummers et al. 2015). Depending on these working conditions, employment officers exercise discretion to apply different ‘coping strategies’ and ‘patterns of practice’ as they try to handle work and meet service users’ needs (i.e. respond to the prerequisites of work) (cf. Lipsky 2010, 18, 81; Tummers et al. 2015; Larsson 2015). The sub-themes presented under each theoretical category in the analysis are those that most frequently reoccurred in the empirical material.

Handling an ‘impossible’ assignment

Now we will focus on the introduction programme and the street-level bureaucrats working in it. In the first two sections, we close in on the working conditions by comparing the political and organizational expectations with the needs of the participants. Here, a pronounced gap between policy and needs is revealed (cf. Larsson 2015; Eriksson 2018, 2019). In the third and fourth sections we highlight two broad ‘patterns of practice’ that the employment officers use to handle the dilemmas of their work. Under each pattern we elaborate on common coping strategies.

Political and organizational expectations

The concept of ‘integration’ – its meaning, definition and implications – has been widely debated and contested among scholars (see. e.g. Kamali 1999; Rytter 2018). The ambition of this analysis is not to engage in the debate concerning the definition or use of the term integration. Rather, we investigate how the existing Swedish integration policy (see Government Missive 2008/09:24; Proposition 2009/10:60; Act 2010:197, 2017:358) is applied within the SPES introduction programme. Levitas (2005) argues that ‘integration’ has been more or less equated with labour-market participation in the contemporary political discourse, and Sweden is no exception. Over the last fifteen years, the Swedish integration policy has been influenced by the workfare logic permeating contemporary social policy, shifting focus from social inclusion, cohesion and counteracting discrimination, towards labour market participation (Qvist 2012; Larsson 2015; Brännström et al. 2018). Consequently, the introduction programme is focused on activation and heavily labour-market oriented. The political expectation is that the participants should be self-sufficient and occupied with work or education after the end of the two-year programme (Proposition 2009, 69). This goal has proven difficult to achieve, and the government has given SPES strict directives to improve this result significantly (Ministry of Labour 2018). When the migrant numbers increased dramatically around 2015, the programme had to expand rapidly. However, the funding did not increase nearly as much as the numbers of participants (SNAO 2015, 129ff; SPES 2017a, 35). This affected working conditions in mainly two ways: (1) the caseloads of every employment officer increased massively, and (2), the system was suddenly considerably under-dimensioned, resulting in bottlenecks and delays to access activities (SNAO 2015; SPES 2017a; SAPM 2019).

In a study by the trade union ST (2012) reaching 2500 employment officers at SPES, 57% stated that they lacked time and resources to properly carry out their assignment. Only 7% stated that they had enough time and resources. This shows that the working conditions at SPES were perceived as harsh well before the increase in participant numbers in the introduction programme, suggesting that the cumulating caseloads aggravated an already difficult working situation, rather than giving rise to the agency’s difficulties. One of our interviewees described being transferred to the introduction programme when the numbers of participants rapidly rose in 2016:

You were supposed to increase gradually over six to nine months, but I was at full caseload after three months, and after six I had more than a hundred cases … we were on our knees … many [employment offices] didn’t want to stay, they just quit. (Focus group 1)

This quote captures the common experience in the focus groups. When the introduction programme was designed, the ideal caseload was set to 30–40 cases per officer, a limit commonly exceeded, sometimes vastly. And as the number of cases double or triple, it is safe to say that this changes the prerequisites of the work that the street-level bureaucrats can carry out.

There is also a strict policy expectation that the street-level bureaucrats assign the programme participants activities equivalent to full-time employment (40 hours per week). This requirement also proved to be a source of problems (cf. Qvist and Tovatt 2014, 59f; Larsson 2015, 182), as there were not always suitable activities for all participants:

One of the most frustrating things is that there haven’t been enough activities for the group … no high-quality activities. (Focus group 1)

Several of the interviewees experienced that the programme ‘runs slow’ due to queues and waiting times to start activities, as well as a general lack of activities to assign to the participants.

