Revisiting professional autonomy and structural constraints: Relational work in anti-drug social work

Professional autonomy in social work

Professional autonomy has long been a central concern in international scholarship on the professionalisation of social work. Extensive scholarship has noted that social workers’ roles have expanded from those of primarily ‘value-oriented caregivers’ to institutional actors responsible for implementing policy objectives, managing risks, and delivering measurable outcomes (Dominelli, 2004). Consequently, professional autonomy is no longer understood as complete independence from organisational or state authority, but as a form of discretion exercised within increasingly dense governance frameworks.

Professional autonomy is commonly examined through the lenses of de-bureaucratisation, marketisation, professionalisation, and ethical independence. Banks (2004) argues that autonomy is reflected not only in practitioners’ value commitments to service users but also in their capacity to exercise judgement and act in accordance with ethical principles when confronted with organisational and policy pressures. These dynamics indicate that, even in liberal democratic welfare regimes, social workers routinely function as agents of the state. Studies of street-level bureaucracy and welfare governance have long shown that social workers’ discretion is embedded within systems of accountability, managerial oversight, and policy conditionality, rendering professional autonomy contingent, negotiated, and situational rather than absolute (Evans, 2010; Lipsky, 2010).

From this perspective, the tension between care and control is not unique to state-led or non-liberal systems, but constitutes a structural feature of social work across global jurisdictions. Social workers’ ethical commitments routinely coexist with responsibilities to implement state priorities, monitor compliance, and regulate behaviour. What varies across political and welfare regimes is therefore not whether social workers ‘do the state’s bidding’, but how discretion, negotiation, and moral justification are organised and enacted in everyday practice (Banks, 2020). This distinction is particularly salient in authoritarian governance contexts, where professional autonomy rarely takes the form of overt resistance or principled confrontation. Instead, autonomy is more commonly exercised through negotiation, strategic adjustment, and relational positioning (Song, 2024). As Gray and Webb (2013) note, social work in non-liberal systems is characterised less by open opposition to policy mandates than by a form of ‘strategic balancing’ between institutional demands and ethical commitments.

Accordingly, professional autonomy should not be understood simply as the capacity to resist power, nor as a binary condition that social workers either possess or lack. Instead, it is more productively conceptualised as a mode of situated agency produced within power relations, in which dependence and autonomy coexist. In this sense, the Chinese context examined in this study does not represent a categorical exception to international debates on social work autonomy.

Chinese anti-drug social work: Professional autonomy and institutional constraints

Chinese anti-drug social work has emerged as a distinct professional field, operating in close cooperation with police officers and primarily within community-based drug rehabilitation programmes shaped by state-led approaches to drug governance (Song et al., 2026). Similar to many Western jurisdictions, drug use interventions are commonly framed through public health or criminal justice agendas rather than a sole care concern (Fraser and Valentine, 2008). In countries such as the United Kingdom, the United States, and Australia, drug or alcohol social workers frequently operate within multi-institutional systems, including drug treatment, criminal justice supervision, and welfare provision (Berridge, 2013; McNeill and Weaver, 2010). Within these settings, social work practice has often been associated with ‘punitive reward’ regimes, such as conditional access to opioid substitution therapy or risk management affected by drug use.

The distinctiveness of Chinese anti-drug social work does not lie simply in the coexistence of care and control, which is a globally recognisable feature of drug-related social work. Rather, it lies in the degree of institutional embedding and functional integration of social workers within a party-state-led penal-welfare apparatus (Jiang and Song, 2025). Since the promulgation of the Anti-Drug Law (2008) and the expansion of community-based rehabilitation and correctional programmes, Chinese anti-drug social workers have been formally positioned as ‘assistants to governance’ (Zhili xiezhuyuan), tasked simultaneously with abstinence monitoring, psychosocial intervention, and coordination with public security and justice authorities.

This embeddedness produces a distinctive mode of practice characterised by close, routine collaboration with law enforcement, project-based procurement mechanisms, and stringent performance evaluation regimes. Unlike many Western contexts, where social workers’ involvement in drug governance is mediated through relatively segmented systems, Chinese anti-drug social workers operate within an integrated framework in which welfare provision, surveillance, and risk control are institutionally fused, under party-state governance. As a result, professional autonomy cannot be adequately conceptualised in terms of independence or discretionary authority; instead, it takes a relational and negotiated form.

