English VersionIrish Version
Search for Click to Search
Advanced Search
Printable Version
All SectionsPractice DirectionsCourt Rules Terms & Sittings
Legal Diary Offices & Maps Judgments & Determinations

Murphy & anor -v- Minister for the Environment Heritage & Local Government & ors
Neutral Citation:
[2007] IEHC 185
High Court Record Number:
2007 2819 P & 2007 3475P
Date of Delivery:
High Court
Composition of Court:
Clarke J.
Judgment by:
Clarke J.

Neutral Citation Number: [2007] IEHC 185

[2007 No. 2819 P]






[2007 No. 3475P]





JUDGMENT of Mr. Justice Clarke delivered 7th June, 2007.

1. Introduction
1.1 There can be little doubt but that the results of the census conducted in April 2006 show very substantial changes not only in the overall population of the State but also in its distribution. In particular it is clear that the population of the constituencies established by law for the purposes of conducting elections to Dáil Eireann have altered in a way which leads to a very significant disparity existing between the proportionality of the population to the numbers of deputies to be elected, at least in certain cases. While it will be necessary to examine the figures in more detail, the scale of the disparity can be gleaned from the fact that the range of population per seat from the highest to lowest is 8,146 and the biggest deviation from the national average is 5,421. Both of those figures have to be seen against the fact that, at a national level, the ratio of population to seats is of the order of 25,000 to 1. In addition five constituencies are more than ten per cent away from the national average with two constituencies being more than ten per cent below that average and three constituencies being that far above average.

1.2 The disparities are most glaringly illustrated by the fact that in a number of instances it is possible to point to pairs of constituencies where more deputies are to be elected from the constituency with a lower population.

1.3 It is as against that general background that both of these proceedings are commenced. The first proceedings in time were those commenced on 5th April, 2007 by Catherine Murphy and Finian McGrath (“the deputies”) who were both members of the 29th Dáil (Deputy McGrath has since these proceedings were heard been re-elected to be a member of the 30th Dáil but Deputy Murphy failed to secure re-election). In the second proceedings the plaintiff (“Mr. Molloy”) is a constituent of the Dublin West constituency which is the most underrepresented constituency by reference to the 2006 census. Its population to seat ratio is 21% above the national average.

2. Procedural History
2.1 As of the time of commencement of the proceedings by the deputies it was clear that a general election was required to be held within a short number of months. With that in mind management of the case was engaged in to see if it would be possible to have the proceedings concluded prior to the anticipated dissolution of the Dáil. That did not prove to be possible. It should be emphasised that it was never part of the deputies’ case that this court had any power or jurisdiction to prevent the general election taking place nor was it part of the deputies’ case that the results of a general election conducted on the existing constituencies could or should be overturned. Nonetheless the original urgency of the case was borne of the fact that the deputies claimed to be entitled to orders, some of which, if granted, might have required an immediate or early redrawing of the constituencies. In those circumstances, while unlikely, it was possible that an order could have been made which would have had an effect prior to the dissolution of the Dáil.

2.2 Notwithstanding the dissolution of the 29th Dáil it remained a matter of some urgency to have the issues raised resolved not least because of the possibility that, for whatever reason, there might be a further dissolution of the new 30th Dáil resulting in an early general election. If aspects of the case made by the deputies are found to be correct then there is the potential that the court might be required to make orders which could have an effect on the constituencies in which any such fresh general election might be fought. For that reason it was arranged that the full proceedings between the deputies and the defendants (“the State”) would be expedited.

2.3 In the meantime Mr. Molloy issued his proceedings which are, subject to one point, identical to those brought by the deputies. It proved possible to greatly expedite Mr. Molloy’s proceedings so that they could be heard at the same time as the proceedings brought by the deputies. This was facilitated by the fact that there was little or no difference in practice between the bulk of the issues which arose in the two sets of proceedings. Mr. Molloy raised an additional point. In substance it was argued on his behalf that, on a proper construction of the Electoral Acts, the first named defendant in each of the proceedings (“the Minister”) had a power to remedy the extreme situation that was found to exist in respect of Dublin West by the simple expedient of adding a seat to the number of seats to be contested in that constituency.

2.4 Against that procedural background it is next necessary to turn to the issues which arose between the parties.

3. The Issues
3.1 I have emphasised that it was never part of the case brought by either the deputies or on behalf of Mr. Molloy that the election which has been completed since the hearing of the case should be halted or that the result of that election, as a matter of constitutional law, could be questioned in anyway. This is accepted to be the case notwithstanding that, it is said, the constituencies by reference to which that election was conducted were inconsistent with the provisions of the Constitution.

3.2 However the deputies and Mr. Molloy argue that the results of the census conducted in April 2006 reveal that the ratio between the number of persons to be elected from, and the population of, constituencies is not, so far as is practicable, the same throughout the country. In consequence it is said that the Electoral (Amendment) Act, 2005 (“the 2005 Act”) (which provided for those constituencies) is inconsistent with the Constitution and should be declared as such.

3.3 It was not seriously contested on behalf of the State that a proper analysis of the figures to be derived from the 2006 census supports the view that the constituencies were no longer, as of the time of hearing before me, drawn in such a way that it could be said that the population to seat ratio was sufficiently proportionate throughout the country so as to remain within constitutional bounds. The proportionality or otherwise of the ratio between population and seats throughout the country was not, therefore, of itself, a real issue in the case although it is, obviously, against the backdrop of the lack of proportionality which I have noted, that the issues between the parties do arise.

3.4 Rather a number of separate issues arose. They are as follows:-

      (a) there is a significant dispute between the parties as to the proper interpretation of a “census” as that term is used in Article 16.2.3 of Bunreacht na hEireann. That dispute is brought into stark relief by a consideration of the timing of certain important events. The census was conducted in April 2006. On 19th July, 2006 the Central Statistics Office (“the CSO”) published a so called “preliminary report”. On 29th March, 2007 the CSO published what are called the “Principal Demographic Results”. Thereafter on 26th April, 2007 the CSO published “Volume 1: Population Classified by Area”. It will be necessary to analyse the contents and status of those three separate publications in due course. However the argument between the parties centres on whether the “preliminary report” or alternatively the “Principal Demographic Results” can be said to be a census for the purposes of Article 16.2.3. The State argues that only the final figures can be properly be regarded as a census for the purposes of the Constitution.

      (b) There is a significant issue between the parties as to the consequence of the publication of a census (howsoever defined) which reveals that the ratio between seats and population in the existing constituencies has become sufficiently skewed so as to no longer represent a level of proportionality consistent with the Constitution. It is argued on behalf of the deputies that, immediately upon the publication of such a census, the ratio between deputies and population is immediately not “the same throughout the country” in accordance with” the last preceding census”. In those circumstances it is said that an Electoral Act which mandates those constituencies becomes immediately inconsistent with the Constitution and ought to be declared to be such by the courts. On the other hand counsel for the State argues that a proper construction of Article 16.2.3 is to the effect that it places an obligation on the Oireachtas, as soon as it may be practicable, to bring about amending legislation so as to bring the constituencies back into conformity with the constitutional obligation of proportionality. In the context of “practicability” reference is made to the Constituencies Commission established by the Electoral Act, 1997 (“the 1997 Act”). It is argued by the State that the establishment of such a commission is a reasonable approach by the Oireachtas to the task of constituency revision. It follows, it is said, that the reasonable time required to allow the commission to do its work must be taken into account in determining what is “practicable”.

      (c) The third issue is, in substance, to be found in a combination of issues (a) and (b). If the plaintiffs are correct in their contention that any Electoral Act providing for constituencies which have become disproportionate by reason of population changes disclosed in a census, becomes unconstitutional immediately on the publication of that census, then it is clear that by the time this case came to hearing the existing constituencies had to be viewed in the light of, at a minimum, the contents of “Volume 1: Population Classified by Area” which had been published a short number of days earlier. On the other hand, if the State is correct as to its contention that the publication of a census places an obligation on the Oireachtas to enact legislation (but does not, at least immediately, make the existing legislation unconstitutional) then the question of the appropriate course of action to be adopted by a court in such circumstances arises. This in turn is dependent on the question of when the population of each constituency can be said to have been ascertained by reference to the census conducted in 2006. If the population can be said to have been so ascertained on the publication of the preliminary report then it follows that the obligation on the Oireachtas to enact new constituencies commenced as of the 19th July, 2006. On the other hand if it cannot be said that the population was ascertained for the purposes of Article 16.2.3 until one or other of the later reports, then the obligation on the Oireachtas would only have commenced in either March or April of this year respectively. Against that background two interconnected questions arise. Firstly if the State argument is correct (and it is part of that argument that the phrase “as far as it is practicable” within Article 16.2.3 relates not just to the mathematical proportionality between constituencies but also to the practicability of putting in place appropriate amended constituencies) then questions arise as to how soon after the publication of a census it can properly be said that the constituencies are no longer, “so far as it is practicable”, within a constitutional ratio. A connected question concerns the appropriate remedy that should be given to a plaintiff in the event that the Oireachtas has not amended the constituencies within whatever timescale might be mandated by the constitution.

