The filtered workforce

The argument across the eight preceding pieces of this series has held the frontline social worker steady as a sympathetic figure. The corporate-stupidity diagnosis explains the failure as institutional rather than individual. The asymmetric-burden argument places the cost of correction on families rather than workers. The class-and-design analysis traces the failures of the system back to the demographic that built it, not the demographic that staffs it. These framings remain right. The workers are not the cause. But the workers are in the room, and a serious investigation has to be able to say something about who is in the room and why the room contains them and not someone else. This essay does that, carefully, because the move from a structural analysis to a workforce analysis is the move where careful journalism most easily becomes its opposite, and the discipline of the previous pieces has to be maintained.

The thesis is structural, not characterological, and it is advanced here as an interpretive model that the available evidence is consistent with rather than as a demonstrated mechanism. A safeguarding workforce that has been operating under the conditions described in the previous pieces — defensive practice, threshold pressure, asymmetric punishment, suppressed challenge, mandatory training as compliance ritual, settlement-agreement exit for dissenting staff — will, on the structural-selection model, tend over time to retain a particular kind of practitioner at the frontline. Not because that practitioner was wrong to enter, but because the practitioner who remains is what is left after two filters operating over the length of a career. The first filter appears to remove many of the practitioners least able to emotionally compartmentalise the work. The second filter removes the dissenting from public discourse after they have already been removed from the workforce.

A note on what is known, what is emerging, and what is interpretive

Six claims run through this essay, at three different levels of evidential support, and it matters to distinguish them before the argument develops.

Established: Social work disproportionately attracts mature entrants and entrants with personal experience of adversity, on the documented Social Work England and DfE workforce data. Secondary traumatic stress in helping professions is well-documented in the peer-reviewed clinical literature (Figley, Bride). UK social-worker burnout has been measured using validated instruments and shows substantial prevalence. Settlement agreements containing confidentiality clauses are in widespread use across UK local authorities, with the Bristol-specific figure of 59 in five years documented through Council FOI disclosures and Bristol Live's mainstream reporting.

Emerging or contested: Harry Ferguson's ethnographic work on contemporary social work practice documents defensive practice, sensory withdrawal, routinisation, emotional self-protection, and the "not seeing" and "not hearing" patterns that practitioners themselves describe. The translation of those ethnographic observations into clinical-trauma vocabulary (dissociation, depersonalisation, affective blunting) is the interpretive move this essay makes, not Ferguson's own framing.

Interpretive hypothesis: That the institution's procedural-rationality demands and incentive structures around defensive practice tend to retain practitioners whose adaptation style is compatible with proceduralism and to lose practitioners whose adaptation style remains affectively engaged. This structural-selection model is consistent with the documented attrition data, with the moral-injury literature in care professions, and with practitioner testimony from those who have left — but it has not been directly demonstrated by the kind of longitudinal workforce study (psychometric comparison of retained versus departed cohorts) that would establish it in strong form. Such a study has not been published for English adult safeguarding specifically. The model is offered here as the analytical contribution of this essay; the reader should treat it accordingly.

A further note on scope. Much of the empirical workforce literature this essay draws on is from child-protection social work — Ferguson's primary fieldwork, the Munro Review, Frontline programme evaluation data — rather than from adult-safeguarding-specific research, which is less developed. The argument extends the better-documented child-protection literature to the adult-safeguarding context on the basis that the workforce training pipeline, the institutional culture, and the procedural architecture are continuous across both. That extension is defensible but should be explicit rather than seamless.

Who enters

The demographic profile of new entrants to social work has been studied carefully in the UK, principally through Social Work England's registration data, the Department for Education's social work workforce surveys, and the academic literature on motivations and pathways into the profession.^[1] Three features stand out.

First, social work disproportionately attracts mature entrants. The proportion of newly qualified social workers in their thirties or forties at the point of qualification is substantially higher than for most graduate professions. The pathway is often a second or third career, frequently a deliberate move from another sector — teaching, nursing, retail, customer service, the voluntary sector — into work the entrant perceives as more meaningful. This is well-documented and has been broadly stable for the last two decades.

