ABSTRACT. Street-level research and theory illuminate the opaque organizational spaces where policies and politics are informally contested. This positions SLOs as important domains in which the promise of democracy — and its perils — take shape. In this period of democratic reckoning, how can research help us understand the mediating role of SLOs in the relationship between individuals and the state?

ABSTRACT. Research in public administration highlights citizens’ trust in frontline workers as crucial for effective service delivery. This trust is often understood as professional trust, based on expertise rather than personal characteristics. However, other scholars stress the importance of personal elements, arguing that trust also depends on citizens’ experiences with frontline workers as individuals. Accordingly, trust has been described as a hybrid phenomenon combining professional and personal dimensions, though the relationship between these remains poorly understood.

This paper therefore asks: “What is the relationship between professional and personal citizen trust?”. It draws on two qualitative studies conducted in distinct empirical contexts that differ in levels of professional expertise, legitimate complexity, and the intensity and frequency of interactions between citizens and frontline workers. This design enables an in-depth examination of the relationship between professional and personal trust and strengthens the robustness of the findings. Study A is an ethnographic investigation of interactions between childcare professionals and parents in Danish daycare centres, drawing in part on 38 interviews with parents. Study B consists of 20 narrative interviews with patients and relatives concerning their experiences with health professionals, including medical doctors, during the COVID-19 pandemic.

Based on an abductive logic of inquiry, the paper develops a core theoretical argument that personal trust shapes professional trust in three ways — reinforcing, reducing, or replacing it — while professional trust appears to have little impact on personal trust. We thus theorize and empirically demonstrate that personal trust can strengthen professional trust, that a lack of personal trust can undermine professional trust, and that personal trust may fully substitute for professional trust. These findings have important implications for how citizens develop trust in frontline interactions, and consequently how frontline workers can foster and sustain that trust — an essential prerequisite for the effective delivery of public services.

ABSTRACT. ***The paper will be presented by Sohad Amaria and Einat Lavee together***

Research on citizen–state encounters has consistently documented predominantly negative emotional experiences in interactions with street-level bureaucrats (SLBs). Existing explanations tend to attribute these emotions primarily to the interaction itself: its tone, power dynamics, and the attitudes of bureaucrats. Yet this literature raises an important puzzle: how is it that similar emotional outcomes, most often negative, are repeatedly observed even when institutional encounters themselves differ substantially? This study revisits emotional experiences in street-level encounters and addresses this puzzle by examining how they are produced. Drawing on institutional ethnography, the study is based on 27 in-depth interviews with low-income Arab single mothers in Israel. The analysis examines how emotional experiences are shaped through their encounters with social workers. The analysis first revealed that the institutional encounter itself differs markedly across sub-categories of single mothers: Widowed mothers often described supportive and proactive encounters, whereas divorced mothers reported stigmatizing and conflictual interactions. Separated mothers referred less to the interaction itself and more to shifting power relations and a loss of control. However, despite these substantial differences in the nature of the encounter, an overarching emotional outcome of exclusion emerged across groups. This convergence was deciphered through additional elements identified in the analysis, particularly the gap between mothers’ needs and the support they actually received. We found that across all sub-categories, unmet expectations and limited practical responses played a central role in shaping emotional experiences, regardless of whether the interaction itself was perceived as positive or negative.

This findings demonstrate that emotional experiences in encounters with SLBs cannot be explained by the interaction alone. Rather, they are produced through the interplay between interactional dynamics and the practical outcomes of support. The study contributes to street-level bureaucracy research by conceptualizing emotions as institutional outcomes of everyday encounters rather than merely individual reactions.

ABSTRACT. Street-level bureaucracy research increasingly pays attention to citizens' agency in encounters with frontline workers. However, there is still limited empirical insight into how institutional context, specific public service domains, and socioeconomic and migrant backgrounds shape coping strategies. This paper begins to address this gap by examining older people's coping strategies in a marketized long-term care system. Drawing on 13 focus groups with 70 older adults in the Netherlands, sampled across socioeconomic and migration groups, we analyse how older people make sense of and position themselves vis-à-vis the welfare state and care organisations, and translate these interpretations into strategies in encounters with frontline workers.

Our findings show that institutional distrust in the long-term care system is widespread among older people in the Netherlands. This distrust is driven by institutional fragmentation, which generates heavy administrative burdens, and by perceptions that profit motives among care providers and weak regulation undermine service quality. While institutional distrust was shared, coping strategies varied across groups. Older adults with higher socioeconomic resources act as active consumers, voicing complaints and/or exiting partially or entirely by paying for private care. Older adults with lower socioeconomic resources act as resigned receivers. When publicly funded services fail to meet their needs, they choose silence over complaint to avoid jeopardizing access to what they perceive as a scarce resource. Older migrants similarly use silence, but additionally rely on intermediaries such as migrant-specific care organisations to cope with distrust and/or administrative burdens.

This paper contributes to street-level bureaucracy research by examining how the institutional context of marketized long-term care is associated with institutional distrust and how this distrust shapes coping strategies in frontline interactions. By analysing variation in coping strategies across socioeconomic positions and migration backgrounds, it advances understanding of unequal citizen agency within a shared institutional setting.

ABSTRACT. For refugees, displacement and unemployment constitute a complex and often time prolonged state of being in-between, which has been described as liminality in scholarly debates (Boland and Griffin, 2015; Alkhaled and Sasaki, 2022; Marotta, 2025). We can expect that unemployed refugees were uprooted from various roles – such as jobholder, organizational member but also neighbor or home owner – without having new roles to identify with (Daskalaki and Simosi, 2018). Based on 35 qualitative, longitudinal interviews with unemployed Ukrainians in Germany, this paper traces processes of identity work under liminality. I find that, despite the multi-layered uncertainties individuals are confronted with, they stabilize their lives by reactivating well known strategies of survival (enacting prior identities) but also by developing new routines, new forms of (collective) engagement and alternative visions for their future selves (enacting new identities). During this process of stabilization, our participants encounter the welfare state and report about various frictions, emerging from e.g. contradictory signals from different layers of the welfare state, long waiting times, or experiences of repeated failure. While some of these frictions trigger stagnation and social withdrawal, others turn out to be productive. This gets most visible when resources developed in the private stabilization process are activated to cope with the frictions. By linking refugees’ experiences with the public administration to their everyday practices of rebuilding stability, the findings advance research on street-level bureaucracy from the client perspective.

ABSTRACT. Across Europe, the demand for social assistance is rising, pressuring welfare agencies to process cases more efficiently. In response, welfare organisations are increasingly using AI to support eligibility assessments. While AI-tools show promise in accelerating eligibility assessments, they might also pose risks to the quality of the decision-making.

Legally, the Principle of Due Care constitutes a minimal, legal quality standard for discretionary administrative decision-making. The principle covers different aspects of the decision-making process, including 1) collecting information at the preparatory phase, 2) weighing different interests when applying a legal norm, and 3) notifying citizens once a decision has been made. However, it remains unclear how the principle should be, and is applied in a decision-making process that is being transformed by AI-tools.

Therefore, this paper examines both the normative and empirical application of the Principle of Due Care within AI-assisted eligibility assessments for social assistance. It does so by means of a qualitative case study of a Belgian welfare agency, which co-developed an AI tool to draft social reports. The study covers interviews with the different actors influencing welfare policies, including private software developers, local politicians, and street-level bureaucrats. Preliminary findings reveal how different actors decide about which facts are relevant to decide upon welfare eligibility, and about how legal norms should be interpreted, disentangling who gets to decide about welfare eligibility in the digital welfare state.

ABSTRACT. AI-mediated documentation technologies are being rapidly introduced across municipal services, fundamentally transforming how frontline workers document citizen encounters. The Danish government's 266.7 million kroner investment in AI-supported solutions (Digitaliseringsministeriet, 2025) visions to free caseworkers from administrative burden, building on the ambition "more time for what's important" (Taskforce for Kunstig Intelligens, 2025, p. 5). Still, we know little about how this transformation affects the frontline workers’ discretion, decision-making and the meeting with and representation of citizens.

