‘All within the margin of the rules’ : how activation workers in Zwolle construct and use discretion under the Participation Act

(1)

‘All within the margin of the rules’

How activation workers in Zwolle construct and use discretion

under the Participation Act

Building of the activation department in Zwolle

Wart Mangnus - 10759956 University of Amsterdam

Master Thesis Political Science: Public Policy & Governance Supervisor: Dr. Imrat Verhoeven

Second reader: Dr. Benno Netelenbos August 2019

(2)

2

Abstract

The Participation Act from 2015 is the latest major reform in welfare and activation policy. It emphasises the need for ‘tailor-made’ policy adjusted to local context and individual needs to get welfare recipients in general, and vulnerable groups in particular, back to work. Other facets of the Participation Act focus on seemingly strict responsibilities and duties for the client and sanctions if those duties are not fulfilled. Altogether, this discourse bestows responsibility on local policymakers and street-level activation workers to bring these ‘motherhood statements’ into practice in a local context. The objective of this thesis is to reconstruct what the reform of the Participation Act means for activation work in practice: what constitutes discretion-as-granted and discretion-as-used from the perspective of street-level activation workers, where are the discrepancies between the two concepts. In the theoretical framework various theoretical perspectives are discussed on street-level work and research in general as well as in activation policy and personalisation in particular. Ten activation workers in Zwolle were interviewed using conversational interviews as part of a qualitative case study. Zwolle proved to be a deviant case where discretion-as-granted only had limited influence on the daily practice of activation workers. Instead, due to limited local policy interpretation and direct management, activation workers amongst themselves have constructed their own interpretation of their role and responsibilities under the Participation Act. I offer the concept of ‘discretion-as-constructed’ to describe this process and use stories from the tradition of interpretative analysis to reconstruct the discretion-as-constructed in Zwolle. Activation workers adapt their practices to the needs and wishes of clients in their discretion-as-used, based on their interpretation of their discretion-as-constructed.

(3)

3

Introduction

In 2015 the Participation Act came into effect, constituting the latest major change in Dutch welfare policy. The act serves as an amalgamation and adaptation of previous policy on unemployment, employment activation and welfare. At the same time, the act transferred the majority of the responsibilities regarding welfare from national government to municipal government (State Secretary for Social Affairs and Employment, 2014; Van Berkel, Caswell, Kupka & Larsen, 2017; Vermeulen, 2015).

One of the main goals of the reform was to reorganise the system in such a way that the unemployed, in particular those with disabilities, could receive tailor-made support in

working towards private employment (Ministry of Social Affairs and Employment, 2019; SER, 2016; State Secretary for Social Affairs and Employment, 2014; Vermeulen, 2015). Nearly all the unemployed with and without disabilities or impairments, ineligible for social insurance, who still have any remaining capacity to work, now belong to the ‘Municipal Target Group’ (gemeentelijke doelgroep), or ‘Target Group’ for short. This places them under the

‘municipal activation responsibility’ (gemeentelijke re-integratieplicht), which means the municipality has a duty to provide them support in achieving optimal activation and societal participation (Bruggeman & Van Rooij, 2018; Sadiraj, Hoff & Versantvoort, 2018; State Secretary for Social Affairs and Employment, 2014). The idea behind this was, again, that the closer care is to the recipient, the better. Municipalities would know best what their regional labour market needs, and frontline workers would work with vulnerable clients to provide assistance to their specific needs (Solvang, 2017; State Secretary for Social Affairs and Employment, 2014).

Other criticism of welfare arrangements pre-Participation Act concentrated on the case of those who require further assistance to make the transition to (paid) employment, but were not deemed eligible for sheltered employment guaranteed by a Wsw-indication (Wet sociale werkvoorziening: Sheltered Employment Act)(Knijn & Van Wel, 2014; Sadiraj et al., 2018; State Secretary for Social Affairs and Employment, 2014). One comprehensive system of welfare, along with decentralising responsibilities to local government is meant to combine austerity with more effective policy of employment activation (Vermeulen, 2015).

(5)

5

The Participation Act offers a mix of existing and new instruments from which to pick and choose to create a policy mix that local authorities deem most suitable for their regional context. The Participation Act also grants local government relative freedom in the selection of ‘instruments’ they wish to employ. Again the underlying assumption was that local

governance, as close to the citizen as possible, would yield the most efficient approach (Bruggeman & Van Rooij, 2018; State Secretary for Social Affairs and Employment, 2014; Vermeulen, 2015). These instruments cover varying facets of the activation process and varying levels of governance in the municipal organisations: from financial stimuli to encourage private actors to hire unemployed welfare recipients with a disability, to

jobcoaching and mediation. Previous to the Participation Act, starting in 2012, municipalities already had the power to demand compensatory employment from ‘regular’ welfare

recipients without disabilities or impairments (Ministry of Social Affairs and Employment, 2013). The Participation Act further stimulated reconsideration and stricter sanctioning of these demands that have to be met by welfare recipients, in order to ensure eligibility (Bruggeman & Van Rooij, 2018; CentERdata, 2017; Raffass, 2017).

By now it should be clear that the Participation Act was meant to realise reforms that change the practical reality of municipalities, activation professionals and welfare recipients alike. It is also the latest push for activation in welfare services, this time mainly focused on welfare recipients with a labour impairment (Sadiraj et al., 2018; SCP, 2016; SER, 2016; State

Secretary for Social Affairs and Employment, 2014).

In September 2018, the first partial evaluation of the Participation Act was published by the Netherlands Institute for Social Research (Sadiraj et al., 2018). The Participation Act focuses particularly on activating vulnerable welfare clients, who were not eligible for a

Wsw-indication, guaranteeing the right to sheltered employment, yet did need extra assistance in finding employment (State Secretary for Social Affairs and Employment, 2014). The report in question gave us a first insight into the effects of the Participation Act. The main takeaway from this report is that the odds of employment for this ‘privileged’ group of disadvantaged job-seekers have decreased after introduction of the Participation Act. Other reports state that the new form of sheltered employment does not nearly generate enough employment to replace the Wsw (Sadiraj et al., 2018; SCP, 2016; Versantvoort & Van Echtelt, 2016). It is impossible and undesirable to draw conclusions about the success of the Participation Act

(6)

6

from these early reports, beyond what was stated above. However, it does emphasise the need to look closely at the workings of the Participation Act in practice, to properly evaluate its successes and challenges, in particular for vulnerable clients. This need seems even

greater after cautions from governmental advisory organisations against the loss of expertise and infrastructure when activating vulnerable clients, as responsibilities shift to different actors in other areas of governance (SER, 2016). Municipalities also gave mixed reports on how well they were able to implement the new instruments (CentERdata, 2017) and signal a large share of untapped employment potential in vulnerable clients (Werk.nl, 2019).

