BAILII is celebrating 24 years of free online access to the law! Would you consider making a contribution?
No donation is too small. If every visitor before 31 December gives just £1, it will have a significant impact on BAILII's ability to continue providing free access to the law.
Thank you very much for your support!
 |
[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback] | |
High Court of Ireland Decisions |
||
|
You are here: BAILII >> Databases >> High Court of Ireland Decisions >> In the Matter of the Referendum on the Proposal for the Amendment to the Constitution contained in The Thirty First Amendment to the Constitiution (Children) Bill 2012, Held on 10th November 2012 [2013] IEHC 458 (18 October 2013) URL: http://www.bailii.org/ie/cases/IEHC/2013/H458.html Cite as: [2013] IEHC 458 |
||
[New search] [Help]
Judgment Title: In the Matter of the Referendum on the Proposal for the Amendment to the Constitution contained in The Thirty First Amendment to the Constitiution (Children) Bill 2012, Held on 10th November 2012 Neutral Citation: [2013] IEHC 458 High Court Record Number: 2012 152 IA Date of Delivery: 18/10/2013 Court: High Court Composition of Court: Judgment by: McDermott J. Status of Judgment: Approved | ||||||||||||||||||||||||
Neutral Citation Number: [2013] IEHC 458 PETITION THE HIGH COURT [2012 No. 152 IA] IN THE MATTER OF THE REFERENDUM ON THE PROPOSAL FOR THE AMENDMENT OF THE CONSTITUTION CONTAINED IN THE THIRTY FIRST AMENDMENT OF THE CONSTITUTION (CHILDREN) BILL 2012, HELD ON 10TH NOVEMBER, 2012
JOANNA JORDAN INTENDED PETITIONER JUDGMENT of Mr. Justice McDermott delivered on the 18th day of October, 2013 1. On 10th November, 2012, a Referendum was held whereby a proposal for the deletion of Article 42.5 of the Constitution and the insertion of a new Article 42A as contained in the Thirty First Amendment of the Constitution (Children) Bill 2012, was submitted for the decision of the people. 2. The Referendum Returning Officer published a provisional Referendum Certificate in Iris Oifigiúil dated 12th November, 2012, on 13th November which confirmed the final results of the Referendum. This certificate was prepared from the reports submitted by the several local Returning Officers in all of the constituencies in the State which set out the following overall result:-
(b) The total number of votes recorded at the Referendum against the proposal was [445,863] 3. A majority of the votes recorded at the Referendum was recorded in favour of the proposal.” Amending the Constitution
5. The law regulating the holding of a Referendum is set out in the Referendum Act 1994. 6. Voting in the Referendum is by secret ballot and s. 8 of the Act provides that:-
7. On 18th September, 2012, the Taoiseach announced in Dáil Éireann that voting on the Children Referendum would take place on 10th November, 2012. The Bill completed its passage through both Houses of the Oireachtas without amendment on 3rd October. On 8th October, the Minister for the Environment, Community and Local Government made an order pursuant to s. 10(1) of the Act nominating 10th November, 2012, as the polling day. 8. Under s. 14 of the Act, the Minister appointed a “Referendum Returning Officer” whose duty it was to conduct the Referendum and to ascertain and declare the result thereof in accordance with the Act. Under s. 15 a person who would normally be the Returning Officer at a Dáil election in a constituency is empowered to act as “the Local Returning Officer” in that constituency for the purposes of the Referendum. It is the duty of the Local Returning Officer to do “such acts and things as may be necessary for effectually taking the poll and counting the votes in the constituency in accordance with this Act”. Section 18 provides that for the purpose of taking the poll the State shall be deemed to be divided into the same constituencies as those applicable to an election to Dáil Éireann and that “the poll shall be taken separately in each such constituency”. Section 37 provides that:-
(a) The number of valid votes recorded in favour of the proposal which is the subject of the Referendum, (b) The number of valid votes recorded against that proposal, and (c) The total number of valid votes recorded at the Referendum in the constituency,” 9. Under s. 40, the Referendum Returning Officer, as soon as he/she has received the s. 34 reports recording the number of votes recorded in each constituency must prepare and sign the provisional certificate in the prescribed form from these reports stating:-
… (c) In every case the number of votes reported by the Local Returning Officer to have been recorded in each constituency in favour of the proposal which is the subject of the Referendum and the number of votes similarly reported to have been recorded in each constituency against the proposal.” 11. In these proceedings Ms. Jordan (hereinafter the petitioner) seeks leave to present a petition pursuant to s. 42 of the Referendum Act 1994, in respect of the provisional Referendum Certificate dated 12th November, 2012, for an order annulling the Referendum result. The proposed respondents are the Minister for Children and Youth Affairs, the Government of Ireland, Ireland and the Attorney General (the respondents). The application is based on the ruling of the Supreme Court in McCrystal v. Minister for Children and Youth Affairs & Ors [2012] IESC 53 delivered on 8th November, 2012, and the several judgments of the court delivered on 11th December. It was established in McCrystal that the information campaign sponsored by the Minister for Children and Youth Affairs in respect of the Children Referendum in advance of the poll constituted a clear disregard of the rights of citizens to a Referendum conducted in accordance with the norms of the democratic process mandated by the provisions of the Constitution. McCrystal Litigation 13. On 19th October, 2012, the Minister for Children and Youth Affairs also commenced an information campaign. The Minister caused a website to be launched, conducted an advertising campaign on television, radio and in the printed media, and produced a booklet said to be for the purpose of informing the electorate about the Referendum proposals. 14. Mr. McCrystal was concerned that the Government’s information campaign constituted a clear disregard of the principles established in McKenna v. An Taoiseach (No. 2) [1995] 2 IR 10 (the McKenna case), which determined that the Government may not spend public monies to promote a result in a Referendum. 15. In McKenna (No.2), Hamilton C.J. in his judgment stated that the Referendum Act 1994, in accordance with which every proposal for constitutional amendment must be put to the people, did not allocate any role to the Government in furnishing information to the electorate or in the conduct of the Referendum. The Government in spending public funds on the promotion of a campaign in favour of a “Yes” vote was not acting in accordance with the executive power of the State. Though the Government was entitled to express its views and urge acceptance of the proposal, the issue was whether the expenditure of public funds to that end constituted an interference with the plaintiff’s constitutional rights. Hamilton C.J. was satisfied that:-
I should think it bordering on the self-evident that in a democracy such as is enshrined in our Constitution (which is not exclusively a parliamentary democracy; it has elements of a plebiciary democracy) it is impermissible for the Government to spend public money in the course of a referendum campaign to benefit one side rather that the other.” (p. 43)
This is an implied right pursuant to Article 40, s. 3 which harmonises with Article 5, Article 6, s. 1, Article 16, Article 40, s. 1, Article 47, s. 3 and is in keeping with the democratic nature of Bunreacht na hÉireann… Power derives from the People, and is exercised under the Constitution through their organs of government (legislative, executive, judicial). Power and decision-making in referenda is with the People. The organs of government are instruments of the People. Thus, the democratic process is fundamental and critical to the exercise of power under the Constitution.” (pp. 53 – 54)
2. If, however, the Government acts otherwise than in accordance with the provisions of the Constitution and in clear disregard thereof, the courts are not only entitled but obliged to intervene. 3. The courts are only entitled to intervene if the circumstances are such as to amount to a clear disregard by the Government of the powers and duties conferred on it by the Constitution.” 20. The Supreme Court concluded that the government funding of the campaign designed to influence voters in favour of a “Yes” vote was an interference with the democratic and constitutional process for the amendment of the Constitution and infringed the concept of equality which was fundamental to the democratic nature of the State. The court granted a declaration encapsulating these findings but declined to issue an injunction. 21. Mr. McCrystal initiated his proceedings, by way of ex parte application to the High Court on 19th October, 2012, in respect of the expenditure of public funds by the Minister for Children and Youth Affairs and the Government in the course of the Children Referendum Campaign based squarely on the McKenna case. He challenged the expenditure of public money on a government information booklet, advertising on radio and television and in printed media, and the creation and maintenance of a website favouring a “Yes” vote in the Referendum. He complained that the Minister had commissioned 2.05m copies of an “information booklet”, the delivery of which to all homes in the State commenced on 19th October, 2012. In the course of Mr. McCrystal’s proceedings, evidence was adduced that the department allocated a budget of €3m to be spent on the Children Referendum. €1.9m was allocated to the Referendum Commission to finance the discharge of its statutory duties. The balance of €1.1m was to be used by the Department of Children and Youth Affairs: “to provide information on the Referendum and encourage members of the public to vote”. Details were provided to this Court of this expenditure as follows:-
This expenditure enabled the government to disseminate the impugned material nationwide and was a large amount compared with the sums spent by others, such as Fine Gael €150,000.00 and the Labour Party €50,000.00. The “no” campaign had very little funding available to it. 22. Affidavits were submitted to the High Court on behalf of Mr. McCrystal supporting his contention that the materials produced by the Government tended to promote a “Yes” vote. He sought a declaration that the defendants acted wrongfully and were not entitled to pursue a particular result under the guise of providing information, an injunction restraining the defendants from promoting a particular result and consequential orders in respect of the ongoing distribution of the booklets, the use of the website and the advertising campaign. 23. The High Court (Kearns P.) in an ex tempore judgment delivered on 1st November, 2012 [2012] IEHC 101, and having considered an extensive body of expert and other evidence submitted on affidavit, was not satisfied to conclude that the material constituted “a clear constitutional abuse or manifest solicitation to vote in a particular way”. 24. The plaintiff appealed to the Supreme Court which commenced the hearing of the appeal on 6th November, 2012, and on 8th November granted a declaration that the various publications including the booklet, the advertising and the material on the website “in places” breached the principles set out in the McKenna case. An injunction was not granted because it was assumed by the court that the Minister and the Government would respect the Supreme Court’s ruling and cease its unconstitutional behaviour. The Supreme Court delivered its decision promptly as the Referendum was due to be held on the following Saturday, 10th November. The court indicated that the judgments would be delivered on 11th December, 2012. The Supreme Court ruling was extensively and immediately reported in the national media and became a source of intense political debate, understandably attracting significant criticism from the “No” campaign and others. It is important to quote the relevant part of the ruling delivered on 8th November because of the claims made by both sides as to its affect or potential affect on the Referendum campaign and on votes cast on polling day two days later. 25. Denham C.J. in delivering the ruling of the court stated:-