My PhD project examines how AI-mediated documentation reconfigures practice in a Danish municipality's children, family and educational department. When conversations are recorded and AI generates transcripts and summaries, frontline workers no longer control and decide what gets documented but rather engage in editing algorithmic outputs. These technologies actively participate in creating the documents that are the basis for welfare decisions. This shift represents a fundamental reconfiguration of documentation practice and professional discretion, and it raises critical questions: What happens to professional judgment? How do frontline workers navigate responsibility for documents they edit but did not create? What consequences does this have for citizen encounters and representation in governmental systems?

Taking frontline workers' perspective as a starting point, this ethnographic study develops an empirically grounded understanding of how AI-mediated documentation shapes daily casework. Through observing documentation practices before and after AI introduction, shadowing 3-5 caseworkers, conducting semi-structured interviews, and analyzing AI-generated outputs alongside final documents, my approach captures the transformation as it unfolds, revealing how technological change affects professional discretion, coping strategies, and the fundamental nature of street-level work in the emerging AI-augmented welfare state.

ABSTRACT. Artificial intelligence is increasingly deployed in public administration, yet its implications for street-level bureaucracies remain conceptually underspecified. Existing research tends to either overstate AI's transformative potential or treat its effects as uniform across administrative contexts. Drawing on secondary analysis of research conducted across German street-level bureaucracies and European local governments between 2020 and 2026 as well as scholarly literature, this paper challenges both assumptions by developing two contrasting theoretical perspectives: the continuation thesis and the reconfiguration thesis. Drawing on street-level bureaucracy theory and comparative public administration research, we conceptualize AI as a new institutional layer that interacts with established governance traditions and administrative practices. We argue that national and organizational starting conditions, particularly administrative cultures ranging from legalistic rule-of-law traditions to cooperative and pragmatic public-interest orientations, shape how AI is introduced, absorbed, and enacted at the frontline. The continuation thesis holds that AI becomes integrated into existing multilevel governance arrangements, reinforcing dominant administrative logics and preserving core street-level practices—especially where physical interaction and relational work remain irreplaceable. The reconfiguration thesis, by contrast, posits that AI may enable recentralization of authority and control, shift public value logics toward efficiency and standardization, curtail frontline discretion, and generate new forms of compensatory and invisible work that fundamentally alter street-level roles. We unpack both perspectives through nine propositions organized at the macro-level (task allocation and public value logics) and street-level (discretion, agency, and coping practices). We conclude by outlining a comparative research agenda to empirically test these propositions across different administrative cultures, policy fields, and levels of government.

ABSTRACT. The integration of information and communication technologies into client-professional interactions is rapidly expanding across social work and public service settings. While these technologies offer new opportunities for accessibility and efficiency in service delivery, they also reconfigure how relationships are enacted when physical co-presence is no longer shared. Drawing on peer-reviewed research, this integrative and generative review examines how relational engagement is understood, articulated and practiced in synchronous digitally mediated client–professional encounters across social work and related fields. A total of 3,545 articles were identified through a comprehensive search across social science databases, of which 40 met the inclusion criteria and were included in the analysis. The findings show that relational engagement is addressed through a range of concepts, including social presence, digital empathy and digital intimacy, each capturing different aspects of relational work in digital settings. Building on this, the article develops an integrative conceptualization of relational engagement as a qualitative aspect of client–professional relationships that emerges through the dynamic interplay of communicative, temporal, spatial and institutional practices. Within this perspective, technologies are not treated as neutral channels but as active participants in the co-constitution of relational processes. By conceptualizing relational engagement as an emergent and situated accomplishment within socio-techno-material practices, the paper offers a framework for understanding the in-situ accomplishment of client/professional relationships.

ABSTRACT. Digital technologies that facilitate telemonitoring are increasingly being implemented to combat current healthcare challenges. Telemonitoring fundamentally alters established clinical practices among frontline healthcare professionals, shifting care from in-person to remote settings. For instance, some scholars argue that telemonitoring primarily challenges healthcare professionals' work (Dyb & Kvam, 2022), whereas others contend that it augments or relieves professionals of certain tasks (Van Grootven et al., 2022).

Although many scholars recognize the significance of healthcare professionals' adoption of telemonitoring, as they function as frontline implementers (Lipsky, 1980; Bagot et al., 2019). Only a few studies examine how professionals adapt to telemonitoring-induced changes in their work (Grossi et al., 2021). As such, less is known about adaptive performance, defined as the adjustments in behaviour that professionals make in response to changes in the work environment (Pulakos et al., 2000; Charbonnier‐Voirin & Roussel, 2012; Krijgsheld et al., 2024). Focusing on frontline healthcare professionals as they adapt to telemonitoring in clinical practice can help identify theoretical explanations for the gap between the promises of digital technologies and daily practice.

We therefore conduct an ethnographic study in a Dutch regional virtual care centre, comprising interviews and observations. We assess the adaptive performance of frontline healthcare professionals in response to changes in their work enabled by digital technologies. Our results show that frontline healthcare professionals rely on creative discretionary practices, such as selecting which technologies to use, as they cope with the effects of managing multiple telemonitoring applications, which increase administrative burden. In addition, the results show the role of frontline professionals as key mediators in realizing the promises of digital technologies through their adaptive performance.

ABSTRACT. As data-driven automated decision-making (ADM) systems are increasingly introduced into Swedish social services, questions arise about how professional social work is represented in the policy discourse surrounding their adoption. Drawing on national policy documents and an analytical framework informed by the What’s the Problem Represented to be? (WPR) approach, this article examines how professional knowledge, conduct, and judgment are articulated in relation to AI-assisted decision-making in the Swedish Personal Social Services (PSS). The analysis identifies three interrelated patterns: a repositioning of the social worker from professional decision-maker to administrative sign-off; a vague and supervisory construction of the competence required for AI-enabled practice; and a treatment of accumulated case file data as neutral and adequate knowledge for training data-driven AI systems. Taken together, these patterns point to a deeper yet largely unexamined epistemological tension that runs throughout the policy discourse.

ABSTRACT. The introduction of digital profiling tools in Active Labour Market Policies (ALMPs) promises standardisation, neutrality, and uniformity in decision-making. Ethnographic observation in Italian Public Employment Services shows that such tools do not eliminate discretion but transform it. This paper examines how discretion is exercised in the use of the GOL (Workers’ Employability Guarantee) profiling tool across two contexts: Lombardy and Campania. Based on over 100 days of ethnographic fieldwork and 400 observed profiling encounters, the analysis shows that discretion assumes context-sensitive forms shaped by organisational settings and labour markets. In Lombardy, where the labour market is relatively dynamic and most users receive standard unemployment benefits, discretion often consists of procedural adjustments that align assessments with institutional expectations while navigating the profiling interface. Operators engage in protective practices toward benefits recipients, yet the cases they handle are less severe and the economic consequences of conditionality more limited. In Campania – characterised by labour market stagnation, long-term unemployment, and widespread reliance on socio-assistance – discretion takes a relational, compensatory form. SLBs focus on mitigating conditionality effects and supporting users facing socio-economic vulnerabilities. Discretion is enacted through mediation, negotiated compliance and strategic deferral, containing administrative conflict while maintaining compliance. As conditionality expands, frontline autonomy shifts toward relational and procedural interventions rather than rule-bending. Crucially, these differences emerge despite the same national profiling tool and similar pressures linked to the Italian Recovery and Resilience Plan. Digital constraints are thus interpreted, absorbed, and adapted locally, giving rise to forms of situated algorithmic governance. Policy is enacted neither solely at the street level nor solely at the code level, but in interactions among interfaces, frontline workers, and organisational contexts. The findings reveal a productive tension between digital profiling and classic interpretations of street-level bureaucracy, showing that algorithmic governance is co-produced through situated institutional ecologies.

ABSTRACT. Imagine a childcare worker making a home visit to provide social support. Arriving at the building, a woman opens the door: pictures cover the walls, a soft smell of food wafts through the air and the TV is flickering in the background. The social worker looks for a place to sit while talking with the woman. She just explained that her husband will run a few minutes late and offers tea in the meantime. At the same time, another childcare worker from the same office is about to start their day. They walk down the office hallway to get to the first of six meetings. It will be in one of the newly renovated rooms, still smelling of fresh paint. The social worker sits down behind the table, facing the door to see the family arriving. A painting of the city skyline is highlighted by bright lights. The clock slips to 10 a.m.; the case file lies open. Yet, no sign of the family.

