A working alliance, as Building the Working Alliance Based on Confidence That Can Be Earned defines it, is measured by what a family and a service actually do together over time — confidence earned through consistent conduct rather than professed goodwill. That measure carries a structural assumption: that the behaviour observed reflects the relationship it is taken to indicate. Disguised compliance is the case in which this assumption fails. A family can attend every appointment, complete every course offered, and answer every question put to it, while the underlying risk to a child remains entirely unaddressed. By every indicator available to a visiting professional, this looks like engagement succeeding. It can, in practice, be engagement that has changed nothing.
This is the confidence measure's hardest case, precisely because it does not present as a problem. A family with low confidence in a service is visibly resistant, and the difficulty is at least apparent to the professional involved. A family performing confidence offers no equivalent signal. Distinguishing the two is not a matter of professional instinct or experience alone; it is a defined problem in UK safeguarding practice, with a specific name, a documented history in serious case reviews, and — as this essay sets out — a growing body of academic critique directed at the concept itself.
What disguised compliance actually is
The term itself was coined in 1993, when the psychiatrist Peter Reder, psychologist Sylvia Duncan and social worker Moira Gray reviewed thirty-five serious child abuse inquiries for their book Beyond Blame: Child Abuse Tragedies Revisited (Reder, Duncan and Gray, 1993). Across that body of cases, they found a recurring pattern: parents responding to professional involvement not by refusing it outright, but by appearing to cooperate in ways that reduced scrutiny and defused the very intervention that had been triggered by concern for the child. Reder and colleagues called this disguised compliance — cooperation offered, consciously or not, as a way of neutralising professional authority rather than as evidence that anything underneath had actually shifted.
The term has since become a standing fixture of UK case-review learning. The NSPCC's own briefing on the subject, drawing on published case reviews from 2014 onward, sets out the pattern in almost identical terms: a parent or carer giving the appearance of cooperating with child welfare agencies to avoid raising suspicion, allay professional concern, and diffuse further intervention (NSPCC Learning, 2019). Both the Climbié and Connelly reviews identify disguised compliance as a contributing feature — professionals recording contact, attendance, and superficial cooperation as reassurance, while the underlying risk to the child continued unaddressed.
YFL's own fictional Midkwell material gives this a face. Kelly Jokden attends appointments whenever her family is on a Child Protection plan. She completes parenting courses. She does not implement anything from them afterward — and, on the account given of her elsewhere in the “Changing People” series, she is not play-acting any of this. She genuinely believes she is a good mother. That detail is not a footnote; it is the essay's actual subject. Disguised compliance is usually introduced to a reader as a deception — something a parent does to a professional. The fictional Kelly Jokden's case asks whether that is always the right way to describe what is happening.
What “risk” actually names
A note on terminology: present harm versus future harm
The language used elsewhere in this essay — that a child's “risk” remains unaddressed while a parent performs compliance — deserves scrutiny, because the word does real conceptual work in child protection discourse, and not always work that helps. “Risk” carries a built-in temporal frame: it names something that might happen, a probability attached to a future event. Statutory language is, in fact, more precise on this point than much everyday professional usage. The Children Act 1989's own threshold criteria distinguish two separate conditions rather than collapsing them into one: a court may act where a child “is suffering, or is likely to suffer, significant harm” (s.31(2)(a)). The first limb is present tense — harm that is already happening. The second is prospective — harm that has not yet happened but might. Practice vocabulary does not always preserve this distinction. “Risk assessment,” “level of risk,” “risk factors” — the language that has grown up around the statute's second limb has a habit of colonising discussion of the first, so that a child living through ongoing neglect or chronic emotional harm is discussed in the same probabilistic register as a child who is not yet, but might in future be, harmed.
Bromfield, Gillingham and Higgins's (2007) concept of cumulative harm names the present-tense version directly: the impact of patterns and circumstances in a child's life that erode their sense of safety, stability and wellbeing, accumulating through repetition rather than announcing itself as a single triggering event. Working Together to Safeguard Children's 2023 statutory guidance now explicitly directs practitioners assessing significant harm to weigh “the cumulative impact of adverse events” — a corrective, in effect, to a risk vocabulary that had drifted toward treating harm as episodic and future-facing by default.
