I. Introduction
Since recognising legitimate expectations as a ground of judicial review, courts in the UK have not settled firmly on a theoretical underpinning for the doctrine. This gap in the law is problematic, with many commentators stressing “that the ‘doctrine’ … is in dire need of a ‘more principled footing’”.Footnote 1 In fact, Mark Elliott claims that the courts’ “failure adequately to engage with the normative basis of legitimate expectations” is “arguably one of the reasons for ongoing uncertainty in this area at the doctrinal level”.Footnote 2 In Coughlan, for example, one of few cases in which the courts have enforced a legitimate expectation against a public authority, the Court of Appeal held that a health authority’s decision to move a patient (Coughlan) from one facility (Mardon House) to another was unlawful. The court said the decision was “an unjustified breach of a clear promise” given to Coughlan by the authority’s predecessor (that she could remain in Mardon House “for as long [she] chose”), thereby frustrating her “legitimate expectation of having a home for life in Mardon House”.Footnote 3 As subsequent jurisprudence and commentary have demonstrated, however, the basis of the court’s conclusion – and how it in turn applies to other cases – remains unclear.
To date, the courts have mostly drawn on the broad principles of fairness, abuse of power and good administration to define legitimate expectations. In Coughlan, Lord Woolf noted that the health authority’s decision constituted “unfairness amounting to an abuse of power”.Footnote 4 And in the later Nadarajah case, Laws L.J. said the doctrine was rooted in “good administration”, requiring public authorities “to deal straightforwardly and consistently with the public”.Footnote 5 Problematically, however, as Elliott explains, these “superficially attractive but ultimately rather empty notions” are “incapable of doing the sort of analytical heavy-lifting that is required if the law in this area is to be placed on an intellectually cogent footing that lends itself to coherent doctrinal development”.Footnote 6
Some scholars have suggested that instead of these broad principles, what underpins the legitimate expectations doctrine is trust – specifically, public trust in government. This “trust conception” posits that the doctrine is concerned with protecting the trust which a citizen has reposed in a public official pursuant to a promise or representation the official has made.Footnote 7 So the court’s conclusion in Coughlan, under this conception, is based on Coughlan’s trust in the health authority, specifically pursuant to its promise that she could remain in Mardon House “for as long as [she] chose”. And trust, the conception’s proponents have claimed, does not suffer from the limitations that fairness, abuse of power and good administration do. This trust conception has even gained some traction in the jurisprudence. In the Court of Appeal’s 2016 Enfield case, Gross L.J., writing for the court, recognised that “[a]n important reason for the protection of legitimate expectations lies in the trust between citizens and officials, upon which good government depends”.Footnote 8 And in Talpada two years later, Hallett L.J., in a judgment with which the other members of the court agreed, cited approvingly a statement of the Upper Tribunal’s then-President that the “two basic ingredients” of substantive legitimate expectations are satisfied “where there is an unambiguous promise or assurance by a public official in which the affected citizen reposes trust”.Footnote 9
Though the trust conception of legitimate expectations has several supporters, it also has its critics. Scholars have argued recently that the conception is, among other things, “artificial”,Footnote 10 disconnected from “how the doctrine applies in reality”Footnote 11 and incomplete.Footnote 12 For these reasons, they claim that trust does not offer the theoretical underpinning needed for legitimate expectations.
While I do not disagree that the trust conception of legitimate expectations is problematic, I argue, in contrast to the conception’s critics, that its fundamental weakness is its lack of conceptual clarity. Proponents of the conception have not done enough to conceptualise its central concept – trust. This has yielded conceptual ambiguity regarding: (1) the form of the citizen’s trust; (2) the object of that trust; and (3) the party of focus in the citizen’s relationship with the public official. This ambiguity has generated problems for the trust conception and opened it up to criticism, ultimately hindering trust from being usefully deployed in this area of administrative law.
In this article, I draw on scholarship on trust from disciplines outside law and integrate it with the legitimate expectations literature, to resolve the above ambiguities and reimagine the trust conception of legitimate expectations. Trust, I suggest, can offer the necessary theoretical underpinning for the legitimate expectations doctrine – and thus, a means to provide certainty to this area of law. But we must first clarify the trust conception, including what we mean by “trust”.
My argument proceeds as follows. In Section II, I review the existing literature on the trust conception of legitimate expectations, outlining the core of the arguments advanced by its proponents and laying the foundation for my argument on conceptual ambiguity. In Section III, I examine the principal arguments made to critique the trust conception. In Section IV, I use the extralegal scholarship on trust to explain the conceptual ambiguity in the legal literature, along with its implications, and present my reimagined trust conception. And lastly, Section V concludes.
II. The Trust Conception and Its Proponents
Christopher Forsyth first advanced a trust conception of legitimate expectations in the UK, more than 35 years ago. In a 1988 article advocating the substantive protection of legitimate expectations (before Coughlan, and UK courts’ recognition of such protection), Forsyth intimated that the theoretical foundation for the doctrine was trust. Forsyth said specifically that the “judicial motivation for seeking to protect [legitimate] expectations was plain: if the executive undertakes, expressly or by past practice, to behave in a particular way the subject expects that undertaking to be complied with”.Footnote 13 In justifying this “judicial motivation”, he stressed that “[p]ublic trust in the government should not be left unprotected”, as trust is a requirement for “good government”.Footnote 14
In his later work, Forsyth has expanded on this fledgling idea, more explicitly promoting a trust conception of legitimate expectations. Forsyth has said unambiguously that “the justification for the protection of legitimate expectations is the simple idea that the law should protect the trust that has been reposed in the promise made by an official”.Footnote 15 He has based this justification on his previously-drawn link between trust and “good government”, explaining that “[g]ood government depends upon trust between the governed and the governor” because “[u]nless that trust is sustained and protected officials will not be believed and individuals will not order their affairs on that assumption”.Footnote 16 This argument is repeated in Forsyth’s Administrative Law text with William Wade.Footnote 17 After considering the broad principles used by the courts to define legitimate expectations (including fairness, abuse of power and good administration), Forsyth and Wade conclude that “a more satisfactory reason for the protection of legitimate expectations lies in the trust that has been reposed by the citizen in what he has been told or led to believe by the official”.Footnote 18
Another key proponent of the trust conception is Paul Reynolds. Like Forsyth, Reynolds has unambiguously argued that “the doctrine [of legitimate expectations] is concerned to protect trust which has been reposed in public officials pursuant to a representation which they have made”.Footnote 19 More precisely, Reynolds contends that the doctrine protects a “specific and narrow trust” – what he labels “specific” trust – to be distinguished from “general” trust.Footnote 20 Reynolds proposes that “[w]hilst the doctrine [of legitimate expectations] is justified because it reinforces general trust, and whilst its role in reinforcing general trust colours our understanding of it, the doctrine does not exist to protect general trust in individual cases”.Footnote 21 For Reynolds, the “relevant trust” for legitimate expectations “is the trust which an individual reposes in a decision-maker, pursuant to some kind of representation, to do what the decision-maker has indicated that it will do”.Footnote 22
Whereas Forsyth and Reynolds have been the most ardent proponents of this conception, the link between legitimate expectations and trust has been recognised by others. Jack Watson, in an argument “related and complementary” to that of Forsyth, has argued that the promise is a “social convention of trust” that invites “an individual to place their confidence in the promise maker”.Footnote 23 Where a public authority makes a promise, according to Watson, it intentionally alters the recipient’s perception of the world, creating “a belief in the occurrence of the event that is the subject of the promise” and “a reliance on that altered perception”.Footnote 24 Watson has therefore suggested that the courts, in enforcing legitimate expectations, ensure the recipient’s “domain of entitlement is not destroyed by the public authority”.Footnote 25 Robert Thomas has similarly employed trust to define legitimate expectations.Footnote 26 He has contended that the protection of legitimate expectations “enforces the trust an individual has placed in the administration and prevents the breakdown of a workable relationship between the two”.Footnote 27 Iain Steele has claimed that “trust in government is impaired” if a representation made by a public official that promises a benefit “is subsequently resiled from”.Footnote 28 And Soren Schønberg has maintained that the protection of legitimate expectations “promotes trust and confidence in authorities” and “their perceived legitimacy and efficacy”, thereby encouraging citizens to “participate, co-operate, and comply”.Footnote 29
