England and Wales Court of Appeal (Civil Division) Decisions

IN THE COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE QUEEN'S BENCH DIVISION
ADMINISTRATIVE COURT
THE HON MR JUSTICE OUSELEY
[2018] EWHC 3520 (Admin)

		Royal Courts of Justice Strand, London, WC2A 2LL
		04/03/19

Lord Justice Hickinbottom:

Introduction

1. On 29 March 2017, under the power granted to her by the European Union (Notification of Withdrawal) Act 2017 ("the 2017 Act"), the Respondent Prime Minister notified the European Union ("the EU") of the United Kingdom's intention to withdraw from the EU under article 50(2) of the Treaty on European Union. In this claim for judicial review, the Applicants contend that the Respondent's decision to notify and the notification itself were unlawful because they were based upon the result of a referendum that was itself unlawful as a result of corrupt and illegal practices, notably offences of overspending committed by those involved in the campaign to leave the EU. Alternatively, it is said that the Respondent erred in law in not responding to the subsequent evidence of those practices as it emerged.
2. On 10 December 2018, following a full day's hearing, Ouseley J refused permission to proceed with the judicial review on the basis of both delay and want of merit, and ordered the Applicants to pay the Respondent's costs summarily assessed in the sum of £17,256.
3. The Applicants applied for permission to appeal that Order. That application came before this court on 21 February 2019 when, after hearing substantial argument, we indicated that we would refuse permission to appeal on all grounds and would give our reasons at a later date. In this judgment, I set out my reasons for that refusal.
4. Before us, Jessica Simor QC and Pavlos Eleftheriadis appeared for the Applicants; and Sir James Eadie QC and Joseph Barrett for the Respondent. At the outset, I thank them all for their respective contributions.

Background Facts

5. The European Union Referendum Act 2015 ("the 2015 Act") provided for a referendum on the UK's membership of the EU ("the EU referendum"). The referendum, although only advisory, was thus made an "an integral part of the process of deciding to withdraw from the EU" (Schindler v Chancellor of the Duchy of Lancaster [2016] EWCA Civ 469; [2017] QB 226 at [19] per Lord Dyson MR).
6. With necessary modifications, the 2015 Act adopted the existing regulatory framework for elections and referendums set out in the Political Parties, Elections and Referendums Act 2000 ("the 2000 Act"), which lays down rules for participation, spending limits and permissible donations, including rules on returns to be submitted to the Electoral Commission in relation to spending and donations, the breach of which constitutes a criminal offence. The Electoral Commission was established by section 1 of the 2000 Act, as the regulator with statutory powers to set and enforce standards in relation to elections and referendums, including the regulation of political finances and campaign spending. Amongst other things, it considers and determines whether there have been breaches of the relevant regulations against the criminal standard of proof. There is a right of appeal against the Commission's findings, on law and facts, to the Central London County Court (paragraph 6(6) of Schedule 19 to the 2000 Act).
7. Where requirements are breached in a binding public vote (e.g. in an election), then in two particular sets of circumstances the Representation of the People Act 1983 ("the 1983 Act") imposes identified consequences so far as the result of the vote is concerned. Section 159(1) deals with the situation where a candidate or his agent has himself been guilty of corrupt or illegal practices:

"If a candidate who has been elected is reported by an election court personally guilty or guilty by his agents of any corrupt or illegal practice his election shall be void."

Section 164(1)(a) covers the circumstances in which an election might be void for general corruption:

"Where on an election petition, it is shown that corrupt or illegal practices or illegal payments, employments or hirings committed in reference to the election for the purpose of promoting or procuring the election of any person at that election have so extensively prevailed that they may be reasonably supposed to have affected the result… his election … shall be void…".

Section 164(2) provides that:

"An election shall not be liable to be avoided otherwise than under this section by reason of general corruption, bribery, treating or intimidation."

8. However, the 2015 Act did not apply those provisions to the EU referendum, which was advisory only. Paragraph 19 of Schedule 3 to the 2015 Act, under the heading, "Restrictions on challenge to referendum result", simply provided:

"(1) No court may entertain any proceedings for questioning the number of ballot papers counted or votes cast in the referendum as certified by the Chief Counting Officer or a Regional Counting Officer or counting officer unless—

(a) the proceedings are brought by a claim for judicial review, and

(b) the claim form is filed before the end of the permitted period.

(2) In sub-paragraph (1) 'the permitted period' means the period of 6 weeks beginning with—

(a) the day on which the officer in question gives a certificate as to the number of ballot papers counted and votes cast in the referendum, or

(b) if the officer gives more than one such certificate, the day on which the last is given."

Paragraph 133 of the Explanatory Note accompanying the 2015 Act said of that paragraph:

"Paragraph 19 relates to how the result of the referendum may be challenged in legal proceedings. It provides that any challenge in respect of the number of ballot papers counted or votes cast as certified by the Chief Counting Officer, a Regional Counting Officer or a counting officer must be brought by way of judicial review (sub-paragraph (1)(a)). In addition, the challenge must be commenced within six weeks of the date of the relevant certificate (sub-paragraphs 1(b) and (2)). The six-week period is intended to ensure that sufficient time is allowed for challenges to be brought while avoiding prolonged delay in the final result of the referendum being known"

However, the 2015 Act did not expressly specify any further legal consequences of any breach of the rules so far as the effect of the EU referendum result was concerned.