The people adopted the Constitution 75 years ago. The Constitution belongs to the people and may be amended only by the people in a Referendum. It is this democratic process which is protected by the McKenna principles. Public funding should not be used in a Referendum to espouse a particular point of view. 4. The McKenna principles may be found in the several judgments in that case. These principles, which are not in dispute, are consistent with standards recognised both nationally and internationally for a Referendum process, such as the European Commission for Democracy Through Law (Venice Commission), Code for Good Practice on Referendums, adopted by the Council for Democratic Elections at its 19th Meeting (Venice, 16 December, 2006) and the Venice Commission at its 70th Plenary Session (Venice, 16 – 17 March, 2007). 5. At issue in this case is the application of these principles to a booklet and a website, both entitled ‘Children’s Referendum’, and advertisements, published and disseminated by the Department of Children and Youth Affairs, on foot of moneys voted by the Oireachtas, which the appellant submits breach the McKenna principles. 6. The Court is required to give its decision promptly, in view of the pending Referendum to be held on Saturday, 10th November, 2012. The substance of that proposal is a matter for the people alone. The Court will give its ruling today and judgements will be delivered on Tuesday, 11th December, 2012. 7. The Court has concluded that it is clear that there are extensive passages in the booklet and on the website which do not conform to the McKenna principles. This material includes a misstatement, now admitted to be such, as to the effect of the Referendum. 8. The Court is satisfied that while not all of the website or the booklet are in breach of the McKenna principles, because of the overall structure of the booklet and website, it would not be appropriate for the Court to redact either. 9. Accordingly, the Court would grant a declaration that the respondents have acted wrongfully in expending or arranging to expend public moneys on the website, booklet and advertisements in relation to the Referendum on the Thirty First Amendment of the Constitution (Children) Bill, 2012, in a manner which was not fair, equal or impartial. The Court does not consider it either appropriate or necessary to grant an injunction, as it is to be assumed that the respondents will cease distributing and publishing the material.” 27. The judgments of the Supreme Court when delivered on 11th December, 2012, contained detailed consideration of the contents of the Government Booklet that had been distributed throughout the State, the website operated by the Minister and the advertising campaign. The various judgments set out detailed reasons for the conclusion that the Minister and Government acted in “clear disregard” of the McKenna principles and highlighted by a close examination of its contents why the Government’s “information” campaign was found to be partial and to favour a “Yes” vote. Of course, that forensic examination of the material was not available to the electorate prior to polling day which the petitioner in this case claims to be a matter of some importance. Whilst it was and remains permissible for the Minister and the Government to spend public money on the dissemination of information, it was not permissible for them to favour a particular outcome or, under the guise of an information campaign to depart from a path of “strict neutrality”. 28. It is clear from the McKenna and McCrystal decisions that if a citizen establishes to the court’s satisfaction on the balance of probabilities that the Minister and/or the Government acted in clear disregard of the constitutional rights of the citizen to a democratic process incorporating the right to equality, fair procedures and freedom of expression, the citizen may apply for and secure judicial intervention in order to protect and vindicate those rights. The Supreme Court judgments recognise the vulnerability of the democratic process and the exercise of those constitutional rights which underpin its strength, integrity and effectiveness, to significant damage if the executive is permitted to spend public money in advancing a “Yes” vote in the course of a Referendum campaign. These proceedings are concerned with the effect, if any, of that unconstitutional behaviour on the poll held on 10th November, 2012, and whether it was such as to “materially affect” the result of the Referendum “as a whole”. The Claim
(2) Use of public funding in a Referendum to espouse a particular point of view also results in a violation of standards, which are recognised nationally and internationally, for a Referendum process. (3) Prior to the Referendum, a booklet (the Booklet) and a website (the Website) (together “the Campaign Material”) both entitled “Children’s Referendum” – and advertisements (“Advertising”), were published and disseminated by or on behalf of the Government and/or the State, on foot of monies voted by the Oireachtas. (4) The Booklet was widely distributed to homes in the jurisdiction. In the premises…given its widespread distribution and readership, the Booklet was such as to affect materially the result of the Referendum as a whole and/or affected the result of the Referendum and/or may have affected the result of the Referendum. (5) The Website was widely viewed during the period. Further and without prejudice to the foregoing…given its widespread readership, the Website was such as to affect materially the result of the Referendum as a whole and/or affected the Referendum and/or may have affected the Referendum. (6) There was widespread awareness of the Advertising. Further and without prejudice to the foregoing…given widespread awareness of it, the Advertising was such as to affect materially the result of the Referendum as a whole and/or affected the Referendum and/or may have affected the Referendum. (7) The Booklet, the Website, and the Advertising espouse the view that voters should vote in favour of the proposed amendment. (8) Following a legal challenge, on 8th November, 2012, the Supreme Court ruled that it was clear that there were extensive passages in the Booklet and on the Website, which were not in accordance with the requirements of the Constitution. The Supreme Court also concluded that said Campaign Material included a misstatement as to the affect of the Referendum (together “the Breaches of the Constitution”). (9) The Supreme Court granted a declaration in respect of the Breaches of the Constitution to the effect that the Minister for Children and Youth Affairs, the Government of Ireland, Ireland and the Attorney General:
(11) Notwithstanding the Supreme Court’s declaration, the defendants proceeded to hold the Referendum as scheduled on 10th day of November, 2012.” 31. The petitioner also relied upon expert evidence to support the proposition that an independent poll carried out after the vote on behalf of the Referendum Commission provided significant evidence that the information campaign had a “material affect” on the Referendum poll result. 32. In addition the petitioner claimed that there was no basis to conclude that the ruling of the Supreme Court on the 8th November, 2012, had any mitigating influence on the effect otherwise caused by the Government information campaign and indeed, that the various Government members who commented upon this ruling tended to undermine its importance and potential affect. The petitioner also contended that the post Supreme Court ruling behaviour of the respondents exacerbated the affect of the unconstitutional conduct, thereby further interfering with the conduct of the Referendum and the democratic process and that it had a further material affect on the Referendum result. The Respondents’ Case 34. The respondents relied upon the McCrystal Supreme Court ruling on the 8th November, 2012, as a factor which had a negative impact against the Government and was widely disseminated through the various organs of the media to the benefit of the “No” campaign. Leave to Present a Petition
(2) A Referendum petition in relation to a provisional Referendum Certificate shall not be presented to the High Court unless that court, on application made to it in that behalf by or on behalf of the person proposing to present it not later than seven days after the publication in Iris Oifigiúil of the Certificate, by order grants leave to the person to do so. (3) The High Court shall not grant leave under subsection (2) to present a Referendum petition unless it is satisfied – (a) that there is prima facie evidence of a matter referred to in section 43 in relation to which the Referendum petition questions the provisional Referendum Certificate concerned, and (b) that the said matter is such as to affect materially the result of the Referendum as a whole. (4) An application for leave to present a Referendum petition may be made by the Director of Public Prosecutions or by any person who is registered or entitled to be registered as a presidential elector.” 37. It was submitted on behalf of Ms. Jordan that the section provides a low threshold for the granting of leave. Reliance was placed upon the judgment of Barrington J. in Hanafin v. Minister for the Environment [1996] 2 IR 321, in which a petition was brought in respect of the result of the Divorce Referendum (a proposal contained in the Fifteenth Amendment of the Constitution Bill (No.2) 1995) following the determination in the McKenna case that the Government was in breach of the Constitution in the conduct of an information campaign conducted in the lead up to that Referendum poll. Leave was granted by McCracken J. in the High Court to present the petition on 7th December, 1995. No issue arose concerning the grant of leave in the Hanafin case. However, in his judgment in the Supreme Court, Barrington J. stated at p. 456:-
39. I am satisfied that in order to grant leave to present the petition in this case the court must be satisfied that there is prima facie evidence adduced which, if accepted and, in the absence of any other evidence, would enable the court to conclude that the unconstitutional conduct may have occurred and that the affect of the breaches of the Constitution of which complaint is made may be such as to have affected materially the result of the Referendum as a whole. The court must be satisfied that the petitioner has established on the evidence advanced in support of the ground upon which leave is sought that there is a fair bona fide or serious issue to be tried. It is not necessary, at this stage, that the court be satisfied of the petitioner’s case on the balance of probabilities: that is the level of satisfaction appropriate to the trial of the petition. If leave is granted, as Barrington J. notes, the petitioner must produce her witnesses and the court must then determine whether it is satisfied to the appropriate level of proof that the petitioner has succeeded in establishing her contention and is entitled to the relief claimed. The threshold established under s. 42 provides a filtering system whereby vexatious or frivolous claims or those based on unarguable, unstateable, or insubstantial grounds or weak or inadequate evidence will be refused. 40. In this case, following the adjournment of the leave application, affidavits were exchanged including affidavits from various expert witnesses in respect of the alleged effect or non-effect of the McCrystal breaches of the Constitution on the Referendum result. It was agreed between the parties that the same evidence grounding the application for leave to present the petition would have to be considered by the court on the full hearing of the petition if leave were granted. At a preliminary hearing it was submitted that, subject to the view of the court, a telescoped hearing of the leave application together with any substantive application necessitated by any grant of leave would afford the parties full opportunity to canvas the relevant issues before the court and enable the court to deal with all such issues in an administratively efficient and fair manner. Dunne J. gave directions to that effect which I adopted when the matter came on for hearing before this Court. 41. I was satisfied, at the request of the parties, to leave the resolution of the application for leave to present the petition until all of the evidence and submissions which the parties wished to tender in relation to the issues in the case were heard. In that regard, it is clear as a matter of fact that the Supreme Court has already determined, that the Minister and the Government acted in clear disregard of the Constitution in the lead up to the Referendum poll, so that there was clear prima facie evidence of the existence of an interference with or irregularity in the conduct of the Referendum under section 43. It remains for the court to consider at the leave stage whether it is satisfied that there was also prima facie evidence that the unconstitutional conduct established “is such as to affect materially the result of the Referendum as a whole”. I will return to this issue later in the judgment when considering the evidence in the case. It is appropriate, before considering the evidence adduced to consider the onus and burden of proof applicable to the trial of a petition and the meaning of “material affect”. The Hanafin Case and s. 43 of the Referendum Act 1994 43. Section 43 sets out the grounds upon which a petitioner may question the validity of a provisional Referendum Certificate as follows:-
(a) the commission of an offence referred to in Part XXII of the Act of 1992 (as applied by section 6 ), (b) obstruction of or interference with or other hindrance to the conduct of the Referendum, (c) failure to complete or otherwise conduct the Referendum in accordance with this Act, or (d) mistake or other irregularity in the conduct of the Referendum or in the particulars stated in the provisional Referendum Certificate.” 45. It was held by the Supreme Court that the only way in which the result of a provisional Referendum Certificate could be challenged was by way of a petition based on the grounds set out in s. 43(1) of the 1994 Act. In construing the section, the court held that the words “conduct of the Referendum” included unlawful conduct on the part of the Government in its Referendum campaign and the consequences thereof. The court rejected a narrow construction of the section based upon the submission that it precluded the High Court from questioning the validity of a provisional Referendum Certificate on the ground that there had been constitutional wrongdoing by the executive because the section limited the courts jurisdiction to grant relief to grounds based on the manner or procedure by which the poll was conducted. Hamilton C.J. stated:-
The Court must, however, presume that the Oireachtas did not so intend but intended that such rights should be defended and vindicated. The defence and vindication of such rights requires that the words “the conduct of the Referendum” contained in s. 43, sub-s. 1 (b) and (d) should be interpreted sufficiently widely to include unlawful and unconstitutional conduct in the Referendum campaign which materially affected the result of the Referendum. There is nothing in the other provisions of the Act which prevents this construction, which construction accords with the provisions of the Constitution. I accept as being correct the statement made by Barr J. when dealing with this issue where he stated that:- ‘It seems to me that the fundamental importance of the concept that the will of the people should be properly ascertained in accordance with law in a Referendum on constitutional change requires that the words ‘the conduct of the Referendum’ in s. 43, sub-s. 1 (b) and (d) should be interpreted sufficiently widely to include unlawful conduct on the part of the Government in its Referendum campaign and the consequences thereof which are alleged to have caused an obstruction, interference, hindrance to or irregularity in the conduct of the Referendum of such gravity as to vitiate its apparent result. This is an issue which the petitioner puts before the Court and on which, in my view he is entitled to a decision.’” 46. Hamilton C.J. also stated that the jurisdiction of the High Court to intervene with the Referendum result existed:
“Material Affect” and the Onus and Standard of Proof
Consequently, the onus lay on the petitioner to establish on the balance of probabilities:- (1) The nature and extent of the obstruction of or interference with or other hindrance or mistake or other irregularity (hereinafter referred to in the circumstances of this case as “constitutional wrongdoing”), and (2) That such “constitutional wrongdoing” materially affected the result of the Referendum as a whole.
