In two recent cases, Jefferies and Finucane, the courts have considered key themes in the modern doctrine of legitimate expectation, including whether an expectation can be created in conditions of strict confidence, and the circumstances in which the government can resile from expectations created in a macro-political context.
The Leveson Inquiry into the behaviour of the British press was originally intended to take place in two parts. The first would inquire into the culture, practice and ethics of the press, and make recommendations for more effective regulation (Leveson 1). This occurred during 2011 and 2012. It generated a 2000 page Final Report, which in turn spawned a new regulator, IPSO.
Following this there was to be a hiatus, during which criminal prosecutions arising from the conduct of the News of the World could take place. Once these were completed, it was proposed that Leveson would reconvene to inquire into the corporate governance of the press, and the conduct of the police, politicians and other public servants in relation to it (Leveson 2).
On 1 March 2018, the current government announced that Leveson 2 would not proceed.
R (Jefferies) v Secretary of State for the Home Department was a judicial review of this decision brought by four people whom Leveson 1 found to have been treated badly by sections of the newspaper industry – a group which included Gerry and Kate McCann, parents of the missing Madeleine. They challenged the decision not to hold Leveson 2 on the ground that they had a legitimate expectation that the second stage of the inquiry would go ahead.
Shortly after Jefferies, the Supreme Court reached a decision in the long-running case of Geraldine Finucane, which also turned on the question of whether the claimant had a legitimate expectation of a public inquiry, in this case into the murder of her husband in one of the most infamous episodes of the Northern Ireland Troubles.
Both cases deal with important themes in the law on when legitimate expectations arise, and when (having arisen) they can lawfully be frustrated.
The Underlying Law
When considering any promise that may give rise to a legitimate expectation, it is important to disentangle two closely related strands – (i) statements of policy as distinct from statements that give rise to a legitimate expectation, and (ii) statements that give rise to a substantive legitimate expectation as distinct from those that create only a procedural legitimate expectation.
Policy v Legitimate Expectation
A policy is binding while it it is in place, unless there is good reason to depart from it in an individual case – see The Duty to Follow Policies (and its Limits). But the policy and political context in which governmental decisions are made evolves constantly. Policies can and do change, entirely lawfully, from time to time.
A legitimate expectation requires there to have been more than a mere statement by a public authority of its existing policy. A legitimate expectation will be created only when there has been a promise that the existing policy will not change.
This distinction was most clearly expressed by Laws LJ in Bhatt Murphy –
‘Authority shows that where a [legitimate] expectation is to run, the promise or practice which is its genesis is not merely a reflection of the ordinary fact (as I have put it) that a policy with no terminal date or terminating event will continue in effect until rational grounds for its cessation arise. Rather it must constitute a specific undertaking…by which the relevant policy’s continuance is assured.‘ 
That was in 2008, and adopts a position subsequently endorsed by the Supreme Court (in Mandalia), which also recognised that the duty to follow existing policy is a discrete legal requirement that exists independently of the doctrine of legitimate expectation.
However, the senior judiciary can still unhelpfully conflate these concepts (for a recent example see Coulson LJ in Save Britain’s Heritage), and over the last few years the courts have been required to decide no shortage of cases in which claimants have sought to rely on things which were only policies as if they were in fact legitimate expectations (I wrote about one of these cases here).
Substantive v Procedural Legitimate Expectation
A legitimate expectation can be based on a promise as to an outcome (substantive) or a promise as to a procedure to be followed before the outcome is determined (procedural). A promise not to close a hospital creates a substantive expectation; a promise to consult before doing so gives rise only to a procedural one.
Both types of expectation must be grounded in a promise that is ‘clear, unambiguous and devoid of relevant qualification’ (Lord Hoffman in Bancoult at ). This is the minimum requirement for a legitimate expectation of either type. But subject to this it is generally easier both to create and sustain a procedural expectation than a substantive one.
There are two particular ways in which this is true.
First, while a procedural legitimate expectation may be created by a promise made to the world at large, the courts prefer the view that a substantive expectation usually requires a promise to have been made only to a small group –
‘Though in theory there may be no limit to the number of beneficiaries of a promise for the purpose of [a substantive] expectation, in reality it is likely to be small, if the court is to make the expectation good…it is difficult to imagine a case in which government will be held legally bound by a representation or undertaking made generally or to a diverse class.‘ (Bhatt Murphy at )
Second, while both types of legitimate expectation create public law rights, those rights will not be enforced by the courts if there is an overriding public interest in the relevant public authority acting inconsistently with its promise; and this is much more likely to be the case in the context of a substantive rather than a procedural expectation.
