Freely Available British and Irish Public Legal Information
[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
Supreme Court of Ireland Decisions
You are here: BAILII >> Databases >> Supreme Court of Ireland Decisions >> Heneghan v Minister for Housing, Planning & Local Government & ors (Approved) [2023] IESC 7 (31 March 2023)
URL: http://www.bailii.org/ie/cases/IESC/2023/2023IESC7(HoganJ).html
[New search] [Printable PDF version] [Help]
AN CHÚIRT UACHTARACH THE SUPREME COURT S:AP:IE:2022:000006 O’Donnell C.J. Dunne J. O’Malley J. Baker J. Hogan J. Between/ TOMÁS HENEGHAN Plaintiff/Appellant and THE MINISTER FOR HOUSING, PLANNING AND LOCAL GOVERNMENT, GOVERNMENT OF IRELAND, THE ATTORNEY GENERAL AND IRELAND Defendants/Respondents JUDGMENT of Mr. Justice Gerard Hogan delivered the 31st day of March 2023 Part I - Introduction Introduction 1. This appeal raises a novel and important issue in respect of the interpretation of constitutional amendments in general and the provisions of Article 18.4.2° of the Constitution in particular. In these proceedings the appellant - a graduate of the University of Limerick - claims in essence that the Oireachtas has failed to give effect to the provisions of the 7th Amendment of the Constitution (Election of Members of Seanad Éireann by Institutions of Higher Education) Act 1979 (“the 7th Amendment”) in the manner required by that constitutional amendment. He contends that as a result, key aspects of the present legislation providing for the election of six university members to the Seanad from the existing (but separate) constituencies of the University of Dublin and the National University of Ireland - namely, ss. 6 and 7 of the Seanad Electoral (Universities Members) Act 1937 (“the 1937 Act”) - have been rendered unconstitutional by reason of the failure of the Oireachtas to extend the franchise to other universities (including the University of Limerick) and institutes of higher education in the manner contemplated by the 7th Amendment. This appeal, accordingly, concerns the contention that the Oireachtas has failed in its constitutional obligation to give effect to the provisions of Article 18 of the Constitution in the manner said to be required by the amendments to that provision effected by the 7th Amendment. 2. In the High Court, a Divisional Court (Irvine P., Simons and O’Moore JJ.) rejected his claim in a comprehensive judgment delivered by O’Moore J. on the 17th November 2021: see Heneghan v. Minister for Housing, Planning and Local Government [2021] IEHC 716, [2022] 1 ILRM 237. Given the manifest importance of these proceedings, this Court granted the appellant leave to appeal directly to this Court in accordance with Article 34.5.4° of the Constitution: see [2022] IESCDET 70. 3. I should say at the outset that I agree with the judgment which Murray J. is about to deliver insofar as we both that the appeal should be allowed. He has comprehensively set out the facts and identified the salient issues in his judgment. I propose simply to address the particular issues of constitutional interpretation raised on this appeal and the remedy for the unconstitutionality thereby identified by way of a separate, concurring judgment. Part II - Is Article 18.4.2 of the Constitution purely permissive in character? 4. As originally enacted by the People on 1st July 1937, Article 18 of the Constitution provided for the election of three university senators by the University of Dublin and three by the National University of Ireland. The original drafting of Article 18 as enacted by the People in 1937 was clear, precise, parsimonious and elegant. The same, unfortunately, cannot be said of the amendments to Article 18, which were effected some forty-two years later by the 7th Amendment following a referendum held in July 1979. I cannot avoid observing that at every level the drafting of this constitutional amendment was hapless, incoherent and confused. In legal terms it was the equivalent of the attempted cleaning of an old master by a careless restoration artist who then proceeded to leave an ink-stain on a Rembrandt. These are admittedly harsh criticisms, but they are, regrettably, fully justified. One could accordingly point immediately to the following drafting deficiencies attending the 7th Amendment. 5. First, while Article 18.4.1.i° and Article 18.4.1.ii° of the Constitution provide in mandatory terms that three Senators shall each be elected by the National University of Ireland and the University of Dublin, Article 18.4.2° (inserted by the 7th Amendment) appears to contemplate the future election in accordance with law of the six university senators by these universities and by other institutions of higher education in the State. Yet Article 18.4.2° does not specify what is to happen to the express wording of Article 18.4.1° in the event that such legislation is enacted and if the franchise is so extended by law. There is, accordingly, in this respect a complete discordance between the wording of these two provisions. 6. Second, Article 18.4.2° speaks of the six members being elected “by one or more of the following institutions”, namely, “i. the universities mentioned in subsection 1 of this section, ii. any other institutions of higher education in the State.” 7. On one view, it would seem that for this purpose both the National University of Ireland and the University of Dublin constitute one institution and the other institutions of higher education in the State constitute the other institution. It is also clear, however, that one of the objectives of the 7th Amendment was to facilitate the possible dissolution of the National University of Ireland given that Article 18.4.3° provides that nothing in this Article “shall be invoked to prohibit the dissolution by law of a university mentioned in subsection 1 of this section.” If this, however, were to occur it is then unclear how the balance of Article 18.4.2° could properly function since - on one view, at least — it seems to identify both the National University of Ireland and the University of Dublin as one single institution for this purpose. 8. Third, Article 18.4.2° refers simply to “any other institutions of higher education in the State.” The drafters presumably simply had in mind institutions such as the modern-day University of Limerick and Dublin City University. But the term actually used is much broader than this. The Royal College of Surgeons in Ireland, the Honourable Society of King’s Inns and the Royal Irish Academy of Music are, for example, all long established institutions of higher education in the State. Dozens - even hundreds - of other similar institutions could be cited for this purpose, yet the wording of Article 18.4.2° gives no real guidance on this point. Was it, for example, intended that the franchise should be extended to all such institutions? If this, however, was not the actual intention, one might ask on what basis could the Oireachtas properly distinguish between these various heterogeneous types of institutions of higher education? 9. It is thus unclear, for example, as to the extent to which the Oireachtas enjoys any real discretion in this matter. If, for instance, legislation was to be enacted giving effect to Article 18.4.2°, would the Oireachtas be obliged (in effect) to extend the University franchise to the graduates of all such higher education institutions? Or could the Oireachtas instead limit the franchise in some way and, if so, how could a distinction be properly made as between the various institutions of higher learning in the State? 