      (d) Finally the discrete issue raised on behalf of Mr. Molloy as to the powers of the Minister to add a deputy to the Dublin West constituency also needs to be determined.

      While the issues were addressed broadly in the above order at the hearing before me, it seems to me that issue (b) is the question to which I should first turn.

3.5 However before going on to consider that issue I should deal briefly with the question of whether it has been established that the population figures revealed by the 2006 census render the existing constituencies, as provided for by the 2005 Act, disproportionate. As indicated earlier this fact, as a matter of mathematical calculation, was not in serious dispute. However it is important, in proceedings such as this, that the court indicate that it is satisfied as to the necessary factual basis for the contentions of the parties even where they are not seriously disputed. It is also important, for the purposes of consistency, to indicate the basis upon which the court came to its conclusions.

3.6 An analysis of the judgment of Budd J. in O’Donovan v. Attorney General [1961] I.R. 114 shows the level of disparity in the constituencies with which the court was concerned in that case. Professor Michael Marsh, who gave evidence on behalf of the deputies, has expressed the opinion that the level of disparity, viewed as a whole, and demonstrated in the figures quoted by Budd J. in his judgment in O’Donovan is comparable to the level of disparity between the ratio of population to deputy under the 2005 Act when considered in the light of the 2006 census. No other witness contested that view it seems to me clear that Professor Marsh engaged in an accurate assessment of the question of the comparable proportionality of the constituencies dealt with in O’Donovan and of the current constituencies. It is important to note, as Professor Marsh agreed, that there is no single measure by reference to which a set of data such as the ratio of population to deputy across the range of constituencies can be uniquely measured.

3.7 It is possible, for example, to identify the extreme cases. As indicated earlier in the course of this judgment five constituencies are currently more than 10% out from the average. However if one were to impose an artificial limitation on the extent to which any one constituency could deviate from the national average (at, say 5%) then it would, of course, be possible to devise a very skewed set of constituencies which nonetheless conformed with that artificial limit. In an extreme case half of the constituencies might be more or less 5% above average with the other half being more or less 5% below the average. No single constituency might breach an artificial 5% barrier but nonetheless the overall level of disparity taken as a whole would be quite substantial.

3.8 An alternative means of looking at the totality of the disparities might be to take an average of the disparity from the mean as a reasonable measure of the overall level of disparity. This again could be misleading in an individual instance. It is possible that virtually all constituencies might approximate very closely to the national average (say less than 1% variance). However there might be (say) 5 or 6 constituencies which varied to the extent of 20%. The average disparity would, in those circumstances, be less than 3% but it could hardly be said that the constituencies represented an appropriate proportionality. I mention those two examples simply to illustrate that no single measure will necessarily give, in all cases, the full picture. The court has to take a broad view as to whether the constituencies, taken as a whole, are, “insofar as it is practicable” proportionate.

3.9 In addition it is, also, necessary to have regard to the other criteria which the Supreme Court identified in Re The Electoral Amendment Bill 1961 (1961) I.R. 169 (“The Reference”) as being properly taken into account. I will refer to those criteria in slightly more detail in due course. However it might well be that on one set of census figures it might be possible (or, more accurately “practicable”) to produce a set of constituencies which broadly conformed with the criteria identified by the Supreme Court in The Reference at a very low level of variation from exact mathematical equivalence. A subsequent set of census figures might make it much more difficult to achieve the same low level of variance from mathematical equivalence simply because of a different distribution of the population by reference to the sort of historic and natural boundaries mentioned in the course of the Supreme Court judgment. While it is clear from the judgment of the Supreme Court in The Reference that this court must afford a significant margin of appreciation to the Oireachtas in determining how best to define a set of constituencies which is proportionate, but which also makes appropriate allowance for those criteria, nonetheless it is possible to envisage circumstances where a court would be satisfied that a particular set of constituencies varied from exact equivalence by a margin which could not be justified by reference to those other criteria. It might be possible to envisage another case where a court would be satisfied that an even greater level of mathematical variance might, nonetheless, be justified because of the greater difficulty which would be encountered in producing a set of coherent constituencies (by reference to the relevant criteria) at any lesser level of variance. This again illustrates the important point that there is no one level of variance which can be specified as being a “magic formula” by reference to which the acceptability or otherwise of constituencies are to be determined. The “fit” of population into natural boundaries is one reason for this. The fact that it is not possible to adopt a single definitive measure of variance is another.

3.10 However on the facts of this case I am satisfied that, applying any reasonable measure of variance, it must be concluded that the disparity with which I am concerned (by reference to the 2006 census) is at least as great as that which led to the constituencies under consideration in O’Donovan being struck down. No basis was advanced on behalf of the State to suggest that there would be any greater difficulty in producing a set of proportionate constituencies (having regard to the distribution of population) today than existed in relation to the constituencies under consideration in O’Donovan. On that basis I am satisfied that the mathematical variance of the existing constituencies is outside constitutionally permissible bounds. I have engaged in this slightly elaborate exercise principally for the purposes of emphasising that the fact that the existing constituencies are at least as disproportionate as those struck down in O’Donovan might not, in an appropriate case, of itself, be sufficient. It would always be open to the State to attempt to persuade the court by cogent evidence that there were difficulties within the criteria identified by the Supreme Court in The Reference which would justify a greater level of variance. That was not, of course, attempted in this case.

3.11 I also agree with the observation of counsel for the State that the Supreme Court in The Reference may not have been prepared to go as far as Budd J. in O’Donovan in requiring a degree as approximation to mathematical equivalence. The definitive ruling is, of course, that to be found in The Reference. The precise approach of Budd J. to the assessment of the degree of proportionality required to satisfy its Constitution needs to be treated, therefore, with some caution.

3.12 Finally, before leaving this topic, I should note one point simply because it was the subject of brief comment by counsel in the course of the case. Attention was drawn to the fact that both in learned legal textbooks and in the reports of a number of Constituency Commissions over the years, reference has been made to a standard of 5% variance as having being regarded in O’Donovan as being acceptable. Leaving aside the issues which I have already addressed as to whether it is possible to identify any such universal threshold as being applicable at all, one further matter of fact is worthy of some note. What Budd J. actually did in O’Donovan was to identify 1,000 members of the population as being a broad estimate of the likely population of an electoral division. On that basis Budd J. felt that it would always be possible to move a single electoral division so as to bring about a greater level of proportionality. He, therefore, indicated that it would be hard to justify any greater level of variance. It is not clear that the Supreme Court in The Reference were prepared to go quite that far. It is, of course, the case that at the time when Budd J. was considering the constituencies in O’Donovan the overall ratio of population to deputy was very close to 20,000:1. It seems likely that the 5% figure had its provenance in the ratio of the 1,000 population figure for an electoral division identified by Budd J. and that overall national ratio of 20,000.

3.13 However it is manifestly clear that any such inference is incorrect. The figure of 1,000 was an absolute figure being the amount of population which could easily be moved from one constituency to the next by the simple expedient of moving an electoral division (having regard to the expected population of an electoral division at that time). In that context the “margin” of 1,000 population (if that is the appropriate way to interpret the judgment of Budd J.) has to be seen in the context of a constituency which would have had a population of between approximately 60,000 (in the case of a three seat constituency) and 100,000 (in the case of a five seat constituency). Thus the 1,000 margin does not represent 5% but rather represents somewhere between 1% and 1.66%. That is not to say that, in general terms, a margin of 5% is, or is not, an appropriate consideration. The true answer is that it may depend on the extent to which all of the other relevant factors can properly be met within that or a smaller margin. I merely note these matters for the purposes of identifying that there does not appear to be any justification for the assertion that O’Donovan is authority for the acceptability of a 5% margin.

4. What are the consequences where a new census shows the constituencies to be disproportionate
4.1 It should be recalled that Article 16.2.3 provides as follows:-

      “The ratio between the number of members to be elected at any for each constituency, as ascertained at the last preceding census shall, so far as it is practicable, be the same throughout the country.”
4.2 As pointed out earlier, the parties disagreed strongly as to the proper interpretation of the Article concerned. Apart from the question as to the proper interpretation of what is or is not a census, for the purposes of the Article, the parties also differed as to the extent of the qualification which is to be derived from the presence of the words, “insofar as it is practicable.” The deputies and Mr. Molloy argue that the term “insofar as it is practicable” relates only to questions directly concerning the ratio of population to deputies.

