Guilty until proven exhausted

The way a wrongly-investigated family loses is, on the available evidence, often not by formal adjudication. It loses by attrition. The council opens an enquiry. The family contests it. The family runs out of money, or time, or energy, or all three, before the matter is formally resolved. The file closes — sometimes with the council recording its findings, sometimes without — and the family is left with whatever they have left. The story is not in the case law. The case law is the small minority of disputes that reached a judgment. The story is in the much larger population of cases that did not, because one side ran out of capacity to fight and the other did not. This is the failure mode that does not appear in safeguarding's official statistics, that no Safeguarding Adults Review will ever investigate, that no Care Quality Commission inspection will ever flag, and that determines, more than most procedural rules, who actually loses.

The architecture of this attrition is the subject of this essay. It is composed of three structural features, each well-documented, none usually discussed together. First, the legal burden of proof in safeguarding is far lower than most people understand, and operates in ways that effectively reverse the lay presumption of innocence. Second, the access-to-justice apparatus that would historically have allowed families to contest unfounded investigations has been substantially dismantled over the last fifteen years, with English family-court legal aid the most badly affected area of an already-eviscerated system. Third, the population of people on whom this combination falls — informal carers for vulnerable adults — is the population least able to absorb additional administrative, financial and psychological burden, because they are already, in the technical sense, exhausted. The three features together produce a system in which being right is, in practice, often no defence at all.

A note on what this essay is and is not. The asymmetric burden described here is one half of a two-sided failure pattern. The other half — the under-reach failure of cases like Steven Hoskin's, where the same regime fails to act against genuine abuse — is examined in the accompanying case study and diagnostic essay in this series. The point of the present piece is not that safeguarding teams should investigate less, or that abuse and exploitation by family members and others are not real (they are, and the social cost of under-reach is severe). The point is that the procedural architecture by which contested safeguarding investigations are resolved imposes its cost on the population least able to bear it, in a way the system does not measure and that no current accountability mechanism is structurally designed to surface.

The burden of proof, accurately stated

The presumption of innocence in English law applies to criminal proceedings. A defendant is innocent until proven guilty beyond reasonable doubt, and the burden of proof rests on the Crown. This is the version of the principle that lay readers know. It is one of the central architectural features of the common-law tradition and is rightly defended.

It does not apply to safeguarding.

A Section 42 enquiry under the Care Act 2014 is a civil-administrative process. Its findings are made on the balance of probabilities, the civil standard.^[1] This is itself unremarkable; almost all administrative processes use the civil standard. What is more consequential is the procedural geometry that surrounds the standard. The local authority is the investigator, the prosecutor and the decider. The accused family member, if they are an adult relative or carer rather than a regulated professional, has no automatic right to representation, no automatic right to advocacy, no automatic right to see the evidence against them in advance, and no automatic right of appeal to an independent body. The Local Government Association's own training materials for section 42 decision-making, used in local-authority safeguarding workshops across England, are explicit about the operational principle that follows from this:

"Not always necessary in safeguarding situations to determine the truth of every allegation if there is sufficient evidence to justify lawful intervention."

— Local Government Association, Decision-making: Section 42 Safeguarding Adults Enquiries — One day workshop, collated slides, 28 November 2018.^[2]

Read that sentence twice. The system does not need to determine whether an allegation is true. It needs only to determine whether there is enough material to justify the intervention it has chosen. The threshold for triggering the intervention is "reasonable cause to suspect," set out in section 42(1) of the Care Act. Once triggered, the enquiry's purpose is not to find out whether the allegation was right; it is to decide what action to take. The action can include referral to the Disclosure and Barring Service, removal of the cared-for adult from the home, refusal of Direct Payments, application to the Court of Protection for authorisation of deprivation of liberty, and (in cases involving regulated professionals) referral to a disciplinary body. None of these actions requires a finding that the original allegation was true. Each is justified by the council's view that the action is in the cared-for adult's best interests, taking the safeguarding concerns into account.