(b) Any consideration of “material affect” necessarily follows a determination that there has been unconstitutional wrongdoing such as to amount to an interference with the conduct of the Referendum and that the logic of the Act demands that the concept of material affect be understood as equivalent to showing that the interference or wrongdoing was not trivial or inconsequential and not a separate matter to be established with almost mathematical certainty by a criminal standard of proof.
Sections 42, 43 and 48, subsection (2) of the Act of 1994 refer to this requirement… From a consideration of these subsections of the Act, it is clear that the Act provided and intended that the result of the Referendum as a whole could only be questioned if it was established to the satisfaction of the court that the result was materially affected by the alleged wrongdoing. The onus of so establishing rests on the petitioner who questions the result of the Referendum. This is not only required by the Act but is in accord with the constitutional right of the citizens to vote in a constitutional Referendum and to have the result thereof accepted, respected and not interfered with unless it is established that such result was materially affected by alleged wrongdoing of such a nature and affect as to vitiate the Referendum.” (pp. 426 – 427)
50. O’Donnell J. added (at para. 40):-
51. It is clear from the McKenna (No.2) and McCrystal cases that the onus was on the plaintiffs in each case to establish that the executive had acted in clear disregard of the McKenna principles on the balance of probabilities. Barrington J., in Hanafin, in considering the onus and standard of proof applicable to proof of a “material affect” on the Referendum result stated:-
At this stage (post leave) the petitioner has to attack a provisional Referendum Certificate purporting to record the decision of the people at a Referendum. The situation is not unlike that which exists when the President refers to this Court a Bill which has been passed by both Houses of the Oireachtas. The court pays the Oireachtas the courtesy of assuming that it has not violated the Constitution. It, therefore, presumes that the Bill is not repugnant to the Constitution until the contrary is clearly established…likewise, this Court will not lightly set aside what appears, prima facie, to be an Act of the sovereign people. Unless, therefore, what has happened is an express and obvious constitutional abuse affecting the outcome of the Referendum, the onus of proof on the petitioner will be a heavy one. This does not mean that the onus is higher than the civil onus of proof but rather that the court will be particularly vigilant in examining serious allegations.” (pp. 456 – 457) 53. O’Flaherty J. noted that:-
An Inquisitorial Aspect?
59. I am, therefore, satisfied that the court is obliged and has ample powers to ensure that the petition process is not abused and that all relevant evidence for the determination of the issues raised on the grounds upon which leave is granted to present a petition is available to the court. These powers are in accordance with the courts overriding obligation to ensure that any issues emerging in the course of the trial related to these grounds are fully explored and considered. “Result of the Referendum as a Whole” Reversal of the Burden of Proof 62. An alternative submission was made that having regard to the fundamental rights at issue and given the seriousness of the breaches of the McKenna principles identified in the McCrystal case, it would be appropriate to place a burden of proof on the respondents to demonstrate that the Referendum result was not affected by the campaign breaches and their conduct post the ruling of the Supreme Court in McCrystal. The petitioner submits that the burden of proof should shift in this case because of the Supreme Court’s finding that the McKenna principles had been breached and thereby deprived the electorate of a constitutional Referendum process. The court is also invited to reverse the onus of proof because the petitioner has difficulties in establishing that the Referendum result as a whole has been affected as a consequence of the unconstitutional conduct. It is submitted that as a matter of fair procedure, the burden of proving that this conduct did not affect the Referendum should fall on the respondents who were the wrongdoers. 63. In Hanrahan & Ors v. Merck Sharp & Dohme [1988] ILRM 629 the plaintiffs claimed damages for nuisance. They claimed to have suffered damage from emissions from the defendant’s factory and had difficulty in establishing that fact, as a result of which they claimed that the onus of proof should be shifted to the defendant. It was held, notwithstanding the difficulties of proof faced by the plaintiffs, that these were not such as would require the shifting of the onus as might be the case in circumstances where a particular matter was within the peculiar knowledge of the defendant. It was also held that the onus of proof placed by the common law on the plaintiffs did not amount to a failure by the state to vindicate their personal rights pursuant to Article 40.3 of the Constitution, since it was not alleged that the tort of nuisance was ineffective to protect their constitutional rights. It was held that the guarantees contained in Article 40.3 were not absolute or unqualified and could not be taken to relieve the plaintiff in a nuisance action of the onus of proving the necessary ingredients of the tort. In that context Henchy J. stated:-
Henchy J. emphasised that the rationale behind the shifting of the onus of proof to the defendant lay in the fact that it would be palpably unfair to require a plaintiff to prove something which is beyond his reach and which is peculiarly within the range of the defendant’s capacity to prove. I am satisfied that whether the unconstitutional conduct “affected materially” the result of the Referendum as a whole, is not a matter which is either within the exclusive knowledge of the respondents or a matter which is peculiarly within their capacity to prove. It is also clear in this case that the remedy sought by the petitioner is not a remedy against the first two respondents whereby they bear the brunt of the relief sought, but a remedy “which seeks to override and reverse the sovereign will of the people as expressed in the provisional Referendum Certificate containing the record of votes cast at the Referendum” (per Hamilton J. in McKenna (No.2) at p. 429). Evidence 66. The non-expert witnesses were the petitioner, Mr. Michael Fitzgibbon, Mr. Nicholas Gargin, Mr. Daniel Ward and Ms. Deirdre Uí Ghoibín. 67. The petitioner took a “no” stance to the proposal and became aware of the government materials in the weeks leading up to 10th November, polling day. She received the Government Booklet ten days before polling at her home. She was also aware of the government advertising. She was concerned that both sources of information were partisan and slanted towards a “yes” vote. The arguments that might be raised against the proposed amendment were not included in the materials. She canvassed for a “no” vote and believed that people had a mistaken understanding of the potential impact of the Referendum proposal. She believed that it was obvious that the booklet, the website and advertising were succeeding in winning “yes” votes. The people she canvassed trusted the government material and were satisfied that it was balanced. They appeared to assume that the government would not act contrary to its obligation to be impartial. She became aware on 8th November of the McCrystal ruling by the Supreme Court. She believed that it came at too late a stage to be useful to the “no” campaign. However, she noticed when campaigning in O’Connell Street in Dublin, in the last two days, that people were more receptive of the “no” campaign, which she assumed was due to the Supreme Court ruling. 68. Nicholas Gargin was a part-time student of Sports Management. He voted “yes” in the Referendum and had read the Government Booklet before voting. He discovered that the Supreme Court ruling had been made during his Christmas holidays. He stated that had he found out about the ruling before polling day, he would have been angry at the government’s behaviour particularly in the light of the judgments of the Supreme Court which he read. Had he known about the government wrongdoing, he would have voted “no” in the Referendum. 69. Mr. Michael Fitzgibbon campaigned for a “no” vote in North Kerry and West Limerick. He spoke to hundreds of potential voters during the course of the campaign. He believed that the “no” campaign faced a “juggernaut” in favour of the proposal, which ranged across the political spectrum and the media. His impression was that people believed government materials had been professionally prepared and were convincing. He believed they had a real influence on voters who expected a non-biased view in the Government Guide. He discerned a huge difference in the campaign after the Supreme Court ruling in that people were coming up to him saying that they would vote “no” and were very suspicious of the government and disillusioned. He acknowledged, however, that at that stage, he was campaigning in an area which he knew would vote “no” in significant numbers. 70. Mr. Daniel Ward, a farmer on Aranmore, Co. Donegal, intended to vote “no”. He read the booklet and was aware of the advertising by the government. He assumed the government would be impartial and neutral in the provision of information. When he read the materials he became uncertain as to whether he should vote no, and eventually decided not to vote at all. Because he lived on Aranmore he was obliged to vote by 7.00pm on 8th November, 2012. He became aware of the Supreme Court ruling at 6.00pm. He decided not to vote presuming in the light of the Supreme Court ruling that the Referendum would be postponed. 71. Ms. Deirdre Uí Ghoibín was also a “no” campaigner. She first saw the Referendum Booklet in October, 2012 and looked at the website. She considered that there was not enough factual information available in the government materials. She believed these materials had an influence, but she could not give details of any specific discussion that she had about the booklet with any voter. She heard about the Supreme Court ruling on the radio. She did not campaign in the last two days before the poll but family and friends with whom she discussed the matter were surprised at the government’s misconduct. 72. Mr. John Waters submitted three affidavits in support of the petitioner. He is an author, journalist and political commentator and has written a number of books on politics. He has been a columnist in the Irish Times for over twenty years. He was a member of the Broadcasting Commission of Ireland from 2003 to 2009 and the Broadcasting Authority from 2009 to date. He has covered all of the elections and referendum campaigns for over thirty years, and gave evidence that he was familiar with techniques of persuasion in public communications and the manner in which government used language in attempts to persuade rather than inform, a knowledge he acquired with experience. He favoured a “no” result in the Referendum. He felt that though the “no” campaign started from a low base there was an opening of the public mind to the “no” arguments in the last ten days of the campaign. He first became aware of the controversy over the government information campaign over the weekend of 20th – 21st October when he was contacted by lawyers for Mr. McCrystal and asked to review the government website and booklet for the purpose of providing a critique for the High Court. That analysis formed part of the evidence considered ultimately by the Supreme Court in the McCrystal judgments. He formed the view that the contents of the booklet failed to have any regard to the arguments against the proposal and failed to present the reality of the existing constitutional jurisprudence in respect of the rights of the child and the family. He was concerned that those who wished to acquaint themselves with the issues might turn to the booklet as an important source of information on the propositions for and against the proposal, which it was not. Instead, the reader was given a highly tendentious view of the affect of the proposal. The booklet was sent to every household in the state and was intended to be read in the quiet of one’s home. It was calculated to override most of the other information which might have offered a general balanced position of the issues in the debate. He was concerned that the Government Booklet acquired a high level of credibility and that voters would be disposed to view it as neutral. It implicitly called for a “yes” vote which, in a sense gave it added strength. Mr. Waters also viewed the website and online advertising contracted by the Department, and found a similar approach. He described the highly emotive presentation of issues in the booklet, but especially on the website in relation to identifying the changes proposed in the Constitution and, in particular, took issue with the suggestion that all sorts of abuses and wrongs suffered by children would be remedied by the Constitution. Anecdotally, he encountered people who had been given this impression. He believed that this expenditure of money was highly effective. He thought the information campaign was the “foundation stone” of the government campaign. He could not quantify its affect, but he believed that to say it was not a central factor in the result would stretch credulity. 