4.3 As is clear from O’Donovan and The Reference, exact mathematical equivalence is not required. In delivering the decision of the Supreme Court in The Reference Maguire C.J. expressed the following view:-

      “The sub-clause recognises that exact parity in the ration between members and the population of each constituency is unlikely to be obtained and is not required. The decision as to what is practical is within the jurisdiction of the Oireachtas. It may reasonably take into consideration a variety of factors, such as the desirability so far as possible to adhere to well known boundaries such as those of counties, townlands and electoral divisions. The existence of divisions created by such physical features as rivers, lakes and mountains may also have to reckoned with. The problem of what is practical is primarily one for the Oireachtas, whose members have a knowledge of the problems and difficulties to be solved which this court cannot have. Its decision could not be reviewed by this Court unless there is a manifest infringement of the Article. This Court cannot, as suggested, lay down a figure above or below which a variation from whit is called the national average is not permitted. This, of course, is not to say that a Court cannot be informed of the difficulties and may not pronounce on whether there has been such a serious divergence from uniformity as to violate the requirements of the Constitution.”
4.4 It is, therefore, established and common case, that the phrase “insofar as it is practicable” qualifies the obligation that the ratio be the same throughout the country so that the Oireachtas is entitled to take into account a variety of factors which may permit, as a consequence, a movement away from exact mathematical equivalence. That much is clear.

4.5 However the issue here is as to whether the phrase “insofar as it is practicable” also governs the timescale within which any amendment to the constituencies (as may be required to reflect new populations figures ascertained in a census) must be put in place. It is important to note that this argument arises, in principle, irrespective of which publications of the CSO may be regarded as being a census for the purposes of Article 16.2.3.

4.6 Whatever may be regarded as, properly speaking, a census for those purposes, there will always be a situation where, from time to time, such publication will reveal that it is no longer possible to stand over the existing constituencies as having (even paying appropriate regard to the various factors identified in the judgment of Maguire C. J. from which I have quoted above), “insofar as it is practicable”, a ratio of deputies to population which is “the same throughout the country”. It is equally obvious that it would not be possible for the Oireachtas to consider the newly revealed population figures, the appropriate variations to existing constituencies required to deal with the problems shown up by those figures, and enact appropriate legislation, on the very next day. In those circumstances the argument in this case brings into stark relief the proper interpretation of Article 16.2.3.

4.7 On the plaintiffs’ argument, as soon as a census establishing a sufficient disparity is published, the existing Electoral Act becomes immediately inconsistent with the Constitution and must be declared as such by the court. On the State’s argument the publication of such a census places an important obligation upon the Oireachtas to take appropriate measures to remedy the situation that has been revealed by that census. However on the State’s argument, the existing Electoral Act does not, automatically, and immediately, become inconsistent with the Constitution.

4.8 The argument in favour and against both points of view centred on both the proper approach to the construction of the text of Article 16.2.3 itself and questions concerning the harmonious interpretation of the Constitution taken as a whole. It is with the latter aspect that I propose to start.

4.9 While the substantial population changes experienced in the last number of years may not have been anticipated at the time when the people adopted Bunreacht na hEireann in 1937, it was nonetheless clear at all times that there was a real possibility that, from time to time (if not at every census) there would be revealed sufficient population changes that would mandate an alteration in the constituencies. In those circumstances counsel for State argues that it could not have been the intention that the constitutional model adopted in Article 16 should create the virtual certainty that, from time to time, there would be no valid set of constituencies in being. As was pointed out by counsel for the State, a number of further questions arise in the event that the argument addressed on behalf of the plaintiffs under this heading is correct. If the then existing Electoral Act is unconstitutional, and must be struck down, then what happens in the event that there is a requirement for an early general election.

4.10 In O’Malley –v- An Taoiseach [1990] ILRM 461 this court made clear that there could be no barrier to the disillusion by the President of Dail Éireann on foot of a request by a Taoiseach who had not lost the confidence of the house. Neither could there be any court barrier to the Taoiseach giving the appropriate advice to the President which would lead to such dissolution. Neither side in this case suggested that O’Malley was, in any respect, wrongly decided. In those circumstances, it seems clear that there is an absolute entitlement on the part of the Taoiseach of the day (provided that he or she has not lost the confidence of the House) to seek and obtain a Dissolution of Dail Éireann from the President. In those circumstances the question arises as to what is to happen in the event that the Taoiseach seeks such dissolution in the immediate aftermath of the publication of a census showing that the population per deputy is outside constitutionally permitted parameters. Article 16.2.1 requires that constituencies be determined by law. There is, therefore, an obligation on the Oireachtas to provide for constituencies. That Article is mandatory rather than permissive.

4.11 As pointed out by counsel on behalf of the State one of two possible consequences might flow from a declaration of inconsistency of an Electoral Act (providing for constituencies) with the constitution. The first possibility is that the Act, as a whole, may be taken to be unconstitutional (including a section which would undoubtedly have been present in such an Act repealing previous provisions specifying constituencies). In those circumstances the next proceeding set of constituencies would “spring up”. In all likelihood those constituencies would be even less proportionate to the now existing population structure and would themselves, undoubtedly, be unconstitutional. In the likely absence of any previous set of constituencies which would actually meet the constitutional proportionality requirement, then it follows that all previous Electoral Acts providing for constituency boundaries would also necessarily have to fall so that there would, in that eventuality, be no valid constituencies.

4.12 The alternative view (which is to the effect that section providing for the repeal of the previous provisions in relation constituencies would not itself fall) also leads to a situation where there would be no law providing for constituencies.

4.13 It, therefore, follows that, if the argument put forward on behalf of the deputies and Mr. Molloy is correct, the Constitution necessarily envisages a situation where there will, from time to time, be no valid constituencies under which an election could be fought. This will occur on each and every occasion when a new census is published and where the population distribution per constituency as disclosed in the new census shows that the ratio between deputies and population has moved beyond a constitutionally permissible variation having regard to all appropriate factors. It is said, correctly in my view, that such an eventuality is one which is unlikely to have have been contemplated by either the people when adopting Bunreacht na hEireann or, indeed, by the framers.

4.14 To get around that difficulty it was argued by counsel on behalf of the deputies that this court might be minded to adopt the practice identified by the Supreme Court of British Columbia in Dixon –v- Attorney General of British Columbia [1989] 4 W.W.R. 393. In that case the petitioner sought an order declaring invalid the British Columbia legislation establishing provincial electoral districts. Having concluded that the relevant provisions were unconstitutional, the court went on to consider what remedy, if any, was appropriate. Having considered argument in that regard McLachlin C.J. expressed her view in the following terms:-

      “It is my conclusion 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 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 required.”
4.15 That such a constitutional crisis would occur in those circumstances is undoubtedly the case. In the case of Bunreacht na hEireann Article 16.2.1 requires that there be legislation providing for constituencies. If the only legislation providing for such constituencies is struck down (or, indeed, if all previous such legislation suffers a similar fate) then there will be no valid constituencies. On the other hand an election may be required and it is clear from O’Malley that no legal barrier can be placed in the way of such an election. How then can such an election be conducted. The answer, it is said on behalf of the deputies’, is that the existing constituencies, even though unconstitutional, could be continued in place for a period of time. In those circumstances, it is said, any election called could be conducted on the old constituencies. The period of time would, of course, be defined by reference to whatever period was considered necessary to a enable the Oireachtas to enact new legislation.

4.16 While an order of that type would undoubtedly solve the constitutional problem, it has to be said that the researches of counsel on both sides could not come up with any case in this jurisdiction where there has even been a suggestion that it is possible for the courts to make a declaration of the type made by the Supreme Court of British Columbia in Dixon. While there has been some significant debate in this jurisdiction (see most recently A. v. Governor of Arbour Hill Prison [2006] IESC 45) as to the extent to which court orders made or other actions taken on foot of a statute declared to be unconstitutional (which were made or taken prior to the declaration of its inconsistency with the constitution) can nonetheless be regarded as valid, there has never, so far as I am aware, been even a suggestion that actions taken on foot of a statute subsequent to a declaration of inconsistency with the constitution could be regarded as valid. The making of an order of the type made by the courts of British Columbia in Dixon would, therefore, be a radical and novel departure in Irish constitutional law. That is not to say that there may be difficult circumstances which might require a court to be creative in the form of order which might be required so as to retain constitutional harmony. I would not, therefore, on the ground of lack of precedent alone, rule out the possibility that, in unusual and exceptional circumstances, an order of the type made in Dixon might have to be contemplated. However it is a form of order which is so out of kilter with the general jurisprudence in respect of declarations of inconsistency with the constitution that, in my view, such an order could only be made, if at all, if there were no other “solution” to the constitutional difficulty which had arisen.

