Practices of activation in frontline interactions: Coercion, persuasion, and the role of trust in activation policies in Germany

Soc Policy Admin. 2018;1–14. Abstract Following Michael Lipsky's well‐known argument that policy is made in the daily encounters between street‐level bureaucracy and citizens, a growing body of research emphasizes that actors and organizations delivering social and labor‐market policy play a crucial role in welfare‐state politics. Using qualitative data collected at three local employment agencies in Germany, this article explores worker‐client relations as a crucial mechanism throughwhich activation policies are translated into practice. The analysis investigates how caseworkers define their role and their relationships with clients. The findings show that it is essential for caseworkers to achieve client compliance. In such a context, building relationships of trust is a strategic instrument in overcoming possible barriers to co‐operation in the caseworker‐client interaction. Caseworkers develop strategies to create the impression of trustworthiness and to motivate both unemployed clients and employers to become trust‐givers in the caseworker‐client relation. While research has often stressed the dichotomy between disciplining and enabling elements of activation policies, our explorative study shows that persuasion and trust‐building are a further important dimension of the frontline delivery of activation policies. These strategies reflect the importance of emotional aspects of frontline work.


| INTRODUCTION
Reforms of unemployment protection systems and labor-market policies have been an important element in the transformation of mature welfare states. Activation policies make welfare benefits increasingly conditional on the welfare recipient's proof of fulfilling certain obligations, namely that he or she is undertaking activities towards (re-)entering the labor market. Although in the development of welfare states many rights have always been linked to individual responsibilities, the extent to which the principle of conditional entitlement has become central to the organization of public welfare is characteristic of welfare reforms encapsulated under the label of "activation" (Dwyer, 2004;Wright, 2012). Activation policies put an emphasis on individual obligations in return for welfare benefits. Policies of "welfare-to-work" or "activation" rely on the notion of an "active, responsible, self-managing 'worker citizen'" (Newman, 2007, p. 371) to replace that of a-presumably-passive, dependent welfare recipient.
Citizens are considered to be responsible for their own lives and expected to invest in their own employability.
Different streams of research have pointed out the policy impacts of this shift towards activation, focusing on legislative and institutional changes, often in a cross-national perspective (e.g., Bonoli, 2010;Clasen & Clegg, 2006;Serrano Pascual & Magnusson, 2007). Moreover, scholars have analyzed the restructuring of welfare administration, identifying trends such as processes of marketization in the provision of activation and employment services, processes of decentralizing policy-making authority, and the introduction of New Public Management (NPM) styles (Gottschall et al., 2015;van Berkel, de Graaf, & Sirovátka, 2011;Weishaupt, 2010).
While institutional change is certainly at the core of welfare reforms, it is important to look at the micro level of policy implementation in order to understand what activation means in practical terms and how activation policies affect the lives of the unemployed in practice. Recently, a growing body of research emphasizes that actors and organizations delivering social and labor-market policy play a crucial role in the politics of welfare states. In particular, research into the tradition of the street-level bureaucracy approach brought forward by Michael Lipsky (2010[1980) has drawn attention to the key role of street-level organizations and frontline workers for the delivery of activation.
Studies have examined the influence of organizational and managerial characteristics on frontline work (e.g., Brodkin & Marston, 2013), and in particular have looked at the nature of frontline workers' discretionary decision-making within the context of activation policies and NPM reforms (Baker Collins, 2016;Brodkin, 2011;Evans & Harris, 2004;Jessen & Tufte, 2014). Moreover, research has stressed the moral dimension of street-level bureaucracy, focusing on frontline workers' judgements and the social construction and categorization of clients (Maynard-Moody & Musheno, 2012;Sainsbury, 2008;Tabin & Perriard, 2016).
While most of this research is based on the theory that the encounters between welfare workers and their clients are central to understanding the implementation and consequences of activation policies, it rarely provides a thorough analysis of inter-individual relationships between welfare administration staff and their clients. Building on the theoretical perspective of street-level bureaucracy along with insights from recent empirical studies on street-level work in the employment and social-policy area, this article explores caseworker-client relationships in the context of the German unemployment insurance scheme. By drawing on qualitative data gathered at three local employment agencies, the analysis explores how caseworkers define their roles, how they interpret their relations with clients, and what resources they develop in the process of policy delivery. Thereby the article seeks to contribute to understanding patterns and practices of activation in the implementation of social and labor-market policy.
