The support a disabled child needs to get an education is a legal right — not a favour to beg for. Getting it has always meant a fight; now councils increasingly reassess an existing plan and quietly strip the support out. What changed is national policy: since 2021, central government has run two programmes — Safety Valve and Delivering Better Value — that reward councils for cutting the number of Education, Health and Care plans and the support in them. The incentive is financial; what it strips away is disabled children's legal rights. It is harming disabled children across England, and here is what the public data shows, council by council.
"Rationing" can sound almost reasonable — budgets are finite, and someone has to prioritise. So look at what is being rationed. Every child is entitled to an education, and for most that means a place, a teacher, a classroom, books — provided as an ordinary part of going to school. Schools do have to manage tight budgets and account for how public money is spent; that is real, and it is not the objection. But the accounting is done for the school as a whole. No head singles out one pupil and makes that child's family prove, case by case, that the chair, the books or the heating are "value for money" before the child can have them, or strips them from that one child when funds are short. The individual pupil is never the line item made to justify itself. A disabled child needs some different things to get that same education — a teaching assistant, a therapy, a quiet space — and the law writes them into a plan, an EHCP. That is not a perk or a prize; it is simply their version of the chair, the books and the teacher — what they need in order to be taught at all. Rationing it does not trim a luxury: it hands a disabled child less of the right every other child gets as a matter of course — and makes that child, alone, prove they deserve it.
If you have a disabled child, you may have had to fight for their support — the assessment refused, the therapy "not available", or a plan you already had quietly cut back at a reassessment. You may have been told it is just pressure on the system, or one difficult officer. It is not. More and more councils reach for the same move — reopen an existing plan, reassess, and strip out support the law required, a move the tribunals overturn almost every time a family can challenge it. It is happening across the country, and it is driven from the centre.
The law is clear. The Children and Families Act 2014 says a child's support is decided by their needs, not by a council's budget — and that the child and family have a right to be involved in the decisions made about them. But central government's SEND finance programmes — Safety Valve and Delivering Better Value — pay down councils' deficits in return for plans to reduce the number of EHCPs and move children out of specialist provision. "Value for money" has become the reason to raise the bar to assessment, refuse plans, and thin the support in them — even though the law does not allow a child's rights to be rationed to save money.
And when families challenge these refusals, they almost always win. Around 99% of decided SEND tribunal appeals go the family's way, and the Local Government Ombudsman upholds roughly nine in ten of the SEND complaints it investigates. A refusal overturned that often was never really a judgement about the child — it was a judgement about the budget.
This page ranks all 149 English councils on how hard the published data suggests it is to get, and keep, a plan. It is built entirely from public government figures and councils' own documents. Search for your own council below — but read "how to read this" first: a high score is a reason to ask questions, not a verdict.
You might think your council is uniquely difficult, or that you were simply unlucky with one officer. The pressure comes from the centre — but, to be fair to government, its stated aim is a reasonable one.
Put that way, it is hard to disagree: help a child sooner and you meet their needs better and spend less. A budget saving that falls out of acting early is a good outcome, not a scandal. And the government makes exactly this case — the Secretary of State, launching the reforms, forecasts that the number of plans will eventually fall, and insists the fall will come from better support, not from cuts:
Take that at face value — and one test settles it. If plans were falling because children's needs were being met earlier, families would rarely have cause to appeal, and would rarely win when they did. Instead:
When nearly every contested refusal is overturned, the fall is not coming from needs being met sooner — it is coming from the door being shut. The finance deals reward exactly that: they are judged on refusals and plan numbers, not on how fast help arrives. The stated aim is early help; the incentive on the ground is refusal; and the gap between the two is the story — a national one, not your council's alone.
But national pressure is not the same as no local choice. Faced with the same incentive, some councils hold the line on children's rights and others ration hard — and the difference between them is real, and large. That is exactly what this page measures. The pressure comes from the centre; a council that chooses to break the law is still accountable for that choice, whatever the pressure. Look up your own council below to see where it sits.
The lever is the high-needs budget — the money that pays for SEND. Many councils run large deficits on it. To pay those deficits down, central government has run a series of programmes — Safety Valve (2021–2026), Delivering Better Value, and now the High Needs Stability Grant — that hand over money in return for plans to cut spending. In practice that means fewer plans, higher thresholds to assessment, and more children kept in mainstream without the extra support they need.