The goal of this thesis is providing insight in the perception of frontline or street-level workers on how the transition to the Participation Act affects their discretion in activating vulnerable clients. The Participation Act potentially changes many contextual factors in the work of street-level activation workers (Bruggeman & Van Rooij, 2018; SER, 2016;

Vermeulen, 2015). Firstly, regulatory frameworks change with the amalgamation of

unemployment assistance laws. Secondly, the Participation Act changes the organisational structures and responsibilities within street-level organisations. Finally, the Participation Act implies a change in the nature of client-work: more diverse tools at the disposal of the street-level bureaucrat, different clients and changing goals of activation (from public to private employment) (Sadiraj et al., 2018; SCP, 2016; State Secretary for Social Affairs and Employment, 2014). Street-level research makes a case for researching the reality of street-level bureaucrats to uncover how contextual factors influence the way these public officials function as co-policymakers, through the use, or refraining from use of discretionary power as-granted and as-used (Hupe, Hill & Buffat, 2016; Van Berkel, Caswell, Kupka & Larsen, 2017). ‘Vulnerable groups’ lack a formal definition in the Participation Act. In evaluating the Participation Act, the Dutch Social and Economic Council (SER, 2016) has chosen to

‘operationalise’ this term as referring to all who require additional assistance or services to obtain or maintain employment.

(7)

7

The goal described in the previous paragraph translates into the following research question for this thesis:

How has the participation act affected the work of street-level activation workers, and by extension their discretion, in providing personalised services in activating vulnerable unemployed?

To answer this question I have formulated three subquestions:

- What constitutes discretion-as-granted under the Participation Act? - What constitutes discretion-as-used under the Participation Act?

- How do street-level workers use discretion to offer a personalised approach to vulnerable groups?

Discretion-as-granted is the room for decision-making by street-level actors as intentionally created by a formal context, such as rules and control mechanisms (Hupe, 2013). Discretion-as-used is the room for decision-making street-level actors actually use in practice.

This thesis consists of three main chapters. The first chapter sets out the theoretical framework for this thesis. In this chapter the various academic perspectives on street-level research, activation policy and personalisation will be discussed as well as their relevance to researching the Participation Act. The second chapter will concern the research design of this thesis as well as the process of case selection and the impact of the deviant case of Zwolle for the remainder of this thesis. In the third chapter, the empirical research of this thesis will be analysed. As we will find out, the two concepts of discretion-as-granted and discretion-as-used describe only part of the findings in the case study, and I introduce an additional concept (discretion-as-constructed) through the use of stories. After this final chapter, the implications for the empirical findings in Zwolle for the research question will be discussed, as well as how they relate to other theoretical perspectives.

(8)

8

1. Theoretical Framework

In this chapter on the theoretical framework for this thesis I will discuss the fields of street-level research and frontline work in general and the concept of discretion specifically. After that, the fields of activation policy and personalisation will be discussed. Some of these academic fields are more developed than others, yet all of them consist of a variation of perspectives. In this chapter I will specify which perspectives in those field I consider to be relevant to this thesis, and how these theoretical perspectives relate to the Participation Act.

1.1 Street-level Bureaucracy & Frontline work

Lipsky coined the term ‘street-level bureaucracies’ nearly half a century ago. His work is part of the broader corpus of the second generation of implementation studies (Pülzl & Treib, 2017, p. 89). These authors started building theoretical frameworks and hypotheses, based on the observations of their predecessors that policy is not always carried out as intended by officials. Underlying both these approaches is the assumption that solely studying policy ‘as designed’ leaves significant blind spots for academics and policymakers: ‘[…]Knowing policy inputs (…) does not allow ‘reading off’ policy outputs (…), let alone outcomes’ (Hupe et al., 2016, p. 9). Within implementation studies a distinction can be made between top-down and bottom-up theories. Lipsky’s work and consequent street-level bureaucracy research from other authors is a bottom-up implementation theory (Pülzl & Treib, 2017, p. 90). These theories view street-level actors as central in policy delivery. From that perspective

policymaking is a ‘[…] negotiation process within networks of implementers’ (Ibid., 2017, p. 90). Top-down theories on the other hand focus on central policymaker’s ability to control the hierarchical implementation process (Hupe et al., 2016, p. 9). Lipsky showed the importance of studying the organisations at the frontline of the state, where they interact with the citizens subject to policy, as well as the public officials that embody this contact. In particular the street-level bureaucrat is a concept that continues to find its use in

contemporary research, discussion and criticisms aside (Durose, 2011; Hupe et al., 2016, pp. 6–8). Lipsky justified his research as ‘a search for the place of the individuals in those

services called street-level bureaucracies’ (Lipsky, 2010, p. xi). This attention to the role of the individual public official within institutions that bring policy into practice, is why the framework of street-level bureaucracies suits this research so well. Lipsky’s street-level focus

(9)

9

stood in stark contrast to the focus of other public policy research traditions, such as public administration, which studied managers and top-level officials instead (Maynard‐Moody & Musheno, 2012, p. 16).

While government evaluations of the Participation Act have not specifically targeted

managerial behaviour, focussing on system-level issues drawn from data and observations at a higher level of abstraction seems to be the popular approach. It has become a trend in governance discourse to refer to ‘tailor-made’ solutions and designing policy close to, or with involvement of those affected (Rice, Fuertes & Monticelli, 2018). The assumption is that each situation requires different accentuation in policy, specific to the context. Therefore, it has become increasingly important to research the behaviour of actors at the ground-level. Lipsky envisioned a key role for street-level bureaucrats in fitting policy to practice, by using their own discretion. Lipsky realised that these deformations of policy, while undesirable for those designing policy, were unavoidable and thus needed to be studied. In the past decade, Dutch policy discourse has seemingly embraced this autonomy at the lower levels of

government, which puts increasing pressures on the shoulders of those in charge of carrying out the frontline work (Jessen and Tufte, 2014; Raven, Achterberg & Van der Veen, 2015; Van Berke, Van der Aa & Van Gestel, 2010). In turn, it is increasingly important to keep studying the behaviour of these workers, and the context they work in, as Lipsky’s theories prescribe.