The political and organizational leadership tried to enhance the introduction programme, for instance by increasing user participation (see Eriksson 2018), the implementation of new digital tools (SPES 2018), and allocating new personnel to the programme. There were also initiatives to streamline the structure and working tasks of the programme.¹⁰ Replying to our question of what has improved, caseloads were the primary topic raised by the interviewees:

It is not as stressful now, we have fewer cases again and more employees. But for the participants, the programme hasn’t got any better. (Focus group 1)

Even though some of the interviewees felt that their caseloads had eased somewhat, they still expressed doubt concerning their possibility to properly support the participants. Some were concerned that the increased streamlining would prevent necessary individual considerations (cf. SAPM p. 75), while others stated that reduced caseloads do not aid the participants if the programme is still not able to provide them with appropriate activities.¹¹

The needs of the participants

Even though the topic of the interviews was the employment officers’ perception of working in the programme, the interviewees repeatedly directed the discussion towards the situation of the participants. They were clearly concerned about their clientele, and stated that many of their clients had basic needs that they as employment officers felt unable – and through their assignment not expected – to meet. Housing recurred as one such issue:

Some people buy an address on paper/ … /but actually they roam about a little bit anywhere, with many people living at the same address. So, there are many challenges related to housing, and their life puzzle in general. It is a huge obstacle … all the things they need information about or help with. (Focus group 2)

As indicated in the last sentence, the participants’ need for practical guidance repeatedly appeared in the interviews as another need that cannot be met. Another recurring need was health, where the interviewees stressed that many participants suffer physical and mental ill-health that prevents them from making use of the services of the introduction programme. The interviewees sometimes showed considerable frustration when discussing the needs of the participants, and they seemed to experience a distinct conflict between organizational expectations and clients’ needs:

We just don’t have enough time. There are all these administrative duties you are expected to do, while you rather would like to do something concrete for the participants. (Focus group 2)

There are frameworks and rules to follow, and these [organizational] rules conflict with the rules of your heart, and you find yourself in a trap. (Focus group 1)

Conclusively, the street-level dilemma predicted by Lipsky is pronounced in our study, and in the two following paragraphs we investigate how the employment officers handled this situation.

A client-centred pattern: ‘doing more’, bending rules and reformulating goals

Throughout the interviews, several interconnected coping strategies were identified, that could be sorted into two broad patterns of practice: a client-centred pattern and an authority-centred pattern. Signs of both these patterns appeared in all three focus groups, and the patterns should be understood as analytical constructions. Even if they represent two distinctive positionings, they are not mutually exclusive, and in practice they presumably coexist, coincide and intersect (cf. Maynard-Moody and Musheno 2000).

The first pattern of practice – that was somewhat more pronounced in the empirical material – is characterized by employment officers acknowledging the needs of the participants, trying to take a holistic approach and to see ‘the whole picture’ of the participants’ lives. Within this pattern, several interviewees refer to ‘compassion’ as a foundation for their actions, and the moral assumption is that if you as a professional meet a person in need, you are obliged to help them as best you can. This approach typically includes coping strategies ‘moving towards clients’ (Tummers et al. 2015, 1103), where coping is exercised to benefit clients. It resembles the position of the ‘citizen agent’ (Maynard-Moody and Musheno 2000), and what Larsson (2015) identified as a ‘social worker’ position within the introduction programme, where dedication to help is a central ambitions.

One coping strategy adopted here is characterized by ‘doing more’ than what the organizational guidelines specify, such as accompanying participants to the primary care unit, offering guidance concerning housing or other societal issues, or handling heavy caseloads by working (unregistered) overtime (cf. Dubois 2010). Through such a strategy, the street-level bureaucrats try to compensate for underfunding by using their private time to work, thereby trying to avoid negative outcomes for the clients (cf. Lipsky 2010, 173). One of the interviewees described it as follows:

We all stress out, trying to make it work. This is not like any other job; I need to do this as a fellow human. I can´t just say that my working hours are full. (Focus group 3)

As indicated, this strategy might lead to consequences such as work-related stress and fatigue. However, ‘doing more’ based on a desire to ease the situation for the participants does not always imply that the participants get extensive supplementary support:

We really don’t have that much to offer them/ … /but at least I try to create a good mood, make it pleasant for them [at the meeting]. Sometimes I offer coffee and cinnamon buns. (Focus group 3)

This interviewee recognizes the needs of the participants, but has seemingly resigned to the fact that she, or rather, SPES, cannot assist them sufficiently. Maynard-Moody and Musheno (2000, 352) stress the relationship with clients as central to the citizen agent positioning, and, consequently, when the interviewee feels she has nothing else to offer, she puts great value in maintaining a good relationship.

