High Court of Ireland Decisions

BETWEEN

1. The present case illustrates some problems of statutory interpretation relating to repeals, commencements, savings and transitional provisions: a cocktail with the capacity to chill the heart of even the most seasoned drafter.

2. The problem at issue involves the interplay of four enactments in particular:

(i). the Refugee Act 1996, now repealed;

(ii). the Immigration Act 1999, which contains a reference to the repealed 1996 Act;

(iii). the Interpretation Act 2005, which specifies how references to repealed Acts are to be construed; and

(iv). the International Protection Act 2015, which repeals the 1996 Act and contains new provisions for processing protection claims.

4. On 19th July, 2013 she applied for subsidiary protection. That application was again refused by the Refugee Applications Commissioner on the basis that there were no substantial grounds to show a real risk of harm. On 28th July, 2016 an appeal to the tribunal against that decision was refused.

5. The Minister then made a proposal to deport the applicant on 14th October, 2016. The applicant submitted representations on 4th November, 2016 pursuant to s. 3 of the 1999 Act including further documentary evidence in support of her claim of risk of harm in the form of an “attestation”. The making of a deportation order was analysed in the Department of Justice and Equality on 16th December, 2016 but a deportation order was in fact not formally made until 13th January, 2017. This was not until after the law had changed by the commencement of the 2015 Act on 31st December, 2016.

6. After the making of the order there was then an unexplained delay in serving the order on the applicant, which ultimately occurred on 29th August, 2017.

7. The applicant now seeks certiorari of the deportation order, and I have heard submissions from Mr. David Leonard B.L. for the applicant and Ms. Siobhán Stack S.C. (with Ms. Grainne Mullan B.L.) for the respondent.

How the problem arose
8. The problem here is that the applicant’s refoulement situation was considered under s. 5 of the 1996 Act, as of 16th December, 2016, but that provision was repealed by the 2015 Act on 31st December, 2016, prior to the formal making of the order. The 2015 Act included a similar provision in s. 50, so the issue comes down to whether the applicant should have been considered under s. 5 of the 1996 Act, or s. 50 of the 2015 Act, or alternatively whether neither of these applied.

9. A point that will become important later in the discussion is that the protection now available under s. 50 of the 2015 Act is wider and more beneficial than that previously available under s. 5 of the 1996 Act. Section 5(1) of the latter Act provides that “A person shall not be expelled from the State or returned in any manner whatsoever to the frontiers of territories where, in the opinion of the Minister, the life or freedom of that person would be threatened on account of his or her race, religion, nationality, membership of a particular social group or political opinion.” However while s. 50(1)(a) of the 2015 Act is in similar terms, the section goes on to add “, or (b) there is a serious risk that the person would be subjected to the death penalty, torture or other inhuman or degrading treatment or punishment.” Thus there is an extension of the protections available to an applicant under the terms of the new provision. Now it may be that s. 50(1)(b) does little more than replicate protections under the ECHR in any event but that is not hugely material for present purposes. The main point to note for when it becomes relevant later is that s. 50 provides greater protection for applicants than s. 5.

The nature of statutory repeal
10. To put the matters at issue in the appropriate setting it seems helpful to set out a few contextual matters regarding the nature of statutory repeal.

11. An enactment by definition can only be fully repealed once (Francis Bennion, Statutory Interpretation 4th ed. (London, 2002) says that the effect of repealing a (fully) repealed Act is “none at all” (p. 252)). By contrast an enactment can be partially repealed (e.g. for various different purposes or in relation to various different provisions) a number of times. Repeals can be looked at in a number of ways (express or implied, for example) but perhaps are best viewed as coming in four varieties.

(i). Firstly there is the full repeal without express qualification. A repeal simpliciter without express additional qualification must be taken as a full repeal, as to do otherwise would create significant uncertainty. The effect of a repeal is as if the statute had not been enacted except in respect of matters past and closed (see Surtees v. Ellison (1829) E.R. 594 (1829) 9 B. & C. 750, Eton College v. Minister of Agriculture [1964] Ch. 274). Full repeal is thus subject to the inherent intention that past transactions are not disturbed, and in addition, in relation to past matters, is subject to general savings under the 2005 Act.

(ii). Secondly, one can have a repeal with express qualification, but where such qualification refers only to past matters, above and beyond the specific past matters catered for in the 2005 Act. And indeed, express savings provisions normally do deal only with past transactions (including transactions “in the pipeline”), where more detail is thought necessary than is provided by the 2005 Act. Such a repeal can also be regarded as a full repeal.

(iii). Thirdly, there may be specific savings provisions in the repealing Act which provide that the Act continues to have ongoing effect for new future transactions or matters. If such provisions are included then the original Act has not been fully repealed (whatever the ostensibly full wording of the repealing section may be). This is, in reality, a partial repeal even though it may look full at first sight. Such enactments may have “an uneasy existence as repealed provisions which are somehow still in force” (Bennion, 4th ed., p. 258).

(iv). Fourthly, a repeal can be expressly partial. A partial repeal may have effect in relation to certain provisions of the statute, leaving others intact, or more complicatedly may repeal the original provision for certain purposes or in respect of certain of its applications, leaving other purposes or applications intact.

13. A repealing Act may or may not seek to sweep up all references to the repealed Act that have been made in other previous legislation. Generally, in Irish drafting that has not always been thought necessary, although it seems to happen with greater frequency in the U.K. Failure to tidy up such references could not in and of itself be held to mean that there is an intention on the part of the legislature that the repealed Act should continue in force for some purpose. That would create extraordinary uncertainty and I have no doubt would fundamentally undermine the actual intentions of the Oireachtas in many cases. The real legislative intention is signposted in s. 26 of the 2005 Act - the construction of such references depends on whether there is a corresponding provision in the replacement enactment.