The text is structured as follows. After a short conceptual note on caseworker-client relations, we provide a brief description of the German labor administration and its transformation in the course of activation reforms, thereby outlining the legislative framework and the organizational context of our study. Subsequently, we offer a brief account of data and methods. The findings highlight that in the process of activation delivery, achieving client co-operation is essential for caseworkers. The results suggest that practices of coercion and disciplinary instruments in such a context have a strategic importance for frontline workers. In caseworkers' narratives, however, strategies of persuasion-and in particular establishing relationships of trust-play an even more prominent role. In describing their relationships with clients in connection with trust, frontline workers refer to themselves as trust-takers and to clients as trust-givers. We conclude that establishing relationships of trust is an important resource for frontline work and may be a key instrument in the implementation of activation policies. Lastly, we discuss the need for further research on the emotional aspects of frontline work.

| THE CASEWORKER-CLIENT RELATION IN ACTIVATION POLICIES
In his seminal work Street-level bureaucracy. Dilemmas of the individual in public services, Michael Lipsky (2010[1980 argued that everyday interactions between state agents and citizens play a key role in the implementation of policies and the delivery of public services. With the notion of street-level bureaucrats, Lipsky was referring to practitioners in the public sector such as teachers, policemen or social workers who interact directly with citizens in the course of their everyday work. He argued that these professional groups have considerable room for discretionary decisionmaking. Street-level bureaucrats typically work in an environment where resources-such as time or informationare inadequate to respond in a satisfactory manner to each individual case and develop practices and instruments in order to cope with these working conditions. According to Lipsky, the decisions of street-level bureaucrats and the routines that they establish "effectively become the public policies they carry out" (Lipsky, 2010(Lipsky, [1980, p. xiii, italics in original). Thus street-level bureaucrats act as mediators between the individual and the state. They hold "the keys to a dimension of citizenship" (Lipsky, 2010(Lipsky, [1980, p. 4).
According to Lipsky, street-level bureaucrats' routines and decisions are situated within and shaped by the organizational context of public services. He argues, for example, that worker-client relations in street-level bureaucracies are characterized by a considerable power asymmetry, resulting from the fact that clients are usually non-voluntary (Lipsky, 2010(Lipsky, [1980. While this seems obvious for institutions with a focus on the coercive dimension of the relationships, such as police departments, it also holds true for welfare agencies and other types of street-level bureaucracies where citizens apply for services that they cannot obtain elsewhere. Within the context of the welfare state, the street-level bureaucracy approach draws attention to direct encounters between welfare recipients and the representatives of welfare organizations as the moment when welfare policies become concrete for the individual. It is through these day-to-day physical encounters between state agencies and citizens that citizens experience the concrete functioning of the welfare state (Dubois, 2014). The re-organization of welfare policies according to the activation paradigm has created specific ambivalences and tensions at the street-level of policy delivery. Studies have, for example, shown that frontline workers may experience conflicts between disciplining and enabling aspects of their work, between realizing short-term or long-term outcomes with clients, or ethical and moral dilemmas (Caswell, Larsen, van Berkel, & Kupka, 2017, p. 183 f.).
An analysis of caseworker-client relations that builds on the contextualized understanding of frontline work that the street-level bureaucracy approach adopts therefore needs to systematically take into account context factors and the various ambiguities and tensions related to them. For the analysis we present in the following, the interplay between the normative principles underlying activation policies and organizational context factors is of particular importance. Policies based on the concept of activation are characterized by individualized interventions directed at changing individual behavior, motivation, and attitudes in order to increase individual employability and promote participation in the labor market (Serrano Pascual, 2007). However, it has been noted that these normative assumptions underlying the activation paradigm and the organizational rules and performance measures derived from them largely ignore the fact that service delivery is a co-production between caseworkers and jobseekers (Bartelheimer, 2008). Moreover, whether organizational performance targets such as labor-market entry are realized depends on a variety of factors which partly lie beyond the immediate control of frontline workers, such as specificities of the local labor-market or recruitment decisions of employers.
Relying on this conceptual understanding of caseworker-client relations as being embedded in the political and organizational context, our study seeks to explore how caseworkers themselves define their roles and how they interpret their relations with clients, what resources they develop in policy delivery, and which shared patterns of activation emerge.