The perception of not having enough quality activities to offer might also result in strategies where street-level bureaucrats use discretion to bend the rules in favour of the participants (cf. Evans 2013). An example, emanating from the policy demand that every participant must have an introduction plan filled with 40 hours of activities per week, is that employment officers fill up this schedule with ‘empty activities’ (Larsson 2015, 207). Such activities are not what the policy directives intend, yet some employment officers interpret the rules creatively when they assess that no suitable activities are available:

There has been a possibility for us to say that, well, these kinds of activities work: “going to the library”, or “taking a walk”./ … /Not all officers have used this opportunity though, some have been stricter, saying ‘no, the rules say 8 hours of activities’, while others have been more like: ‘we’ll make it work for you’. (Focus group 1)

Here, the discretion of the street-level bureaucrats becomes evident, and the divide between the client-centred and the authority-centred patterns of practice is made explicit – where some employment officers are more prone to rule-following than others (cf. Oberfield 2010). Focusing on the client-centred pattern for now; completing the introduction plan with ‘empty’ activities is another way of trying to compensate for perceived deficiencies in the programme. Indeed, the strategy does not really support the participants entering the labour market, since they do not engage in any ‘real’ activities. Still, the strategy can be interpreted as client-centred as the employment officers neglect to force participants to attend activities assessed as unsuitable, ensuring them the full introduction benefit despite not engaging in full-time activities.

A final strategy adopted in this pattern of practice involves formulating personal goals that depart from the official goals of the agency (Lipsky 2010, 144). The overt goal of the introduction programme is that every participant should be self-sufficient after two years. However, for some participants this is perceived as unrealistic:

You can only get so far during the introduction programme, depending on where you are when you enter. We must have this attitude, that we try to establish them as far as possible. If you go from being illiterate to learning Swedish, that is a huge step, just achieving that. (Focus group 2)

For some participants, a more distant labour-market integration is predicted, exceeding the two-year time frame. Learning Swedish, for instance, is formulated as a more realistic and appropriate goal. Such reformulation of goals enables street-level bureaucrats to feel that they are doing a satisfactory job, while not reaching formal policy expectations.

In sum, in the client-centred pattern of practice, the coping strategies aim at assisting the participants take their individual circumstances as a starting point. This involves ‘doing more’ than what the organizational mandate dictates (cf. Dubois 2010; Borelli and Lindberg 2018), using discretion to bend the rules (cf. Evans 2013), as well as reformulating official goals (Lipsky 2010, 144). In the interviews, these actions are commonly dressed in a humanitarian terminology. However, as shown in the extracts – and in consent with the citizen agent narrative (Maynard-Moody and Musheno 2000, 353) – practice within this pattern can also be driven by pragmatic reasoning and discretional judgements of what is proper treatment or realistic to achieve under given organizational restrains. While these coping strategies all involve departing from the scripted procedures within SPES, the practices might still align with the overarching (political) goal of the introduction programme: to help the participants become self-sufficient. Thus, since organizational policy is typically inconsistent and contradictory (Lipsky 2010, 41, 164; Brodkin 2011, 256), street-level bureaucrats might simultaneously act in agreement with some policy directives, but in conflict with others. In this case, the employment officers typically do their best to pursue the central political goal, while transgressing the (politically and organizationally formulated) principles on how to act to reach this goal. Lipsky (2010, 83, 147) stresses that when street-level bureaucrats are forced to adopt coping strategies, this typically delivers negative outcomes for clients. Yet, at least in some sense, the strategies sketched here work in favour of the participants (cf. Maynard-Moody and Musheno 2000; Tummers et al. 2015). However, since all employment officers do not – and cannot, certainly not with all their clients – use these coping strategies, the introduction programme risks becoming inherently unequal (cf. Larsson 2015, 220). Based on their moral – or perhaps professional – judgement of deservingness, even those employment officers that adopt the client-centred pattern will inevitably have to decide in which situations and with which participants they ‘go the extra mile’ (cf. Maynard-Moody and Musheno 2000, 351).