Structural constraints in party-state-led welfare

In the Chinese context, the development of social work has been deeply shaped by national institutions and policy logics. Since the 2000s, the state has promoted the institutionalisation of social work through mechanisms such as service procurement, project-based management, and performance evaluation (He, 2017). This state-led welfare regime emphasises policy goal orientation and organisational compliance, positioning social workers in a dual role situated between ‘service’ and ‘governance’. On one hand, practitioners are expected to mobilise professional knowledge to intervene in and support vulnerable groups; on the other, they are required to assist the state in fulfilling social management objectives and maintaining social stability (Yan, 2020). Through this dual positioning, social work functions simultaneously as a ‘provider of social services’ and an ‘extension of political order’. Such an arrangement not only constrains social workers’ professional autonomy but also compels them to develop complex strategies of negotiation and survival to sustain meaningful practice within institutional boundaries.

In recent years, new political dynamics have reshaped social workers’ position. The establishment of the Central Department of Social Work in 2023 formally integrated social work into the Party Central Committee’s organisational structure, positioning it as a key component of ‘Party-building-led governance’ (Mao and Yu, 2025). This institutional update reinforced the political attributes of social work, requiring practitioners not only to serve communities and local populations but also to ‘serve the central work of the government, and the Party’. This emerging configuration, which can be termed a party-state-led welfare regime, further strengthens party ideology and political integration from the existing state-led framework. Under this regime, local social organisations are required to establish Party branches and operate through a dual mechanism of ‘Party-building leadership + professional services’, with Party-building indicators incorporated into project design, funding allocation, and performance evaluation. These reforms have deepened social work’s dependence on party-state institutions, rendering professional practice increasingly institutionalised and politicised (Guo, 2025). ‘Political performance’ thus became a crucial dimension of organisational survival and resource competition.

The financial and human-resource stability of social work organisations remains heavily dependent on government funding and project-based contracts. The dominance of project-based funding has produced a ‘short-term’ and ‘performance-oriented’ professional logic, placing social workers in a condition of ‘passive dependence’ (He, 2017). Simultaneously, new requirements related to Party-building assessments and political loyalty further intensify the administrative and hierarchical tendencies of social work. Practitioners must not only meet quantitative performance indicators but also cultivate and maintain trust relationships with government departments and party organisations to secure project renewals, policy endorsement, and funding flows. These structural pressures further weaken professional autonomy. Yet they also contribute to what Wang (2024) describes as a ‘combined construction’ of professionalisation: within institutional constraints, social workers carve out limited but sustainable practical initiative through strategies such as flexible adaptation, symbolic compliance, relational coordination, and political signalling. In this way, autonomy is reconfigured as a negotiated and relational process rather than an independent professional entitlement.

This professional ‘combination’ in the Chinese language reflects the new normal of social work under the party-state-led welfare regime, specifically, political loyalty operates alongside professional service, and performance evaluation coexists with client-centred practice. It captures the inherently tense position of social work within this governance structure, functioning simultaneously as the ‘front-end tentacles’ of policy implementation and as a mechanism of ‘soft governance’ tasked with maintaining community stability. This forms professional autonomy boundaries of anti-drug social work.

Research design

This study adopts a qualitative method to develop an in-depth understanding of social workers’ relational professional autonomy. The research focuses on two prominent anti-drug social work organisations, A and B, in western China, and two competing agencies, Y and Z, in eastern coastal cities, analysing their organisational trajectories, relational strategies, and project development paths.

These four organisations were selected because they represent two ideal-typical models of anti-drug social work: (1) ‘Professional organisations’, such as A and Y characterised by strong professional identity, adherence to social work standards, and an emphasis on service ethics; and (2) ‘Bureaucratic projects’, such as B and Z whose operations rely heavily on government trust, administrative networks, and project-based resource allocation. Contrasting these two models allows us to illuminate how social workers cultivate and sustain relationships, secure organisational resources, and enact daily practice under the same party-state-led governance.