4.17 Counsel for the deputies’ also places reliance upon the fact that this court, in O’Donovan, did in fact make an order striking down the constituencies which had been enacted by the Oireachtas in 1959. The precise position which obtained between the decision of this court in O’Donovan and the final passage into law of the Electoral (Amendment) Act, 1961 (after the Supreme Court had determined in the reference that it was consistent with the constitution) is, therefore, open to debate. However, counsel for the State makes the point that the situation with which this court was faced in O’Donovan is, in principle, different in a number of respects to the situation with which I am faced in this case. In O’Donovan the Oireachtas had enacted a new set of constituencies which were then challenged. In this case no constituencies have been enacted and the court is invited to declare invalid constituencies whose enactment predated the census by reference to which they are now said to lack proportionality.

4.18 A further distinction is drawn between the underlying circumstances giving rise to the challenge in O’Donovan and the present situation. The case made on behalf of the State in defending O’Donovan was to seek to justify the constituencies which had then recently been enacted on the basis of an asserted entitlement to allow a lower level of population per deputy in rural constituencies with a view to preventing those constituencies from becoming too large. Budd J., in this court, held that such a consideration could not, properly, be taken into account in determining the constituency boundaries. O’Donovan was, therefore, a case in which the only issue was as to the entitlement of the Oireachtas to take into account such a factor in drawing constituency boundaries. While, doubtless, the previous constituencies which had been formulated in 1947, would have been, by the late 1950s and early 1960s, somewhat disproportionate by reference to the then last available census (that taken in 1956) nonetheless this court was not asked to consider what the consequences of a declaration of unconstitutionality would be in the event that circumstances led to a general election before new constituencies had been enacted.

4.19 Likewise two aspects of current relevant constitutional jurisprudence were not clarified until the decision of this court in O’Malley. Firstly it had not been clarified that the courts had no jurisdiction to prevent the dissolution of the Dáil. Secondly it is also clear from O’Malley (although obiter) that there is an obligation on the Oireachtas to revise constituencies as soon as it becomes clear from figures ascertained in a census (however defined) that the existing constituencies no longer have the level of proportionality which the constitution requires. It would appear that, for much of the earlier history of the State, the view had been taken that a constituency review was not required in that way.

4.20 Article 16.2.4 does, of course, require that, in any event, there be a constituency review at least every 12 years. The previsions of the Free State Constitution in relation to constituencies are identical to those contained in Bunreacht na hEireann save that the provision requiring a periodic review specifies a period of 10 rather than 12 years. Up to the late 1950’s it would appear that constituency revision only occurred at the outer limit of what was constitutionally permissible with revisions taking place in respectively 1923, 1935, 1947, and 1959/61. Indeed it would appear that the revision in 1935 was outside the time provided for in the Free State Constitution. In addition the revision in 1959/1961 was delayed by reason of O’Donovan and The Reference. Therefore up to that time it would seem to have been considered that it was not necessarily the case that the Oireachtas was required to conduct a review after each census. Rather it seems to have been the case that it was considered that a review was only required at or around the time specified as the default period provision in the respective constitutions.

4.21 While the comments of Hamilton P. in O’Malley, in relation to the obligation on the Oireachtas to revise constituencies, are obiter, it was not suggested by counsel for either side that they do not represent a correct statement of the law. I agree. I am, therefore, satisfied that the Oireachtas has an obligation, as soon as significant and unjustifiable disproportionality is revealed by census figures, to act to remedy that difficulty.

4.22 One final consideration that needs to be taken into account under this heading stems from the fact that the constituencies which currently exist under the 2005 Act were, at the date of their enactment, entirely consistent with the last census then available (that is the census taken in 2002 whose results became available throughout 2002 and 2003). There can be no suggestion, therefore, that the 2005 Act was unconstitutional as of the time of its enactment.

4.23 In all those circumstances I am satisfied that a more harmonious construction of the various sub clauses of Article 16, favours the view that the immediate consequence of the publication of census figures which show that the constituencies then currently provided for by law are no longer proportional “so far as it is practicable”, leads to a situation where the Oireachtas has an immediate and pressing obligation to put in place measures designed to remedy that problem by enacting appropriate amending legislation. Unless, therefore, the alternative construction argued for by the plaintiffs was mandated by the clear text of the Constitution, I would not favour the view that an existing Electoral Act is, in those circumstances, immediately unconstitutional. I would leave to a case where it specifically arises, the question of what the proper approach of a court should be to circumstances where, in purported reliance on that obligation, the Oireachtas enacts constituencies which, themselves, fail to meet the proportionality test.

4.24 However, I am satisfied that, in fact, the approach which I have indicated, is in conformity with a proper construction of the wording of the Article itself. While it is true to state that the Article requires that the “ratio” be proportionate “insofar as it is practicable”, it seems to me that the ratio can only be provided for by law as the ratio is determined by the constituencies which must themselves be determined by law. Therefore, the ratio can only be changed by the enactment of a new law. Thus it is not practicable to change the ratio until such time as a reasonable opportunity has been given to the Oireachtas to bring about that change. Even though the ratio itself might be disproportionate it is not a breach of the obligation that it be proportionate “insofar as it is practicable”, if it can be shown that there was no reasonable opportunity, after the results of a relevant census had become available, to change the ratio. In all the circumstances I am satisfied that an existing Electoral Act does not, as a matter of principle, become unconstitutional immediately upon the publication of census figures showing a disproportion between the level of representation from one constituency to another. Such a situation gives rise simply to an obligation on the Oireachtas to take appropriate remedial action.

4.25 However as a consequence of that finding it is necessary, now, to determine the time at which that constitutional obligation commenced for it is only against that starting point that the actions of the Oireachtas in meeting the constitutional obligation which I have identified can be judged. In order to assess that timeframe it is necessary to have regard to the nature of the various figures published by the CSO. I now turn to that question.

5. The CSO Publications.
5.1. As pointed out earlier the CSO published three sets of figures specifying the population of the State as measured by reference to the census conducted in April, 2006. The census of population concerned was taken on the night of Sunday 23rd April, 2006. It should be noted that the population as ascertained in a census is therefore a “snap shot” of the population on that night. In particular the distribution of the population throughout the country is as it was on that night. The census does not record, for population calculation purposes at least, where persons might be said to be “ordinarily resident” or the like. While it might be expected that much of the variations that could, theoretically, occur because of people being recorded in a different part of the country to that where they ordinarily reside (by reason of their being in that different location on the night in question), might cancel each other out, nonetheless it should be noted that a census, whenever and however conducted, does run the risk that there will be some distortion between the figures produced and what might, in one sense, be said to be the “true” population of an area. Doubtless those charged with conducting the census choose a time for its conduct which is likely to minimise the risk of any significant distortion occurring in this way. However the possibility that there may be at least a minor disparity between the population as recorded and the “normal” population of an area (however one might define such a concept) remains. It was in that context that Professor Michael Marsh, who gave evidence on behalf of the deputies, expressed the view, which I accept, that a census is, to a material extent, only an estimate, no matter how well and how thoroughly it is conducted. I should emphasis that there was no criticism of any type whatsoever as to the manner in which the CSO conducted the most recent census or indeed any previous exercises.

5.2 While it is true, therefore, to say that there is a sense in which a census remains an estimate (albeit a very good one) it does not seem to me that this is a very relevant issue in this case. It is the Constitution itself which uses the yardstick of the census as the basis upon which population is to be calculated for the purposes of determining the proportionality of that population to deputies. The fact that it may, to some extent, therefore, be an estimate does not, it seems to me, affect the issue.

5.3 In any event the “preliminary report” was published on the 19th July, 2006. That report contained, amongst other things, two tables as follows:-

      “Table 3 – Persons in each constituency, defined in the Electoral (Amendment) Act, 2005 for elections to Dáil Éireann

      Table 4 – Population of each province, county, city, urban area, rural area and electoral division, 2002 and 2006.”

5.4. It should also be noted that the preliminary report drew attention to the possibility of the figures being revised and went on to state that they should therefore not be regarded as having statutory force. The latter appears to be a reference to the provisions of s.45 of the Statistics Act, 1993 (“the 1993 Act”) which permits certain figures published by the CSO to be admitted in evidence in Court without formal proof. It is clear that CSO itself did not regard the preliminary figures as published as being suitable for having that status attached to them.

5.5 Mr. Aidan Punch is the senior statistician in the CSO who has been in charge of the last four censuses and has 35 years experience in the CSO. Having regard to the independent and neutral position of the CSO (see for examples s.13 of the 1993 Act) it was agreed between the parties that, while notionally tendered on behalf of the State, Mr. Punch would be regarded as a neutral witness who could be cross examined by both sides.