This is not, in itself, a scandal. There are good reasons why safeguarding regimes operate on a lower threshold than criminal law. The harms they are designed to prevent are often irreversible, the evidence is often circumstantial, the people they protect are often unable to give the testimony a criminal trial requires. A system that demanded "beyond reasonable doubt" before taking protective action would fail catastrophically and predictably. The threshold has to be set somewhere lower than that. And procedural fairness obligations apply: the same LGA workshop materials note that public-law principles require local authorities to give people an opportunity to put their case, citing R(AB and CD) v Haringey London Borough Council [2013] as authority, with departures from that requirement needing "very careful justification."^[2]

What is consequential is what happens at the other end — when the family on the receiving end of an enquiry believes the council has it wrong and wants to contest. The structural choice the English system has made is that contestation is not symmetric with investigation. The council can investigate at no marginal cost to itself; the family can defend itself only by accessing a legal apparatus that has, since 2013, been progressively closed to it.

The dismantling of access

The Legal Aid, Sentencing and Punishment of Offenders Act 2012 (LASPO) came into force on 1 April 2013. Its effect on civil legal aid was profound and is well-documented in academic, parliamentary and regulatory literature.^[3] The relevant numbers, in headline form:

The official policy justification for LASPO was a £350 million annual saving on the legal aid budget, achieved by removing whole categories of civil law from "scope" — that is, no longer eligible for legal aid funding. Family law in particular was taken almost entirely out of scope, except where domestic violence could be evidenced (a requirement which has its own well-documented set of access problems^[6]). Welfare benefits, housing, employment, debt and clinical negligence were heavily affected. Areas of law where the state is, by definition, the most powerful party — and where citizens most need representation to assert basic rights against it — were the areas most heavily removed from public funding.

Sir James Munby, then President of the Family Division of the High Court, was unusually direct in his criticism of LASPO and what followed. In a 2014 adoption case, Re D (A Child), where parents facing the permanent removal of their child found that legal aid had been refused, his assessment was that the absence of representation in cases of that gravity raised serious questions about the compatibility of the regime with Article 6 of the European Convention on Human Rights.^[7] His broader assessment, delivered repeatedly over his tenure and after his retirement, was that LASPO had turned the family court into an increasingly lawyer-free zone, with growing numbers of litigants appearing unrepresented and without legal advice.^[8] The withdrawal of legal aid from private family law disputes was, in his framing, the central failure.

The Court of Protection is, formally, a different jurisdiction from the family court. In practice it is subject to the same access constraints. Cases concerning deprivation of liberty for adults lacking capacity are heard there; the legal complexity is substantial; legal aid is means-tested and merit-tested and, for adult relatives or carers not personally accused of a criminal offence, very rarely granted. The result is that families who want to contest a Court of Protection application — for example, an application by a local authority for authorisation of arrangements they believe are not in the cared-for person's best interests — usually do so as unrepresented litigants, often with the assistance of a McKenzie Friend.^[9]

McKenzie Friends are the workaround the system has produced for the access gap. They have no automatic right of audience and cannot ordinarily conduct litigation, under the judicial Practice Guidance (McKenzie Friends: Civil and Family Courts) issued by Sir Nicholas Wall as President of the Family Division in July 2010 and still the governing framework.^[10] Their ordinary role is to assist a litigant in person by taking notes, prompting, offering moral support, and explaining what is happening; courts may, in limited circumstances and on application, grant rights of audience, but this is the exception rather than the rule. The Society of Professional McKenzie Friends exists to professionalise the role; charges typically run between £30 and £80 an hour against solicitor rates of £200–£350 plus VAT.^[10] Many work for free. They are, in practice, the only legal support available to most unrepresented families in safeguarding-adjacent proceedings, and the volunteer end of the McKenzie Friend ecosystem is the closest thing to a civil legal aid system the country still has.

The combined effect is that the apparatus for contesting a wrongly-opened safeguarding enquiry, or a Court of Protection application that the family considers disproportionate, is — for most families — not available at the level of professional legal representation that an outsider would assume is the minimum required to make the system fair. The accused family can have a McKenzie Friend, can read the procedure rules, can write submissions, can fill in the forms. They cannot, in any practical sense, fight the case on equal terms with a council that has dedicated lawyers, dedicated procedures, and the institutional time to wait.