| FRONTLINE WORK IN GERMANY'S LABOR ADMINISTRATION: LEGISLATIVE FRAMEWORK AND ORGANIZATIONAL CONTEXT
The study reported upon in this article was carried out within the scope of the German unemployment insurance scheme. The unemployment insurance scheme is one pillar of a two-tier unemployment-support structure whose division into two parts was cemented during the so-called Hartz reforms introduced between 2003 and 2005 (Dingeldey, 2011). The insurance-based benefit system supports unemployed people with an adequate contribution record for up to one year. 1 The second pillar, a tax-based basic income support scheme (Grundsicherung für Arbeitsuchende), provides a social minimum on a flat-rate basis and is directed at unemployed people deemed capable of working but not eligible for insurance-based benefits. This basic income support scheme is dependent upon a household-related means test and includes particularly strong activation requirements.
The unemployment insurance scheme represents a classical element of Germany's conservative-corporatist welfare system (Esping-Andersen, 1990). Unemployment insurance traditionally rests on the principle of status protection, reflected by benefits which are proportional to previous earnings and employment integration in accordance with qualifications and work experience. Also, in the post-reform system of social security, the insurance principle remains a significant element of unemployment support in Germany (Clasen & Goerne, 2011, p. 801).
Nonetheless, incremental policy changes have, overall, weakened the principle of status protection (Betzelt & Bothfeld, 2011;Bothfeld & Rosenthal, 2018). While income replacement rates remained largely untouched during the reforms, 2 individual responsibilities and obligations in return for unemployment insurance benefits have been stressed. A new definition of the concept of "suitable employment" requires unemployed people covered by the unemployment insurance scheme to accept a job wage equal to the benefit level after six months of unemployment.
Moreover, job-search activities are strictly monitored. The unemployed are obliged to give an account of their job-seeking efforts and to participate in labor-market programs. Non-compliance can lead to a withdrawal of benefits for a certain period of time.
Another aspect of the Hartz reforms was a reorganization of the Federal Employment Agency (Bundesagentur für Arbeit). The Federal Employment Agency-a self-governing public body that acts according to statutory goals and tasks on behalf of the Federal Government and is subject to the legal supervision by the Federal Ministry of Labor and Social Affairs-provides public employment services through a network of local employment agencies across the country (currently 156 agencies with about 600 branch offices). Next to the provision of unemployment benefits, employment services include job and training placement, occupational counseling, employer counseling, and the promotion of vocational and further training.
A core aspect of the organizational reforms were changes in management and governance, such as new systems of performance management and output control (Eichhorst, Grienberger-Zingerle, & Konle-Seidl, 2008, pp. 42-46;Kaltenborn et al., 2010). Moreover, the reforms addressed the placement process itself. Local employment agencies were reorganized according to a unified business model, establishing new standards for dealing with clients (Schütz & Oschmiansky, 2006). Common procedures and tools for placement work were introduced, and these are now used in all employment agencies across the country.
In sum, with regard to its underlying normative principles, frontline work in German employment agencies is situated somewhere between the poles of social rights "earned" through contributions, status protection, and activation policy. It is embedded in the organizational context of a public administration that has transformed its governance and management structures according to market logic over the last two decades. In particular, performance measures and quantifying practices of accounting have been shown to be crucial factors shaping frontline work (Sowa & Staples, 2014).

| DATA AND METHODS
The empirical data referred to in this article were collected as part of commissioned research within the framework of a pilot project carried out by the German Federal Employment Agency. Over a period of two years (2015-16), three local employment agencies were asked to initiate a bottom-up process of organizational change and to develop local solutions for daily tasks in service delivery. The employment agencies involved were to critically examine existing administrative rules and regulations in the area of counseling and job placement, and thereby identify unnecessary bureaucratic routines. When implementing new solutions and re-organizing their work processes, the employment agencies were entitled to ignore administrative regulations issued by the Federal Employment Agency headquarters.