Second, social work attracts a high proportion of entrants with personal experience of the issues the profession addresses. This is sometimes called the wounded healer route, after the Jungian concept developed in the 1950s and substantially elaborated in the contemporary peer-reviewed clinical-training and counselling-psychology literature.^[2] The pattern is documented across the helping professions — psychology, counselling, psychotherapy, social work — and is more pronounced in social work than in most. People who have experienced care, mental health crisis, addiction recovery, family breakdown, bereavement at a young age, or analogous adversities are statistically over-represented among new entrants. The qualitative research on motivations consistently finds these entrants citing direct or indirect personal experience as the reason for their career choice.

Third, the educational pathway is increasingly the postgraduate conversion route — an undergraduate degree in a related discipline (psychology, sociology, social policy) followed by a Master's in Social Work, often funded by a Department for Education bursary or by the local authority itself through a sponsored route. Step Up to Social Work, Frontline, Think Ahead and the various fast-track programmes have, since their introduction, supplied a growing fraction of the new-entrant pipeline.

None of this is, on its own, problematic. The wounded-healer route in particular is reliably correlated, in the outcomes research, with the very qualities families on the receiving end of social work consistently report valuing: empathic register, the ability to hold non-judgemental space, lived understanding of what state intervention feels like from the other side. The strongest social workers, as reported in the families' qualitative testimony, are disproportionately drawn from this pathway. The route in is not the problem.

The problem the rest of this essay describes is the route through.

What unprocessed adversity does, clinically

The clinical literature on trauma and helping professions has, since the 1990s, identified two broad outcomes of personal adversity carried into professional life. The distinction is not about the adversity itself; it is about whether the adversity has been substantially processed before, or during, entry into the work.

The first outcome is the wounded-healer pattern proper. Adversity that has been processed — through therapy, time, support, reflection, the writing-it-down or talking-it-through that constitutes integration in the trauma literature — produces practitioners whose understanding of suffering is direct and whose empathic capacity is widened by what they themselves have come through. Bessel van der Kolk's The Body Keeps the Score, Judith Herman's Trauma and Recovery, and the practitioner literature descending from both describe the integrated state in detail.^[3] The professional who has done this work is, in the outcomes research, among the more effective practitioners in the helping professions. The wounded-healer pathway, when it terminates in this state, has been associated with improved outcomes for the people the professional then works with.

The second outcome — less discussed in the lay literature on social work, but extensively documented in trauma research — is the unprocessed state. Adversity that has not been integrated remains in the body and the affective system in particular forms. The clinical names are dissociation (when the person separates affectively from the source of distress), affective blunting (when emotional response is reduced as a protective measure), depersonalisation (when the person experiences themselves as detached from their own actions), and compartmentalisation (when distressing content is held in an inaccessible mental compartment that does not connect to ordinary functioning).^[4] Each is a defensive adaptation. Each is, in its own context, protective. The person carrying it has survived something that would otherwise have been unsurvivable.

The adaptations are often only partially visible to the person carrying them, with the degree of self-awareness varying across the practitioner population. This is the central clinical point. Dissociation, by its nature, frequently operates outside the conscious awareness of the dissociating person — that is, in significant part, what dissociation is. The person can experience themselves as functional, as competent, as appropriate to the situation. The protective adaptation often operates beneath the level of self-knowledge, though some practitioners do describe partial awareness of emotional numbing, detachment, exhaustion or shutdown when asked about it directly. The clinical literature on practitioner self-awareness is mixed on the precise degree to which dissociated practitioners can identify their own state.

When such a practitioner enters a helping profession that requires them to be present with the suffering of others, the adaptation can produce a specific phenomenon, which the trauma research has documented in detail. The practitioner is able to perform the tasks of the role — the visits, the assessments, the form-filling, the conversations — while the emotional dimension of those tasks may not register in the way it would for a non-dissociated person. The completion of a Section 42 enquiry against a loving spouse, for a non-dissociated person, is an experience that would normally produce significant affective response: doubt, discomfort, the felt sense that something is wrong with the procedure. For a dissociated practitioner, the same act can be completed cleanly, because the affective channel that would deliver the doubt is muted. This is not callousness. It is, on the structural-selection model advanced here, the same protective adaptation that allowed the person to survive whatever they originally survived, redirected into the work.

