The previous article in this series closed by noting that Circular 2026/09 had made the right of withdrawal workable for the first time in forty years, but that the harder question — whether the curriculum from which parents may now more easily withdraw will be reformed to the point where withdrawal is no longer necessary — remained open. That question is now being answered from two different directions. The Minister of Education, addressing the Assembly on 3 February, set out the Department’s procedural and curricular response. David Armstrong, writing in the News Letter on 11 May, has now proposed an alternative architecture for the controlled sector. Both address the wrong question.
The right question, established by Lord Stephens at paragraphs 116 to 120 of Re JR87 [2025] UKSC 40, is whether the architecture through which religious provision reaches children produces stigmatisation, compelled disclosure, and deterrent effects sufficient to render illusory the rights of children and parents under Article 2 of Protocol No. 1 (A2P1) of the European Convention on Human Rights. The evidence strongly suggests that the controlled sector now presents precisely such a systemic risk, and that the risk persists regardless of whether the mechanism families navigate is called withdrawal or opt-in, applied to a syllabus that is denominational or pluralistic, layered on top of a Christian default or alongside one. The problem lies less in the form of the choice than in the structure that makes such a choice necessary. JR87 does not require a secular curriculum, and the Court was explicit that “it is within the Department’s margin of appreciation in planning and setting the curriculum for the greater part of religious education to focus on knowledge of Christianity”. What it does require is that the state should not rely on withdrawal, non-objection, or parental signalling to cure a religious arrangement that otherwise exposes children and parents to stigma, disclosure, and deterrence. On that reading, Armstrong’s proposal is less secure than it first appears, and the Department’s response still leaves the structural question unresolved.
What the Purdy report shows
The most striking thing about the available evidence base is that the central finding does not come from critics. It comes from a study commissioned and funded by Scripture Union Northern Ireland (SUNI), conducted by the Centre for Research in Educational Underachievement at Stranmillis, and led by Professor Noel Purdy. Purdy is also chair of the Department’s Drafting Group for the revised Core Syllabus, appointed by the Minister on 3 February. Faith in our Schools, dated February 2026, surveyed pupils, parents, principals and church leaders before the Supreme Court judgment was handed down. The authors are explicit that their samples are small and illustrative rather than representative, and that engagement from the Catholic maintained sector was limited. With those caveats, the report’s data closely resemble the conditions that JR87 treated as legally significant.
Three findings stand out. On disclosure: 97% of principals reported informing parents of the right to withdraw, but only 79% of parents reported being aware of it, and only 38% felt sufficiently informed about the content of religious activities. One parent told the focus group she had been reassured by a principal that religious involvement was minimal, “only to later discover that the information she gave me wasn’t accurate, and there were many religious visitors.” On stigma: a pupil told the researchers it would be “awkward going to a teacher and saying I want to sit out, because you’d be the only one doing it. And this is a Christian school, so I guess everyone would be wondering, ‘Why did you come to this school if you don’t want to hear anything Christian?’” Another said she would feel “left out if I wasn’t allowed to go to assembly.” On deterrence: the report cites the Parents for Inclusive Education finding that just 1.2% of pupils are withdrawn from RE or collective worship — a figure the authors treat approvingly, as evidence that few parents wish to withdraw, but which Lord Stephens’s reasoning permits us to read in precisely the opposite way. A 1.2% withdrawal rate against a backdrop of 21% of parents in the report’s own sample describing their child’s religion as ‘None’, and 47.4% of controlled primary enrolment not recorded as Protestant, is consistent with the suppressed exercise of a right rendered illusory by stigma and disclosure.
The Scandinavian experience is worth reflecting on. Norway, Sweden and Denmark have all wrestled, in different ways, with exemption from religious education. Lomsdalen and von der Lippe’s January 2025 comparison in the Journal of Beliefs and Values suggests that where Christianity remains embedded as cultural inheritance, exemption systems repeatedly struggle with disclosure and stigma. That matters because Northern Ireland is now drifting towards something with a family resemblance: a Christianity-centred default, moderated by opt-out or opt-in mechanisms. None of the three Scandinavian jurisdictions has adopted the layered arrangement Armstrong proposes.
Closer to home, the NICIE/IEF Faith and Belief in School guidance, published in February 2025, documents what opt-in denominational provision actually looks like in the sector best equipped to manage it. Integrated schools already accommodate Catholic sacramental preparation alongside their inclusive ethos; the guidance devotes fifteen pages to the operational complexity, records parents reporting that opting out meant “separating their child from whatever the rest of the class or school was doing — and not an option they chose lightly”, and rests on an infrastructure of steering groups, commissioned research, and 124 pages of guidance. That is the realistic floor for what an opt-in scheme in controlled schools would actually require. The Northern Ireland data, the comparative analysis, and the most institutionally committed inclusive sector all point in the same direction.