73. Each of the witnesses gave useful and honest accounts of their experiences during the course of the Referendum campaign and their impressions of the government materials and information. These impressions were vindicated by the ruling and judgments of the Supreme Court in the McCrystal case. Some of the more generalised conclusions reached by the witnesses about the affect of the materials are highly speculative and subjective. I have considered all of this evidence and it is clear that without something more tangible and objective, this evidence would be entirely insufficient to justify the granting of leave to present a petition or to establish that the overall Referendum result had been materially affected by the government information campaign. However, the petitioner also relied upon expert testimony in respect of the government information campaign. The Government Information Campaign 75. Preparation for the Children Referendum was the responsibility of Ms. Elizabeth Canavan, Assistant Secretary of the Department and a relatively small staff. It is clear from the discovered material and documents presented in the course of evidence that an enormous amount of preparatory work was completed prior to the publication of the wording of the proposed Referendum on 19th September, 2012. It was clear from surveys carried out in 2012 on behalf of the Department that people were associating the general proposal to amend the Constitution with issues with which it was not concerned. The information campaign was the result of a strong government view that, due to complaints about lack of information in previous Referendum campaigns, as reflected in post poll surveys by the Referendum Commission, it was necessary to put forward information to the people explaining the proposal. Ms. Canavan explained that the Department was conscious of the potential for confusion between its material and that of the Referendum Commission because of previous Referendum Commission reports to that effect. It was also acutely aware of the McKenna principles. Those tendering for contracts in respect of the information campaign were obliged to comply with those principles and there was a strict communication protocol which ensured that no draft could be finalised unless signed off by the Assistant Secretary following legal advice from the office of the Attorney General, which was consulted as necessary from time to time. The Booklet 77. The choice of the logo for the booklet was made following the submission of a number of potential logos by an outside company and was extensively discussed within the Department and finally, chosen by the Minister. Unsurprisingly, many of the Department’s publications contain logos of children or photographs or silhouettes of children. There was some concern that the Department’s logo should not be similar to that used by other children’s organisations in their literature. Ms. Canavan told the court that there was no discussion of the use of the words “Children’s’ Referendum” in the booklet. Four headlines were used on the cover of the booklet which related to content within. The communication company retained did not have a lead role in the compiling of the booklet but made suggestions concerning the simplification of the language used in its contents. The booklet was also uploaded onto the website created by the Department in the days following its publication. 78. In the McCrystal judgment the contents of the booklet were found to have advocated a “Yes” vote. Denham C.J. pointed to a number of matters including slogans used in the booklet such as “Protecting Children” on the cover page and at pp. 1, 6 and 14. The phrase “Supporting Families” which she deemed not to be impartial also appeared on the same pages. The question posed “Why do we need a Referendum?” was found to infer a need for the Referendum. The title “Children’s Referendum” in the booklet was juxtaposed with a silhouette that appeared to be three children linking hands. Mr. McCrystal had argued that these visual representations were designed to induce an emotional response and advocated a yes vote in a subliminal fashion rather than being a neutral and objective visual representation. 79. Murray J. found that the material as a whole was characterised by the four slogans contained on the cover of the booklet namely:-
Supporting Families Removing inequalities and adoption Recognising children in their own right”
(2) Protecting children and supporting families; (3) Removing inequalities; (4) Adoption: a second chance for children; (5) Recognising children in their own right.” 80. Fennelly J. having set out the history of the preparation of the various materials, and having outlined in summary the evidence given by Ms. Canavan in the McCrystal case, which were substantially the same as given to this Court, also concluded that the booklet was written with a view to providing support for the objectives of the Referendum proposal. He noted that the booklet was circulated shortly after the opening of the website and it largely repeated the same material, albeit in different language. 81. O’Donnell J. considered the website material to be “undoubtedly the most important” in the case. He noted that:-
82. Denham C.J. and O’Donnell J. gave detailed consideration to an important error contained in the booklet. O’Donnell J. (at para. 15) stated:-
‘It will continue to be the case that the power given by the Constitution in this area can only be used by the State in very well defined circumstances. Key requirements will continue to be underlined as follows:
• If failure of parental duty towards the child must exist – “where the parents regardless of their marital status, fail in their duty towards their children”; • Any failure must involve harm or risk to the child’s safety or welfare – “to such extent that the safety or welfare of any of their children is likely to be prejudicially affected”; • The actions of the State must be in balance with the harm or risk to the child that needs to be addressed – “by proportionate means”; and • The actions the State can take must be set out in law – “as provided by law” (emphasis added) 84. Ms. Canavan in evidence to this Court stated that she believed at the time that the material contained in the “admitted error” had been cleared and that it was not a mistake. She stated that the first time she realised it was an error was on 31st October, when it was conceded in court. It came as a surprise to her that the change had been advised to them and that they had not made it. She described it as a complete shock. When the papers on the file were reviewed, she discovered that she had actually annotated a final version of the booklet herself in the light of legal advice received and had gone through it by hand with a number of other officials and had put a line through the “admitted error”, but this had not been executed. She acknowledged that quicker action should have been taken to correct the error. However, the booklet was in the course of distribution and it was impossible to take back. It simply did not occur to her that the booklet was also in PDF form on the website and though she took steps to have it taken down and corrected, the material was not removed entirely from the site until 9th November. 85. The affect of this booklet on the result of the Referendum formed a core element of the petitioner’s claim. It was submitted that it was state financed in clear disregard of the Constitution and entirely one sided in advocating a yes vote, albeit not explicitly. It was received by every household in the country. It was presented as neutral and therefore to be trusted, when it was not. It was claimed, that a survey compiled on behalf of the Referendum Commission by “Behaviour and Attitudes”, demonstrated that the booklet had a material affect on the Referendum result. Both sides relied upon the evidence of a number of eminent expert witnesses in respect of this issue. 86. Dr. Michael Bruter is a reader in political science specialising in electoral psychology at the London School of Economics and Political Science. He seeks to understand what goes on in people’s minds when entering a polling station and what leads them to cast their vote in a particular way. Dr. Bruter was retained by the petitioner’s solicitor and asked to provide an expert opinion on two matters:-
(ii) whether delaying the Referendum vote following the decision of the Supreme Court in the McCrystal case on 8th November, 2012, would likely have resulted in a materially different result. 88. Dr. Bruter highlighted a number of specific elements contained in the information which had already been considered in the McCrystal judgment but which were in his view, particularly relevant in terms of the likely psychological impact on voters. He identified what he considered to be the four essential elements which went beyond a simple lack of neutrality and clearly leant towards a yes argument. These were:-
(ii) the description of the proposal that explicitly represented it as stemming from and/or legitimised by non-partisan, neutral, and prototypically competent sources and iconography and that implicitly associates its source with children themselves; (iii) the narrative of the proposal as answering a (possibly urgent) need; and (iv) the use of a rhetoric which explicitly focuses on emotions and was likely to induce a possible feeling of guilt amongst those not supporting the proposal.
(b) In every Referendum, a voter’s choice is likely to be influenced by two important elements (i) cues; and (ii) influences. Cues are sources that collect, summarise and report information on the pros and cons of a yes or no vote. The source could be regarded as neutral (the Referendum Commission), providing information without necessarily being neutral (the media) or partisan (political parties or yes or no campaigners). External influences are elements not directly related to the question in issue but which are accepted as having a strong impact on the way people vote, for example, government popularity: political scientists are not agreed on whether “cues” or “influences” matter most. However, Dr. Bruter was of the view that the perception of whether the source of a “cue” was neutral or partisan had a “major effect on the way voters would treat the information that it provides”. A partisan source for a cue would likely be heavily discounted whereas maximum credit will be afforded information from a neutral cue source: it would be regarded as a “trusted or dominant” cue by the prospective voter. (c) The lower the salience of the issues, the more likely the voter is to compensate for this by voting on the basis of more general principles received or adopted through various cues and perhaps by acting on other principles such as disapproval of the government.
(ii) Secondly, the perceived nature of a source is a major determinant of how influential the information it provides will be. Thus, information perceived to come from a “neutral” source would be significantly more influential than information perceived to come from a “partisan” source. The government information was presented as and perceived to be neutral. (iii) Thirdly, in terms of the potentially strong impact of external influences in the context of a low salience referendum, any element which, during the campaign, was likely to give additional credit or discredit to government would have a strong effect on the result.