The mechanism for this is that an overriding public interest is more likely to arise where the promise exists in the ‘macro-political field‘, in which case the court’s supervision of the public authority’s actions will be ‘less intrusive‘ (Begbie at ) and ‘the expectation’s enforcement in the courts will encounter a steeper climb‘ (Nadarajah at ).
In some cases, a promise may be one that ‘lies so deep in the macro-political field‘ that the courts will not recognise it as capable of creating a legitimate expectation at all (Wheeler at  – the promise in that case being to hold a referendum).
More usually, however, the macro-political context comes into play only in identifying whether an overriding public interest has arisen. And substantive expectations are more vulnerable to this override than procedural ones, because they are more likely to be macro-political in nature. Promises not to close hospitals entail a commitment of public resources that can easily be categorised as macro-political; mere promises to consult will fall far short of this threshold.
Jefferies – Legitimate Expectation and Confidentiality
The Leveson 1 report was published on 29 November 2012. On the same day, the then Prime Minister, David Cameron, made a statement to the House of Commons indicating that it was the policy of the government to go ahead with Leveson 2 in due course.
The Public Statement
The claimants, however, did not rely on this statement as the basis for their legitimate expectation claim.
In giving the judgment of the Divisional Court, Lord Justice Davis expressed the view (at ) that this was because the expectation claimed was substantive in nature, and David Cameron’s statement in Parliament was not (by definition) made to a small group.
If this is right, then counsel were mistaken (as the judge certainly was) as to the nature of the expectation being claimed. A legitimate expectation that an inquiry will be carried out is procedural rather than substantive in nature.
Shortly after Jefferies was decided, this was clarified definitively by the Supreme Court in Finucane. In that case, the appellant sought to rely on statements made in Parliament which appeared to commit the government to a public inquiry into state collusion in her husband’s murder. Lord Kerr held that –
‘Here the promise made did not partake of a substantive benefit to a limited class of individuals…it was a policy statement about procedure, made not just to Mrs Finucane but to the world at large.’ 
However, since there is no rule that a procedural legitimate expectation can be created only by a statement made to a small group, there was no reason why Parliamentary statements could not be relied upon in Finucane, as indeed they were. The Supreme Court found that they constituted a sufficiently ‘plain and explicit undertaking‘ () to create a legitimate expectation.
In principle, David Cameron’s House of Commons statement on Leveson 2 could have been relied on in the same way in Jefferies. However, whether or not for the wrong reasons, the claimants’ counsel were right not to place reliance on it. Properly analysed, what Cameron expressed in Parliament was merely the government’s policy at that time – ‘One of the things that the victims have been most concerned about is that Part 2 of the investigation should go ahead…that is fully our intention‘ . This fell short of the clear and unambiguous language of promises as to the future.
The Private Statement
While Cameron did not undertake to Parliament that the policy of holding Leveson 2 was guaranteed to continue, the claimants believed that he had made such a promise to them in private.
Just over a week before the Leveson 1 report was published, the Prime Minister held a private meeting with a select group of supporters of ‘Hacked Off’ an organisation which campaigns for greater accountability of the press. Around 30 people were present, including Evan Harris, the Executive Director of Hacked Off, Kate McCann (representing herself and her husband) and the other two claimants.
The evidence of the claimants, given in their witness statements for the purposes of the case, was that Cameron had been more forthcoming in private than in public, and had made a clear promise that Leveson 2 would occur (‘He gave me that specific assurance. He said to me in terms that once the criminal trials were concluded Part 2 would take place‘ ).
Unfortunately for the claimants, it came to light after the claim had commenced that one of the Hacked Off supporters had secretly recorded the meeting. In accordance with the duty of candour that applies to both sides in judicial review, a transcript of this recording then had to be disclosed. It revealed that Cameron’s statements were more guarded, and far more ambiguous, than their witness evidence had suggested.
There is no need to conclude that the claimants lied to the court. More likely Cameron, as an experienced politician used to working a room, had learned the trick of saying things in such a way that his listeners heard what they wanted to hear.
In any event, what he actually said as revealed by the transcript was very different from what had been claimed (‘I think what we are seeing is, you know, the wheels of justice as such do turn quite slowly and these court cases take a very long time, so I think we will have to return to this issue‘ ). Taken as a whole, what Cameron said in the meeting fell a long way short of the clear, unambiguous and unqualified assurance that would have been required to create a legitimate expectation of Leveson 2.