The scope of s. 50 of the 2015 Act
14. Section 50(1) of the 2015 Act provides for non-refoulement of “a person” but sub-s. (7) defines “person” as “a person who is, or was, an applicant”. An applicant is defined in s. 2(1)(a) as a person who “has made an application for international protection in accordance with section 15, or on whose behalf such an application has been made or is deemed to have been made” and who has not ceased to be an applicant (s. 2(1)(b)).

15. This is convoluted drafting. There is no valid reason why the reality of the scope of the section should not be brought out more into the open by referring in subs. (1) to the principle that “a person who is or was an applicant shall not be expelled …” etc.

16. In addition, the effect of the drafting is that the present applicant, Ms. G., is not a “person” for the purposes of s. 50 of the 2015 Act. Thus she must be a “nonperson”, or an “unperson”, as Orwell would have had it. This is ugly drafting by any standards.

17. Furthermore, it is tendentious drafting. In referring to “a person” not being expelled from the State, the section gives the impression that it is mirroring the language of international law (art. 33(1) of the Geneva Convention; art. 3(1) of the U.N. Convention against Torture). But by going on in the small print to define “person” as meaning a person who is or was an applicant, the drafter has defined an ordinary word in extraordinary terms. Subsection (1) seems designed to give the impression that effect is being given to international law in an unqualified way, whereas in fact there is a very significant qualification and a very unnatural definition lurking later in the innards of the section.

18. Persons whose proceedings are in the pipeline are deemed to be applicants under s. 70, but each of the categories of persons covered by that section are the subject of applications that have not been finalised. The only exceptions are s. 70(7) and (8):

(i). Under subsection (7), certain applicants whose applications are not deemed to be made under the 2015 Act will continue to be dealt with under the 1996 Act, but it is clear that that only applies if “his or her application under the Act has not been determined”. That is clearly not this applicant.

(ii). Subsection (8) provides that a person who is either an applicant under reg. 2(1) of the European Union (Subsidiary Protection) Regulations 2013 (S.I. No. 426 of 2013) or a person entitled to make an application for subsidiary protection thereunder whose application is not deemed to be made under the 2015 Act continues to be dealt with under the 2013 Regulations. Again, that provision does not seem to apply to this applicant: certainly it has not been contended that it does. (The fact that I have therefore not been called upon to extract any meaning from the particular pit of obscurity that is s. 70(8) is one of the brighter spots on the canvas of this case).

Categories of deportation order
20. Originally the sole power to make a deportation order was set out in s. 3 of the 1999 Act, subject to the non-refoulement provisions of s. 5 of the 1996 Act. The fact that it was for some reason thought necessary to overlay a new type of deportation order created by s. 51 of the 2015 Act on top of the previous system for deportation orders in s. 3 of the 1999 Act, which continues for some purposes, has created five categories of deportation order. They are as follows:

(i). Deportation orders made pre-commencement of the 2015 Act. Those orders continue unaffected, because s. 3 of the 1999 Act remains in force.

(ii). Deportation orders made post-commencement of the 2015 Act against persons who were never protection-seekers. Such orders must now be made under s. 3 of the 1999 Act.

(iii). Deportation orders made against persons who had protection applications in progress on the commencement of the 2015 Act. Because those applications are deemed to be made under the 2015 Act by virtue of s. 70 of that Act, a deportation order against any such person is now made under s. 51 of the 2015 Act, subject to the non-refoulement provisions of s. 50 of that Act.

(iv). Deportation orders made against persons whose protection applications were fully finalised pre-commencement of the 2015 Act but where the order is made post-commencement. In such a case a person is not deemed to be a protection seeker under the new Act, so the order is made under s. 3 of the 1999 Act. The question then is where, if anywhere, the protection against refoulement in such a case is to be found. This applicant falls into category 4.

(v). Deportation orders made against persons who make protection applications post-commencement of the 2015 Act. Like category 3, a deportation order against any such person is now made under s. 51 of the 2015 Act, subject to the non-refoulement provisions of s. 50 of that Act.

	D.O. made pre-31/12/16	D.O. made post-31/12/16
Applicant not a protection seeker	Category 1 - 1999 Act	Category 2 - 1999 Act
Applicant applied for protection pre-31/12/16 and protection application finalised pre-31/12/16	Category 1 - 1999 Act	Category 4 - 1999 Act
Applicant applied for protection pre-31/12/16 and protection application pending as of 31/12/16	Not applicable	Category 3 - 2015 Act
Applicant applied for protection post-31/12/16	Not applicable	Category 5 - 2015 Act

The form of the deportation order under the 1999 Act
22. The form of the deportation order under the 1999 Act was set out in the Immigration Act 1999 (Deportation) Regulations 2005 (S.I. No. 55 of 2005) which were made under s. 7 of the 1999 Act and which prescribed the form of order for the purposes of s. 3(7) of that Act. On 31st December, 2016, as noted above, the 1996 Act was repealed in its entirety by the commencement of the 2015 Act. That entire repeal of course included the statutory non-refoulement provision in s. 5 of the 1996 Act. The 2015 Act enacted a new, and as we have noted, somewhat broader refoulement provision in the form of s. 50 which, as noted above, was expressed not to apply to everybody but only to applicants under the 2015 Act. Despite the repeal of s. 5 of the 1996 Act on 31st December, 2016 the deportation order made in this case on 13th January, 2017 recites s. 3(1) of the 1999 Act, stating that a deportation order may be made subject to s. 5 of the 1996 Act and also recites that s. 5 has been complied with.