10. Fourth, Article 18.4.2° also provides in its concluding sentence that: “A member or members of Seanad Éireann may be elected under this subsection by institutions grouped together or by a single institution.” 11. The natural meaning of this phrase suggests that the words refer to an individual educational institution such as the University of Dublin or the University of Limerick. Yet, as I have already pointed out, in the preceding sentence Article 18.4.2° had earlier defined - or, perhaps, it might be more correct to say, appeared to define — the University of Dublin and the National University of Ireland as one single institution for this purpose. It would seem, therefore, that in order to give full meaning to the word “institutions” in the final sentence of Article 18.4.2° it is necessary to ignore that special meaning which this provision gives to this very word in the preceding sentence and to give this word its ordinary meaning. 12. Fifth, Article 18.6 provides that the senators to be elected “by the Universities shall be elected on a franchise and in the manner to be provided by law.” (Emphasis supplied). Yet Article 18.4.2° recites that “Provision may be made by law for the election, on a franchise and in the manner to be provided by law” (emphasis supplied) by one or more of the following “institutions”, one of which is defined by reference by Article 18.4.2. i° as the National University of Ireland and the University of Dublin. There is here - regrettably, yet again - a clear contradiction between the mandatory obligation of Article 18.6 on the one hand and the permissive language of Article 18.4.2° on the other so far as these two Universities are concerned. One is also obliged to say that it is far from clear how Article 18.6 could or would function in the event that the National University of Ireland were ever to be dissolved. 13. Despite these significant drafting difficulties, it nonetheless falls to this Court to endeavour to give a sensible, workable interpretation of these provisions insofar as such is possible. The key provision here is Article 18.4.2° which states that provision “may be made by law for the election, on a franchise and in the manner to be provided by law” of the election of the University senators on an extended franchise. The critical question governing the outcome of this appeal is whether the reference to “Provision may be made by law” (“Féadfar foráil a dhéanamh le dlí”) is purely permissive? 14. So far as the principles of statutory interpretation are concerned, it is clear that phrases such as “hereby authorised”, “it shall be lawful” or “may be made” can be construed as being mandatory in nature, albeit that this is dependent on the precise context in which these phrases are used: see, e.g., the decisions of this Court in cases such as Dolan v. Neligan [1967] IR 247 at 275, per Walsh J.; Re Dunne’s Application [1968] IR 105 at 116-188, per Walsh J.; Duffy v. Dublin Corporation [1974] IR 33 at 42-44, per Henchy J. and Doyle v. Hearne [1987] IR 601 at 607, per Finlay C.J.; Bakht v. Medical Council [1990] 1 IR 515 at 523 per Griffin J. and the decision of the Court of Appeal in McK v. Minister for Justice and Equality [2018] IECA 110. 15. Thus, for example, in Dolan, this Court held that the provisions of certain customs legislation which provided that the Revenue Commissioners were “hereby authorised to return any money which have been overpaid…” were in truth mandatory in nature. While Walsh J. acknowledged that, prima facie, these words “import a discretion”, he nonetheless stressed that “the general context must be examined to see if there is anything in the subject matter to indicate that these words are intended to be imperative”: [1967] IR 247 at 274-275. 16. As it happens, Walsh J. considered that there was indeed something in the subject matter which negatived the suggestion that these words were simply permissive or discretionary, saying ([1967] IR 247 at 275) that: “It would be difficult to conceive that the legislature would have authorized a refusal to return moneys acknowledged or established to have been erroneously or wrongly demanded or exacted by the State, or by any organ of the State…Such an intention on the part of the legislature would have required to be expressed in the clearest and most unambiguous terms…” 17. In Duffy, by contrast, Henchy J. held ([1974] IR 33 at 41) that the provisions of the Dublin Improvement Act 1849 providing that “it shall be lawful” for the Corporation to hold a cattle market could not be construed as imposing a perpetual obligation to do so “regardless of the cost to the ratepayers or the absence of public demand or its unsuitability.” 18. In the same vein I concluded in my judgment in McK that the provisions of the Transfer of Sentenced Persons Act 1995 (which provided that the Minister “may” accept the transfer of a prisoner from another jurisdiction). should not be construed as imposing a mandatory obligation. As I put it: “…the entire context of the 1995 Act also strongly re-inforces the conclusion that the scheme was intended to be a discretionary one. The control of prisons and prison administration generally are clearly executive functions: see, e.g., Re Gallagher’s Application [1991] 1 I.R. 31. The decision to accept a transfer back is so clearly dependent on issues such as prison security and the availability of prison places that it would be surprising if the Oireachtas ever intended to oblige the Minister to accept every such request, even if the requirements of the 1995 Act were all otherwise satisfied.” 19. Doyle v. Hearne is a decision which falls on the other side of the line to that of Duffy and McK. In that case Finlay C.J. held that the words “and may adjourn the pronouncement of his judgment or order in the matter” pending the outcome of a case stated from the Circuit Court to this Court which were contained in s. 16 of the Courts of Justice Act 1947 in fact imposed a mandatory obligation on the Circuit Court judge to do so. As Finlay C.J. observed ([1987] IR 601 at 607): “Any other construction would create a total absurdity for it would be giving to a Circuit Court judge a power to consult the Supreme Court as to the determination of a question of law but leaving him free to decide the case in which it arose and, thus, presumably, the question of law as well prior to that determination.” 20. One can also see the same approach in Bakht. Here Griffin J. traced the history of the free movement of services in the (then) European Economic Community before concluding that the provisions of the Medical Practitioners Act 1978 enabling the Medical Council to make rules governing the full registration of medical practitioners were indeed mandatory in nature. This is yet another example of where the general statutory context led to the conclusion that ostensibly permissive language was indeed mandatory in character. 21. The key point to emerge from this set of cases is that there is generally no ex ante rule governing the question of whether and when the use of apparently permissive language may be interpreted as importing a legal obligation. While the choice of language is certainly important, the case-law plainly demonstrates that it is generally the particular context in which this language appears which is the decisive consideration in determining the question of whether the apparently permissive statutory language should in truth be regarded as mandatory. I propose to return later to this point. 