Administrative burden as policy choice

The academic apparatus that names this mechanism most precisely is not in social-care literature. It is in American public administration scholarship: specifically, Pamela Herd and Donald Moynihan's 2018 book Administrative Burden: Policymaking by Other Means, expanded in the Russell Sage Foundation series and now standard in public-policy graduate education.^[11]

Their framework decomposes administrative burden into three categories of cost imposed on the citizen by the state:

• Learning costs — the time and effort required to understand a programme's existence, its eligibility rules, its procedural requirements, and the language in which they are stated. For a family contesting a safeguarding enquiry, this includes reading the Care Act, the Mental Capacity Act, the relevant council procedures, the Court of Protection rules, the Family Procedure Rules where relevant, and the procedural decisions in their own case.
• Compliance costs — the time, effort and money required to file documents, attend meetings, provide evidence, and produce the procedural outputs the system demands. For a family in safeguarding proceedings, this can involve hundreds of hours of work spread across many months — and, in contested Court of Protection welfare cases, often spanning eighteen months to three years, on the case-length patterns reported in Court of Protection annual reports.^[12]
• Psychological costs — the emotional and cognitive cost of dealing with the uncertainty, the perceived hostility, the stigma of being investigated, and the cumulative stress of contesting a decision made by a body with more power than you have. Herd and Moynihan identify this as often the largest of the three categories, and the most likely to cause eligible citizens to give up.

Herd and Moynihan are explicit that administrative burdens are sometimes deliberately designed to deter take-up. American welfare reform in the 1990s is their canonical case: the work requirements, documentation rules, and recertification cycles introduced by the 1996 Personal Responsibility Act were, they argue, designed at least in part to reduce caseloads by making participation arduous enough that some eligible people would drop out. The same mechanism is identifiable in many areas of US policy where, formally, eligibility is unchanged but participation has collapsed.

The English safeguarding system is not, as far as the available evidence shows, deliberately designed to deter contestation. Nobody sat down in the drafting of the Care Act and chose burden levels with the goal of suppressing family resistance. But Herd and Moynihan's framework does not require deliberate design. It captures the structural effect: a system in which the cost of participation falls on the citizen, and falls in a pattern that disproportionately excludes those least able to bear it, will produce predictable outcomes whether or not anyone intended them. The intent is not the mechanism. The mechanism is the asymmetry of cost.

This is the asymmetric burden in its formal shape. It is not unique to safeguarding. The same architecture characterises welfare benefits appeals, immigration cases, housing disputes, and a long list of other proceedings in which an individual contests a state decision. What makes the safeguarding case distinct is the identity of the people on whom the burden falls.

The population on whom this falls

The people contesting wrongful safeguarding investigations are, by definition, family carers for vulnerable adults. The empirical literature on the health and capacity of this population is unambiguous. Carer burnout — defined as physical, mental and emotional exhaustion produced by sustained caring responsibilities — is an extensively documented clinical phenomenon. Carers UK's State of Caring 2023 survey, drawing on responses from over 12,000 carers across the UK, found that 49% of unpaid carers reported a mental or physical health condition that had worsened over the previous year, with over a quarter reporting bad or very bad mental health.^[13] Successive years of the survey have shown a consistent picture: roughly half of UK carers face documented health disadvantages produced by, or substantially compounded by, their caring responsibilities.

The clinical literature describes carer burnout as occurring in identifiable stages, with chronic exhaustion — exhaustion that does not respond to a single rest or a single good night's sleep — characterised by emotional detachment, cognitive impairment, increased physical illness, sleep disturbance, and a sustained inability to manage tasks that would, in a non-burned-out state, be straightforward.^[14] The clinical literature describes the condition explicitly as a state in which sustained additional cognitive load becomes very difficult to absorb.