Data collection

Data collection drew on multiple qualitative sources to ensure depth, breadth, and triangulation. We conducted semi-structured interviews with 22 participants, including 16 social workers across the four organisations, 4 project managers, and 2 local government officials. The interviews explored career trajectories, project operation logics, relational maintenance strategies, and forms of work stress (see Table 1). Each interview lasted between 60 and 120 minutes. To protect confidentiality, all names, locations, and organisational identifiers are pseudonyms.

OPEN IN VIEWER

Table 1.

Information on participants.

Organisations	Locations	Participants	Gender	Age	Roles	Years on anti-drug service
A	Western China	Lee	Male	36	Founder	7
		Bin	Male	28	Co-supervisor	5
		Ying	Female	26	Frontline worker	3
		Feng	Male	30	Frontline worker	4
		Hong	Female	29	Co-supervisor	4
B		Cheng	Male	31	Project manager	4
		Ming	Female	26	Frontline worker	2
		Zong	Male	28	Project manager	5
		Qun	Male	32	Frontline worker	3
		Neng	Female	25	Frontline worker	2
		Ping	Female	41	Former staff	5
–		Chun	Male	47	Official	11
Y	Eastern China	Bing	Male	35	Manager and policymaker	8
		Anne	Female	28	Project manager	4
		Ting	Female	24	Frontline worker	2
		Van	Male	31	Frontline worker	3
		Kun	Female	29	Frontline worker	5
Z		King	Male	33	Director	6
		Ing	Female	35	Project manager	4
		Jiang	Male	29	Frontline worker	3
		Bill	Male	34	Manager and policymaker	6
–		Song	Male	50	Official	15

Researchers conducted 10 months of participant observation at drug-control service sites in two localities and all four organisations. Observational settings included organisational meetings, outreach activities, community visits, and inter-departmental collaborations. These observations focused on everyday organisational atmospheres, relationship dynamics, and the enactment of emotional labour in routine interactions. Field notes totalling approximately 130,000 words document multi-layered organisational processes and practices. To supplement primary data, we analysed relevant policy documents, government procurement contracts, service delivery guidelines, project evaluation reports, and publicly available organisational materials. These documents provided contextual grounding, supported triangulation, and helped trace the institutionalisation of organisational–government relationships.

Analytical processes

The analysis employed a reflexive thematic approach (Braun and Clarke, 2019), informed by the conceptual lens of relational work. All interview transcripts and fieldnotes were analysed manually. The first and second authors conducted multiple rounds of close reading to familiarise themselves with the data and to generate initial codes inductively. Coding was iterative rather than blind, allowing continuous movement between data, theory, and emerging interpretations. This reflexive engagement enabled the analysis to remain sensitive to contextual meanings while progressively refining analytical categories.

In a second stage, related codes were clustered into provisional themes and examined comparatively across organisational and institutional contexts. These themes were then interpreted through a relational work framework to explore how boundaries were negotiated, how emotions and power intersected in everyday practice, and how relational strategies were mobilised under structural constraint. Throughout the analytic process, regular discussions were held among the research team to keep a reflexive approach, review coding decisions, refine theme definitions, and challenge emerging interpretations. Furthermore, this study was approved by the Ethics Committee of University of New South Wales. All participants provided informed consent, and strict anonymity and organisational confidentiality were upheld throughout the research process.

‘Professionalism as isolation’: The fragility of depoliticised autonomy

Organisation A is a drug rehabilitation social work agency widely recognised for its ‘professionalism’, which was established in 2012 by a group of young, social work-educated practitioners. Its founding vision was to promote a rehabilitation model grounded in humanistic care and social integration. As stated in its mission, ‘We are not the ones managing individuals; we are the ones accompanying them’. In a policy environment dominated by control-oriented and performance-driven logics, this positioning was distinctive, reflecting an idealistic professional ethos that viewed service recipients as subjects rather than mere objects of governance.