22. The decision of this Court in The State (Sheehan) v. Government of Ireland [1987] IR 550 provides perhaps the closest analogue to the issues presented in this appeal, albeit that this decision was given in the context of statutory - rather than constitutional - interpretation. This case concerned the question as to whether the Government was under an obligation to make an order commencing the provisions of s. 60(1) of the Civil Liability Act 1961. That sub-section provides for the abolition of the traditional common law misfeasance rule. Section 60(7) provides, however, that the sub-section shall come into operation “on such day, not before 1st April 1967, as may be fixed therefor by order made by the Government.” 23. The complaint of the applicant in Sheehan was that the Government had failed to bring the sub-section into force. (As it happens, the sub-section remains un-commenced to this day.) A majority of this Court concluded, however, that the sub-section was purely permissive. As Henchy J. put it ([1987] IR 550 at 561): “I am satisfied that s. 60, sub-s. 7 is merely enabling. The uses of “shall” and “may”, both in the sub-section and in the section as a whole, point to the conclusion that the radical law-reform embodied in the section was intended not to come into effect before the 1st April 1967, and thereafter only on such day as may be fixed by an order made by the Government. Not, be it noted, on such date as shall be fixed by the Government. Limiting words such as “as soon as may be” or “as soon as convenient”, which are to be found in comparable statutory provisions, are markedly absent. If the true reading of s. 60, sub-s. 7 were to the effect that the Government were bound to bring the section into operation, it would of course be unconstitutional for the Government to achieve by their prolonged inactivity the virtual repeal of the section. In my opinion, however, s. 60, sub-s. 7 by vesting the power of bringing the section into operation in the Government rather than in a particular Minister, and the wording used, connoting an enabling rather than a mandatory power or discretion, would seem to point to the parliamentary recognition of the fact that the important law reform to be effected by the section was not to take effect unless and until the Government became satisfied that, in the light of factors such as the necessary deployment of financial and other resources, the postulated reform could be carried into effect. The discretion vested in the Government to bring the section into operation on a date after the 1st April 1967, was not limited in any way, as to time or otherwise.” 24. It might be thought at first blush that this decision provides support for the State parties’ contention in the present case that Article 18.4.2° is purely permissive and that it simply enables - but in no sense compels - the Oireachtas to re-organise the configuration of the University seats should it consider it appropriate to do so. 25. There is admittedly force to this submission. There are, after all, many other constitutional provisions which, by a parity of reasoning, are equally permissive in character. Thus, for example, Article 15.2.2° is plainly permissive in character, stating as it does that the Oireachtas “ may however” provide by law “for the creation or recognition of subordinate legislatures and for the powers and functions of such legislatures.” (Emphasis supplied). This sub-section was, of course, obviously intended to allow for the recognition of a separate parliament for Northern Ireland in the event of the ultimate political reunification of the island. No one would, I think, suggest that the Oireachtas could presently be mandated by judicial proceedings to take some positive step in that direction in the event that it failed to provide for such subordinate legislatures. 26. One might also observe that some constitutional provisions are self-executing and do not require statutory vesture. This was held to be true of Article 41.3.2° (dealing with divorce) following the repeal of the previous constitutional ban by the 15th Amendment of the Constitution Act 1996. In RC v. CC (divorce) [1997] 1 IR 334 Barron J. held that these provisions were self-executing, so that the High Court had jurisdiction to grant a divorce even though the provisions of the Family Law (Divorce) Act 1996 (which sought to give effect to the (then) new constitutional amendment) were not then yet in force. 27. The central question in this appeal, therefore, is whether the changes effected to Article 18 by the 7th Amendment fall into this former category of purely permissive provisions. Yet apart from the fact that Sheehan concerned issues of statutory interpretation, there are, I think, limits to this approach in the case of a constitutional amendment. Even if one accepts that Henchy J.’s analysis of s. 60 of the 1961 Act in Sheehan was correct - and, speaking for myself, I find there is much to be said for the contrary position of Costello J. in the High Court, and the dissent of McCarthy J. in this Court in that case - this analysis cannot realistically be applied without at least some important qualifications in the case of a constitutional amendment such as in respect of the changes effected to Article 18 by the 7th Amendment. 28. Here one may note that, of course, Article 46.1 provides that any provision of the Constitution can be amended “whether by way of amendment, variation, or repeal”. It is also true that as Barrington J. said in Finn v. Attorney General. [1983] IR 154, at 163-164, this means that the People “intended to give themselves full power to amend any provision of the Constitution and that this power includes a power to clarify or make more explicit anything already in the Constitution.” 29. This means in turn that the People have the right to amend the Constitution in order, for example, to include a provision by way of constitutional amendment that was either never brought into force or which was otherwise ineffective. Should they think well of it the People could, of course, accordingly vote in favour of a new constitutional provision which was either ineffective or which lay dormant in some way or which, on its true contextual interpretation, was entirely dependent in some way upon the decision of the Government and the Oireachtas to trigger the operation of the new clause at some uncertain stage in the future. That, in essence, is the argument advanced by the State respondents in this appeal so far as the construction of Article 18.4.2° is concerned. 30. Yet while the People certainly could elect to do this, it seems to me unlikely that they would in fact do so. By deliberate design the process of constitutional amendment is not intended to be a straightforward one. Any Bill to amend the Constitution must commence in Dáil Éireann (Article 46.2) and a Bill containing a proposal or proposals “for the amendment of the Constitution shall not contain any other proposal” (Article 46.4). The Bill must then pass both Houses and be approved by a majority of voters (Article 47.1). 