9. The EU referendum was held on 23 June 2016. The result was declared the following day, the UK voting 51.9 percent (17,410,742 votes) to 48.1 percent (16,141,241 votes) in favour of leaving the EU. Although the EU referendum was only advisory, the Conservative Party Manifesto had said that the result would be honoured; and, following the referendum result, the Government announced its intention to give notice of withdrawal from the EU.
10. The Government intended to give that notice under Royal Prerogative powers; but its ability to do so was challenged on the basis that notice to leave the EU could not be given without an Act of Parliament. That challenge was successful in the Divisional Court and then, on appeal by the Secretary of State, in the Supreme Court (R (Miller) v Secretary of State for Exiting the European Union ([2016] EWHC 2768 (Admin) affirmed [2017] UKSC 5; [2018] AC 61). Miller thus confirmed that Parliament retained control over the ultimate decision as to whether the UK left the EU.
11. As a result, on 16 March 2017, Parliament passed the 2017 Act. Section 1(1) provided that:

"The Prime Minister may notify, under Article 50(2) of the Treaty on European Union, the United Kingdom's intention to withdraw from the EU."

By that provision, as confirmed in R (Webster) v Secretary of State for Exiting the European Union [2018] EWHC 1543 (Admin), Parliament gave the Prime Minister the power to give notification of withdrawal and a discretion as to when to give notice.

12. Under the power granted by the 2017 Act, on 29 March 2017, in a letter to the President of the European Council, His Excellency Donald Tusk, the Respondent gave notification of withdrawal.
13. Since the notification, on 26 June 2018 Parliament has enacted the European Union (Withdrawal) Act 2018 ("the 2018 Act") repealing the European Communities Act 1972 on "exit day". Subject to section 20(4), which permits the date to be amended by regulation "to ensure that the day and time specified in the definition are the day and time that the Treaties are to cease to apply to the United Kingdom", "exit day" is defined as 29 March 2019. The lengthy section 13, headed "Parliamentary approval of the outcome of negotiations with the EU", requires Parliament to approve any negotiated withdrawal deal or a withdrawal without a deal. That again illustrates that, whilst the executive government has been undertaking the withdrawal negotiations, Parliament has retained ultimate control over the UK's withdrawal from the EU.
14. The present challenge focuses on alleged misconduct in the EU referendum by various permitted participants who campaigned for leaving the EU, notably the designated lead campaigner Vote Leave Limited ("Vote Leave"); another leave campaigner Leave.EU Limited ("Leave.EU"), its director Arron Banks and its campaign organiser Better for the Country Limited; and Darren Grimes as well as BeLeave, an unregistered campaigner with which Mr Grimes was associated. The essential alleged misconduct comprised of exceeding campaign spending limits and rules as prescribed by the 2000 Act.
15. I pause to note that, in March 2017, the Electoral Commission published its "Report on the regulation of campaigners at the referendum on the UK's membership of the [EU] held on 23 June 2016", which noted that campaigners registered for the UK to remain in the EU reported an aggregate spend of £19.3m and those registered to campaign for the UK to leave the EU reported a spend of £13.3m. Remain campaigners therefore reported spending about £6m more than leave campaigners. However, of course, "overspend" in the context of the EU referendum campaign refers not to absolute or comparative figures, but to money spent in excess of that allowed under the relevant referendum rules.
16. The Electoral Commission carried out initial investigations into Vote Leave's spending. The first investigation closed on 4 October 2016; and, although there is no direct evidence as to when it was opened, as Ouseley J indicated (at [15] of his judgment), it must have been shortly after the referendum. Concerns were therefore expressed about the leave campaign's spending very soon after the referendum had been held. The opening date for the second investigation – which was somewhat wider in scope – is also unevidenced, but it concluded on 21 March 2017. In respect of each investigation, the Electoral Commission found that there were no reasonable grounds to suspect that there had been any incorrect reporting of campaign donations or spending. That was the position when the withdrawal notice was given on 29 March 2017.
17. However, that was not the end of the matter. First, the Good Law Project judicially reviewed the conclusion of the Electoral Commission that there had been no incorrect reporting of campaign finances so far as Vote Leave was concerned, on the basis that certain payments it had made in respect of campaign expenses incurred by Mr Grimes were referendum expenses incurred by Vote Leave and should therefore have been included in its expenses return. The Electoral Commission had in fact advised Vote Leave that these need not be included. A Divisional Court (Leggatt LJ and Green J as he then was) concluded otherwise (R (Good Law Project) v Electoral Commission [2018] EWHC 2414 (Admin); [2019] 1 All ER 365. The Commission are currently seeking to appeal that decision.
18. Furthermore, following the commencement of the Good Law Project challenge, the Electoral Commission announced that it would reopen the investigation. In fact, it opened three further investigations.
19. The first to report, on 11 May 2018, was an investigation into Leave.EU and its associates. As a non-designated campaigner, its spending limit was £700,000. It had reported spending of just under that figure. However, the Electoral Commission found that it had failed to include a minimum of £77,380 in its spending return being the fees it had paid Better for the Country as its campaign organiser, so that it had overspent by at least 10%. In addition, there were other transactions which probably represented some further overspending, and some loans were not properly reported. It identified a number of offences that appeared to have been committed by both Leave.EU and its responsible person, Ms Elizabeth Bilney. Those findings are currently the subject of an appeal to the County Court. The Commission also imposed a fine on Leave.EU of £77,000 for multiple breaches of electoral rules; and referred Ms Bilney and the campaign generally to the police for further investigation.
20. Second, the Electoral Commission reported on its further investigation into Vote Leave and its associates on 17 July 2018, which focused on payments made in June 2016 to a Canadian data analytics firm called Aggregate IQ which had been the focus of one of the original investigations. On a different basis from that looked at in the original investigation, it found that there had been a breach of the regulations. Vote Leave, as the designated campaigner, had a spending limit of £7m. However, the Commission found that all of Mr Grimes' and BeLeave's spending on referendum campaigning was incurred under a common plan with Vote Leave, so that, for the purposes of Vote Leave's return, their spending should have been aggregated. The aggregate spend was just under £7,450,000. Further, the Commission made other findings adverse to Vote Leave and its associates, e.g. that BeLeave had exceeded its spending limit of £10,000 as an unregistered campaigner. The Commission fined, amongst others, Vote Leave £61,000 and Mr Grimes £20,000. I understand that Vote Leave has exercised its right to appeal to the County Court.
21. Third and finally, on 1 November 2018, the Electoral Commission published a further, short report confirming that it had reasonable grounds to suspect that Arron Banks (who was involved with Better for the Country), and Ms Bilney and others from Leave.EU and its associates had concealed the true details of various campaign financial transactions; and it had referred the matter and handed over the evidence to the National Crime Agency.
22. In the meantime, on 5 July 2018, in the light of leaks of the Electoral Commission's findings that found their way into its July 2018 report, the Director of Let's Take Back Control Limited trading as the Fair Vote Project ("the Fair Vote Project") wrote to the Respondent seeking either another referendum or at least an independent investigation into the funding of the leave campaign. The letter sought a response by 12 July; and, no response having been received by that date, on 20 July 2018 solicitors on behalf of the Fair Vote Project wrote a pre-action protocol letter to the Respondent, challenging both the decision to give the withdrawal notification, and the decision not to take action in the light of the evidence of corruption and illegality in relation to the referendum that had come to light. On 31 July 2018, solicitors on behalf of the Applicants wrote their own pre-action protocol letter adopting Fair Project Vote's grounds. The Respondent responded to those grounds through the Government Legal Department on 10 August 2018. That letter refused any of the relief sought.
23. This claim for judicial review was issued on 13 August 2018. In addition to various declarations, the relief sought included orders quashing both the Respondent's decision to serve the notification and the notification itself. On 21 September 2018, Supperstone J refused permission to proceed on the papers, on grounds of delay and on the merits. In relation to the latter, he said this (paragraph 4):