23. Somewhat belatedly, the Minister then made the Immigration Act 1999 (Deportation) (Amendment) Regulations 2017 (S.I. 74 of 2017), on 8th March, 2017, which deleted reference to s. 5 of the 1996 Act from the form of the deportation order. On the face of it, due to the unfortunate delay involved in making those regulations, there was an interregnum period between 31st December, 2016 and 8th March, 2017 where deportation orders contained an error on the face of the record, stating that a repealed statute had been complied with. The State denies that there is any error and I will deal with this issue later in the judgment.

Is there a statutory protection against refoulement for the purposes of s. 3 of the 1999 Act and if so where is it to be found?
24. Ms. Stack informs me that for persons the subject of orders under s. 3 of the 1999 Act, the matters set out in the repealed s. 5 continue to be considered in practice as to their substance, but the section is not mentioned expressly in the Minister’s analysis. Thus the Minister is not actually pinning his colours to the mast in the decision itself as to which statutory provision refoulement has been considered under. This perhaps retrospectively explains the situation that I encountered in H.A.A. (Nigeria) v. Minister for Justice and Equality [2018] IEHC 34, where I noted that the deportation analysis failed to refer to any specific statutory provision in terms of refoulement which was one of the factors, combined with other factors, that led me to quash the decision in that case. That was also a case where the parties agreed that the reference to s. 5 should be read as a reference to s. 50 of the 2015 Act, but the Department in this case is contesting that position. Whichever way one looks at the matter, it is not appropriate or acceptable having regard to basic rule of law considerations for the Department to be intentionally obscuring the nature of the statutory exercise they are engaged in. What seems to be happening is that they believe on legal advice that they should be applying the repealed s. 5 (whether that advice is correct I will address later), but clearly they are in sufficient doubt about it not to mention s. 5 explicitly. That is not an appropriate mode of procedure on any view. What is needed is express clarification of what the appropriate section to be applied is, and the present case now provides the vehicle for that clarification to be provided. Future analysis by the Department will need to expressly refer to the appropriate statutory provision as so clarified.

25. To summarise the difficulty from another point of view, the problem is that if one looks at the five categories of deportation order I mentioned above, categories 2 and 4 do not on the face of it have a specific refoulement provision to protect applicants, the reason being that the order in those cases is made under s. 3 of the 1999 Act which refers to s. 5 of the 1996 Act, which has now been repealed.

26. Frequently one simply has to adjudicate between two competing submissions. That in itself is often complex enough. However, both sides in the present case have advanced alternative theories, the applicant offering two possible theories and the Minister, at various points in the submissions made, suggested up to three theories. I can endeavour to summarise the five resultant alternatives as follows:

(i). One option is to say that the reference to s. 5 in the 1996 Act simply falls away because the section has been repealed. The statutory protection against refoulement must be found elsewhere, in particular in the European Convention on Human Rights Act 2003, insofar as it implements arts. 2 and 3 of the ECHR.

(ii). A second option is to deem s. 5 of the 1996 Act as continuing in force for the purposes of s. 3 of the 1999 Act on the basis that it is mentioned in s. 3 and that mention has not been repealed by the repealing Act.

(iii). A third option is to apply s. 27(1)(c) of the 2005 Act and hold that the right of the applicant to rely on s. 5 of the 1996 Act must be preserved notwithstanding the repeal of that Act.

(iv). A fourth approach would be to interpret the definition of “applicant” in s. 2 of the 2015 Act or of person in s. 50(7) of that Act or the transitional provisions in s. 70 in such a way as to encompass persons in the lacuna category into which this applicant falls such that this applicant would be deemed an applicant within the 2015 Act and would be subject to s. 50 of that Act.

(v). The final option is to interpret the reference in s. 3 of the 1999 Act to s. 5 of the 1996 Act as referable to s. 50 of the 2015 Act, pursuant to s. 26(2)(f) of the 2005 Act.

Option 1
28. Option 1 can be dismissed immediately as it ignores the provisions of s. 26 of the 2005 Act. Furthermore, it would significantly diminish the statutory protection against refoulement in a manner that it is hard to imagine was intended by the Oireachtas because prior to the enactment of the 2015 Act, applicants enjoyed the protection of both the 2003 Act and a specific provision in relation to refoulement. Articles 2 and 3 of the ECHR are on one view (both because they apply only to the death penalty, torture and inhuman treatment and because of the qualifications in the European Convention on Human Rights Act 2003) potentially narrower than a situation where an applicant has the protection of both the ECHR and a specific non-refoulement section.

Option 2
29. The second option, which was only faintly pressed at one stage in argument by Ms. Stack, would create massive uncertainty. The statute book has many examples of references in previous legislation to repealed Acts that have not been fully tidied up by the repealing Act. One of four situations can arise there:

(i). Firstly, an Act is repealed without replacement, in which case a reference to that repealed Act in some other legislation can normally be taken to be impliedly repealed.

(ii). Secondly, the Act can be repealed with a replacement, but the replacement does not have any provision corresponding to the particular provision at issue. In such a case, a statutory reference in some other Act to that latter provision can be taken to also be impliedly repealed (or it is to be “disregarded” to use the term in s. 26(2)(f) of the 2005 Act).