As already noted, support for this trust conception is also found in the jurisprudence. In Mehmood, the then-President of the Upper Tribunal (Immigration and Asylum Chamber), McCloskey J., endorsed Forsyth’s statement that “[g]ood government depends upon trust between the governed and the governor”, adding that “unless that trust is sustained and protected officials will not be believed and the Government becomes a choice between chaos and coercion”.Footnote 30 Following on from this, McCloskey J. concluded that “[t]he two basic ingredients of what the law has come to recognise as a substantive legitimate expectation are satisfied where there is an unambiguous promise or assurance by a public official in which the affected citizen reposes trust”.Footnote 31 This statement was repeated by the Tribunal in the subsequent Iqbal caseFootnote 32 and cited approvingly by the Court of Appeal in Talpada.Footnote 33 And in Enfield, Gross L.J., writing for the Court of Appeal, cited Wade and Forsyth, identifying as “[a]n important reason for the protection of legitimate expectations … the trust between citizens and officials, upon which good government depends”.Footnote 34
The trust conception’s proponents contend that trust offers a better underpinning for the legitimate expectations doctrine than fairness, abuse of power or good administration. While these principles are not incorrect, they suffer from two problems. First, they are overly abstract, offering courts and administrative decision-makers little guidance on what the doctrine requires.Footnote 35 On abuse of power, for example, Laws L.J. noted in Nadarajah that while it “catches the moral impetus of the rule of law … it goes no distance to tell you, case by case, what is lawful and what is not”.Footnote 36 It is what Matthew Groves terms a “motherhood statement”, providing no “clear explanation of what might constitute an abuse of power or whether a new ground of review can be said to fall within the scope of that term”.Footnote 37 And similar remarks, Wade and Forsyth say, can “be made about the other reasons for protecting legitimate expectations”, like fairness and good administration.Footnote 38
Trust, the conception’s proponents claim, does not share this problem. Forsyth has emphasised (independently and alongside Wade) that “it is a simple concrete question of fact whether trust has been reposed in an official’s promise”.Footnote 39 He has therefore argued that in contrast to the “unhelpful” principles of fairness, abuse of power and good administration, trust does go “‘some distance’ to indicate which expectations should be protected and which should not”.Footnote 40
Second, fairness, abuse of power and good administration are “over-inclusive”, failing to distinguish legitimate expectations from other grounds of review.Footnote 41 For instance, Farrah Ahmed and Adam Perry contend that fairness, while relevant in legitimate expectations cases, “is relevant in other kinds of cases, too”.Footnote 42 Whereas it “favours keeping a promise and, at least sometimes, following through on a policy or practice”, it “also favours other things … which play no part in the doctrine”.Footnote 43 Paul Daly has likewise suggested that abuse of power “is probably better understood as an overarching principle of judicial review of administrative action in general, a ‘rational’ rather than a free-standing rule”.Footnote 44 And good administration is equally problematic in this respect.Footnote 45 As Reynolds says, “[i]t is hard to disagree with the assertion that good administration is a relevant principle in the doctrine, but it is not the specific and distinct principle that we need in this area”.Footnote 46 Good administration is “too vague and overarching to provide any concrete delimitation or guidance”, sitting “on the same plane of abstraction” as fairness and abuse of power.Footnote 47
Trust, the conception’s proponents claim, also does not share this issue. More precisely, if we understand trust in the “specific and narrow” way that Reynolds has outlined, trust, unlike its counterparts of fairness, abuse of power and good administration, is unique to the legitimate expectations doctrine. In fact, Reynolds explains in distinguishing “specific” from “general” trust that a restrictive definition “is necessary to avoid the doctrine becoming uselessly overextended”.Footnote 48
III. Critiques of the Trust Conception
Although the trust conception of legitimate expectations has enjoyed widespread support, it has not been universally accepted, receiving a good amount of criticism – especially in recent years.
Joe Tomlinson, for instance, has criticised the conception for being “artificial” and not connecting “with how the doctrine applies in reality”.Footnote 49 He bases his critique on a distinction drawn between “trusting” and “expecting”. This distinction stems from the “general, ordinary language” definitions that Tomlinson adopts, with “trust” defined as “a firm belief that something is reliable, true, or able” and “expect” defined as “to regard something as likely”. Using these definitions, Tomlinson argues that the conception “ignores that one can expect something without trusting that it will happen” – in other words, “one can regard something as likely to happen (and maybe even hope that it will happen) without holding the firm belief (or trusting) that it will happen”.Footnote 50
Tomlinson accepts that two rejoinders to his argument are possible; but, neither, in his view, saves the conception. First, although it may be argued that a legitimate expectation should require “that a claimant has actually placed trust in a public authority”, this rejoinder fails for Tomlinson because it forces an “arbitrary” and “unjustified distinction, potentially between similar cases, to be made between those individuals who have actually placed trust in a promise, policy, or practice of an administrative body and those who did not (and merely had an expectation)”.Footnote 51 Coughlan, Tomlinson claims, offers an illustration. He points out that while Coughlan sought reassurances about the authority’s promises, evidencing that she was not convinced of those promises and “never actually placed trust in them”, this “would surely not vary the legal analysis”.Footnote 52
Second, Tomlinson acknowledges that a distinction can be drawn between “‘trust’ in an administrative body and ‘trusting’ that the body will adopt a certain course of action”.Footnote 53 He notes, for instance, that “an individual may ‘trust’ that someone will betray them, but, under those conditions, it would be unwise to ‘trust’ them”.Footnote 54 This distinction, Tomlinson says, aligns with Reynolds’s distinction between “specific” and “general” trust. However, he insists that even if we restrict legitimate expectations to cases of “specific” trust, or trust that an administrative body will adopt a certain course of action, the distinction he draws between trusting and expecting persists.Footnote 55
Alexander Brown has also criticised the trust conception. He argues that the conception implies that there “cannot be a legitimate expectation without trust”.Footnote 56 But this, he claims, “needlessly and unhelpfully conflates the bases of the existence of legitimate expectations … with the normative support or grounds for principles of administrative justice that lay down obligations and duties to protect legitimate expectations”.Footnote 57 Aligning with Tomlinson’s trust-expect distinction critique, Brown explains that we can imagine a citizen having “a legitimate expectation that a governmental administrative agency will do or not do something whilst at the same time lacking trust, that is, confidence and faith, that it will actually do or not do that thing”.Footnote 58 Brown further criticises the conception for incompleteness. He highlights that in addition to the “specific” trust Reynolds identifies (trust that the public official will do what they indicated they would do), the protection of legitimate expectations enhances “trust in administrative courts themselves”.Footnote 59
Another critic of the conception is Benny Tai who has criticised it specifically in the context of Hong Kong’s comparative experience. In Ng Siu Tung, where the Hong Kong Court of Appeal recognised the protection of substantive legitimate expectations, Bokhary P.J. recognised trust in public administration as a rationale for legitimate expectations. Rejecting this rationale, Tai says the conception presents an “internal contradiction”.Footnote 60 For him, “if trust in public administration is to be the rationale for recognising [substantive legitimate expectations], then the force for upholding it is stronger, rather than weaker, with large classes of people, for more people may lose faith in government as a result of reneging on government promises”.Footnote 61 But this goes against how the doctrine applies in reality, with courts, including in the UK, only willing to protect substantive expectations where the promise or representation is made to one person or a small group.Footnote 62
And Joanna Bell has criticised the conception for its inability to explain certain aspects of the legitimate expectations jurisprudence.Footnote 63 She highlights, in particular, that the conception cannot explain instances where the courts have enforced legitimate expectations despite no apparent reposing or placing of trust by a citizen. Bell uses the example of Greenpeace,Footnote 64 where the court held the Government to an assurance in a White Paper that its policy on nuclear power would not be altered without the “fullest public consultation” even though “it would not seem to be an instance in which any identifiable individual invested ‘specific trust’ in the Government”.Footnote 65
Moreover, the trust conception has been criticised indirectly, often in tandem with the above direct critiques. Some scholars have argued that no single principle – including trust – underpins the legitimate expectations doctrine, suggesting that it is underpinned by a plurality of values.Footnote 66 Bell, for example, alongside her direct critique, has proposed such a pluralist approach. For her, “[i]n adjudicating under the legitimate expectations ground, the courts accommodate and protect many different normative considerations” – consequently, legitimate expectations, she explains, “cannot be neatly characterised as being ‘about’ protecting a singular value or cluster of values”.Footnote 67 And Daly, Tomlinson and Rebecca Williams have advocated similar pluralist approaches.