The research project utilized qualitative interviews and group discussions with caseworkers; qualitative interviews with local managers and team leaders; document analysis; and occasional participant observation of team meetings, workshops, and caseworker-client interactions in order to explore the dynamics of bottom-up innovation and to examine how frontline work was organized and experienced within the scope of the project (Freier, Kupka, Senghaas, & Wuppinger, 2017 This article is primarily based on a qualitative content analysis of the interviews and group discussions conducted with caseworkers (21 interviews and seven group discussions). Interview partners were all employees of the three employment agencies taking part in the study. The majority of them were working as so-called "employment oriented" placement officers, meaning that they were responsible for providing job placement services and counseling to jobseekers. In addition, the sample included so-called "employer oriented" placement officers who provided counseling and support to employers (see section 5.4). Regarding frontline workers' educational background, the sample included both employment counselors who had completed an internal training within the Federal Employment Agency (vocational training or an academic degree from the Federal Employment Agency's University of Applied Labour Studies) and employment counselors with other professional qualifications (such as business studies, social sciences, or engineering). Group discussions consisted of four to six frontline workers from the same employment agency; both employment and employer oriented placement officers were represented in the groups.
For the analysis, interview transcripts were divided into analytical units and assigned to inductively developed categories, reflecting the main research questions (Mayring, 2000). Based on the thematic coding of interviews that was carried out with computer-assistance using the software MAXQDA, sequences dealing with caseworkers' role definitions and caseworker-client relations were identified and typical patterns of argumentation were reconstructed.
Interviews with local managers and participant observation of caseworker-client interactions were used to contextualize the findings.
Our study does not aim to compare the three local employment agencies included in the sample. Instead, the analysis seeks to identify frontline workers' common perceptions and shared patterns in service delivery across different regional settings. As mentioned above, in the pilot project that provided the context of data collection, caseworkers were asked to think about possible changes in their daily work. We observed that this encouraged them to talk about their role definitions, their relations with clients, and their interpretations of the context in which frontline work took place. Interviews and group discussions conducted in the later stage of the research project explicitly addressed the caseworker-client relationship and the way caseworkers dealt with clients. The interview sequences relevant to the issues addressed in this article focused on general aspects of frontline work rather than on the specific approaches and instruments developed in the pilot project. For this reason, the analysis does not address these specific aspects in more detail. 4

| CASEWORKER-CLIENT RELATIONS IN THE GERMAN ACTIVATION SETTING
In the following, we present our main findings. Through the analytical lens of a street-level perspective which understands frontline work to be shaped by factors related to the political and organizational context, we show how caseworkers in the activation setting interpret their role and their relations with clients and what resources they developed in policy delivery.

| Activation between support and control
As mentioned above, German labor-market reforms have, in general, strengthened the individual responsibilities of the unemployed while at the same time expanding sanctioning possibilities. Nonetheless, more supportive and empowering functions-such as job-placement services, assistance in transition into the labor market, or the provision of training programs and further training-have for the most part been retained and, in some cases, even expanded. 5 Frontline work within the policy context of activation therefore encompasses a variety of forms ranging from more regulatory and disciplinary, to more enabling and empowering elements. Combining and reconciling these various elements produces a specific role ambiguity: while caseworkers provide counseling and support in (re-)integrating into the labor market-acting quasi as the "gatekeepers" to enabling instruments such as further-training activities-they at the same time guide and monitor job-search activities and decide whether financial sanctions should be imposed in the case of non-compliance.
In previous research on caseworkers' role perceptions, a large majority of caseworkers in German employment agencies (more than 60%) depicted themselves as a "service provider" while about 10% either considered themselves a "broker on the labor market", a "social worker", or a "bureaucrat" (Osiander & Steinke, 2011; see also Kupka & Osiander, 2017, p. 95). 6 In line with these previous findings, when the caseworkers we interviewed talked about their understanding of placement work, many of them put the supporting, service-providing elements of their job in the foreground. One spoke for many when she explained: "What I actually appreciate most is when…clients who leave my room tell me that they had expected something completely different: that they had been afraid to come but that everything had worked out well and that I had been able to help them. I think that is a sign for us that we are doing our job well." (Interview 53) Support is, however, typically understood as being dependent on a client's (assumed) willingness to work. In the words of another caseworker who participated in a group discussion: "Well, yes: people have needs. Those who do want to work-they have requirements: they want to get some kind of support. They want a tip about where to go. Some also need more than just a tip. And that's really what our work is about." (Group discussion 32) The interaction between frontline workers and their clients in local German employment agencies takes place within an institutional context governed by public law. In their encounters with clients, caseworkers work with legal concepts and apply legal and administrative rules to individual cases. Being confronted with a heterogeneity of individual circumstances, caseworkers have to interpret these rules within a given situation and thereby shape the actual meaning of activation for the individual client. In this regard, it is instructive to look at how the concept of "suitable employment" is used in practice. From a legal point of view, any employment in line with legal requirements and other relevant provisions is defined as "suitable". There is no protection for occupational status in a strict sense as the definition of suitable employment does not take individual skills and qualifications into account. There is, however, an income-related protection of status, meaning that during the first months of unemployment an employment is not deemed suitable if the wage is considerably lower than the previous income. After six months of unemployment, claimants must accept a wage equal to the benefit level. Within the scope of suitable employment, welfare recipients must also provide evidence that they have actually applied for job opportunities passed on by the labor administration.