A further complication works in the opposite direction. Whether an ongoing harm is visible in a child's presentation depends heavily on what Rutter (1985) established as the buffering effect of protective factors: at least one secure attachment relationship, access to support beyond the immediate family, and a degree of temperament and circumstance some children simply carry more of than others (Werner and Smith, 1992). A child who is resilient, or well held by a grandparent, a teacher, or an older sibling, can absorb a considerable amount of ongoing harm without showing it — not because the harm is not real, but because something else in that child's world is currently doing the work of absorbing it. Read through a probabilistic risk lens, an absence of visible impairment looks like evidence that the feared outcome has not materialised. Read through a cumulative-harm lens, it can just as easily mean the opposite: that harm is ongoing and is being masked by protective buffering that may not hold indefinitely.
This distinction bears directly on disguised compliance, because it supplies the second condition under which the pattern can go unchallenged for a long time. A family whose compliance is entirely performed, paired with a child who happens to have strong protective factors elsewhere in their world, together produce exactly the profile a probabilistic risk assessment is least equipped to catch: no episode to point to, no visible deterioration in the child, and a parent who, on every observable measure, appears to be engaging. The absence of visible harm in this scenario is not reassurance. It may be a resilient child and a protective grandparent doing, quietly and precariously, what the compliance in the room was supposed to be doing and is not.
Why it is not always deliberate
Reder and colleagues' original formulation, and much of the case-review literature that followed it, is written from the professional's vantage point: here is a pattern to notice, here is how it can fool you, here is what it costs the child when it does. That framing is necessary and has almost certainly protected children in the three decades since — but it also tends to assume a fairly clear-eyed actor on the other side of the table, someone choosing to perform cooperation as a strategy. The fictional Kelly Jokden's case suggests that assumption does not always hold, and it turns out the academic literature itself has already begun to say so.
Patricia Crittenden's work on what she terms compulsive compliance offers one route into why. Crittenden and DiLalla (1988) describe how some children, particularly those who have experienced neglect or maltreatment, learn to suppress their own needs and become hypervigilant to what an authority figure expects, as a survival strategy rather than a conscious calculation — and this pattern, later developed as part of Crittenden's dynamic-maturational model of attachment (Crittenden and Landini, 2011), does not simply disappear in adulthood. A parent who was themselves raised to read and satisfy an authority figure's expectations before that figure could turn dangerous is not necessarily deploying a new strategy when a social worker arrives at the door. They may be running the only script they have ever had for managing exactly this kind of scrutiny — outwardly compliant, inwardly unreachable by what is actually being asked of them, because compliance was never, in their own history, about the content of the request.
More directly still, Leigh, Beddoe and Keddell (2020) subjected the concept of disguised compliance itself to critical discourse analysis and found real reason for caution about how the term gets used in practice. Their argument is not that the underlying behaviour Reder and colleagues identified is imaginary, but that labelling a parent's response “disguised compliance” can function as a way of pre-judging intent — reading an outcome (concerns not resolving) back into a claim about motive (the parent never meant to cooperate) that the available evidence may not actually support. A term coined to sharpen professional vigilance can, used carelessly, do the opposite: close down curiosity about why cooperation is not translating into adaptation, by supplying a ready-made, already-suspicious explanation before anyone has actually established one.
Taken together, this suggests disguised compliance is not a single behaviour with a single explanation. It can be produced by at least two quite different mechanisms — deliberate strategy in one case, a learned, largely automatic response to being watched and judged in another — while looking identical from the outside in both. YFL's existing Learning to Survive framework, which treats certain patterns of behaviour under scrutiny as adaptive survival strategies shaped by a person's own history rather than assembled as a conscious tactic in the moment, sits comfortably alongside Crittenden's account of compulsive compliance specifically.