You can read the commitments in the councils' own documents. Thirty-eight councils signed Safety Valve agreements; several commit, in writing, to cut EHCP numbers — some to a benchmark ("towards national averages"), some to a set figure — targets to hit rather than needs to meet. The walk-through that follows, and the "their own words" column in the table below, show this verbatim, council by council.
One legitimate-sounding lever does much of the work: moving children out of independent and non-maintained special schools, which are costly. Cutting that reliance can be sound value for money — the objection is not to that. It becomes rationing when a council thins the specialist support a child is entitled to in order to force the move into mainstream before mainstream can meet the need. The families caught in the middle often don't want an independent placement; they need the specialist help, wherever their child is taught. The independent-placement column in the table below shows where that reliance is highest — Surrey (12.2%) and West Sussex (11.2%) lean on it most — and it is precisely this that the finance programmes press councils to reduce.
And the squeeze is tightening. An accounting rule that let councils keep these deficits off their books is being withdrawn (now to March 2028), and the newest programme ties the bailout to council "reform plans" whose criteria the government has not published — so the incentive continues, but with less of the paper trail.
And this is not one party's policy. The Conservative government created Safety Valve in 2021; the Labour government elected in 2024 has continued and extended it, and its 2026 White Paper would move many children off EHCPs altogether. Under both, central government has incentivised a rationing model in place of the disability-rights model the law is built on.
Read the documents in order and a process comes into focus — not a conspiracy, but a chain of small, deniable decisions that add up to a child losing support. It starts with a number to hit.
An average is only a target if you assume every area has average need. They don't — need tracks deprivation and how well it is picked up — so "towards the average" is a quota, not an assessment. And the quota reaches into the plan itself:
"Specificity" is the legal heart of a plan — the precise hours and provision a council is bound to secure. Thin it and you save money and dissolve the duty in the same stroke. Support already in place is re-costed and clawed back:
At the front door, cutting the number of applications becomes the aim — sometimes so plainly that a council's own scrutiny committee flinches:
Plymouth's scrutiny committee later replaced that phrase with the gentler "minimising the need for new EHCPs through enhanced early intervention" — explicitly, in the minutes, "to protect stakeholder relationships with SEND families." They understood exactly how "curbing applications" reads. (Note the discipline this needs: "expanding local provision" and "reducing reliance on independent placements", in the same Plymouth sentence, are legitimate aims — it is the word "curbing" that crosses the line, not the whole plan.)
Refusing is not treated as a failure. In the finance paperwork it is a target — one council's Safety Valve monitoring lists, as a deliverable:
Why do families fight so hard for the plan in the first place? Because, as the Secretary of State conceded, it is the only thing that actually compels a council to act:
Hollow out the ordinary support in a mainstream school and you manufacture the scramble for the plan — then blame the queue on the parents. And when a family finally forces the question in front of an independent judge, the refusals collapse:
That is the system as it stands. The reform now on the table would formalise the shift the paper trail already reveals — from a right you can enforce to a resource that is allocated:
Which brings it back to the child. None of this is really about a document. It is about whether a disabled child gets the support they need to learn — and whether that support is a right, or a favour they and their family have to beg for, one panel at a time.
The word the system uses is "provision" — but it just means the things a child needs in order to learn: a teaching assistant, a speech and language therapist, a quiet room to regulate in, a piece of equipment. For a disabled child these are not extras. They are the equivalent of the chair, the desk, the textbook and the whiteboard — what makes education possible at all.
So look at what happens to them. Even once a child has an Education, Health and Care plan — a document that legally obliges the council to secure that support — the support can be routed back through the council's own funding panel to be argued again and tested for "value for money". There is no such step in law: the plan is meant to be the decision, not the start of another one. And it is rarely put to families as an honest, formal "apply again" — because there is no lawful basis to make them. Instead a family is sent away to "make a case", while the real form is filled in and pleaded internally, by an officer, to a panel the family never sees. You are left having to persuade an official to beg, on your behalf, for something your child is already owed by law. Imagine that rule applied to every other child: a school made to justify, pupil by pupil, that the chair, the textbook, the heating and the water were affordable — and to withhold them if the budget was tight. It would be unthinkable. Yet for the very things a disabled child needs to sit in the same classroom, it is routine. The right becomes a bid.
Some of this is genuinely fair, and councils say so: a good deal of SEND money is discretionary help offered before any plan exists, and authorities are honest it is not an entitlement. The problem is when the same machinery — the application, the panel, the value-for-money test — is turned on support already written into a statutory plan, and so owed as of right.
