THE SUPREME COURT
[Appeal No. 432/13]Denham CJ
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 held on the 10th November 2012
The Minister for Children and Youth Affairs, Government of Ireland, Ireland and the Attorney General
Notice Parties /Respondents
[Appeal No. 342/2014]
The Minister for Children and Youth Affairs, Government of Ireland, Ireland and the Attorney General
Defendants /RespondentsJudgment of Mr. Justice Clarke delivered on the 24th April, 2015
1.1 I agree fully with the judgment delivered by O'Donnell J. in these two appeals. I write this concurring judgment for the purposes of explaining briefly some aspects of the reasons why I support the conclusions set out by O'Donnell J. in his judgment. I also include some observations on a number of the issues which arose in the course of these proceedings.
1.2 For the avoidance of doubt, I wish to emphasise that nothing which is said in this concurring judgment should be taken as in any way indicating that, in my view, the test set out at para. 85 of the judgment of O'Donnell J. should, in any way, be altered or modified. It is important, for the purposes of absolute clarity, that there be no doubt but that the test set out in that paragraph represents the basis on which a court faced with a referendum petition should analyse the question of whether a material effect on the result of the referendum in question, resulting from an established breach of law, has occurred. Indeed, it might well be said that one of the difficulties which arose on the appeal in respect of the referendum petition in this case was identifying with precision the test which this Court applied in Hanafin v. Minister for the Environment  2 I.R. 321. The respective judgments of the members of this Court in that case seem to express the test in somewhat different terms, thus leading to a potential difficulty in identifying a test which was supported by a majority of the members of the Court. In my view, the proper test is that identified by O'Donnell J. in his judgment in these appeals.
1.3 As the facts which underlie the issues which arise in both of these appeals are very fully set out in other judgments I do not propose to rehearse them here. I turn first to the issues.
2. The Issues
2.1 There were three main issues which arose on these appeals. The appeals themselves, of course, encompassed both the referendum petition brought before the High Court challenging the provisional result of the Referendum in respect of the 31st Amendment of the Constitution (Children) Bill (“the Referendum”), and the plenary proceedings which challenge the consistency of certain aspects of the Referendum Act 1994 (“the 1994 Act”) with Bunreacht na hÉireann. The issues which arose in that constitutional challenge were, to a significant extent, interlinked with some of the issues which arose on the referendum petition, not least the issue of the proper test or standard by reference to which it might be appropriate for a court to interfere with a provisional result.
2.2 However, the broad issues were as follows:
2.3 As appears from other judgments, the constitutional issue was concerned with the second of those questions for, it was argued, if the test was as submitted on behalf of the defendants/respondents (“the State”) then, it was said, the 1994 Act was inconsistent with the Constitution in that regard. I turn first to the technical question concerning the proper interpretation of the procedures provided for in the 1994 Act for the conduct of a referendum petition.
(a) An issue concerning the proper interpretation of the 1994 Act, insofar as it relates to a technical question of interpretation, arising out of the fact that the procedure contemplated involves an application by the petitioner for leave followed, if leave is granted, by a trial. The specific issue concerned whether the materiality of any established irregularity only properly arose, as the petitioner/appellant (“Ms. Jordan”) argued, at the leave stage;
(b) The proper test to be applied in determining whether any established irregularity in the conduct of a referendum had, as specified in s.43(1) of the 1994 Act, materially affected the result;
(c) Whether, applying the appropriate test, a material affect had been established in this case.
3. The Leave/Materiality Issue
3.1 On that technical question of interpretation of the 1994 Act, insofar as it governs the conduct of a referendum petition, I agree with the reasoning set out in the judgment of O'Donnell J. I would add only two further observations.
3.2 It is a fundamental requirement of fair procedures that any party potentially affected by a finding adverse to their rights and obligations should have an opportunity to be heard in an appropriate way on the issue in question. That principle applies just as much in the context of an election or referendum petition as it does in any other form of litigation. In the context of a referendum, it must be recalled that those who support the provisional result have just as much of an entitlement to have the argument in favour of upholding that provisional result fully considered by the Court as those who wish to challenge the provisional result have to present the opposite case. It must also be recalled that it is possible that a provisional result which led to the rejection of the proposal contained in a referendum can also be the subject of a challenge. It may well be that such a challenge would not be likely to be of the type which arises here, for it is unlikely that it would be contended that improper government expenditure influenced the result against the passage of a referendum. It is, after all, unlikely that a proposal to amend the Constitution, which is required to be passed by both the Dáil and the Seanad before being put to the people, would be likely to be opposed by the government of the day (although there is no formal reason why that might not be so).