While both social workers are working within the same organization, following comparable policy goals and being confronted with seemingly similar situations, their service provision varies distinctively in 'where' they deliver it. In interactions, street-level workers move within different spaces, exposing them to diverse (un)structured settings of public and private spheres. Arguing that knowledge mobilizes and transforms along with street-level workers’ spatial movement, this paper discusses how street-level workers actually interpret spaces and how space shapes interactions with citizens. We look at situated practices of Dutch municipal care workers of a neighbourhood-team in Rotterdam, to ethnographically explore how social workers mobilize professional knowledge across different spaces of interaction. This way, we aim to unravel how spatial and situated knowledge are interpreted in everyday practices, unfolding how street-level workers make sense of local neighbourhood complexities in service delivery.

ABSTRACT. Across Europe, the mandates of social work professionals are becoming progressively entangled with policing and criminal justice systems. This comprehensive criminalization of social policy can translate into formal legal procedures, but also be seen as an alignment of policy initiatives across domains. In some welfare areas, there are (new) formalised obligations to report cases to the police or to the justice system, such as in cases of welfare fraud, illegalized migrants and child abuse or neglect. In others, there are traces of less formalized, more opaque entanglements between social work institutions and policing entities that can go unnoticed to participants. These include interagency data sharing and data matching, and the use of risk prediction technologies.

In this paper, we address the expanding interfaces between social work and policing and their impact on the everyday practices of frontline workers, presenting preliminary findings from the “Social Work as Police Work?” project. The paper draws on a comparative analysis of laws and protocols governing interagency collaboration between social services and policing entities in Ireland, Sweden, and Switzerland, as well as preliminary interviews with professionals. The paper aims to conceptualize different forms of interagency collaboration and reporting and set out their implications for frontline workers’ discretion, decision-making, and knowledge mobilization. While interagency collaboration has been a long-standing feature of governance, research has tended to focus on public goods provision and coordination around ‘wicked problems’ such as sustainability or digitalization. Our case, by contrast, concerns collaboration between frontline organizations to marshal state capacity for law enforcement and crime prevention. In this context, we ask how frontline workers, as important decision-makers, interpret, negotiate, and resist different forms of (non-)knowledge to navigate their everyday practice.

ABSTRACT. The implementation of data protection regulations in hospital relies on a wide range of frontline and managerial actors whose everyday decisions shape how sensitive health data are handled for care and research purposes. While prior street-level bureaucracy studies typically focus on a single professional group within one national system, this study compares how different forms of knowledge interact in the implementation of data protection rules in Swiss and French hospitals. Because data protection is a shared responsibility, every category of actors involved in the data lifecycle (medical, technical, and legal) must contribute to safeguarding personal data, particularly in healthcare settings where sensitive data require heightened precautions. Understanding how each professional group perceives data protection and mobilizes knowledge to navigate complex situations is therefore essential. Hospitals offer a particularly rich environment to explore these dynamics, as multiple professional groups intervene in the collection, processing, storage and sharing of patient data, often with overlapping or competing priorities. Drawing on 48 semi structured interviews conducted between November 2024 and December 2025 with medical, technical and legal staff in hospitals and healthcare centres in Switzerland and France, the study uses a thematic analysis to examine how professionals combine, prioritize and negotiate different types of knowledge when confronted with ambiguous or contradictory objectives related to data protection. By adopting a knowledge-centered perspective, the paper sheds light on the micro level mechanisms through which data protection regulations are translated into practice within multi-professional hospital environments in different national settings.

ABSTRACT. Networking with external partners in delivering policies is an essential requirement for street-level bureaucrats (SLBs) in times of network governance, organizational fragmentation, and interdependence. In the case of employment services, such partners include, for example, language training providers, employers, or social counseling organizations. This article investigates the factors that determine the extent and characteristics of external networking at the frontline. Based on a representative survey of frontline workers in German job centers, we examine antecedents of different forms of external networking. We find strong evidence that complex clients’ needs, fewer resource constraints, a network-friendly professional identity contribute to front level networking. The observed correlations hold across all our network indicators: the diversity of external contacts, participation in formal networks, and support for clients in taking up third-party support. By contrast, factors such as work experience, part-time work, or local labor market conditions do not influence SLBs networking.

ABSTRACT. The study of public encounters generates new and innovative insight into citizen-State relationships. Its rapid growth as a sub-field of public administration demonstrates that the relational dynamics in-between and around citizens and agents of the State shape experiences, power, and outcomes. Studies of public encounters have mainly focused on what is said and how in interactions and to what effect. We extend this emerging relational understanding of public encounters by analysing the fundamental yet overlooked role of sensory experiences of how sound is co-produced by and around citizens and agents of the State.

This paper offers original empirical and conceptual insight into how sound shapes public encounters. We conceptualise the soundscape of public encounters as an embodied experience that is co-produced over space by sonic agency. Based on an ethnographic study of Swedish prisons, we examine how sound impacts relational dynamics of co-production and conflict. Prisons offer a unique context because their spatial confinement makes sound unavoidable and key to how prisoners and prison officers engage in dynamics of co-production and conflict.

Our findings are threefold. First, sound forms an arena for interaction by oscillating between built silence for punishment, control, and moral introspection on the one hand and the production of noise for coercive and social communication. Second, dynamics of conflict emerge from the use of listening for control and sound as an appropriation tactic. Third, dynamics of co-production emerge from experiencing spaces of therapeutic sound and silence. By adding a sonic layer to the relational nature of public encounters, the paper contributes to moving beyond “language-as-word” and developing a more holistic understanding of how citizen-State relationships are co-produced.

ABSTRACT. While securitisation policies often respond to real security threats, their expansion raises concerns about the narrowing of democratic space and critical debate (Virta & Branders, 2016). At the same time, participatory approaches to security governance face difficulties in operationalising active citizenship and transparency. From the perspective of the social groups to be “secured,” however, security emerges as the result of multiple, seemingly disconnected and uncoordinated interactions. This article develops a conceptual framework for understanding how perceptions of security are co-produced at the street level through the combined actions of state organisations, civil society actors, and citizens. Building on street-level bureaucracy and co-production scholarship, the framework shifts attention from security as an institutional output to security perception as a socially embedded outcome shaped in everyday encounters (Brodkin, 2012; Jaspers & Tuurnas, 2023). Street-level actors such as police officers and social workers are treated as key mediators through whom the state becomes experientially present and security is interpreted in practice (Brodkin, 2012; Lotta & Kirschbaum, 2021). At the same time, third-sector, religious, private, and community actors are conceptualised as co-producers of local security arrangements and complementary shapers of security perceptions (Mosselson, 2019; van Eijk et al., 2018). The framework links this multi-actor perspective to social sustainability, approached through equity, inclusion, trust, and civic engagement (Eizenberg & Jabareen, 2017). Together, these elements connect street-level interaction and co-production systems within a single analytical lens. The framework enables analysis of whether security arrangements recognise diverse social groups and contribute to a broadly shared sense of security, or instead shift responsibility toward individuals in ways that erode collective accountability (Gearon, 2019). Particular attention is given to tensions between institutional security logics and citizens’ lived experiences, including conflicts between groups’ security perspectives (Alves, 2014).

ABSTRACT. Vulnerable groups often face intensified administrative burdens due to limited "administrative literacy", the ability to navigate complex bureaucratic systems, understand official jargon and fulfil procedural demands (Döring, 2021). When these burdens are too high, individuals risk administrative exclusion, effectively losing access to rights and services, thereby deepening marginalisation (Brodkin & Majmundar, 2010).

This paper examines how individuals experiencing housing exclusion (PEHE) in Belgium engage with public services, focusing on their everyday encounters with street-level bureaucracy across multiple institutions. It shifts the lens from frontline professionals to clients’ lived experiences: how people interpret, endure, and navigate bureaucratic demands over time.

Based on a two-year qualitative study with 20 adults facing housing exclusion, the research combines repeated in-depth interviews with go-along methods, following participants’ interactions with (inter alia) local welfare offices, housing agencies, shelters, and NGOs. The analysis is guided by three questions: (1) how do PEHE experience administrative burdens across the housing rights chain?; (2) how do they mobilize and develop administrative literacy?; and (3) what role do intermediaries play in mitigating or compounding these burdens?