(ii) International research demonstrates that the major sources of information used by voters in elections and referenda campaigns are the mass media and in particular television and radio. (iii) In the minds of most voters, the government booklet or advertisement, even if published by a department, is one produced by the government, one run by the Fine Gael and Labour parties. It would be wrong to think that most voters would necessarily see such material as objective. He concluded by reference to published authorities and his own experience that government supporters would most likely have paid more attention to the booklet and most likely have voted yes in the Referendum. The general level of satisfaction with the government also had an impact on how government published information would be perceived by voters, and he noted that satisfaction with the government was low throughout the course of the Referendum. (iv) While the issues surrounding the proposal are a factor in referendum voting, they are not the only consideration. In referenda, voters would often place their trust in the people making the arguments rather than analyse the arguments themselves. These may be experts, party figures, media commentators but voters may also decide on the basis of whom they mistrust the most. In many referenda and particularly in those in which the campaign has been weak and voters are uncertain of the issues, significant numbers of voters often vote no in what is effectively a signal of dissatisfaction to the government. (v) Data suggested that the electorate felt poorly informed about the proposal. A comparatively low level of understanding measured before and during the children referendum campaign suggests that there was a high potential for change as people started to interest themselves more in what the amendment was about. (vi) Data also suggested that the referendum campaign worked in favour of the no side, by converting opinion or persuading undecided voters disproportionately in favour of a no, or mobilising the no voters rather than those inclined to a yes and the campaign may have achieved all three. 93. Prof. Marsh also noted that the approach adopted by most researchers is to use surveys of electors and employ multivariate statistical methods to ascertain the likely affect of certain events upon voting. Those who report reading the booklet could be compared with those who do not. Statistical methods may be used to “match” the two samples because two groups could be different in any number of ways beyond their experience of the booklet. He noted that there were particular problems with this approach relating to the limited number of characteristics that can be used in the matching. A causal sequence can be very difficult to establish. There may be indirect effects which are almost impossible to capture such as whether a voter discussed the Referendum with anyone else who had read a booklet or whether that other person had been influenced by it. He was of the view that there was no reason to believe that any one factor will have the same weight in any campaign and that there is a great deal of evidence to suggest that any given factor may vary in its effect across different campaigns. 94. The petitioner also relied upon the evidence of Prof. Paul Whiteley, Professor of Governance at the University of Essex. He furnished a report in January, 2013. His expertise is in the study of elections, political participation and public opinion, including the methodology or design of surveys conducted to ascertain political opinion. He argued that the Referendum was of low salience as did Prof. Marsh and Dr. Bruter. He acknowledged that if all political parties in the state were in favour of a particular aim, people are likely to vote for it taking their cue from the party which they support. He stated that people look for assistance from a truthful or trusted source when undecided and confused about a proposal. In this instance, he stated that the information offered from a Department of State having expertise in the area would be influential, especially when invoking the opinion of other trusted sources such as Mrs. Justice McGuinness, a former Supreme Court judge. He also viewed the message carried by the information campaign as drawing upon the emotions of the reader of the booklet referring to the slogans which were discussed in the judgments of the Supreme Court. The booklet did not contain any counter arguments. He accepted that this material would not influence those who had already made their minds up or who were uninterested in the issues. The undecided voter would have attached more importance to the Department’s information than information coming from an obviously partisan group or party. He also acknowledged that though he believed the information influenced the decision of voters, he could not give a precise assessment of the extent of that influence because the data was simply not available. He was, however, satisfied that biased information would help one side of the campaign and, if one had data which measured the extent to which the Irish people trusted their government at the time of the Referendum and the extent to which they absorbed the material, the extent to which it influenced the outcome would be ascertainable. He was of the view that notwithstanding the absence of a detailed study capable of giving a precise estimate of the extent of the influence of the general campaign, this did not prevent him forming “a judgment of a rough kind about what the effects were”. He was clear, however, that it was not possible to determine the percentage affect of the booklet’s influence. He was satisfied its influence could not be proven either way. He was only able to conclude that the information “may very well have been quite influential”. He also believed that the general trend in the Referendum reflected trends seen in other Referenda, that the “yes” campaign started out from a position of strength but saw its support leak away during the course of the campaign. This led him to believe that a further negative trend would have continued to a somewhat greater extent had the government information campaign been neutral. 95. It is an interesting feature of Prof. Whiteley’s evidence that he relied upon a study which he and his colleagues carried out during the Alternate Vote System Referendum Campaign in Great Britain in 2011. It is clear from that study which was advanced as support for the proposition that voters will follow a trusted source in a Referendum that the trusted sources in that campaign were the party political leaders, and not a neutral government booklet which had been published to assist electors in making up their minds. Prof. Whiteley accepted that in the Children Referendum, the unanimous support of all the political parties in favour of the proposal must have resonated with the electorate and would have been a significant contributory factor to the yes vote. He accepted that all the experts were agreed that the supporting data was not available to enable a study to be carried out to determine exactly the extent of the influence of the government’s publication. 96. The court is satisfied that this evidence was completely insufficient to establish that the votes cast in the Referendum had been materially affected by the information distributed by the government (including the booklet). Dr. Bruter’s initial evidence set out a number of broad propositions which he acknowledged, required significant qualification and was the subject of reasonable criticism by Professor Marsh, some of which was based on the realities of the political campaign conducted in the lead up to the vote. Prof. Marsh and Prof. Whiteley importantly, agreed that there was insufficient data to enable the type of analysis that might make it possible to determine what influence, if any, the information had on the course of the campaign and to what precise level. Dr. Bruter also acknowledged the absence of this information but was satisfied to reach conclusions based on the limited data available. Dr. Bruter advanced additional material in support of his contentions in two further stages based on his analysis of a post Referendum survey. The Booklet: Statistical Evidence 98. The “Background and Objectives” section of the poll states that following the vote on 12th November, “market research was required to measure sources of information relating to the Referendum and the perceived role and efficacy of the Referendum Commission. Fieldwork took place from 20th November to 10th December, 2012, in the form of a nationally representative survey of 2,014 aged 18+ who were eligible to vote”. A number of questions were put to the respondents and the results were set out in tabulated form. The survey does not contain any lengthy analysis or commentary on these results save for a short summary at pp. 97 – 99 which gives some understanding of the data which interested the Commission and states:-
• 50% claim to have voted in the Referendum; claimed voting is lower amongst youths (under the age of ) 35. • 18% spontaneously suggest their non-voting was driven by a lack of understanding of this issue, whilst 16% mention that they did not know enough to vote. • 62% claim to have voted yes; dropping to below 51% amongst under 25s. • Amongst no voters, just over one in five put it down to not understanding the Referendum. 2. Level of Understanding of Referendums • 57% felt they understood the Referendum, thirteen stating very well. • This compares positively with the Fiscal Stability Treaty Referendum and Oireachtas Inquiry Referendums. • 79% claimed to have received the Referendum Commission Guide, a significant improvement on the 61% level seen in October 2011 and the 73% in May 2012. • 26% read all or most of the Referendum Guide, rising to 72% who read at least some of the guide. This overall figure is in line with recent referendums. Summary • In terms of the complexity of the guide, there is a slightly lower incidence of people who found the Referendum Commission Guide too detailed and complicated, compared to the May 2012 referendum. • Perceived helpfulness of the guide is in line with May 2012 levels also. • 56% correctly identified the Referendum Commission as being responsible for that guide with the government guide correctly attributed by 68%. 3. Referendum Commission Communications • TV dominates perceived sources of communications on the Referendum, as is generally the case. National radio advertising and outdoor advertising both emerged strongly too. • In terms of Referendum Commission TV and recall, 75% claim to have seen the ad which was ahead of October 2011 and May 2012 levels. • 39% claimed to have seen the free to air TV ad (in line with the May 2012) and 46% recall the free to air radio. • 75% agree that the ad was effective in letting people know there was a referendum about to happen and 67% agreed encouraged people to vote in the Referendum; 59% agree that the ad was effective in encouraging people to find out more about the Referendum.” Analysis of the Post Children Referendum Poll
101. The Referendum Commission was interested to ascertain the level of recognition amongst the electorate of its booklet. Dr. Bruter stated that the poll figures indicated an improved level of correct identification for both the Referendum Commission Guide and the government guide as against the level of recognition achieved in the Financial Stability Treaty Referendum. 56% correctly identified the source of the Referendum Commission Guide with 23% attributing it to the government, and the balance of 21% to other sources: 68% attributed the government guide to the government, 9% to the Referendum Commission and 23% to others. Dr. Bruter juxtaposed this finding with the figure concerning the level of importance attached by the respondents to the receipt of information that was independent of the yes or no side: 88% felt it was important. He had regard to the figures for the level of readership of the respective guides by those who recalled receiving them: 21% of those who received the government guide read all or most of it: 45% read part of it, but “by no means all”: 29% did not read it at all and 5% did not know. The same question when asked in respect of the Referendum Commission Guide provided figures of 26%, 46%, 27% and 1% respectively. When those who had received both guides were asked how much of the government guide they had read, 27% indicated that they had read all or most of it, 46% that they had read certain parts of it but by no means most of it, and 26% that they had not read it. 102. The witness laid considerable emphasis on the results obtained from answers to a number of questions in the poll, such as “how did you vote in the Referendum on Children?” The results indicated that 62% of those who received the Referendum Commission Guide (total 921) voted yes and 38% no: 64% of those who received the Referendum government guide (668) voted yes and 36% no, and 65% of those who received both guides (636) voted yes and 35% no, while 64% of those who received neither guide voted yes and 36% no. He accepted that these results could suggest that there was no difference between people who received the government guide and those who did not, but concluded that this particular result did not coincide with his raw data analysis. Dr. Bruter was satisfied that there was “a statistically significant relationship between saying you received the Government Booklet and saying that you voted yes in the Referendum”. 103. Prof. Marsh did not disagree that the figure of 9.3% could be obtained from the unweighted data presented. However, he carried out a similar statistical analysis, but this time using weighted data from the underlying raw data supplied by the pollsters in their working papers. He also concluded that those who said that they received the government guide were statistically represented as voting yes more than those who did not, but the difference was approximately 5.5% as against 9.3%. For the purposes of this analysis Dr. Bruter was happy to accept that one could apply weighted data to the calculation. 104. Counsel for the petitioner submitted that Dr. Bruter’s findings demonstrated that there was a 9.3% increase in the number of people who voted yes when they received the government guide than when they did not, and that given the 58:42% split in the vote, the increase in support was sufficient to carry the Referendum for the yes side. Counsel for the respondent submitted that Dr. Bruter’s evidence did not go that far and that Dr. Bruter accepted that it simply demonstrated the statistically significant relationship set out above. Prof. Marsh’s analysis indicates that when using weighted data, the statistical relationship is 5.5%. He was strongly of the view that this did not demonstrate a casual link between receiving the Government booklet and voting yes in the Referendum, and that more detailed data concerning other influential factors in the campaign must be available and accounted for, before such a conclusion could be reached. 