5.6 Mr. Punch described the manner in which the preliminary figures were calculated. Each census enumerator (there appeared to be approximately 4,400) conducted the exercise of ensuring that each household within his or her enumeration area properly fills in the census form by reference to the night in question. Prior to sending the completed individual census forms back to the Census Office in Swords, the enumerator is required to extract from each census form a limited amount of information being simply the total number of persons broken down as and between males and females and to sum that information on a street/townland basis in each of the electoral divisions within their remuneration area. This exercise is called a clerical summary and is done for each emuneration area. That information is, in turn, remitted to the census office. It is the summation of that extracted information which gives rise to the figures contained in the preliminary report. The preliminary report does not involve the Census Office in carrying out any exercise of reviewing or emunerating the individual census returns from individual households. Those individual returns form the basis of the subsequent reports.

5.7 It is as against that background that the status of the “preliminary report” needs to be considered. Mr. Punch, in the course of examination, described the figures as “standoverable”. In other words he conveyed that it would be reasonable for those who wished to, for example, formulate policy (in the public or private sector) based on population trends, to place reliance on the figures as being a close approximation to the figures that would ultimately be revealed when the full reports became available.

5.8 An analysis was also conducted and put in evidence (upon which Mr. Punch was asked to comment), of the differences, on an electoral division basis, between the figures contained in the preliminary report and those which may be gleaned from the subsequent final reports. Those differences were set out in a table which was established in evidence. However the most glaring examples of difference can be readily explained. For the purposes of its own analysis the CSO gives population for what are described as “urban areas”. As all will be aware many urban areas have, in practice, outgrown their formal boundaries in recent years. However by reference to certain relevant criteria the CSO attempt to define the population of what might, in real terms, be properly described as, for example, a provincial town. That exercise necessarily involves allocating some part of the population in what are, for administrative purposes, still neighbouring rural areas to the town in question. Almost all of the significant variations (that is variations in excess of approximately 800 between the figures revealed in the preliminary returns and the subsequent final figures) can be explained by reference to this phenomenon.

5.8 I am satisfied, therefore, that, as a fact, the figures set out in the “preliminary returns” are accurate to quite a high degree. Obviously their accuracy is greater at the larger scale. The variance in the population of the State as and between the preliminary and final figures was a total of 4,923 or 0.12%. As some of the variances were positive and some negative the total amount of the differences disregarding whether they are negative or positive was somewhat higher at slightly over 7,500 or close to 0.18% when taken at a constituency level. This fact of positive and negative variances cancelling each other out obviously makes it more likely that somewhat larger differentials (on a percentage basis) may be found in smaller areas. The fact that 50 persons may be allocated to the wrong townland in a rural area could give a very significant percentage difference between the (wrong) preliminary figures and the (correct) final figures for that townland. The difference might be of no relevance at the level of even an electoral division if both townlands were in the same electoral division as the difference would then cancel out. It is all the more likely that such differences will cancel themselves out as there are larger aggregations at the level of constituency, counties and the like.

5.9 Therefore, as a matter of practicality, it seems unlikely that there would be very significant variance at the level of an electoral division between the preliminary figures and the final figures. To the extent, therefore, that constituencies are made up of a number of electoral divisions, it seems equally unlikely that there would be significant variation between the preliminary and the final figures at the level of a constituency. In the case of the 2006 census the most extreme variance appears to have occurred in relation to the constituency Meath East where the difference was 554 (or 0.63%) while the variance in respect of Dublin South West was 612 or 0.62%. All in all 33 of the constituencies (over three quarters of the total) had a variance between preliminary and final figures of less then 0.2%.

5.10 A comparable exercise was engaged in in analysing the differential between the preliminary and final figures derived from the 2002 census which gave a similar result.

5.11 The final figures are compiled by a detailed emuneration of the actual household forms filled out and returned to the Census Office. That analysis is, undoubtedly, therefore, more accurate and must be regarded as the definitive figures being derived from the same data base as all of the additional detailed census information that will, over time, come to be published.

5.12 One final matter that needs to be noted is the distinction between the reports published in, respectively March and April, 2007. The report published on the 29th March, 2007 (“the March Report”) is described by Mr. Punch as the first in a serious of reports containing final census figures. The March Report contains, in Table 3, a list of the number of persons in each constituency.

5.13 Thereafter the CSO publishes a whole series of volumes containing the information to be derived from the 2006 census in relation to the range of information gathered on the census form. The first such volume concerns population. It is important to note that the population (or its distribution) contained in the April 2007 Volume (“the April Report”) is not, in any respect, different from the population contained in the March report. It is simply that the information is set out in more detail. While, for present purposes, the most detailed breakdown contained in the March report was at the level of a constituency, the April report contains details of the population of each urban area, rural area and electoral division. There is no doubt but that some of the additional information published in the April report is likely to be necessary for the purpose of formulating constituencies. While the information contained in the March report will give a definitive and final account of the disparity between constituencies it does not give information sufficient to allow the sort of relatively fine turned adjustment that may be necessary to bring the constituencies back into order. At a minimum, information at the electoral division level is required for that exercise. Indeed in urban areas it would appear that it is frequently the case that information at a level below that of electoral division is necessary to allow an appropriate exercise to be carried out. In rural areas this is also, sometimes, necessary with information at a townland level being required.

5.14 In that context it is important to note that sub electoral division information is never, in fact, published, but is made available directly by the CSO to those framing constituency boundary proposals under the provisions of the 1997 Act.

5.15 It was also accepted by Mr. Punch that the electoral division level information was necessarily available (even though not published) as of the time of the publication of the March report because information at that level of particularity would, in fact, be necessary to enable the population to be divided on a constituency basis. It is also clear that information at the sub electoral division level (should same prove necessary) was as capable of being provided after the publication of the March report as at any later stage. The reason for this is that the methodology adopted is that all of the information contained in the individual household census forms is placed on computer (a very painstaking task) so that, once properly entered onto the computer and available for analysis, the more detailed breakdowns that might be required to fine tune constituency boundaries will be available at least by the time that it is possible to publish a report such as that which was published in March.

5.16 Having described the various reports it is now necessary to consider which reports may be regarded as having the status of a census for the purpose of Article 16.2.3. I now turn to that question.

6. What is a census for constitutional purposes?
6.1 A more rigorous way of putting that question is to ask what might properly be regarded as the population “as ascertained at the last preceding census”. It is accepted by all sides that the population figures specified in the April report determines, for any time after that report had been published, the population as ascertained by the last preceding census.

6.2 The next question which arises is as to whether a similar conclusion should be reached (contrary to the submissions of the State) about the figures contained in the March Report. It is pointed out, correctly so far as it goes, that the March Report does not contain a sufficient level of detail in relation to the breakdown of the figures such as would enable a re-drawing of constituencies to take place. However, that does not seem to me to be an end of the matter. The reality is that in at least some cases even the April Report does not contain that level of detail. It is both permissible in theory, and actually occurs in practice, that further fine tuning is required to be conducted by reference to the population of areas defined below that of the level of electoral divisions. This information is never published. It is ascertained by the Constituencies’ Commission requesting it directly from the CSO. In those circumstances it does not seem to me that the fact that information at any particular level is published or not published is, of itself, determinative of whether the population has been ascertained by a census. It seems to me, therefore, that on the evidence, all of the information necessary to conduct a review of constituencies was available (even though not all of it was published) by the time that the March Report was publicly available on the 29th of that month. The fact that it might be that more of the information required for constituency revision was unpublished as of that stage than would have been the case after the publication of additional information by means of the April Report does not, it seems to me, affect the issue. The figures are there. They are available to be given to whoever might be charged with the task of reviewing constituencies. They are, therefore, in my view, ascertained for the purposes of Article 16.2.3. If one were to take the view that they could not be said to be “ascertained” because they had not been published, then this would mean that there would never be sufficiently “ascertained” figures because it is clear that some of the figures necessary to formulate proper constituencies are likely never to be published.

6.3 I am, therefore, satisfied that there was, as of the 29th March at a minimum, a set of figures available from which any exercise necessary to ascertain the population of both existing constituencies and any proposed constituencies could have been calculated. It seems to me, therefore, that at all times after that date, at a minimum, the population of any relevant area that required to be ascertained by reference to the “last preceding census”, required to be so calculated by reference to the 2006 census.

6.4 The remaining question is as to the status of the “preliminary report”.

6.5 The starting point has to be to note that it would be a strange result indeed if figures which the CSO itself describe as being “of no statutory force” were nonetheless to have constitutional force. That the figures are, in Mr. Punch’s words, “standoverable” cannot be doubted. Any policy maker or commercial entity attempting to formulate policy in the latter part of 2006 would have been extremely foolish to formulate that policy on the basis of the results of the 2002 census rather than the preliminary results of the 2006 census. But that is not the issue in this case.