An authority-centred pattern: delimitation and formality

The main tasks of the employment officers in the introduction programme are to chart new participants’ needs and elaborate an introduction plan, to assign activities to participants, and to check the progress of the plan. Apart from this, working with participants is formulated as giving information and guidance to facilitate labour-market matching. These key duties are stipulated in political documents (e.g. Government Missive 2008/09:24, 43; Proposition 2009/10:60) and policy directives (see Ministry of Labour 2018, 2), as well as by legal legislation (Regulation 2017, 820) and intra-organizational policy (e.g. SPES 2016, 2018). As elaborated above, some employment officers transgress these instructions and take different measures to assist the participants. The employment officers adopting the second pattern of practice, however, stay closer to their formal assignment and strive to meet intra-organizational expectations. This authority-centred pattern comprises a more restrained logic, often resting on a literal interpretation of rules (cf. Oberfield 2010). Thus, it resembles and confirms the strategies that Larsson (2015, 203) identified as ‘boundary setting-strategies’ and has similarities to the ‘state agent’ narrative proposed by Maynard-Moody and Musheno (2000). However, in the state agent narrative, it is argued, discretion is mainly used to serve the self-interest of the street-level workers, to make work ‘easier, safer and more rewarding’ (Maynard-Moody and Musheno 2000, 339). This might be true for our interviewees as well, but when expressed by themselves they rather seem to perceive their actions as a way of preserving the bureaucratic order and serving SPES’s interests by maintaining organizational stability.

A common coping strategy within this pattern involves trying to limit caseload by referring participants to other organizations. By giving one’s assignment a strict interpretation – in itself, an act of discretion – the employment officers withdraw from responsibility, referring ‘the problem’ to someone else. This is a typical example:

Generally, you could say this: if a person isn’t at the disposal of the labour market, they are not actually our target group. SPES shouldn’t work with them. It’s others – the municipality, healthcare, or whatever – that must handle it, until they are able to enter the labour market./ … /There is not much to say, I think, because that’s our assignment. That’s what we are supposed to stick to. (Focus group 2)

Sometimes the perception that a participant needs the aid of another organization might indeed be an appropriate assessment. Nonetheless, SPES is responsible for coordinating the introduction of every individual migrant. Thus, it could be argued that SPES ought to make sure that proper support is available, and help the participants access this.

Even when participants are acknowledged as the responsibility of SPES, employment officers can use coping strategies that deny discretion (cf. Lipsky 2010, 149; Larsson 2015, 198) by clearly defining their (limited) obligations. Statements like ‘that is not our responsibility’, ‘there’s nothing I can do’ and ‘we have rules to follow’ reoccur in the interviews. Rather than moving towards clients and their perspective, the coping strategies of this pattern tend to move the street-level bureaucrats away from clients (Tummers et al. 2015, 1103), avoiding meaningful interactions. An example:

A participant came into the room crying, and my colleague said to her: ‘What do you want me to do? We can’t do everything, we can’t be social assistants, we can’t be police officers or your trustee. We can’t be everything, you must understand that … But they come to us with everything. No matter how many times you tell them. (Focus group 1)

In general, the informants express understanding of the participants and their predicaments. Still, signs of emotional and psychological detachment occur, which can be understood as a coping strategy used to ease feelings of inadequacy (cf. Lipsky 2010, 142). Even though the interviewees commonly constructed the participants as capable and hardworking, such distancing was occasionally done by constructing individuals as difficult or dubious (cf. Lipsky 2010, 152). Usually this took subtle expressions, but was sometimes articulated more explicitly:

Some [participants] say they can’t work, with the excuse that they have three or four impairments – but you don’t really know if that’s true. It’s just: ‘I can’t do this work because this hurts.’/ … /‘Well okay, can you work with numbers then?’. ’No, I don’t know numbers, my eyes hurt, I can’t see … ’. They just want the benefits. (Focus group 1)

These kinds of constructions of clients relate to the broader issue of determining deservingness that permeates most welfare practices (see Handler and Hollingsworth 1971; Schneider and Ingram 1993; Van Oorschot et al. 2017), where target groups and clients are constantly constructed as deserving or undeserving of access to welfare. In this, migrants as a group – and especially asylum seekers and unemployed migrants – are particularly at risk of being represented as undeserving (Sales 2002; Ataç 2019). At an individual level, vulnerability and need as well as ‘good behaviour’ are constructions usually used to motivate deservingness (Jilke and Tummers 2018), and the needs of the participants are generally not questioned in our study. However, constructing clients as behaving badly – as in the extract above – might make it easier to interpret them as undeserving (Lipsky 2010, 152), justifying differentiation between clients. This practice also relates to the following of rules, as such constructions imply that participants do not adhere to their introduction plans and try to avoid the rule of being at the disposal of the labour market.