IV. A Reimagined Trust Conception
For the above reasons, critics of the trust conception claim that trust does not provide the theoretical underpinning needed for the legitimate expectations doctrine. I propose, however, that trust can provide this underpinning. But first we must better understand what trust means in this context. Proponents of the conception have insufficiently conceptualised trust; and this has yielded ambiguity in the conception, generating problems for it and opening it up to the above criticism. To remedy this, I draw on scholarship on trust from outside law to reimagine the trust conception. This reimagined conception addresses the above issues that the conception’s critics have identified. And I claim that under this reimagined conception, trust can explain what underpins the doctrine.
A. The Under-Conceptualisation and Resulting Ambiguity of Trust
I will start by considering three specific areas of conceptual ambiguity in the trust conception. For each, I will identify the ambiguity and then describe how it is problematic for the trust conception.
1. The form of the citizen’s trust
The first area of conceptual ambiguity concerns the form of the citizen’s trust. Put simply, the conception’s proponents have not made sufficiently clear what it means for the citizen to “trust”. Forsyth never offers a clear definition of trust. Reynolds draws the distinction between “specific” and “general” trust but never defines either. He intimates, rather unhelpfully, that “general” trust denotes “trust” in a public authority “acting fairly, pursuing good administration or not abusing its power”, and “specific” trust means “trust” in the authority, “pursuant to some kind of representation, to do what [that authority] has indicated that it will do”.Footnote 68 But in both instances, it is unclear what it means for the citizen to “trust”.
In particular, the conception’s proponents have not expressly specified the level at which trust operates in the conception. The scholarship on trust from disciplines outside law, including philosophy, sociology and political theory, makes evident that trust arises on several different levels, including affective, cognitive and behavioural levels.Footnote 69 Affective trust is emotional in nature, indicating that the “truster” or the actor who trusts (i.e. the citizen in legitimate expectations cases) has “trustful affects, emotions or motivational structures” towards the “trustee” or actor being trusted (i.e. the public official).Footnote 70 Cognitive trust refers to trust at the level of thoughts, beliefs or expectations that are held by the truster about the trustee.Footnote 71 And behavioural trust represents the manifestation of either the affective or cognitive trust that the truster has in the trustee – in the form of actions. Behavioural trust signals to the trustee, and the world at large, that the truster trusts the trustee.Footnote 72 In the legal literature on the trust conception, however, there is no explicit mention of trust being affective, cognitive or behavioural.
We are required instead to infer the meaning of trust – including whether it is affective, cognitive or behavioural – from the context in which the conception’s proponents use trust. One source of context is the terminology adopted. But this is of little assistance. Most proponents use the language of trust being “reposed” or “placed”. For Forsyth, legitimate expectations protect the trust that “has been reposed in the promise made by an official”.Footnote 73 He also speaks of protecting the trust “placed in public officials”.Footnote 74 Reynolds similarly says legitimate expectations protect trust which “has been reposed in public officials pursuant to a representation that they have made”.Footnote 75 The language of “reposed” or “placed” does not offer clarity because it provides little information about what it means for the citizen to trust, ultimately capable of referring to trust at any of the above levels. We must therefore consider trust’s broader context to infer its meaning.
This broader context includes the statements surrounding proponents’ discussion of trust in the conception. From this broader context, we can infer that both Forsyth and Reynolds understand trust cognitively, reflecting more specifically the citizen’s expectations. Forsyth states that “[t]he individual to whom a promise is made reposes trust in that promise, and consequently expects the promise to be fulfilled”.Footnote 76 This implies that when a truster “reposes” trust in a promise, they hold an expectation, specifically that that promise will be fulfilled. And consequently, the citizen, in reposing trust in the public official, expects that official to deliver on the promise (or representation) made. The same is true in Reynolds’s account. In describing “specific” trust, for instance, he explains that when a public authority makes a representation, the citizen “place[s] trust in that representation and expect[s] it to be fulfilled”.Footnote 77 In parallel to Forsyth’s comments, this suggests that when the citizen “places” trust in a public authority (particularly in a “specific” sense) they have an expectation of that authority, again that they will fulfil their representation.
But Forsyth’s and Reynolds’s cognitive understandings of trust demand reading between the lines. And this conceptual ambiguity is problematic for the trust conception. It makes it unclear what the claimant, under the conception, must do to succeed in their legitimate expectations claim. Forsyth, in his account, says the claimant must demonstrate that they did in fact repose or place their trust in the public official. It will be recalled that for Forsyth, this is an advantage of the trust conception as it indicates which expectations the doctrine ultimately protects – namely those where trust has been reposed.Footnote 78 Reynolds’s account, though less onerous for the claimant, nonetheless treats the claimant’s trust as the basis for their legitimate expectations claim. He proposes specifically that the claimant’s trust can be presumed where they show that “a relevant representation has been made by the public authority and that it has been received by the claimant”.Footnote 79 However, this presumption may be rebutted with evidence showing “that the claimant has exhibited distrust”.Footnote 80 Accordingly, in both accounts, the claimant’s reposing or placing of trust is pivotal to their claim’s success: in Forsyth’s, the claimant must demonstrate it; and in Reynolds’s, the exhibition of their “distrust”, which indicates they did not repose or place trust in the official, disproves their claim. For this reason, it is critical that trust be well defined. It indicates to the claimant what they must establish (or what constitutes “distrust” to disprove their claim) and it guides the courts in determining whether a legitimate expectations claim has been made out.