"(a) The Claimants' case is that the Prime Minister's decision to give Article 50 notification is rendered unlawful because of the conclusions of the Electoral Commission published in May and July 2018. The Electoral Commission made findings that some bodies and individuals involved in the Referendum campaign breached spending limits or committed other breaches of campaign financing requirements. The Claimants contend that the Referendum result is vitiated by reason of such conduct which would fall within the definition of 'corrupt and illegal practices' in the Representation of the People Act 1983 and other legislation, and that the decision of the Prime Minister and notification are vitiated for the same reason.

(b) The Prime Minister cannot be said to have acted unlawfully in making the decision to give notification on 29 March 2017 because of the findings of the Electoral Commission that were not made or published until May/July 2018.

(c) The Claimants acknowledge that at the date of the Prime Minister's decision allegations that there had been breaches of campaign finance limits during the Referendum campaign had been in the public domain for some time, and that such allegations were being investigated by the Electoral Commission. The relevant facts were matters of public knowledge. The contention that the decision to give Article 50 notice was founded on a premise that there had been compliance with campaign finance requirements is therefore incorrect.

(d) Neither of the Electoral Commission reports establish that the breaches of campaign finance or other requirements identified in the reports mean that the result of the Referendum was 'procured by fraud', or that the outcome of the Referendum was affected by any wrongdoing or unlawful conduct.

(e) The decision that the UK will leave the EU on 29 March 2019 has now been approved by Parliament. The issues raised by this claim are therefore academic."

24. The Applicants sought reconsideration of that decision at an oral hearing, which Ouseley J heard on 7 December, delivering his judgment on 10 December 2018. He refused the application and awarded the Respondent her costs. It is that order which the Applicants now seek to appeal.

The Grounds of Challenge: Introduction

25. The Applicants rely upon seven grounds of appeal: four go to the contention that Ouseley J erred in relation to the merits of the claim, two go to his conclusion that in any event the claim is too late, and one goes to the contention that the costs order he made was in any event wrong.
26. I propose dealing with the grounds relating to the merits first. The merits threshold at this stage in a judicial review claim is low, namely that there is an arguable ground of review which has a realistic prospect of success, which thus warrants a full investigation at a substantive hearing with all parties being given an opportunity to file relevant evidence and make submissions. This being an application for permission to appeal, the threshold is equally low.

Ground 1: Invalidity of the EU Referendum at Common Law

27. As her first ground, Ms Simor submits that the EU referendum, upon which the decision to notify was made, failed to comply with the requirements of the common law; and therefore it could not properly be taken to express the democratic will of the people which the common law was designed to uphold. She thus originally sought a declaration that the EU referendum result was "vitiated". However, during the course of argument before us, Ms Simor reconstructed the relief she sought into a declaration that the referendum involved corrupt and illegal practices such that, had it been subject to the relevant provisions of the 1983 Act (see paragraph 7 above), it would have been void. As the decision to notify withdrawal and the notification itself were based on the referendum result, the Applicants also seek an order quashing both.
28. Ms Simor's main argument ran as follows.

i) As I have explained, other than paragraph 19 of Schedule 3 to the 2015 Act (which does not apply in this case), Parliament did not prescribe any consequences for breaches of the donation and spending rules governing the EU referendum.