Drawing on theories of administrative burden, street-level bureaucracy, and welfare conditionality, the paper examines how housing exclusion accumulates gradually, as procedural frictions such as missed appointments, unfiled documents and contradictory rules, compound over time. A single misstep can cascade into broader exclusion (Chudnovsky & Peeters, 2022; Robben, Peeters & Widlak, 2024), turning bureaucratic missteps and technical errors into life-altering consequences. Housing exclusion is approached not as a single point of failure but as a trajectory of citizen-state interactions across settings. Rather than a fixed individual skill, this paper operationalises administrative literacy as a relational and evolving capacity, shaped by institutional tolerance for error, access to supportive intermediaries, and the degree of transparency and continuity in citizen-state encounters.

ABSTRACT. Street-level encounters are experienced first by clients, who must interpret, navigate, and respond to frontline workers’ discretionary practices to secure support. Research on street-level bureaucracy has documented how SLBs’ perceptions of deservingness shape resource allocation. Yet far less is known about how clients perceive these processes and act within them. This study shifts the analytical lens from workers to clients, examining how service users understand and navigate discretion in everyday interactions with frontline providers. Recent literature demonstrated that SLBs’ discretion is exercised not only in distributing formal organizational resources, but also in providing informal resources beyond official mandates. These practices have become increasingly central in contexts of scarcity, where SLBs respond to unmet needs through personal efforts. Focusing on such informal personal resources (IFRs) is therefore critical, as they are shaped by interactional dynamics. The study asks: how do clients perceive the factors shaping SLBs’ provision of IFRs, and how do they adjust their conduct in response? Findings are based on 35 in-depth interviews with clients receiving social services from SLBs in Israel. The analysis shows that clients operate under uncertainty and develop practical expertise in “learning the system”. We identified three justification frames through which deservingness is communicated: alignment with a discourse of personal responsibility, emphasizing cooperation and self-effort; claims of “true deservingness” grounded in narratives of hardship and need; and adherence to interactional codes, including appreciation and cultivation of a personal relationship with the worker. These patterns illuminate how bureaucratic encounters are experienced as relational and strategic from the clients’ standpoint. They also reveal a widening gap between those who succeed in deciphering implicit expectations and accessing informal support, and those who remain without essential resources. By focusing on clients’ strategies towards IFRs, we highlight a novel layer in current understanding of how street-level encounters may reproduce inequality.

ABSTRACT. Qualitative research on street-level bureaucracy commonly relies on interviews and ethnography. However, a growing body of studies complements these methods with visual data, observation, and life histories to address the limitations and challenges associated with interviews and ethnographic approaches. This paper seeks to deepen the debate on the use of visual data, observation, and life stories in studies of street-level bureaucracy, arguing that these approaches are particularly fruitful for capturing citizens’ lived experiences in their encounters with the state. Research in this field has focused on understanding discretion and decisions of street-level bureaucrats. Recent studies have addressed this gap by examining the administrative burdens individuals face when interacting with the state, the ways personal attributes such as race and gender shape interactions and perceptions of access to rights, how clients strategically construct narratives to emphasize particular experiences, and experiences of waiting, uncertainty, and limbo. These studies emphasize how individuals experience, comply with, or contest public action, but the analysis of the relationship between street-level bureaucrats and citizens and the embedding in a citizens´ life trajectories are still relatively limited methodologically. This article contributes to this debate by reviewing and comparatively discussing different methodological approaches that employ visual data, observation, and life stories and also mixed-methods, to analyze citizens’ perspectives, and especially perspectives of very vulnerable people. These methods share an understanding of the social world as dynamic, in which individuals actively interpret and give meaning to events. As such, they are particularly well suited to capturing meanings, representations, frames, and narratives, and to illuminating everyday interactions, spaces, and burdens that shape citizens’ encounters with the state. The paper also reflects on the analytical potential of these methods, as well as on their operational challenges and the ethical dilemmas they raise when researching vulnerable populations and asymmetric power relations.

ABSTRACT. In the field of social care the AI developments may be seen as beneficial in relation to the stronger professions but potentially problematical in other occupations. That judgement is based upon an analysis influenced by Abbott’s objective/subjective distinction (1988).

The objective category is dominant in professions with a strong knowledge base. As Suskind and Suskind show (2021) AI here reinforces open decision making and challenges professional closure.

The subjective category encompasses more intuitive work. Where this is dominant routinisation is being pushed by authorities eager to make personnel savings and indifferent to the human encounters involved. Routinisation is particularly problematical for the needs of people at the beginning and end of life, or when well- being is threatened by poor health and disability.

Exposition of this argument will be explored through an examination of the way professions and professionalism are defined. Arguments in defence of professions primarily involve the subjective competencies, and secondarily involve defence of values: trust and reliability, a moral order, ‘craft’, personal interaction, empathy, pride in ‘good work’ etc. The latter are not intrinsic to the defence of professions per se but are particularly important for many tasks which are too readily dismissed as routine. Therefore the protection of values should be seen not as a reason to protect particular professions but as a body of principles about discretion within work practice. This leads to a conception of care work highlights the need for commitments to public service (job security, training, collective management of work, recognition of labour force needs and cultural differences) without any need for occupational closure.

ABSTRACT. Most research on AI in the administrative state focuses on its use by officials. This paper addresses a comparatively neglected issue: the impact of the use of AI tools by the public themselves. Drawing on a qualitative survey with over 500 UK civil servants - including those working in the digital welfare state - it argues that AI use by the public is increasingly disrupting the front-line of the administrative state. The argument is put in three parts. First, as the administrative state increasingly relies on text-based inputs, officials use correspondence as a proxy for individuals – a longstanding assumption of ‘autographic authorship’ that the diffusion of generative AI tools unsettles. Second, the problem can be characterised as 'AI slop': AI-generated correspondence with high volume and velocity that adopts a particular voice in interactions with officials. Third, AI slop causes three forms of disruption: to the signals officials draw from correspondence, to trust between officials and the public, and to systems ill-equipped to deal with it. The paper concludes by arguing that the status quo itself fails to adequately meet the needs of many people accessing services, and that systems need redesigning not only for a society with AI slop, but in ways that address the drivers of it too.

ABSTRACT. Street-level bureaucrats (SLBs) interact directly with citizens and exercise substantial discretion in the execution of their duties. This discretion allows them to make context-sensitive judgments within the legal and institutional boundaries of their roles (Lipsky, 1980). The literature examining how digital tools influence the work of SLBs identifies three main perspectives. The curtailment perspective argues that technology reinforces hierarchical control, thereby reducing discretion. The enablement perspective sees digital tools as empowering resources that support SLBs in their daily tasks. The continuation perspective, in turn, highlights acts of resistance by bureaucrats who reinterpret or challenge the goals embedded in digital technologies (Marienfeldt, 2024). Despite these advances, as Marienfeldt (2024) notes, the complexity of frontline work calls for deeper inquiry into how digital tools—particularly those based on AI—affect the core functions and values of classical street-level bureaucracy. As artificial intelligence (AI) and other digital technologies become increasingly embedded in public administration (Wirtz et al., 2021; Zuiderwijk et al., 2021), they begin to intersect directly with the discretionary work of SLBs. AI-based tools are introduced to enhance service quality, improve efficiency, and optimize decision-making processes (Van Noordt & Tangi, 2023). However, these innovations also redefine the very nature of bureaucratic discretion, potentially altering how frontline officials interpret rules, interact with citizens, and perform their responsibilities. Consequently, digitalization profoundly transforms the everyday work of public servants, reshaping organizational dynamics and administrative practices (Bullock, 2019). In light of these considerations, the present study seeks to answer the following question: How does artificial intelligence affect the work of street-level bureaucracy? To address this question, we conduct a systematic literature review drawing on primary empirical studies published in English, Portuguese, and Spanish, collected from three high-quality scientific databases: JSTOR, ScienceDirect, and Scopus.