In practice, however, caseworkers handle the concept of suitable employment in a more differentiated way. A caseworker talked about a client's obligation to apply for a placement offer in the following way: "Generally, you have the obligation to respond to placement offers.…But when someone is, let's say, in his mid-50s and would have to apply to H&M, then I can understand that it is just not suitable. Of course, you could insist on it and say the job is reasonable and acceptable, so he had to apply for it. But I think it's more humane to say it's fine if you don't respond to that one, but please apply for other jobs."

(Interview 76)
This example illustrates the leeway that caseworkers have in enforcing rules. It suggests that when applying legal concepts also, caseworkers create their own balance between the supporting and regulating elements of social policy in general and activation policies in particular. Caseworkers' narratives put the supporting elements of their work in the foreground. They refer to a self-perception of service providers who assist labor-market re-integration on the basis of job counseling. The interpretation of the concept of suitable employment takes into account individual qualifications and preferences rather than relying solely on what is defined as "suitable" from a legal point of view.
That caseworkers emphasize the supporting elements of their work possibly reflects the traditional principles of social rights and status protection. At the same time, all quotes show a clear reference to re-employment and the jobseekers' willingness to work as a pre-condition for support, thereby reflecting normative principles of the activation paradigm.
Although caseworkers' narratives place supporting elements in the foreground, caseworkers also consider disciplinary and regulating elements as an equal part of their work. When they describe frontline work, they stress in particular the disciplinary function of such instruments. Implementing and enforcing rules against welfare recipients is, however, described in a rather distanced way. For example, caseworkers illustrated that they may turn to the legal framework to communicate, rationalize and justify their actions. This was illustrated by a caseworker who explained:

"But when I have clients who say 'No' from the beginning, who have little motivation, then I have to go more in the direction of the legal [he laughs] framework and say, look, this is what the legislator says.
This is where my discretion ends." (Interview 42)

| Encounters between caseworkers and clients: The importance of information
Both supportive and disciplinary aspects of activation policies become concrete in personal encounters between caseworkers and their unemployed clients. Since pre-established general standards and processes became more important during the organizational reforms, the assessment of clients with regard to providing the appropriate type of activation service has become partly standardized. In the first interview with clients, caseworkers routinely apply a computer-assisted assessment of skills, competences, and individual needs for action, choosing from a given set of categories (Hielscher & Ochs, 2009, pp. 54-62). Based on this data, clients are categorized into different groups reflecting their level of employability. The assessment concludes with an "integration agreement", specifying integration objectives, the duties of the client as well as assistance to be provided by the labor administration (Weinbach, 2012).
Follow-up interviews are meant to monitor job-search activities and possibly re-align activation strategies.
The organizational routines which structure the caseworker-client interactions are part of the social construction of clients which Lipsky describes as a fundamental mechanism in the relations with clients (Lipsky, 2010(Lipsky, [1980, pp. 59-70). These routines can be considered as tools for processing people into clients, as they assign the unemployed who register with the employment agency the role of a client and classify the individuals into bureaucratic categories. Using a specific software, caseworkers can subsequently match jobseekers' profiles and job offers technically to identify both suitable jobs for unemployed clients and suitable candidates for open positions.
These routines which form part of the placement process support the drawing of a demarcation line and assigning jobseekers the role of clients. Nonetheless, using them requires a minimum of co-production between caseworkers and clients. To conduct the assessment and for the subsequent processes, eliciting information from the client is therefore essential for frontline workers. Obviously, caseworkers need to be informed about the formal skills and qualifications of their clients. At the same time it is important that they find out about individual circumstances which may have an impact on labor-market integration, such as care responsibilities, health conditions, work-related attitudes, and factors that may interfere with employment. Moreover, caseworkers explained that knowing about the individual situation, motivation, and interests of their clients was helpful to them since this type of information enabled them to search for suitable job offers.