ii) However, Ms Simor submitted that the fact that there are no statutory consequences does not exclude or otherwise affect the jurisdiction of the court to apply long-established common law rules in such circumstances.

iii) At common law, there is not just a right to vote, but a fundamental right to vote in accordance with "the modalities of the expression of democracy" (see Moohan v Lord Advocate [2014] UKSC 67; [2015] AC 901 per Lord Hodge JSC at [34]), i.e. a right to participate in a vote in accordance with the relevant regulations that apply and not in a vote that was (e.g.) the subject of corrupt and illegal practices.

iv) By reference to such cases as Ashby v White (1703) 1 Smith's LC (13th ed) 253; 2 Ld Raym 938; 92 ER 126, Faulkner v Elger and Newby (1825) 4 B&C 449, Bradford Case (No 2) (1869) 1 O'M&H 35; 19 LT 723 and Morgan v Simpson [1975] 1 QB 151, she submitted that, in respect of a binding election, the common law considered that a vote was void if either (i) the candidate or his agents were guilty of any corrupt or illegal practice, or (ii) where there was no such guilt, but the corrupt or illegal practices committed in reference to an election to promote or procure the election of a candidate has so extensively prevailed that they may reasonably be supposed to have affected the result. The common law would not want for a remedy, and so, in an appropriate case, it must be open to a court of law to quash or declare a vote void.

v) Ms Simor submitted that the common law position was helpfully set out in the judgment of Lord Denning MR in Morgan v Simpson, which concerned the issue of whether a local election was voided as a result of certain ballot papers not having been stamped with the official mark by the polling station. That turned on the construction of section 37(1) of the Representation of the People Act 1949 (the predecessor of the relevant provisions in the 1983 Act) which provided:

"No local government election shall be declared invalid by reason of any act or omission of the returning officer or any other person in breach of his official duty in connection with the election or otherwise of the local elections rules if it appears to the tribunal having cognizance of the question that the election was so conducted as to be substantially in accordance with the law as to elections and that the act or omission did not affect its result."

Having reviewed the authorities which had considered that section and its predecessors, Lord Denning distilled the following three propositions from them:

"1. If the election was conducted so badly that it was not substantially in accordance with the law as to elections, the election is vitiated, irrespective of whether the result was affected or not....

2. If the election was so conducted that it was substantially in accordance with the law as to elections, it is not vitiated by a breach of the rules or a mistake at the polls - provided that it did not affect the result of the election….

3. But, even though the election was conducted substantially in accordance with the law as to elections, nevertheless if there was a breach of the rules or a mistake at the polls - and it did affect the result - then the election is vitiated…".

These (Ms Simor submitted) effectively set out the requirements of the common law which Parliament has simply adopted and codified in a succession of Acts ending with the 1983 Act (see paragraph 7 above).

vi) Although for binding elections, the consequences of corrupt and illegal practices to the validity of an election at common law have been codified in statute – originally in the Ballot Act 1872, but now in sections 159 and 164 of the 1983 Act – which effectively exclude the common law (expressly in section 164 of the 1983 Act), Ms Simor submitted that the common law still applies to non-binding votes such as the EU referendum not covered by that legislation.

vii) Therefore, where either limb of the common law rule applied, the result was void. In this case, she submitted, the referendum failed to satisfy the first – or, alternatively, the second – limb. It therefore failed properly to represent the democratic will of the people.

viii) Insofar as it failed to satisfy the first limb because the registered campaigners (including the designated campaigner) in the leave camp were themselves guilty of breaches of the relevant regulations, the referendum was unlawful even in the absence of any evidence of materiality, i.e. even if there was no evidence that the result would probably have been different if the breaches had not occurred.

29. Ms Simor submitted that, in this case, Ouseley J erred in not concluding that the EU referendum result failed to comply with the common law requirements for a valid vote; and, at least, the referendum should be declared unlawful or (on her redrafted declaration) declared as not satisfying the statutory provisions for a valid binding vote which do no more and no less than reflect the common law requirements.
30. However, I am unable to accept that submission as even arguable.
31. It is based on the following set of assumptions:

i) prior to the intervention of Parliament, the courts of law had a right at common law to hold and/or declare a binding vote invalid;

ii) the scope of the grounds on which that right could be exercised are now replicated in the 1983 Act; and

iii) that common law right and remedy can be transposed from elections that are binding (now taken over by statute) to the different context of the EU referendum which was advisory only.

In my view, these assumptions do not bear the weight of the arguments made by Ms Simor.