3.3 But whichever side may challenge the result, those holding the opposite view are entitled to a procedure which ensures that the case against the challenge is fully and properly heard. In passing, I should observe that the provisions of the 1994 Act, and in particular ss. 44 and 46, which provide for the service of a petition in respect of which leave has been given and the conduct of the trial of such a petition respectively, are silent as to the parties, other than the petitioner and the Director of Public Prosecutions, who may be represented, although s.46 clearly contemplates that there may be a number of parties to the petition. Likewise, s.53 contemplates costs being awarded against a party without addressing whom such a party may be. The Rules of the Superior Courts which apply in respect of all forms of election petitions (Order 97) seem to expressly contemplate that any person who is required to be served with a petition, or any person who is directed by the court to be so served, can appear. In substance, that would seem to require that the Minister for the Environment is to be properly regarded as a Notice Party, together with any one else whom the court directs.
3.4 It should also be noted that the only parties who, in election type petitions, are regarded as respondents are electoral candidates affected or potentially affected by the petition in question. It does appear that the Minister for Children and Youth Affairs, the Government of Ireland, Ireland and the Attorney General were directed to be served in this petition and should, therefore, be regarded as Notice Parties. It is not clear as to whether the Minister for the Environment took any formal position. However, any such technical deficiency is more than overcome by the fact that the Minister for the Environment is a member of the Government of Ireland which was itself a Notice Party. However, it is worth commenting that it is important that there be clear precision as to who is to be regarded as a party to any specific election petition. This is particularly so in the context of a referendum petition, where the interests of citizens generally are engaged to a greater extent and where there are no candidates as such. Furthermore, it would, in my view, be useful if some attention were to be paid to putting in place procedures (whether at the level of legislation or rules of court) to ensure that the position of parties with a legitimate interest in the outcome of a referendum petition can be properly represented at the trial of such petitions.
3.5 Be that as it may, it is against such a background that the construction of the 1994 Act urged on behalf of Ms. Jordan needs to be considered. It is suggested that a proper interpretation of the relevant provisions of the legislation leads to the conclusion that the issue of whether an alleged irregularity in the conduct of a referendum had a "material" effect is to be considered only when an application for leave is brought before the Court. It must also be recalled that, while on the facts of this case the leave application was heard after the State was put on notice and at the same time as the substantive hearing, the 1994 Act contemplates that leave may be granted ex parte and having heard only those who might wish to challenge the provisional result. To suggest that, assuming leave is given, those who might wish to oppose the challenge and uphold the provisional result would be precluded, at the full hearing, from addressing the issue of materiality, would be to suggest that such parties would be denied a fundamental entitlement to be heard on an important part of the case.
3.6 Whatever may be the appropriate test to justify court interference with a provisional result, it involves some assessment of the material effect of the asserted irregularity. To require that an assessment of the effect of the irregularity be made once and once only at the leave stage, and thus, in a manner which did not give to those who wished to uphold the provisional result any opportunity to be heard, would be a significant breach of fair procedures.
3.7 To suggest that this could be cured by giving an opportunity to seek to have leave set aside on the grounds that materiality had not been properly established does not seem to me to provide an answer. It is true that it is always possible for a party who is affected by an order made ex parte to seek to have that order set aside. However, where the ex parte order is simply a means of introducing a filter to the commencement of proceedings which would, where leave is given, be followed by a full hearing, it is normally only considered to be appropriate to seek to revisit the ex parte order (rather than to contest the substantive hearing) where it can be suggested that the ex parte order was wrongly obtained by, for example, the placing of misleading evidence before the court or a failure in the duty of candour which lies on every applicant for an ex parte order. It would be open to significant doubt, in my view, whether it would be appropriate, at least in most circumstances, to seek to revisit an ex parte leave order by inviting the court to conduct what would, in substance, be a full hearing on the issues within the framework of an application to set aside leave.