There is a third mechanism worth adding to these two, and it does not depend on any history of childhood maltreatment at all. YFL's own essay The Epistemology of Safeguarding makes the case directly: a parent's apparent cooperation can be rational self-protection under present-tense interpretive uncertainty, not a strategy and not an old survival pattern, but a reasonable response to a genuine unknown. The essay's own illustration is worth setting out, because it does not rest on any special vulnerability in the parent at all. Put a professional with twenty years inside the system in the parent's chair, and ask whether they would be fully open with a visiting social worker. The honest answer is no — not from any wish to deceive, but because there is no way to know in advance how their own exhaustion, frustration, or an ordinary explanation for an ordinary thing will be read on the other side of the table. If someone with full insider knowledge of how these judgements are made cannot predict how their own honesty will land, an ordinary parent, with none of that knowledge, has even less basis for assuming that transparency is safe. What follows is not deception but a strategic decision about what to share and how to present it, made because one person's interpretation of the encounter becomes institutional fact, not because the parent set out to manage the professional.
This gives disguised compliance an entirely different shape from the one the term itself suggests. It is not a family managing a professional; it is a family managing an asymmetry the professional relationship itself creates — full disclosure asked of one side, opaque interpretation held by the other. Read this way, the term's own critics have a real point. The NSPCC's more recent guidance on the concept cautions directly against overusing it, on the grounds that it can create an atmosphere of suspicion rather than collaboration, and can obscure a professional's own role in failing to build the relationship that would have made openness safer (NSPCC Learning, 2024). Forrester and colleagues' research on parental resistance backs this from the other direction: what looks like non-cooperation frequently traces back to inadequate communication on the professional's side rather than to concealment on the family's (Forrester et al., 2012). Labelling a parent “disguised compliant” can, on this account, do the opposite of what the term was coined to achieve — closing down the very curiosity it was meant to prompt, by supplying a ready explanation for the behaviour before anyone has actually established what is producing it.
That behaviour is not always withdrawal. It is easy to picture disguised compliance as a family gone quiet — offering little, volunteering nothing, present but closed. In practice it is at least as often the opposite: warm smiles, enthusiastic nodding, a parent who telephones between visits with questions, who seems, if anything, more engaged than most. This presentation is arguably the harder of the two to read correctly, because it resembles exactly the alliance a confidence measure is meant to detect. Where it is Crittenden's compulsive compliance at work — the learned, hypervigilant eagerness to satisfy whoever holds authority, carried forward from the parent's own childhood and described earlier in this essay — the warmth is not evidence that a working relationship is forming. It is the same adaptive pattern already discussed, wearing a friendlier face than silence does, and no easier to distinguish from the real thing for being pleasant to sit across from.
This matters practically, not just as a point of nuance. A parent who is consciously managing professional attention responds differently to being challenged than a parent who has no clear internal account of what they are doing, and differently again from a parent who is making a reasoned, defensible calculation about what it is safe to disclose. Treating all three as the same problem, to be solved the same way, risks getting the response wrong in every direction: too suspicious of a parent who is not gaming the process, not curious enough about a parent who is making a rational bet about how they will be read, and not searching enough for the parent who genuinely is concealing risk.
None of this is an argument that disguised compliance is therefore harmless, or that professional vigilance about it should soften. YFL's own house position — feed the solution, starve the problem, adaptation rather than instruction — does not extend to concluding that a child's safety can wait on a full explanation of why a parent's cooperation is not translating into anything real. The point is narrower, and it is the point The Epistemology of Safeguarding and Leigh, Beddoe and Keddell both make in their own ways: correctly identifying disguised compliance as sometimes non-deliberate changes what a professional does next — it asks for more curiosity about mechanism and more honesty about the asymmetry the relationship itself creates, not less concern about outcome — but it does not change whether they act.
What the fictional Thomkden case adds
The fictional Kelly Jokden's case shows disguised compliance succeeding against a single visiting professional, over an extended period. The Thomkdens, another fictional family from the same Midkwell material, show what happens when it succeeds against the system's own final safeguard.