A final coping strategy in this pattern of practice – touched upon in the previous paragraph – occurs when participants are appointed to available activities, whether or not these activities are appropriate. Here, the formal rule of 40 hours of activities per week becomes the imperative concern, trumping the fact that the plan is supposed to be individualized. Larsson (2015, 182) observed similar tendencies, and the logic can be identified as ‘any activity over appropriate activity’. This strategy is a form of routinization (Lipsky 2010, 83, 99) that can be understood in the light of the workfare logic and activation policy spreading through welfare politics (see Greer 2016), where the individual must do something – anything – in return, to be deemed as deserving of welfare benefits. As illustrated by Floros and Bak Jørgensen (2020, 8), activation policies can be equally about controlling as about supporting migrants, perhaps explaining why some employment officers in our study tend towards more authority-centred coping strategies.

In sum, the authority-centred pattern of practice involves coping strategies where the employment officers rely on a rigid interpretation of rules and organizational decrees to create a manageable work situation. Doing this seemingly also creates a need to produce an emotional and psychological distance to the participants (Lipsky 2010, 142), as well as denying discretion (Lipsky 2010, 149; Larsson 2015, 198pp). Within this pattern of practice – with its emphasis on the importance of formality and rules, and following statements like ‘first and foremost, we are employment officers’ and ‘we have to stick to the SPES assignment’ – the interviewees approach the state agent narrative elaborated by Maynard-Moody and Musheno (2000). Also, denying discretion, which is one of the coping strategies used within the authority centred pattern of practice, resonates with the ideal of the state agent. Nonetheless, our authority-centred pattern of practice does not equate the state agent narrative.¹² Firstly, the moral judgement and reasoning that Maynard-Moody and Musheno (2000, 351) associate with the citizen agent narrative is present in our authority-centred pattern as well, which highlights that street-level bureaucrats who strive to adhere to rules also actively exercise moral judgement. Even if strictly implementing policy is in accordance with regulation, the politically determined regulations are also based on moral judgements of deservingness (cf. Schneider and Ingram 1993), making street-level bureaucrats conforming to rules a moral practice as well. Secondly, as Maynard-Moody and Musheno explain, the state agent is an ideal typical elite narrative of the civil servant, separated from street-level workers’ own perception of their work. In contrast, the authority-centred pattern of practice outlined here is expressed by the employment officers themselves, still placing emphasis on some of the characteristic ideals of the state agent narrative. This could indicate that, as working conditions worsen and elements of control and workfare logics increase in policy, (some) street-level bureaucrats become alienated from their clients (cf. Lipsky 2010, 79), positioning themselves closer to the authority and perceiving themselves increasingly as state agents rather than citizen agents.

Discussion

Summary

The analysis has shown that the interviewees perceive working conditions within the introduction programme as difficult, characterized by a pronounced tension between organizational demands and participants’ needs. Heavy caseloads, inadequate available activities and extensive administrative duties have been identified as working conditions that obstruct proper treatment. Moreover, several of the participants’ identified needs (such as insufficient housing, a need for practical guidance, and health related issues) are perceived as impossible to meet within the boundaries of SPES assignment. The employment officers handle this dilemma by applying coping strategies within two patterns of practice. Within the client-centred pattern, employment officers attempt to use discretion to assist the participants in accordance with their needs. Within this pattern we identified three coping strategies: (1) doing more to assist the participants than the formal assignment requires; (2) bending rules to ease the predicaments of the participants; and (3) reformulating official goals to fit what participants can realistically achieve. Within the authority-centred pattern, employment officers applied a formal and rule-oriented understanding of their assignment, concentrating their efforts on maintaining the functionality of the introduction programme. Within this pattern we identified four coping strategies: (1) delimiting caseload by referring participants to other organizations and denying discretion; (2) emotional and psychological detachment from participants; (3) differentiating between participants through constructions of deservingness; and (4) routinization of work by assigning all participants to the same activities. Our results confirm that coping strategies not only serve the self-interest of street-level bureaucrats, but can be applied more regularly as attempts to assist clients (cf. Maynard-Moody and Musheno 2000; Tummers et al. 2015, 1103).