One could make a valid argument that top-down implementation studies also offer a framework for studying policy beyond the policy programme. Top-down implementation studies, according to Hupe, Hill and Buffat, study ‘what happens after a bill becomes a law’ (2015, p.11) and maintain a vertical perspective on policy implementation (down through the levels of governance). Street-level bureaucracy, however, invites research from a

horizontal perspective, which allows the conceptualisation of the interactions and structures within a street-level bureaucracy and between street-level bureaucrats and citizens. Policy reforms in general demand study after their implementation, the Participation Act is no exception. This also goes for the level dimension of the policy, and the role of street-level bureaucrats. In the case of the Participation Act, from research from the street-street-level framework, the relative lack of attention to activation work (Hupe et al., 2016; Kampen & Tonkens, 2018; Van Berkel et al., 2010) and personalisation (for exceptions, see; Kampen &

(10)

10

Tonkens, 2018; Solvang, 2017) on a street-level, are some of the motivations for choosing this ‘bottom-up’ perspective, as it offers a framework better suited to study the dilemmas that street-level bureaucrats themselves face. As Durose stated: ‘Top-down accounts

underestimated the importance of negotiation with front-line staff, service users and others in the establishment of the meaning of policy and its implementation’ (2011, p. 979).

Street-level bureaucrats or frontline workers

Recent decades have seen discussion whether Lipsky’s characterisation of street-level bureaucrats (and in turn street-level bureaucracies) is still relevant for those who work at the frontline of government agencies today. One pertinent critique comes from Catherine Durose (2011). She argues that frontline work in the UK is increasingly characterised by the task to unite new, emerging demands from local communities with old systems of

governance. In this context, frontline workers are required to display a sense of civic entrepreneurship and look for opportunities ‘at the periphery’ of governance. Contrary to Lipsky’s observations, Durose paints a picture where frontline workers are required to think outside of the box, in order to develop new strategies to meet the changing demands. In Lipsky’s portrayal of frontline work, the context is more set and defined, and discretion and autonomy exerted by street-level bureaucrats follows more strict procedures. Durose argues that it is exactly at the ‘periphery’ of traditional government where frontline workers need to display these skills. In this research project, it is important to take into account these changing perspectives on frontline work, in order to judge the context of the Participation Act properly.

Discretion

Discretionary power is an essential element in Lipsky’s framework on how street-level bureaucrats shape policy (Brodkin, 2007; Lipsky, 2010). It is the freedom street-level

bureaucrats (need to) wield in a context where indeterminate policy and its interests meet a practical reality with the street-level bureaucrat as a mediator. Discretion is essential to the work of Lipsky’s street-level bureaucrats (teachers, police officers). It is needed to perform their job, since it forms the link between the ideal-typical designed policy, and the practical reality upon which this policy needs to fit. In that sense it is an almost technical concept, adjusting for the technical limits and indeterminate nature of policy. Moreover, discretion is

(11)

11

also a normative concept. The assessment of whether citizens are eligible for services for example is inherently normative, coloured by experiences, education and biases towards clients. Lipsky recognised the opportunity for dilemmas to arise at the frontline, precisely because of this double loyalty (Lipsky, 2010). Other researchers have elaborated how these dilemmas can also extend to conflict between world-views, education, experience job instructions (Kjørstad, 2005; Solvang, 2017). Processes of activation and personalisation in welfare policy only increase the potential for such dilemmas, as a result of their normatively loaded character (Raffass, 2017; Solvang, 2017).

Other authors have made contributions to further elaborate on discretion as an analytical concept. One of those additions is the distinction between discretion-as-granted and discretion-as-used (Hupe, 2013; Hupe et al., 2016, pp. 17–18). Discretion-as-granted describes the freedom of a street-level bureaucrat as it is ‘prescriptively granted by a rule-maker to an actor supposed to apply the rules from the latter’ (Hupe, 2013, p. 12).

Discretion-as-used is the actual behaviour of street-level actors. This distinction recognises that even though frontline workers might technically be granted certain room for

manoeuvre in a design on a macro level, the practical, micro-level reality might be less positive. A lack of familiarity with new practices could discourage frontline workers from using this discretionary space to its full potential. The other way around I would argue that discretion-as-used can also refer to the use of freedom by frontline workers in areas not incorporated in discretion as-designed. The distinction between these two dimensions of discretion, as well as how they relate to each other, proves very important in the empirical research of this thesis. A second valuable addition for analytical purpose is a differentiation between discretionary power for each task of a frontline worker (Hupe et al., 2016). In their research, Hupe and Hill find differing levels of perceived discretionary power with frontline workers depending on the task. While discretion in one part of frontline work may be limited, another facet could warrant high levels of discretion. To reflect the behaviour and context of the frontline worker properly, it is important to take this into account and be sensitive to the different responsibilities and tasks of frontline workers.

(12)

12 1.2 Activation

The active welfare state has not just been introduced in the Dutch welfare state with the introduction of the Participation Act. Earlier, around the ending of the 20th_{century the}

relationship between state and citizen was already re-evaluated (Jessen & Tufte, 2014; Ossewaarde, 2007; Raffass, 2017). The state providing universal benefits and the citizen playing passive recipient was deemed no longer feasible or desirable. Using this relationship as the base of government action led to criticism of a financially heavy, patronising and ineffective government (Bruggeman & Rooij, 2018; Ossewaarde, 2007; Van Berkel et al., 2017; Vermeulen, 2015). For more than two decades now the relationship between government and citizen has shifted progressively towards one where government is more minimal, and the citizens position is characterised not only by rights, but also by certain duties. Although this representation might come off as a pessimistic perspective, the activation of citizens can also prevent them from being held back in developing to their full potential (Jeene, Van Oorschot & Uunk, 2014; Knijn & Van Wel, 2014; Vermeulen, 2015). This is particularly relevant for socially vulnerable groups, whom under the ‘old’ system ran the risk of ending up in a system which could institutionalise them (Knijn & Van Wel, 2014). Hence, activation is a rather ambiguous term that, depending on the chosen perspective, can lead to quite diverging results in policy. A prime example of this is welfare provision by the state (Kjørstad, 2005; Ossewaarde, 2007; Rice, 2015). Activation has been used as an argument to reduce welfare benefits of unwilling recipients and to create mandatory volunteer programs as compensation to the broader society (Raffass, 2017). Other authors and policymakers adhere to a broader perspective on activation in welfare (Solvang, 2017; Van Berkel & Van der Aa, 2012). They argue for more constructive governance that allows for everyone to participate in every dimension of society, as well as the labour market. The discussion in government and the academical field surrounding activation as a concept does not limit itself to the frontline of current policy implementation. At the beginning of this century this more active and individualised perspective on citizenship was incorporated in New Public Management initiatives (Ibid., 2017; Van Berkel & Borghi, 2007).