105. While the figures establish some statistical correlation between receipt of the Government booklet and voting yes, the court is not satisfied that receiving the booklet meant that the recipient was thereby caused to vote yes. It is also clear that notwithstanding receipt of the Government booklet, most of the electorate did not vote at all and that a substantial proportion of those who received the booklet voted no. I am not satisfied that these figures established as a matter of probability that receiving the booklet caused an increase in the “yes” vote wholly attributable to that factor, thereby affecting the result of the Referendum as a whole. Dr. Bruter’s Second Report 107. Dr. Bruter explained that regression assesses the effect of a number of causes (independent variables) on a consequence (dependent variable) by attempting to ensure a levelling of conditions using control variables. Binomial logistic regression is a specific type of regression applied when the dependent variable (the factor one is trying to explain) is dichotomous (i.e. has only two possible values – voting yes or no). Bivariate regression is a method of measuring the impact of one variable on another. Multivariate regression assesses the effects of many independent variables on a dependent variable, thus permitting the identification of the variable that is really responsible for changes on the dependent variable – voting yes or no. 108. Dr. Bruter, therefore, applied binominal logistic regression in this case because the dependent variable was dichotomous, namely voting yes or no. He then tested for other independent variables including, he claimed, exposure to information or other elements of the Referendum campaign, and also receiving the Referendum Commission Booklet together with various demographic and social variables such as age, gender and social class. He stated that he also tested for exposure to information on the general campaign in two different ways: total exposure (aggregating every possible source included in the survey) and separate exposure (including each source of information or campaign element measured in the survey as a separate variable) and measured each as to whether they were more likely to make a person vote yes or no. In this way, he claimed to have controlled for potential substantive interference on the way individuals voted such as receiving the Referendum Commission booklet, exposure to television advertisements, and internet campaigning and radio advertising. He produced four regression equations and these factors were entered individually in the third and fourth equations and in aggregate form in the first and second equations. The specific questions used to create the variables used in the regression equations were set out in the appendix to the report. 109. The dependent variable was determined by Question 4a “how did you vote in the Referendum on Children?”. The independent variable of interest was determined by Question 6 “do you remember receiving this (Government Guide) in your own home?”. Other variables were contained in Question 13 which asked “apart from the Referendum Commission Guide, which of the following specific types of Referendum Commission advertising or Communication did you see, read or hear during this year’s Childrens Referendum campaign” (emphasis added). The respondents were then provided with a list of alternative media sources through which they may have received Referendum Commission information. In addition, Dr. Bruter sought to take account of other factors such as age, gender and social class. 110. Dr. Bruter concluded that each of the regression equations indicated that receiving the Government booklet was a statistically significant predictor of voting yes. He accepted, in the course of evidence, that in fact the other sources and factors referred to by him did not include data concerning other influential factors in the campaign because it was not available: the data which he used, refers only to exposure to Referendum Commission material during the campaign. He then contended that this material was an acceptable “proxy” variable for campaign exposure because people who saw that material were more likely to be following the campaign: it was the best imitation of the campaign available. Though he acknowledged that the overall campaign would have had a “politically huge” influence on voting “yes” or “no”, he did not think it would add anything to the impact of specifically saying that you received the Government Booklet. Prof. Marsh was of the view that the absence of this important data from Dr. Bruter’s analysis created a difficulty about drawing impressions of causality from the association posited by Dr. Bruter. It was important to control for other events that could have influenced the voter or, indeed, whether one received the booklet. 111. Dr. Bruter later explained that he did not mean to include this material as a proxy for the rest of the information received by voters in the course of the campaign, but only as a proxy in respect of exposure to the campaign in a more limited sense in that if they paid some attention to the balance of the Referendum Commission’s campaign, voters would likely take account of other elements of the campaign. It is fair to say that having regard to Dr. Bruter’s previous acceptance that it would have been preferable to have full data on the effect of the main elements of the campaign for the purposes of a study such as this it may be regarded as a very insubstantial proxy. 112. Dr. Bruter calculated a co-efficient, (b), which was the increase in the odds that one would vote yes rather than no having received the Government Booklet as opposed to not receiving it. A figure was also provided for a “standard error” (SE) which represented a measure of “dispersion” of the sample. If the coefficient was to be regarded as strongly reliable, it was necessary that it be “quite a bit bigger” than the standard error. The “rule of thumb” used was that the (b) coefficient should be twice the value of the standard error (SE) if it were to be considered strong enough to be relied upon. Thus, in regression 1, the (b) coefficient was 0.20 and the standard error 0.08 for the variable of receiving the Government booklet. In regression 2, the coefficient (b) was 0.31 and the standard error 0.14. Thus, the coefficient (b) could be regarded as a strong and reliable indicator of the increase in the odds that one would vote yes rather than no having received the Government booklet. Similar exercises were carried out in respect of the other variables such as receiving the Referendum Commission Booklet, gender, age and social class and what was referred to as “total other exposure”. The statistical significance of the results was also calculated as a “P” value. These and other figures enabled Dr. Bruter to conclude that his findings in respect of receiving the Government Booklet in each regression were statistically significant. 113. In his third and fourth regressions, which were similar to the first and second regressions, Dr. Bruter included each individual source “of campaigning” individually. These included figures concerning the influences of variables such as communications seen, read or heard during the campaign broken down into television advertising, television coverage, radio advertising, radio coverage, internet advertising and internet coverage. As noted earlier, it later emerged that these figures related to coverage of the Referendum Commission’s materials and not the general Referendum Campaign. Dr. Bruter concluded “The results confirm that even when controlling for all the additional variables…identified, receiving the Government booklet made respondents consistently and significantly more likely to vote yes in the Children Referendum”. 114. Prof. Marsh urged a degree of caution in using the results of the “Behaviour and Attitudes” poll. He stated that the pollsters were not attempting to obtain a rounded view as to why the particular outcome occurred in the Referendum. They were trying to analyse the effectiveness of the Referendum Commission’s campaign. The Referendum Commission was seeking information about the reception of its booklet and other sources of information. It was not concerned and did not take account of other sources of information and their significance to the Referendum campaign such as how voting intentions were affected by political allegiances or government popularity. 115. Prof. Marsh pointed out that demographic weighting is often a feature of surveys and is unquestionably acceptable when single variables are under consideration. However, there is amongst political scientists a debate as to whether weighting is appropriate to multivariate analysis. That controversy was explored in evidence and, in effect, both experts agreed that the application of weightings should not make a difference in multivariate analysis. Prof. Marsh pointed out that there is no guidance on which form of analysis should carry more weight when weighting produces a difference. He stated that this means that there may be known biases in a survey which may be ignored or not taken into account. While weighting, even in this case would not make a great deal of difference to the figures in the tables produced by Dr. Bruter, Prof. Marsh was of the view that, in this case, it made a great deal of difference to the inferences to be drawn from the results obtained. 116. Prof. Marsh re-ran the regression model provided by Dr. Bruter and provided for weightings in respect of gender and social class. As expected the figures were substantially the same. However, there was a significant difference in respect of those in receipt of the Government guide, which made it unsafe to say that there was a significant relationship between having recalled the Government guide and voting choice. Dr. Marsh does not contend that there are major differences in the figures obtained for weighting, but that what should be an inconsequential weighting on the re-run on regression 2 produced a (b) coefficient for receiving the Government booklet of 0.185 and a standard error figure of 0.152. That is clearly a figure that is not twice the standard error. Dr. Bruter stated that a re-run of his model with weighting should not change the results significantly and Prof. Marsh agrees: however, what was of concern to Prof. Marsh was the ease with which a slight adjustment for weighting changed the result and produced the above figure and a significantly different but related statistical significance value “P” and other figures that undermined the reliability of Dr. Bruter’s model and results. Dr. Bruter disagreed with this analysis and argued that the preponderance of opinion amongst political scientists was that weighted opinion should not be used in regressions of this kind. 117. Prof. Marsh also conducted an experiment to test Dr. Bruter’s model. He postulated that the inclusion of the Government booklet in the model should make a significant contribution to explaining voting “yes” and if removed from the model it should be clear that it has much less explanatory power. However, when this exercise was carried out, the predictive power of the model improved by only a trivial margin when including the receipt of the Government guide as opposed to not receiving it. Prof. Marsh, therefore, concluded that as a variable this “tells us almost nothing about the way people voted”. Dr. Bruter disagreed substantially with the statistical technique used by Prof. Marsh to measure how one variable explains another and said it was not applicable to binomial linear regression, but is more usually used in other types of regression. 118. Prof. Marsh also postulated a theory that the Government guide and the Referendum Commission Guide could each have affects which are not entirely independent of one another, and that the regressions should be adapted to allow for that fact. He concluded that when that is done there are no significant effects of recall of any kind. In a table that emerged from that study he demonstrated that those who recalled receiving neither guide, just the Referendum Commission Guide, or both the Referendum Commission and the Government Guides, were more inclined to vote yes: by implication receiving only the Government guide led to a smaller probability of voting yes compared to either getting no guides at all or only the Commission’s guide, or both. He concluded that recalling either guide or neither guide made no difference to the likelihood of a yes vote. Dr. Bruter contended that the methodology of the analysis was flawed: the witnesses, once again, disagreed. 119. Prof. Marsh questions the confidence to be placed in estimates of the affect of recalling the Government booklet which treats data as a simple random sample. The data in this case came from multiple samples in which areas were first chosen and then individuals. In a random sample every person has an equal chance of being included. In this case, individuals are chosen from 80 sample points. Interviewers then select respondents so as to provide a sample that fits age, gender and social classes. This “clustering” should mean that sampling variants (the degree to which identically drawn samples yield somewhat different results as a consequence of chance) will be greater than if the sample has been a “simple random sample”. The software employed to assemble the data underestimates “standard errors and confidence intervals” that surround estimates from the sample. This requires a correction of one and a half times that which is provided in a typical software statistical package which assumes a simple random sample. Dr. Bruter disagreed with this proposition which was based on the authoritative work by Gelman & Hill, Data Analysts using Regression and Multi-level/hierarchical Models (Cambridge University Press 2006: Chap. 20). He said that the 1.5 times adjustment would depend on the quality of the sampling and could only be a rough instrument which was inappropriate to psychologically based work. He claimed that the Gelman & Hill work concerned a very specific type of data which applied to a very specific type of model namely, multi-level models which were used primarily when dealing with comparative studies between one country or jurisdiction and another in which referenda, for example, had been held. This principle was not applicable to issues which arose within a single jurisdiction. Prof. Marsh applied a 1.5 times and a 1.1 times adjustment to Dr. Bruter’s figures in respect of the error. When these adjustments were made the result and figures demonstrated no basis for the existence of any affect of the recall of receipt of the Government booklet on voting. Prof Marsh concluded as a result that the affect was not significant. 120. Prof. Marsh also focused upon the fact that the variable referred to in Dr. Bruter’s analysis was in respect of those who “received Government guide”. He noted that only 51% of the weighted sample reported that they had received the guide compared with 79% who received the Referendum Commission guide. However, similar numbers of each guide were printed and distributed through An Post. Therefore, all households got the government’s guide which indicates that the recollection of respondents was faulty. The actual question 6(b) asked “Do you recall having received the Government guide into your home?”