ABSTRACT. Contemporary welfare reforms are premised on assumptions of greater efficiency in benefit administration. Research on the “digital turn” in welfare governance shows how automated processes and digital interfaces reshape discretion and citizen-state relations (Bovens & Zouridis 2002; Ranerup & Henriksen 2020). Rather than eliminating human judgement, automation produces hybrid forms of bureaucracy in which professional reasoning is entangled with system constraints and algorithmic logic. While this shift is well documented in frontline activation work, less is known about how automation transforms discretion and competence requirements in back-office administration of income replacement schemes. This paper examines how automation restructures caseworker practice across sickness, parental and care-related benefit schemes, administered by the Norwegian Labour and Welfare Administration. Drawing on 29 qualitative interviews with caseworkers, we show how work has shifted from paper based assessments to system driven processing. Automated routines handle standard tasks, leaving caseworkers responsible for complex, ambiguous, and exception heavy cases while adapting to constraints of the automated systems. To succeed, caseworkers must combine legal interpretation, technical understanding, economic and analytic processing, and possess personal capacities such as rapid learning, precision, and coping with high production pressure. In addition, caseworkers must apply discretionary assessments on eligibility for users with complex work-histories and vulnerable life situations based on fragmented, written information. No existing educational pathway covers this hybrid skillset, making internal training essential, yet increasingly challenging, as simpler cases disappear. Our findings suggest that automation generates both efficiencies and new burdens: it reduces repetitive tasks but intensifies cognitive demands, fragments case information, but limits organisational learning, and increases distance between caseworker and users. These findings reflect a broader shift in digital welfare states, where policy is shaped at the intersection of human expertise and opaque digital infrastructures, raising new questions about discretion, accountability and procedural fairness.

ABSTRACT. The digital turn in welfare governance has transformed the information architecture of the welfare state in ways that both advance and unsettle its protective aims. We conceptualize visibility as a triad comprising three dimensions: access (recognition as a rights-holder), oversight (exposure to monitoring and fraud control), and epistemic standing (the capacity of claimants’ accounts to shape street-level decisions).

In the “old” welfare regime, these visibility dimensions were produced and negotiated through episodic, face-to-face encounters in which claimants disclosed personal information to street-level bureaucracies. Visibility enabled access to rights while simultaneously exposing claimants to scrutiny and control. Yet it retained an expressive dimension. Within discretionary encounters, claimants could narrate their circumstances, contest interpretations, and influence how their cases were understood. This paper argues that automated welfare systems reconfigures and intensify this visibility paradox. Automated eligibility may expand access and reduce administrative burdens. At the same time, they intensify oversight through continuous data integration and background monitoring. Most significantly, they erode epistemic standing: narrative explanation is displaced by standardized data inputs and algorithmic classifications that pre-structure frontline judgment. Claimants become hyper-visible as data subjects yet diminished as speaking agents.

The article contends that governing visibility - not merely protecting privacy - must become a central concern of welfare governance. It concludes by outlining legal, institutional, and participatory frameworks capable of rendering automated systems transparent and contestable, and of realigning digital welfare with commitments to dignity, fairness, and social citizenship.

ABSTRACT. The digitalisation of welfare governance has generated hybrid administrative environments in which discretionary judgment is distributed across street-level interaction, digital interfaces and encoded decision rules. While SLB scholarship on screen- and code-level bureaucracy demonstrates how digital systems restructure discretion, less is known about how these transformations reshape epistemic authority, power and possibilities for resistance in Public Employment Service (PES) encounters. This paper examines the embedding of digital epistemic governance within PES through hybrid digital–caseworker interactions, arguing that such systems preserve existing authority and fail to realise the disruptive potential of ethical algorithms and big data analytics. Recent European policy has framed caseworker override of digital PES systems as a positive and necessary move (Pieterson, 2025). This paper argues that, in practice, this mandated override can limit the capacity of digital systems to democratise knowledge and expertise (Mol, 2008). Rather than redistributing discretionary authority, these arrangements risk re-entrenching established power relations between caseworkers and clients, constraining collaborative decision-making and limiting opportunities to return labour-market data to citizens in ways that might enhance autonomy in career decision-making. The analysis draws on qualitative data from the HECAT EU Horizon 2020 project, which developed an ethical algorithmic platform visualising labour-market information for caseworkers and clients. Data derive from vignette-based focus groups and exploratory discussions with caseworkers in Denmark, France and Ireland (n = 37). Scripted scenarios elicited professional interpretations, judgments and anticipated practices in digitally mediated welfare encounters. Findings indicate that caseworkers interpret ethical algorithmic tools primarily as advisory resources that reinforce rather than redistribute decision authority; as a ‘tool in their armoury’ often filtering what information is shared with clients. The study provides empirical insight into emerging forms of digital epistemic governance and contributes to street-level bureaucracy debates on hybrid discretion, infrastructural power and constrained contestation within algorithmically supported welfare systems.

ABSTRACT. One of the most significant aspects of contemporary public sector governance is the enormous increase in production and use of performance data on public service delivery. Whether this performance data is being used to improve the delivery and management of public services or instead is misused to more “perverse” effects (e.g. goal displacement or data manipulation) is, however, more contested. This paper argues that an important - but hitherto largely overlooked - precondition for how performance information is used or misused, is the trust that public managers and -employees have in the performance information. We thus still know little of the factors shaping actors’ trust or mistrust in performance information and how this influences their use of said information.

To fill this lacuna, this paper develops a framework for studying trust in performance information within public service organizations. Based on ethnographic field work in four Danish municipal job centers and interviews with frontline workers, managers and data analysts in these job centers, this paper finds four distinct criteria for trust in performance information: 1) The validity of the information (i.e. whether the measured performance reflects actual performance); 2) The reliability of the information (i.e. whether the performance is measured correctly); 3 The relevance of the information (i.e. whether what is measured captures the most important aspects of performance); 4) The consequentiality of the information (i.e. whether the measurements have implications for actual practice). Furthermore, we find that the factors shaping trust - and the relative importance of each of these criteria - differ across the different types of actors (frontline workers, data analysists and managers).

By developing this framework, the paper thus furthers our understanding of the way trust in performance information ultimately shapes whether and how such information has perverse- or purposeful effects for public service delivery.

ABSTRACT. Social policy implementation is increasingly contracted out to non-governmental organizations. Street-level bureaucracy research has highlighted how relationships between governmental and non-governmental actors influence frontline implementation (Lotta & Considine 2025). Existing studies focus on discretion within contracted organizations, while paying less attention to how inter-organizational relationships between commissioning agencies and contracted non-governmental providers shape discretionary decisions and actions. Because frontline decisions depend on which forms of knowledge are mobilized (Møller 2021) and governmental and non-governmental organizations possess different kinds of knowledge, it is necessary to examine their interaction under contracting arrangements.

This study examines joint deliberations where commissioning agencies and contracted non-governmental providers participate, asking what forms of knowledge are mobilized and how they shape frontline decisions. It draws on a case study of the implementation of the Act for Supporting the Self-reliance of Persons in Need (SSPN) in Japan. Although local governments are responsible for implementing the SSPN, some support services are allowed to be contracted out. The SSPN also mandates regular support coordination meetings where both a contracted provider and the responsible government agency review support plans and discuss the future direction of support. These meetings are suitable for observing the process in which each actor mobilizes their knowledge through case-based deliberation.

Based on participant observation of these meetings and interviews, this study finds that public servants mainly drew on experiential knowledge accumulated through personnel rotations within the administrative organization, while provider staff relied on practical knowledge circulating within their parent corporation across municipalities. Governmental and non-governmental staff thus tended to mobilize different forms of organizational knowledge rather than credential-based professional knowledge, translating diverse difficulties into a client’s “chief complaint” and an actionable support plan. The study further considers power asymmetries, biases in the plans constructed through deliberation, and implications for the responsiveness of support for individuals in need.

ABSTRACT. Knowledge mobilization plays a significant role in both the design and implementation of public policy. While a long-standing body of research has examined how policymakers mobilize knowledge in policy design (e.g., Nutley et al., 2007), studies in the street-level bureaucracy literature have only recently begun to examine how street-level professionals (SLPs) mobilize different forms of knowledge in everyday decision-making (e.g., Møller, 2022; Lotta et al. 2023). Drawing on qualitative data from ethnographic fieldwork in a Dutch childcare organization and municipality, this research develops a nuanced understanding on the bidirectional mobilization of knowledge between policy design and policy implementation in street-level organizations. Building on the literature on knowledge in frontline work, this research shows how knowledge is mobilized through interactions between SLPs, managers, and policymakers.