The fictional Dean Thomkden presents with what is described elsewhere in the Changing People material as a narcissistic presentation — confident, articulate, persuasive in exactly the register a courtroom rewards. The fictional Jade Thomkden's presentation is close to the opposite: dissociated, withdrawn, offering little that a judge can easily read. Between them, a family court judge was persuaded to refuse a Care Order and instead direct that the family receive “more support” — this despite the social worker on the case already knowing the family to be dangerous.
What makes this fictional case worth its own treatment, rather than a more extreme version of the fictional Kelly Jokden's, is what it reveals about where disguised compliance can succeed. It is not only, or even primarily, a matter of fooling an inexperienced or under-resourced visiting worker. A confident, articulate presentation is itself a performance skill, and the courtroom is a setting built to reward exactly that skill regardless of its source. The system's own procedural safeguards — expert reports, the presumption in favour of keeping a family together wherever safely possible, the instinct to give a family “one more chance” before the most serious order available — can be turned into further camouflage by a parent skilled enough to use them. The very features of the system designed to protect against injustice toward parents are not immune to being used by a parent working, consciously or not, against the child's interest.
This is the sharpest version of the caveat this essay exists to make. An earlier essay in this series describes a working alliance as something that “resists being faked for long by either party” — and for the great majority of cases, across enough contact and enough time, that holds. The fictional Thomkden case is the demonstration that it does not hold universally, or does not hold quickly enough. A performance skilled enough, deployed in front of an audience built to be persuaded by exactly that skill, can be faked for far longer than a single home visit, and can be faked all the way up to and including the point of highest formal scrutiny the system has.
How good practice tells the difference
If disguised compliance cannot reliably be told apart from a genuine working alliance by what happens at the front door, or even by what happens in a courtroom, the question becomes what can distinguish them. UK case-review learning offers a small number of consistent markers, none of them individually conclusive, but considerably more reliable in combination than any single visit or hearing can be on its own.
Consistency over time, rather than concentration around review points. Genuine adaptation tends to be visible unevenly across the whole period between reviews. Disguised compliance tends to cluster conspicuously in the run-up to a scheduled visit, a case conference, or a court hearing — a sudden increase in school attendance, a run of kept appointments, a flurry of engagement that does not extend backward or forward from the date that mattered. The NSPCC's own briefing lists exactly this clustering as one of the pattern's recognisable signs.
Corroboration from more than one source. A parent's own account, however consistent and however convincingly delivered, is one account. The fictional Thomkden case shows what happens when a single persuasive presentation is allowed to substitute for corroboration. Good practice cross-checks a family's own account against school attendance records, health visitor contact, extended family, and any other source with independent visibility into the same household — not because a family's account is assumed to be false, but because disguised compliance specifically targets the credibility of that single account, and only that account, as its route to reducing scrutiny.
The child's own presentation, which tends to outlast a parent's performance. A parent can manage what a professional sees of themselves for the length of a visit. A child's presentation is harder to manage in the same way, and tends to keep communicating what a household's spoken account does not. Destiny's scratching and Kenzie still being in nappies at five — both details from the same fictional Midkwell material — are exactly this kind of signal: neither is something a parent's performance in the room can straightforwardly override, because neither depends on what the parent says.
Professional curiosity, sustained rather than satisfied. Munro's 2011 review of child protection practice named the erosion of what she called professional curiosity — a readiness to keep asking why an account holds together, rather than accepting a plausible one at face value — as a recurring feature of cases where disguised compliance went unchallenged for too long. This is not a demand for permanent suspicion of every family; it is a discipline of continuing to test an account that fits, rather than treating its fit as itself the evidence of adaptation having occurred.
None of these four markers is a test a family can pass or fail at a single point in time, which is itself part of the point. Disguised compliance specifically exploits any measure that can be satisfied in a single sitting. A measure built instead around consistency, corroboration, the child's own presentation, and sustained curiosity is harder — not impossible, as the fictional Thomkden case shows, but harder — to perform against successfully for as long as it would need to be performed against to matter.