. Prof. Marsh viewed this as a measure of recall not of delivery. The meaning of “recall” in that context may involve a number of propositions. He queried whether the failure was random, whether it reflected non-delivery, whether it was selective and on what basis did some recall and others not. The figures indicated that those who recalled being in receipt of one guide were more likely to recall the other. Of those who said they received the Commission guide, 86% said they received the Government guide and of those who did not get the Commission guide only 38% said that they got the Government guide. Prof. Marsh stated that this was a strong association leaving a difference between the two figures of 48 percentage points. He, therefore, concluded that it was wrong to treat recall as a random treatment variable. There may be factors that prompt recall which, if excluded from a model, are likely to bias the estimates – the coefficients – adopted by the model signifying the effect of recall on the vote. Dr. Bruter was of the view that this was simply semantics and that recall indicated that the household had received the document. 121. I have considered all of the econometric evidence from the expert witnesses. There were numerous differences between them, but what was missing from the discussion was any real engagement with what was happening in the Referendum campaign. The cut and thrust of that debate and the strength of influence of the various participants in favour of the proposal or, indeed, against it, were not afforded any substantial part in the academic narrative provided by the witnesses. To a large extent this was dictated, as was readily acknowledged by Dr. Bruter, by the fact that there was an enormous absence of information which was highly important to an understanding of the dynamics of the campaign. The assessment by Dr. Bruter of the data concerning the affect of recall of the Government referendum booklet has been the subject of a great deal of criticism by his colleague, Prof. Marsh. It appears to be accepted that he may be correct in his conclusion that there was some statistical connection between recall of the receipt of the booklet and voting yes. The figures which he relies upon contain some variables but nothing remotely sufficient to inform the court authoritatively of the affect of the other elements of the campaign. The “proxy” variable offered is clearly insufficient to measure the reality of the affect of other factors in the course of the campaign. It is significant that political and social discourse over the last thirty years has been obliged to focus from time to time very heavily on issues concerning the welfare and rights of children. The Houses of the Oireachtas, the Courts, the Churches and virtually every organ of the media have been concerned, if not convulsed, with a whole range of issues concerning child abuse, child protection, homelessness, custody and access issues between children and parents, the detention of children, education rights of children, and many other child and family centred issues. These matters clearly informed the debate that led to the formulation of the proposal to amend the Constitution. Understandably, neither Prof. Whiteley nor Dr. Bruter had any knowledge of this background. 122. Prior to the publication of the wording, the polls indicated that there was a substantial positive view in favour of a proposal to amend the Constitution. That was reduced over time following the publication of the wording, but the majority obtained on polling day was 16%. It would require clear and cogent evidence to establish the proposition that the Government Booklet influenced the Referendum as a whole. I am not satisfied on the basis of the evidence of Dr. Bruter, and having heard the evidence of Prof. Marsh which I prefer, that his work on the “Behaviour and Attitudes” survey is sufficient to enable me to find for the petitioner on the balance of probabilities. The experts have disagreed over fundamental elements of the survey, the meaning and significance of the questions posed, interpretation of the percentage figures given and how best to assemble and interpret the data underlying the figures set out in the survey. I am not satisfied that the petitioner has established, notwithstanding the booklet’s widespread distribution, that it materially affected the result of the Referendum as a whole. The Website 124. In the McCrystal ruling delivered on 8th November, 2012, it was held that there were extensive passages in the website which did not conform to the McKenna principles. Though not all of its contents were in breach of the principles, the court considered that it was not possible to redact the material for the website. The court was satisfied that the respondent had acted wrongfully in spending public monies on the website. 125. The nature and extent of the breach of the McKenna principles contained in the website was addressed in the judgments of the Supreme Court delivered on 11th December, 2012, by Denham C.J. and Fennelly and O’Donnell J.J. It is not necessary at this stage to set out in detail the various criticisms of the material, which were extensive. The issue in these proceedings is what affect, if any, the contents of the website had on the Referendum result. 126. Mr. Roger Jupp, Vice Chairman of Millward Brown Lansdowne, one of the most experienced market research companies in Ireland, noted that the website had 23,309 “unique” visitors over the full length of its life, 10% of whom were from outside the country. A “unique” visitor is a computer visitor who logs onto a website but may return to it on other occasions. 23,309 was well under 1% of the total population over 18 years of age. He noted that there was no evidence, research based or otherwise, as to what people believed the website was telling them, either in favour or against the proposal or, whether it changed their opinions from one side to the other, or whether it reinforced their opinions. The website’s busiest day was on 19th September. Thereafter, the average number of daily visits was three to four hundred visitors, frequently dropping to much lower than that. The website had a minimal reach into the general population. Mr. Jupp stated that there was no evidence that the website caused a change in voting intention or had a material affect on the Referendum outcome. 127. I am not satisfied having regard to the established number of visitors to the website that there is any evidence that the website had a material affect on the overall result of the Referendum. It is clear that given the size of the majority, it could not have been decisive, and if it had any affect, it could only have been minimal. Print, Radio and Television Advertising 129. Brindley Advertising, which holds the government print advertising contract, was commissioned to develop proposals and an advertising schedule. The key messages to be conveyed were similar to those in respect of the broadcast media. Advertisements were carried in national newspapers, regional/local papers, weekend newspaper magazines (on 3rd and 4th November), the Irish Farmers Journal, the Metro Herald and the Big Issue during a period 22nd October and 8th November. Though every effort was made to withdraw all advertising after the Supreme Court ruling in the McCrystal case, an advertisement was published in ten local/regional newspapers with publication dates of 9th or 10th November, because printing had been completed in advance of the Supreme Court ruling on 8th November. 130. Dr. Robert Heath, an Associate Professor lecturing in Advanced Advertising Theory at Bath University, gave evidence on behalf of the petitioner on the likely emotional influence of government advertising materials used during the course of the Referendum. He compiled a report dated 28th January, 2013, upon which he elaborated in the course of evidence. The report goes somewhat beyond the issue of advertising in that he also reviewed the effect of the cover on the Government booklet. Dr. Heath’s research speciality is in the processing of emotion in advertising with a special emphasis on how this occurs at low levels of attention and influences intuitive decision making. He is the author of “Seducing the Subconscious- the Psychology of Emotional Influence in Advertising” (Wiley – Blackwell, 2012). He favours the view that the message in advertising has little if any influence on the favouring of a product, but that emotive non-verbal communication such as body language, tone of voice and emotional behaviour that accompanies and qualifies the communication induces a favourable response. He contends, based on work in the area by himself and others, that emotive communication or, “meta communication” in publicity material is processed automatically, instantaneously and regardless of how much attention is paid to the communication. It exerts a powerful influence on our “favourability” and in situations where decisions are hard to make, it subconsciously influences the direction of the decision through its affect on human intuition. It is claimed that because its influence is not as obvious as verbal claims or written messages, people tend not to counter-argue meta communication. He cited experimental evidence to suggest that the slightest exposure to advertising material can influence behaviour even when no recall or knowledge that the material was seen, remains. He examined government advertising materials and offered his professional opinion on the likely emotional influence of these materials on the Referendum vote. The Booklet
• Bullet point supporting families. • Bullet point removing inequalities in adoption. • Bullet point recognising children in their own right.” 133. He stated that the cover’s influence was enhanced by the logo illustration of three very young children hand in hand. This was another meta communication which would be processed instantaneously and automatically in the following way:-
• They are holding hands, which children do when they fear being separated. • The child on the left appears to be pausing or recoiling slightly; the one in the middle has half length trousers that seem either too small or worn and bare feet; the child on the right appears especially small and lonely and also has bare feet. • The children are shown in silhouette with their faces blank. This suggests these children are not proper people with identities and views, but more akin to “non-people” who do not really count. • My opinion is that most people would feel that these were three “lost waifs”, wandering alone and defenceless to the world”. The Television Advertisement Printed Advertisements 136. Dr. Heath noted that the booklet had been distributed to every household in the state and that the vast majority of voters had seen the cover. The printed advertisements, according to the government’s figures, would have been seen on average 3.6 times by 85% of adults and 2.2 times by 78% of adults in the last week before polling. He believed the logo was foremost amongst the devices used in the material and that it would have become for the vast majority of people, one of the defining “emotionally competent association(s)” connected to the Referendum. An “emotionally competent association” is an element which is repeatedly exposed alongside a brand or an event and which carries powerful emotional significance. It is a term taken from a book of the same name by Damasio (2003). Dr. Heath stated that the power of such devices has been demonstrated in many case studies of advertising effectiveness. It is particularly effective in influencing those who are unsure of which way to make a decision. In that regard, he noted that there was a high level of lack of understanding or no understanding of the issues being addressed in the Referendum. He, therefore, expected that an emotionally competent association such as the three children logo had a powerful influence on voter behaviour. He emphasised that such a subconscious influence can occur even when exposure is so brief that people cannot recall ever seeing the material. Thus, evidence that people do not recall receiving a booklet or seeing a print advertisement or seeing a television commercial does not mean that they did not see it, and were not influenced by it. Nobody who saw the logo would have seen it as an attempt to influence their behaviour and thus, significantly, hardly anyone would have tried or, indeed, been able to counter argue the idea that the image conjured up in their subconscious. He, therefore, concluded that the government publicity material contained not just verbal messages, but important and highly influential meta communication that influenced voters’ behaviour. 137. John Fanning is an Adjunct Faculty Member at University College Dublin Michael Smurfit Graduate Business School and currently Chair of Bord Bia Brand Forum. He worked in the advertising industry in Ireland for more than forty years at the highest level and was Managing Director of McConnell’s Advertising and subsequently Chairman between the years 1980 and 2005. During his career he worked on a number of Referendum campaigns including for the Referendum Commission on the second Nice Treaty Referendum in June, 2001. He gave evidence on behalf of the respondents and was invited to offer an expert view on whether the advertising and website material sponsored by the Department in the Referendum had an impact on its outcome and, if so, to what extent. He stated that it was notoriously difficult to determine the effect of a marketing communications campaign. He relied on a widely cited authority “Testing to Destruction: A Fresh and Critical Look at the Uses of Research in Advertising” by Alan Hedges, which posed the question whether research can do anything to prove that advertising money has been well spent. The author concluded that it could not be proven in a direct sense, although the more one knows about what goes on in terms of mass communication the better placed one is to make an intelligent judgment about whether advertising has worked. Mr. Fanning stated that in order to come to a definitive assessment of the impact of the material published by the Department, specific survey data generated before, during and after the Referendum campaign would be required. At a minimum that would require data on the degree of knowledge which potential voters had of the issues surrounding the Referendum, the degree of exposure they had to campaign literature and advertising from the Department and other sources, the degree of exposure they had to traditional media, digital media and word of mouth comment. It would also be necessary to assess resulting attitudes to the issues involved and the likelihood of voting. Though there was some limited pre-campaign polling data, he concluded that there was nowhere near enough data to enable one to evaluate the effect of the materials published by the Department when examined separately or in the context of the wider campaign. He noted that even if one had this type of information, two problems remained namely, the difficulty of untangling the effects of the Department’s campaign from other “noise” going on in the wider Referendum campaign, for example, in newspapers and on television and radio in the blogosphere, and in private discussion about the Referendum which people had with friends, family and work colleagues. In addition, in seeking to assess the impact of material published by the Department, current theories on marketing communications by Dr. Heath suggest such marketing or advertising activity often operates at very low levels of involvement, thereby making recall difficult. A voter may be processing information from different sources at different levels and at different points during a Referendum campaign. He concluded that it would be impossible to separate the effect of the material published by the Department from this surrounding “noise”. 