In the summer of 2022, during the election process which Liz Truss would eventually win, Tortoise Media wrote to the Conservative Party’s CEO, Darren Mott, asking nine questions. Considering Tortoise Media had just successfully registered a tortoise called Archie, two overseas nationals, and one fictional person called ‘Margaret Roberts’ to vote for Britain’s next prime minister, they were understandably concerned about the integrity of the Party’s internal election processes. The Conservative Party declined to answer Tortoise Media’s questions replying, ‘the Party is not a public body and does not carry out public functions’ (Tortoise Media v Conservative Party [10]). Lang J, and then Fordham J, refused applications for judicial review,  and now the Court of Appeal has as well.

This post will argue, as I have elsewhere, against this approach. Tortoise Media do have an Article 10 ECHR right to the information they sought. The Court of Appeal is of course right to say that it is important ‘in a free and pluralistic society’ to ‘permit parties to adopt their own rules, for example as to how they elect their leader, without undue interference by the state.’ [50] Nothing in any of Tortoise Media’s submissions denied this. Instead, the submissions of Tortoise Media are better understood as the logical correlate of this claim. In the UK parties are free to organise their affairs, within the limits of equality, anti-terrorism and other applicable laws, and then voters can vote for them or not. This political approach to parties cannot be sustained however, without some transparency about how internal party processes work. That transparency is all that Tortoise Media was asking for. Their requests are consistent with both common and Convention law, and the public had a right to the answers to Tortoise Media’s questions.

This post has four parts. First, it explains the law and facts of Tortoise Media. Second, it traces how the reasoning in the Court of Appeal moves on from the oral hearing for permission. The third part challenges the reasoning of Singh LJ in the Court of Appeal, and acknowledges some of the very useful points put forward by Dane Luo on this platform. The fourth and final part of the post emphasises the importance of the Article 10 arguments in this case, particularly given the ECtHR decision in Magyar v Hungary, and the Government’s concession that the law of information rights may need to be revisited given the Grand Chamber’s decision.

The Law and Facts of Tortoise Media

The question at issue in Tortoise Media was whether, when the Conservative Party elected its leader in 2022, the Party was exercising a public function. Political parties are not public authorities within the meaning of the Freedom of Information Act [22] and so Tortoise Media sought instead to invoke the positive obligation arising under Article 10 of the ECHR, via section 6 of the Human Rights Act. To satisfy s.6 of the Human Rights Act, however, Tortoise Media had to demonstrate that the Conservative Party was exercising a public function. The information that Tortoise Media requested from the Conservative Party included: (i) the number of attempts at party infiltration which had been caught, (ii) the compliance mechanisms for verifying that voters were who they claimed to be, (iii) anonymised member demographics, particularly of overseas and underage voters, and (iv) an explanation for GCHQ’s offer to provide advice on ballot distribution security. [9] The Conservative Party declined to answer these questions saying it was ‘not exercising a public function.’ [10] The Party also said:

The election of the Leader of the Conservative Party is a private matter for the members of the party under its Constitution. [10]

The Court of Appeal agreed with the Conservative Party, finding it would be wrong to ‘impose constraints on the autonomy of political parties which Parliament has not thought fit to impose.’ [50]

Before turning to Singh LJ’s reasoning, one point about his conclusion is material. First, Tortoise Media were not asking for any limits on the autonomy of political parties. They were asking only for the Conservative Party to explain its own processes so that voters could make up their own minds. Such an approach ensures autonomy not only for parties, but for the voters who vote for them. The approach Tortoise Media were advocating, therefore, is necessary for political accountability. Finding a function is public does not mean finding it is unlawful. 