32. In respect of (i), the authorities upon which Ms Simor relied do not give her much, if any, support. Some of the cases concerned the individual right to vote. For example, Ashby v White concerned the issue of whether an action in law for damages by a citizen against a returning officer lay where the citizen had been denied a personal right to vote. It does not suggest that the court could interfere with the result of the vote. The report of Bradford Case (No 2) is short, and it concerns mainly what amounts to "treating" in an election. It does say that, at common law, an election would be void where there was general treating or bribery, but it seems that that was obiter as no sufficient treating was in fact found. It is unclear from the report – and the date of the case (given as merely "circa 1869") – whether the court was considering powers given to the courts by Parliament; but, in any event, I do not read that report as holding that the courts of law may set aside an election or referendum at common law.
33. Ms Simor relied most heavily on Faulkner v Elger; but that case concerned the nature of the custom by which, in a particular parish, the curate was elected, the Court of Common Pleas concluding that the Assize Judge (Gaselee J) was wrong to leave an issue as to whether the restriction of those entitled to vote to exclude those who had not paid church rates was either the custom or the custom as lawfully changed. It concerned parish custom, not elections conducted by or under the authority of Parliament. In any event, none of the judges went as far as finding that the election was void – they each expressly declined to do so, and they simply remitted the matter to the Assize Court.
34. Morgan v Simpson (which concerned the true construction of section 37(1) of the Representation of the People Act 1949) does not provide any support for the proposition that the courts have a freestanding jurisdiction outside that granted by Parliament to set aside elections and referendums. Far from it. As Lord Denning described in that case (at pages 161-2), before statutory intervention, polls were taken by open ballot; and any redress was not through the courts of law but through an election petition to Parliament. Petitions were first diverted to the High Court by the Parliamentary Elections Act 1868, and various statutes starting with the Ballot Act 1872 (and including the 2015 Act) have set out the (very restricted) circumstances in which the courts can interfere with the results of elections and referendums.
35. Relying on [63] and [67] of his judgment, Ms Simor submits that Ouseley J regarded it as arguable that there is a residual common law jurisdiction. It is not entirely clear that he did (see [65]). I am myself highly doubtful that there is. If there be such a power, Counsel before us were unable to find a single example of its exercise. The cases relied upon by Ms Simor related to binding elections and an individual's right to vote and are not to the point.
36. Whether, absent power granted by Parliament, the court has a residual common law power to interfere in an election in any circumstances is an issue which in my view we are not required to determine for the purposes of this application. In circumstances such as this – where, in paragraph 19 of schedule 3 to the 2015 Act, Parliament has set out the circumstances on which the court may interfere with the EU referendum result – there must a strong presumption that it did not intend the court to interfere otherwise. In any event, where, as here, the vote is only advisory, and Parliament has retained ultimate control over the question of whether the UK leaves the EU, I am entirely unpersuaded that the courts have any residual common law power to interfere.
37. But, even assuming hypothetically there were such a common law power, the criteria by which its scope is determined could not simply be read across from those adopted in an entirely different context by Parliament in (e.g.) the Representation of the People Acts as Ms Simor suggested. I agree with Sir James Eadie (and Ouseley J) that a minimum requirement for the exercise of any common law power in this new context of non-binding referendums would be that any breach of rules is material. It would be inconceivable for the common law to adopt a principle that requires or even enables a court of law to interfere with the democratic process where any breach of the voting rules is proved but not such as to affect the result; and, in my view, as Ouseley J properly said, in this case, there is no evidence that gives rise to any soundly based ground for believing the outcome of the referendum result would have been different if the breaches of the rules had not occurred (see [67] of his judgment).
38. There are two limbs to the materiality argument. First, before breaches could rationally be said to be material for the purposes of the court interfering with the democratic process, there must be certainty and finality in relation to the findings of misconduct. Whilst I accept that the Electoral Commission make findings to the criminal standard of proof, their relevant findings in respect of these matters are currently subject to appeal to the County Court – and other investigations, including police investigations, are on-going. Whilst of course Parliament or the executive could take steps on the basis of concerns that there has been wrongdoing, it is difficult to see how the court could require them to take action at a time when the facts are not fully and finally established.
39. In any event, even if the breaches are taken at their highest – i.e. as found by the Electoral Commission – there is no evidential basis for the proposition that they are material in the sense that, had they not occurred, the result of the EU referendum would probably have been different.
40. Before us, as a basis for the proposition that had there been none of the alleged breaches of the referendum rules then the result of the referendum would probably have been different, Ms Simor relied upon only (i) the fact that the overspend of Vote Leave was about 8%, and (ii) the swing required to reverse the result was only about 600,000 votes. However, that provides no evidential basis for her proposition.
41. On the issue of materiality, before us, Ms Simor did not seek to rely on Prof Philip Howard's report dated 30 November 2018 (and its addendum dated 6 December 2018). Prof Howard is Professor of Internet Studies at Oxford University and Director of the Oxford Internet Institute. In his opinion, without the extra (unlawful) expenditure by the campaign to leave on digital advertising it is "very likely" that the EU referendum result would have been different. However, the application to have that evidence admitted was issued on 3 December, for a hearing on 7 December 2018. Prof Howard's addendum report was not produced until 6 December 2018. The evidence was so late that Ouseley J refused to admit it; and Ms Simor does not now submit that he was wrong to do so. But, in any event, Ouseley J considered the report was not of any arguable weight because it was essentially speculative and based on propositions that were patently unsound (see [67]). Prof Howard's evidence sought to show that the EU referendum result would probably have been different if the overspend had not occurred, by reference to general voting behaviour and Vote Leave's digital strategy of identifying sympathisers and "persuadables" from analysis of Facebook pages, and then by applying a "conversion rate" of 0.1% to the likely attracted audience. On the basis that 80m Facebook users would have seen Vote Leave's social media advertisements, Prof Howard considered that 800,000 would have been converted and persuaded by the campaign, i.e. enough to overturn the Remain shortfall in the EU referendum. Ouseley J was, very clearly, unimpressed by this evidence. It is unnecessary for me to consider this evidence in any depth, because Ms Simor eschewed reliance on Prof Howard's evidence before us: but it should not be thought that I disagree with Ouseley J's view on the strength of this evidence. As Sir James Eadie observed, the evidence is clearly highly problematic.
42. Thus, there is simply no evidential basis for the proposition that the breaches, or any of them, are material in the sense that, had they not occurred, the result of the referendum would have been different. In my view, that is fatal to the assumption upon which this part of her argument is based.
43. However, Ms Simor's difficulties do not end there; because her third assumption is also clearly untenable. Even if, contrary to my view, there were a common law principle that a court of law could hold or declare a binding vote to be void if it fell within one of the two limbs I have described, it could not be assumed that that principle can simply be transferred across from binding elections to non-binding votes. An advisory referendum is a very different animal from a binding election. All of the authorities to which we were referred concerned binding elections. As Sir James Eadie submitted, there are fundamental objections to introducing a new duty of the kind suggested, which directly cuts across the will of Parliament as now expressed in (e.g.) the 2017 and 2018 Acts; and requires the court to enter a political arena to determine questions well outside its proper remit and ones which moreover it is ill-equipped to determine. I do not accept Ms Simor's submission that, in this case, she is relying upon uncontentious and well-established public law rights and remedies. In my view, Ouseley J was right to observe that, however it was sought to present the Applicants' grounds, at the heart of their case is an attack on the democratic process, i.e. on the referendum itself and its outcome (see [45] of his judgment).
44. For all those reasons, this first ground is unarguable.