(iii). Thirdly, the Act can be repealed and replaced with something similar (“provisions … relating to the same subject-matter”), in which case s. 26 of the 2005 Act may apply.

(iv). Finally, the Act may be repealed and replaced with something so different from the original that s. 26 can have no application.

31. It was suggested at one stage that there was some analogy with a case where a contract such as a mortgage document sets out the intention of parties in such a way that it was to be construed as having ongoing effect after repeal of a statute referred to (see Dowdall v. O’Connor [2013] IEHC 423 (Unreported, McDermott, 11th September, 2013). However, the public law context of the interpretation of statutes is completely different and not only requires much more certainty but also takes place within very clear rules of interpretation as set down in the 2005 Act. The second option misunderstands principles and practices of statutory drafting, and would create far more problems than it would solve.

Option 3
32. There are a number of fundamental difficulties with the third option. First of all, it creates a further anomaly in of itself in that this interpretation provides no protection whatsoever for persons who are not protection seekers and who are now made the subject of deportation orders. (Such persons are in category 2 of the five categories of deportation order to which I have referred.) Such persons have no vested “right” on the State’s theory and thus can benefit neither from s. 5 of the Act as allegedly continued by s. 27 nor from s. 50, as they are not applicants under the 2015 Act.

33. More fundamentally however, the State’s reading of s. 27(1)(c) fails to comport with what was stated by O’Donnell J. in Minister for Justice Equality and Law Reform v. Tobin (No. 2) [2012] IESC 37 [2012] 4 I.R. 147 at p. 353 relying on Bennion 4th edn. (London, 2002): the type of right at issue for s. 27 to be activated “must not have been a mere right to take advantage of the enactment now repealed" (Bennion, p. 259). O’Donnell J. also cited Craies on Legislation, 9th edn. (London, 2008) para. 14.4.12: “the saving does not apply to a mere right to take advantage of a repealed enactment (clearly, since that would deprive the notion of a repeal of much of its obvious significance). Something must have been done or occurred to cause a particular right to accrue under a repealed enactment."

34. Bennion’s phrase quoted by O’Donnell J. in turn cites, at p. 259, n.5, three cases on the subject:

(i). The first is Abbott v. Minister of Lands [1895] AC 425. Lord Herschell L.C. said at p. 431 that “It has been very common in the case of repealing statutes to save all rights accrued. If it were held that the effect of this was to leave it open to any one who could have taken advantage of any of the repealed enactments still to take advantage of them the result would be very far reaching. It may be, as Windeyer J. observes, that the power to take advantage of an enactment may without impropriety be termed a “right.” But the question is whether it is a “right accrued” within the meaning of the enactment which has to be construed. Their Lordships think not, and they are confirmed in this opinion by the fact that the words relied on are found in conjunction with the words “obligations incurred or imposed.” They think that the mere right (assuming it to be properly so called) existing in the members of the community or any class of them to take advantage of an enactment, without any act done by an individual towards availing himself of that right, cannot properly be deemed a “right accrued” within the meaning of the enactment.” The reference to acts done towards availing of the “right” does not I think mean that merely doing anything to rely on a statute creates a shadow ongoing existence for that statute after its repeal. The act done must be such as to make it unjust not to apply the Act. In the present context, merely to disappoint an applicant by changing the law does not constitute injustice such that making representations under the former Act constitutes an accrued right that cannot be interfered with.

(ii). Secondly, Bennion cites Hamilton-Gell v. White [1922] 2 K.B. 422 at 431, which appears to be a reference to the comment of Atkin L.J. as he then was that “the tenant in this case has acquired a right to claim compensation under the Act of 1908 on his quitting his holding” . That instances the sort of property-type right that is intended by the reference to accrued rights in s. 27.

(iii). The third case cited is Chief Adjudication Officer v. Maguire [1999] 1 WLR 1778 [1999] 2 All E.R. 859, where the Court of Appeal held that a claimant who before the repeal of social welfare legislation satisfied all the preconditions to entitlement to the benefit save only that of making the requisite claim, such claim then being made within the prescribed period, albeit after repeal, had an accrued right preserved notwithstanding the repeal. Again that is the sort of property-type financial entitlement to which s. 27 is directed. Clarke L.J. (as he then was) commented at 1791 that “It is surely far better for the statute to state clearly what rights are to survive and what rights are not, so that fine distinctions and the costs of endless debate as to whether a particular alleged right has been acquired or not can be avoided.”

36. The problem for the State here is that it is not unfair to “deprive” an applicant of the benefit of s. 5 of the 1996 Act by declining to regard the right to make submissions under that section as a vested right for the purposes of s. 27 of the 2005 Act, as the State suggests at para. 22 of their submissions. That is for two reasons. More generally, an applicant does not have any “right” to take advantage of legislation after its repeal absent some more specific injustice. But more specifically, the alternative to unconvincingly applying s. 27 is to apply s. 26, which provides more extensive protection for applicants through the mechanism of construing the reference to s. 5 as being one to s. 50 of the 2015 Act, which provides wider protection for applicants than s. 5, as we have noted. The State’s deeply touching and tearful concern that unfortunate applicants might be deprived of the benefit of their “rights” actually amounts here to a concern that applicants would be deprived of the “right” to be processed under a less favourable procedure than that that would apply if the State’s position was rejected. With rights like that, who needs duties? The State submissions could only be more cynical if they were accompanied by an alternative argument that the “right” was that of the State to subject the applicants to the “liability” of being processed under s. 5 of the 1996 Act rather than s. 50 of the 2015 Act. Thus there is in fact no detriment to applicants. But even if there was such detriment, that would not be an injustice. Everyone has to live with the potential for changes in legislation and the possible refinement or reduction of statutory entitlements from time to time, within constitutional boundaries. Bennion goes on to say that “Being able to avail oneself of a statutory defence is not a ‘right’ for this purpose” (p. 259), which perhaps illustrates the point further.