3.8 In any event, even if such were possible, same would create an extraordinarily complicated procedural mire. As others have commented in their judgments in this case, it is of significant importance that a final determination of the result of a constitutional referendum is not held up by a lengthy court process. If the provisional result is to be confirmed then that should be done as quickly as possible so that the constitutional amendment approved by the people becomes effective without any delay. Likewise, if grounds are established for requiring a recount, a rerun or other intervention by the court, then those measures should be put in place as a matter of urgency to ensure that a final and definitive result is confirmed in the shortest possible time.
3.9 The interpretation urged on behalf of Ms. Jordan would run the risk of either or both of significant procedural unfairness, by excluding or minimising the entitlement of those supporting the provisional result to be heard on the question of materiality, or significant procedural over-complication almost inevitably leading to delay in reaching a final conclusion. As the process for amending the Constitution is itself a process with a high constitutional value, an interpretation which led to what I think can be described as, at a very minimum, giving rise to very significant difficulties, could only be favoured if no alternative interpretation were available. Even if, therefore, I had come to the view that there was some technical merit in the argument advanced on behalf of Ms. Jordan under this heading, I would, nonetheless, have favoured the interpretation adopted in the judgment of O'Donnell J. However, in any event, as noted earlier, in my view the proper construction can be arrived at by means of the analysis adopted by O'Donnell J. in the course of his judgment.
3.10 Before leaving this aspect of the case I wish to make one final observation. It is a point which may well have some relevance to other issues in these appeals as well. While there are many similarities between the voting process which is designed to elect public representatives, on the one hand, and that adopted in the conduct of referendums, on the other, there is, in my view, one important distinction.
3.11 Both processes involve the people exercising their constitutional entitlement to vote in the selection of their representatives or in determining the content of the Constitution. However, the election process also involves the rights and, potentially, obligations of those who seek office. While the electorate has an entirely legitimate and constitutional entitlement to vote on the question of who should be elected to office, it is fully open to the Oireachtas to put in place measures which impose obligations on those who seek office and which further provide, within certain boundaries which do not need to be explored for the purposes of this case, for the consequences of electoral misbehaviour, including the deprivation of office in respect of those in relation to whom significant electoral misbehaviour is established. There is, in my view, no reason at the level of principle why it may not be constitutionally possible to deprive a candidate of victory as a consequence of established serious misbehaviour in the electoral process even if it may not be clear that the relevant electoral misbehaviour actually affected the outcome of the election in the sense of affecting those who were elected.
3.12 However, different considerations apply in the case of a referendum. While it may, as a matter of practical politics, be possible to identify those, whether within the party political or elected representative system or otherwise, who support or oppose a particular referendum, the people have a right to have their views on the constitutional amendment in question accepted and enforced independent of the conduct of those who may support or oppose the referendum concerned. A referendum which the people genuinely support should not be denied passage even if some of those who support a yes vote have acted improperly. It is only where such improper actions have actually affected the outcome or, for reasons which I will briefly turn to, have potentially done so, that it is appropriate for the courts to intervene, for it is only in such circumstances that it can be said that there is a real risk that reasonable and informed members of the public would not actually trust the provisional outcome of the referendum as truly representing the will of the people. That analysis has some affect on the proper test to be applied. However, for present purposes, it is also important to emphasise that those who take a contrary view to a challenger to a provisional referendum result must be entitled to be assured that the case in favour of that provisional result will be fully and fairly heard. A referendum result should not be displaced simply because misconduct or error has been established. It should only be displaced where there is a basis for suggesting that what went wrong has actually affected the outcome.
3.13 The next major issue which arises in these cases is as to the test by reference to which that judgment should be formed. I, therefore, turn to that question.