In its early years, Organisation A gained significant government recognition for its standardised procedures and high degree of professionalisation, becoming a local exemplar. It was emblematic of the ‘winner’ at the early phase of the state service-procurement system, securing legitimacy through adherence to professional standards rather than through administrative compliance. As the organisation’s founder recalled,

In those years, we were practically a model; many places came to visit. We believed that as long as we did a good job professionally, the state would naturally see it. (Lee, male, 2024)

However, this professional belief system became increasingly fragile as the broader political climate shifted. With the rise of ‘party-building-led’ and ‘penal-welfare’ paradigms, trust based solely on professional competence became insufficient. Organisation A did not establish stable political relationships promptly. In this evolving political ecosystem, absence effectively translates into marginalisation. As co-supervisor Bin candidly reflected,

We worked on this for ten years, and now the project is suddenly gone. We know it’s not because we did a bad job, but because they were ‘closer’ to the government. (Bin, male, 2024)

Within Organisation A, there is a pervasive sense of loss and powerlessness. During a team meeting, a young social worker quietly remarked,

Sometimes we feel like we’ve been forgotten by policy. We’re still talking about professional ethics, and they’re already talking about political performance. Professionalism used to make us proud, but now it’s become a burden. The government thinks we’re too ‘idealistic’ and not ‘pragmatic’ enough. (Ying, female, 2024)

As the political climate shifted, ‘professionalism’ became a marker of organisational isolation. Organisation A continued strict routines, such as archived interviews, standardised assessments, and supervised plans, reinforcing internal identity but deepening external ‘closedness’. This surfaced when officials invited the agency to a Party-building session. The manager recalled,

‘We’d rather focus our time on case work. We have our own schedule and professional standards to maintain, and these activities don’t directly relate to our service outcomes. If we step away from our cases, it affects the continuity of care for our clients’. The officials responded curtly, ‘Then you probably don’t understand the current priorities. These arrangements are not optional, and participation reflects your organisation’s stance’. (Fieldnotes, 052024)

What appeared to be an ordinary exchange in fact revealed a deeper fissure between the organisation’s professional ethos and the political logic of contemporary governance. As one staff member later reflected,

At first, we thought professionalism was all about relationships. But now we realise that relationships are the prerequisite for professionalism to exist. (Feng, male, 2024)

Organisation A’s internal resistance appeared ‘incongruous’. While most agencies embraced Party-building and political signalling to secure projects, it maintained ‘professional neutrality’, avoiding politicised expression. Although echoing Western ideals of ethical independence, this stance had limited traction in China’s party-state environment. As one participant reflected,

We always thought we were upholding professionalism, focusing only on service delivery and ethical boundaries. But looking back, we were actually avoiding engagement with the system and its people. In this environment, relationships are part of how work gets done. What we saw as professionalism was, in practice, a form of resistance to relationships. (Hong, female, 2024)

Field observations suggest that this resistance is not a silent protest but an emotionally charged commitment. In internal trainings, social workers repeatedly invoked principles such as ‘professional ethics’ and ‘people-centredness’, reinforcing a collective identity anchored in the moral significance of professional practice. At a debriefing meeting, a project manager declared,

We cannot abandon our professional identity, even if it means being marginalised within the system. If we start adjusting everything to fit external expectations, then what we do loses its meaning. We are trained to uphold certain values, and those values should guide our decisions. We do . . . social work, not just tasks assigned by others. (Fieldnotes, 092025)

This commitment carries a tragic undertone: it is simultaneously a self-protective ethical defence and a soft symbol of rupture from the prevailing order.

From a relational work perspective, Organisation A’s predicament stemmed not from inadequate competence but from its refusal to cultivate and maintain strategic relationships. This stance limited opportunities to rebuild trust or visibility within the governance system. Professionalism was recoded as a moral boundary – purity, rationality, neutrality – yet it deprived the organisation of flexibility in an increasingly politicised environment. Over time, this voiceless professionalism produced detachment from key administrative networks, reflected in project losses, staff turnover, and declining policy support. The result is a structural paradox: once professionalism is read as rejecting politics, it loses legitimacy and becomes a dissenting voice, rendering the organisation simultaneously ‘more professional’ yet ‘more vulnerable’.