37. While the State may think it is getting itself out of a scrape by launching the argument that s. 27 applies here, in fact, were I to accept this misconceived submission, it would create huge difficulty for the organs of public administration because it would frustrate the entitlement of the Oireachtas to repeal legislation going forward without having to face endless complaint that individuals had a right to take advantage of the repealed legislation in some way, shape or form. The “right” to make submissions or to have one’s submissions under the repealed section considered or even to take advantage of the section without more is simply not the sort of “right” that s. 27 is meant to apply to. All that happened in the 2015 Act was that a procedure set out in a particular statute was replaced with another procedure. The fact that, but for the repeal, the person would have been subject to the former procedure is not a “right”. There is no right to have a statute applied to one after its repeal merely because one has made representations about it prior to the repeal. Section 27 was clearly meant to cover vested rights such as property rights or matters of that nature, or situations where it would be clearly unjust to preclude the person from completing a process under the repealed legislation. None of those situations apply here.

38. Obviously in the present context there are certain human rights limitations on refoulement. But the key point is that those limitations arise from the Constitution and international law insofar as it is relevant, and not from the mere fact that a statute had been enacted giving such rights express embodiment in a particular form. In other words, the mere former existence of the statute combined with having made submissions under it, or still less having had the opportunity to do so, does not create a vested “right” to continue to take advantage of the procedures provided under its wording, despite its repeal. That is so anyway but is even more emphatically so where the new provision is even more beneficial to applicants than the old provision, and can be applied to them via s. 26 of the 2005 Act.

Option 4
39. The fourth interpretative action can also be dismissed immediately. Absent any basis (such as option 5) to modify the wording of the 2015 Act by virtue of the 2005 Act, to hold that this applicant is an “applicant” under the 2015 Act would fly in the face of the express wording of the latter Act.

40. I would also note that this option would fail to resolve a number of the anomalies that also remain unsolved by option 3.

Option 5
41. The only remaining option is option 5, which is clearly the most satisfactory one for a range of reasons (and not simply on the approach that “when you have eliminated the impossible, whatever remains, however improbable, must be the truth” (Arthur Conan Doyle, The Sign of the Four (1880) Ch. 6)).

42. Firstly, there is absolutely no basis to consider that the intention of the legislature was to leave potential deportees without the benefit of a statutory refoulement provision. Indeed the object of the 2015 Act is to enhance the level of protection made available by streamlining procedures in a number of ways. Furthermore, given that the protection under s. 50 is wider than that under the original s. 5, it seems to me that an interpretation that provides a more consistent level of protection for human rights by affording the benefit of s. 50 to all deportees is more likely to be in keeping with the legislative intention.

43. Furthermore, no reason whatsoever was put forward as to why the Oireachtas would have legislated for a situation where failed protection seekers under the 2015 Act would benefit from a more extensive refoulement provision than other persons such as this applicant (who is a failed protection seeker under previous legislation), and indeed why category 2 deportees would have no specific non-refoulement provision at all above and beyond the ECHR. It is hard to see how such an approach is in keeping with either the letter or spirit of equality before the law.

44. Section 26(2)(f) of the 2005 Act applies when a “former enactment” [the 1996 Act] is repealed and re-enacted with modification by a “new enactment” [the 2015 Act] and the court is then called upon to interpret a reference to the former enactment in “any other enactment” [the 1999 Act]. The main modification for present purposes is that the refoulement provision is set in the context of new application machinery and is confined, apart from any application or modification inherently necessitated by the application of s. 26 of the 2005 Act, to applications made or deemed to be made under the 2015 Act. Section 26(2)(f) provides that “a reference in any other enactment to the former enactment shall, with respect to a subsequent transaction, matter or thing, be read as a reference to the provisions of the new enactment relating to the same subject-matter as that of the former enactment”.

45. The State in para. 38 of the written submissions made the extraordinary argument that “the 2015 Act is not a repeal and re-enactment of the 1999 Act and consequently s. 26(2)(f) has no application”. Even by the sometimes debased standards of forensic illogicality, the sentence quoted from achieves the not-inconsiderable feat of scaling new depths. Given the technicality of the matter at issue, this may be perhaps more clearly seen if the State’s submission is rephrased in syllogistic form:

Major Premise: Where an enactment (“former enactment”) (the 1996 Act) is repealed and re-enacted, with or without modification, by another enactment (“new enactment”) (the 2015 Act), a reference in any other enactment (the 1999 Act) to the former enactment shall, with respect to a subsequent transaction, matter or thing, be read as a reference to the provisions of the new enactment relating to the same subject-matter as that of the former enactment [S. 26 as applied to the present context].

Minor premise: The new enactment (the 2015 Act) did not repeal the other enactment (the 1999 Act) [Para. 38 of State submissions.]

Conclusion: Therefore the section has no application [Para. 38 of State submissions.]

47. The 2005 Act applies when the “former enactment” is repealed. By obvious definition it has nothing to do with a reference in the repealed enactment itself. If to look at the other side of the coin of the point absurdly made by the State, if the 2015 Act was a repeal and re-enactment of the 1999 Act we would not be concerned with what references to other Acts in the 1999 Act mean because that Act would have been repealed. For the purposes of s. 26, the 1999 Act is not what the 2005 Act calls the “former enactment”, it is the “other enactment”. The fact that the 1999 Act has not been repealed is totally irrelevant. The submission made completely misunderstands the nature and effect of s. 26.