Understanding these interactions is important, as SLPs possess specialized knowledge on clients’ circumstances and the practicalities of implementing policies. When this knowledge is not mobilized into policy design, policies risk misalignment with frontline work. Conversely, when knowledge from policy design is not adequately mobilized toward policy implementation, SLPs’ understanding of policy goals, rationales, and available resources stay limited, potentially increasing policy alienation and constraining effective implementation.

By not only analyzing how knowledge is mobilized toward SLPs, but also how their knowledge is shared, interpreted, and integrated into policy processes, this research seeks to contribute to scholarship on the role of SLPs in policy design (e.g., Cohen, 2021). It also speaks to scholarship on street-level management (e.g., Gassner & Gofen, 2018) by revealing the role of managers in mobilizing knowledge in both directions: shaping which forms of policy knowledge reach the frontline and which forms of SLPs’ knowledge feed back into policy design.

More than 200 hours of observations have been conducted, and the remaining fieldwork, including interviews, is expected to conclude in April 2026.

ABSTRACT. The article is concerned with the concept of value, central to Osbourne's definition of co-creation. While the complexity of value formation is recognized, there is remarkably limited empirical evidence on how such processes unfold at the micro level of (inter)action. In this article, we ask how value is co-created and co-destroyed in encounters between young service users and public youth employment services. Within the specific context of public youth employment services, we give priority to the service user perspective, and study value formation through lived experiences of young people NEET. We apply the Public Service Logic (PSL) as our main theoretical lens and combine this with inspiration from the field of youth studies and theories on belonging. This allows us to study how young peoples’ experiences align/conflict with the understandings of value reflected in services offered in public youth service. Together, the analysis nuances the understanding that ‘value-in-context’ is created “beyond the control of the public service organisations” (Strokosch & Osborne, 2024; Grönroos & Voima, 2013). Exploring the relation between value-in-use and value-in-context, we demonstrate how service use without integration of youth context involves experiences of value co-destruction and a dominating sense of normative and conditional belonging. On the contrary, the integration of context in use enables young people to experience value as co-created, reflecting a sense of belonging in relations and in the everyday. This finding reveals how ‘value-in-context’ is indeed within the reach of public service organizations to both control and destroy value but also to help create it. Transforming the public youth employment services towards increased co-creation thus calls for further responsiveness towards the narratives of youth

ABSTRACT. Edri-Peer and Cohen (2025) identify “moving with” citizens as a distinct strategy for street-level bureaucrats to cope with authority, procedures, time pressure, and citizen demands. Based on direct observations and semi-structured interviews with justice professionals and citizens in two Belgian judicial settings – probation services and proximity courts – we show that frontline officers may “move with” citizens who display voluntarism or vulnerability. We refine a sequence – trusting, listening, and engaging – through which “moving with” emerges. Mutual trust is built when agents value relational support and communication, while citizens show basic procedural understanding and willingness to engage. Active listening enables shared sense-making, distinguishing constraints from negotiable elements, and supports conditional empowerment that reframes, rather than undermines, judicial authority. Finally, “moving with” is facilitated by settings which foster a citizen-agent narrative. We draw some theoretical and practical implications of these findings.

ABSTRACT. In research on street-level bureaucracy (SLB) and value co-creation (VCC) alike, the frontline interactions in public service delivery are of paramount conceptual importance. Both research traditions, however, are 1) primarily concerned with dyadic frontline encounters, and 2) tend to focus on either the service provider (SLB research) or the service user (VCC research) side of the dyadic equation (Gofen et al., 2025). Accordingly, central questions of power, agency and value as these unfold in group-based (involving multiple service users) frontline encounters, and as these are simultaneously experienced on both sides of the interaction, remain empirically and theoretically underexplored in both research traditions (Rønshaugen et al., 2025). Drawing on ethnographic data generated at the frontline of a group-based youth activation initiative in Danish public employment services, the article asks what agency positions the group-based setting affords its participants, and what kinds of value that can emerge from the group-based co-creation process. The analysis reveals three ideal-typical modes of interaction: service provider dominated interaction, equal interaction, and service user dominated interaction, each yielding different agency positions and value outcomes. In the empirically most prevalent mode of interaction, equal interaction, service providers leverage their professional expertise to facilitate the exchange of experiential knowledge among service users regarding the practical, emotional and social consequences of youth unemployment, thereby affording the service users to act as both resource providers and recourse integrators. Despite this collaborative dynamic, normative conflicts over rights and duties do arise. In these situations, the service users may constitute a normative majority. The article concludes by proposing future avenues for research on group-based frontline encounters in SLB and VCC research.

ABSTRACT. Living with and caring for multiple chronic conditions demands continuous navigation across clinicians, treatment, and information sources. Digital tools offer opportunities to support this work, but their effectiveness depends on how well they align with patients’ needs, skills, and everyday contexts. The way patients and clinicians navigate both the opportunities and challenges of blended care and what this means for how care fits the daily lives of patients remains poorly understood. This qualitative interview study examines how patients with multiple chronic conditions and clinicians navigate care in a context of digitalization and how this navigation plays a role in making care fit patients’ needs, priorities, and capacities. Results show how patient and clinician experiences interact and how they relate to the sensitizing concepts. For instance, [xxx]. Understanding how patients and clinicians navigate these possibilities and challenges is essential for digitally supported care that is in alignment with patients’ lived realities.

ABSTRACT. The aim of the paper is to explain a research project about how citizens perceive responsiveness in encounters with professionals who use AI, both on-the-spot and in the long run. The core idea of the project is to examine how and why the use of artificial intelligence decision technology in public service delivery impacts responsiveness to citizens. Various theoretical frameworks by Abbott, Goffman, and Latour are integrated, as I explore how several interconnected dimensions of the citizen-professional interaction may take part in shaping the citizen’s experience of responsiveness of the encounter. I also integrate the theoretical proposition that responsiveness impact is conditioned by target group profile where citizens are seen and targeted in groups with varying levels of societal deservingness and political power. As the question addresses both meaning-making and structural conditions, the project integrates an interpretative and variance-based methodology to provide context-rich data on causal mechanisms. During observations not only professionals will be observed, but also citizens’ verbal and behavioural reactions in their interactions with frontline professionals. After their first contact with a frontline professional, citizens will be asked to participate in qualitative longitudinal research. They will be provided with a diary allowing them to log their experiences during and after frontline encounters. This will guide later follow-up interviews after decisions have been made. The results will be used to develop a typology of artificial intelligence responsiveness from the citizen perspective, comparing the role of different target groups’ political power and deservingness, to theorize the citizen perspective in AI-informed encounters, and to examine how policy intent on citizen outcomes and actual citizen experiences align. To examine the influence of professionals’ use of artificial intelligence on citizens perceived responsiveness a qualitative vignette experiment will be conducted. Results will be used to assess the influence of different frontline uses of artificial intelligence on how citizens perceive responsiveness, and whether these are different across target group profiles.

ABSTRACT. The aim of the paper is to explain a research project about how citizens evaluate responsiveness and artificial intelligence use by frontline professionals, both on-the-spot and in the long run. The core idea of the project is to examine how and why the use of artificial intelligence decision technology in public service delivery impacts responsiveness to citizens. The theoretical proposition is that this impact is conditioned by target group profile where citizens are seen and targeted in groups with varying levels of societal deservingness and political power. As the question addresses both meaning-making and structural conditions, the project integrates an interpretative and variance-based methodology to provide context-rich data on causal mechanisms. The project zooms in on the encounters between frontline professionals and citizens. The aim is to explore the different ways in which artificial intelligence-generated knowledge could affect frontline professionals’ responsiveness and to develop more nuanced theoretically expectations. First, frontline professionals from the different target group profiles will be observed in their encounters with citizens and office work to examine how they use artificial intelligence towards citizens. The empirical results will be used to answer the question how this may affect the responsiveness of frontline professionals in the different target group profiles, to examine how policy intent and actual artificial intelligence uses align, and to conceptualize the role of different types of citizen-generated knowledge in discretionary artificial intelligence contexts.

ABSTRACT. This paper reflects on the politics of design and governance of digital welfare infrastructures in Germany. Drawing on scholarship in Science and Technology Studies (STS) and Information Infrastructures Studies, we analyse the sociomateriality of digital welfare infrastructures and make visible different strategies of repair work of volunteer advisors.