The conception’s ambiguity about the form of trust also opens it up to criticism. Both Tomlinson’s and Brown’s critiques are rooted in confusion concerning what it means for the citizen to “trust”. Tomlinson uses a “general, ordinary language” definition of trust – which he finds in the dictionary – to distinguish it from “expect”. This leads him to characterise trust as an expectation of especial strength such that it rises to the level of a “firm belief”. Brown similarly draws a distinction between trusting and expecting, defining trust as a strong expectation. He states that when “trust” is applied to expectations, it “implies that the expectations are held with a strong degree of confidence and perhaps even faith”; and applied to the administrative law context, it suggests “that a governmental administrative agent or agency will do or not do something simply because it says it will or because of some other feature of its conduct, without seeking further verification or evidence for this besides the assurance or conduct itself”.Footnote 81 In both instances, the distinction between trusting and expecting is employed to challenge the trust conception, claiming that the citizen can expect an official to do something without necessarily trusting them to do so. The under-conceptualisation and ambiguity of trust’s form under the conception makes these critiques possible. Without a precise definition of trust, it remains open to critics like Tomlinson and Brown to impute their own definition and to use it as the basis for criticising the conception.
2. The object of the citizen’s trust
A second area of conceptual ambiguity in the trust conception concerns the object of the citizen’s trust. By this I mean ambiguity about whom or what the citizen is trusting. In some instances, the conception’s proponents have described the object of the citizen’s trust as the official. Both Forsyth and Reynolds speak of trust being reposed or placed “in public officials”.Footnote 82 Steele refers to trust “in government”.Footnote 83 Thomas discusses the trust placed in “the administration” and “government”.Footnote 84 And Watson talks of confidence “in the promise maker”.Footnote 85 But in other instances, the object of the citizen’s trust is, instead of the official, the promise or representation the official has made. Forsyth, it will be recalled, states that “[t]he individual to whom a promise is made reposes trust in that promise”.Footnote 86 Reynolds, at some points, defines the trust conception in terms of “the trust of the governed in the public authority’s representation”.Footnote 87 And McCloskey J., in setting out his “two basic ingredients” for substantive legitimate expectations, identifies as the second ingredient, the citizen’s having reposed trust in the “promise or assurance by a public official”.Footnote 88
This inconsistent terminology yields ambiguity over whether trust, under the conception, is procedurally or substantively oriented. To explain, I will start by assuming that trust under the conception denotes cognitive trust. I will justify later why I think this should be so. But for now, I will assume it to be true based on the inferences I drew earlier from Forsyth’s and Reynolds’s accounts. The scholarship on trust from various disciplines recognises that with respect to trust at a cognitive level, two categories of expectations operate. The first is procedural expectations. When a truster trusts a trustee (like a public official), the truster holds positive expectations regarding the trustee’s goodwill and competence, expecting the trustee to exercise both in their interaction with the trustee.Footnote 89 As philosopher Annette Baier explains, trust is a “belief in the trusted’s goodwill and competence” that “grounds the willingness to be or remain within the trusted’s power in a way the distrustful are not”.Footnote 90 Applying these expectations to legitimate expectations cases, trust therefore means that the citizen expects the official to exercise goodwill and competence with respect to the promise or representation they have made. For example, the citizen expects the official to make best efforts not to over-promise or over-represent and to make best efforts to fulfil that promise or representation. These expectations of goodwill and competence are procedural in nature because they focus on the procedure followed by the trustee or public official – both in making and fulfilling the promise or representation. There is additionally, however, a second category of expectations: substantive expectations.Footnote 91 Rather than an expectation about the procedure followed by the trustee, a substantive expectation is an expectation held by the truster that their interaction with the trustee will have a favourable outcome. In legitimate expectations cases, it translates into an expectation that the public official will ultimately fulfil the promise or representation made.
To be clear, the distinction I am drawing between procedurally and substantively oriented trust does not necessarily map onto the distinction in the legitimate expectations literature between procedural and substantive expectations.Footnote 92 Substantively oriented trust denotes the citizen’s expectation that the public official will fulfil their promise or representation – that is, that they will do what they have indicated they would do. However, that promise or representation (i.e. what the official indicated they would do) may be procedural. This was the case in GCHQ, for instance, where the representation in question was that the Minister for the Civil Service would consult GCHQ staff on significant changes in their conditions of service (based on established practice).Footnote 93 Procedurally oriented trust, in contrast, refers to the citizen’s expectation that certain procedural steps will be taken by the official, even though those steps have not been promised or represented.
When the trust conception’s proponents refer to the promise or representation itself as the object of the citizen’s trust, it suggests substantively oriented trust. The citizen trusts the official’s promise or representation to be true and thus expects it to be fulfilled. When they refer to the public official as the object, however, it is less clear: trust could be substantively oriented, with the citizen expecting the official to fulfil their promise or representation, or it could be procedurally oriented, with the citizen expecting the official to make best efforts, both not to over-promise or over-represent and to fulfil the promise or representation. And by fluctuating between these two objects of the citizen’s trust, proponents have introduced ambiguity into the conception.
The distinction between procedurally and substantively oriented trust is what I think Reynolds seeks to address when he distinguishes “specific” from “general” trust. In fact, that Reynolds saw it necessary to draw the distinction evinces the ambiguity in the conception. For Reynolds, “specific” trust is substantively oriented, reflecting the citizen’s trust in the official “to do what [they have] indicated that [they] will do”.Footnote 94 This is essentially an expectation of a favourable outcome, namely the fulfilment of the promise or representation. And what Reynolds calls “general” trust is, it seems, procedurally oriented. To repeat from earlier, Reynolds says “general” trust reflects the citizen’s trust “in the authority acting fairly, pursuing good administration or not abusing its power”.Footnote 95 However, Reynolds’s description of “general” trust, including whether it involves procedural (as opposed to substantive) expectations, is unclear.
In fairness, proponents of the trust conception have attempted to clarify this ambiguity about the object of trust. Reynolds, in stressing that the conception is concerned with “specific” trust rather than “general” trust, has pushed trust under the conception towards a substantive orientation. And we can infer from Forsyth’s account that his focus is likewise on substantively oriented trust. This conclusion follows on from his statement that “[t]he individual to whom a promise is made reposes trust in that promise, and consequently expects the promise to be fulfilled”.Footnote 96 Expecting the promise to be fulfilled, in line with what I have said, amounts to a substantive expectation. But much like the conception’s ambiguity about form, its view of trust as substantively rather than procedurally oriented requires some reading between the lines.
This ambiguity is again problematic for the conception. In particular, the ambiguity makes it unclear what it means to protect the citizen’s trust – the objective, under the trust conception, of the legitimate expectations doctrine. If trust is procedurally oriented, the doctrine requires the courts to enforce procedural safeguards against officials making and seeking to fulfil the promises or representations they make. If trust is substantively oriented, however, the doctrine obliges the courts to enforce the promises or representations themselves. Hence, a clear indication of trust’s object is necessary if what we seek to do under the conception is to protect the citizen’s trust.
3. The party of focus in the citizen–government relationship
Lastly, there is ambiguity in the trust conception about which party in the citizen–government relationship – the citizen or the public official – is the focus. Trust arises in a relationship between two parties: the truster and the trustee or, in legitimate expectations cases, the citizen and the official. When we examine trust in that relationship, we can focus on, or consider trust from the perspective of, either. In the trust conception, however, the relevant party of focus is unclear.