Elsewhere it is an "Exceptional Circumstances Funding Application Form" put to a "weekly complex needs panel" (Barnet); a panel that meets "twice a term to consider applications for SEND funding" (Norfolk); or SEND co-ordinators trained "how to make good applications" (Leeds), so that a child's support turns on how well an adult can write a bid.
None of this should turn on making the case for a child's support a second time — through an application, a value-for-money test, or a "statement of impact". The plan already exists because the child's needs were assessed and met the legal test. Re-justifying the support in it, by impact or by cost, is a second gate the law does not sanction: needs were the question, and they were already answered.
And this removes the one person the decision is actually about. That a disabled child or young person takes part in decisions that affect them is not a nicety — it is a legal right, and one of the foundations of the whole system. The Children and Families Act 2014 requires councils to have regard to the child's and young person's views, wishes and feelings and to help them take part; the United Nations Convention on the Rights of Persons with Disabilities, which the United Kingdom has signed, requires disabled people to be involved in the decisions that affect their lives. A financial rationing decision, taken by an internal panel, does the reverse: the child is not there, the family is not there, the case is gathered by an officer and argued behind closed doors — and if it is refused, families are seldom given the reasons in any form they can answer, let alone a chance to put their side to the people who actually decided. The principle the law makes central — nothing about a disabled child settled without them — is exactly what the rationing process strips out. Support a child is owed by law should not have to be applied for, argued, or justified for value for money. It should simply be secured.
The "golden ticket". The idea that families game the system for a prize was put on the parliamentary record by Ofsted's then-Chief Inspector, who told MPs in 2019 that "an EHCP is the golden ticket, so there is a great deal of focus among parents to do anything to get a plan." But if plans were golden tickets wrongly claimed, the tribunals — independent judges — would refuse them. Instead councils lose around 99% of the appeals they fight. A plan is only "golden" because the ordinary support around it has been hollowed out.
The "national average". Some councils set out to cut EHCP numbers "towards national averages" or "in line with statistical neighbours". But there is no lawful basis to drive plan numbers to an average unless the needs of the area are average — and need is not uniform; it tracks deprivation, demographics and how well needs are picked up. An area with above-average need should have above-average plans. Targeting the average is a quota, not an assessment. And if every council chases the mean, the mean itself falls — a ratchet that lowers support everywhere.
Two independent bodies reach the same verdict. Families win around 99% of the SEND tribunal appeals that are decided. And the Local Government Ombudsman upholds roughly nine in ten (91–94%) of the SEND complaints it investigates — far above its average for everything else. When either watchdog actually looks, councils are overwhelmingly found at fault.
Yet the one body meant to inspect councils proactively — Ofsted — doesn't look at this at all. Its Area SEND inspections judge experience and process, not whether a council is refusing assessments or rationing plans to save money; the words "refusal", "gatekeeping" and "Safety Valve" don't even appear in its framework. A favourable Ofsted verdict is not evidence that a council isn't gatekeeping.
The finance side, meanwhile, treats the opposite as success: one council's Safety Valve monitoring records "increasing the number of refusal to assess and refusal to issue decisions" as a deliverable. One arm of government rewards the refusals; the inspector doesn't count them; and families are left to fight for the law, one tribunal at a time.
In February 2026 the government published its Schools White Paper, "Every child achieving and thriving." EHCPs are kept and reformed, not abolished, and not restricted to special schools. But a new plan — the Individual Support Plan — is proposed to take over for many children, and the DfE projects the share of children with an EHCP to fall over the next decade.
The government says the new plans will be legally binding. The question families and MPs are asking is whether they will carry the same enforceable rights an EHCP does: the right to appeal to a tribunal, and the duty on the council to actually secure the support. It is still at consultation stage, with no law yet — which is exactly why it is worth asking the question now.
This is not an argument against saving money, or against inclusion. Reducing reliance on expensive independent schools can be sensible value for money, and a good mainstream school is often the right place for a child. The line is crossed when "inclusion" becomes a label for removing the support a child needs, or when plans are rationed to hit a budget. The table keeps these apart: a council that is only reducing independent placements reads "placement-cost", not gatekeeping — the count against a council is for rationing the entitlement or the support.
And a score here is a starting point for questions, not a verdict on any child or council. It is built from public data with known limits — please read "how to read this" below before drawing conclusions.