4. The Test
4.1 It is important to emphasise that the types of irregularity which may potentially affect the result of any electoral process are many and varied. But so also is the extent to which it may be possible to analyse with precision the effect of any established irregularity. Some irregularities involve purely technical or “mechanical” questions concerning aspects of the casting of ballots, the counting of votes or the assessment of whether a ballot paper should properly be included in the count. In many such cases it will also be possible, if the court is persuaded that there was an irregularity, to determine the precise effect of that irregularity with absolute precision. If votes have been counted which should not have been included then the relevant ballot papers can be looked at and an exact adjustment made. Likewise, if votes have been excluded which should have been taken into account, a similar adjustment can be made.
4.2 In an intermediate category, cases may be found in which there has been an established irregularity in relation to votes cast, or which should have been cast, but where it is not possible to tell with precision how the relevant votes would have been cast had it not been for the irregularity. Examples in this category would be cases where voters were wrongfully deprived of the opportunity to vote, or where votes were cast but for some reason were not included in the count, and where the relevant votes are no longer available to ascertain how they were, in fact, cast. In such cases, it will normally be possible to identify, with some reasonable level of precision, the maximum extent of the possible effect of the irregularity. It should be possible, for example, to identify the number of voters wrongly excluded from the opportunity to cast their ballot or the number of votes actually cast which can no longer be found. These issues raise the so-called “magic number” question which has been addressed in some of the Canadian jurisprudence to which we have been referred, and which is cited in other judgments of the Court. In some cases the answer may be straightforward. If the maximum number of votes potentially affected by the irregularity is 1,000 and the margin at the end of the day is more than 1,000 then the irregularity could not have made a difference and no basis for upsetting the original result would exist. The more difficult question arises where the number of votes affected is less than the margin, but where, as a matter of practicality, it is unlikely, or even highly unlikely, that all of the relevant votes affected by the irregularity in question would have been cast in the same way, so as to change the result.
4.3 However, the third, and most difficult category of case is one in which the irregularity is not so much concerned with practical questions, such as legitimate voters being denied the opportunity to vote in a proper way, or votes which have been properly cast not being taken into account, but rather where improper activity has been established which might be argued to have affected the way in which people voted. Some mention in that context should be made of cases involving an allegation of personation or other forms of voter fraud. It may be possible to establish that certain votes were fraudulently cast and to invite the court to determine, for the purposes of assessing materiality, whether it is possible to say how those votes were cast and whether the removal of those votes from the process would affect the result. Such cases do not really fit into the third category which I have sought to identify. They cannot properly be described as cases where there has been, on the face of things, a proper ballot and a proper count, but where the potential problem may stem from the fact that people may have been improperly influenced in the way in which they actually voted.
4.4 While a case involving unconstitutional expenditure, such as the present, is an example of the third type of irregularity, there can be others. Cases of, for example, bribery come to mind. A result may have been declared on the basis of an unimpeachable process for the casting and counting of votes. However, it may be established that bribery or other electoral malpractice of that type was deployed. In such a case the effect is more difficult to establish, for it requires an assessment of how people might have voted were it not for the irregularity. There is, therefore, a very real distinction to be made. On the one hand there are cases which focus simply on whether all the votes which should have been cast and counted (but only those votes) were properly assessed, but which do not involve any assessment of the way in which any individual voter cast their ballot (except to the limited extent of determining whether the total number of votes affected by any irregularity could have made a difference). On the other hand, there are cases where the irregularity concerns whether votes which were, on their face, properly cast and counted, may have become tainted by a more general form of irregularity.
4.5 This case is clearly of that third kind. There is no suggestion but that the provisional result properly represents the calculation of all of the votes which were correctly cast and counted, and there is no suggestion that voters were improperly deprived of the opportunity to cast their votes or to have votes which were cast properly counted.
4.6 In such a case it must, at least to some degree, be uncertain as to what the precise result, in terms of votes, would actually have been had it not been for the irregularity. It is not simply a question of counting votes which were wrongly excluded, or estimating the maximum number of persons affected, for example, by a wrongful failure to open a polling station. It is a question of estimating how the electorate would have cast their votes had it not been for the relevant irregularity. Did people actually vote who might not have voted had it not been for the irregularity, or did people not vote who might have voted in its absence? Did people actually change their vote? All of these, and doubtless more, are possible consequences of irregular expenditure, but they are in the nature of hypothetical considerations. The Court is being asked to assess what would, hypothetically, have happened in the event that the irregularity concerned had not occurred.