6.6 It seems to me that what Article 16.2.3 creates is a definitive benchmark by reference to which the proportionality of constituencies are to be determined. In that context it is worth noting some aspects of the evidence of Dr. Eoin O’Malley of Dublin City University who gave evidence on behalf of the deputies. Dr. O’Malley set out an interesting theoretical analysis, from the prospective of a political scientist, of the approach to proportionality between elector and public representative. As he pointed out there are a number of bases upon which that proportionality can be considered. One of the reports from the United Kingdom which he cited noted a disproportionality which derives from the tendency of electors in some constituencies to vote in greater numbers than others. Thus, in one sense, the vote of the elector in a constituency where a smaller turnout can be expected may, in practice, be more valuable than the vote of an elector in a constituency where a higher turnout could be expected, even if the proportionality between the electorate and the public representatives to be elected had exact mathematical equivalence.

6.7 In addition, in the course of the hearing, it was noted that there may well be a difference in proportionality between constituencies based on the population of an area, on the one hand, and the electorate in that area, on the other hand. Historically, in this jurisdiction, such a situation may well have occurred by reference to a disparity in the number of children (or other categories of citizen not entitled to vote) who may have been present in each of the constituencies. In more recent times that disparity has the potential to be even greater by reference to a potential disparity between the percentage of non-citizens from constituency to constituency.

6.8 Thus it will be seen that there are a whole range of issues which may intervene between the population of an area and the number of people who actually vote.

6.9 A proportion of those who form part of the population properly calculated may not be entitled to vote. A proportion of those who may be entitled to vote, in all probability, will not vote. A cursory glance at the quotas at the most recent General Election will show that the number of votes which it, in practice, took, on the last effective count, to elect deputies, varied very significantly throughout the country. Some of this was, doubtless, due to the population disparities which are the heart of this litigation. However, other aspects of the difference may well be due to some of the other factors which I have identified. To those may be added the fact that it is inherent in multi-seat constituency proportional representation by the single transferable vote that there will be a difference in the number of electors required to elect a deputy (i.e. the quota) by reference to the number of seats in that constituency even where everything else is exactly proportionate. An example will illustrate. Using the current approximate ratio of deputies to population one might envisage that an entirely proportionate 5-seat constituency would, therefore, have a population of 125,000 while an equally proportionate 3-seat constituency would have a population of 75,000. If one assumes that those not entitled to vote or who do not chose to exercise their franchise amounted, in each case, to 40%, one would be left with a valid poll of 75,000 in the 5-seat constituency and 45,000 in the 3-seat constituency. The quota in the 5-seat constituency would, therefore, approximate to 12,500 (75,000 divided by 6) while the quota in the 3-seat constituency would approximate to 11,250 (45,000 divided by 4). It is, therefore, clear that, even if everything else is equal, PR itself creates a situation where a somewhat higher number of electors are required to elect a deputy in a larger constituency than a smaller constituency.

6.10 I note all of these factors not because they are, strictly speaking, relevant to the issue with which I am concerned. They are, however, indirectly relevant in this way. Freed from any specific constitutional requirement as to how proportionality between population/electorate on the one hand and numbers of elected representatives on the other hand was to be determined (as is, for example, the case in the United States and Canada) there are a whole variety of ways in which one might seek to achieve an appropriate level of proportionality. That debate may be of great academic interest to political scientists. In that context the discussion in his statement of evidence by Dr. O’Malley is interesting. However, it seems to me that it is also wholly irrelevant. Bunreacht na hÉireann selects one model and one model only by reference to which proportionality is to be determined. It is the population as ascertained by the last census. Whatever may be the merits or otherwise of that benchmark it is the benchmark that has been adopted for this State.

6.11 The merits or otherwise of any other way of looking at proportionality do not, therefore, in my view, arise. If the Constitution had not specified that form of benchmarking then there might well be a whole area of legitimate constitutional debate as to the right way to go about determining proportionality. It might well be that constituencies would be regarded as constitutionally acceptable provided that they were proportionate on one acceptable basis of calculation even though they might be less acceptable if a different method of comparing constituencies was to be carried out. However, all of that possible debate in this jurisdiction disappears in the face of the express terms of Article 16.2.3, which chooses population by reference to census as the benchmark. It is important, in that context, to view the comments in the various authorities concerning the value of an individual’s vote as being subject to that overall limitation.

6.12 It is clear that, even if every constituency in Ireland had an exact mathematical proportionality between number of deputies and population, the number of electors that would be required to elect a deputy could vary quite significantly from one constituency to another by reason of, amongst other things, the factors which I have identified above. In that sense it could be argued that the vote of the elector in the constituency that required the higher number of votes to elect a deputy was less valuable than the vote of the elector in the constituency that required the lower number of votes. Against that it might equally be said that there is a sense in which those who trouble to vote should not be penalised by reference to those who do not vote and that those who do vote might be said to be acting, in a general sense, in a surrogate way for those who live within their constituency but are not entitled to vote. Whatever may be the merits or otherwise of all of the arguments that might arise under that heading, the Irish Constitution is clear. Proportionality is only to be tested by reference to population as ascertained by a census.

6.13 I emphasise this aspect of the argument because it seems to me that it demonstrates that what the Constitution does in Article 16.2.3 is to pick a benchmark for better or worse. There can be little doubt that, as of the day before the publication of the “preliminary report”, there was significant information available to show that there was likely to have been quite a significant shift in the balance of population between the 2002 census and 2006 census. However, on any view, no constitutional question mark could have been raised in relation to the constituencies formulated on the basis of the 2002 census prior to the publication of the preliminary report in July 2006. This is so even though it may have been manifestly clear that significant population changes had occurred. To apply that analysis to the argument put forward on behalf of the deputies, it seems equally clear that a policy maker in the public or private sector attempting to formulate policy in the first half (rather than the second half) of 2006 would have had to have treated the figures from the 2002 census with significant caution. He or she might not have had the sort of detailed information available from the preliminary report which a counterpart addressing a similar question in the latter part of the same year would have had. However, he or she is still more likely to have relied on the best estimates of the actual position in early 2006 rather than 2002 census. It seems to me, however, that the constitutional scheme set out in Article 16.2.3 is designed to present a definitive and identifiable benchmark by reference to which the proportionality of constituencies is to be determined. The fact that it may be (and may demonstrably be) out of date does not stop it being the benchmark. One could envisage a situation where there was no census for (say) ten years. It might be demonstrably clear that significant population changes had occurred in the intervening period. It might even be the case that by reason of the absence of a census more detailed estimates of the likely population breakdown by area might have been attempted. Those figures might even, in Mr. Punch’s words, be “standoverable”. However, those figures are not the definitive benchmark census figures which the Constitution mandates as the basis by reference to which constituency proportionality is to be determined.

6.14 There is little direct authority on this point. It is true to state that, in the reference, the Supreme Court expressly rejected an argument put forward, by counsel assigned, that the overall level of the population as of the time of the passage of the 1961 Bill through both houses did not justify 144 deputies. The argument was based on the absence of any reference in Article 16.2.2 to a census. That Article requires that the ratio of population to deputies cannot be less than twenty thousand to one. By reference to the last preceding census the population was just enough to sustain a Dáil of 144 deputies. It was argued that there was cogent evidence available to suggest that the population had decreased since the time of the last available census such that it was then below the figure of 2,880,000 which would have been necessary to sustain a Dáil of 144 members. In rejecting that argument, and notwithstanding the absence of any reference to the population being required to be ascertained by a census in the provisions of Article 16.2.2 (as opposed to Article 16.2.3) the Supreme Court held that it should not engage in a consideration of what the actual population might be. It seems to me that the approach of the Supreme Court in that case is consistent with the view that the Constitution was acknowledged as imposing a definitive benchmark. That benchmark was the census. Even though there might be evidence that the population had changed since the last census, that fact was not to be taken into account. The benchmark approach has the merit of being definitive. I am, therefore, satisfied that what Article 16.2.3 requires is an ascertainment which meets the benchmarking requirement which is implicit in the Article. I am not satisfied that the “preliminary report” meets that criterion. It is doubtless “standoverable”. The figures contained in it are, doubtless, the only ones which any sensible planner would have regard to. They are not, however, the definitive census figures upon which the proportionality benchmarking exercise mandated by Article 16.2.3 is to be carried out.

6.15 I am not, therefore, satisfied that the figures contained in the “preliminary report” can be said to represent the population as ascertained by a census. It follows that, at all times up to the publication of the report of the 29th March, 2007, the last available census report was that published in respect of the 2002 census.