Second, Singh LJ took the absence of legislation in this area to be evidence that Parliament did not intend to intervene, saying courts should not intervene when ‘Parliament has not thought fit to impose’ limits [50.] Diggory Bailey, from the Office of Parliamentary Counsel, has persuasively demonstrated the risks of this approach in ‘Interpreting Parliamentary Inaction.’ Bailey rightly argues that the absence of action is not evidence of parliamentary intention. In addition to this approach being untenable as a matter of logic, Bailey convincingly shows that such an approach has been rejected as a matter of law, for instance by the UKSC in R v. Newham and Lewisham.

Stages, Functions, Consequences, and Links in the Chain

In the prior oral hearing for permission, Fordham J’s key move was to say that the election of a leader was not a public function because functions should not be conflated with consequences. Fordham J concluded that: ‘on the evening of 5 September 2022…the Party’s function was completed. Its function had an outcome. But there was not a new Prime Minister.’ ‘A function,’ he held, ‘is not to be equated with a consequence.’ [35] The Court of Appeal made the same point, although using different language, through invoking the idea of stages. The three stages outlined in the Court of Appeal were:

1. The election of the Party leader.
2. Advice given by the incumbent Prime Minister to the Sovereign as to which person was likely to have the confidence of the House of Commons…
3. The appointment by the Sovereign of the Prime Minister. [31]

As in the oral hearing for permission, Tortoise Media sought to demonstrate the inextricable connections between Stages 1, 2, and 3. In the Court of Appeal, Tortoise Media sought to demonstrate this through providing evidence from Professor Robert Hazell that the last time the Sovereign attempted to appoint a Prime Minister who did not have the support of a majority of members of the House of Commons was in 1834. [14]

Singh LJ was not persuaded of the inextricability of stages 1-3, finding that stage 1 was a private matter. More than that, though, the reasoning in the Court of Appeal is different than that of the oral hearing for permission because Singh LJ says that even if it were true that the leader of the largest party automatically became Prime Minister, that would not be enough for the selection of the leader of the largest party to be a public function. [41] Fordham J in the oral hearing denied that public law consequences would follow from the election of a Prime Minister. In the Court of Appeal, by contrast, even if there were public law consequences from choosing a leader, Singh LJ said those consequences would not make the election process public. Singh LJ said that the election of a leader is private because even “public law consequences” [44] are not necessarily enough to show that an underlying function is public. This is because of the ‘nature’ [41] of the internal Party process, particularly that the function being exercised by the Party was not ‘judicial.’ [48] More on the meaning of ‘judicial’ in a moment.

While Luo agrees with the outcome of Tortoise Media, he is worried about the reasoning in the case. Luo argues that Singh LJ should not have accepted Tortoise Media’s premise that it is ‘well-established by constitutional convention that the Sovereign will appoint the Prime Minister on the advice of the incumbent Prime Minister.’ [36] Luo offers many compelling reasons to think that the appointment of the Prime Minister is a ‘reserve’ power which the sovereign can make, he says, ‘contrary to the recommendation of the outgoing Prime Minister.’ Luo’s points matter because in the oral hearing for permission the case was thought to hinge on how tightly the chain is constructed from the selection of the leader of the largest party to the appointment of Prime Minister. In the Court of Appeal, however, Singh LJ argues that even if it could be demonstrated that the leader of the Conservative Party would automatically become Prime Minister, the underlying function of the party’s selection of the leader would still be private.

Moving the Goalposts

The shift in the Court of Appeal away from Fordham J’s functions and consequences distinction to an analysis of stages is reflected in the Court of Appeal’s acknowledgment that the ‘high point’ [43] of Tortoise Media’s case is the dictum by Lloyd LJ in Datafin. Datafin says:

If the body in question is exercising public law functions, or if the exercise of its functions have [sic] public law consequences, then that may… be sufficient to bring the body within the reach of judicial review. [847C]

This passage matters enormously because, as Singh LJ acknowledges, it is the strongest point Tortoise Media made. This passage suggests that public law consequences could be enough to satisfy the public function test. However, Singh LJ stresses the word ‘may’ in the passage above, arguing that the fact that an action can have public law consequences does not necessarily mean the underlying function is public. Therefore, even if the chain outlined above in three stages is tightly linked, and the sovereign did necessarily appoint the leader of the largest party as the Prime Minister, the Court of Appeal concludes that the underlying function of selecting that leader would still not be public.