4.7 Against that background, I agree with the analysis to be found in the judgment of O’Donnell J. as to why the test proposed in his judgment properly meets the constitutional values at stake. A case in which a reasonable and informed citizen could not reasonably be expected to trust that the result would necessarily have been the same in the absence of whatever irregularity is established is a case in which the provisional result cannot be said to necessarily represent the will of the people properly exercised in accordance with law.
4.8 To the reasoning of O’Donnell J. I would merely add one further observation. The test of the balance of probabilities which is used in all civil proceedings applies in its pure form to the assessment by a court, or other decider of fact, of events which actually did or did not occur, but where uncertainty arises as to the facts because the evidence is weak, inconclusive or contradictory. In such a case, the decider of fact is required to assess all of the available evidence and to decide what actually happened on the balance of probabilities. But in such cases the court, or other decider of fact, seeks to reach a conclusion about an event which did or did not actually occur. In cases where a court has been asked a somewhat different and hypothetical question about what might have been, a somewhat different approach has sometimes been adopted. Where a court is, for example, asking itself as to what would have happened in the event that a particular act of wrongdoing had not taken place, the court is, by definition, asking itself a hypothetical question. The wrongdoing did occur and has been established. In order to assess the appropriate remedy, the court is required to compare what actually happened as a result of that wrongdoing with what might, hypothetically, have happened if the wrongdoing had not occurred. In some cases the answer may be obvious and straightforward. If the car had not crashed, the plaintiff would not have broken their leg and would not have suffered an interruption in their work with consequent loss of wages. If the defendant had supplied the contracted goods, then the plaintiff would have had those goods for sale and would, doubtless, have been able to secure the market price and achieve whatever profit that reflected.
4.9 But other cases may be more complex. What would the patient or client have done had they been properly advised by their lawyer, accountant or doctor? In many such cases there may be a range of possibilities, no single one of which may even be probable (in the sense of being more than 50% likely). Many other examples could be given. In such cases, the courts frequently adopt an analysis which accepts that there is a range of possible consequences for a hypothetical event, all of which need to be properly taken into account in assessing the overall consequence. For a brief discussion of this question note might be taken of the comments which I made, in my article, writing extra judicially, "Frances E. Moran Memorial Lecture: ‘Do Hard Sums Make Bad Law? Mathematics in the Courts’", (2014) 37 D.U.L.J 1.
4.10 It is not, of course, the case that the standard of proof to be applied in civil proceedings is ever anything other than the balance of probabilities. However, it is important to analyse carefully the precise question which needs to be resolved by reference to that standard. Where the question is as to whether a disputed fact is correct then the application of that standard in its pure form is easily understood. However, where the question raises the range of hypothetical possible consequences of an event which did not, in fact, occur, then the question itself may be different, and may involve the court assessing a range of realistically possible consequences. A court may, in such circumstances, reasonably conclude on the evidence that there was, for example, a realistic possibility of a particular course of action having been adopted by a party to whom negligent advice was given. That possibility may fall short of a probability but nonetheless the client or patient may still be entitled to damages.
4.11 Without pushing the analogy too far, it seems to me that a like consideration arises in respect of the third category of voting irregularity which I have earlier identified. Assuming that the irregularity is established, then the question which must be asked is as to what would have happened had the irregularity not occurred. That is necessarily a hypothetical question. The irregularity did occur and people voted whatever way they did in the light of that irregularity. How they would have voted had it not been for the irregularity is necessary hypothetical, and it is, equally necessarily, the case that there is a range of possible outcomes to such analysis, no one of which is “probable”. The test proposed by O'Donnell J. is not one which requires a pure balance of probabilities approach because there is, in reality, a whole range of possible consequences which may be capable, within a margin of error, of being assessed, but none of which actually involves an assessment of facts which either did or did not occur.