6.16 In the light of that finding it is necessary to turn to the obligation which I have found the Oireachtas has in relation to the alteration of constituencies.

7. The obligation of the Oireachtas
7.1 It is important to note, firstly, that the obligation is one on the Oireachtas rather than the Government. While it is, in practical terms, the case that almost all legislation which passes through the Oireachtas is promoted by the Government, or a relevant Minister, it nonetheless remains the case that the obligation to determine constituencies is one which requires the constituencies to be determined by law. It follows that it is an obligation of the Oireachtas rather than of the Executive.

7.2 How is this obligation to be exercised? The manner of its exercise has, of course, a particular relevance in the context of considering how soon it may be regarded as “practicable” to alter the ratio of population to deputies so as to bring it within constitutional norms. The Oireachtas has chosen to enact the 1997 Act which provides, in Part II, for the establishment of a Constituency Commission and the reporting by that commission to the Chairman of the Dáil of the results of its enquiries together with its recommendations. The report is to be presented “as soon as may be” but in any event not later than six months after the establishment of the Commission. The practical merit of such a proposal hardly needs to be emphasised.

7.3 During the 1960s and 1970s there was much controversy over the drawing of constituency boundaries. Up to that point in time the practice had been that the Government (normally through the Minister for Local Government) formulated a proposal to review constituencies and, in substance, procured its passage through the Dáil and Seanad and thus its enactment into law. Notwithstanding that all constituency models adopted after O’Donovan complied, it would appear, with the constitutional proportionally requirements, this did not mean that there were not significant political controversies over the fairness or otherwise of the constituencies determined. It is not only the proportionality of population to deputy that has the potential to skew an election. Article 16.2.5 mandates the system of proportional representation by means of a single transferable vote. The minimum number of deputies per constituency cannot be less than three. However the Constitution contemplates that there may be a variation in the number of deputies per constituency. It is clear that the larger number of deputies per constituencies the more likely that a proportionate result will ensue. On the other hand very large constituencies give rise to difficulties associated with the representation of individual citizens. A balance is normally struck. However, the precise fine tuning of that balance has, in theory, and it was alleged in practice, the capacity to alter the outcome of elections, particularly where there is a close balance between the completing blocks.

7.4 It was against that background that, on the occasion of each census after 1977, an ad hoc constituencies commission was established to make recommendations for all subsequent constituency revisions. Thereafter, the former ad hoc arrangement was put on a statutory basis by the enactment of the 1997 Act.

7.5 I have no doubt but that Dr. Richard Synnott was correct when he gave evidence concerning the overall importance of the public having confidence in the fact that constituency boundaries have been drawn in an even handed way. I am satisfied that it is more than reasonable for the Oireachtas to put in place appropriate measurers deigned to maintain and ensure a high level of public confidence in the objectivity of the constituency formulation process. To the extent that it may, therefore, be necessary to go through a certain process to ensure that the constituencies as enacted into law have that high degree of public confidence, then I am satisfied that it is reasonable that the process be engaged in. To that extent I am satisfied that it cannot be said that the ratio of deputies to population is outside constitutional norms “insofar as it is practicable” until such time as a reasonable period has elapsed from the availability of relevant census figures to enable such a public confidence maintenance process to be gone through. It is, of course, the case that the 1997 Act cannot alter the Constitution. However, it does seem to me that measures, such as those contained in the 1997 Act, while not expressly mandated by the Constitution, are more than consistent with the overall constitutional imperative of establishing and maintaining legitimate public trust in the institutions of State. I am, therefore, satisfied that the Oireachtas is not in breach of the constitutional imperative to alter constituencies for such reasonable period of time as might be required to allow proper public confidence maintenance measures in the constituency revision process to be gone through.

7.6 Against that overall finding it is necessary to turn to the facts of this case.

8. Application to this case
8.1 For the reasons which I have set out I am satisfied that the term “population as ascertained by the last preceding census” did not refer to the 2006 census until the 29th March, 2007. I am, however, also satisfied that the obligation to review constituencies is one which places an urgent burden on the Oireachtas. This burden is all the more so when the constituencies can be demonstrated to have as high a degree of lack of proportionality as was undoubtedly demonstrated to be the case on the facts of this case. That urgent burden is all the more important in the light of the undoubted entailment of a Taoiseach to seek a dissolution of the Dáil without court interference. The very fact that an election may require to be called at short notice makes all the more important the obligation of the Oireachtas to act with all due expedition to ensure that amending constituencies, necessitated by a new census, are put in place at the earliest possible time.

8.2 On the facts of this case it does not appear to me that it would have been practicable for the Oireachtas, in the period between 29th March, 2007 and the date of the dissolution of the Dáil, to have procured that the process set out in the 1997 Act (or any like appropriate process) could have been completed in time to ensure that, in this case, the election to the 30th Dáil could have been conducted on the basis of new constituencies. For that reason it does not appear to me to be appropriate to declare that the Oireachtas is in breach of its constitutional obligations on the facts of this case.

8.3 However, I do feel that a number of additional comments are called for. Firstly I have emphasised the urgent nature of the obligation on the Oireachtas. It seems to me that that urgent obligation in turn calls for a consideration of whether measures can be put in place to minimise the gap between the availability of census figures and the enactment of legislation.

8.4 It has become abundantly clear from the evidence given in this case (if it was not clear before now) that, for practical purposes, it is almost certain that the preliminary report figures can be treated as being highly likely to be very close to the figures which will ultimately be determined as the final figures (at least at the level of figures needed to properly formulate constituencies). On that basis there does not seem to be any reason why, as was cogently argued by counsel on behalf of Mr. Molloy, much of the preliminary work geared towards the formulation of new constituencies could not be conducted after the publication of the preliminary figures and before the publication of the final figures. While the findings which I have made earlier in the course of this judgment make it clear that, in determining the constituencies which are actually enacted into law, the Oireachtas is required only to work on final figures, that does not mean that the preliminary work required to formulate a view as to the appropriate constituencies and to engage in significant public consultation as to those constituencies, could not be conducted on the basis of preliminary figures with a final determination awaiting the publication of the final figures. On the basis of the evidence, it seems highly improbable that any significant fine tuning would need to be done unless there happened to be (contrary to recent experience) a very significant variation between the preliminary figures and the final figures.

8.5 Given the urgent nature of the obligation on the Oireachtas to ensure that the period between the publication of a new census disclosing a population disparity in the constituencies and the enactment of new constituencies to deal with that disparity, is kept to a minimum, it seems to me that the Oireachtas should give urgent consideration to whether it would be appropriate to amend the 1997 Act to enable the processes set out in that Act to commence (though not complete) after the publication of a preliminary report. There does not, on the evidence presented, appear to be any good constitutional reason why such a process could not be adopted. The constituencies that would ultimately be put in place would be based on final census figures. The fact that a significant portion of the preliminary work necessary to formulate those constituencies might have been conducted prior to the publication of the final figures would not, it seems to me, give rise to any constitutional infirmity. Obviously if the final figures revealed a significant deviation from the figures published in the preliminary report then it might well be that the process would take some significant additional time. In those circumstances the persons or body charged with formulating the new constituencies (under current law the Constituencies Commission) would have to reassess matters in the light of the new figures. However, in the much more likely event that, as previous experience demonstrates, the final figures are very close to the preliminary figures, then it would seem unlikely that any significant adjustments would be required after the publication of those final figures.

8.6 In those circumstances it seems clear that new constituencies could be determined upon and recommended to the Oireachtas within a very short timescale of the publication of final figures and, undoubtedly, within a much shorter timescale than is currently necessary under the 1997 Act when the process is only triggered by ministerial order subsequent to the publication of the final figures. Given what I have found to be an urgent constitutional imperative on the Oireachtas to deal with any disparity as a matter of urgency then, in the absence of any significant countervailing factor, it seems to me that there may well be an obligation on the part of the Oireachtas to consider whether there are more expeditious ways (such as those which I have outlined) of formulating new constituencies, while at the same time adopting a process designed to assure proper public confidence in the determination of the appropriate constituency boundaries.

8.7 In the circumstances of this case I am not satisfied that it would be appropriate to conclude that the Oireachtas has failed in its constitutional obligation. I do not, therefore, propose making a declaration in those terms. However it does appear to me to be a case in which it is appropriate to adopt the position taken by the Supreme Court in MacMenamin v. Ireland [1996] 3 I.R. 100 in which the courts view as to the general constitutional obligations which arise are set out and the Oireachtas is invited to take whatever detailed measures it might consider appropriate to deal with the issue which has arisen. The precise methods to be adopted in the formulation of new constituencies is, of course, a matter for the Oireachtas. The only role of the court is to intervene if the methods adopted are in breach of the constitutional obligations of the Oireachtas. For the reasons which I have indicated, I am not satisfied that that position has been reached. However it seems to me to be clear that if, without justifiable reason, the Oireachtas did not take appropriate steps to ensure the minimum delay between the finalisation of the ascertainment of the population in a census and the determination and enactment of a law providing for new constituencies, then it might be appropriate for the court to take further action.