Legitimate Expectation and Confidentiality
That ought to have been sufficient to dispose of the claim in Jefferies. The court, however, went further. As the transcript showed, at the beginning of the meeting Evan Harris told the Prime Minister that ‘what is said in this room stays in this room‘ (). Davis LJ took a dim view of the notion that statements made by David Cameron under these conditions could be capable of constituting a legitimate expectation –
‘I find it unacceptable that this claim is, in substance, formulated by reference to a breach of that assurance…
That point of itself, therefore, causes me to hold against this claim. Any expectation engendered by what Mr Cameron said in this meeting, conducted on the basis that it was, cannot, in my judgment, be recognised or protected as a legitimate expectation for this reason alone.‘ , 
At one level, it is possible to sympathise with this view. Clearly the claimants were in the unattractive position of having to reveal that, while promising the Prime Minister that the meeting would take place under conditions of strict confidence, one of their group was surreptitiously recording it. Hacked Off, which campaigns against press conduct of a not dissimilar sort, could no longer occupy the moral high ground.
Nonetheless, the court’s reaction was intemperate and ill-considered. Properly analysed, the statement ‘what is said in this room stays in this room‘ was merely a rhetorical way of saying that the meeting was confidential, and in particular – hardly surprising in context – was not going to be leaked to the media.
Many legitimate expectations – especially substantive expectations created by statements made to small groups – arise from meetings, correspondence or email exchanges which occur under formal conditions of confidentiality. Was the position of the court in Jefferies that this should negate the expectation in each case? If so, it had no foundation in prior case law.
Davis LJ attempted two justifications for his conclusion. The first was rooted in fairness –
‘Indeed, given that the doctrine of legitimate expectation is rooted in fairness (in the public law sense) it seems to me that it would be contrary to fairness to allow what is said in such a meeting, conducted on an agreed private basis, to be permitted to ground the claim.‘ 
The problem with this is that the public law principle of fairness governs how the State is required to behave towards its citizens, not the other way round. There is no public law requirement on individuals to behave fairly towards government. In principle at least, the nature of their conduct might have a bearing on the question of whether the State has treated them unfairly – public law fairness is always highly contextualised – but this is not how the judge expressed the point.
In any event, what exactly was unfair about the claimants’ behaviour? Agreeing that a meeting takes place on a confidential basis is not the same as waiving all legal rights in relation to what happens in it.
The second justification was based on reliance –
‘Mr Cameron could not possibly, in consequence, have thought (or, objectively, could be expected to have thought) that what he was saying could or would thereafter be relied upon directly by those present as creating a legitimate expectation.‘ 
This assumes that legitimate expectation functions on a kind of quasi-contractual basis, in which the State is only bound by a promise if it had reason to assume that the persons to whom it was directed were likely to place reliance on it. With respect to the judge, this is both conceptually wrong and unsupported by authority.
It follows that neither of these justifications withstands close scrutiny. The court reached the wrong conclusion both on the law and the evidence – an agreement to confidentiality is (whether express or implied) a feature of many meetings and much correspondence. It is not the same as a waiver of rights, analagous to a ‘without prejudice’ meeting between lawyers. If it was, then it would be likely to have a serious and narrowing effect on the law of legitimate expectation, for which there is no warrant in prior authorities.
In consequence it must be suggested that Jefferies should not be treated as good authority on this issue. At the very least, it should be confined to its highly specific facts.
Finucane – Legitimate Expectation and Politics
Even by the standards of the Northern Ireland Troubles, the 1989 murder of the Belfast solicitor Pat Finucane, at home in front of his wife and children, marked a notorious low point.
In the three decades since then, his widow Geraldine has, in the words of the Supreme Court, ‘waged a relentless campaign…to have a proper investigation conducted into the circumstances in which he was murdered‘, seeking in particular a public inquiry into the undoubted collusion between his killers and agents of the British State .
By means of various assurances given by the government during the period 2004 to 2008, a promise to hold a public inquiry was made – ‘It is quite clear that, individually and cumulatively, they amount to an unequivocal undertaking to hold a public inquiry into Mr Finucane’s death‘ . As already explained above, these created a procedural legitimate expectation.
The Macro-Political Context and Procedural Expectations
The assurances that created this expectation were all given during the period of the Blair and Brown administrations. At the same time, Lord Saville was in the middle of what, in the end, was to turn into a twelve-year inquiry into the Bloody Sunday killings of 1972 – an inquiry which took so long, and cost such an elaborate sum of money, that it poisoned the political well in relation to any future exercises of the same type.