Harry Ferguson's ethnographic research on contemporary English social work practice — primarily on child protection rather than adult safeguarding — documented in detail what practitioners do under sustained safeguarding pressure. Ferguson physically accompanied social workers on home visits over a period of years, recording what happened in the room.^[5] His published findings describe defensive practice, sensory withdrawal, routinisation, emotional self-protection, task-oriented detachment, and what he calls "not seeing" and "not hearing" patterns. Practitioners attending serious incidents, removing children, conducting capacity assessments, performed those tasks while displaying observable detachment from what was in the room. The talking-without-hearing pattern. The visiting-without-seeing pattern. The form-completion as cognitive refuge. Ferguson's own framing is ethnographic and is careful not to pathologise practitioners in clinical-trauma terms; his argument is that the institutional pressures (caseloads, procedural demands, time pressure, structural impossibility of being fully present to every case in a thirty-case workload) produce these defensive patterns in workers who would otherwise function differently. Translating those ethnographic observations into the trauma-clinical vocabulary above — reading Ferguson's "not seeing" through the dissociation literature, his routinisation through affective blunting — is the interpretive move this essay makes. It is a defensible move because the surface phenomena Ferguson describes and the clinical phenomena the trauma literature describes are recognisably similar. It is also a move whose clinical confidence exceeds anything Ferguson himself claims, and the reader should weight it accordingly.

Who stays — the institutional selection

The institution does not screen for the difference between processed and unprocessed adversity at the point of entry. The screening would require reflective supervision time, clinical resources, and a culture of psychological depth that English local authority adult social care has largely not had since at least the early 2000s. The 2011 Munro Review and the subsequent literature identified the loss of reflective supervision as one of the central deteriorations in the profession over the previous twenty years.^[6] The institution receives both states of entrant — the processed and the unprocessed — without distinguishing between them, and then subjects both to working conditions that produce, on the documented evidence, a substantial rate of secondary traumatic stress.

Charles Figley's research on secondary traumatic stress, developed across the 1980s and 1990s and now the standard reference, describes the condition as functionally similar to primary PTSD in symptom profile, transmitted through professional exposure to the trauma of others.^[7] Brian Bride's 2007 study of social workers in the United States found roughly 15% of practitioners meeting full diagnostic criteria for secondary PTSD, with substantially higher proportions exhibiting subclinical symptoms.^[8] UK studies, including Queen's University Belfast research using the Maslach Burnout Inventory with English social workers via Community Care, and the Doherty et al. 2020 study of N=1,257 UK social workers, have found high prevalence of emotional exhaustion and depersonalisation across the workforce.^[9] The combination is what the literature has begun calling compound trauma: practitioners with unprocessed primary adversity, working in conditions that generate secondary trauma, often operating without the institutional supports that would allow either to be processed during the working life.

The structural outcome over a career, on the model advanced here, is interpretively predictable. The cohort of entrants whose original adversity was processed, and whose empathic capacity has therefore remained substantially intact, can experience the working conditions as increasingly intolerable. The mismatch between their reasons for entering and the daily reality of the work has been documented in the British Journal of Social Work — what the Sundkvist and Zorn 2023 study called the dislocating routines phenomenon, the practitioner's sense that staying too long produces a person they dislike.^[10] These practitioners disproportionately leave. The 48% of newly qualified social workers who anticipate leaving within five years, in the Social Work England / YouGov 2020 workforce research, are on this model disproportionately drawn from this cohort, though the survey data do not directly establish the personality-adaptation linkage. The 4,000 social workers who failed to renew their registration in 2022 are, on the same interpretive frame, drawn disproportionately from this cohort.^[11]

The cohort of entrants whose original adversity was not processed, and whose dissociative or compartmentalising adaptations are already in place, can experience the working conditions differently. The conditions that drive the empathically-intact cohort out are, for the compartmentalising cohort, more manageable. The form-filling that costs the empathically-engaged practitioner emotional energy costs the compartmentalising practitioner less, because the form-filling is consonant with how that practitioner already operates. The protective adaptation that allowed them to survive earlier life is, on the structural-selection model, well-suited to the work as the institution operates. They are more likely to survive. They progress. They become senior. They train the next cohort.