The asymmetry Armstrong does not acknowledge
Armstrong’s proposal turns on a parity argument: Protestant parents in controlled schools should get what Catholic parents in Catholic schools already get. The premise is that the controlled sector is the Protestant equivalent of the Catholic maintained sector, and that what is permitted in one should be permitted in the other.
That parity does not exist in law. Catholic maintained schools have a recognised denominational character. Their religious character is reflected in governance, sacramental preparation, and the Grow in Love programme that sits alongside the Core Syllabus. According to Borooah and Knox’s 2026 figures, their pupil population is 94% Catholic. Institutional character and pupil body match.
Controlled schools do not have the same denominational character. Article 21(2) of the 1986 Order requires their RE to be undenominational, and Lord Stephens explicitly noted this distinction at paragraph 54 of JR87, contrasting “undenominational” controlled and integrated schools with Catholic maintained schools, where denominational RE is permitted because the school itself is denominational. The historic role of the transferor churches in founding controlled schools is real, but it was transformed by the transfer settlement of the 1930s and 1940s. The transferor churches transferred schools into state ownership and retained statutory governance representation — including, in a standard nine-member controlled primary board, four transferor governor places — and accepted that the schools would teach undenominational RE. That settlement made the schools publicly owned controlled schools with church governance representation.
This matters for Armstrong’s argument in two ways. His parity claim assumes a symmetry that no longer exists, since the transfers have been completed. And his proposed remedy would, in effect, cut across the settlement that the transferor churches struck, reintroducing denominational instruction into schools whose statutory character is undenominational, on the strength of a governance arrangement that was the consideration for accepting that character in the first place.
Why Armstrong’s remedy is legally fragile
Armstrong proposes “a voluntary, opt-in programme outside the statutory curriculum” of denominational Christian RE for controlled schools whose parents request it. The proposal is presented as restoring parity with the Catholic sector, where denominational instruction continues through the ‘Grow in Love’ programme alongside the statutory syllabus. The argument is internally coherent on its narrow legal premise: JR87 targets the Core Syllabus, not denominational instruction more broadly, and Article 21(7) of the 1986 Order already permits ministers and “other suitable persons” to deliver denominational instruction in controlled schools subject to parental non-objection.
The difficulty is that Armstrong’s model may recreate the same burdens from the other side. In a controlled primary where 47.4% of pupils are not recorded as Protestant — and in 39% of controlled primaries where pupils not recorded as Protestant are now the majority — an opt-in denominational programme would mark out the children whose families had positively declared Christian adherence as a separately identified cohort within the pupil body, and in many schools, potentially a visible minority. Lord Stephens’s finding at paragraph 118, that being the only child differently provided for in a small school environment was inherently stigmatising, applies whichever side of the line the child sits on. Opting in to denominational instruction discloses parental conviction more directly than opting out of a multi-faith curriculum, because the act of consenting is the declaration. The school maintains a register of children participating because it has to organise around them — a register from which parental religious preference could readily be inferred. And the deterrent effect operates on opt-in mechanisms as it does on opt-out mechanisms: families weigh the social cost of either acting on or refraining from their preferences, and the Purdy data suggest that, in current Northern Ireland conditions, many families are likely to avoid choices that carry an obvious social cost.
Armstrong appears to anticipate this objection by locating the denominational programme “outside the statutory curriculum”. This does not work. Lord Stephens recorded at paragraphs 26 and 64 to 68 that the access of ministers and other suitable persons under Article 21(7) at JR87’s own school had been exclusively Christian, was Bible-based, and was directed at proselytising. Far from falling outside the A2P1 analysis, it formed part of the factual matrix on which the breach was found. At paragraph 130, the court flagged Article 21(7) as one of three issues on which it expressed concern but declined to rule definitively, “absent full argument on these points” — including specifically whether the parental-non-objection safeguard was sufficient to protect Convention rights. Building further denominational provision on a foundation the Supreme Court has just identified as potentially defective is not a reliable route to compliance.
Why the Department’s response may not be enough
The Department’s response takes a different route. The Core Syllabus will be revised under Professor Purdy’s chairmanship to deliver content that, in the Department’s terms, is “objective, critical and pluralistic,” while keeping Christianity central on cultural-heritage grounds. Withdrawal has been overhauled through Circular 2026/09 to be frictionless. Inspection arrangements are to be brought forward through legislation. A final assessment is not possible until the revised syllabus exists. But there are reasons to think it may not, on its present trajectory, be sufficient.
Where Christianity is retained as cultural inheritance, it can become the route by which Christian centrality re-enters a supposedly pluralistic system. The Terms of Reference for Northern Ireland’s review couple “Retain Christianity as the central focus” with the pluralism requirement, justified by “the historical, cultural and legal context of Northern Ireland and the enduring role of Christianity”. That places the review closer to the Norwegian trajectory after 2015 than to the Welsh one.