Activation as a concept of policy is currently dealing with two interlinked issues in its implementation. First, there is the issue of knowledge and expertise (Breit et al., 2018; Van Berkel et al., 2010; Van Berkel and Van der Aa, 2012). ‘What works’ in activation is subject to

(13)

13

debate because it is a relatively young field of practice; activation work is not (yet) a

profession in the sense that there is a generally accepted body of knowledge upon which to act, both for policymakers as well as frontline workers. The second issue is partially related to the first: critics argue that activation in practice leads to undesirable politicisation of certain government services (Brodkin, 2007; Lipsky, 2010). These critics emphasise that activation contains an inherently ideological component prescribing desirable behaviour for citizens in interacting with the government. Certain rights become conditional: welfare assistance can be reduced when a recipient does not comply, in the eyes of the state. The right to a certain standard of living suddenly becomes trial for deservingness. If not in itself, then differing perspectives on what activation is, or should be, can lead to issues in

implementation. Because there is relatively little uncontested knowledge, discretionary decisions in activation policy are susceptible to external influences. Personal biases or experiences as well as ideological views could influence the behaviour of a street-level bureaucrat in their interaction with citizens (Solvang, 2017). While his research did not concern street-level workers, Van Oorschot (2008) did establish five criteria citizens use to assess whether they deemed certain groups more or less deserving than others of social assistance in welfare reform. While those working at the street-level in a professional capacity are deemed to be neutral, his findings do offer an interesting insight when researching the daily practice of working with clients in activation work, as will be further elaborated on in the empirical section of this thesis.

In public policy research, there are various perspectives one can take in research relating to activation. In frontline research one of those perspective, and the perspective most

interesting to frontline research, focuses on the sources which influence the discretional behaviour of frontline workers. From studying the weight of ethical dilemmas and considerations, to studying the relative importance of different sources of knowledge in influencing frontline workers (e.g. education, policy goals, job instructions) (Breit, Fossestøl & Pedersen, 2018; Kjørstad, 2005; Solvang, 2017). Activation is present from abstract levels of policy discourse (Yerkes & Braken, 2019), to conversational practices of frontline workers (Kampen & Tonkens, 2018; Van Berkel et al., 2017).

The academical and policy debates around activation show how ambiguous the concept is, and also how normatively loaded (Jeene et al., 2014). These characteristics tie this policy to

(14)

14

Lipsky’s discretion for frontline workers. There is an essential role for the frontline worker to shape activation policy, because so much in policy is unknown or underdefined or

dependent on the situational context. This implies an essential role for the frontline worker in the Participation Act, as the latest embodiment of activation in Dutch welfare to give this interpretation required by the policy.

1.3 Personalisation

Finally, a major goal of the Participation Act is providing tailor-made services or services adjusted to the individual’s needs to stimulate activation (Bruggeman & Van Rooij, 2018; State Secretary for Social Affairs and Employment, 2014). Like activation, the Participation Act is not the first move towards personalisation: it is however one of its main focal points to produce better policy (SER, 2016). It is also a concept that is often used in popular discourse (Rice, et al., 2018). Like many of the concepts in the large-scale reform that is the

Participation Act, ‘tailor-made policy’ (Aanpak op maat) is an ambiguous concept. It instils an image of policymaking as close to the recipient as possible, and to acknowledge the

heterogenous characteristics of recipients in the types of policy they encounter. Solvang (2017, p.537) defines personalisation as ‘[…] the principle (…) that services should respond to the individual instead of the individual having to fit with the service’. Personalisation in public policy can express itself in multiple ways. From personalised conversations with clients to intake systems sensible to individual characteristics to divergent policy instruments depending on the client (Kampen & Tonkens, 2018; Solvang, 2017; Toerien, Sainsbury, Drew & Irvine, 2013).

Personalisation as a concept is hard to pin down in a definition. It is hailed as a means to increase efficiency and reduce costs, produce flexible policies, sensible to a heterogenous clientele and the character of wicked problems at hand. Meanwhile relatively little research has been done as to what a personalised approach means in practice (Kampen & Tonkens, 2018). The Participation Act provides more diverse instruments for municipalities to use in welfare, and aims to reduce red tape and clients falling between the cracks of the system (SER, 2016). Again, frontline workers have significant discretionary power in this process of personalisation (Solvang, 2017; Van Berkel & Van der Aa, 2012). Their position is potentially pivotal in matching the services to the needs of the client. The proper working context is

(15)

15

important in enabling the frontline worker to provide these personalised services, for example through providing adequate education for the tasks at hand, or sufficient available resources. (Rice, 2015; D. Rice, et al., 2018; Solvang, 2017; Van Berkel et al., 2017).

Activation policy and personalisation are relatively young fields (Nothdurfter, 2016). They are also heavily present in public discourse. Despite this attention, research into the street-level manifestations and processes of activation and personalisation has lagged behind (Van Berkel et al., 2017). Yet this is right where particularly these processes should be studied. Lipsky brought attention to the (re-)shaping of policy at the street-level. Activation and personalisation are both concepts that emphasise developments at the level of the individual. It is a simple step to connect these concepts. Research that does focus on the street-level processes of activation and personalisation either chose the perspective of the client, while this thesis elaborates on the perspective of the frontline activation worker, (Eikenaar, De Rijk & Meershoek 2016; Kampen and Tonkens, 2018) or they do not focus on the Dutch context. (Jessen & Tufte, 2014; Solvang, 2017; Van Berkel, Larsen & Caswell, 2018).

(16)

16

2. Methods & Research Design

2.1 Research strategy

The theoretical chapter of this thesis discussed various theoretical fields ranging from street-level research to activation and personalisation research. While some of the discussed theories supplement each other, others are partially mutually exclusive (such as the characterisation of street-level bureaucracies vis-à-vis frontline work and

professionalisation). The function of the empirical research in this thesis is to generate a reflection of frontline workers’ perceptions and interpretations of working with the Participation Act. The variation in theoretical and empirical findings in other research reflects the need for sensitivity to local context when researching the frontline practices of activation work under the Participation Act. The challenge in choosing a research design and methods thus lies ensuring sensitivity to local context and to frontline workers’ own

interpretation of said context. This chapter discusses the research design for this thesis project, as well as the process of case selection, data collection, recruiting respondents and finally methods of analysis including the adoption of the interpretative approach and the use of ‘stories’ to relate frontline workers’ experiences (Polletta, 2009).