This is the first filter on the structural-selection model. The strong version of the claim — that this cohort is "the numb" and the leaving cohort is "the empathic" — overstates what the available evidence supports; the institution does not measure the personality-adaptation profile of its retained workforce, and no direct longitudinal study has demonstrated the differential-retention mechanism in the form the strong version requires. The weaker version of the claim — that institutional conditions tend to select for compartmentalisation-compatible adaptation styles — is consistent with the documented attrition data and the moral-injury literature, and is the version this essay advances.

What else is operating alongside the workforce-psychology argument

The structural-selection model this essay advances is one explanatory layer in a multi-causal landscape. The workforce dynamics described above operate alongside, and in interaction with, a long list of other documented pressures on the English adult social care workforce: austerity and the local-government funding crisis that has produced sustained understaffing relative to demand; managerialism and New Public Management reforms that translated practitioner-client relationships into contractual ones; the procedural and digital-system load that consumes time previously available for reflective practice; the risk-aversion and legal-liability culture that has built up around regulatory and inspection regimes since the 2000s; the inspection pressure from the CQC and the safeguarding-board governance architecture; safeguarding-policy incentives that reward enquiry-completion over enquiry-quality; and the long-term effects of caseload size in workforces that have been understaffed for a decade or more. Each of these is independently consequential and the workforce-psychology argument should not be read as a substitute for any of them. The honest version of the argument is that the structural-selection model is one explanatory layer among several, and that the layer is doing analytical work the others are not — naming what tends, on the interpretive frame, to happen to the practitioner population over the length of a career under conditions the other layers describe. The interaction matters more than any single factor in isolation.

The second filter — settlement agreements and the loss of internal critics

Not every practitioner with sustained empathic engagement leaves quietly. Some try to stay and change the institution from inside. A smaller cohort develops the political and professional skill to articulate what they are seeing, document the failures they witness, and push back through formal channels. This cohort is the institution's most acute internal management problem, and English local government has, over the last decade, developed a specific apparatus to handle it: the settlement agreement containing confidentiality clauses.

The Bristol Live report of 25 January 2025, based on the answer given by Bristol City Council to a question put by Councillor Graham Morris (Conservative, Stockwood) at the council member forum meeting of 14 January, established on the public record that the Council signed 59 settlement agreements with departing staff in the five years to December 2024.^[13] An average of twelve a year. The figure was given by the Council itself in response to the formal question and is in the documentary record. The Bristol Live coverage notes that the figure excludes school staff and excludes further settlements signed under the Green administration since December 2024.

The structural concern that flows from these figures is not that every individual settlement agreement conceals institutional wrongdoing — many such agreements arise from interpersonal disputes, redundancy negotiations, sickness-absence resolution, discrimination claims, performance disputes, and various other sources that do not involve concealment of organisational misconduct. The structural concern is the aggregate effect: settlement agreements containing confidentiality clauses can have the effect of removing experienced internal critics from public discourse, regardless of the merits of any individual case, and the resulting silencing operates systematically across the population of departing staff. Where the Bristol-specific figure is unusual — twelve agreements per year for a single local authority is at the upper end of the comparable distribution — the aggregate concern is correspondingly larger.

The Bristol pattern is not unique. Comparator data from other UK councils shows similar use of settlement agreements with confidentiality clauses, with figures varying by size of authority. Gwent's five Welsh councils signed 227 such agreements in five years at a cost of £3.2m; Caerphilly alone signed 116 at a cost of £1.6m, averaging £14,000 per settlement.^[14] The York City Council spent £251,471 in a single year on ten such agreements.^[15] Glasgow City Council spent £377,256 over five years on 35 such agreements.^[16] The pattern is national. The use of public money to settle staff departures under confidentiality terms is one of the more consistent operating features of contemporary English and Welsh local government, and one of the least scrutinised.

For the structural argument of this series, the relevant feature is the documented effect on public discourse. Settlement agreements with confidentiality clauses systematically remove from public availability the testimony of departing staff — including, in the Bristol case, staff who had raised institutional-racism concerns following the Bijan Ebrahimi independent review.^[17] Whether or not any individual agreement concealed institutional wrongdoing, the aggregate effect is that the cohort whose departing testimony would most usefully inform public understanding of the council's conduct cannot, by the terms of the agreements they have signed, contribute to that understanding. The cohort with the most direct access to what happened inside the institution is the cohort the institution has paid not to describe what they saw.