48. A further argument advanced by Ms. Stack is the submission that to accept Mr. Leonard’s submission in relation to s. 26 would “open the floodgates” because she said the 2015 Act is not a re-enactment with modifications of the 1996 Act but an entirely new system. Ms. Stack submits that if s. 26 is applied here, the power of the Oireachtas to restart an area of law by creating a new system would be “completely restricted” and that such an interpretation would “prevent efforts to create a new system”.

49. Even overlooking the fact that bogeyman floodgate arguments from the State have become rather less effective over time through promiscuous and indiscriminate over-use, the submission made is unfortunately a complete misunderstanding of s. 26. The Oireachtas can create whatever new system it wants, whether it updates old statutory references or not. Only if the legislature fails to update references to the old, repealed, system, does s. 26 come into play. But that does not mean that there cannot be a new system or that the new system cannot work effectively. Quite the reverse. In that situation, the court in an interpretative capacity has an obligation to apply s. 26 of the 2005 Act to give effect to the intention of the Oireachtas as expressed in that section, which is to promote continuity of public administration whereby any stray references to a repealed Act should be read as normally referring to the corresponding provision of the replacing Act. That is not only common sense but overall is massively to the advantage of the executive and to the orderly carrying out of public administration generally.

50. A system may well be a completely new system but may still have provisions, to use the terms of s. 26, “relating to the same subject matter as that of the former enactment”. The fact that a completely new procedure is put in place by the new enactment is not a reason not to apply s. 26. For example, preliminary examination in s. 8 of the Criminal Procedure Act 1967 was replaced by a completely new procedure under the Criminal Justice Act 1999, but it was held nonetheless that reference to the 1967 Act in another enactment should be read as referable to the 1999 Act by virtue of s. 26(2)(f) (see Lavole and Carvida Limited v. O’Donnell (Unreported, O’Neill J., 3rd March, 2005).

51. Reliance is also placed on the commentary in s. 17(2)(a) of the Interpretation Act 1978, in Craies on Legislation, 11th edn. (London, 2017), that a rule of this nature is a “rule of thumb that will suffice in comparatively simple cases, but where the conversion of references is at all complicated the re-enacting legislation will be well advised to make express provision about the conversion of references (or to give power for consequential provision of that kind to be made)” (p. 710). That is all well and good as advice to drafters, which is what that passage purports to be, but what is to happen if the legislature fails to do that? The court must interpret the reference to the repealed enactment in some way and it seems to me the obvious way is to interpret it as referring to the corresponding provision of the 2015 Act, namely in this case s. 50 by virtue of s. 26 of the 2005 Act.

52. Having said that, the note of caution in Craies is also echoed in Bennion, 4th ed., who argues for “great caution” in applying the corresponding U.K. principle, “because of the vagueness of the word ‘modification’… If it is held to cover anything more than minor modification it may alter rights and liabilities in unintended ways” (p. 260). Bennion cites, at n. 5, R. v. Secretary of State for Social Security ex p. Britnell [1991] 1 W.L.R. 198, per Lord Keith of Kinkel, and Stevens v. General Steam Navigation Co. [1903] 1 KB 890; but those cases do not in fact seem to support his proposition and rather lean very strongly in the other direction. Indeed, Lord Keith of Kinkel endorses a very wide definition of modifications at p. 203: “I am of the opinion that the word “modifying” here used is capable on a proper construction of covering the extension of the scope of an enactment. In Stevens v. General Steam Navigation Co. Ltd. [1903] 1 KB 890 it was held that the word “modification” in section 38(1) of the Interpretation Act 1889 included an addition. The Factory and Workshop Act 1901 added to the definition of “factory” in the Factory and Workshop Act 1895 , which it repealed, machinery used in the unloading of ships in a navigable river. The definition of “factory” in the Workmen's Compensation Act 1897 referred to the Act of 1895. That reference was construed, by virtue of section 38(1) of the Interpretation Act 1889, as being a reference to the Act of 1901, thus incorporating the wider meaning thereby attributed to the word.” He went on to endorse the view of Collins M.R. in Stevens at pp. 893-894, that “there is no reason to limit the word ‘modification,’ which is equally applicable whether the effect of the alteration is to narrow or to enlarge the provisions of the former Act.”

53. These authorities on the wide scope of modifications seem to me to be a complete answer to the State’s submission (para. 40) that “the provisions of s. 50 are broader in scope than s. 5 of the Refugee Act, 1996, and contain a further basis on which refoulement can be claimed. As such, s. 50 does not relate “to the same subject matter”, as required for the application of s. 26(2)(f) and therefore that provision does not apply.” Unacceptably, the submission seeks to shrink s. 26(2)(f) to Lilliputian proportions contrary to a purposive interpretation, and indeed contrary to authority. But at the same time, one cannot but be struck by the short-termism of this submission. The State is seeking to emasculate the very useful statutory principle of legislative continuity embodied in s. 26(2)(f), which generally operates very much in favour of the machinery of public administration and is capable of bridging many a gap in the garment of the statute book. The notion that this should be cast aside or reduced to bonsai proportions just to provide a quick fix for this one case is unattractive in the extreme. The role of the court is to give effect to the intention of the Oireachtas subject to the Constitution, and I can discern no intention that s. 26 be construed in such a narrow, mean-spirited and artificial manner. Section 26(2)(f) promotes the important principle of continuity of the statue book and should be construed to do so to the fullest appropriate extent, by analogy with the way Article 50.1 of the Constitution continues previous law, as did Article 73 of the Free State Constitution and indeed Article 8 of the Articles of Union set out in the Act of Union (Ireland) 1800. The same subject matter means the same broad subject matter, not literally the same words or precisely the same coverage. There may inevitably be addition, subtraction or amendment, but if the broad subject matter - in this case, refoulement - remains the same, then the principle of continuity embodied in the section may have application.