The Court of Appeal take great effort to distinguish Tortoise Media from Datafin. This concerned the question of whether the Panel on Take-overs and Mergers in the City of London was exercising a public function. The Court of Appeal held the panel on Take-overs and Mergers was exercising a public function, even though the source of the Panel’s power was not case law or statute. In Tortoise Media, the Court of Appeal thought that what was central to Datafin was that the Panel on Take-overs and Mergers was performing a ‘judicial’ function, and that it was this feature of the Panel on Take-overs and Mergers that ultimately made Datafin distinguishable from Tortoise Media. [48]

The Court of Appeal’s reading of Datafin is too narrow. Parts of the Court of Appeal’s own analysis in Tortoise Media demonstrates why. Key to the conclusion in Datafin was that, as Sir John Donaldson MR said, (i) there was no compelling alternative remedy available, as was also true in Tortoise Media, and (ii) the decisions of the Panel on Take-overs and Mergers had an indirect effect on others. Indeed, in Datafin, Sir John Donaldson MR said the panel was exercising a public function because ‘the rights of citizens are indirectly affected by its decisions.’ (838F-H) Such indirect effects are clearly present for voters in Tortoise Media as well. Singh LJ moves the goalposts on Tortoise Media to say that: public law consequences are not enough; it is the nature of the underlying act that matters and that the public function test is trying to assess. Even when the goalposts are moved, however, Tortoise Media can win on this ground. This is because like the Panel on Take-overs and Mergers, the Conservative Party was ‘lacking any authority de jure…[but] exercise immense power de facto.’ [826C]

Datafin is about the context-contingency of the public function test. In the context of Datafin, the term ‘judicial’ speaks to the impact of a body in the context of a regulatory ecosystem, particularly the capacity of the Panel of Take-overs and Mergers to issue sanctions. Political parties can issue sanctions, for instance, by deciding who can run to be party leader or is eligible to vote in such an election. This holistic approach is the best way of understanding what Sir John Donaldson MR describes as acting ‘judicially.’ Singh LJ said that the Conservative Party was not acting ‘judicially’ [48] but judicially should not be read as ‘court-like’ – the construction of the public function test in Datafin, which establishes the need only for a ‘public element’ [838E]is too expansive for that. The better reading of Datafin is that ‘judicial’ means a body must be ‘impartial for purposes of a regulatory ecosystem.’ Such a requirement of independence is true of the Conservative Party in the selection of the Leader and Prime Minister, and so the underlying function should be understood be public. Further, it is the independence of internal process that Tortoise Media were trying to establish with their questions, particularly given, as was said in Datafin, ‘the rights of citizens are indirectly affected by its decisions.’ (838F-H) The same was true in the circumstances of Tortoise Media as well.

Further, if, as Singh LJ says, the question in the case should be about the ‘nature’ of the act of selecting a party leader, rather than its political and legal consequences, then there are other features of the public function test which are material but were not considered in Tortoise Media. The fact that political parties receive public subsidies has made a difference in other domestic public function cases such as Weaver, in which the fact that a housing trust was in receipt of public money did help establish that a public function was being exercised. Further, the importance of public subsidies for the public function of political parties was dispositive in the Ontario Superior Court case of Graff v NDP which rightly held that: ‘The fact that political parties receive considerable amounts of public funds ties them more closely to government than would be the case for other voluntary associations.’ [14] Graff was followed in Canada by the latter Divisional Court case of Trost v Conservative Party which came to the opposite conclusion to Graff for similar reasons to the Court of Appeal in Tortoise Media. What the Graff-Trost debate in Canada demonstrates, however, is that the question of the integration of the Conservative party into the governmental framework is not just a matter of the inevitability of the appointment of the Prime Minister; there are a range of considerations which are material given the context-contingency of the public function test.