This is the second filter on the structural-selection model. Settlement agreements do not, on their own, demonstrate institutional misconduct, but they reliably reduce the population of departing staff who can publicly inform external scrutiny of the institution's operation. Combined with the first filter, the effect is that the practitioner cohort the institution loses tends to be the cohort with the most acute critical view of its operations, and the public availability of that critical view is systematically reduced at the point of exit.

What is left in the room

The combined effect of both filters, on the interpretive model advanced in this essay, is that the frontline workforce a family encounters in any given safeguarding interaction is composed substantially of practitioners who have remained inside the institution under conditions that the structural-selection model predicts will, over time, retain particular adaptation profiles. The strong claim — that the remaining cohort is "the dissociated and procedural" while the integrated cohort has gone — overstates what direct evidence the literature provides. The weaker claim — that the institutional conditions tend to retain workers compatible with proceduralism and to lose workers whose adaptation style does not compartmentalise the work — is consistent with the documented attrition data and is the version this essay advances.

The clinical and structural argument up to this point has been advanced in the cautious register the academic literature requires. The phenomenon being described is, however, one that does not require academic vocabulary to be seen clearly. People without the technical apparatus see it, and have been seeing it for years, in language that is sometimes more direct than the literature precisely because it does not need to navigate the literature's methodological caveats.

An African-born care worker, employed by a contracted domiciliary provider in a British city, working alongside the safeguarding apparatus of an English local authority, was asked her view of two of the social workers she had been working with. Her answer:

This sentence, delivered by someone with no training in trauma psychology, no professional vocabulary for dissociation, no academic register for the structural arguments of this series, names in eight words what the previous four parts of this essay have spent thousands of words approaching. The clinical literature on dissociation is, on one possible reading, an extended technical paraphrase of this sentence. The body is present. The relational person — the part of someone that registers what is happening to the human being in front of them — is, on the carer's observation, less present than she would expect. The professional functions. The form is completed. The decision is made. The family on the other end of the encounter, on the carer's description, experiences something like a structural absence behind a procedurally-present body.

The reason the sentence is more direct than the literature is that the carer is comparing what she is seeing against a baseline the literature does not have. She comes from a setting in which the people doing care work are, in her description, present in the relational sense as well as the bodily one. The contrast between what she expected and what she encountered is the data. She does not need the trauma research to tell her what she is seeing, because she has the comparative frame the research itself does not have. Whether the practitioners she observed would, on clinical assessment, meet diagnostic criteria for dissociation in the technical sense is a separate question; what she is naming is the relational quality of the encounter, and that is itself the observation the structural argument of this essay turns on.

That she could only deliver the observation in this country, in this register, to someone outside the institution she was observing, is its own structural fact. A British social worker noticing the same thing about a colleague would find few institutional channels available for saying so — and the channels that exist (formal supervision, raising concerns, whistleblowing) lead, on the available evidence, to the second filter described above. The carer's observation is something the carer can say partly because she is not inside the system that would respond to her saying it.

What follows from this

The structural consequence is the point. If the workforce a family encounters has been filtered, on the interpretive model advanced here, across the length of a career, for capacity to do the work under conditions that produce documented secondary trauma without the institutional supports that would allow it to be processed, then no amount of training, guidance, or procedural reform will produce humane outcomes from that workforce in isolation. The workers performing the procedures are the workers who have remained under the conditions; the workers who could not perform them under those conditions have, on the model, disproportionately left or been managed out. Asking the remaining workforce to be more humane is asking them to undo, individually, an adaptation that the institutional conditions have produced — and to do so without the supports that would make the undoing possible. The structural reform that would change this is not at the level of the individual practitioner. It is at the level of the conditions the institution provides for the practitioner.

The reforms proposed across the earlier pieces of this series — the AAIB-equivalent investigation body, restoration of civil legal aid, an Independent Family Advocate, regulated rights of audience for trained McKenzie Friends — each address part of the system. None of them addresses the workforce condition directly. The reform that would address the workforce condition is one that has not been seriously proposed by anyone in the contemporary debate, because its preconditions are politically and financially substantial: properly funded reflective supervision for every safeguarding practitioner, with clinical supervision available to those carrying compound trauma, in a culture that treats the workforce's mental health as a precondition of safe practice rather than as an HR overhead. The Royal College of Psychiatrists has been calling for something analogous in NHS mental health staffing for two decades. The equivalent argument has not, to date, been made forcefully in adult social care. The argument deserves to be made.