Because using standardized tools such as the IT-based assessment and matching procedures requires co-production, caseworkers consider appropriate communication skills to be an important resource of frontline work. Some interviewees, for instance, emphasized the importance of asking an open question or of being patient during longer breaks in the conversation in order to prompt the client to respond. Another caseworker described the relation with his unemployed clients as a sort of strategy game: "You have to [he laughs]-I don't want to say play-with the client, but you need to tease something out of him.…So that you can draw a picture." (Interview 39) The quote also points to the fact that co-production in collecting information requires the caseworker to relate to the client. In order to do so, interviewees explained that establishing a level of trust is of high relevance for them. In particular with regard to the first interview with jobseekers, caseworkers stressed the importance of establishing a trust-based relationship. Given their need to collect information within a restricted timeframe, they presented trust-based relationships as a strategy to encourage the client to open up and to get him or her to speak. This, in turn, is a precondition for assessing the client. One caseworker summarized it like this: "Counselling can only start when I have a picture of the client in my mind, and when the client has opened up a bit. Establishing that level of trust is very, very important. You know, a client doesn't tell you everything right from the beginning." (Interview 87)

| Persuasive strategies in activation delivery
Next to eliciting information from the client, ensuring that clients act according to the rules of the institutions is another crucial function of personal encounters between caseworkers and the unemployed. Through personal meetings with clients, caseworkers are expected to ensure that clients meet the terms and conditions set out by the labor administration. If clients do not keep these appointments, fail to prove job-search activities as agreed in the integration agreement, or reject work or training programs offered by the employment agency, financial sanctions may be imposed.
Some interviewees in our study referred to coercive instruments as a strategy to motivate client commitment and compliance. One caseworker emphasized the importance of disciplinary elements such as a two-week job-application training for her relations with clients: "I know that these programs are not necessarily 'chic'. But they're supposed to be disciplinary. And for that purpose we need them" (Interview 73). However, our interviews consistently showed that more "persuasive" approaches were equally important in frontline work. When describing their interaction with clients, caseworkers laid strong emphasis on the importance of getting through to clients and establishing a rapport with them. In particular, relating to clients in a trustful manner was viewed as a way to literally "move" and influence the client. As one frontline worker put it: "If I get their trust...then we've won. I always say we're all in the same boat, and we're rowing in the same direction. That's the important thing. If I don't get to that point across, then the boat won't move." (Interview 39) In this quotation, relating to clients and building trust with them is shown to be a strategy for overcoming possible barriers to co-operation. Trust-building is presented as a mechanism to encourage clients to take an active part in placement work and to co-operate with the caseworker in order to achieve labor-market integration.
Compliance is thus ensured in a persuasive rather than a coercive way.
Relational aspects of frontline work-the need to get through to the client-were emphasized by many frontline workers in our study. These references reflect that worker-client relations in activation policies are not only a vehicle to assess clients and monitor client progress, but also in itself "a central feature of the workers' intervention tool kit" (Hasenfeld, 2009, p. 407). In order to reach the overall objective of labor-market participation which is-as outlined before-influenced by a variety of factors beyond the immediate control of the caseworker, the personal encounter is aimed at bringing about behavioral change. Our interviews and group discussions suggest that coercive instrument fulfill a rather strategic function within that process. In many cases, frontline workers seem to rely primarily on a persuasive approach, based on trust-building. In view of the fact that realizing organizational performance targets such as an agreed rate of integration depends, amongst other things, on jobseekers' co-operation, caseworkers may consider this strategy as functional in many situations.

| Relations with employers
Maintaining good relations with employers is an important (although less studied) dimension of activating labor-market policies. In the course of the Hartz reforms, the German labor administration increased its employer orientation (Sowa, Reims, & Theuer, 2015). Specific teams providing employer-oriented services were established, the so-called Employers Service (Arbeitgeberservice). This was accompanied by a differentiation of tasks: placement officers were now either specialized in working with unemployed persons or in carrying out employer-oriented placement work. Employer-oriented placement officers were responsible for registering job vacancies and for finding suitable candidates for these job offers; and more generally, they were supposed to maintain contacts with employers and to act as their contact persons for all sorts of concerns. Employer-oriented placement officers typically describe networks with employers as particularly helpful in placing unemployed persons with poor job prospects in a job ("to somehow get in those who might not be first-class goods any more" [Group discussion 34]).