‘Being trusted to obey’: Relational work in bureaucratic survival

Organisation B operates with an almost opposite logic. Founded in 2016 under the civil affairs department, it is led by a director deeply familiar with government procedures and evaluation systems. The organisation openly defines itself as a ‘Party-led social work entity’, with office slogans such as ‘Party-building leads livelihood services’ and ‘Obedience is the greatest professionalism’. At internal meetings, the director repeatedly stresses that ‘administrative trust matters more than professional ability’, a principle that anchors the organisation’s daily operations:

We must respond to the district’s instructions immediately, without delay. Remember, ‘professionalism’ here means the ability to implement requirements, not your own ideals. (Fieldnotes, 042024)

What may appear as casual guidance succinctly captures the organisation’s cultural logic: administrative loyalty supersedes professional judgement, and political performance is exchanged for institutional security. Within this organisational culture, cooperation is elevated as the highest value. As one project manager explained,

We may not be the most professional, but we are the most ‘cooperative’. The government needs trust, not questioning. (Cheng, male, 2024)

Another social worker echoed this:

As long as the leadership is satisfied, the project is stable. We don’t have to worry about funding running out. (Ming, female, 2024)

Here, ‘trust’ stems not from professional merit but from mutually beneficial political ties. This cooperative relationship has ensured repeated contract renewals and expanded assignments, demonstrating that political embeddedness, rather than expertise, drives organisational stability and resource growth in the current governance landscape.

In addition, social workers in Organisation B proactively visit community officials, attend meetings, organise group photos, and send images of their activities to demonstrate ‘effective execution’. These symbolic practices have become embedded in performance evaluation. As the manager, Zong frequently reminded staff,

Send the photos to the district as soon as possible; they like to see us ‘following up’. Make sure the pictures are clear and show interaction, not just attendance. It’s not only about what we do, but how it can be seen and reported. Otherwise, they won’t count it as real progress. (Fieldnotes, 082024)

Such visual reporting has replaced earlier forms of professional reflection and supervision; ‘evidence of compliance’ communicated through images now functions as the most effective organisational signal.

Yet this deep administrative embedding does not fully suppress practitioner initiative. A ‘grey zone’ persists within Organisation B. When oversight from superiors is relaxed, social workers employ flexible strategies to adjust their work. As one participant explained,

What the higher-ups value is the quantity of activities; we can arrange the details ourselves. For example, in case of counselling, we can secretly have a few more sessions, as long as the report is well written. (Qun, male, 2024)

This form of ‘tolerated flexibility’ exemplifies relational autonomy in practice rather than openly resisting the system; social workers carve out space for discretion within the boundaries of administrative trust. The director tacitly endorsed this arrangement:

As long as nothing goes wrong, I don’t care how it’s done. You can adjust your approach depending on the situation, just make sure the outcomes are acceptable. We need to keep things running smoothly and avoid drawing unnecessary attention. If the work is delivered and no problems arise, that’s what matters. (Fieldnotes, 082024)

However, this grey-zone autonomy is fragile. The organisation lacks systematic training and mechanisms for professional reflection, leading to widespread burnout. A young social worker commented,

We do a lot of activities every day, but we don’t know what the meaning of these activities is. Reports are written, and performance reviews are passed, but nobody cares about the service recipients. (Neng, female, 2024)

Within this pragmatic organisational culture, ‘professionalism’ becomes redefined as the ability to comply with instructions and avoid risks, rather than as an ongoing process of reflection, improvement, and ethical engagement.

Through ongoing administrative engagement, Organisation B’s staff internalised a notion of ‘autonomy through trust’. As one put it, ‘Our autonomy is growing from being trusted’. This reflects a dependent, system-granted agency permitted only within sanctioned institutional boundaries. As a former staff member summarised,

Success here isn’t about how professional you are, but about how well you know how to get along with people. You need to understand what officials expect and how to respond without creating friction. Professional skills still matter, but they come after relationships. If you can build trust, then you are given more space to act. (Ping, female, 2024)

Organisation B exchanges political loyalty and administrative closeness for operational space within the system. Its ‘relational autonomy’ is not confrontational, but a constrained agency generated through deep political embedding. This creates limited grey zones of practice, yet such breathing room remains contingent on shifting political climates and bureaucratic preferences. In effect, Organisation B survives by trading compliance for existence: its autonomy is relationally produced, power-shaped, and bounded by governance logics of trust – an everyday oscillation emblematic of social work in China’s party-state-led, judicially driven drug control system.