This ambiguity stems from the two ways in which the conception’s proponents have used trust. First, proponents have used trust descriptively to describe what the legitimate expectations doctrine protects. Under the conception, the doctrine protects the citizen’s trust; and following on from this, the question we must ask, they suggest, is whether the citizen has reposed or placed trust – either in the official, or their promise or representation. This “simple concrete question of fact”, Forsyth says, determines “what is lawful and what is not”.Footnote 97 But proponents have also used trust normatively to justify why legitimate expectations should protect trust. They argue that protecting the citizen’s trust in legitimate expectations cases helps foster public trust in government – a valuable resource for government as it encourages public cooperation. Forsyth, it will be recalled, claims that “[g]ood government depends upon trust between the governed and the governor”. He says we need public trust in government because without it “officials will not be believed and the government becomes a choice between chaos and coercion”.Footnote 98 Reynolds likewise contends that “trust in public officials is essential to a properly functioning political system”.Footnote 99
In proponents’ first use of trust, the party of focus is the truster or citizen – and trust is the primary concern: we focus on the citizen and whether they have actually reposed or placed trust in the official (or their promise or representation). In proponents’ second use of trust, however, the situation is different. Rather than the truster or citizen, the party of focus is the trustee or official. This is because the aim of legitimate expectations under the trust conception is not to foster all forms of public trust in government. Trust can sometimes be misplaced. The aim is to foster what political theorist Mark Warren terms “warranted” trust in government (or philosopher Onora O’Neill describes as “intelligent” trust) – that is, trust where government actors are worthy of public trust.Footnote 100 So, under the trust conception, the legitimate expectations doctrine, rather than foster public trust in government per se, seeks to create the conditions under which the citizen is able to trust their government. And the trust conception’s proponents seem to recognise this point. Bokhary P.J.’s discussion of trust in Ng Siu Tung says the protection of legitimate expectations “facilitates the task of governance” because “people feel able to put their faith in what their government says and does”.Footnote 101 Reynolds notes that “when we feel unable to trust [public authorities], we place no weight on their representations to us, thereby denying them respect, legitimacy and the ability to govern well”.Footnote 102 Thomas indicates that the doctrine of “[l]egitimate expectations compels the administration to be trustworthy”.Footnote 103 And Schønberg explains that “public authorities are unlikely to be perceived as trustworthy if they are free to, and do in fact, go back on representations with impunity”.Footnote 104 Because of this, in this second use of trust by the conception’s proponents, the citizen’s trust is a secondary concern. The primary concern is the trustworthiness of the official, ensuring that they are worthy of the citizen’s trust in them. In fact, empirical scholarship on trust has demonstrated that it is this trustworthiness, with the public’s warranted or intelligent trust that follows on from it, on which the above-noted valuable ends of “[g]ood government” and “a properly functioning political system” depend. Public attributions of the trustworthiness of an authority, this research shows, are connected with the public’s willingness to accept the authority’s decisions, the public’s feelings of obligation to obey laws and the public’s performance evaluations of the authority – resulting in public cooperation with government.Footnote 105
This ambiguity about the party of focus in the citizen–government relationship likewise poses problems for the trust conception. First, it makes it unclear whether, under the conception, trust is required to make out a claim for legitimate expectations. Forsyth and Reynolds, it will be recalled, both say trust is required for a successful legitimate expectations claim. However, if the party of focus is the official – with the aim of the legitimate expectations doctrine, under the conception, being to promote government trustworthiness – it is unclear, despite what Forsyth and Reynolds say, whether the citizen must actually repose or place trust. Government trustworthiness can be promoted even though the citizen may not have reposed or placed trust.
Second, this ambiguity opens the conception up to additional criticism. Both Tai’s and Brown’s critiques of the conception arise out of this ambiguity. Tai’s critique that the conception presents an “internal contradiction” follows on from an assumption he makes: that the party of focus is the citizen and the primary concern of the conception is their trust. More precisely, Tai assumes that the aim under the conception is to foster as much public trust in government as possible. Based on this assumption, he proposes that, in contradiction to what the courts have held, the case for protecting substantive legitimate expectations is “stronger … with large classes of people, for more people may lose faith in government as a result of reneging on government promises”.Footnote 106 In other words, we can expect that protecting the expectations of larger classes will yield the greatest increase in public trust in government. This “internal contradiction” does not exist, however, if the aim under the conception is government trustworthiness rather than public trust in government. In that case, as I explain shortly, the number of people to whom the promise or representation is made is irrelevant, because it has no bearing on the official’s trustworthiness.
Somewhat similarly, Brown’s critique of the conception on incompleteness grounds stems from his making the same assumption as Tai: that the citizen is the party of focus and their trust the primary concern. And assuming as much, Brown challenges the conception for ignoring the citizen’s trust in other actors, namely the courts. Again, however, this critique does not hold water where the aim under the conception is government trustworthiness. Whereas the citizen’s trust in the courts may be significant, it has no bearing on the objective of government trustworthiness.
B. Reimagining the Trust Conception
Given these areas of conceptual ambiguity and the problems they pose for the trust conception, it is imperative that we more clearly conceptualise trust under the conception. I will therefore turn now to this task and, in the process, offer a reimagined trust conception of legitimate expectations.
1. Trust is cognitive, not affective or behavioural
We should begin by clarifying that the trust conception is concerned with cognitive trust. More specifically, and in line with the doctrine it seeks to explain, it is concerned with expectations.
Affective trust has little relevance for the legitimate expectations doctrine – or indeed administrative law more broadly. It usually arises in relationships of shared interests between the truster and trustee, merging into a shared identify (e.g. parent–child or marital relationships).Footnote 107 This does not apply to legitimate expectations cases where the relationship at issue is between the citizen and public official. Affective trust is also less contingent on external circumstances, like legal constraints imposed on the trustee. It is thus of limited use as a guiding concept for law.
Behavioural trust, although relevant for both the conception and the legitimate expectations doctrine, is inappropriate for the purpose of defining trust in the trust conception. First, as I highlighted earlier, it does not fit with how proponents understand trust. From the context in which they use the term, there are strong indications that they understand trust cognitively. Second, adopting a behavioural interpretation of trust has problematic consequences for the conception. I said above that behavioural trust has relevance for the trust conception and the legitimate expectations doctrine. I say this because behavioural trust aligns with what the legal literature calls “reliance”. Like trust, reliance is neither defined nor conceptualised in the legitimate expectations literature. However, we can again infer from context that reliance is understood (at least usually) as behaviour.Footnote 108 Put simply, the citizen “relies” on the public official’s promise or representation where they act on their expectation that that promise or representation will be fulfilled. And if we understand reliance as such, it essentially amounts to behavioural trust.
A behavioural interpretation of trust is problematic, however, because courts in the UK have recognised that reliance is not a requirement of a successful legitimate expectations claim. In Bancoult, for example, Lord Hoffmann said it “is not essential that the applicant should have relied upon the [public official’s] promise to his detriment”; it is, rather, “a relevant consideration in deciding whether the adoption of a policy in conflict with the promise would be an abuse of power”.Footnote 109 If trust under the conception is understood behaviourally, the conception effectively translates into a requirement of reliance. As the conception protects the trust that the citizen has reposed or placed in the public official (or their promise or representation), requiring the citizen to have reposed or placed such trust, interpreting trust behaviourally means the citizen must have acted on their expectation that the official’s promise or representation would be fulfilled. This is not in line with the jurisprudence – and hence, would offer the conception little explanatory power.
Aside from this problem of explanation, interpreting trust behaviourally is also normatively problematic. In Bibi, Schiemann L.J. emphasised that “reliance” cannot be made a requirement of legitimate expectations because it would introduce unfairness to the doctrine. As he explains, it would “place the weakest in society at a particular disadvantage” since “those who have a choice and the means to exercise it in reliance on some official practice or promise would gain a legal toehold inaccessible to those who” do not have that choice.Footnote 110 Put simply, legitimate expectations would be limited to those who can afford to act on the promises and representations of officials.
The better interpretation is thus to understand trust cognitively. Not only does it align with the jurisprudence and fit with the inferences drawn from Forsyth’s and Reynolds’s accounts, but it makes sense in the context of the relevant doctrine – legitimate expectations. A cognitive understanding means the citizen, in trusting, has certain expectations about their interaction with the public official even though they may not have necessarily acted on those expectations.