Grounds 2 and 3: The Respondent's Decision to Notify

45. As Ground 2, relying on the line of jurisprudence running from the judgment of Carnwath LJ (as he then was) giving the judgment of this court in E v Secretary of State for the Home Department [2004] EWCA Civ 49; [2004] QB 1044 at [66], Ms Simor submits that the Respondent's decision to notify withdrawal from the EU was unlawful, because it was based upon an error of fact, namely that the EU referendum had been lawfully conducted. In similar vein, as Ground 3, she submits that the Respondent's notification was ultra vires, because, in giving her the power to notify, Parliament must have intended her to have exercised the power only on the basis of a lawful "leave" referendum result.
46. However, in my view, these grounds too are unarguable.
47. First, it cannot be said that the Respondent erred in law in exercising the discretion given to her by Parliament in the 2017 Act to notify withdrawal from the EU on 29 March 2017, at a time when she was unaware – and could not have been aware – of the illegalities that were discovered later. In substance, these grounds amount to a contention that the Respondent ought to have taken action when that further evidence came to light and the further findings of the Electoral Commission were made, i.e. Ground 4 (which is unarguable for the reasons set out below).
48. In any event, second, Carnwath LJ's fourth requirement for an error of law to amount to a head of public law challenge is that the mistake must have played a material part in the decision-maker's reasoning. Here, for the reasons I have already given, there is no evidential basis for concluding it was material. That alone is fatal to Ground 2. In my view, the want of materiality also undermines Ground 3.
49. Third, each of these grounds is based upon the alleged unlawfulness of the EU referendum. However, as I have described, the Respondent's power to notify withdrawal from the EU derives, not from a lawful referendum, but from the 2017 Act. As Ouseley J said (at [70]), the discretion given to the Respondent by Parliament to give notification of withdrawal did not require her to await the outcome of any and all future investigations into actual or potential irregularities in the EU referendum. The referendum was merely advisory; and it was up to those that it advised to consider whether, in all the circumstances, the EU referendum result was and continues to be sufficiently robust as to reflect the will of the people on the question of withdrawal from the EU. Crucial to both of these grounds – and, in my view, to the appeal in respect of the merits as a whole – Parliament has exerted and maintained control over the withdrawal process, as Miller made clear it was and is fully entitled to do. Thus, section 13 of the 2018 Act (see paragraph 13 above) provides that any negotiated withdrawal deal or a withdrawal with no deal has to be approved by Parliament. In the circumstances, for the court to declare void the decision to notify withdrawal or the notification itself would clearly be a constitutionally inappropriate and unlawful interference in the due democratic process.

Ground 4: The Respondent's Continuing Failure to Respond to the Developing Evidence of Illegality in the EU Referendum

50. The 2017 Act received Royal Assent on 16 March 2017. The Respondent gave the EU notification of withdrawal on 29 March 2017. At that time, there were no outstanding investigations by the Electoral Commission; and the investigations which had been conducted and concluded had found no irregularities in the leave campaign donations or spending. However, as I have described, the Electoral Commission conducted further investigations, which reported in May, July and November 2018 to the effect that there had been illegal and corrupt practices in the form of breaches of the rules concerning donations and spending.
51. Ms Simor submits that the Respondent's decision to serve the notification to withdraw was based upon two factors, namely (i) the result of the referendum constituted an expression of the democratic will of the people and (ii) the Government had made a manifesto commitment to give effect to the result of the referendum. Section 1(1) of the 2017 Act merely gave the Respondent the power to notify the EU of withdrawal – it did not require withdrawal after the notification had taken place. It is open to the UK unilaterally to revoke the notification (Wightman v Secretary of State for Exiting the European Union (2018) (Case C-621/18)).
52. In the circumstances, she submits that the Respondent is therefore in a position to take steps in response to the illegalities established by the Electoral Commission after 29 March 2017; and she has acted – and continues to act –irrationally in failing to take any such steps and failing even to have any regard to the illegalities which have now been established.
53. However, for the reasons I have already given, the Respondent cannot arguably have acted irrationally in giving the notification at a time when there were no outstanding investigations; nor is there any force in the contention that the Respondent acted under a mistake of fact falling within the scope of E. Nor did she arguably act irrationally in not seeking to withdraw the notification – or further consider whether the notification should be withdrawn – in circumstances in which there is no evidence that any irregularities were material. And, I repeat, the fact that Parliament has maintained control over withdrawal makes it patently inappropriate for the court to intervene.
54. This ground fails: it too is unarguable.