‘Knowing the language of power’: Navigating competition through relational autonomy

In Western cities, anti-drug social work operates within a competitive yet coexisting landscape of professionalism and relationship-based practice. Organisations must demonstrate both technical compliance and political reliability in bidding, evaluation, and assessment (Wang, 2024). This institutional configuration intensifies the tension between professional practice and political embeddedness. As one policymaker observed,

One needs expertise, and one needs connections. What we need now, in fact, is both. (Bing, male, 2024)

Organisation Y first gained prominence as a ‘model of professional drug-control social work’, but as Party-building and governance indicators grew central to evaluation, its professionalism was reinterpreted as political indifference. During a municipal anti-drug project presentation in 2023, the director delivered a meticulously prepared report detailing service outputs and case achievements, yet the recipient remarked,

The report is good, but it seems to lack political warmth. It focuses too much on technical details and not enough on alignment with current priorities. We need to see how your work connects with the broader agenda and reflects the right stance. Professional results are important, but they should also be expressed in a way that resonates politically. (Fieldnotes, 102024)

Since then, Y has repeatedly lost competitive anti-drug bids, and its professional stance has been labelled ‘idealistic’, ‘detached’, and even ‘too academic’. As participant Anne recalled,

Our drug service proposals have always been logically clear, but now reviewers keep saying, ‘It lacks political orientation’. (Anne, female, 2024)

Y’s predicament illustrates a structural reality that under a bureaucratised and politically inflected project-review regime, ‘professional competence’ is no longer a decisive asset. Anti-drug government officials prioritise organisations’ responsiveness to Party-building requirements, participation in political activities, and ability to demonstrate political reliability within the broader discourse of ‘social security’. As one frontline worker privately reflected,

We thought that as long as we delivered good services and solid data, we could win. But now, ‘who understands the political language’ matters more than ‘who understands the drug users’. (Ting, female, 2024)

In stark contrast to Organisation Y, Organisation Z achieved rapid and sustained expansion through a flexible strategy of relational adaptation. The organisation cultivated close ties with local party bodies and the municipal anti-drug committee, while also embedding itself in a broad range of cross-departmental activities. As its director remarked,

We don’t just do anti-drug work; we also do party-building activities, mental health services, and community patrols. The logic is simple . . . whoever has a stable relationship has a larger scope. (King, male, 2025)

Field observations reveal that Organisation Z strategically replaced professional expression with political responsiveness:

At reporting meetings, project achievements were framed within narratives of ‘serving the masses’ and ‘spreading positive energy’, rather than standardised professional indicators . . . This discursive alignment secured Z’s recognition as a key municipal project in early 2024. (Fieldnote, 092025)

Compared with Organisation Y, Z’s organisational ‘success’ is rooted less in its professional capacity than in its acute reading of institutional logic and its fluency in power language. Social workers in Z adeptly weave political terminology, such as ‘party-building leadership’ and ‘supporting the modernisation of grassroots governance’, into their reports to secure governmental approval. As one participant noted,

We all understand these words are a ‘password’. Speaking their language allows you to breathe a little more air. (Ing, female, 2025)

This practice of discursive language translation is not mere conformity but embodies a form of relational autonomy. A strategic negotiation in which symbolic expressions of loyalty are exchanged for greater flexibility in practice. One employee was explicit:

We write ‘Party building activities’ in our reports, but the actual activities are mostly psychological counselling. The higher-ups want a political flavour, so we add a little. That way, everyone is comfortable. (Jiang, male, 2024)

Through such calibrated symbolic compliance, Organisation Z maintains operational leeway and sustains its project survival even under tightening administrative oversight.

This competitive ecosystem in coastal anti-drug work reveals the logic of China’s project-based governance: professional ethics, administrative trust, and political loyalty are collapsed into a single evaluative regime. ‘High-quality’ institutions are defined less by service outcomes than by political visibility and recognisability. Institution Y, keeping ‘professional distance’, lost political oxygen, whereas Organisation Z sustained support through relational embeddedness. As one policy maker succinctly put it,

What we need is not who understands social work best, but who understands our language best. You have to know how to translate your work into terms that fit our priorities and reporting logic. It’s not just about doing things well, but about making them legible to the system. If we can’t recognise it, it’s hard for us to support it. (Bill, male, 2025)

Crucially, this ecosystem is not zero-sum but structurally symbiotic. Professionalised institutions provide the system with technical legitimacy, while relationally adaptive organisations translate and stabilise policy implementation. Together, they enact the dual logic of Chinese social governance, technocratic rationality alongside political loyalty, within which relational autonomy is continually practised and reproduced.