Now apply Herd and Moynihan's three cost categories to a person in this state. The learning cost — reading and assimilating the Care Act, the Mental Capacity Act, the procedural rules — requires extended focused cognitive work. Burnout produces a measurable deficit in exactly the cognitive operations that work requires. The compliance cost — filing documents, attending meetings, gathering evidence, producing submissions — requires sustained executive function and time. Burnout produces a deficit in executive function and consumes the time that might otherwise be available, because the underlying care work continues. The psychological cost — bearing the stigma of being investigated, the uncertainty about outcome, the anticipated consequences of losing — produces an emotional load that compounds the existing emotional load of the care work itself.

The system has selected, for the imposition of these costs, the population least able to bear them. This is not a metaphor. It is the structural fact. A family caring for an adult with dementia, intellectual disability, post-stroke cognitive impairment, severe mental illness or a similar condition is, by the time a safeguarding investigation lands on them, frequently already operating at or near the limit of their absorptive capacity. Any additional burden — including the burden of contesting the investigation — comes out of a fixed reserve that is, on the survey evidence, already substantially depleted in roughly half of all UK carers before the investigation begins.

The behavioural prediction is straightforward. A non-trivial proportion of families in this position will, regardless of the merits of their case, eventually stop fighting. The cost of continuing exceeds their capacity. The exit may take the form of accepting a council's proposed care plan even though they consider it wrong; of agreeing to a Section 42 finding even though they dispute it; of consenting to a Court of Protection arrangement they regard as disproportionate; of withdrawing a complaint; or simply of allowing the process to run its course without the energy to contest each procedural step. In each case, the file closes. In each case, the system records the outcome as agreement, compliance, or non-contestation. In none of these cases does the official statistical record capture what actually happened, which is that one side ran out of capacity to fight and the other did not.

The Local Government Ombudsman as a partial proxy

The Local Government and Social Care Ombudsman's data, which the second piece in this series analysed in detail, contains one statistic that bears directly on this argument. The LGO's published annual review notes that self-funders — people paying for their own adult social care — make up roughly 23% of the care user population, but submit only 11% of the complaints the Ombudsman receives.^[15] Self-funders are, on average, materially better-resourced than people receiving council-funded care. They are the population most likely to have the capacity to complain. They are also the population most likely to have other options — alternative providers, private legal advice, family advocacy — and therefore the population least dependent on the formal complaints process.

The 23%/11% disparity is, on the most natural reading, the empirical footprint of the attrition mechanism, visible in one specific dataset. It does not prove that the gap between the population share and the complaints share represents people who experienced poor service and gave up. It is consistent with that, and consistent with other explanations (self-funders may have less interaction with statutory processes, may complain through other routes, may be older and less complaint-inclined as a generation). But every comparable dataset, in every adjacent policy area, shows the same pattern: the gap between people who experience a problem and people who formally complain is large, is systematically biased against those with fewer resources, and produces an official statistical record that systematically undercounts harm.

The LGO itself is candid about this. Its own commentary on its complaint data routinely notes that the headline figures capture only the cases that reached it, which is a subset of cases that survived the council's two-stage internal complaints process, which is a subset of cases where the complainant knew the process existed and had the energy to pursue it. The eighty per cent national uphold rate among investigated complaints is therefore a measurement of fault in the subset of cases brought by people determined and resourced enough to fight. Whatever the rate is among everyone who had a bad experience and gave up, it is presumably much higher than that — and we have no way to measure it.

The legal-philosophical question

It would be possible to defend the current arrangement on the grounds that the rule of law does not, in any administrative-civil context, require equality of arms. The state is permitted to be more powerful than the citizen in its interactions with the citizen, because the state has functions the citizen does not, and these functions require operational latitude. The European Convention on Human Rights, in Article 6, requires a fair hearing, not an equally-resourced one. The lay presumption of innocence is a feature of criminal law, not of every state-citizen interaction.

This defence is real, and not foolish. It is also incomplete. The substantive question is not whether the state is permitted to be more powerful; it is whether the asymmetry, as currently configured, produces outcomes that are systematically wrong in identifiable ways. The argument of this series is that it does. The corporate-stupidity diagnosis explains why a system that has lost its capacity for proportionate judgement produces both under-reach and over-reach errors; the comparative LGO data documents the scale of the resulting fault-finding by an independent adjudicator; the AAIB-transplant proposal addresses the failure to learn. The asymmetric-burden argument set out in this essay completes the picture by explaining why the system's wrong decisions are, in practice, very rarely corrected: because the people on whom they fall are structurally selected for inability to mount the correction.