54. This takes us to a more fundamental point, that s. 26 of the 2005 is in wider language than s. 20(1) of the Interpretation Act 1937, which required a “re-enactment” of the previous statute, with or without modification. It provided that “Whenever any statute or portion of a statute is repealed and re-enacted, with or without modification, by an Act of the Oireachtas, references in any other statute or in any statutory instrument to the statute or portion of a statute so repealed and re-enacted shall, unless the contrary intention appears, be construed as references to the portion of such Act of the Oireachtas containing such re-enactment.” The 2005 Act simply refers to there being provisions relating to the same subject-matter. Thus the 2005 Act is significantly more flexible than either the previous Irish law or the U.K. law, flexible and all as the latter actually seems to be.

55. In any event, here, it seems to me that firstly the tweaking in the refoulement provision as between s. 5 of the 1996 Act and s. 50 of the 2015 Act, while important, is firmly in the category of “minor” change compared to many amendments that one could envisage, keeping us well within the concept of the same subject-matter, and furthermore applying s. 26 to the situation by reading s. 5 as meaning s. 50 is far more likely to represent the intention of the Oireachtas than to hold that certain categories of applicant are now to have either no statutory provision regarding refoulement, or for some equally unexplained reason are to have a different and lesser level of protection.

56. This brings us to the serious difficulty with the State’s submission insofar as it involves distinctions being drawn between different categories of applicants and the creation of other anomalies. I have noted above that category 2 deportees will be without protection other than under the ECHR as applied by the European Convention on Human Rights Act 2003. But that is not the only anomaly.

57. One must also consider any other references to s. 5 of the 1996 Act, for example that in s. 5(1) of the Immigration Act 2003. On the State’s logic it is hard to see how that can be shoe-horned within the confines of s. 27 of the 2005 Act. Again, to make sense of that provision, the best way to interpret it is to apply s. 26 there as well and to construe that reference to s. 5 of the 1996 Act as referable to s. 50 of the 2015 Act. Thus the application of s. 26 to the present problem also provides coherence in respect of immigration law generally.

58. Of course, the provisions of s. 50(7) do not apply, by necessary implication, in any context where s. 5 is read as meaning s. 50 in respect of a person who is not an applicant under the 2015 Act. It is the earlier provisions of s. 50 (i.e., sub-ss. (1) to (6)) that must be read as being the new provisions to which the references to s. 5 must now be taken as referable. While s. 26 does not use the phrases “with any necessary modifications” or “save where the context otherwise requires”, these are inherent in the nature of the exercise being endorsed by the section, namely reading a reference to one statute as if it referred to another. By definition the language of the target statute must be applied mutatis mutandis rather than with some sort of literalistic pretence that no modifications are permissible. That is the fundamental fallacy in the State’s argument that “even if the references in s.3 of the 1999 Act to s.5 of the 1996 Act were read instead to mean references to the 2015 Act, this would still not avail the Applicant, as, by definition, she is not a “person” to whom s. 50 applies” (submissions para. 37). It is the nature of the exercise engaged in by the application of s. 26 itself that inherently involves any necessary modifications being read in, the pertinent one being the disregard of sub-s. (7) of s. 50. Think about the alternative. If no necessary modifications were impliedly allowed to be applied, then the reference to the “former enactment” could only be read as referable to the “new enactment” if every last fine micro-detail of the new enactment could be directly mapped onto the former enactment in a word-for-word manner. That would totally defeat the clear statutory intention in s. 26 that all that is required is that the new enactment deals with “the same subject matter”. It would also render s. 26(2)(f) all but meaningless as the condition of exact correspondence would only very rarely be satisfied.

59. This interpretation creates no absurdity and leaves no applicant bereft of protection. Nor does it, contrary to the completely overblown rhetoric of the State, create any floodgate problem. On the contrary. Were the court to support the alternative options being advocated by the State, severe interpretative difficulties would arise in other cases, even if it got the State over the hump of the present case. But the court must be concerned with giving effect to the legislative intention and making the system of public administration work, even and indeed especially if the executive has failed to think through the ramifications of any particular ostensibly utilitarian quick fix.

60. The solution of applying s. 26 is simple and elegant, preserves and promotes the rights of applicants, ensures comprehensive and equal protection against refoulement, avoid anomalies and discrimination, opens no floodgates, avoid the severe downstream uncertainty created by other interpretative options, provides coherence in immigration law in general, provides equivalent workability in the context of s. 5 of the Immigration Act 2003, and gives effect to the intention of the legislature, as expressed both in the 2005 Act in general and the 2015 Act in particular. It provides a sustainable, beneficial and purposive reading of the 2005 Act in its application to legislation in general, and avoid undesirable interpretations that would thwart either the effectiveness of repeals by over-expanding s. 27 of the 2005 Act, or the principle of continuity by under-cutting s. 26 of that Act. It provides a clear position that will work in practice and will enable the 1999 Act and indeed the 2003 Act to be operated effectively into the future. It is simply the obvious answer.