Our study confirms the positionings of the employment officers within the introduction programme that Larsson (2015) revealed, and our client-centred and authority-centred patterns of practice are reflected in the citizen agent – state agent narratives developed by Maynard-Moody and Musheno (2000). However, the two patterns of practice we propose do not correspond directly to Maynard-Moody and Musheno’s two narratives.¹³ Within our authority-centred pattern, at least to some extent, the interviewees actually appear to identify with the state agent narrative, which differs from the theorization of Maynard-Moody and Musheno – where street-level workers are understood solely as bearers of the citizen agent narrative. Moreover, within the authority-centred pattern, using discretion is typically represented as acting in the best interest of the authority and administration, rather than based on self-interest, as in the case of the ideal-typical state agents (Maynard-Moody and Musheno 2000, 339). This difference might be a result of our interviewees not wanting to acknowledge self-interest as motivator, but it can also indicate a genuine identification with a role similar to the state agent, and a sincere belief that they are working righteously to maintain the system of the introduction programme. Finally, our results also show that decisions based on moral judgement are not exclusive to the client-centred pattern, but indeed occur within the authority-centred pattern as well. In relation to Maynard-Moody & Musheno’s terminology, where moral judgements are referred to the role of the citizen agent (cf. Maynard-Moody and Musheno 2000, 351), our results highlight that street-level bureaucrats who take a position close to the rule-obedient state agent exercise moral judgements as well. Wherever street-level bureaucrats have the opportunity to exercise discretion, the decision to make a strict interpretation of formal rules and guidelines is in itself an individual moral judgement, that is sometimes explicitly justified in terms of morality. In our study, client-centred and authority-centred street-level bureaucrats alike appear as pragmatists, however the former use pragmatism to orient towards participants’ needs, while the latter use pragmatism to sustain the operation of the organization. Our interviewees more often seemed to embody a role resembling the citizen agent narrative rather than the state agent narrative. Nonetheless, the study shows that street-level bureaucrats indeed can, and at least to some extent do, orient towards the state agent narrative as well.

Conclusion

If, as in the case of the introduction programme, working conditions are difficult, street-level bureaucrats must try to handle work as best they can. Regardless of pattern of practice, adopting coping strategies will typically make work manageable and keep the agency’s operations running (cf. Holm Vohnsen 2017), but will inevitably also affect quality of work. There is an impending risk that neither political nor organizational, not to mention clients’, goals will be met. When street-level bureaucrats are forced to use discretion mainly to compensate for inadequate resources or working conditions, rather than to make professional judgements that enable individualized treatment (see Handler 1986; Wallander and Molander 2014), the risk of unequal and poor-quality practice increases (Lipsky 2010; Astvik and Melin 2013; Brodkin 2012). Further, Schierenbeck’s (2003) study reminds us that – at least when policy is adequately formulated and funded to handle a certain problem – a practice predominated by discretional individual judgement (rather than following of formal rules) might also create difficulties for clients and deter desirable policy outcome. On this note, perhaps the most important implication of this study is that the interviews reveal a mismatch between the politically formulated integration policy and the actual needs of the participants in the programme (cf. Eriksson 2018, 11, 2019), suggesting that the introduction programme might constitute an inadequate support structure to achieve labour-market integration (cf. Larsson 2015; Abdulla 2017). Following the political assignment, the employment officers are instructed and expected to provide support and guidance directly connected to labour-market matching (cf. SPES 2016). While such labour-market oriented support is vital, the study shows that participants have other needs as well. This corresponds to the results of studies focusing on migrants’ perspectives, which similarly have shown that migrants’ perceived needs differ from the support offered by the public (see e.g. Tovatt 2013, 27f; Ikonen 2015; Brännström et al. 2018; Eriksson 2018, 2019). Currently, the participants of the introduction programme lack adequate support to handle e.g. practical, health-related and psychosocial issues related to secure living conditions, that would indirectly facilitate labour-market participation (cf. Lindencrona 2008, 86ff). Some employment officers develop coping strategies to make up for this deficiency and try to support clients in a more holistic way, thereby actively affecting policy as it is implemented. However, the basic mismatch between political policy and needs will persist if political integration policy – and instructions to SPES – does not change. Even if caseloads were ideal and current activities in the programme were of satisfactory quality, the street-level bureaucrats would still have difficulty in adequately supporting participants, simply because the integration programme does not include activities that target the full range of participants’ needs. Thus, this study encourages politicians and policymakers to adhere to the challenges perceived by the employment officers, and to formulate a more holistic integration policy (cf. Eriksson 2019, 25). Such policy should also consider how the currently applied temporary residence permit affects the migrants and their possibilities to integrate and establish themselves in the labour market – which also is an urgent subject for future research on Swedish integration policy and practice.