2.2 Research design

A qualitative case study offers the best conditions to ensure this sensitivity as well as the ability to reflect upon the prevalence of the various theoretical approaches and empirical findings described in the theoretical framework of this thesis. Frontline research, in

particular research that focuses on the perspective of the frontline worker, is incompatible with ontological and epistemological assumptions underlying what Stone calls the

‘rationality project’ (Stone, 2012). This rationality project assumes that policy and by extension policy analysis are created on the basis of rational, step-by-step decisions. This thesis focuses on the individual and shared perceptions and experiences of frontline workers and how these interact with the local context to inform daily practice of frontline workers. By definition of the research question and accompanying theories this thesis is based upon, I reject these assumptions of the rationality project when it comes to studying the

Participation Act. The Participation Act is not implemented step-by-step or linearly and uniformly. The activating welfare state and a decentralisation of governance ask for

(17)

17

increasingly more local and situational interpretation of policy. Adopting a constructivist approach to the meaning of such policy reform in a qualitative case study allows this thesis to properly reflect the conflicting factors between a policy reform as it is designed and how frontline workers create policy in practice in their work. The fields of frontline work,

activation work and personalisation are yet to be extensively connected to the current context of the social domain in the Netherlands under the Participation Act. A case study with elaborate conversational interviews allows for flexibility in a relatively unexplored field as well as sensitivity to new issues raised by activation workers themselves. As such, the broad and open approach chosen in this research design minimalizes the risk of omitting important theoretical approaches for the basis of data collection. The choice for this research design prioritises a broad and open view over in-depth testing of theories or the generalisation of findings.

2.3 Case selection

When initially selecting a case for this project, I set out to find a typical case (Seawright & Gerring, 2008, p. 229). Originally it was the goal of this research to explore friction between policy as designed and policy in frontline practice. While generalisation was not the main goal, a typical case would grant as close to a baseline measure for these frictions and discrepancies as possible. This baseline measurement was, and still is required, as research evaluating the Participation Act at the frontline level is still relatively scarce. As such, Zwolle was selected as a case. Cities in the ‘Randstad’ (Amsterdam, Rotterdam, The Hague and Utrecht) were dismissed as these cities face their own unique labour market conditions, which are not representative for most of the rest of the Netherlands. For the same reason, cities in rural or ‘remote’ regions of the Netherlands (Zeeland, Groningen) were dismissed, as well as cities in labour market regions consistent solely of smaller towns. Zwolle is a city of fairly average size (±130.000 inhabitants) with no clearly deviating characteristics regarding the labour market. Further statistical information about the employment and activation services, gathered from the UWV, showed a fairly average if not slightly higher rate of employment for people with labour impairments. Due to a lack of further information about individual labour market regions, this last observation was taken to imply that if frontline workers in Zwolle felt certain challenges and frictions working under the Participation Act, these were more likely to be present in other regions as well. Whereas a lower employment

(18)

18

rate could imply that frontline workers in that region might experience (additional) external factors influencing their work which were not present in other regions.

What I encountered in Zwolle, however, was a completely different type of case. This research project initially intended to find discrepancies between a formally designed policy reform along with policymakers and managers attempting to implement it on one side, and frontline workers coping with a practical reality and their own values and ideas on the other side. More specifically: what room for own interpretation and judgement was available in the form of discretion in each of these two situations. Instead of a typical case, Zwolle turned out to be a deviant case (Seawright & Gerring, 2008, p. 302), insofar that

professionals did not work under systematic guidelines of which they were fully aware, while being monitored and supervised by management. Instead, activation workers in Zwolle enjoyed very limited restrictions on their discretion and were granted great freedom in defining their own tasks and responsibilities. This autonomy was sometimes formally granted to them by municipal policymakers through deliberation or local decrees, while other times it simply arose because exhaustive interpretation of policy and frameworks by superiors was absent. Moreover, frontline workers were not universally aware of formal regulations to the high degree that was required for them to relate to me how their own behaviour was informed by, or related to formal regulation. Instead, frontline workers related their work, responsibilities and decision-making to a much more ambiguous amalgamation of (partial) knowledge about formal rules, personal experiences, organisational norms and personal interpretations of policy frameworks. I create the concept of ‘discretion-as-constructed’ to describe this amalgamation. The original intended objective of this thesis was to provide some systematically structured overview of exactly which regulations caused friction between a policy design and the practical implementation and interpretation by frontline workers. The deviant case of Zwolle does not allow for such a systemic, almost positivist analysis. Instead it offers an opportunity to show a case where activation workers have had a high level of discretion and responsibility (or need) to create their own meaning and interpretation of the Participation Act and to create accompanying practice. While this type of case might reduce the possibility for generalisation of findings, it offers a unique opportunity to research how frontline workers create meaning with the

(19)

19

Participation Act on their own as a ‘community of meaning’ (Yanow, 2000, pp. 5–6) when external influences and restrictions are only remotely present.

2.4 Methods and analysis

The deviating case I encountered in Zwolle is essentially home to an institutional void (Hajer, 2009, pp. 33–36). The Participation Act is an inherently inexhaustive policy framework in need of local interpretation by municipal policymakers and situational interpretation by frontline workers. As local policy interpretation in Zwolle is often lacking or left to the activation worker and direct managerial supervision is absent too, there is no clear and generally accepted set of rules in the domain of activation work in Zwolle. As such there is an institutional void ‘[…] where there are no generally accepted rules and norms according to which politics is to be conducted and policy measures are to be agreed upon.’ (Hajer, 2009, p. 34). Hajer argues that authority in an institutional void stems from a negotiated framing of the context and the issues at hand. In other words, how meaning is given to the context and to the issues at hand becomes increasingly important when there is an institutional void.

The interpretative approach to policy analysis offers the appropriate toolset to reconstruct how activation workers in Zwolle have created this frame in their institutional void. As Dvora Yanow argues: ‘The role of the interpretative policy analyst is to map the ‘architecture’ of the debate relative to the policy issue under investigation, by identifying the language and its entailments (understandings, actions, meanings) used by different interpretive

communities in their framing of the issue’ (2000, pp. 12–13). In this case, the activation workers in Zwolle are the interpretive community, or community of meaning, that is the focus of this thesis. The empirical data on discretion-as-constructed (the concept I use to reflect the shared understanding of the context and the rules in Zwolle in the absence of proper discretion-as-granted) will be analysed with the use of ‘stories’. Stories are a tool from the interpretative approach to policy analysis and allow to reconstruct the abstract narrative that is discretion-as-constructed in Zwolle. The concept of stories, or narratives, offer the possibility to properly portray how the activation workers in Zwolle all adhere to a shared system of values and ideas, without the assumption that the meaning of this story is identical for all of them.