Empirically, we build on a study of the German pension system as a comprehensive social security infrastructure. In recent years, this infrastructure has undergone significant digital transformation, encompassing the digitisation of pension applications, consolidation of diverse data sources, and automation of work processes. Since the creation of the German welfare state in the late 19th century, self-administration has been an important pillar. This implies that those paying into state pension (employees/employers) and those receiving state pensions (pensioners) elect representatives who participate in the organisation of the pension system. The elected representatives nominate about 4500 volunteer advisors who support citizens in questions about or applications for state pension. Through their close connection between civil society and the state, volunteer advisors are a core element of a functioning German welfare infrastructure. Yet, their work remains largely invisible.

Drawing on 30 qualitative interviews with volunteer advisors and pension staff, 5 focus groups and a survey with 2300 volunteer advisors, we present findings on (1) how digitalisation is positioned by the pension fund as a way to fix/repair pension delivery in the face of demographic ageing and staff shortage. (2) We consider how volunteer advisors experience, perceive, and assess digital transformation in their everyday consulting work. (3) We discuss how volunteer advisors themselves engage in continuous tinkering and repair to keep sociodigital welfare infrastructures running. While digitalisation may help providing social welfare in times of demographic change, the continuous work of volunteer advisors and pension staff is required to maintain a “good” welfare infrastructure.

ABSTRACT. Administrative codification does not eliminate bureaucratic discretion but displaces it within welfare systems. As eligibility criteria become embedded in ever-more fixed classification systems, legislative thresholds and standard operating procedures, frontline interpretive flexibility contracts and tensions emerge between administrative categories and complex social realities. This article conceptualises institutional responses to those tensions through the concept of bureaucratic repair work: the retrospective reinterpretation of formally valid welfare decisions whose application produces substantively unjust outcomes.

Drawing on interpretive document analysis of five years of data from Ireland’s Social Welfare Appeals Office (2020–2024), including 325 published appeal case summaries and longitudinal statistics across welfare schemes, the article advances two arguments. First, bureaucratic repair work is a structural feature of codified welfare systems because administrative classification frameworks cannot fully accommodate the diversity, fluidity and contingency of lived social circumstances. Second, the displacement of discretion to appellate mechanisms has important political and administrative consequences. It concentrates contextual judgement within institutions that are unevenly accessible, extends the burden of welfare interaction into a second administrative encounter, and stabilises the legitimacy of rule architectures whose failures it locally corrects without systematically resolving.

The article further argues that many dynamics commonly associated with algorithmic welfare governance predate digitalisation and are instead constitutive features of administrative codification itself. By examining Ireland’s historically layered and shock-responsive welfare system, the article contributes to debates on administrative burden, hybrid bureaucracy, welfare governance and democratic accountability.

Keywords: administrative burden; street-level bureaucracy; discretion; welfare administration; codification; formal rationality; interpretive policy analysis; appeals

ABSTRACT. One of the defining characteristics of street-level work is “endless” demand for services, creating chronic resource problems in street-level bureaucracies (Lipsky, 2010). Yet, digital technologies are often implemented to make public services more efficient, saving them time and reducing their workload. In this paper, I explore how workers in the Norwegian labor and welfare administration (Nav) perceive and manage demand in digital service channels. Nav is a ‘one-stop shop’ for labor, welfare, and social services. Its call center (NKS) responds to client inquiries over the phone, in digital chats (synchronous and asynchronous), and even trains Nav’s chatbot.

This study draws on data from NKS, including interviews with frontline workers, chatbot trainers, NKS experts, and one client (n=16), as well as observations of frontline workers in phone conversations with clients. Through abductive analysis, I aim to establish mechanisms generating and counteracting demand in these service channels.

Preliminary findings show a fundamental duality in digital service provision. On the one hand, digital service provision can enhance demand. One example of this is how new service channels tend to increase overall demand, even when channels are digital. Digital channels can also contribute to “failure demand.” This demand stems from public agencies not being able to solve clients’ problems, resulting in them “pushing” for service in additional channels (Seddon, 2005). On the other hand, NKS also uses digital service channels to manage demand. One example is their chatbot Frida, who responds to inquiries around the clock. Frida seems particularly useful in absorbing acute failure demand, when crises create sudden spikes in demand. Providing automated responses, Frida handles sudden influxes, making Nav less sensitive to fluctuations in demand.

This paper contributes to discussions about the often-invisible structures of digital service provision, arguing that it reshapes rather than resolves chronic resource problems in street-level bureaucracies.

ABSTRACT. Lipsky (1980) challenged what this paper terms policy realism from below, showing that policy is made through street-level discretion. This paper argues digitalisation challenges it from within. As algorithmic systems become embedded in Public Employment Services (PES), discretionary judgement relocates from street-level encounters to code-level architectures (Bovens and Zouridis, 2002). It creates forms of opacity, persistence and path dependency that amplify the problems Lipsky identified, requiring a rethinking of street-level governance in digital contexts and new analytical tools to study it.

Grounded in research on digital PES in Ireland within the EU digital strategy, the paper adopts a post-structuralist orientation integrating Foucauldian analytics, Deleuzian assemblage theory and Actor-Network Theory. Extending critical policy ethnography (Dubois, 2009) and SLB research, it develops assemblage ethnography as a method for studying hybrid bureaucracies, following policy across scales, sites and temporalities, tracing how algorithmic systems reshape frontline encounters in ways exceeding traditional SLB frameworks and orthodox policy analysis.

Ireland's Probability of Exit (PEX) algorithm demonstrates both claims. Conceptualised via Foucault's dispositif, PEX categorises jobseekers into risk profiles pre-determining caseworker interaction, conditionality and sanctions, making up populations (Hacking, 1986) through computational sorting while appearing as a neutral instrument. PEX reveals how digitalisation reconstitutes street-level practice, embedding governmental power in technical architectures resisting democratic oversight. Methodologically, assemblage ethnography reveals what conventional SLB approaches cannot: how PEX persists a decade beyond its crisis-era justification, an 'object of concern' yet simultaneously an 'object of ignorance' (O’Doherty and Neyland, 2019) and how resistance emerges within organisations rather than from citizens.

The paper contributes to SLB scholarship on two fronts: a substantive argument that digital welfare governance transforms where discretion resides and how power operates in state-citizen encounters, and a methodological innovation, assemblage ethnography, providing tools for studying hybrid bureaucracies where street-level and code-level governance entangle.

ABSTRACT. Public organizations are inhabited by human beings with thoughts and feelings, which influence their actions and decisions. Yet in comparison to the amount of scholarly attention directed towards the cognitive aspects of rational decision-making and concepts such as bounded rationality and cognitive biases, the role of emotions in bureaucratic knowledge-seeking and decision-making has received little attention. Emotions have traditionally been considered problematic in relation to knowledge-seeking and decision-making, as something that arises from the body rather than the thinking mind and distract, bias, disrupt, or undermine rational action. However, the dichotomy of rationality and emotions is increasingly being challenged, as scholars are recognizing the crucial role of emotions in forming motivations, giving direction to reason, building engagement, and sustaining efforts despite obstacles. In this paper, we seek to expand existing perspectives on emotions and provide a foundation for future scholarship on the role of emotions in frontline work by introducing the concept of epistemic emotions, and demonstrating its analytical potential and relevance. Rather than focusing on emotions as something that arises from or in citizen-state interactions, the concept of epistemic emotions directs our attention to the back-office aspects of frontline work and the practices guiding frontline knowledge-seeking and decision-making before and after encounters with citizens. To further develop our understanding of the role of emotions in shaping frontline inquiries, we introduce two key distinctions: First, a distinction between foregrounded and backgrounded emotions; and second, a distinction between integral and incidental emotions. On this basis, we demonstrate how these concepts can be used analytically, drawing on empirical examples from our own field research. We discuss the importance of epistemic emotions in knowledge-mobilization, including how the role of emotions in the “doing” of rationality and objectivity may be affected by contextual conditions, such as peer-relations, organizational and professional contexts, and institutional embeddedness.