4.12 It might, of course, be possible to attempt to answer the question as to whether more than 50% of the weighted range of possible consequences would have resulted in a different outcome to the referendum to the one contained in the provisional result. But any such analysis would, necessarily, involve questions of margins of error and statistical significance. The extent to which the absence of perfect information (a point to which I will return) might increase the range of potential error adds a further complication. To test this question, one might envisage a case where there was an unimpeachable set of survey data available which all experts agreed gave rise to a 51% probability that the relevant irregularity did not affect the result (in the sense of turning a yes into no or a no into yes) and a corresponding 49% probability that it did affect the result. Should such a result stand? On one view, it is probable (even if only just) that the result was not affected by the relevant irregularity. But another way of looking at exactly the same figures is to say that an informed citizen would be faced with being asked to trust the result of a referendum where we cannot approach the issue of the effect of the relevant irregularity with precision, and where the best we can do is to say that there was only a 49% chance that the irregularity changed the result. In my view, it is sufficient to state that proposition to understand why it cannot represent the proper test.
4.13 On that basis, I have no hesitation in agreeing with the test as formulated in the judgment of O'Donnell J. So far as the application of that test to the facts of this case is concerned, I have little to add to the analysis found in other judgments of the Court. I would only emphasise one point. The margin in this case was very significant. The evidence concerning the effect of the undoubtedly unconstitutional governmental expenditure was limited. But it is important to emphasise that, in a case where it was otherwise, and the evidence of effect greater and the margin smaller, a very different finding could be reached. I do, in the context of the evidence in this case, however, wish to make some brief observations on the use of survey evidence in litigation.
5. Survey Evidence
5.1 One of the significant disputes which emerged before the High Court, and which was the subject of significant debate before this Court, concerned the use of survey evidence coupled with the expert analysis of that survey evidence by political scientists. It is not my intention to comment on all of the issues under that heading which were canvassed both at the trial and on appeal. Those issues have been addressed both in the judgments of the trial judge and in other judgments of this Court on these appeals. However, there are a small number of observations which I wish to make.
5.2 First, I should record that it does not seem to me that there is any reason in principle why survey evidence and expert views derived from that survey evidence cannot form an appropriate evidential basis for certain types of finding of fact. While there may well be a legitimate debate as to the extent to which the sort of survey evidence frequently deployed in the social sciences can be compared to the type of experimental evidence used in the physical sciences, it must be recalled that courts should have regard to any potentially probative and admissible evidence which is material to any question of fact which the Court needs to determine to decide the case before it. Ordinarily, questions concerning the reliability of scientific evidence (whether of general theory or specific to the facts of the individual case) go to the weight to be attached to that evidence in reaching an overall conclusion. If, and to the extent that, a court is satisfied that survey evidence can have a bearing on a factual decision which the court is required to make, then that evidence should be considered. To the extent that there may be legitimate questions over the survey methodology, the interpretation of the results of the survey or the conclusions which can be reached as a result of the relevant findings, those are matters which affect the weight which the court should attach to the evidence in question and may, in that way, have a significant effect on the ultimate outcome of the court's factual findings.
5.3 But in that regard it does not seem to me that survey evidence is any different to any other form of forensic or experimental evidence. The circumstances in which forensic evidence was obtained or maintained, the methodology used in its collection or analysis, the scientific basis for its interpretation and the expert views as to what can be gleaned from the relevant information are all matters which may, to a greater or lesser extent, be the subject of debate in any case in which expert evidence based on an analysis of physical or other materials may be of assistance in reaching a conclusion on the facts.
5.4 But a second matter, which must also be borne in mind, is that it is important to keep clearly in focus the precise question with which the court is concerned in any particular case. Many disciplines (not least the social sciences) have appropriate protocols for the conduct of research and for reaching conclusions based on that research. It would, indeed, be foolish for any court which is asked to take such evidence into account not to have regard to whether relevant protocols or recommended practises were complied with. But it must be recalled that the task of a court is not the same as the task of those charged with considering whether, for example, an article should be published in a learned peer review journal. A court is required to make a decision on the facts and is invited to take certain evidence into account in assessing those facts. The precise, factual question which the court may have to address may not raise exactly the same as the issue which might typically be required to be addressed in a journal article. A lack of compliance with what some might consider to be best practice might legitimately lead to an article which reflected the results of relevant research not being accepted for publication. But provided that a court is of the view that the findings were of some assistance in determining facts which needed to be found to decide the case, then the court should consider those findings while, of course, taking into account, in considering the weight to be attached to same, any substantiated criticism of the methodology or analysis.