A note on sourcing: the government quotes above are drawn from the parliamentary record and should be checked against the original transcripts before being quoted onward. Everything in the table comes from named public documents.
National figures across all councils — a full count of every request, assessment and plan issued, not a survey, so these are exact rates over tens of thousands of cases a year and the year-on-year moves are real, not sampling noise. Each panel is a separate gatekeeping signal on its own scale — the vertical axis is cropped to each metric's range and does not start at zero, to make the shape legible — so higher = harder to get support; the points-change printed beneath each panel states the true size of the move, so the zoom can't dress a small change up as a big one. "Plan issued late" is the inverse of the 20-week timeliness column. Refuse-to-assess begins in 2023 — the DfE did not collect the decision split before then.
The direction of travel is a gradual worsening: plans are issued late far more often than in 2019, and front-door refusals have risen since 2023. This is the national backdrop; the table below shows how individual councils sit within it, and each council's own two-year trend.
All 149 councils, worst-first. Type a name to find yours. Click any heading to sort. Hover "their own words" for the verbatim quotes; hover any figure for what it means.
| # | Local authority | Score | Trend | Their own words | Refuse to assess | Assess, no plan | Appeal rate | 20-wk timely | Indep. placements | Safety Valve |
|---|
Composite gatekeeping score. Bars right of centre (warm) = more gatekeeping than the national average; left (cool) = less.
Each dot is a local authority. Vertical axis = gatekeeping score (up = more gatekeeping). Horizontal axis = high-needs funding per head of the 2–18 population (right = better funded), context only, never in the score. The cross marks the national average gatekeeping (score 0) and the median funding per head. Read the quadrants:
Top-left = underfunded & gatekeeping (systemic pressure). Top-right = well-funded & gatekeeping (a choice, not a budget story). Bottom = below-average gatekeeping, whatever the funding. Funding never excuses gatekeeping — it only explains where the pressure sits.
Everything on this page is built from what people actually said or wrote. Search and filter the record below — government ministers and Parliament on one side, councils' own budget and strategy documents on the other. The full parliamentary record of 46 quotes, with citations, is in the accompanying dossier; council quotes link to the source document.
Everything here is built from public data. No figure is invented, estimated or placeholdered; every quote is verbatim with a live source; and the whole ranking regenerates from open code, so anyone can check it.
The gatekeeping score combines four official DfE measures, each oriented so higher = more gatekeeping, standardised across the 149 scored councils and weighted: forced-appeal rate 40%, refuse-to-assess 22.5%, assess-but-no-plan 22.5%, and 20-week timeliness 15%. Timeliness only earns a council credit if it is not already a heavy front-door refuser — so fast processing can't disguise gatekeeping. Councils are joined across datasets by name, not code (two DfE codes are swapped), and the build checks for zero mismatches.
Coverage. 149 councils are scored. A few are omitted where the DfE suppresses or omits a required 2025 figure for them (named at the foot of the page, not fabricated). Cumbria and Northamptonshire don't appear because they were reorganised; their successor councils do. High-needs funding, independent-placement share, Safety Valve status and the "their own words" column are shown as context and never form part of the score — underfunding may explain pressure, but it is not a lawful reason to gatekeep.
Direction of travel. Each council's trend compares its latest year (2025) with the year before (2024) on the three flow measures — refuse-to-assess, assess-but-no-plan and timeliness. Because the DfE figures are a full count, the honest question for a single council is whether a change is bigger than small numbers could produce by chance, so each measure's year-on-year change is tested with a two-proportion z-test and counts as a move only at 95% confidence (with the standard validity check, so a council with a handful of cases cannot register one). The significant moves are then combined: ↑ worsening (only worse), ↓ improving (only better), ↕ mixed (significantly worse on one measure and better on another — typically a backlog clear-out that speeds issuing while front-door refusals rise), → stable (nothing beyond noise). The appeal rate is pooled with no single-year value, so it is excluded. The hover shows each measure's change and the case numbers behind it; the trend sits next to the score but is not part of it.
The stance tags (+ / e / c / ~ / −) are a judgement about each quote's mechanism, made in context: does it ration a right, or genuinely meet needs earlier? They are context only, never scored — reducing reliance on expensive independent placements, or offering real early help, is legitimate; refusing, thinning and capping the support a child is owed is not.
Sources, all public:
A gatekeeping score is a signal, not a verdict: it flags where the published data suggests families face the hardest time — a starting point for questions, not a conclusion about any one council or child.