This produces a particular philosophical scandal. A system that gets some decisions wrong is unremarkable. A system that gets a substantial fraction of its decisions wrong, lacks any mechanism for systematically identifying which ones, and imposes the burden of correction on the people least capable of mounting it — that is a system whose self-described function (protection of the vulnerable) and operational function (administrative throughput at minimal challenge) have decoupled. The protection language remains; what has emerged in its place is something else: an institutional equilibrium oriented, in the technical systems-theory sense, toward the system's own continued operation rather than toward the function it was designed for. This is not, on the available evidence, anyone's design — no one decided that safeguarding should function this way. It is the equilibrium that the structural choices made over the last fifteen years have produced. The function the system performs is, in operation, distinct from the function it states. That is a structural diagnosis, not a moral accusation.

LASPO was not designed to harm carers; it was designed to save £350 million on the legal aid budget. The threshold standards in the Care Act were not designed to disadvantage families; they were designed to allow protective intervention in cases where formal proof would arrive too late. The Court of Protection is not designed to be inaccessible; it is designed to handle complex capacity decisions with the procedural care they require. Each individual choice has its own defence. The combined effect of all of them — the system the choices produce — is what this essay has tried to describe.

What would fix it

The structural reform proposed in the supporting article earlier in this series — an independent investigation body modelled on the Air Accidents Investigation Branch, with statutory powers, safe-space testimony, family-initiated investigation rights and sector-wide propagation — addresses the failure to learn but does not, on its own, address the asymmetric burden. The two are different problems and require different fixes. An AAIB-equivalent investigates after the fact; the asymmetric burden is what determines whether a wrongly-opened case is ever corrected at the time. Both reforms are needed.

The fixes for the burden side of the problem fall into three categories, each with precedents in adjacent areas of English law and policy.

Restore civil legal aid in safeguarding-adjacent proceedings

The simplest move, and the most politically resistant. Legal aid for adult relatives or carers facing Section 42 enquiries, Court of Protection applications, or DBS referrals based on safeguarding concerns would restore a baseline of representation that LASPO removed. The means and merits tests would still apply; the gateway is not what is wrong with the current system. What is wrong is that even meeting the means and merits tests no longer brings safeguarding-adjacent cases into scope. Bringing them back into scope would be a legislative amendment of modest scale and substantial effect. The cost is real but not prohibitive; the £350 million annual saving LASPO sought has, on most independent assessments — including the February 2024 NAO report which found the Ministry of Justice "lacks an understanding of the full costs and benefits of its reforms" — been substantially offset by knock-on costs to other parts of the justice system and to the safeguarding bodies whose decisions go uncorrected.^[5]

Independent advocacy for the accused

The Care Act already requires local authorities to arrange an Independent Mental Capacity Advocate for the cared-for person where they would otherwise be unsupported. There is no equivalent statutory provision for the accused carer or family member. Extending the IMCA-equivalent right to provide an Independent Family Advocate for the accused — at the local authority's expense, with statutory powers to access records, attend meetings, and submit representations on behalf of the family — would address a substantial fraction of the asymmetric burden problem at a fraction of the cost of full legal representation. Several Western European systems already operate something analogous; the principle is well-established.

A regulated right of audience for Court of Protection McKenzie Friends

The current McKenzie Friend framework was developed for civil cases in which the litigant is broadly competent and assistance is supplementary. It is not adequate for capacity proceedings, where the litigant's capacity may itself be in question or where the litigant is acting on behalf of an incapacitated relative. The Society of Professional McKenzie Friends has, since 2014, called for a regulated right of audience in Court of Protection proceedings, with mandatory insurance and qualification requirements. The Legal Services Board's ongoing review of lay representation in legal services is the relevant policy vehicle.^[16] A regulated, insured, capped-fee right of audience for trained advocates in safeguarding-adjacent Court of Protection proceedings would substantially close the access gap without restoring the full legal-aid apparatus.