The Purdy report’s own data suggest that even with revised content, the structural conditions will not disappear. The 1.2% withdrawal figure is unlikely to be moved very far by a syllabus rewrite alone, because the constraint is not the syllabus but the social cost of being identified, in a small classroom, as the family that withdraws. The Circular requires that withdrawn pupils receive “meaningful, age-appropriate and supervised” alternative activities; the NICIE/IEF research records that in practice, even in the Integrated sector, this typically means reading, educational games, or helping a teacher. Nelson and Yang’s 2022 finding that the 2007 world-religions reform at Key Stage 3 was implemented with such variability across controlled schools that it would be impossible to say what an education in world religions consisted of in any given school applies with even greater force to non-curricular alternative provision. The Department’s response improves the floor but leaves the ceiling untouched.
Wales as a comparator
The Welsh Curriculum’s Religion, Values and Ethics subject was designed so that no family would need to withdraw from it. RVE must be taught objectively, critically and pluralistically, and explicitly not as a means of making learners religious or non-religious; the syllabus must reflect that religious traditions in Wales are mainly Christian while also taking account of other religions and non-religious convictions. The parental right of withdrawal has been removed in Welsh schools without a religious character because the curriculum is designed to meet the human rights standard without requiring any child to opt out. Wales is not a secular model — Christianity remains historically and culturally central. What Wales demonstrates is that Christian content can occupy a central place in RE without withdrawal serving as the mechanism that protects rights.
JR87 does not require Northern Ireland to copy Wales. But Wales better answers the structural question JR87 poses. Architectures that depend on opt-in or opt-out mechanisms to manage the rights of children whose families do not share the institutional default carry recurring risks. Architectures that design the default to be defensible for all pupils do not. Armstrong’s hypothesis fails because it asks how to preserve confessional content for those who want it, rather than how to design a curriculum that no family needs to withdraw from. The Department’s response is more sophisticated but pulls in the same direction: it asks how to make withdrawal frictionless, whereas JR87 asks how to design a curriculum in which withdrawal is unnecessary.
The Minister has been clear that “Christianity will remain central” and that he has “concluded that it is neither desirable nor easily achievable to bring about fundamental changes to collective worship”. These commitments are politically intelligible. They are also the same commitments that produced the settlement the Supreme Court has just declared incompatible with the Convention, as it was implemented in the school before it. A review whose terms of reference instruct it to retain Christianity centrally on cultural-heritage grounds does not appear, at this stage, to be on a trajectory to deliver something Welsh-comparable. At present, it looks closer to the Norwegian post-2015 settlement: a revised arrangement that may hold until someone tests it again.
That is not the worst outcome. But it would leave Northern Ireland managing withdrawal when it could be designing a curriculum that no longer requires it.
This is the twelfth article in a series examining educational governance in Northern Ireland. Previous articles: ‘The Transformation Majority That Doesn’t Count’ (I); ‘It’s Not Just Protestant Schools’ (II); ‘Take Down the Hurdles’ (III); ‘The Irony of Integration’ (IV); ‘Time to Flip the Switch’ (V); ‘Beyond Indoctrination’ (VI); ‘Eight Per Cent After Forty Years’ (VII); ‘Good in Parts’ (VIII); ‘Gone Girls’ (IX); ‘New Wine, Old Wineskins’ (X); ‘Epistle to the Parents: RE Withdrawal’ (XI).
Sources: Re JR87 [2025] UKSC 40; DE Circular 2026/09; DE (2026) Oral Statement: Religious Education and Collective Worship, 3 February; DE (2026) Terms of Reference: Review of the RE Core Syllabus; Armstrong, D. (2026) ‘RE ruling will hit Protestant teaching more than Catholic’, News Letter, 11 May; Borooah, V.K. and Knox, C. (2026) ‘Education, inequality and integration in Northern Ireland’, International Journal of Inclusive Education; Purdy, N., Symington, E., Orr, K. and Ballentine, M. (2026) A Critical Evaluation of how Christian Churches and Organisations engage with Schools in Northern Ireland, Centre for Research in Educational Underachievement, Stranmillis University College; Lomsdalen, C. and von der Lippe, M. (2025) ‘The right to opt out: a comparative analysis of exemption schemes in Scandinavian public schools and the impact of cultural bias’, Journal of Beliefs and Values, 46:3, 539–553; NICIE/IEF/APTIS (2025) Faith and Belief in School; Parents for Inclusive Education (2024) On a Mission: Religious Practices in NI Primary Schools; Nelson, J. and Yang, Y. (2022) ‘World Religions in Religious Education in Northern Ireland’, Religion & Education, 49:1.
El Cavador is a Slugger reader from Belfast with a particular interest in education
Discover more from Slugger O'Toole
Subscribe to get the latest posts sent to your email.