5.5 It must also be acknowledged that the precise factual question which the court has to answer can have a very significant bearing on the extent to which any legitimate criticism of the methodology or methods of analysis might lead to a significant devaluation of the expert conclusions reached for the purposes of the case in question. If, for example, a court was required to be satisfied that there was a more than a 50% likelihood that a particular state of affairs persisted, then it might well be necessary to conduct a very rigorous statistical analysis to ascertain whether that threshold had been reached. It might, in that context, be necessary to resolve debateable questions over the precise statistical methodology which should be adopted. Questions concerning the extent to which the raw data was appropriate or sufficient to allow a sufficiently rigorous analysis to be conducted might also loom large. As in all other cases, the question of the onus of proof may, where the evidence is sparse, prove to be decisive. A legitimate criticism as to the methodology adopted might lead to a finding that the party on whom the onus of proof rested had failed to discharge that onus.
5.6 In the context of this case, the evidence which might be regarded as sufficient to establish that, as a matter of mathematics, it was more than 50% likely that the unconstitutional expenditure of the government changed the result of the referendum from a no to a yes could be very different from the kind of evidence which would be material in deciding whether the test which O'Donnell J. has set out in his judgment has been met.
5.7 That leads to the last point. Much of the criticism which was to be found in the evidence of Professor Marsh as directed towards the evidence presented on behalf of Ms. Jordan was concerned with the ideal conditions in which one might structure survey evidence and data in order to reach conclusions on the materiality of the unconstitutional government expenditure which occurred in this case. But it must be recalled that every day, in all sorts of cases, and in relation to all sorts of evidence, courts are called on to do the best they can with the evidence which happens to be available. Doubtless it would make the determination of liability in motor accident cases a lot easier, and rarely controversial, if the circumstances of all accidents were clearly caught on CCTV. Doubtless an unfortunate but explained gap in the records of events may make it much harder to assess business conduct or professional activity which may be the subject of litigation. Very many other examples could be given. But in the messy world of real litigation, courts rarely have the luxury of having available to them evidence obtained in what might be called the real world equivalent of laboratory conditions. But courts are nonetheless required to do the best they can in assessing such evidence as may be available. Of course, it remains the case that it is open to a defendant to assert that the plaintiff has just not put forward sufficient evidence to discharge the onus of proof. Survey data and its analysis are potentially as subject to that type of criticism as any other type of evidence. It may be said that it just falls so far short of the kind of materials on which any sort of safe conclusion could be reached so that the party presenting the evidence in question has failed to meet the onus of proof. But the fact that the evidence falls a long way short of the ideal which might be put in place, in advance, by experienced professionals in the social sciences, does not mean that the evidence may not, nonetheless, be sufficient, in an appropriate case, to discharge the onus of proof.
5.8 That question, in the context of survey data and analysis, is no different to that which arises in the context of any other kind of evidence. The question is not whether there might, theoretically, have been better evidence. There almost always will. The question is whether there is enough.
5.9 In the light of those general observations, I wish to conclude with a few points concerning their application to the facts of this case. First, I should say that I have no reason to doubt but that Professor Marsh was correct when he said that the proper way to conduct a rigorous analysis of the effect of an intervention, such as the government expenditure in this case, would be to have a series of surveys from which the impact of the relevant expenditure might be inferred. But to impose such an obligation would be to require a person to anticipate a breach of the Constitution in the course of a referendum campaign. How else would a potential challenger be able to have the evidence which Professor Marsh identified as being ideal unless they had decided, just in case, to commission the relevant laboratory condition surveys in advance? Such surveys might well be the appropriate standard for a professional political scientist setting up, in advance, a project designed to analyse the impact of various events in the course of an election or referendum campaign. A failure to adopt such practice might well be a legitimate criticism of someone who set themselves the task, for research purposes, of conducting such an analysis, but without putting in place the appropriate periodic surveys. But that is not the question with which a court is concerned. The question which a court must answer is whether the available evidence is sufficient to discharge the burden of proof.