(20)

20

The benefits of a single case study and employing an interpretative approach are that they allow for high levels of flexibility and sensitivity to all aspects of the local context. It is inadvisable to employ a stringent approach in an attempt to classify a deviant case such as Zwolle into pre-existing categories or theories. This could nullify the potential to learn from the unique opportunity to study all the ways activation workers employ to construct their discretionary space in an institutional void. The research design for this thesis does however put limitations on the potential for generalisation of the findings in this thesis. It is not possible in this thesis to uncover to what extent the conditions of the deviant case in Zwolle are representative for other activation services in other municipalities.

2.5 Data collection

The primary data for this thesis are hour-long conversational interviews (Yanow, 2000, pp. 31–33) with nine frontline activation workers in Zwolle, one half-hour interview with another activation worker and one interview with two newly appointed managers. The use of conversational interviews allowed to incorporate the broad basis of theoretical and empirical views established in the theoretical framework. Simultaneously conversational interviews also allow room for follow-up questions when activation workers indicate a certain topic was relevant for them. All interviews started by creating a personal and professional profile of the respondents. They were asked about previous work experiences and education and how long they had been working in activation in Zwolle. In the rest of the interview respondents were asked to describe their daily practice of decision-making in working with clients. This started from the intake process, to how they monitor and counsel their caseload and finally in the possible endings of their work with a client. Throughout the interview respondents were asked to relate the challenges they encountered to the topics of formal regulations, job characteristics, training and education, and the influence of

managers and policymakers. When all these topics had been discussed, respondents were asked if they perceived any obstacles or challenges in their work that had not yet been discussed.

Policy documents functioned as a secondary source of data. The Participation Act (Ministry of General Affairs, 2014; Ministry of Social Affairs and Employment, 2019) and the

(21)

21

(Municipality of Zwolle, 2014) were used to provide the context of the ‘discretion-as-granted’ to activation workers under the Participation Act. The governmental evaluations ‘Municipal experiences with the Participation Act’ (CentERdata, 2017), ‘Disabled youths under the Participation Act’ (Kok et al., 2018) and ‘Social infrastructure vulnerable groups under the Participation Act’ (SER, 2016) were consulted for insight into the current national context of evaluations regarding the Participation Act. The local policy document ‘General Outlook Municipality Zwolle 2018-2021’ provided a general context for the municipal discourse regarding the current state of, and future plans for activation work. These documents have served to reconstruct the context of the Participation Act in the introduction and theoretical framework and as background information for the conversational interviews.

Entering the field and contacting respondents

Upon first contact with Zwolle one activation worker was appointed as liaison and coordinator of the interviews by the management team. Respondents were recruited through invitations passed on by this liaison in four consecutive rounds. An effort was made to recruit multiple respondents from each of the four different teams (Zwolle Works, Zwolle Develops, Zwolle Moves and Team Target Groups). Eventually ten activation workers

responded to one of the invitations: four activation workers from Team Target Groups, four from Zwolle Develops, one from Zwolle Moves and one from Zwolle Works. Zwolle Works is a team which only contains two activation workers, neither of whom were initially willing to participate in an hour-long interview. As a placeholder, a shorter, half-hour interview with an activation worker from Zwolle Works was arranged. In this interview, questions were more specifically targeted to verify statements made by colleagues at other teams about the nature of work at Zwolle Works. Other activation workers from Zwolle Moves were individually invited by the liaison, but unwilling to participate. Each of these respondents have been given a code: AW1 through AW 10. Finally, two newly appointed managers at ‘Tiem’ were interviewed, the new governmental organisation which houses the former activation department and activation workers of the municipality, as well as the former organisation responsible for the Sheltered Employment Act (Wet Sociale Werkvoorziening) which was replaced by the Participation Act. The transition to ‘Tiem’ and the assignment of these managers had only occurred in the spring of 2019 and as such, little had yet changed

(22)

22

in the activities of activation workers. The interview with the managers served to cross-reference some of the statements made by respondents and to get a more complete perspective on the current state of activation work in the municipality of Zwolle. The managers have each been given a code: MT1 and MT2.

(23)

23

3. Empirical data & Analysis

The activation department in Zwolle stands at the beginning of a period of change when I arrive there. The modern building, located at a business park on the outskirts of Zwolle, wears the name ‘Tiem’. This new organisation is the product of the merger of the activation department and the department in charge of implementing the now repealed Sheltered Employment Act [Wet sociale werkvoorziening: Wsw]. The clients of the activation workers are welcomed at the reception and directed to the waiting room, consisting of three benches next to the reception, to wait for their activation worker. When ready, one of around 35 activation workers then take their client to the area where most of the

counselling meetings take place. This area consists of a row of around ten cubicles and a few of secluded rooms for the meetings that require a bit more privacy. This counselling area is but a small part of the building. On the upper floor are the offices and meeting rooms for staff, where most of my interviews took place, but the largest part of the building is dedicated to the sheltered workshop. Here, clients from various backgrounds work in

production, packaging and shipping jobs. Altogether, this is the new heart of activation work in Zwolle.

In this chapter, I will discuss and analyse the empirical findings of the case study in Zwolle. First I will establish a general overview of the framework in place to grant activation workers their discretion-as-granted. In the interviews with activation workers, it quickly became apparent that this framework was not the most important structure in steering activation workers actions. Instead, I introduce the concept of discretion-as-constructed to describe an abstract construct which activation workers have created among themselves to define their role and their responsibilities. This discretion-as-constructed will be recreated through the use of stories (Polletta, 2009). After that, each separate part of the story, beginning, middle and end will separately be discussed an elaborated upon. In the final section of this empirical chapter, the discretion-as-used by activation workers will be discussed as they employ it in daily practice with clients.

(24)

24 3.1 Discretion-as-granted, or as-constructed?

The Participation Act is a policy reform that crosses levels of local, regional and national government. As such, there are multiple stakeholders, as well as boundaries between governmental and non-governmental organisations. Therefore, discretion is influenced by multiple sources, and as such can become quite ambiguous.

Even though decentralisation is a prime part of the Act, with a greater responsibility for the municipalities, regional and national regulative frameworks partially form the institutional context in which these municipalities and their professionals work. In Zwolle, one of these frameworks is created by a regional cooperation between Zwolle and (smaller)

municipalities around Zwolle. This cooperation, to an extent, decides what instruments are available to frontline workers and coordinates labour market policy between municipalities in the region. For example, they deliberate on who qualifies for jobcoaching and for how long. Jobcoaching, in this case, is one of the special instruments available to clients with a ‘Target Group-indication’. While ‘jobcoaching’, or ‘coaching’ are often used interchangeably, or for every form of counselling a professional offers to their clients, only those with a Target Group indication are eligible for counselling after they have found work. Other clients are only eligible for coaching up until they find employment.