None of these reforms is in the current programme of government. The first — restoration of civil legal aid — is the largest and would have the most effect; it is also the least likely to be enacted in the next parliamentary cycle. The second — the Independent Family Advocate — is the smallest, the most achievable, and could be introduced through amendment to the Care Act regulations rather than through primary legislation. The third — the regulated McKenzie Friend — depends on the LSB's review and the willingness of the senior judiciary to support a managed expansion of rights of audience.

The structural argument of this series is that adult safeguarding in England fails in two opposite directions because it has lost its capacity for proportionate judgement; that the regime's inability to learn from its failures is structurally caused by the absence of an independent investigation body of the kind every comparable safety-critical sector has built; and that the regime's wrong decisions are very rarely corrected because the people on whom they fall are structurally selected for inability to mount correction. The asymmetric-burden argument is the hardest of these to land in public discourse, because its primary evidence is the population the system has already filtered out. Cases that make the news are either the under-reach failures that ended in death, or the wrongly-investigated families with the resources to take their case public. The much larger population that this essay has tried to characterise — families who lost not by adjudication but by attrition, whose cases never reached an outcome because they ran out of capacity before the system did — is structurally invisible. The point of naming the mechanism is that it is the mechanism by which a largely unmeasured proportion of safeguarding disputes are actually resolved.

The lay phrase for this is the title of the essay. Guilty until proven exhausted. The system does not, in practice, require evidence to find against a family. It requires only that the family give up first.

Sources

Sources tagged by evidence tier: [R] peer-reviewed research; [S] statutory, regulatory or judicial source; [J] journalism; [C] community/anecdotal.

1. [S] The civil-standard rule for section 42 enquiries is set out in successive local-authority adult-safeguarding procedures; see, for example, Trafford Adult Safeguarding Procedures, Section 4: Safeguarding Enquiries (2024–25): https://www.staffordshire.gov.uk/ssaspb/safeguarding-procedures-section-42/section-4-safeguarding-enquiries . The statutory framework is at Care Act 2014, section 42: https://www.legislation.gov.uk/ukpga/2014/23/section/42 . Statutory guidance: Department of Health and Social Care, Care and Support Statutory Guidance, chapter 14.

2. [S] Local Government Association, Decision-making: Section 42 Safeguarding Adults Enquiries — One day workshop, collated workshop slides, 28 November 2018: https://www.local.gov.uk/sites/default/files/documents/S42%20collated%20slides%20MC%20final%20version%20for%2028%20Nov%202018.pdf . The quoted passage on the operational principle ("not always necessary in safeguarding situations to determine the truth of every allegation if there is sufficient evidence to justify lawful intervention") and the procedural-fairness reference to R(AB and CD) v Haringey London Borough Council [2013] are both in this document. See also LGA, Making Decisions on the Duty to Carry Out Safeguarding Adults Enquiries (2024): https://www.local.gov.uk/sites/default/files/documents/25.130%20Making%20Decisions%20on%20the%20duty_06%20WEB.pdf

3. [R] Smith, L. (2024). 'LASPO 2012: ten years and beyond — a socio-legal study of the impact of legal aid cuts on service providers in England and Wales.' Journal of Social Welfare and Family Law. DOI: 10.1080/1460728x.2024.2399921.

4. [S/J] Amnesty International UK (2016). Cuts that Hurt: The impact of legal aid cuts on access to justice. Report on the first-year impact of LASPO: https://www.amnesty.org.uk/files/aiuk_legal_aid_report.pdf

5. [S] National Audit Office (9 February 2024). Government's Management of Legal Aid. The headline real-terms reduction is £728 million (28%) between 2012–13 and 2022–23, from £2,584 million to £1,856 million in 2022–23 prices. The 40%-of-family-cases-unrepresented and 73%-to-64% legal-aid-housing-provider statistics are from the same report. NAO press release: https://www.nao.org.uk/press-releases/governments-management-of-legal-aid/ . Full report: https://www.nao.org.uk/reports/governments-management-of-legal-aid/ . The May 2024 Public Accounts Committee follow-up, Value for Money from Legal Aid, confirmed the figures: https://committees.parliament.uk/work/8153/value-for-money-from-legal-aid/