31. In these circumstances there must at least be a presumption - admittedly a rebuttable one - that every constitutional amendment was intended to be effective and to produce legal or political or symbolic consequences. The referendum provisions give practical effect to the statement of popular sovereignty in Article 6, namely, that it is right of the People “in final appeal” to decide all questions of national policy. In this respect popular sovereignty is the ultimate Grundnorm of the Irish constitutional system. The reference to “in final appeal” in Article 6 suggests that the Constitution does not readily contemplate casual or unnecessary constitutional change and, conversely, that, as I have just stated, it is to be assumed that any amendment which has been passed via the referendum process was in fact intended to be effective and to have legal, political or symbolic consequences. It is for that very reason that the reasoning in Sheehan has little or no application in the context of a constitutional amendment of this kind. 32. An example of all of this is supplied by the new Article 42A effected by the 31st Amendment of the Constitution Act 2015. In Re JJ [2021] IESC 1, this Court noted the similarities in wording between the “old’ Article 42.5 and the “new” Article 42A (which was inserted by the 31st Amendment of the Constitution Act 2015) such that the differences between the two provisions were quite subtle. In their joint judgment O’Donnell, Dunne, O’Malley and Baker JJ. nonetheless rejected the argument that these changes were not real or effective or did not otherwise produce substantive effects. The leitmotif underpinning JJ is that the People must nonetheless be taken to have intended to effect enhanced protection for the rights of children by enacting the new Article 42A, the similarity with the wording of the pre-existing Article 42.5 notwithstanding. 33. It seems to me that a similar approach ought to be adopted with regard to the changes effected by the 7th Amendment. As Murray J. has demonstrated in his judgment, all the official documentation attending the passage and enactment of the Amendment conveyed the unambiguous message to the electorate that the University franchise would in fact be extended to other institutions of higher education. After all, the very title of the 7th Amendment when still in Bill form at the time of the referendum vote in July 1979 - the Seventh Amendment of the Constitution (Election of Members of Seanad Éireann by Institutions of Higher Education) Bill 1979 - clearly conveyed the impression to the electorate that the franchise was to be extended to other institutions of higher education. 34. This is underscored by the wording of the actual voting card provided to each voter at the time of the referendum in accordance with s. 1 of the Referendum (Amendment) Act 1979. This section provided that the Appendix to the voting card would contain the following information: “The Seventh Amendment of the Constitution (Election of Members of Seanad Éireann by Institutions of Higher Education) Bill, 1979, proposes the election by universities and other institutions of higher education specified by law of such number of members of Seanad Éireann, not exceeding 6, as may be specified by law. Those so elected would be in substitution for an equal number of the members elected at present (3 each) by the National University of Ireland and the University of Dublin. The Bill also proposes that nothing in Article 18 of the Constitution shall prohibit the dissolution by law of those Universities.” 35. I stress, of course, that this Court could only properly look at official documentation of this kind and then only for the purposes of ascertaining the true context of Article 18 (as so amended). While, given the primacy of the text, the Constitution can generally only properly be interpreted by reference to the objective meaning of its actual words in their proper context, one can, I think, nevertheless look at official documentation to ascertain that very context where the wording of a subsequent constitutional amendment is - as here - strikingly unclear. At the same time the subjective beliefs of, or statements by, the members of the Oireachtas or other participants in that particular constitutional debate as to what the effect of any particular amendment would be are quite irrelevant for this purpose since to have regard to such subjective statements or predictions regarding the effect of the proposed change would be at odds with the primacy of the constitutional text itself: see, e.g., the judgments of this Court in Crilly v. T & J Farrington Ltd. [2001] 3 IR 251. 36. In the present case I find myself coerced to finding that, in view of the profound uncertainties attending the drafting of the wording of the 7th Amendment and, by extension, the proper interpretation of Article 18.4.2°, the context of that Amendment as disclosed by this official documentation really admits of no conclusion other than that the People did indeed intend that the franchise for the University seats should be extended beyond the pre-existing University of Dublin and National University of Ireland constituencies. The suggestion that the 7th Amendment did no more than permit the Oireachtas to extend or otherwise vary the scope of the University franchise should it elect to do so at some future and ill-defined point seems to me to be an implausible one. 37. Even allowing for the fact that the decisions in cases such as Dolan and Doyle concerned the interpretation of statutes — rather than the Constitution - the reasoning in the cases can nonetheless be adapted for the purposes of this somewhat different task of constitutional interpretation. The underlying principle, after all, is that the context of the measure will often determine whether apparently permissive language used in a legal instrument should in truth be regarded as having imposed a mandatory obligation. So it is here. Just as this Court said that in Dolan that it was unlikely that the parliament enacting the relevant customs legislation ever intended that the Revenue Commissioners were to be given a discretion to retain monies which had been unlawfully collected or improperly exacted, one may equally say that there is little in either the language of Article 18.4.2° or the actual context of that amendment which justifies the conclusion that when the enacting the provisions of the 7th Amendment in 1979 the People did in fact intend to give the Oireachtas a purely permissive power which could lie dormant indefinitely. 38. It follows that Article 18.4.2° must accordingly be interpreted as requiring that the Oireachtas must revise the Seanad University constituencies by enacting legislation for this purpose within a reasonable period from the date of the passage of the 7th Amendment. To that extent, therefore, these provisions must be interpreted as imposing a mandatory obligation on the Oireachtas to extend the University franchise to at least certain other institutions of higher education in the State. Any other conclusion would effectively mean that the sovereign will of the People as democratically expressed through the referendum process in the manner envisaged by Article 46 and Article 47 - perhaps the most fundamental and core feature of our constitutional identity as a State - would be thwarted by an obvious failure of legislative inertia. 