8.8 For the reasons analysed in some detail in O’Donovan and also, at least in this regard, affirmed by the Supreme Court in The Reference, there is a high constitutional value to be attached to ensuring that each election, conducted in accordance with the Constitution, affords proportional representation to each part of the country by reference to its population. That high constitutional value can only be met by ensuring that the quickest possible response is made to any ascertained alteration in the population, and its distribution, consistent only with ensuring that there is in place an appropriate process to sustain public confidence in the constituency revision model.

8.9 Having concluded that there is no established failure to comply with the constitutional obligation placed on the Oireachtas it is now necessary to turn to the separate argument made on behalf of Mr. Molloy.

9. The Minister’s power
9.1 Under this heading it is suggested that the Minister can exercise the power conferred on him by s. 164 of the Electoral Act, 1992, (“the 1992 Act”) to be to deal with what is said to be the emergency situation which has arisen. As indicated earlier Mr. Molloy is a constituent in the Dublin West constituency which is the most underrepresented constituency in the country with a variance of 21.2% above the national average in terms of its ratio of population to seats. That that constituency is significantly out of line cannot be doubted.

9.2 The provisions of s. 164(1) are as follows:-

      “The Minister may, in any case in which it appears to him that there is an emergency or a special difficulty, by order make such adaption or modification of any statute order, or regulation relating to the registration of Dáil electors, presidential electors, Seanad electors, European electors or local government electors or the conduct of Dáil elections, presidential elections, Seanad elections, European elections, local government, elections direct elections under Part V of the Local Government Act, 2001 or referenda as may in his opinion be necessary to enable such registration to be duly carried out or such election or referendum to be duly held subject to compliance with the principles laid down in the relevant Acts taken as a whole”.
9.3 Orders made under the section must be laid before each House of the Oireachas as soon as may be after the making of the order and are subject to annulment without prejudice to the validity of anything previously done under the order (see s. 164(2)). The question which arises under this discrete aspect of Mr. Molloy’s case is as to whether the Minister had, in fact, the power to “cure” the difficulty created by the extreme disparity in the case of Dublin West by, for example, the simple expedient of adding an additional deputy to that consistency.

9.4 Before going on to consider the legal issues which arise in relation to this aspect of the case, a brief comment on the facts is appropriate. If an additional deputy had been added to Dublin West, then that constituency would have moved from a position as the most underrepresented constituency to one where it would, in fact, have been one of the more overrepresented constituencies. It is true to say that the degree of its variance from the national average would have been reduced but it still would have been significantly out of line with that national average. The point was also made in the course of the evidence that it would have been possible to apply a similar regime to Dublin North which was the second most underrepresented constituency. Similar considerations would have applied in its case as well. All in all, I am not, therefore satisfied, on the facts, that even if it were to be legally permissible in theory, there is demonstrated to be a factual basis for adopting the measure suggested.

9.5 Indeed the very difficulties thrown up by those facts demonstrate the problems, at a purely practical level, of adopting measures of the type suggested. It is clear on all the evidence that, in putting together a set of coherent and proportionate constituencies, there are, necessarily, knock on effects from one constituency to its neighbours. In that context it is very difficult, in many cases, to introduce an alteration in one constituency without also having to consider alterations in neighbouring constituencies. Frequently, it would appear, a Constituencies Commission (or indeed anyone else charged with the task of formulating constituencies) will have to consider whether it is appropriate to add or take away an area from a constituency so as to bring it into an appropriate level of proportionality. Either course of action may be mandated even in the case of underrepresented constituencies. The two constituencies which were the centre of debate in this case illustrate that fact. To produce properly proportionate constituencies based on the existing Dublin West and Dublin North, it would, in all probability, be necessary either for those constituencies to cede some territory to a neighbouring constituency and keep its existing level of representation with a reduced population or have areas ceded to it by a neighbour and increase the level of representation by one. This fact simply illustrates the interlocking nature of the formulation of appropriate constituencies. The addition of a deputy to one or more constituencies is an extremely crude way of addressing the problem.

9.6 In any event I am not satisfied that the Minister has a power of the type argued for on behalf of Mr. Molloy. Article 16.2.1 of the Constitution provides that constituencies are to be determined by law. The Constitution therefore delegates the power of determining constituencies to the Oireachtas. An inherent part of the determination of constituencies is to determine the number of deputies to be elected from them. That too, is, therefore, a matter which is delegated to the Oireachtas. It does not appear to me that the Oireachtas has the competence to sub delegate that power to any other person or body. Whatever may be the extent of the power conferred on the Minister under the section to amend statutory provisions in emergency cases, it does not seem to me that it can apply, at all, to interfering with the constituency boundaries or the number of deputies provided for each constituency as set out in express terms in the Electoral Acts. That is a power confined to the Oireachtas and it must be exercised by the Oireachtas and the Oireachtas alone.

9.7 I am not, therefore, satisfied that the Minister does have any power of the type asserted. If he had such a power then it is difficult to see any reason, in principle, why its exercise would be confined to the relatively straightforward surgery recommended in this case of the addition of a deputy in one or two constituencies. Why could it not, in principle, apply equally to the creation by the Minister of an entirely new set of constituencies? It is manifestly clear that the best solution to the constituency disparity which was revealed by the 2006 census is the creation of new constituencies which meet the requirements of constitutional proportionality and are consistent with the other criteria specified in The Reference. If the section had the meaning suggested for it on behalf of Mr. Molloy, why would the Minister be confined to imposing a crude, and on any view, only marginally effective solution when a much better solution would be present in the form of a complete revision? However such a consideration brings into stark relief the contention made on behalf of Mr. Molloy. It is wholly inconsistent with the constitutional scheme, set out in Article 16, that the Minister (as opposed to the Oireachtas) should have the task of drawing constituencies. If the Minister cannot draw up constituencies then why should he be allowed to play around with them. The consequences for confidence in the electoral system of it being perceived that a Minister (or more likely a Minister’s party) had benefited by the exercise of such a power hardly needs to be explained. On the facts it would seem most likely that, had the Minister got the power to increase the number of representatives in Dublin West by one, and had exercised that power, it would, as it happens, have been almost certain that Deputy Higgins of the Socialist Party would have been the beneficiary. If, however, the facts were different and it were to transpire that a deputy of the Minister’s party were to be the potential beneficiary, then the Minister’s exercise of the power would, not surprisingly, be treated by many of the electorate with extreme scepticism. It is possible to envisage circumstances (not too far removed from the parliamentary arithmetic which has emerged in the 30th Dáil) where the exercise of such a power by a Minister favouring the creation of an extra seat which was in turn won by a deputy of the Minister’s party, might have a significant effect on government formation. The consequences for legitimate public confidence in the electoral process of such situation do not need emphasis.

9.8 In all the circumstances I am not satisfied that the Minister has the claimed power. Even if I had been so satisfied, it is clear that the question of whether the Minister had or did not have that power was not raised, in any meaningful way, in the correspondence which preceded the issuing of proceedings by Mr. Molloy. While there is authority for the proposition that this court can grant declarations as to ministerial power in circumstances where the Minister concerned has expressly indicated that he or she does not believe that they have a power (see the judgment of Costello J. in Sherwin v. Minister for the Environment [2004] 4 I.R. 279 which relates specifically to s. 164), that jurisdiction arises in a case where the Minister has been asked to exercise a power and declines to do so on the basis of an allegedly mistaken belief that the Minister does not have the power in the first place. The Minister was not asked to exercise any power under the section in this case. He could not be said, therefore, to have declined to exercise the power on any basis, let alone a basis of an allegedly mistaken belief as to the absence of the power concerned. In those circumstances I am not satisfied that I would, in any event, have had jurisdiction to make a declaration of the type sought on the facts of this case even if I had been satisfied that the relevant power existed.

9.9 For all of those reasons I am not satisfied that the additional issue raised on behalf of the Mr. Molloy leads to any positive conclusion from the point of view of the plaintiffs in these proceedings.

10. Conclusion
10.1 It follows that both sets of proceedings must be dismissed. That dismissal should, however, be seen in the context of the comments which I have made concerning the urgent obligation on the Oireachtas to deal with the disproportionality which has now emerged and the additional comments which I have made about the possible need to adopt further measures to ensure that the time lag between the final ascertainment by census of the population and its distribution, and the enactment of new constituencies to reflect that population distribution, is reduced to an absolute minimum.

Back to top of document