ABSTRACT. This article advances the understanding of discretion in street-level bureaucracy by linking classic street-level bureaucracy theory with the growing literature on organizational ignorance. Organizational ignorance refers to situations in which not-knowing is neither an error nor a consequence of frontline workers’ lack of competence, but rather a structural and constitutive element of organizational knowledge. From this perspective, ignorance exists alongside knowledge and must be actively managed rather than eliminated. While theories of street-level bureaucracy have extensively addressed how knowledge, rules, and discretion shape frontline decision-making, the role of ignorance has received comparatively little theoretical attention. This article contributes to this gap by developing a conceptual framework that integrates organizational ignorance into street-level bureaucracy theory. Empirically, the article draws on an in-depth single-case study conducted in Denmark, comprising 75 semi-structured interviews with professionals from municipalities, police, and children advocacy centers involved in cases of suspected child physical abuse, extensive document analysis, and longitudinal observations over a five-year period. The analysis identifies interrelated forms of organizational ignorance in street-level practice such as for example: (1) ignorance as selective attention, enabling action under chronic resource constraints and time pressure; (2) ignorance produced by jurisdictional boundaries that formally delimit what frontline workers may act upon; (3) ignorance of practitioners’ own experiential knowledge, as standardized rules and procedures render aspects of complex phenomena invisible; and (4) strategic ignorance, which emerges when acquiring knowledge is likely to trigger obligations, interventions, or sanctions, leading frontline workers to avoid further inquiry or documentation. The article concludes by discussing the ethical implications of organizational ignorance in child physical abuse cases and reflects on how such ignorance may both protect and endanger children, highlighting the need for reflexive and transparent management of ignorance in street-level bureaucracies.

ABSTRACT. This paper reflects on the implementation of recovery capital (RC) in Swedish alcohol and other drug (AOD) treatment, drawing on findings from a doctoral dissertation comprising four studies (Härd, 2024). While RC is widely regarded as a conceptually useful framework, its translation into practice reveals tensions that extend beyond the well-documented relationship between standardised assessment and professional discretion. Findings suggest that professionals frequently adapted standardised tools to fit local treatment contexts — not as an obstacle, but as an expression of ongoing jurisdictional work. More fundamentally, RC's conceptual framework was already embedded in existing practice – pointing to a broader pattern in which practice-based knowledge exists and is actively used in social work, but remains unrecognised within formal knowledge frameworks. Local knowledge initiatives, though lacking formal evidential status, proved more practicable than externally validated tools. Drawing on Abbott's (1988) theory of professional jurisdiction, these findings are interpreted as expressions of a jurisdictional conflict that is not exceptional but constitutive of professional work. This conflict is particularly visible in the AOD field, where forthcoming legislative changes in Sweden will shift principal responsibility for treatment from social services to regional healthcare – a development that risks displacing the social and structural perspectives that have long characterised social work's approach to AOD-related problems. The paper concludes that the conditions for meaningful knowledge development in social work may lie not in more effective implementation, but in a more fundamental reckoning with what social work is.

ABSTRACT. Co-creation is often described as a fruitful way to support research informed practice. This paper aims to analyse and discuss the involvement of SLBs and Artificial Intelligence (AI) in co-creative research. It reports experiences of a co-creative research process where SLBs and researchers, supported by a Large Language Model (LLM), developed a factorial survey experiment for a research project about work capacity assessments of Swedish benefit recipients. In focus groups, SLBs were asked to describe common personal characteristics and life circumstances among long-term unemployed clients. Subsequently, the suggested characteristics were categorized into over-arching themes that were reduced into specific dimensions (e.g.health, education, social network etc.). Thereafter, an LLM was used to suggest and refine multiple realistic phrasings of descriptions of life situations. These phrasings were assessed and reduced by researchers and validated by SLBs. Together, the SLBs and the LLM enrichened the vignette construction, the former by providing up-to-date practice-based experience, and the latter by enabling variation through its capacity to propose alternative phrasings based on the processing of vast amounts of text data. Together, they provided the researchers with richer data for the vignettes, covering numerous problems of long-term unemployed clients within social services and employment services. We conclude that scholars could benefit from teaming up with SLBs and AI to obtain data that are both more practice-relevant and richer in scope. However, it is important to consider the mediating role researchers take in such collaboration. SLBs are a heterogenous group with different experiences and understandings of clients. LLMs can be an important tool to assist researchers in developing vignettes but needs to be handled with care. AI-generated suggestions must always be reviewed and revised by researchers to avoid stigmatizing or non-realistic scenarios

ABSTRACT. Along with the new trends in workplace design, also Norwegian labour and welfare services (Nav) has undergone structural and conceptual transformations regarding everyday work for the frontline staff in the services. During the last few years local Nav-offices have relocated into modern buildings adapted to an activity-based workplace concept with free seating and flexible spatial solutions that is expected to enhance creativity, collaboration, and productivity. In this paper we explore how the changes in work-place and -space have influenced frontline workers’ mode of working, including collaboration and co-production with service users. The study is based on data from two ongoing research projects in the Norwegian labour and welfare services in which the services had recently moved into buildings designed with the purpose of implementing the activity-based workplace concept. Data consists of field visits, observations, interviews and workshops with staff and service user, over a period of nearly two years. Using a thematic analysis strategy, we find that staff experienced enhanced flexibility regarding where one works: at home, in the office or in other places as a positive token of autonomy in their work. However, they also experienced enhanced isolation and lack of collective responsibility in the new work environment due to decrease in collaboration and interactions with co-workers in the daily work. The findings also point to less face-to-face encounters with service users and more contact via digital communication systems after moving into the new workspace. This was experienced as a loss regarding the opportunity to perform relational work and co-production with service users. Resulting from the changes in work-place structures, these changes in frontline staff’s ‘work-modes’ may in the long run affect their identity as labour and welfare service workers – from social workers to screen-level bureaucrats.

ABSTRACT. Much of the work on co-production suggests that it can lead to public services that are more democratic and empowering for service users, centred around the argument that involving service users in service delivery gives them more control over the services they use. This has particularly been argued in the case of more relational social services, where the interaction between service user and professional is key.

At the same time, there has been increasing attention to the potential dark sides of co-production: that it can actually re-create power imbalances, ‘responsibilize’ citizens, and be used as a cover for cuts. Furthermore, co-production calls into question the hegemony of professionals’ knowledge, potentially making them feel as though their expertise and identity are being undermined. This creates a set of paradoxical tensions: professionals have a duty of care for vulnerable service users, and their training, knowledge and expertise as street-level bureaucrats is centred around ensuring that services can lead to the best possible outcomes. Though engaging service users in co-production is argued to be a way to support this, transferring the risk and responsibility of service delivery to users could be disempowering for both service users (who do not want to take on this responsibility, or feel tokenized, or insufficiently skilled) and professionals (whose time and expertise become undervalued).

In this paper, I explore the tensions and paradoxes of co-production through the lens of relational leadership and ethics of care. Relational leadership recentres the notion of leadership away from the individual and towards the interactions and relationships between individuals, while the ethics of care provides a moral framework for why and how leadership can support mutual well-being in a co-production context. Drawing upon illustrative examples from case studies, I explore the dynamics of how a care ethical approach can help us to understand the tensions of responsibility, empowerment/disempowerment, and professional expertise versus expertise-by-experience that are inherent to co-production.

ABSTRACT. Activating labour market policy is implemented in everyday interactions between beneficiaries and job centres, where two logics intersect: welfare conditionality, linking benefit entitlement to behavioural compliance, and co-production, that relies on beneficiaries’ active involvement in service delivery. These logics are often considered difficult to reconcile and create a constitutive tension. The article examines the following questions: How do beneficiaries experience their interaction with the job centre within the structural tension between welfare conditionality and co-production? How is this tension interactionally processed, and with what outcomes? The analysis draws on 26 in-depth interviews with beneficiaries and ethnographic observations of 30 counselling interactions in German job centres. The findings show that beneficiaries predominantly interpret job centre interactions through the lens of welfare conditionality, while expectations of co-production are limited. Experiences related to co-production emerge only under specific conditions, i.e. whether beneficiaries perceive certain measures as useful and supportive and when beneficiaries are confident that they possess relevant resources. Three modes of navigating the structural tension in encounters between frontline workers and beneficiaries are identified: (1) when both logics are productively balanced through negotiation, (2) when welfare conditionality forecloses co-production through institutional distancing and adaptive compliance, (3) when both logics fail and create a deadlock situation. The results underscore the relevance of understanding the complexities of welfare interactions and their perceptions, highlighting the interplay of structural tensions and its navigation.