39. All of this is to say, adapting the words of Walsh J. in Dolan, the general context of Article 18.4.2° plainly indicates that these words are intended to be imperative. While one may acknowledge that apparently permissive language was used, this was fundamentally to enable the Oireachtas to reconfigure the University seats and to give it a reasonable time to effect these far-reaching changes. On the fundamental issue, however, Article 18.4.2° is not permissive in that on its proper contextual interpretation it required the Oireachtas to act within a reasonable time of the enactment of the constitutional amendment in order to revise and extend the franchise beyond the existing University constituencies and this it has not done. Part III - The Article 40.1 issue 40. So far as the potential application of the equality provisions in Article 40.1 are concerned, it may be observed that the requirement of equality and equal treatment in the representative democratic process - such as elections for Dáil Eireann and referenda - does not apply to elections for Seanad Éireann. It is not a directly representative body in the way that Dáil Éireann is. To that extent, Article 40.1 cannot be directly invoked so as, for example, to require an approximate equality in terms of the size of Seanad constituencies in the same manner as would be the case in respect of Dáil constituencies as decisions such as O’Donovan and Murphy v. Minister for Environment [2007] IEHC 185, [2008] 3 IR 438 illustrate. There, is, in any event, no equivalent of Article 16.2.3° (which requires that the ratio between the Deputies elected in respect of the Dáil per constituency and the population of such constituency “shall, so far as it is practicable, be the same throughout the country) applicable to the Seanad. 41. Accordingly, while Article 18.4.2° clearly vests the Oireachtas with a wide discretion regarding the re-organisation of the University Seanad seats, this discretion is not, perhaps, entirely without its limits. There might possibly be instances where the differing treatment of various Universities and institutions of higher education was manifestly indefensible on basic rationality grounds such that the application of Article 40.1 could not be excluded. 42. As matters stand, these issues remain entirely hypothetical and they were not, in any event, the subject of any argument on this appeal. I would accordingly leave over these questions unless and until they arose. Part IV - The consequences of the conclusion that Article 18.4.2° is not purely permissive 43. It remains to consider the consequences of this conclusion. The first and most obvious consequence is that the key provisions of the 1937 Act (namely, ss. 6 and 7) have been rendered unconstitutional by virtue of the enactment of Article 18.4.2° and the subsequent failure by the Oireachtas to give effect to this provision in the manner required by its terms. As I have already indicated, on its proper contextual interpretation the effect of Article 18.4.2° is that the Oireachtas was obliged to revise the University constituencies within a reasonable time, a time which I would measure in terms of several years. After all, the extension of the franchise in the manner contemplated by the amendment was likely to present a series of logistical and other practical difficulties which went beyond the immediate difficulties of drafting new legislation. It is equally clear, however, that the reasonable period in question has long since expired. 44. This Court has already had occasion to address the question of the obligations of the Oireachtas in circumstances where a constitutional obligation has not been complied with by reference to the time period stipulated by the Constitution itself. 45. This issue of non-compliance with constitutionally prescribed time limits was first considered by this Court in Re Article 26 and the Electoral (Amendment) Bill 1961 [1961] IR 169. Article 16.2.4° requires that the constituencies be revised every twelve years. The Oireachtas had sought to comply with this requirement when it revised the constituencies with the Electoral (Amendment) Act 1959. That Act had been signed by the President on 26th November 1959 which was just within the 12-year period since the passage of the Electoral (Amendment) Act 1947 which had itself been enacted on the 27th November 1947. 46. The 1959 Act was subsequently found to be unconstitutional in O’Donovan v. Attorney General [1961] IR 114. The Oireachtas then sought to pass new legislation revising the constituencies in the wake of this decision. The President then referred this new Bill - the Electoral (Amendment) Bill 1961 - to this Court under Article 26. By the date this matter came before this Court in July 1961 it was clear that the twelve-year limit had been breached by some two years and eight months. This Court held that this did not in itself now mean that the new 1961 Bill was unconstitutional. Rather, as there was a “satisfactory explanation” for the delay, the obligation was now to carry out the constitutional obligation “as soon as possible”: see Re Article 26 and Electoral (Amendment) Bill 1961 [1961] IR 169 at 180, per Maguire CJ. 47. This issue was also in view in this Court’s recent judgment in Glann Mór Céibh Teoranta v. An tAire Tithíochta, Pleanáil agus Rialtas Áitiuil [2022] IESC 40. Here, one of the questions was whether the State had failed in its constitutional obligation pursuant to Article 25.4.2° to effect a translation into Irish of a particular Act of the Oireachtas. Holding that in the instant case there had been such a failure, I said [at 65]: “While the context of the Electoral (Amendment) Bill reference is very different to the present one, the same underlying principle regarding the issue of compliance with constitutional time limits holds true. It is clear that the State is under a constitutional duty to effect a translation of all Acts of the Oireachtas as soon as possible but the extent to which such non-compliance is excusable will naturally depend on all the circumstances.” 48. For all the reasons I have already expressed in this judgment I consider that the Oireachtas has failed in its constitutional obligation to revise the Seanad Universities constituencies in the manner required by Article 18.4.2°. In the light of this Court’s decisions in both the Electoral (Amendment) Bill 1961 and Glann Mór it is clear that this constitutional obligation still subsists. It must now be complied within a reasonable time (which I would measure in terms of years rather than months) from the date of the delivery of this judgment. I propose to return presently to this point. Part V - The consequences for the past of this conclusion 49. The question of how to address the consequences of a finding of the unconstitutionality of a law is one with which the US Supreme Court has wrestled for the best part of 220 years. It was not for nothing that in 1940 Hughes C.J. acknowledged that this issue was among the most “are among the most difficult of those which have engaged the attention of courts, state and federal, and it is manifest from numerous decisions that an all-inclusive statement of a principle of absolute retroactive invalidity cannot be justified”: see Chicot Company Drainage District v. Baxter State Bank 308 US 371 at 374 (1940). 50. The Chief Justice had earlier stated in that case that: “The actual existence of a statute, prior to such a determination, is an operative fact, and may have consequences which cannot justly be ignored. The past cannot always be erased by a new judicial declaration. The effect of the subsequent ruling as to invalidity may have to be considered in various aspects -- with respect to particular relations, individual and corporate, and particular conduct, private and official. Questions of rights claimed to have become vested, of status, of prior determinations deemed to have finality and acted upon accordingly, of public policy in the light of the nature both of the statute and of its previous application, demand examination.” 