The Casey Commission, whose 2026 work has the Health Secretary's agreement to create a National Safeguarding Board, is the operational venue where this argument can land in the near term. The new body, if constituted with the structural features the third piece in this series described, would be in a position to investigate the workforce conditions as well as the procedural ones. The recommendations that flowed from such investigations could include the reflective-supervision-and-clinical-support reforms that the workforce literature has been pointing to for fifteen years. Whether they do — whether the National Safeguarding Board takes the workforce-condition question seriously, or whether it treats the question as outside its remit — is one of the early markers of whether the body has been built in the AAIB mode or in the soft-coordination mode the fourth piece in this series warned against.

The argument of this series across nine pieces has accumulated, in its careful structural way, toward a single picture. Adult safeguarding in England fails because it cannot judge proportionately; it cannot judge proportionately because it has lost the institutional learning apparatus every comparable safety-critical sector has built; it cannot correct its failures because the burden of correction falls on the people least able to bear it; it cannot see the families it serves because its design priors were written by a class whose family life was different from theirs; and, on the structural-selection model advanced in this essay, it tends not to improve from within because the workforce it has produced under sustained operating conditions is, in significant part, the workforce that has remained under those conditions. The picture is consistent. The diagnosis is precise on the structural points and interpretive on the workforce-psychology points, and the essay has tried to mark the difference. The reforms that would address the structural diagnosis are visible. The reform that would address the workforce condition is harder and not yet on the table.

The condition for any of this to change is recognition. The recognition is what the carer's eight words supply, more directly than nine pieces of analysis have, and what no policy document has yet been willing to write down. The body is here. The person is, on her observation, less present than she would expect. Until the institution that has produced this outcome is asked, plainly, why the conditions it sustains produce it, and is required by something independent of itself to answer, the bodies will continue to be present and the persons will continue to be less present, and the families who encounter them will continue to live with the consequences of a system whose human dimension has been, on the interpretive model advanced here, structurally reduced.

Sources

Sources tagged by evidence tier: [R] peer-reviewed research; [S] statutory, regulatory or official source; [J] journalism; [I] interpretive synthesis.

1. [S] Skills for Care (2024). The state of the adult social care sector and workforce in England. Annual workforce intelligence report: https://www.skillsforcare.org.uk/Adult-Social-Care-Workforce-Data/Workforce-intelligence/publications/national-information/The-state-of-the-adult-social-care-sector-and-workforce-in-England.aspx . See also Department for Education (annual), Children's social work workforce statistics: https://www.gov.uk/government/collections/statistics-childrens-social-care-workforce . Social Work England registration data: https://www.socialworkengland.org.uk/

2. [R] The wounded-healer concept descends from Jung, C. G., The Practice of Psychotherapy, Collected Works Vol. 16 (1954). For the contemporary peer-reviewed development, see Wheeler, S. (2007), 'What shall we do with the wounded healer? The supervisor's dilemma,' Psychodynamic Practice 13(3): 245–256; Newcomb, M., Burton, J., Edwards, N., and Hazelwood, Z. (2015), 'How Jung's concept of the wounded healer can guide learning and teaching in social work and human services,' Advances in Social Work and Welfare Education 17(2): 55–69. The Barr 2006 master's dissertation (University of Strathclyde) is also a useful supplementary source on the British context.

3. [R] van der Kolk, B. (2014). The Body Keeps the Score: Mind, Brain and Body in the Transformation of Trauma. Viking. Herman, J. L. (1992). Trauma and Recovery: The aftermath of violence — from domestic abuse to political terror. Basic Books.

4. [R] American Psychiatric Association (2022). Diagnostic and Statistical Manual of Mental Disorders, 5th edition, text revision (DSM-5-TR). Dissociative disorders sections. See also Putnam, F. W. (1997), Dissociation in Children and Adolescents: A Developmental Perspective, Guilford Press.