However, these regional arrangements were only partially effective in demarcating the discretionary space for frontline workers:

On top of everything you have the Regional Labour Organisation, which makes regional arrangements regarding our services, regarding jobcoaching: óúr field of expertise. So that’s also something you have to deal with: local policymakers saying ‘no, that’s not the way we are going to do this’, while regional policymakers will then say ‘No, no, that ís how we are going to do it. So, you know, it’s a bit of mist [referring to policy], which I tend to disregard. That’s why I enjoy the implementation, where it’s just me and my colleagues working it out together. (AW6)

In this section I want to briefly sketch the formal framework that frontline workers did in fact adhere to in telling their stories. This framework is the source of the discretion-as-granted as Hupe (2013) describes it: formally granted by a higher level of governance. After this, I will reconstruct how frontline workers created their own discretion, partially built out of

(25)

25

influencing policy on their own, and partially by interpreting existing policy through a shared story.

A minimal formal framework

First, it is necessary to reconstruct more or less formal structure of the social domain in Zwolle. In Zwolle clients can enter the system of welfare through multiple canals. Most important is ‘The Portal’ (De Poort). Clients who apply for social support get invited for a conversation at this municipal organisation. It functions as a point of triage where they are assessed on their distance from the labour market. To help this triage, a DARIUZ-evaluation can be used (an evaluation method for basic labour and life skills), among other instruments to help construct a report for referral to other municipal organisations. Beside the activation services where this research takes place, those can be income support, who monitor the eligibility of welfare recipients, and focus on the financially administrative side of welfare support.

A counsellor at The Portal writes up a report which includes a client’s personal and labour history, as well as a summary of other troubles and obstacles in a client’s life, resulting in an assessment on what needs to be done or resolved to help a client back to work. Along with it, they try to assess how long a client will need to be counselled to return to work. Frontline professionals at the activation service themselves have varying insight on how this

assessment is made. A term that floats around is ‘life sectors’, and whether there are issues in a client’s life in these sectors. Examples of these sectors are: financial situation, familial or domestic situation and health status.

When they are referred, clients end up at either Zwolle Works, Zwolle Develops, or Zwolle Moves. Each of these, respectively, is meant for clients that require increasingly more counselling or other services to return back to work. The general distinction maintained by many frontline workers at the activation service is: Zwolle Works is for clients who are ready for work; Zwolle Develops is for clients who have issues in one of their life sectors, but are ready to be counselled towards work; Zwolle Moves is for clients who are not yet ready for work, caused by multiple or severe issues in the life sectors, but are ready for some light form of activation. Each of these groups has their own expected term, within which they are expected to be able to find paid work:

(26)

26

• Zwolle Works: 6-9 months

• Zwolle Develops: 6-12 months

• Zwolle Moves: up to 24 months

Outside of the activation service, clients can also be referred to either the Social Workplace, or a Social District Team. The Social Workplace was meant for welfare recipients who fell under the now expired Sheltered Employment Act, or Wsw. The Wsw was meant for individuals who were unable to earn a living wage in regular, private employment. As discussed earlier, it was replaced by arrangements under the Participation Act, but the clients already employed in the Social Workplace retained their positions.

When a client was deemed to have extensive issues in their life sectors, activation towards work could yet be unfeasible. In this case clients were referred to a Social District Team [Sociaal Wijkteam]. These teams offer highly decentralised support with general life issues, as these clients tend to have multiple interconnected issues and require assistance with establishing a basic life routine, in cooperation with other services such as debt counselling.

Team Target Groups

Aside from these three groups, since the introduction of the Participation Act in 2015, there is the ‘Team Target Groups’. This subsection of professionals was tasked with handling clients from the ‘Target Group Register’. This register is meant to replace the Wsw, in so far that it is meant for those who are permanently deemed to be unable to earn a minimum living wage, in an entry level job, due to mental or physical impairments.

When a client is eligible for the Target Group, they obtain an ‘indication Target Group’. This can be obtained in several ways. The two main ways of obtaining such an indication are either through applying for it at the UWV, or when youths finish or leave schools for special-needs education. In theory, individuals in the first group apply personally for the indication, after which they are assessed on eligibility. In practice, this group already receives care in some form, and is often aided and stimulated by these caregivers to apply. Frontline workers in Zwolle do occasionally help their clients to file the application, if they expect a client has more structural issues and could benefit from the additional instruments. The goal for this team is to employ their clients within a year. However, as the indication Target Group in

(27)

27

principle is given to a client for life, these clients do not leave the caseload of an activation worker. Rather they continue working, and their activation worker monitors and updates the relevant subsidies.

Instruments

The instruments available to frontline professionals in activation are loosely described in the Participation Act. Regional and local policymakers then give further local interpretation to the broader framework offered by the Act. The majority of the instruments directly related to work is dependent on private partners in the region.

There is a variety of instruments available in Zwolle, broadly divided into five groups:

1. coaching or counselling, 2. diagnostic instruments, 3. educational instruments,

4. internal or external placements, and 5. referring to external services.

This distinction is not formally present in Zwolle. Some instruments can also serve multiple purposes, as I will explain. In an attempt to create some sort of order and oversight I derived the aforementioned categories from the way respondents refer to their instruments in the interviews. First, counselling is the most important red line through all the programs activation workers take their clients through. Counselling includes helping with creating a resume or advising on how to approach a job interview, but can also extend much further into helping with more personal or domestic issues. I will return to this later, but for now it suffices to state that it includes most of the services directly delivered by the frontline worker to the client in one-on-one conversations. Second there are some diagnostic

instruments, varying from asking other (municipal) organisations to carry out evaluations on their client’s skills and possibilities, but can also include carrying out a career choice test. Thirdly, external placements are one of the more important instruments on the road towards work. There are varying conditions under which such placements take place. An activation worker and an employer can agree on a trial period, or employment while the client maintains their welfare. Instruments like this can either provide work experience, or

(28)

28

be a step-up to regular employment. Internal placements are also used as diagnostic instruments, although they are not directly intended for that purpose. Sometimes when a professional has difficulties finding out a client’s strengths and weaknesses, they let them work internally at the municipal sheltered workshop while monitoring the situation. This serves both the purpose of helping someone develop labour skills, as well as revealing possible obstacles on the road to work in a controlled environment.

Lastly, there are external services to which activation workers can refer their clients. These services range from jobcoaching by external, private parties (in the case of lacking expertise or an overflowing caseload), to various other professional services relating to other facets of life, such as psychological counselling. These external services can be used for varying reasons, but are mostly used to fill in for services that are not or cannot be offered by the frontline professionals, or by other municipal services.

