State Protection Requires Effectiveness, Not Mere “Improvement”: COI Must Rationally Support a Section 31 Finding

1. Introduction

The applicant, a Georgian national anonymised under the International Protection Act 2015, sought international protection in Ireland on the basis of sustained stalking and threats (including threats to kill) by a man (“GA”) between 2018 and 2021. The International Protection Office (IPO) accepted her account, accepted a well-founded fear of persecution, and accepted a Convention nexus (membership of a particular social group: women victims of stalking/gender-based violence), but recommended refusal because it concluded that effective state protection was available in Georgia.

On appeal, the International Protection Appeals Tribunal (IPAT) again accepted credibility, accepted the persecution risk and the particular social group, but refused refugee status and subsidiary protection on the basis that Georgia had an “effective legal system” for gender-based violence and that protection was “effective and of a non-temporary nature”.

The judicial review concentrated on the legality of the Tribunal’s state protection assessment under section 31 of the 2015 Act, and in particular whether IPAT’s handling of country-of-origin information (COI)—which it described as presenting a “mixed picture”—could rationally sustain a finding of effective protection for this applicant.

2. Summary of the Judgment

The High Court quashed the Tribunal’s decision and remitted the appeal to a differently constituted Tribunal, granting relief on the applicant’s principal COI/state protection ground.

• The Court rejected the complaint that IPAT failed to consider the 2023 USDOS report, applying the presumption from GK v Minister for Justice [2002] 2 IR 418 (as developed to Rana v Minister for Justice [2024] IESC 46) that a decision-maker’s statement that it considered all materials is accepted absent evidence to the contrary.
• The Court also held that, even if a newer USDOS report existed, the applicant had not shown a material difference sufficient to invalidate reliance on the 2022 report, endorsing Phelan J.’s approach in A.S. v. IPAT [2025] IEHC 184.
• However, on the central issue, the Court held that IPAT’s finding of “effective state protection” did not flow rationally from the COI and that the Tribunal unlawfully conflated “improvements” with “effectiveness”, thereby misapplying section 31 of the 2015 Act.
• The Court rejected a standalone challenge to IPAT’s “police oversight” reasoning (the Special Investigation Service) and found the decision adequately reasoned as a matter of explanation, even though the state protection conclusion was legally flawed.

3. Analysis

3.1 Precedents Cited (and how they shaped the result)

A. The section 31 framework and “particularised” analysis

• B.C. v IPAT [2019] IEHC 763: IPAT cited Barrett J.’s “four-part” breakdown of section 31 and the need to focus on the applicant’s particular circumstances. The High Court accepted that IPAT identified the correct test, but held that it failed in its application—illustrating the familiar judicial review distinction between “correct legal test stated” and “test correctly applied”.
• Idiakheua v Minister for Justice [2005] IEHC 150: Cited for the proposition that the question is whether the country provides “reasonable protection in practical terms”. The Court treated “reasonableness” as central but emphasised that section 31 requires effectiveness, not merely effort or trend improvement.

B. COI conflict, selectivity, and rational explanation

• D.V.T.S. v Minister for Justice [2008] 3 IR 476: Used as the paradigm where conflicting COI must be acknowledged and the resolution explained. O’Donnell J. held the present case did not involve the same kind of impermissible “selectivity” found in D.V.T.S., but nonetheless fell foul of a related illegality: the ultimate conclusion was not rationally supported by the COI.
• B.A. v. IPAT [2020] IEHC 589: Central to the Court’s approach. In B.A., the Tribunal’s own summary of COI was predominantly negative, yet it reached a positive state protection finding. O’Donnell J. applied the same logic: where the COI “general tenor” indicates inadequacy, a conclusion of “effective” protection requires more than citing improvements—it must be demonstrably supported by the evidence as assessed.
• FM. & R.M. v. IPAT [2021] IEHC 817: Deployed to distinguish between state “endeavours” and an “effective response”. The High Court drew a direct analogy: section 31 is not satisfied by “doing its incompetent best”; effectiveness is the statutory threshold. This case strongly anchored the Court’s conclusion that IPAT impermissibly equated improvement measures with effective protection.
• BD v IPAT [2024] IECA 186: Cited (Hogan J.) for the proposition that “it is not enough” that a country is making endeavours if those efforts are ineffective. This supported the High Court’s core holding: effectiveness is the key question under section 31.