51. The experience in this State since 1937 has been no different, as courts have, on occasion, struggled to articulate fully coherent principles in their efforts to contain the consequences of a finding of unconstitutionality. Experience has shown, however, that according fully retroactive status to a finding of unconstitutionality would simply result in the triumph of abstract logic over the requirements of justice, often with unpredictable, chaotic and indefensible consequences. 52. As Denham C.J. explained in The People (Director of Public Prosecutions) v. Kavanagh [2012] IECCA 65: “it would be a fallacy to treat the issue of the consequences of a finding of unconstitutionality as if it were some form of abstract quasi-mathematical syllogism, with the courts looking on helplessly as the retroactive application of a finding of unconstitutionality worked inexorably to bring about catastrophic consequences for the legal system and ordered political society. While the first duty of the courts is to secure legal redress for those whose rights have been infringed by unconstitutional action, this duty is, as Article 40.3.1 itself recognises, tempered by considerations of feasibility and practicability. Any other conclusion would mean that the “true social order” envisaged by the Preamble to the Constitution could not be attained.” 53. There is, accordingly, a long history in this jurisdiction of cases where the courts have indicated that they, of necessity, must enjoy the power to restrict the fully retroactive operation of such a finding. Thus, for example, in McMahon v. Attorney General [1972] IR 69 Ó Dálaigh C.J. hinted (at 111-112) that the validity of past elections was beyond challenge, even though a key part of the Electoral Act 1923 dealing with secret ballots had just been found to be unconstitutional. In The State (Byrne) v. Frawley [1978] IR 326 this Court held that the applicant was barred by acquiescence from challenging the validity of a jury verdict even though the previous jury system provided for by the Juries (Amendment) Act 1927 excluding women and non-ratepayers had been found to be unconstitutional in de Búrca v. Attorney General [1976] IR 38. 54. This trend reached its apotheosis in two leading judgments of this Court, Murphy v. Attorney General [1982] IR 241 and A v. Governor of Mountjoy Prison [2006] IESC 45, [2006] 4 IR 88. The decision in A really follows on from Byrne, in that this Court held that an accused who pleaded guilty to an offence without raising a constitutional objection could not re-open that conviction even when the offence in question was subsequently found to be unconstitutional in separate and different proceedings. 55. In Murphy this Court first held in January 1980 that the provisions of the Income Tax Act 1967 which had provided for the aggregation of the income of a married couples were unconstitutional. In a subsequent judgment delivered in April 1980 Henchy J. first demonstrated that these provisions of the 1967 Act was, in theory, at least, invalid ab intio. Noting the language of Article 15.4.1° of the Constitution, he pointed out that ([1982] IR 241 at 310): “…the constitutional disposition of the powers of the State in this respect falls into line with the general principle that, when a constitutional statute gives a specifically confined power off legislation to a legislature, laws found to have been enacted in excess of that delegation are ultra vires and therefore void ab initio. This is a principle which is inherent in the nature of such limited powers, but it is unequivocally spelled out in some constitutional and constitutional statutes.” 56. It followed, therefore, that the relevant sections of the 1967 Act were invalid as of the date of their enactment. I should pause at this point what while this is the general rule it does not necessarily apply to the particular finding of unconstitutionality in the present case given that it arises as a consequence of a constitutional amendment., I propose presently to expand upon this point in a little greater detail. 57. Returning to the judgment in Murphy, Henchy J. stressed that this, however, did not mean that everything that was done by virtue of these section was invalid. Drawing on a consistent theme of his as expressed in earlier judgments such as Byrne, Henchy J. pointed out that the law had always recognised that there were what he described as ‘transcendent considerations’ which prevented the reopening of past events. He also observed that courts of other jurisdictions with comparable constitutional regimes - ranging from the US Supreme Court to the Court of Justice - had rejected such full and automatic retroactivity. The payment of tax had, in any event, always been a special case, since the State had altered its position by quickly spending the money of taxpayers on the implicit assurance that the unchallenged deduction of such monies was lawful and valid. He continued by saying ([1982] IR 241 at 314-315): “But it is not a universal rule that what has been done in pursuance of a law which has been held to be invalid for constitutional or other reasons will necessarily give a good cause of action: see, for example, the decision of this Court in The State (Byrne) v. Frawley. While it is central to the due administration of justice in an ordered society that one of the primary concerns of the courts should be to see that prejudice suffered at the hands of those who act without legal justification, where legal justification. Is required, shall not stand beyond the reach of corrective legal proceedings, the law has to recognise that there may be transcendent considerations which make such a course of undesirable, impracticable or impossible. Over the centuries the law has come to recognise, in one degree or another, that factors such as prescription (negative or positive), waiver, estoppel, laches, a statute of limitations, res judicata, or other matters (most of which may be grouped under the heading of public policy) may debar a person from obtaining a redress in the Courts for injury, pecuniary or otherwise, which would be justiciable and redressable if such considerations had not intervened…. For a variety of reasons, the law recognises that in certain circumstances, no matter how unfounded in law certain conduct may have been, no matter how unwarranted its operation in a particular case, what has happened and cannot, or should not, be undone. The irreversible progressions and byproducts of time, the compulsion of public order, the aversion of the law from giving a hearing to those who have slept on their rights, the quality of legality - even irreversibility - that tends to attach to what has become inveterate or has been widely accepted or acted upon, the recognition that even in the short term the accomplished fact may sometimes acquire an inviolable sacredness, these and other factors may convert what has been done under an unconstitutional, or otherwise void, law into an acceptable part of the corpus juris.” 58. It followed, therefore, that the plaintiffs themselves were allowed only limited recovery of overpaid taxes, backdated only to the first full year in respect of which they had issued proceedings. The logical corollary of that finding was that the vast majority of other taxpayers — who had not issued such proceedings challenging the constitutionality of the law — were not entitled to any recovery in respect of these unconstitutionally collected taxes. 