6. [S] Equality and Human Rights Commission (2018). The Impact of LASPO on Routes to Justice: https://www.equalityhumanrights.com/sites/default/files/the-impact-of-laspo-on-routes-to-justice-september-2018.pdf

7. [S] Re D (A Child) [2014] EWFC 39, judgment of Sir James Munby P. Discussed in Law Gazette, "Munby's fury as aid denied in adoption fight," 3 November 2014: https://www.lawgazette.co.uk/

8. [S/J] Sir James Munby, The crisis in private law — speech to Shared Parenting Scotland, Edinburgh, 10 February 2020. Published by The Transparency Project: https://transparencyproject.org.uk/

9. [S] Court of Protection Rules 2017 (as amended): https://www.legislation.gov.uk/uksi/2017/1035/contents . The Open Justice Court of Protection Project (Celia Kitzinger, since 2020) maintains attendance records and analysis of welfare hearings: https://openjusticecourtofprotection.org/

10. [S] Senior Courts (Master of the Rolls and President of the Family Division), Practice Guidance: McKenzie Friends (Civil and Family Courts), July 2010, issued by Sir Nicholas Wall P: https://www.judiciary.uk/wp-content/uploads/JCO/Documents/Guidance/mckenzie-friends-practice-guidance-july-2010.pdf . The Practice Guidance sets out the position that McKenzie Friends have no automatic right of audience and cannot ordinarily conduct litigation, but that courts may grant rights of audience in limited circumstances. Society of Professional McKenzie Friends: https://www.mckenziefriends.directory/ . On the Legal Services Board's ongoing work on lay representation, see [16].

11. [R] Herd, P. and Moynihan, D. (2018). Administrative Burden: Policymaking by Other Means. Russell Sage Foundation. Winner of the 2019 Louis Brownlow Book Award. See also Herd, P. and Moynihan, D. (2025), 'Administrative Burdens in the Social Safety Net,' Journal of Economic Perspectives 39(1): 129–50: https://www.aeaweb.org/articles?id=10.1257/jep.39.1.129

12. [S] Court of Protection annual reports and case-length data: HMCTS Family Court Statistics Quarterly and Court of Protection Statistics tables, published by the Ministry of Justice: https://www.gov.uk/government/collections/family-court-statistics-quarterly . Average disposal times for contested welfare cases vary; the eighteen-months-to-three-years range describes a common pattern in contested welfare proceedings rather than a statistical mean.

13. [S/R] Carers UK, State of Caring 2023 (published October 2023), survey of over 12,000 unpaid carers across the UK. 49% reported a mental or physical health condition that had worsened over the previous year; over 25% reported bad or very bad mental health. Report: https://www.carersuk.org/reports/state-of-caring-2023-report/ . Successive State of Caring surveys (2022, 2024, 2025) show consistent patterns: https://www.carersuk.org/policy-and-research/state-of-caring-survey/

14. [R] Clinical literature on caregiver burnout: South London and Maudsley NHS Foundation Trust (2025), Caregiver Burnout: Causes, Symptoms, and Coping Strategies. See also Alzheimer's Association (2024), Caregiver Health materials, and National Alliance for Caregiving & AARP (2020), Caregiving in the U.S. report.

15. [S] Local Government and Social Care Ombudsman, Annual Review of Adult Social Care Complaints 2023–24, including the self-funder share of complaint volume versus care-user population: https://www.lgo.org.uk/information-centre/reports/annual-review-reports/adult-social-care-reviews

16. [S] Legal Services Board, Reshaping Legal Services programme of work on lay representation and consumer access, including the LSB's policy statement on McKenzie Friends and rights of audience: https://legalservicesboard.org.uk/ . See also Bach Commission on Access to Justice (2017), The Right to Justice, Fabian Society.

Compiled · May 2026 · For research and journalism use