5.10 What happened in this case is, of course, that, short of Ms. Jordan commissioning an ex post facto survey of her own, the only raw data available to her experts was the information available from the Referendum Commission. Hardly surprisingly, the focus of the data collected on behalf of the Referendum Commission was designed to assist it in determining whether the various programmes which it had adopted in fulfilment of its statutory remit, to inform the electorate on the referendum and to encourage citizens to vote, had been successful. Using data for a purpose other than that for which it was collected always runs some risk (to a greater or lesser extent depending on the circumstances) of encountering difficulty stemming from the data not being ideal or even fit for purpose. But the fact that the data in question was not ideal did not, in my view, mean that it could not properly be taken into account and weighed in the balance in considering whether it had been established that there was a material effect.
5.11 In that context, I should touch on the constitutional issue. Put at its simplest, the case made on behalf of Ms. Jordan was that the interpretation which the State sought to place on the burden which she bore to establish material effect was so great as to render it impossible, or virtually impossible, or excessively expensive to show materiality even where there was, as here, a clearly established breach of electoral law. It does seem to me to be worthy of some comment that if the Court had been persuaded that the only way in which materiality could be established was in the manner suggested by Professor Marsh, there might well have been very considerable merit in Ms. Jordan's constitutional challenge. Such a level of proof would have rendered a challenge to the result of a referendum, based on the sort of unconstitutional expenditure arising in this case, close to impossible if the standards identified by Professor Marsh were the only ones which could produce evidence sufficient to prove materiality. If a different conclusion as to the test for materiality had been reached and if it had, in that context, been accepted that materiality could only be established in the manner identified by Professor Marsh, then I might well have come to a very different view on the issue of the constitutionality of the legislation.
5.12 Finally, I should comment that it is important that it be kept in mind that expert evidence of the type debated in this case is no different, at least at the level of high principle, to any other type of evidence which may need to be considered on appeal. It will always be possible to argue that the evidence does not bear any inference drawn. It will always be possible to point to any clear error in the way in which the evidence was approached and assessed. It will always be possible to argue that the evidence is insufficient to meet the burden of proof. But where there are competing but legitimate professional views on a topic, then it is really a matter for the trial judge to reach an assessment on those questions. While it is often emphasised that a trial judge has the benefit of hearing witnesses whose credibility may be in question giving evidence, it seems to me that there is another aspect of the proper reluctance of appellate courts to interfere with findings of fact of a trial judge which needs to be emphasised. Where there are competing professional views, the trial judge will have had the benefit of hearing those competing views teased out and challenged, frequently over a significant period of time. While this Court will, of course, have a transcript, there can be little doubt but that a trial judge is, nonetheless, in a better position to reach a conclusion as to which professional view is to be preferred. In the absence, therefore, of established error or findings which are not sustained by the evidence at all, the appropriate reluctance of appellate courts to interfere with findings of fact applies equally in the context of expert evidence such as that tendered in these cases as in respect of any other evidence.
6.1 As noted earlier the purpose of this concurrence is to set out some of the reasons why I fully agree with the judgment of O'Donnell J. delivered today and not least the test for determining the materiality of the effect of established irregularity which he proposes at para. 85 of his judgment. I also add some observations on the use of survey evidence in a case such as this.
6.2 I would finally add that it is important to emphasise that the test identified by O'Donnell J. is one which is far from impossible to meet. The problem in this case was that the margin, at the end of the day, was quite significant, and the evidence concerning the effect of unconstitutional government expenditure did not suggest that there was any realistic basis for fearing that the expenditure concerned could have altered the result.
6.3 However, it requires to be said that, in a different case, the Court might be faced with a narrower result coupled with stronger evidence of the effect of unlawful expenditure. In such a case I have no doubt but that it would be possible to conclude that the hypothetical reasonable and informed observer could not be expected to trust that the result might not have been changed by unconstitutional activity. If such a case were to occur then it would not only be open to this Court, but it would be the duty of the Court, to take action and, for example, require that a poll be reconducted. Such a consequence would, undoubtedly, be highly undesirable. In order that it be avoided it is important that every conceivable care be taken to ensure that a referendum is properly conducted in all respects.