138. Mr. Fanning also deduced from opinion polls carried out before the Referendum leading up to the actual vote that the percentage of the population in favour of change fell steadily from over 90% just under a year before the vote, to the 58% majority on polling day. Thus, voters were significantly more likely to vote “yes” before exposure to the government campaign than after it. This is in accordance with previous Referendum results as noted by Prof. Whiteley and Prof. Marsh. He concluded the Department material had minimal affect in influencing voters. He also suggested that the main reason for the lack of impact of this material during the campaign was its emphasis on public information rather than persuasion. It had a very conservative tone and was conventional in design, style and layout. The fact that it included some phrases and images, the effect of which was to render it a breach of the fairness required by the McKenna principles, did not mean it was effective in terms of persuading voters to vote, and to vote yes. The evidence of the polling data, such as it was, suggested that it did not have that persuasive affect. 139. He maintained that there was no “big idea” in the material in the sense that there was nothing in it to jolt people into a fresh recognition of a product, service or referendum subject. The government and public bodies have used creative ideas to good affect in recent years, such as graphic depictions of car accidents and injuries in road safety campaigns and the health affects of cigarettes in anti-smoking campaigns. He concluded that the materials published by the Department, including the logo of the three children and the four slogans, did not carry that type of affect. He considered the government campaign to be “low voltage”. 140. Mr. Jupp agreed with Mr. Fanning that it was impossible to establish definitively what role the Department’s materials had on the Referendum result because it was part of a broad swathe of information coming to potential voters from a diffuse set of sources. 141. Dr. Heath contended that there was a large proportion of people who did not understand the issues in the Referendum and would have used their intuition to make a decision on how to vote, which would have been very much directed by the emotive elements and materials. He sought to support this proposition by reference to the “Behaviour and Attitudes” survey material. However, I am satisfied that there was insufficient evidence available for him to reach the conclusion that the emotive material caused people to vote yes. Indeed, some of the material put in cross examination suggested that less than half of those who only received the Government guide voted yes, and that a higher proportion of people who only got the Referendum Commission guide voted yes. However, once again, the absence of a multivariate analysis of other influences at work in the campaign makes it difficult to determine the precise causal affect which this material had, if any. I am not satisfied that Dr. Heath’s evidence establishes as a matter of probability that the Referendum result was as a whole materially affected by the meta communication described and I prefer the evidence of Dr. Jupp and Mr. Fanning to that of Dr. Heath. Events Post the Ruling of the Supreme Court – 143. The petitioner contends that following the ruling the respondents did not attempt to remedy the breaches of the Constitution identified by the Supreme Court or negate or dilute their affect. It is claimed that the Minister for Justice, Equality and Defence, Mr. Alan Shatter T.D., failed to issue an apology for the breaches when interviewed and requested to do so on RTE news. It is alleged that the Minister stated that the ruling only applied to the website and booklet and did not apply to the advertising campaign, and thereby further obfuscated its affect. It is said that the Minister drew attention to the High Court ruling when criticised in relation to the Supreme Court ruling. 144. The petitioner also complained that no legislation was enacted to postpone the polling day. It was claimed that, as a result, “no” campaigners were deprived of a right to participate in a constitutionally compliant Referendum process. Dr. Bruter and Professor Whiteley gave evidence about the likely benefits to the Referendum process of the postponement of the poll following the Supreme Court ruling. Dr. Bruter concluded that it was impossible to define a period of days within which a significant event, such as the ruling, will take effect in a campaign. He thought it reasonable to conclude that the affect of the ruling on the electorate would have been stronger and more stabilised if the Referendum had been delayed by a few days or three weeks. He later agreed with Professor Whiteley that a delay of two weeks between the ruling and the polling day would have been sufficient. In that regard, Professor Whiteley noted that according to a European Social Survey, Irish voters had a significant level of trust in the Supreme Court and that the ruling would have been quite influential if given time to diffuse amongst citizens. The Professor relied upon his study of the last British General Election during the course of which polls were taken in relation to significant events. One of these events was the first tri-partite leadership debate between the Liberal Democrat leader, Mr. Clegg, and the other two party leaders, the present British Prime Minister, Mr. Cameron and the then British Prime Minister, Mr. Brown. Mr. Clegg was believed to be the victor of this debate and there was a very significant surge in support for his Liberal Democrat Party which, it was said, took two weeks to have an impact on general trends. However, when the graph produced was examined, it is clear that within approximately 48 hours of the debate the poll ratings of the Liberal Democrats had surged and remained at or about that level until the date of the general election. Prof. Whiteley submitted that the longer period of two weeks was required in order to judge whether the affect of the significant event had “stabilised” in the electorate. This evidence was advanced to support the proposition that a delay in the holding of the Referendum would have favoured the growth of the “No” vote due to an increasing understanding of the misconduct of the government by the electorate. However, it was clear from the post voting survey carried out on behalf of the Referendum Commission that a high percentage of those who voted were aware of the Supreme Court ruling. 145. Prof. Marsh considered that the negative message of the Supreme Court ruling to some extent cancelled whatever positive affect, if any, gained by the government material. Over 61% of adults were aware of the ruling and 39% were unaware according to the Behaviour and Attitudes survey. This was adjusted and recorded as 77% of voters within the survey. It was also submitted on behalf of the petitioners that this case must be contrasted with the facts of the Hanafin case in which a week remained of polling within which the affect of the judgments in McKenna (No.2) was absorbed by the electorate and various steps were taken by the defendants in that case to remedy the breaches. This also permitted the absorption by the electorate of the full extent of the breaches of the McKenna principles and the fact that the material was not impartial and had been wrongfully funded by the government. I am not satisfied that the time remaining for canvassing and campaigning in the Divorce Referendum following the Supreme Court ruling binds this Court in relation to its consideration of the two day period between the ruling and polling day. That period is a factor to be taken into consideration, but I am not satisfied that there is sufficient evidence to indicate that the ruling and later judgments would have had a greater affect if the polling date had been delayed. 146. There was extensive national coverage of the McCrystal ruling. On 8th November, RTE radio’s News at One, beginning at 1.00pm., devoted almost three quarters of its coverage to the ruling and its implications, including the heated exchanges between Mr. Shatter, the interviewer Mr. O’Connell and Mr. John Waters. Mr. Waters later told the court that he believed that the Referendum would have been defeated if the poll had been postponed following the Supreme Court ruling until after delivery of the judgments when the electorate would have fully understood its meaning and affect. The programme had an audience of approximately 360,000 listeners. The ruling was also the main topic on the RTE Radio programme Liveline on the same date at 1.45pm, which included an extensive interview between the presenter, Mr. Duffy, and Mr. Mark McCrystal. Liveline at the time was Ireland’s second most listened to programme with an average audience of 420,000 listeners. The ruling was the first item on the RTE Radio programme Drivetime commencing at 4.30pm, which included legal analysis. It was the first item on the Six One News on television which included a film report on the judgment by the Legal Affairs Correspondent and a filmed report on the reaction of campaigners, together with live analysis by a political correspondent. This was again followed by a live interview and debate on the ruling between the Minister for Justice, Equality and Defence, Mr. Alan Shatter T.D., and Mr. Mattie McGrath T.D. This programme had an average viewership of in excess of 400,000. The ruling was given similar prominence in RTE television news at 9.00pm, which had an even larger audience of approximately 730,000. The RTE Primetime programme on the evening of 8th November, 2012, also covered the matter extensively and had an average audience of approximately 350,000. The matter was then covered in the TV3 programme Tonight hosted by Mr. Vincent Browne, featuring an interview with Mr. McCrystal and a panel discussions with Ms. Patricia McKenna. It was also covered by the presenter Mr. Matt Cooper on TodayFM, which had an average listenership of 160,000, the lunchtime programme and the Right Hook on Newstalk which had an average audience of 130,000. On 9th November, the McCrystal ruling was again the leading story on Morning Ireland on RTE Radio which featured an interview between the Transport Minister, Mr. Leo Varadkar T.D., and Mr. Malachy Steenson on the issue, and the Referendum generally. This programme has an average audience of 441,000 listeners. The McCrystal ruling was also given saturation coverage in the print media over 8th, 9th and 10th November. 147. I am not satisfied on the evidence adduced that the Ministers cited showed disrespect for or sought to obfuscate the effect of the Supreme Court ruling. I have viewed the materials and listened to the recordings produced to the court in respect of these matters. It is clear from these materials that there was a robust engagement between at least one of the interviewers and the Minister for Justice, Equality and Defence in relation to the affect of the ruling on RTE news. Those listening to the exchanges or reading the newspaper reports of the ruling can have been in no doubt of the nature of the Supreme Court ruling. I am not satisfied that the Supreme Court ruling in the McCrystal case could be said not to have had an immediate impact. It was a short, focused and definitive condemnation of the expenditure of public funds by the government on a partisan information campaign as a breach of the Constitution and was reported as such. This gave rise to a robust public debate and engagement on the issues raised by the ruling. 148. The respondents also submit that the Minister for Children and Social Affairs issued a press release in which she acknowledged that the Supreme Court had found that some of the material published by her department in connection with the Children Referendum did not comply with the McKenna principles. I am satisfied that she demonstrated respect for the Supreme Court judgment in its criticism of the unconstitutional expenditure of public funds and undertook that the government would comply fully with the judgment of the court. On the same date, the Minister informed the Seanad of the Government’s intention to act on the ruling and to cease distribution and publication of materials. On the same evening she apologised to the Irish people in clear terms stating that she was “extremely sorry that this happened. I would not have wanted this to happen, absolutely not”. 149. On 8th November, the material on the website was edited by reference to the Supreme Court ruling and it was taken offline completely on the morning of 9th November. Advertisements scheduled to appear in five national daily and one local newspaper on 9th November, as well as proposed publication in another free sheet newspaper, were cancelled. Radio and television advertisements scheduled to be broadcast on 8th, 9th and 10th November were cancelled. On 8th November the department also issued a notice to public bodies, including libraries, citizen information centres and family resource centres, requesting that copies of the criticised booklet be removed from public display. I am not satisfied that there is any evidence upon which to base a claim that the action of any Minister or the Government following the Supreme Court ruling in any way contributed to the obstruction of or interference with or hindrance of the conduct of the Referendum or, amounted to conduct which was unconstitutional, and interfered in any way with the constitutional rights of citizens or the democratic process. 150. The petitioner contends that the government should have sponsored emergency legislation to ensure the postponement of the Referendum in order to guarantee an effective remedy to the petitioner for the breach of her constitutional rights. Apart from the fact that the petitioner never sought any such pre-Referendum relief, I have considerable doubt as to whether the jurisdiction of this Court extends to a review of a government’s failure to introduce emergency legislation to postpone a Referendum. In Fitzgibbon v. Ireland (Unreported, Supreme Court, 8th June, 2001), the Supreme Court refused an application to restrain the holding of three Referenda or alternatively, the counting of votes on grounds of insufficient information and inadequate time for debate. In an ex-tempore judgment Keane C.J. thought it unnecessary to consider whether there might ever be “exceptional circumstances” in which the court could order the postponement of a Referendum poll and considered that such circumstances would be “so rare and so exceptional that it is difficult to conceive them in practice”. The court applied the judgment of the Supreme Court in Slattery v. An Taoiseach [1993] 1 I.R. 286, in which a claim was made that the court should grant an injunction to restrain the holding of a Referendum on ratification of the Treaty of European Union until such time as the government had provided further factual information on the benefits, disadvantages and obligations of the Treaty and until it had clarified its implications for the constitutional protection of the right to life of the unborn. McCarthy J. stated (at p. 301):-
Conclusion 153. I have considered all of the evidence and submissions made in respect of the trial of the petition concerning all of the grounds advanced, and I am not satisfied on the balance of probabilities that the petitioner has succeeded in establishing those grounds for the reasons which I have set out in the judgment. The petition is dismissed.
|