59. What principles, therefore, can be drawn from these decisions? No one has ever come forward to challenge the validity of past elections in respect of the University Senator seats, even though since the enactment of the 7th Amendment in July 1979 Seanad elections have taken place pursuant to the provisions of the 1937 Act in 1981, 1982, 1982-1983, 1987,1989, 1992-1993, 1997, 2002, 2007, 2011, 2016 and 2020. There have, indeed, been occasional bye-elections in respect of the University seats, the most recent of which was held in respect of a University of Dublin seat in February and March 2022. 60. Adopting the general language of Henchy J. in Murphy it may therefore be said that such universal acquiescence to such an “inveterate or widely accepted practice” means that the validity of these elections has thereby acquired an “inviolable sacredness.” To this one might add that as s. 26 of the 1937 Act applies by reference the legislative provisions in force regarding elections petitions to the results of the elections in respect of University Senators, one cannot here overlook the fact that Rule 3(2) of the Third Schedule of the Electoral Act 1992 provides that any challenge to the validity of an election must be presented by way of election petition within 28 days of the result of an election. No such challenge has ever been brought to the result of an election for University Senators. 61. The consequence of all of this is that the validity of the results of all past election since the enactment of the 7th Amendment in 1979 in respect of the University Seanad seats must be regarded as standing beyond legal challenge, this finding of unconstitutionality in respect of these key provisions of the 1937 Act notwithstanding. Part VI: Whether the finding of unconstitutionality in respect of the 1937 Act ought to be suspended? 57. So much for the past. The question arises as to what should be done in respect of the immediate future. It might first be noted that other courts enjoying a comparable constitutional jurisdiction to ours have also suspended - or effectively suspended - declarations of unconstitutionality when faced with roughly similar problems. A few representative examples may be mentioned at this point. 58. In 1974-1975 the Australian High Court was required to deal with a series of cases arising from the exercise by the Australian Governor-General of the so-called double-dissolution power conferred by s. 57 of the Australian Constitution whereby both Houses of Parliament were simultaneously dissolved, and a general election held. This power was designed to resolve an actual deadlock between the two Houses and in earlier litigation the Australian High Court held that the conditions for the exercise of this special constitutional power had not been satisfied in that instance ( Cormack v. Cope (1974) 131 CLR 432). 59. In the subsequent decision in Victoria v. Commonwealth (1975) 134 CLR 81 the High Court held that this non-compliance with constitutional requirements affected the validity of certain legislative enactments passed by the old Parliament. What is of more direct relevance for us is the observation of Barwick C.J. (at 178) that this finding could not affect the validity of the elections of the new Parliament even though the old Parliament had, on this reasoning, been invalidly dissolved: “…. once the Governor-General has in fact dissolved both Chambers, whether or not he is justified in doing so in terms of s. 57, the existing Parliament will have been brought effectively to an end and the new Parliament which results from the issue of writs and the holding of an election following such dissolution will be quite unaffected by whatever may or may not have preceded that dissolution.” 60. In the same vein - albeit, perhaps, even more directly on point — is the judgment of McLachlin C.J. in Dixon v. British Columbia (1989) 59 DLR 4th 247. Here the court had held that the uneven and unequal distribution of seats in the Province of British Columbia violated the equality guarantees contained in s. 52 of Canadian Charter. (This judgment is roughly the equivalent of our decision in O’Donovan). Dealing, then, with the question of a remedy, McLachlin C.J. stated (at 303): “The effect of a declaration that a law is inconsistent with the Charter is to render it of no force and effect under s. 52 of the Charter. In most cases where a particular provision falls under s. 52, the result is to restore the law in question to the status of conformity with the Charter. The effect in this case is arguably the reverse. If the provisions prescribing electoral districts in British Columbia are set aside, the electoral districts vanish. Should an election be required before they are restored, it would be impossible to conduct it. The result would be disenfranchisement of the citizens of the Province.” 61. The Chief Justice went on to elaborate upon the Canadian jurisprudence regarding suspended declarations of this kind before she concluded that: “…. this Court cannot escape its constitutional obligation to review the validity of s. 19 and Schedule 1 of the Constitution Act and must declare those provisions to be contrary to the Charter of Rights and Freedoms. Pending submissions on what time period may reasonably be required to remedy the legislation and the expiry of that period, the legislation will stay provisionally in place to avoid the constitutional crisis which would occur should a precipitate election be called.” 62. One might finally draw attention to the practice of the German Constitutional Court in cases of this kind. In the words of Kommers and Miller, The Constitutional Jurisprudence of the Federal Republic of Germany (Duke, 2012) (at 36): “This strategy of declaring a law or practice unconstitutional but not void is designed to prevent the greater hardship or inconvenience that would flow from the complete voidance of a statute. How long and under what conditions an unconstitutional but still-viable law can remain in force is a matter which the Court reserves itself to decide. The Court usually sets a deadline for corrective legislative action and occasionally directs the Bundestag to adopt a specific solution. More often the Court lays down the general guidelines within which the legislature is required to act.” 63. The approach of this Court has really been no different. We have already indicated that the strict logic of the Murphy cannot, in any event, be applied in what is sometimes called instances of “creeping unconstitutionality”, namely, cases where the law was not unconstitutional upon its enactment but becomes such over time. This is true here. There was no suggestion that the 1937 Act was unconstitutional upon its enactment. It only became unconstitutional at some point after the enactment of the 7th Amendment in 1979 insofar as it failed to provide for a revision of the University Senator constituencies. 