Consequences for the present case
61. The conclusion for the purposes of resolving the present proceedings is that both the deportation order and the analysis supporting it contain an erroneous statement that the order is subject to s. 5 and that that section has been complied with. Instead, the Minister should have read the reference to s. 5 as being a reference to s. 50 (by virtue of s. 26 of the 2015 Act). As the reference to s. 5 constitutes an error on the face of the record, the appropriate order is to quash the decision. That jurisdiction is not dependent on establishing ultra vires (see De Blacam Judicial Review, 3rd edn. (Dublin, 2016) para. 12.02, Bannon v. The Employment Appeals Tribunal [1993] 1 I.R. 500).

62. I can deal briefly with the remaining grounds of challenge.

Failure to have regard to the duration of the applicant’s residence in the State
63. The applicant complains that the analysis says that she was in the State for four years whereas, due to a long delay in service of the deportation order, she was in the State for four years and nine months as of the date of service. That is a completely spurious point because the consideration of s. 3 arises at the time the order is made, not when it is notified. There may be all sorts of reasons why the order is not notified immediately and the fact that information is by definition a snapshot at a given time is not a stateable reason to question the validity of the order. Delay in service does not invalidate an otherwise valid deportation order. If there is some fundamental change in circumstances in the meantime an applicant can make a revocation application.

Failure to consider evidence
64. The decision-maker found that no further evidence had advanced that might alter the mind of the Minister in relation to the question of serious harm. However, the applicant had submitted further documentary evidence in the form of the attestation dated 11th July, 2016.

65. In written submissions, the State retrospectively sought to retrospectively re-characterise the decision-maker’s finding to mean that “the applicant has not put forward any further evidence which was a sufficient weight to cause the Minister to reach a different conclusion to that reached previously on the issue of risk” (para. 53 of submissions) and “that no new material had been submitted which was sufficiently different or weighty to cause the Minister to conclude that the applicant was at serious risk of harm”.

66. This reinterpretation is tinged with esprit d’escalier and in the particular circumstances of this case is a reprogramming of the decision after the event. That is not only not what the decision is but more importantly on the fact-specific context of this particular case, it is not what the decision means. It says that no new information had been advanced that might alter the mind of the Minister, not that the new information advanced did not in fact alter the Minister’s mind. Now it is quite possible that a decision-maker could use such language but that when read in context (for example if the decision adverts more specifically to what was actually received) it would be clear that it was intended to convey the point that the new information actually furnished was insufficient. No such context or intention appears here on the specific facts of this particular case. The content of the new material is not adverted to in any way so as to qualify the literal meaning of the words used. The fact is the applicant did put forward new information which clearly might have altered the mind of the Minister. The fact that the Minister’s submissions attempt to deconstruct the attestation in detail that is contained nowhere in the analysis and suggest that it contains a point already rejected by the Minister only doubly reinforces this conclusion. The fundamental misunderstanding clearly continues, because it is the fact that the point was already rejected that makes the production of new evidence into something relevant to whether the Minister’s mind might be altered. So it seems to me that the decision should be quashed on this ground also.

Order
67. The take-home point for the ongoing operation of the immigration system is that s. 5 of the 1996 Act has been fully repealed and that pursuant to s. 26 of the 2005 Act, references in any legislation to s. 5 (in particular the reference in s. 3 of the 1999 Act) should now be construed as references to s. 50 of the 2015 Act, which provides a protection against refoulement for all persons being deported, whether such deportation is ordered under s. 3 of the 1999 Act or s. 51 of the 2015 Act.

68. If for any reason the Oireachtas wants to provide different statutory levels of protection against refoulement for different categories of persons, it is, of course, at liberty to do so within constitutional constraints, but I can discern no such intention at the present time. Any such hypothetical amending legislation would, of course, need to have an objective basis.

69. Perhaps most importantly for practical purposes, it is, in my view, not sustainable or in accordance with basic rule of law considerations for decisions under the 1999 Act to be made on a basis that refoulement is considered without acknowledging any particular statutory provision. Doubly so when it is now clear that this is a deliberate policy of fudging to avoid committing to whether s. 5 of the 1996 Act should be considered. Given that the position that it should not is as outlined in this judgment, such decisions need to expressly refer to consideration being given by the Minister to whether s. 50 of the 2015 Act precludes deportation.

70. For all of the foregoing reasons, there will be an order of certiorari removing for the purpose of being quashed the deportation order of 13th January, 2017 made against the applicant. I will discuss with counsel any consequential issues arising.

Postscript
71. Since circulation of the foregoing judgment it has come to my attention that para. 54 above (particularly the word “simply”) is unclear. The point intended to be conveyed by that paragraph was not that s. 26(2)(f) of the 2005 Act does not require a context of repeal and re-enactment (it does - a point acknowledged expressly at paras. 44 and 45 and impliedly by the rationale of paras. 47, 48 and 51) but that by contrast with the 1937 Act it provides a gloss on the term “re-enactment” by introducing the concept that re-enactment can include the enactment of provisions dealing with “the same subject-matter” (a phrase not found in s. 20 of the 1937 Act). The (obviously inarticulate) use of the word “simply” in para. 54 was to contrast that broad and simple concept, applicable here, with the respondent’s narrow and technical interpretation of the term “re-enactment” as requiring some sort of complex and difficult-to-satisfy test of precise, literal, near-identical or even exact correspondence, if not word-for-word regurgitation of the repealed enactment."