Understanding trust cognitively also addresses the critiques of the conception rooted in the distinction between trusting and expecting. Understood cognitively, there is no distinction to be drawn between expecting and trusting: trusting is expecting. While both Tomlinson and Brown draw this distinction, neither supports it with the scholarship on trust. Tomlinson’s is based on dictionary definitions; and while such definitions have their place in legal analysis, his use of them here is problematic given the contentious nature of trust, the conceptual scholarship available and the centrality of trust to the conception he criticises. Brown’s characterisation of trust as expectations held “with a strong degree of confidence and perhaps even faith” is likewise not based on the scholarship, and simultaneously conflates trust with confidence and faith – concepts which writers, from sociologists to political scientists, have distinguished from trust.Footnote 111 With neither Tomlinson nor Brown basing their distinction on the scholarship, their arguments that the citizen can expect without trusting the official (or their promise or representation) are unconvincing.
2. Trust is substantively oriented but rooted in procedurally oriented expectations
We should also be precise about the nature of the citizen’s trust expectations. I agree with Reynolds that the relevant trust under the trust conception is the citizen’s trust in the official to do what they have indicated they will do – what he calls “specific” trust. But such trust (i.e. in an actor to fulfil a promise or representation) is not simply a substantively oriented expectation. As some philosophers have proposed, it is better represented as a hybrid of expectations. Whereas this trust does denote the truster’s substantively oriented expectation that the trustee will fulfil the promise or representation made, that expectation is rooted in the truster’s procedurally oriented expectations about the trustee’s goodwill and competence.Footnote 112 The truster expects the trustee to fulfil their promise or representation because they expect the trustee to exercise goodwill and competence towards them. As philosopher David Owens says, such trust “is not just to rely on the promise being fulfilled for whatever reason; it is to rely on the promisor’s conscientiousness in particular”.Footnote 113 If the truster expects the trustee to fulfil their promise for reasons other than goodwill and competence (e.g. fearing repercussions or due to incentives), it does not reflect trust.
Applied to legitimate expectations, the citizen’s trust in the public official to do what they have indicated they will do thus involves both substantively and procedurally oriented expectations. In trusting the public official, the citizen expects the official to fulfil their promise or representation because they expect the official to exercise goodwill and competence towards them.
This perhaps explains the distinction Tomlinson and Brown draw between trusting and expecting. Rather than trust representing an especially strong expectation, as both of them suggest, it reflects expectations of a particular nature. Trust, in these cases, is an expectation of a favourable outcome rooted in expectations of the trustee exercising goodwill and competence. So, to say the citizen can expect without trusting, as Tomlinson and Brown do, is misleading. More accurately, the citizen can expect the official to fulfil their promise or representation without having expectations that the official will exercise goodwill and competence towards them. If the citizen has the former expectation for other reasons, it does not reflect trust. Nevertheless, their critique becomes moot if the trust conception’s focus is really trustworthiness, as I propose shortly.
The above said, because the citizen’s substantively oriented expectation follows on from their procedurally oriented expectation, trust under the conception, we can conclude, is effectively procedurally oriented. In trusting the public official, the citizen expects them to exercise goodwill and competence. Where the official does so, they have fulfilled the citizen’s trust. But where they do not exercise goodwill and competence, they have breached or betrayed that trust. Now, as I have said, the citizen’s expectation that the official’s promise or representation will be fulfilled follows on from their expectations of goodwill and competence. However, the fulfilment of that expectation is neither necessary nor sufficient for the citizen’s trust to be fulfilled. The official’s failure to fulfil their promise or representation does not mean the official has breached or betrayed the citizen’s trust because they may have not fulfilled their promise or representation for various reasons, despite exercising goodwill and competence towards the citizen.Footnote 114 This is implicitly recognised by Steele, explaining that “in some cases public confidence in government may in fact be enhanced by judicious departure from, rather than slavish adherence to, past promises”.Footnote 115 Put simply, the official’s exercise of goodwill and competence towards the citizen may lead them to conclude that their promise or representation should not be fulfilled. At the same time, their fulfilling their promise or representation does not necessarily mean they have fulfilled the citizen’s trust, if, for example, they have fulfilled that promise or representation for reasons other than their goodwill and competence.
To be clear, the citizen’s expectations of the official’s goodwill and competence, to which I am referring, do not equate with Reynolds’s category of “general” trust. First, unlike “general” trust which seems, based on Reynolds’s description, to relate to the public official generally in their role as official (i.e. how they make decisions, broadly speaking, as an official), the expectations of goodwill and competence relate specifically to the promise or representation the official has made to the citizen. The citizen expects the official – in making the promise or representation as well as seeking to fulfil that promise or representation – to exercise goodwill and competence towards them. Second, these expectations are not expectations of compliance with broad administrative law principles, including fairness, good administration and non-abuse of power. Recognising that the official’s promise or representation gives rise to a right or interest in the citizen, the expectations of goodwill and competence go beyond what is normally available at administrative law. I will consider the citizen’s expectations of goodwill and competence in more detail shortly.
Conceptualising trust in this way, as both substantively and procedurally oriented, helps to resolve one of the greatest problems, in my view, for the trust conception as currently formulated. If we understand trust as only substantively oriented – simply as an expectation that the official will fulfil their promise or representation – trust can only explain part of the legitimate expectations doctrine. This is because when a legitimate expectations claim is made, two questions arise: “whether an expectation has arisen and, if so, how (if at all) it should be protected.”Footnote 116 Substantively oriented trust can explain the first question – that is, which expectations are engaged by the doctrine (i.e. where the citizen has reposed or placed trust). It cannot, however, explain which expectations the courts will protect. A public official is entitled to resile from their promise or representation, despite it giving rise to a legitimate expectation, where it is in the “public interest”. In Finucane, for instance, Lord Kerr, writing for the Supreme Court, explained: “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.”Footnote 117
Trust, if substantively oriented, offers no guidance on this second question. It does not indicate when the public official so resiling will be in the public interest and when it will not be in the public interest with the courts choosing instead to protect the citizen’s legitimate expectation. Owing to this, Forsyth acknowledges that trust cannot provide “the whole answer”, noting that reposing trust “is a necessary but not a sufficient condition for the protection of an expectation”.Footnote 118
A substantively oriented understanding of trust also has troubling implications. Where a court permits a public official to resile from their promise or representation in the public interest, substantively oriented trust suggests that the official has breached or betrayed the citizen’s trust. They have not delivered on their promise or representation and so, have failed to fulfil the citizen’s expectation that it would be fulfilled (i.e. their trust). This is troubling for two reasons. First, it leads to unreasonable conclusions. There can be any number of reasons why an official may not be able to fulfil their promise or representation. As Schiemann L.J. recognised in Bibi, the official or their superior, focusing on the “wider background”, may realise that “the making of the promise was unwise or that, in any event, its fulfilment seems too difficult”.Footnote 119 It is unreasonable to state that in not fulfilling the promise or representation in such circumstances, the official has breached or betrayed the citizen’s trust. Simultaneously, this understanding conveys a problematic message to the citizen: that a public authority’s decision to pursue the wider public interest over fulfilment of their promise or representation constitutes a breach or betrayal of the citizen’s trust – a message that is detrimental not only to the citizen–government relationship but also wider social solidarity.