By the time of the 2010 general election, the Conservative Party manifesto contained a statement to the effect that there should not be any more long-running, open-ended and costly inquiries into the past in Northern Ireland. When the Conservatives formed the senior partner in the Coalition government following the election, they immediately set about the process of resiling from all promises given by their predecessors in relation to the circumstances of Pat Finucane’s death.
What they established as an alternative – a review undertaken by Sir Desmond da Silva – was found by the Supreme Court in Finucane to fall short of the standard required for an inquiry that was compliant with the UK’s obligations under Article 2 of the European Convention on Human Rights – ‘the conclusion that an article 2 compliant inquiry into Mr Finucane’s death has not yet taken place is inescapable‘ .
But was this failure to conduct a full public inquiry also a breach of Geraldine Finucane’s legitimate expectation? The answer is yes, but according to the Supreme Court not in a manner that involved any unlawful action by the government.
For context, it is important to note three things.
First, legitimate expectations both can and do survive changes in government. A promise made by the ministers or officials in one administration is quite capable of binding their successors.
Second, however, if the promise exists in the macro-political space, it is subject to being overridden by public interest considerations, and a post-election change of government may supply all the features needed to justify resiling from the expectation, especially if its subject-matter has been a campaign issue during the election. In Jefferies, where the 2017 Tory Manifesto had stated that Leveson 2 should not proceed, the court held (obiter at ) that this would have defeated any legitimate expectation of an inquiry had one existed.
Third, while it is generally easier to cast substantive expectations as occupying macro-political territory, and therefore susceptible to a supervening public interest, there is no rule against procedural expectations receiving the same treatment. If the expectation is merely about consultation, it is very likely to be upheld, even in a post-election context (see, for instance, Luton Borough Council v Secretary of State for Education at  – ). But proposed public inquiries – particularly in a highly politicised environment such as Northern Ireland – are more obviously amenable to being overridden.
And this was in fact the basis for the outcome in Finucane, where Lord Kerr summed up the difficulty of enforcing a legitimate expectation (whether procedural or substantive) created in a macro-political context –
‘Where political issues overtake a promise or undertaking given by government, and where contemporary considerations impel a different course, provided a bona fide decision is taken on genuine policy grounds not to adhere to the original undertaking, it will be difficult for a person who holds a legitimate expectation to enforce compliance with it.‘ 
This is a clear but also sweeping statement, which strikingly posits a low bar for resiling from promises in a macro-political context – the mere fact that politics has overtaken the original promise.
Aside from the need to take into account the existence of the promise that is being frustrated, is this actually all that different from a statement of the basis on which the government may change any policy from time to time? It certainly appears to accord very little weight to individual rights in the face of much larger political imperatives.
Accordingly, this is sure to become a much-quoted touchstone of the law, frequently cited by the government in the defence of future judicial reviews.
A Brief Footnote – Detrimental Reliance
In the earlier development of the law on substantive legitimate expectation – proceeding to some extent by analogy with the law of promissory estoppel – it was considered that detrimental reliance on a promise by its beneficiary was essential if the expectation was to be binding and enforceable.
This has long since ceased to be the prevailing view. The better position, as understood for the last decade or so, is that detrimental reliance is not necessary, though it may still be a relevant consideration in deciding whether or not it is fair for a public authority to resile from an expectation in the public interest (see Bancoult, cited above, at ).
Even so, there remained sufficient room for argument that the government could plead in Finucane that detrimental reliance was required, relying among other things on words of Lord Carnwath in the recent Privy Council case of United Policyholders (at ).
Although it was obiter in the context of Finucane, Lord Carnwath took the opportunity to do some resiling of his own, distancing himself from his previous opinion – ‘On reflection…I accept that…the proposition may have been too narrowly stated‘ .
Taken together with the view of the other members of the court as expressed by Lord Kerr – ‘a matter sounding on the question of fairness is whether the alteration in policy frustrates any reliance which the person or group has placed on it. This is quite different, in my opinion, from saying that it is a prerequisite of a substantive legitimate expectation claim that the person relying on it must show that he or she has suffered a detriment‘  – this should, obiter or not, put the point to bed.
Legitimate expectation is a key principle of modern public law, but it is much easier for a claimant to assert than to rely on.
Both Jefferies and Finucane demonstrate how many steps there are in judicial review between a promise and an enforceable legal right, and how easy it is to trip over any one of them. This is so even though both cases concerned procedural expectations; it is true a fortiori of their substantive equivalents.
However, while Finucane adopts an apparently orthodox (albeit, for the claimant, hard to swallow) view of the law, Jefferies occupies much more contentious territory, and in its treatment of expectations created under circumstances of confidentiality may better be seen as a guide to judicial attitudes than as a statement of what the law actually is.