C. Decisions upholding Tribunal evaluations of mixed COI (and why this case differed)

• X.S. v. IPAT [2022] IEHC 100: Cited by respondents for burden and the permissibility of concluding state protection exists despite crime/corruption. The High Court accepted the general restraint principle but held it did not rescue a conclusion not rationally supported by the COI.
• X.S. and J.T. v. IPAT and Minister for Justice and Equality [2022] IEHC 100: Discussed as an example where the Tribunal lawfully preferred a specific, methodologically robust COI source (the “CEDOCA report”) over more general material and provided cogent reasons. This provided the contrast: IPAT may weigh mixed COI, but it must show a rational evidential path to “effective protection”.
• W.P.L. and BP v. The Minister for Justice and Equality [2024] IEHC 184: Used to reaffirm “effective not perfect protection” and the limited judicial role where conclusions are within the range open on the materials. O’Donnell J. adopted the principle, but found the present case crossed the line because the COI did not, on fair reading, substantiate the “effective” finding.
• ES v IPAT [2022] IEHC 613: Cited for the proposition that evidence need not show crime/corruption is “completely” or “substantially” controlled before a tribunal may find state protection. The High Court again accepted this in principle, but treated it as distinct from the requirement that the COI must still rationally support effectiveness under section 31.

D. Presumption that materials were considered; “up-to-date COI” complaints

• GK v Minister for Justice [2002] 2 IR 418 and Rana v Minister for Justice [2024] IESC 46: The Court applied the presumption that a tribunal’s statement that all materials were considered is accepted absent evidence to the contrary. Because IPAT expressly referenced the 2023 USDOS report elsewhere (medical issues), the Court rejected an inference that it was excluded from the state protection assessment.
• A.S. v. IPAT [2025] IEHC 184: Adopted for requiring a “qualitative, comparative exercise” to show the newer COI is materially different such that non-reference (or reliance on slightly older material) would matter. The applicant failed to establish a material difference affecting the section 31 outcome.

E. EU and contextual authorities on gender-based violence claims

• MG (Georgia) v. IPAT [2026] IEHC 151: The Court noted overlap with a near-contemporaneous Georgian gender-based violence protection case and addressed some shared issues more briefly here.
• WS v. Intervyuirasht organ na Darzhavna agentsia za bezhantsite pri Ministerskia savet C-621/ 21, ECLI:EU:C:2024:47, 16 January 2024 (W.S.): O’Donnell J. treated the importance of W.S. as acknowledged and incorporated by reference (via MG (Georgia) v. IPAT [2026] IEHC 151), but held that it did not decide this case because IPAT accepted persecution risk and nexus; the defect was in the state protection/COI analysis.

F. Reasons / adequacy of explanation

• G.K. v IPAT [2022] IEHC 204: Cited for the proposition that reasons need not be discursive or address every point, but must engage core claims.
• YY v Minister for Justice and Equality [2017] IESC 61 and Middelkamp v Minister for Justice and Equality [2023] IESC 2: Cited for the requirement that reasons must allow the person affected and a reviewing court to understand the decision-maker’s reasoning process. The Court held the Tribunal’s reasoning was understandable even though legally flawed on the merits of section 31 application.

G. Comparative “effectiveness not effort” articulation

• R (Atkinson) v. Secretary of State for the Home Department [2004] all England Reports (D) 14: Quoted (via FM. & R.M. v. IPAT [2021] IEHC 817) for the principle that it is no answer that a state is “doing its incompetent best” if it falls below the required standard of protection—supporting the High Court’s insistence on effectiveness.

3.2 Legal Reasoning

(i) The Court’s role on judicial review

The judgment tightly reasserts the judicial review boundary: the High Court does not re-decide the protection claim, re-weigh COI, or substitute its view for IPAT’s. The legality inquiry is whether IPAT’s conclusion was:

• reached using the correct statutory test (section 31 of the 2015 Act),
• based on a fair reading of relevant materials (COI), and
• rationally supported by those materials (i.e., a conclusion “open to” the Tribunal).

(ii) “Mixed picture” COI must still support an “effective protection” conclusion

IPAT characterised the COI as mixed: improvements existed, but shortcomings remained. The High Court accepted that “mixed” evidence can lawfully be weighed to reach either conclusion, provided the tribunal explains how it resolves tensions and provided the outcome is rationally supported.

