Will courts uphold legislative attempts to “channel” challenges to executive or administrative decisions through statutory avenues of redress before allowing access to judicial review? If so, what conditions must be met – when and to what extent will this type of provision be given effect? In New Zealand, the answer has been that channelling is permissible provided it does not deprive those affected by the relevant decisions of an effective avenue of recourse. The 2019 Supreme Court decision in H (SC 52/2018) v Refugee and Protection Officer provides an example of a situation where that condition was not satisfied.
Orthodoxy on ouster or privative clauses
Every student of Administrative Law knows the presumption against ouster of judicial review established in Anisminic: courts give short shrift to ouster or privative clauses in statutes, invariably insisting that clearer words are needed to exclude review. The essential reason is that ouster of judicial review offends the rule of law. A plurality of the United Kingdom Supreme Court in an obiter dictum in Privacy International recently went further, finding a strong case for holding this is more than a presumption: any ouster that offends the rule of law simply cannot be given effect at all, regardless how clear the words (at para 144).
Modern qualifications to this orthodoxy include allowing “channelling”
Yet in modern case law, there are qualifications to this orthodoxy. In Privacy International itself, this strong reassertion of orthodoxy was qualified by recognition that there are limits to when and how much judicial review is required to satisfy the rule of law – as I noted here. At least some forms of ouster clause were thought likely to be acceptable on the basis that they do not offend the rule of law (at paras 131-134). The main example given were time limit clauses, which indeed have long been accepted by the courts (see Ostler). “Channelling” of challenges through statutory procedures was not among the types of ouster mentioned as likely to be upheld – but perhaps there is little need for that in the UK, given that permission for judicial review will usually be denied where other remedies have not yet been exhausted.
In some other jurisdictions, however, there is case law in favour of allowing such “channelling” – as has been pointed out by Paul Daly. New Zealand is one of those jurisdictions. Indeed, in the leading Tannadyce decision, the Supreme Court gave effect to a clause that went beyond mere channelling and instead substituted statutory procedures as the exclusive remedy. This was considered acceptable (subject to very narrow exceptions) on the basis that the statutory avenues were equally effective as judicial review, in that they could accommodate all the same grounds of challenge and (importantly) included access to the High Court. (For more detail on this, and discussion of other recent New Zealand decisions, see my 2013 and 2019 reviews of developments in Administrative Law in the New Zealand Law Review, available here and here.)
Doubt cast on the effectiveness of channelling provisions in NZ?
However, in H (SC 52/2018) v Refugee and Protection Officer, the New Zealand Supreme Court’s most recent decision on this question, the Court overturned one of the decisions allowing channelling that had been reported by Paul Daly. That might be considered to cast doubt on the New Zealand courts’ tolerance for channelling or exclusive remedy provisions.
In this post, I will argue to the contrary. Based on my analysis of the reasons in H, my conclusion will be that the decision is consistent with the approach established by cases like Tannadyce. The acid test for giving effect to channelling or exclusive remedy provisions has always been whether this would deprive those affected by the relevant administrative decisions of an effective avenue of recourse. H is best understood as an application of that test.
Analysis of H: an application of the established test for allowing channelling
H concerned the channelling clause in s 249 of the Immigration Act 2009, which requires unsuccessful refugee status claimants to exercise a statutory appeal to the Immigration and Protection Tribunal before seeking leave to apply for judicial review. The Court acknowledged the intention to “prevent duplicative proceedings”, but insisted that it must “prevent injustice occurring when a statutory process fails” (para 63). In the particular situation in this case, the channelling clause could not be given effect, for two connected reasons.