5. [R] Ferguson, H. (2018). 'How social workers reflect in action and when and why they don't: the possibilities and limits to reflective practice in social work.' Social Work Education 37(4): 415–427: https://www.tandfonline.com/doi/full/10.1080/02615479.2017.1413083 . See also Ferguson, H. (2017), 'How children become invisible in child protection work: Findings from research into day-to-day social work practice,' British Journal of Social Work 47(4): 1007–1023; Ferguson, H. (2010), 'Walks, home visits and atmospheres: Risk and the everyday practices and mobilities of social work and child protection,' British Journal of Social Work 40(4): 1100–1117. Note: Ferguson's primary fieldwork is in child protection rather than adult safeguarding; the translation of his ethnographic findings into adult safeguarding context is an extension of the model rather than directly evidenced research.

6. [S] Munro, E. (2011). The Munro Review of Child Protection: Final Report — A child-centred system. Department for Education, May 2011: https://www.gov.uk/government/publications/munro-review-of-child-protection-final-report-a-child-centred-system . Specifically on the loss of reflective supervision in safeguarding practice, see chapters 4–6.

7. [R] Figley, C. R. (1995). Compassion Fatigue: Coping with Secondary Traumatic Stress Disorder in Those Who Treat the Traumatized. Brunner/Mazel. See also Figley, C. R. (2002), 'Compassion fatigue: Psychotherapists' chronic lack of self care,' Journal of Clinical Psychology 58(11): 1433–1441.

8. [R] Bride, B. E. (2007). 'Prevalence of secondary traumatic stress among social workers.' Social Work 52(1): 63–70: https://academic.oup.com/sw/article-abstract/52/1/63/1937540

9. [R] McFadden, P. and Campbell, A. (2014). Measuring burnout among UK social workers: A Community Care study (Queen's University Belfast / Community Care): https://www.qub.ac.uk/sites/media/Media,514081,en.pdf . See also Doherty, A. S. et al. (2020), 'Measuring Burnout in Social Work: Factorial Validity of the Maslach Burnout Inventory — Human Services Survey,' European Journal of Psychological Assessment — using a UK sample of N=1,257 social workers. The Maslach Burnout Inventory is the standard validated instrument and has been applied to UK social work populations in successive studies; emotional exhaustion has been the most consistently elevated subscale in UK samples.

10. [R] Sundkvist, T. and Zorn, T. (2023). 'Newly Qualified Social Workers' Careers: A Mix of Turbulence, Goals and Safety.' British Journal of Social Work 53(2): 794–812. Anka, A., Thacker, H., Penhale, B., Lloyd-Smith, W., and Booth, B. (2024). Professional Curiosity in Safeguarding Adults. Routledge.

11. [S/J] Social Work England / YouGov (2020), workforce research, reported in Community Care, 'Four in ten social workers anticipate quitting profession within five years,' 4 September 2020. The 48% figure is the survey's headline finding on stated intention to leave within five years; actual leave rates will differ from stated intention.

12. [S/J] Social Work England register data, year to November 2022, reported in Community Care, 30 March 2023: https://www.communitycare.co.uk/

13. [J] Prince, D. (2025). '"No accountability" in fury at council's gagging orders.' Bristol Live / Bristol Post, 25 January 2025. Reporting the answer given at the Bristol City Council member forum meeting of 14 January 2025, in response to a question by Cllr Graham Morris (Conservative, Stockwood). The 59-settlement-agreement figure is from the Council's own answer to the formal question and is in the documentary record. https://www.bristolpost.co.uk/

14. [J] Thomas, N. (2024). 'Councils spent £3.2m on confidentiality deals with staff.' Nation.cymru, 6 February 2024. FOI data from five Gwent councils: https://nation.cymru/

15. [J] York Mix (2021). 'York Council spends £250k in a single year on gagging orders to stop former staff speaking out.' 30 March 2021.

16. [J] Glasgow Times (2019). '£400,000 gagging city staff.' 2 January 2019. FOI data from Glasgow City Council.

17. [J] Postans, A. (2020). 'More work to do on institutional racism at Bristol City Council.' Bristol Cable, 3 June 2020: https://thebristolcable.org/ . Context: complaints raised by BAME staff-led group following the Bijan Ebrahimi independent review.

Compiled · May 2026 · For research and journalism use