(iii) The decisive error: conflating improvement with statutory effectiveness

The Court’s central holding is that IPAT’s state protection conclusion was unlawful because:

• The finding of “effective state protection” did not flow rationally from the COI: the Court observed “very little” in the COI plainly supported a conclusion of effectiveness; the “general tenor” was that the response to gender-based violence was inadequate.
• IPAT treated “improvements” as if they equated to “effectiveness”: section 31 requires more than evidence of reforms, legislation, units, training, or initiatives. Those may demonstrate effort and trajectory, but the question is whether, in practice, the system reaches an effective and non-temporary threshold for the applicant’s risk.

(iv) Non-selectivity finding, but still an irrational/misdirected outcome

Notably, the Court did not rest its decision on a classic “selective quotation” problem (as in D.V.T.S. v Minister for Justice [2008] 3 IR 476). It found the error to be more structural: despite acknowledging shortcomings, IPAT moved from “mixed picture/improvement” to “effective protection” without the COI providing a stable evidential bridge.

(v) Stalking as a sub-type issue within broader gender-based violence COI

The Court recorded that the COI did not expressly address stalking and that it was unclear how Georgian law addressed it. IPAT treated the case as repeated psychological harassment within the broader domestic/gender-based violence context. The judgment signals that where the feared harm is not directly covered by available COI, decision-makers must exercise particular care before concluding “effectiveness” from generalised materials.

3.3 Impact

• Sharper discipline in section 31 decisions: Tribunals must demonstrate, not merely assert, that reforms and “steps” translate into an effective legal system and effective non-temporary protection for the particular feared harm.
• COI reasoning must show an evidential pathway: Where COI is mixed (improvements plus persistent shortcomings), a conclusion of effectiveness must be anchored in concrete COI support; otherwise it is vulnerable as irrational or as a misapplication of section 31.
• “Safe country” designation does not settle state protection: Although Georgia was a designated safe country and the appeal was papers-only accelerated, IPAT itself treated state protection without that overlay. The judgment reinforces that “safe country” procedure cannot dilute the section 31 analysis.
• Limits on “latest report” challenges: The decision strengthens the line that applicants must show materiality when relying on newer COI—courts will not quash solely because the tribunal relied on a report one year older if the difference is not demonstrated to matter.
• Gender-based violence claims: In cases where persecution risk/nexus is accepted (as here), litigation focus may shift to the robustness of state protection reasoning. This judgment provides a template for challenging “reform narrative” decisions that do not meet the effectiveness threshold.

4. Complex Concepts Simplified

• Certiorari: a High Court order quashing an unlawful decision (here, IPAT’s refusal decision), usually followed by remittal for reconsideration.
• COI (Country of Origin Information): objective reports (e.g., USDOS, Freedom House) used to assess conditions in the applicant’s home country, including whether the state can protect them.
• “State protection” (section 31, International Protection Act 2015): the question is not whether the state is perfect, but whether it takes reasonable steps and operates an effective legal system such that protection is effective, accessible to the applicant, and non-temporary.
• “Mixed picture” COI: evidence showing both positive and negative features (e.g., reforms alongside persistent under-enforcement). A tribunal may decide either way, but must explain its resolution and stay rationally tethered to the evidence.
• “Reasonable steps” vs “success rate”: a state need not prevent all harm, but the statutory requirement still demands effectiveness. A state’s efforts cannot be treated as sufficient if the evidence shows the system remains ineffective in practice.
• “Safe country” accelerated procedure: a procedural track allowing papers-only appeals for certain designated countries; it does not remove the need for a lawful, individualised state protection analysis.

5. Conclusion

T.M. [Georgia] v The International Appeals Tribunal and Anor [2026] IEHC 150 reinforces a disciplined approach to section 31 of the International Protection Act 2015: decision-makers must not treat improvement in a country’s response to gender-based violence as synonymous with effective protection. Where COI presents a mixed or generally pessimistic account of enforcement and outcomes, a finding that protection is “effective and non-temporary” must be demonstrably supported by the COI, not reached by optimistic inference from reform measures.

The case also consolidates two boundary principles: (i) judicial review does not re-weigh COI but will intervene where conclusions are not rationally open on the materials; and (ii) “newer COI” arguments require a demonstrated material difference, not merely recency. The remittal to a differently constituted Tribunal underscores that the legal defect lay in the state protection evaluative method, not in credibility or persecution findings.