What this means for the confidence measure
An earlier essay in this series sets out working alliance and confidence as the right replacement for trust as the construct underpinning good practice in child protection contact — forward-looking, behaviourally groundable, and not requiring a family to relitigate whoever has failed them before. Nothing in this essay argues against that replacement. What it adds is a limit on what the measure alone can tell a professional, at any single point in time, about whether a working alliance is real.
A working alliance built through consistent, corroborated, sustained contact over time is a genuinely different thing from an appearance of one produced for the length of a visit or a hearing. The two can look identical at any single point of observation. They generally do not look identical once observed the way the markers above require — across time, across sources, and against a child's own presentation rather than a parent's account of it alone. The practical implication is not that confidence-based practice should be abandoned or treated with generalised suspicion; it is that confidence, like any other measure that can be gamed, needs to be corroborated rather than simply observed at the door.
Close
An earlier essay in this series closes on the image of two families, the same knock, different responses — one door opening slowly, earned over time; another closing out of habit, formed by history rather than by anything this particular service has done. Disguised compliance describes a third door, harder to place in that picture than either of the first two: a door that opens readily, smiles, offers tea, and changes nothing behind it. It is not the easiest case for a confidence measure to handle. It may be the case that measure most needs to be built to withstand.
References
Bromfield, L., Gillingham, P. and Higgins, D. (2007). Cumulative harm and chronic child maltreatment. Developing Practice, 19 (Winter/Spring), 34–42.
Children Act 1989, s.31(2), s.31(9)–(10). London: HMSO.
Crittenden, P. M. and DiLalla, D. L. (1988). Compulsive compliance: the development of an inhibitory coping strategy in infancy. Journal of Abnormal Child Psychology, 16(5), 585–599.
Crittenden, P. M. and Landini, A. (2011). Assessing Adult Attachment: A Dynamic-Maturational Approach to Discourse Analysis. New York: W. W. Norton.
Department for Education (2023). Working Together to Safeguard Children. London: HM Government.
Forrester, D., Westlake, D. and Glynn, G. (2012). Parental resistance and social worker skills: towards a theory of motivational social work. Child & Family Social Work, 17(2), 118–129.
Healy, K. (2000). Social Work Practices: Contemporary Perspectives on Change. London: Sage.
Leigh, J., Beddoe, L. and Keddell, E. (2020). Disguised compliance or undisguised nonsense? A critical discourse analysis of compliance and resistance in social work practice. Families, Relationships and Societies, 9(2), 269–285.
Munro, E. (2011). The Munro Review of Child Protection: Final Report — A Child-Centred System. Cm 8062. London: Department for Education.
NSPCC Learning (2019). Disguised Compliance [Learning from Case Reviews briefing]. London: NSPCC.
NSPCC Learning (2024). Why Using the Term 'Disguised Compliance' Can Be Problematic. London: NSPCC.
Reder, P., Duncan, S. and Gray, M. (1993). Beyond Blame: Child Abuse Tragedies Revisited. London: Routledge.
Rutter, M. (1985). Resilience in the face of adversity: protective factors and resistance to psychiatric disorder. British Journal of Psychiatry, 147, 598–611.
Spratt, T. and Callan, J. (2004). Parents' views on social work interventions in child welfare cases. British Journal of Social Work, 34(2), 199–224.
Tobis, D. (2013). From Pariahs to Partners: How Parents and Their Allies Changed New York City's Child Welfare System. Oxford: Oxford University Press.
Werner, E. E. and Smith, R. S. (1992). Overcoming the Odds: High Risk Children from Birth to Adulthood. Ithaca: Cornell University Press.
Topics: #DisguisedCompliance #ChildProtection #Safeguarding #WorkingAlliance #ProfessionalCuriosity #SocialWork #ChildrenAct1989 #CumulativeHarm