We eliminate these issues by understanding trust as procedurally as well as substantively oriented. Doing so, first, permits the trust conception to explain when a citizen’s legitimate expectation will be protected. As I describe next, the expectations of goodwill and competence can encapsulate considerations of the wider public interest. Despite exercising goodwill and competence, the official may not be able to fulfil the promise or representation made to the citizen. Second, it leads to a more reasonable conclusion and conveys a more suitable message. Procedurally oriented, it is only where the official does not exercise goodwill and competence vis-à-vis their promise or representation that the citizen’s trust is considered breached or betrayed.
3. Trustworthiness – not trust – is the conception’s primary concern
Lastly, I suggest that under the trust conception, the party of focus is the public official rather than the citizen; and the primary concern is the former’s trustworthiness rather than the latter’s trust.
Focusing the trust conception on the citizen’s trust is problematic for many reasons. First, it renders the conception impractical. Assuming that by trust we mean cognitive trust (as I have argued), “it is difficult (if not impossible) to have access to the state of mind of each person”.Footnote 120 How do we know whether the citizen has in fact reposed or placed trust in the official (or their promise or representation)? We can use the citizen’s behaviour, or “reliance” on the promise or representation, to support a conclusion that the citizen did trust. However, there are several issues with relying on the citizen’s reliance or lack thereof, including that reliance is not a requirement for the legitimate expectations doctrine, the citizen may behave as if they trusted the official without trusting them and the citizen may “emotionally” rely on the official (trusting them without any corresponding action).Footnote 121 Reynolds circumvents the difficulty of proving the citizen’s state of mind by presuming the citizen’s trust – a presumption that, we will recall, is rebuttable.Footnote 122 But even if we incorporate this presumption into the trust conception, the inclusion of a requirement that the citizen has reposed or placed trust poses the previously described problem of unfairness that Tomlinson has identified.Footnote 123 And Reynolds’s rebuttable presumption penalises claimants who are less trusting and more likely to exhibit distrust in the promise or representation.
Additionally, in line with Bell’s critique of the trust conception, focusing on trust does not align with the jurisprudence. The requirement that the citizen reposed or placed trust in the public official (or their promise or representation) means that the citizen must have had an expectation of the official – based on my earlier argument, an expectation that the official will exercise goodwill and competence towards the citizen vis-à-vis their promise or representation. However, in legitimate expectations cases, the courts are “not concerned with what the [citizen] actually expected”.Footnote 124 They are concerned with “what could reasonably be expected”.Footnote 125 In Begbie, for example, the court focused its attention on how the “reasonable informed reader” of the newspaper article at issue in that case would have interpreted it, specifically if they “could believe that it was the announcement of a change of the [relevant] policy” on class sizes.Footnote 126 The court did not concern itself with what the applicants actually expected as a result of the article. Consequently, a trust conception that centres on trust is inconsistent with the courts’ approach to the doctrine.
Focusing the trust conception on the official’s trustworthiness eliminates these problems. Trustworthiness reflects the likelihood that the trustee will fulfil the truster’s trust.Footnote 127 As political scientist Paul Bauer says, “implicit behind many theoretical elaborations” is the idea that trustworthiness describes “a probability, namely the probability that a trustee acts as expected by the trustor”.Footnote 128 So, applied to legitimate expectations cases, it reflects the likelihood that the official will fulfil the citizen’s trust, exercising both goodwill and competence. If the official is likely to exercise goodwill and competence towards the citizen, they are trustworthy. Thus, focusing the trust conception on the official’s trustworthiness means that in applying the legitimate expectations doctrine, the courts require the official to act trustworthily towards the citizen, exercising both goodwill and competence towards the citizen vis-à-vis their promise or representation. Where the official acts trustworthily towards the citizen vis-à-vis their promise or representation, the citizen’s claim against them fails; where they do not act trustworthily, however, the citizen’s claim succeeds.
In this reimagined trust conception, trustworthiness operates both retrospectively and prospectively. It operates retrospectively as a standard to which the public official is held by the courts. The official is required, in making and seeking to fulfil their promises or representations to citizens, to act trustworthily towards those citizens. Simultaneously, however, it operates prospectively as a normative justification for focusing the legitimate expectations doctrine on government trustworthiness. By holding public officials to the above standard, the courts promote government trustworthiness. They make it more likely that officials, when making and seeking to fulfil their promises or representations, will exercise goodwill and competence towards citizens. Officials know that if they do not, they will be held accountable by the courts. And this promotes, in turn, public trust in government because it enables citizens to feel more able to trust, thereby generating the valuable end of cooperation that the trust conception’s proponents have noted.
Understood in this way, the trust conception does not require the citizen to have trusting expectations. It is concerned with creating the conditions under which the official is likely to fulfil the citizen’s expectations of goodwill and competence, regardless of whether they have them. Hence, the legitimate expectations doctrine does protect the trust of citizens; it is not, however, the trust they have actually reposed or placed in the official, but the trust they may repose or place. Owing to this, all the above problems arising out of a requirement that the citizen trust fall away.
Focusing the trust conception on trustworthiness rather than trust also addresses Tai’s concern about the conception presenting an “internal contradiction”. If the aim of the legitimate expectations doctrine is promoting the official’s trustworthiness, the number of people to whom the promise or representation is made is irrelevant; it has no bearing on official trustworthiness. The focus is the public official, no matter how many people received the promise or representation. Nevertheless, it makes sense for the courts to consider the number of people to whom the promise or representation was made in substantive legitimate expectations cases because it has relevance for certain parts of the courts’ analysis. For example, the smaller the number of people to whom the promise or representation was made, the more reasonable it would be for the citizen to regard it as “clear and unambiguous” as well as to expect the promise or representation to be fulfilled.Footnote 129
Lastly, focusing on the official’s trustworthiness addresses a broader problem with the legitimate expectations doctrine. The legitimate expectations terminology “is rather odd because it is described from the claimant’s perspective – it is [they] who [have] the legitimate expectation”.Footnote 130 This is at odds with other grounds of review which “tend to tell the decision-maker how to act – for example, reasonably, proportionately, taking all relevant considerations into account and discarding all irrelevant considerations”.Footnote 131 It has therefore been argued that “the citizen’s state of mind is less important than encouraging careful deliberation as to what decision-makers promise, and their careful deliberation before breaking those promises”.Footnote 132 Shifting our focus under the trust conception from the citizen’s trust to the official’s trustworthiness shifts our perspective from that of the claimant to that of the official, telling the decision maker how to act – namely trustworthily.
That said, how does an official act trustworthily? A thorough account of government trustworthiness is beyond this article’s scope. However, for the purpose of my argument that trust can provide the theoretical underpinning for legitimate expectations, I want to illustrate how trustworthiness can explain the principles laid down by the courts in their jurisprudence as well as the arguments advanced by academic commentators. In line with what I have said, trustworthiness demands that the official exercise goodwill and competence towards the citizen vis-à-vis their promise or representation – both in making that promise or representation and then fulfilling it.
Collectively, goodwill and competence reflect conscientiousness.Footnote 133 The official recognises that owing to their promise or representation, they have an obligation to the citizen, which they fulfil in a diligent, responsible and careful manner. I thus suggest that goodwill translates into the official making best efforts vis-à-vis their promise or representation: they do their best not to over-promise or over-represent to the citizen and once they have promised or represented, take that promise or representation seriously, doing their best to fulfil the promise or representation. And competence reflects appropriate knowledge and skills, demanding that the official exercise such knowledge and skills both when making and when seeking to fulfil their promise or representation.