Substantive Grounds: Conclusion

55. For those reasons, I consider each of the substantive grounds of appeal fails on its merits. On those merits, I agree with the judgment of Ouseley J; and with the succinct, but in my view perfectly adequate and commendable, observations of Supperstone J in refusing permission to proceed on the papers. I too do not consider that any of the substantive grounds has any real prospect of success.
56. The Applicants clearly oppose the UK leaving the EU; and hold strong views to that effect. Others hold strong views in favour of leaving the EU. The subject matter raises passions on both sides. However, consideration of this claim must be focused exclusively on the question of whether the Respondent has acted in accordance with the law. The courts are not concerned at all with the merits of leaving or remaining in the EU. As Gross LJ observed of the issue before him in Webster (at 24]):

"Put bluntly, the debate which the claimant seeks to promote belongs firmly in the political arena, not the courts."

Or, as Singh LJ and Carr J put it recently in R (Hoareau and Bancoult) v Secretary of State for Foreign and Commonwealth Affairs [2019] EWHC 221 (Admin) at [326]:

"Judicial review is not, and should not be regarded as, politics by another means."

As will be clear from this judgment, I consider those observations equally applicable to this case.

Grounds 5 and 6: Delay

57. Leaving aside the appeal against the costs order, my conclusions on the merits are sufficient to dispose of the substantive application by refusing it. However, before Ouseley J, the Respondent submitted that, irrespective of the merits, permission to proceed should not be granted because of delay, on two bases. First, it was submitted that, under paragraph 19 of Schedule 3 to the 2015 Act and/or CPR rule 54.5(1), the claim was out of time – and time should not be extended. Second, it was submitted that, under section 31(6)(a) of the Senior Courts Act 1981, the claim was not brought without undue delay and the grant of the relief sought would likely be detrimental to good administration. Ouseley J considered those submissions had been made good. As Grounds of Appeal 5 and 6, the Applicants contend that he was wrong to have done so.
58. I have already set out the relevant paragraph from the Schedule to the 2015 Act, which imposes a time limit on judicial review challenges to the EU referendum of six weeks from the date of certification of the number of votes, i.e. six weeks from 24 June 2016. That time limit is incapable of extension.
59. CPR rule 54.5(1) provides that, where there is no specific provision as there is in the 2015 Act, judicial review proceedings have to be issued promptly and, in any event, not later than three months after the grounds to make the claim first arose. That time limit is capable of being extended by the court, subject to an extension being warranted on good grounds.
60. Section 31(6)(a) of the Senior Courts Act 1981 provides, so far as relevant to this application:

"Where the High Court considers that there has been undue delay in making an application for judicial review, the court may refuse to grant… leave for the making of the application… if it considers that the granting of the relief sought would be likely to… be detrimental to good administration."

This does not provide for any specific time limit – indeed it is without prejudice to any such time limit (see section 31(7)) – but it provides a separate, freestanding basis upon which the court may refuse permission to proceed with a judicial review.

61. In response to the Respondent's submission that this claim falls foul of both of these requirements – and Ouseley J was not only entitled, but right to dismiss the claim on the basis of delay in any event – Ms Simor submits:

i) In respect of the grounds that relate to the EU referendum and the notification, although the claim is out of time, Ouseley J was wrong not to extend time because he did so primarily on the basis that to allow the challenge to proceed now would be detrimental to good administration and he should have considered that this would leave a flawed and illegal referendum result undisturbed which (it is submitted) cannot be conducive to good administration. The claims could not reasonably have been made any earlier, because the evidence and Electoral Commission's findings of illegality were not public until July 2018. The claim was commenced promptly, and certainly within three months, after that. There was no unreasonable or undue delay.

ii) Ouseley J erred in holding that the challenge to the ongoing refusal to consider the evidence of illegality of the referendum was out of time at all, because the Respondent first refused to consider the evidence set out in the Fair Vote Project letter of 5 July 2018 – and the claim was issued promptly and within three months of then – and is it in any event a continuing failure.

iii) For the same reasons, for the purposes of section 31(6) of the Senior Courts Act 1981, he was wrong to conclude that there was undue delay and that to allow the challenge to proceed now would be detrimental to good administration.

62. Given my findings in the merits, it is unnecessary for me to consider at length issues arising out of alleged delay; but it does seem to me that the Respondent's submissions are compelling. The referendum result was determined and announced on 24 June 2016. There is a good argument that paragraph 19 of Schedule 3 to the 2015 Act limited the manner and time in which a challenge to the result could be made in the High Court; but, even if that is not so, time for a judicial review claim challenging the EU referendum undoubtedly began to run then. Time to challenge the notification started to run from 29 March 2017. The claims challenging each are on any view hugely out of time.
63. Ms Simor submits that they could not reasonably have been made sooner, because it was not until mid-2018 that the Electoral Commission made findings of illegality; but claims of such impropriety were made much earlier, and Ouseley J properly took that into account when considering whether to extend time. In this case, the importance of the issues and the potential consequences of the challenge, far from making it easier to obtain an extension of time, made it all the more important that any claim be made promptly. Contrary to Ms Simor's submission, in my view, the delay and granting the relief sought would now very clearly prejudice good administration.
64. Although the absence of merit is in itself determinative of the application before us, in my view Ouseley J was not only entitled but right to refuse to extend time for the judicial review; and to conclude that the delay which has occurred was undue, and the relief sought – e.g. quashing and/or declaring unlawful the EU referendum and the notification of withdrawal – would obviously involve detriment to good administration of the most serious kind.

Ground 7: Costs Order

65. The general approach to costs orders in judicial review proceedings is now well-established.
66. By section 51(1) of the Senior Courts Act 1981, subject to any other enactment and the rules of court, the costs of and incidental to all proceedings in the High Court "shall be in the discretion of the court". As section 51(3) emphasises:

"The court shall have full power to determine by whom and to what extent the costs are paid".