The structural position of employers differs from the typical worker-client relationship in street-level bureaucracies. For employers, co-operation with the public labor administration is only one of several alternatives in recruiting staff, and they are largely free in their decision whether to interact with the labor administration or whether to rely exclusively on other means such as job advertisements in newspapers or online recruitment. In the words of an employer-oriented placement officer: "We work with employers, and they use our service on a voluntary basis of course.
There are no legal consequences for the employer" (Interview 40).
In a similar way as their colleagues who work with the unemployed, employer-oriented placement officers stress the relational element of their work, while obviously these relations are far more stable and long-term in nature.
Establishing permanent, personal relationships with employers is thus regarded as a crucial aspect of employer-oriented placement work. Also in interviews with employer-oriented placement officers, trust was an important reference point. One employer-oriented placement officer expressed the view that counseling aspects played an increasingly important role in relations with employers and emphasized the importance of trust in this regard:  ' ." (Interview 40) Establishing relationships of trust are presented in this quotation as a strategy to encourage the employer to "open up" and as a way of obtaining essential information. Similarly as with the caseworker-client relation, trust is presented as a mechanism to overcome barriers in information exchange and eventually support the caseworker in achieving the objectives of activation policy.

| Frontline workers as "trust-takers"
Our findings suggest that building a rapport with clients is an important resource in frontline work. In order to elicit information from the client and to motivate client compliance and co-operation, strategies relying on persuasion seem to be of great relevance to caseworkers. In this regard, trust becomes a particularly meaningful dimension of caseworker-client relationships. Establishing a level of trust is-from the perspective of caseworkers-a crucial objective of personal encounters with clients.
The interviews in our study did not address trust explicitly. Instead, the notion was brought up by the interviewees themselves, which stresses its subjective relevance for the caseworkers interviewed. A closer look at the situation-specific meaning of trust revealed that caseworkers referred to trust in a rather one-sided way: typically, they depicted themselves as trust-takers, while unemployed clients as well as employers were addressed as trust-givers. Our interviews suggested that the caseworkers' role-definition included the typical attributes of the trust-taker in trust-based relationships (see Beckert, 2005): caseworkers describe strategies for setting up circumstances that could promote trust, such as creating a good atmosphere at the beginning of an interview, addressing clients in a personal, individual way, or leaving them enough time to express their concerns. An experienced caseworker made this point when he talked about his methods of introducing younger colleagues to the tasks of the job: This quotation clearly illustrates that clients are expected to become trust-givers in the caseworker-client relationship. The role of the caseworker, on the other hand, implies signaling that she or he deserves to be trusted.
This involves making the client feel comfortable and reducing anxiety. These actions intend to produce the appearance of trustworthiness, and, as a consequence, to overcome barriers of co-operation in the caseworker-client interaction. When the client has been successfully motivated to become a trust-giver, the frontline worker can expect her or him to act according to the rules of the institution.
As mentioned above, producing trust may be a specific strategy to reach the aims of activation policies. On the other hand, producing trust may also be a rational strategy with regard to the frontline worker's position in his relation with clients. Even though there is a structural power asymmetry between street-level bureaucracies and their clients, individual encounters between caseworkers and their clients are likewise shaped by the capabilities of the clients (Ludwig-Mayerhofer, Behrend, & Sondermann, 2014). Moreover, research on sanctioning practices within the German basic income support scheme has shown that many caseworkers tend to avoid sanctions and prefer to rely on practices of non-sanctioning in interactions with welfare recipients (Karl, Müller, & Wolff, 2011). The authors conclude that imposing sanctions may restrict the caseworkers' scope for action within the caseworker-client relation (Karl et al., 2011, p. 124). Against this background, producing trust could be a more functional strategy for the frontline worker in order to control the situation. Given that caseworkers refer to disciplining elements in a fairly distanced way, they may consider structuring the interaction in a manner that promotes trustworthiness as a more effective way to retain their authority and to remain in the more powerful position.