3.2 Discretion as constructed

The framework described above, offers only limited guidance for professionals in daily practice. As such, more interpretation is required by professionals, but it is not always given by higher-ups. In Zwolle, this lack of further guidance became quite extreme through a combination of causes. First, and more specific to Team Target Groups and the Participation Act, there was very little preparation before 2015. When the Act was about to come into effect, three frontline professionals were hired, with expertise regarding working with clients such as those who would be eligible for the Target Group Register. These frontline

professionals were essentially given free rein in setting up instruments, work planners and intake procedures. One of those had previously assisted in setting up these arrangements in another municipality:

What you’ll see, is that policymakers and implementers each don’t know what the other is doing. That’s also what happened in Zwolle: we just began. Our department leaders also gave us a sort of ‘carte blanche’: go do it, and make the numbers transparent, so I can follow it. (AW6)

Second, professionals also lacked direct supervision on their day to day work: ‘As Zwolle Develops, we are very free to do as we see fit, I would say. We are almost a self-steering

(29)

29

team.’ Another frontline worker at Team Target Groups who joined in 2017, when asked what their job description was, stated:

That’s the common thread: there is just no policy on a municipal level. We’ve never actually received that assignment. Yes, we got ‘you will do the Target Group Register’, but that’s it. Nothing about how or what. We made our own policy, and they

[municipal policymakers] agreed with that, but no official assignment. (AW8)

While it is not the case that there was no supervision at all, it was inconsistent and subject to personnel changes and reorganisations. As such, the frontline workers lacked a point of contact, and a supervisor who regularly reflected on their work with them. Professionals who only joined the Zwolle social service in the past few years were used to having such forms of supervision at their previous workplaces. They, and professionals who had worked in Zwolle longer, both indicated they lacked a point of contact when rules and directives were incomplete or inconclusive. Only as recently as spring this year, 2019, a new

managerial board has been appointed. As a result of this unusual situation, professionals felt the need to fill in the blanks amongst themselves.

In implementation research it can be functional to distinguish between

discretion-as-granted, and discretion-as-used, to symbolise how daily practice does not necessarily reflect practice as it was designed. There is definitely added value in this distinction for analytical purposes. It is, however, a distinction that cannot be made in the case of Zwolle and the Participation Act, as it assumes two conditions that were absent: the presence of a clear and exhaustive policy design, as well as sufficient understanding of this policy design by the professionals carrying it out. As stated before, the Participation Act is a framework of a policy reform, which requires further interpretation. In Zwolle, professionals often were tasked directly or indirectly with creating that interpretation themselves. How this was done exactly depended on the context of the specific situation, but the common theme was that professionals resolved issues amongst themselves. By this I mean that they used whatever resources were available, and often shared the insight they gained with their peers. As such they built a framework constructed out of both formal and informal rules to inform their day-to-day routine. As informal norms constructed in cooperation with peers are as important in dictating this behaviour, as externally designed procedures, it would be

(30)

30

incorrect to refer to designed’ instead, I coin the term ‘discretion-as-constructed’, to describe this amalgamation of rules, norms and ad-hoc solutions, from which these frontline professionals in Zwolle draw to inform their daily actions.

The new concept of ‘discretion-as-constructed’ also better reflects the ambiguity of this discretion, as opposed to ‘discretion-as-granted’, which assumes a more linear and top-down ‘granted’ discretion as if it solely comes from clearly distinguishable sources higher up in the implementation process. Instead the interpretation and creation of policy and practice within the framework of the Participation Act goes back and forth between different

organisations, actors, and levels of governance. As a result, particularly in Zwolle, this bigger picture of ‘what are our tasks, roles, responsibilities and limits’ that informs activation workers in their daily practice, is much more abstract and ambiguous. When a frontline worker in Zwolle encounters a situation which they do not know how to approach, they can turn to formal rules and guidelines (discretion-as-granted) for more clarification. But just as often, they turn to colleagues, or to the larger question of ‘what is our assignment here’ to inform their actions in specific situations.

Frontline workers in Zwolle have constructed their own story, or narrative as to what their general purpose is, both as individuals and as an organisation. This bigger picture can both align with, as well as contradict formal regulations. As a group of individuals, frontline workers can agree to a certain story or narrative, however, the exact interpretation of this story in an everyday context can vary from person to person. Discretion-as-constructed is the story activation workers refer to when discussing their work and their position in a setting of an interview, or when they reflect on their responsibilities with peers. As such it partially substitutes a lack of clear discretion as it is formally granted to them by a higher authority (Hupe, 2013). Discretion-as-used, then, is the practice of activation workers when they interpret discretion-as-constructed into concrete actions and decisions to employ their discretion in their work. As such, as-constructed is different from both discretion-as-granted, as well as discretion-as-used.

(31)

31 3.2.1 Through stories

Discretion-as-constructed is more a set of vague notions, than it is a concrete set of norms. As such, the framework of stories lends itself well for capturing discretion-as-constructed for analytical purposes, as well as reflecting the ambiguity at the heart of the concept, when interpreted by individual professionals. This section focuses on recreating the red line through the stories the frontline workers used, to inform their decisions while working with candidates.

Stories are a tool from the tradition of interpretative analysis. The concept hails from sociological research, in particular into social movements, but has its use outside that arena as well. In social movements these stories hold the power to unite those whom relate to the story, as well as steer action in a certain direction, as they display desirable (and

undesirable) futures (Polletta, 2009, pp. 8–9). Stories can be an alternative to

all-encompassing ‘narratives’ which ought to reflect one, universal truth. Stories assume that there are multiple, partial truths: ‘All stories have characters and points of view from which events are experienced’ (Ibid., 2009, p.9). Individual people’s stories will differ, is the underlying assumption. What I found in my interviews, is that the stories the activation workers told, deviated from the stories told in public discourse, national media and political debate on a national level. What stories allow us to do, is to: ‘[…] make explicit, the

underlying cultural schemas that underpin institutional practices’ (Ibid., 2009, p. 13).

On a technical level, what separates stories from ideologies, is that stories have a beginning, a middle and an end. They are not meant to serve as universal guidelines for multiple settings. Stories start by painting a setting or a context in which they take place. In this following section I will describe the story as it serves to describe the discretion-as-constructed for activation workers in Zwolle:

A willing, but often troubled candidate comes to us, full of potential, amidst the harsh world that is the social domain. Through working with, as well as adapting to a client’s needs, and building a relationship of trust we do whatever is necessary, within the boundaries of the rules, to improve their readiness for (work-)life, and their general well-being with the eventual goal of durable employment. Through attending the needs of the

‘All within the margin of the rules’ : how activation workers in Zwolle construct and use discretion under the Participation Act