In Datafin, Sir John Donaldson MR said: ‘I should be very disappointed if the courts could not recognise the realities of executive power and allowed their vision to be clouded by the subtlety and sometimes complexity of the way in which it can be exerted.’ [838H-839A] Tortoise Media is a clear example of the ways in which the realities of executive power can be clouded by the complexity of the process. Political parties are a way of being ‘in power’ and assuming ‘executive power’, therefore their internal processes are the public’s business and are conducted – in part – on the public’s dime. Therefore, their underlying functions should be understood to be public. Tortoise Media is not a marginal case. It is a paradigmatic case where: ‘the public, in the shape of the state, have assumed responsibility, at public expense…and in the public interest.’ (YL at [65]).

The Next Step: An Injustice in Information Rights

As Lewis Graham rightly points out in the Administrative Court Blog, the UK Government recently accepted that the law here may need to change. In a friendly settlement decision, it was accepted that the analysis in Kennedy v Charity Commission was obiter and ‘may be revisited’ in light of the Grand Chamber decision in Magyar v Hungary,  which lays out principles for access to information [156]. In the leading judgment in Kennedy, Lord Mance concluded that Article 10 does not impose a freestanding positive duty of disclosure on public bodies. (Although Article 10 was not engaged in the case, which is why the findings were obiter [Kennedy, 57]).

Magyar was a case brought by a respected Hungarian civil society organisation, the Helsinki Committee, about an access to information request in respect of the appointment of public defenders. [1] The Hungarian Supreme Court ruled that a defence counsel was not performing public duties because they did not inherently have any public law powers. While the ECtHR acknowledged that the text of Article 10, focused on freedom of expression, does not expressly include a right to seek information, the Court nevertheless held that sometimes a right to information was required for the ‘practical and effective’ [155] implementation of Article 10. The question of whether information must be released was established by a two-part test, depending on (i) the nature of the underlying information and (ii) the purpose for which it was sought. [164] On the second purpose question, the ECtHR said that:

The Court considers that an important consideration is whether the person seeking access to the information in question does so with a view to informing the public in the capacity of a public “watchdog.” [168]

Tortoise Media’s submissions are powerful because they were acting like a watchdog in this case. They were trying to alert the public to an information injustice, not for ‘voyeuristic’ or ‘sensationalist’ reasons, [Magyar 162] or to undermine political contestation. Magyar demonstrates that Convention law has the resources to respond to the information injustice in Tortoise Media without undermining the autonomy of political parties or the norms of democratic contestation.

Conclusion

In the oral hearing for permission, Fordham J said of the case of Tortoise Media that ‘if there is an injustice [here] it is an injustice of the limits of the law on information rights.’ The injustice, Fordham J said, ‘is not actually directed to the specific function of the political party selecting the new leader.’ [43] Fordham J was right on the first point, and wrong on the second. He was right to say the injustice here is of information rights, and wrong that that injustice is severable from the public function question. There is an information injustice here because this election is choosing the Prime Minister. Arguments made for party autonomy are untenable when, as in Tortoise Media, those arguments undermine the autonomy of voters.

Many thanks to Weronika Galka, in her personal capacity, for a helpful discussion of some of these issues and to Hayley J. Hooper for outstanding section s.6(3)(b) feedback. Thanks as well to the editors of the UKCLA blog for exemplary comments. The usual disclaimers apply.

Leah Trueblood, Worcester College, Oxford

(Suggested citation: L. Trueblood, ‘An Injustice in the Law of Information Rights: Tortoise Media Ltd v Conservative Party and Unionist Party’, U.K. Const. L. Blog (11th June 2025) (available at https://ukconstitutionallaw.org/))

*Editors’ note – the original post was amended on 13 June 2025 to correct a typographical error in the quotation from para [31] of the Court of Appeal’s judgment.