67. Section 51 gives the court a discretion to award costs on an application for permission to judicially review (R v Camden London Borough Council ex parte Martin [1997] 1 WLR 359). In exercising that discretion, the general rules in CPR rules 1.1, 44.2 and 44.4 apply. The court's overriding concern is to do justice between the parties, taking into account all relevant circumstances and, on assessment, to ensure that costs ordered to be paid were reasonably incurred and are reasonable and proportionate in amount.
68. However, in exercising that broad discretion in relation to applications for permission to proceed, the court must take into account the nature of such applications. As Sedley J put it in Martin (at page 364G), the point of the requirement for permission in judicial review claims is "to afford applicants a simple and inexpensive way of finding out whether they [have] a worthwhile case". The whole purpose of requiring permission to be obtained would be defeated if the court were to go into the matter in depth at that stage, the proper place for full exploration of the evidence and argument ordinarily being at the substantive hearing of the claim which has been shown to be arguable at the permission stage. Thus, the part played by a defendant to such a claim at that stage is restricted. The relevant public body may of course file an acknowledgement of service with short summary grounds of resistance; but, to do so, it should generally not be necessary for it to do much additional work. As this court has said, its proper course is to explain its decision and any further grounds of opposition in short form; and not take an active part in any oral hearing, but simply wait and see if permission is granted and, if it is, then and only then deploy a full defence (R (Davey) v Aylesbury Vale District Council [2007] EWCA Civ 1166; [2008] 1 WLR 878 at [12]-[13] per Sedley LJ). Thus, paragraphs 8.5 and 8.6 of CPR PD 54A provide that:

"8.5 Neither the defendant nor any other interested party need attend a hearing on the question of permission unless the court directs otherwise.

8.6 Where the defendant or any other party does attend a hearing, the court will not generally make an order for costs against the claimant."

69. The circumstances in which a claimant may be ordered to pay a defendant's costs of attending a permission hearing were considered by this court in R (Mount Cook Land Limited) v Westminster City Council [2003] EWCA Civ 1346; [2004] 1 PLR 29.
70. Having considered the history of the relevant provisions in the CPR, Auld LJ said (at [72]):

"Generally - that is, save in exceptional circumstances - costs of and occasioned by such attendance should not be awarded against a claimant."

He continued (at [73]):

"It follows that judges before whom contested permission applications are listed, and in their conduct of them, should discourage long hearings and/or the filing by both parties of voluminous documentary evidence for consideration at them. In short, they should not allow the court to be sucked into lengthy and fully argued oral hearings that transform the process from an inquiry into arguability into that of a rehearsal for, or effectively, an expedited and full hearing of the substantive claim."

71. In [76], he indicated that "exceptional circumstances" in this context "may consist… of…" (i) the hopelessness of the claim, (ii) persistence in the claim by the claimant after being alerted to its hopeless nature, (iii) the extent to which the court considers the claim is abusive because it had been used for collateral ends and (iv) whether, by as a result of the deployment of full evidence and submissions, the claimant has effectively had the benefit of a full hearing.
72. These were, clearly, not intended to be an exhaustive list, but rather examples of what might comprise circumstances that may warrant an award of costs of a permission hearing against an unsuccessful claimant. But, in my view, Auld LJ's comments about the usual, brief documentation and consideration in a permission application is also informative. As Sedley LJ said in Davey (at [13]), it is "not ordinarily necessary for a public body on which a claim for judicial review is served to do much additional work before completing its acknowledgement of service" (emphasis added); and it will usually be neither proportionate nor reasonable for the public body to attend the permission hearing to contest it. However, it is implicit in what he said that there may be occasions when it will be appropriate for the defendant to attend, and appropriate for the court to make a costs order in his favour if permission is refused.
73. Each case will of course depend on its own facts. In this case, Ouseley J clearly had the factors set out in Mount Cook in mind. He referred to a number of the grounds being "hopeless", and of the Applicants continuing with the oral hearing although Supperstone J had indicated he considered the claim particularly weak. Ouseley J considered, with some force, that the Applicants were inappropriately pursuing what was effectively a political campaign through the courts. In any event, the permission hearing was not a short event. It lasted a full day, with judgment reserved over the intervening weekend. The Applicants' skeleton argument, drafted by two Leading Counsel and three Juniors, occupied 48 pages. Ouseley J heard oral submissions from Leading Counsel for the Respondent, whose skeleton argument (at eight pages) was modest in length. For the Respondent, the issues raised were of great – indeed, constitutional – importance. I do not consider that Ouseley J can possibly be criticised for allowing the Applicants every opportunity in making good their contention that their claims – or some of them – were arguable; or for allowing the Respondent a modest opportunity to respond orally. He was clearly assisted by those submissions on behalf of the Respondent. He was in the best position to assess whether this permission hearing was such as to warrant a costs order against the Applicants.
74. In my view, Ouseley J approached the issue of costs correctly, and the costs order he made was well within the margin of proper discretion and judgment left to judges in matters of costs. I do not consider that the contrary is arguable.

Conclusion

75. It was for those reasons that I refused permission to appeal, on all grounds.
76. Given that the application raises issues of wider importance, this judgment should have a neutral citation so that it can be reported (and, where appropriate, referred to in other cases) in accordance with Practice Direction (Citation of Authorities) [2001] 1 WLR 1001.

Lord Justice Haddon-Cave:

77. I agree.