64. This issue has, in any event, been pre-figured in some recent judgments of this Court. In PC v. Minister for Social Welfare [2018] IESC 57. Here O’Donnell J. referred with approval to the Canadian practice of suspending declarations of unconstitutionality and then added [at para. 15]: “Even more extreme circumstances could be envisaged and have arisen in other jurisdictions. One example is where the flaw relates to the election, or legal constitution, of a legislature, and where immediate invalidity might be simply incapable of remedy, or worse, might remove the only mechanism for remedying the flaw. But the fact that the solution proposed, of suspending the declaration of unconstitutionality in such circumstances in other jurisdictions seems reasonable and sensible, even necessary, does not mean it is constitutionally permissible under the Irish Constitution. Indeed, it is argued here that what is sensible and reasonable must give way to principle, no matter how inconvenient the result. It would not be fair to dismiss this argument as merely a narrow and inflexible absolutism. Part of the appeal of constitutional guarantees of rights is the belief that they contain enduring truths which cannot and should not be compromised. This is indeed a familiar argument for the rule of law: justice must be done whatever the consequences. But that in turn only leads to a deeper question as to what the doing of justice entails. This may be a particularly troublesome question where the issue involves not just the resolution of litigation between the parties but where the outcome may directly affect many others not before the court.” 65. In effect, therefore, as O’Donnell J. noted, while it is important that constitutional rights are vindicated, this should not - where possible — be done be the expense of causing uncontrolled consequences for society and the institutions of State which support it. It is in this context a case of fiat justitia, dum maneat caelum. 66. In her judgment in PP v. Judges of the Dublin Circuit Court [2019] IESC 26, [2020] 1 IR 123 at 200-201 O’Malley J. alluded first to the creeping unconstitutionality cases (including those cases where - as here - a statute “become[s] unconstitutional as a result of an amendment to the Constitution by the People”). She then added: “As the year 1937 recedes further into history, and the Constitution is subject both to amendment from time to time by the People and to interpretation by the judiciary in the light of developing jurisprudence, it seems clear that a strict application of Henchy J.'s theory [in Murphy] will encounter increasingly formidable objections. The courts could legitimately find in one era that a particular statute was consistent with the Constitution, while coming to the opposite conclusion at a different time. A statute that is found to be unobjectionable in one particular case, other than one that has been the subject of a reference to this Court under the provisions of Article 26, does not thereby acquire an immunity from future challenge. It seems to me, therefore, that the theory may need to be refined at least to the extent of distinguishing between pre-1937 legislation that was not consistent with the Constitution as enacted in 1937, and legislation that is inconsistent with the Constitution as it stands at the time the Court considers the matter. I would respectfully adopt Walsh J.'s analysis [in McGee v. Attorney General [1974] IR 284] and suggest that in the latter category, the legislation ceases to have legal force only when the finding of inconsistency is made.” 67. All of these factors combine in the present case such that the case for an immediate suspension of the finding of unconstitutionality is well-nigh overwhelming. The democratic character of the State as described in Article 5 is an inviolate feature of the Constitution’s identity. Yet this democratic character would be compromised if elections (whether for the Dáil or Seanad) could not safely be held or they were otherwise conducted under the shadow of unconstitutionality with the attendant risk of invalidity. 68. It would not be practical or realistic to make that declaration immediately effective since it would effectively render our democratic system positively unworkable, as it would be all but impossible in such circumstances to conduct a general election for Seanad Eireann in the manner required by Article 18.8. Since Seanad general elections must take place “not later than ninety days after a dissolution of Dáil Éireann”, such a constitutional impasse would also , inter alia, frustrate the right of the Taoiseach to seek a dissolution of Dáil Éireann in accordance with Article 13.2.1° of the Constitution. The Oireachtas, moreover, could not be expected quickly to complete this task given the practical and other difficulties and complexities attending the extension of the University franchise, not least given the difficulties associated with the assembly of entirely new electoral registers associated with other institutes of higher education. 69. One must also have regard to the fact that the unconstitutionality in the present case arose by reason of the enactment of a constitutional amendment enacted by the People. To that extent, therefore, the unconstitutionality crystallised only with this pronouncement of this judgment: see PP [2020] 1 IR 123 at 201, per O’Malley J. 70. Echoing, therefore, the approach of the Australian High Court in Victoria, the practice of the German Constitutional Court with its “unconstitutional but not void” jurisprudence; that of the Canadian Supreme Court in Dixon, and the previous decisions of this Court in cases such as PC and PP, it is accordingly appropriate to suspend this declaration of unconstitutionality in the first instance until 31 July 2023. This suspension of the declaration is in the first instance to enable the parties to address this Court on the length of the final suspension of the declaration of unconstitutionality which the Court might ultimately make having heard the parties on this issue. 71. In these circumstances it is sufficient for this Court simply to declare that the 1937 Act is unconstitutional while suspending that declaration of unconstitutionality in the first instance pending further order until the 31st of July 2023. The effect of this suspension is that the 1937 Act remains effective until that date. In particular and for the avoidance of any possible doubt, I would like to make it clear (i) that all the existing University Senate elections held since 1979 are now legally impregnable from challenge and (ii) that any University Seanad elections (whether general elections or bye-elections) may validly be conducted under the provisions of the 1937 Act up to that date. Part VI - Overall conclusions 72. In summary, therefore, I am of the view that the appeal should be allowed and would principally conclude as follows: 73. First, on its proper, contextual construction, Article 18.4.2° imposes a constitutional obligation to revise and to extend the University Seanad franchise within a reasonable time from the date of the enactment of the 7th Amendment of the Constitution Act in 1979. That reasonable time period has long since expired with the result that the key sections of the 1937 Act have been rendered unconstitutional. 74. Second, in the light of the decision of this Court, the Oireachtas is under an obligation to revise these Universities constituencies in accordance with law. That obligation must now be discharged within a reasonable period. 75. Third, given the need to safeguard the proper operation of the democratic process, together with a variety of practical and logistical reasons, I would suspend that declaration of unconstitutionality in the first instance until 31st July 2023 pending a further order from this Court following a further hearing on this point. This means that the 1937 Act remains effective and valid until that date and, in particular, a general election and/or a bye-election may validly be held under the terms of that 1937 Act up to and including that date.