First, the Refugee and Protection Officer’s first instance process was fundamentally flawed: “there was no consideration of the merits of the claim at all” (paras 64, 77). The Officer had issued a decision declining a refugee status claim without interviewing the claimant, and as a result also without evaluating the material that had been submitted in support of the claim (see paras 9-15). While there was no clear statutory right to an interview, it was normal practice to conduct one, because an assessment of the claimant’s credibility is crucial. Indeed, the Officer concluded that without this opportunity to assess the claimant’s credibility, the claim “cannot be determined” (para 14). Yet he proceeded to decline the claim. The Officer’s reason for proceeding this way was that the claimant had “failed to attend” a scheduled interview. If correct, this would under s 149(9) of the Act have provided grounds for proceeding without that interview. However, in fact the claimant’s lawyer had sent a medical certificate when the claimant fell ill the night before the interview. While the certificate failed to comply with the requirement to state the illness, the lawyer also sent the medical notes which contained PDF format. No-one disputed that this process was seriously flawed.
Secondly, the Court did not accept that the statutory appeal to the Tribunal was an appropriate next step in such circumstances of a fundamentally flawed decision at first instance. This point was put in terms of a “curing” analysis (this term is used at paras 43, 53, 65 and 82). This usually applies where there has already been an appeal, and is used to determine whether remedies are still needed for the breach at first instance (see Calvin v Carr). But the question whether the appeal would cure the fundamental flaw at first instance was also relevant in the present context, and the Court held it would not do so.
On the Court’s reading, the statutory scheme was intended to afford refugee claimants two hearings on the merits (paras 43, 70-74). The process for the Officer’s first instance decision involves a set of significant safeguards that are not replicated in the appeal process (paras 41, 67-69). The first instance safeguards are requirements for the Officer to summarise the interview and to give the claimant three weeks to correct the record or deal with concerns raised, and for the Officer to then issue a reasoned decision. While the Tribunal at the second tier conducts a de novo appeal, its process does not replicate these features. A fair hearing at that level therefore was not considered sufficient by itself. Nor does the Tribunal have power to restart the whole process by remitting the matter for reconsideration by the Officer (paras 71-74). The Act also appears to preclude the Officer reconsidering his own decision (paras 17, 29-30). Hence the statutory process, once it miscarries at the first level (as it did in this case), does not include the means for correcting itself.
The Court did not quite spell out how this satisfied the usual test for refusing to enforce channelling or exclusive remedy clauses: it did not quite say why enforcing the clause would deprive the applicant of effective avenues of recourse. One might perhaps have thought that the entitlement to two fair hearings could have waited until after an appeal. However, arguably such a process would indeed not have amounted to effective avenues of recourse in these circumstances: it would have imposed an unnecessarily onerous burden on the applicant. Also relevant is that it might not have served the statutory purpose of avoiding duplication of proceedings.
One puzzling aspect of the decision may still cast some doubt on the Tannadyce approach. The Court commented that s 249 was in effect an absolute ouster clause, not a mere channelling (or “deferral”) provision. Its reason was that judicial review following an appeal would very likely scrutinize the decision on appeal rather than the first instance decision: review of the first instance decision was therefore in practice entirely excluded (paras 59-62). That reasoning effectively denies any significant distinction between channelling clauses and absolute ouster or privative clauses: it would apply to many or most clauses requiring appeal before judicial review, and entail treating these as absolute ouster clauses. But it involves an unduly narrow definition of channelling: on the better view, a clause involves channelling rather than ouster of judicial review so long as it preserves access to judicial review at some stage, even if only in relation to an appellate decision in the matter.
The main message from H, however, is quite consistent with Tannadyce. Tannadyce was distinguished on the basis that the appeal process there included access to the High Court and was sufficiently comprehensive to render judicial review unnecessary (para 87). What H rejects is any notion that the reluctance to allow ouster of judicial review is automatically inapplicable when dealing with a channelling clause. The constitutional concerns are still in play, and hence “the courts approach privative clauses cautiously and in particular will give anxious consideration to their interpretation and application” (para 63, see also 78). That is why the effect of channelling or exclusive remedy clauses depends on an assessment of whether the particular statutory remedial scheme represents effective avenues of recourse in the particular circumstances.
Hanna Wilberg (Associate Professor, University of Auckland Faculty of Law)