| CONCLUSIONS
This article has examined the role of caseworkers as intermediaries of the welfare state in the context of the German unemployment insurance scheme. The analysis of qualitative interviews and group discussions with frontline workers has illustrated how caseworkers act as "policy mediators" and shape the actual meaning of activation in practice. Our findings strengthen the argument made by other scholars that activation is to be understood as a fairly vague concept which encompasses a variety of different forms in the day-to-day processes of policy implementation (Brodkin, 2011;Tabin & Perriard, 2016). Activation thus involves a mix of supportive and coercive elements, and frontline workers have to combine and reconcile these different elements and roles in their interaction with clients (Sainsbury, 2008).
Research across different national contexts has stressed this multifaceted character of frontline work and the demands for reconciling disciplining and enabling aspects of work (see, for a summary, Caswell et al., 2017). When it comes to the concrete practices of activation, coercive elements are well-known methods of exerting greater control over client behavior and of bringing it into line with requirements. These coercive elements of activation policies have received broad attention in the corresponding research. In particular, sanctioning practices have been a focal point in welfare-state research, both in the German (e.g., Grüttner, Moczall, & Wolff, 2016;Zahradnik et al., 2016) as well as in other institutional contexts (Caswell & Høybye-Mortensen, 2015;Soss, Fording, & Schram, 2011). Studies that focus on support and counseling, on the other hand, are far less frequent and knowledge is still patchy (Caswell et al., 2017, p. 182).
The results of our study indicate that practices of activation are more complex than the dichotomy between disciplining and enabling elements suggests. Our focus on worker-client relations revealed that what we termed a "persuasive" approach of activation plays a prominent role in frontline interactions. Our findings highlight that achieving client co-operation and compliance is of utmost importance for workers on the frontline, given that activation delivery requires co-production. When assessing and monitoring clients and matching them with available programs and strategies to assist labor-market re-integration, frontline workers depend to a great extent on the information provided by the client. Moreover, frontline workers can influence labor-market integration only to a limited extent.
Relating to clients can therefore be regarded as a means to influence and control the interaction. Caseworkers in our study emphasized the importance of establishing a rapport with clients in order to overcome possible barriers to information exchange and ease co-operation in the caseworker-client interaction.
Even though our study is surely explorative in nature, our findings suggest that it is important to pay more attention to the emotional dimension of frontline work. Recently, scholars from various disciplines have pointed to the role of emotions in policy-making and public service delivery (e.g., Graham, 2002;Hunter, 2015;Jupp, Pykett, & Smith, 2016). With regard to the transformation of the German welfare state, it has been argued that welfare reforms have engendered emotional states of fear, and that these feelings of anxiety, in turn, enforce compliance with the expectations of increased individual responsibilities (Betzelt & Bode, 2017). Engendering and mobilizing emotions has also been shown to be of growing importance in welfare organizations. With the transformation of state bureaucracies into customer-oriented service providers, these organizations increasingly expect welfare workers to guide, motivate, and monitor clients by using affective means (Penz et al., 2017). In such a context, addressing jobseekers as partners, showing empathy and understanding for them, and establishing relationships of trust are key strategies to motivate and activate clients (Penz et al., 2017).
In our study, frontline workers in German employment agencies referred to the term of "trust" to characterize these relational aspects of frontline work. In doing so, caseworkers typically identified with the ideal-typical role of trust-takers. They described strategies to motivate clients to become trust-givers in the caseworker-client relationship. As we have shown, strategies to produce the impression of trustworthiness are effectively integrated into frontline work. Frontline workers rely on these strategies to motivate clients to co-operate and to comply with the rules of activation policies. From a frontline perspective, establishing relationships of trust may thus not only be an instrument of activation, but may also eventually help frontline workers to control the work situation and to retain their authority.
Due to the explorative character of our study, it is not possible to generalize these findings. The importance of emotional work and the relevance of persuasive approaches to activation delivery may possibly be linked to the specific normative tradition of a continental welfare state where activation requirements are generally less pronounced than in means-tested welfare systems. However, the issue of control is a fundamental mechanism of worker-client relations in street-level bureaucracies (Lipsky, 2000(Lipsky, [1980, in particular pp. 57-59). We may therefore hypothesize that the relational dimension of frontline work and strategies to turn clients into "trust-givers" may also play a role in other institutional and national settings. Further research should therefore examine in more detail the affective dimension of frontline work and address the social and organizational mechanism of relationships of trust.