Acknowledgments

During the writing of this manuscript, my valued colleague and co-author, Associate Professor Kerstin Johansson, unexpectedly passed away. Kerstin: this article would not have been possible without you, and I wish that you could have been here to see its completion. You will be remembered and greatly missed.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This study was funded by Linköping University and co-funded by the Swedish Public Employment Service.

Notes

1. Hungary received more asylum-seekers per capita than Sweden; but as the Hungarian asylum process has been deemed as violating human rights, most of them were allowed into Germany and Austria instead (Zaun 2017).

2. In line with the theoretical framework, in this article the term ‘implement’ is used quite openly, as a synonym to ‘execute’ or ‘carry out’, rather than as a main theoretical concept or object of study.

3. See e.g. Tummers et al. (2015) for a comprehensive overview.

4. A treaty between the Social Democrat/Green minority government and the liberal opposition parties in January 2018 states that SPES will be fundamentally reformed. During 2019–2020, half of the 200 local SPES offices will be closed. By 2021 all matching and supporting activities should have been transferred to private entrepreneurs (see Government decision A2019/00923/A). Thus, SPES as it was constituted when this study was conducted, will no longer exist.

5. Integration and establishment are equally inherently complex and contested terms. See e.g. Gustafson (2004, 36) and Eriksson (2019, 13–21) for elaborated discussions on these concepts in the Swedish context.

6. There is a broad variety of activities aimed at preparing the participants for labour-market participation, such as trainee positions, labour-market training and coaching. See SPES (2016) for an overview of current activities.

7. The figures include subsidized jobs.

8. Since Schierenbeck’s (2003) study, the Swedish integration policy has been reformed and practice has transformed. However, this study highlights the important notion that – depending on context and how policy is formulated – rule-following might sometimes be more adequate than individual discretionary judgement when it comes to reaching satisfactory outcomes for clients.

9. Scholars have predicted that the legislation that normalizes temporary residence permits will hamper the possibilities of integration and labour-market participation (e.g. Engdahl 2016). This concern is consistent with international research (e.g. Dustmann et al. 2017), and SPES itself foresees the temporary permits as a complication of the introduction programme, as they aggravate long-term planning and risk labour-market mismatches (SPEC 2017a). However, as almost all participants of the programme held permanent residence permits at the time of our study, the analysis cannot address the effects of the temporary permits.

10. E.g. through standardized guidelines on how to chart participants and set up their introduction plan, as well as by increased performance management (see SPES 2018, 22; SAPM 2019).

11. This refers to e.g. a perceived lack of qualified trainee placements and high-quality labour-market education.

12. There is also a conceptual distinction between the ‘patterns of practice’ we suggest, and the ‘narratives’ suggested by Maynard-Moody and Musheno. Theoretically, the patterns of practice more narrowly represent the street-level bureaucrats handling of work, while the narratives represent a broader recount of the role of the street-level worker. In this, a certain pattern of practice might correspond to a certain perception (narrative) of the role as a street-level worker.

13. This might, for instance, depend on differences in the welfare settings of the two studies (cf. Hupe and Buffat 2014). Indeed, if both studies were conducted in an American setting (as Maynard-Moody & Musheno’s), the linkages between the patterns of practice and the narratives might have been even more pronounced.