If we understand the official’s trustworthiness in this way, it explains the balancing exercise in which the courts engage in legitimate expectations cases, balancing the interests of the citizen in having the promise or representation made to them fulfilled against the wider public interest. In so balancing, the courts ensure that the official has exercised both goodwill and competence in their attempt to fulfil the promise or representation. Because goodwill requires the official to take their promise or representation seriously and make best efforts to fulfil it, where an official refuses to fulfil it, they must justify it. Without sufficient justification, it cannot reasonably be said that the official has taken their promise or representation seriously or made best efforts to fulfil it. Accordingly, trustworthiness explains the various circumstances in which the courts have decided that public authorities are entitled to resile from their promise or representation, including where what was promised or represented threatens national security,Footnote 134 conflicts with the rights or interests of other citizens which take priorityFootnote 135 and has unreasonable costs or implications for government finances.Footnote 136 These circumstances constitute sufficient justification from the authority, evidencing that they took their promise or representation seriously and made best efforts to fulfil it. And similarly, because competence requires the official to exercise appropriate knowledge and skills, they must base their attempt to fulfil their promise or representation on relevant information. If they do not do so, it cannot reasonably be said that they have exercised appropriate knowledge and skills. This explains the various considerations that the courts have required public authorities to take into account when seeking to fulfil their promise or representation, including the legitimate expectations of the citizenFootnote 137 and the views of affected citizens via consultation.Footnote 138
I will use Bibi to illustrate. There, the London Borough of Newham, misunderstanding its statutory obligations, promised the applicants legally secure accommodation within 18 months. When Newham refused to keep its promise, the applicants challenged the decision as a breach of their legitimate expectations. The lower court agreed and issued an order declaring that Newham was “bound to treat the duties originally owed by [it to the applicants under statute] as not discharged until the applicants [were] provided with suitable accommodation on a secure tenancy”.Footnote 139 The Court of Appeal, however, engaging in a balancing exercise, concluded that while Newham had “legally committed itself to providing the applicants with suitable accommodation with secure tenure” (which gave rise to a legitimate expectation in the applicants), it should not be held to its promise.Footnote 140 Writing for the court, Schiemann L.J. recognised that there was a potential conflict between “the ‘legitimate aspirations’ of those who have been told where they are on the housing waiting list and what [Newham’s] allocation scheme is” and “the ‘legitimate expectations’ of those to whom promises have been made by [Newham] the fulfilment of which conflicts with the priorities contained in the allocation scheme”.Footnote 141 And Newham, he reasoned, was “the appropriate body” to resolve this conflict, given that such “invidious choices” are informed by “social and political value judgments as to the priorities of expenditure”.Footnote 142 That said, Schiemann L.J. concluded that Newham was obliged to properly take into account – in its decision-making process – the promise it had made and the legitimate expectation it had created.Footnote 143 And following on from this, he issued a declaration that Newham was “under a duty to consider the applicants’ applications for suitable housing on the basis that they have a legitimate expectation that they will be provided by [Newham] with suitable accommodation on a secure tenancy”.Footnote 144
Bibi can be explained in trustworthiness terms – that is, the court, via its judgment, required Newham to act trustworthily towards the applicants. First, Schiemann L.J.’s reference to the potential conflict between the “legitimate expectations” of the applicants and the “legitimate aspirations” of other citizens speaks, although indirectly, to the exercise of goodwill by Newham. It recognises that Newham may have a justification for not fulfilling the promise it had made to the applicants – that justification stemming from the “legitimate aspirations” of other citizens based on its allocation scheme. In fact, Schiemann L.J. said that the lower court judge’s order “went too far” because “it seems implicit in his declaration that there can not be factors which inhibit the fulfilment of the legitimate expectations”.Footnote 145 Granted, the court could have been more direct, explicitly requiring Newham to justify its failure to fulfil the promise made to the applicants. In Nadarajah, Laws L.J. stated this requirement explicitly, noting that “[w]here a public authority has issued a promise or adopted a practice which represents how it proposes to act in a given area, the law will require the promise or practice to be honoured unless there is good reason not to do so”.Footnote 146 And as we will see shortly, this was the approach of the House of Lords in Bancoult. While Laws L.J. rooted this requirement in good administration, I think that rather than the overly abstract and over-inclusive principle of good administration, the requirement is more precisely captured by trustworthiness, specifically under goodwill and its definition in terms of best efforts.
Second, Schiemann L.J.’s requirement that Newham consider the promise it had made to the applicants, and the legitimate expectation its promise had created, further speaks to Newham’s exercise of goodwill as well as the competence of its decision-making process. If goodwill reflects, as I have indicated, an authority taking their promise or representation seriously and making best efforts to fulfil it, it seems reasonable that an authority, like Newham, to exercise goodwill, must take into account the promise it made and legitimate expectation it created. Such taking into account shows the authority is taking its promise or representation seriously. At the same time, a decision-making process that does not take into account a promise made, and legitimate expectation created, cannot reasonably be considered competent. The promise and corresponding expectation are relevant to the decision being made, reflecting appropriate knowledge. Thus, the court’s requirement that Newham take the applicants’ legitimate expectation into account can be seen as a demand that it exercise appropriate knowledge in making its housing allocation decision.
Bancoult offers a further illustration. On agreement with the US government that a British Indian Ocean Territory (BIOT) island would serve as a US military base, the UK Government secured the removal of the islands’ inhabitants, the Chagossians. And the BIOT Commissioner, in 1971, made an Immigration Ordinance prohibiting the Chagossians from visiting without a permit.Footnote 147 But in 2000, in response to a court judgment, the Foreign Secretary issued a press release indicating that the Government was putting “in place a new Immigration Ordinance” which would allow the Chagossians “to return to the outer islands”; and the Commissioner revoked the 1971 Ordinance and made a new one which did not include the restrictions.Footnote 148 However, the Government subsequently decided that resettlement of the islands was not feasible and that the territory was still needed for defence purposes. The Queen, exercising her prerogative, thus made two Orders in Council in 2004 which prevented the Chagossians from returning to the islands.
The applicant claimed that the press release and 2000 Ordinance were a promise which created a legitimate expectation that the islanders would be free of immigration controls. The House of Lords majority dismissed the applicants’ legitimate expectations claim for failing the preliminary step of establishing a “clear and unambiguous promise” from the Government (giving rise to a legitimate expectation).Footnote 149 It said, however, that if a “clear and unambiguous promise” had been established, the claim would nonetheless have failed since the Government had “sufficient public interest justification” for resiling from the promise.Footnote 150 Lord Carswell, for example, accepted the Government’s argument that if it were obliged to resettle the Chagossians, “it could give rise to friction with the United States”.Footnote 151 Lords Mance and Bingham, in dissent, found that the press release and 2000 Ordinance did constitute a “clear and unambiguous promise” and concluded that the Government should be held to its promise.Footnote 152 Lord Bingham said the “Government could not lawfully resile from its representation without compelling reason” – which, for him, had not been shown.Footnote 153 And Lord Mance emphasised that the Government could not go back on its promise “without any consultation and without strong cause” – which again, for him, had not been demonstrated.Footnote 154 He noted that there had been “no consultation with the Chagossians or anyone” before the 2004 Orders were issuedFootnote 155 and said the Government’s justification for the Orders had not been “substantiated in legal or, to any realistic extent, in practical terms”.Footnote 156
Both the majority and dissenting judgments in Bancoult can likewise be explained in trustworthiness terms. Their references to the Government having “justification”, “compelling reason” or “strong cause” for resiling from its promise speak to the Government’s exercise of goodwill. They effectively require the Government to make best efforts to fulfil the promise made to the Chagossians, only permitted to resile from it with sufficient justification. The Law Lords in the majority concluded that there was such justification; and those in the dissent found there was not. Further, Lord Mance’s requirement that the Government consult with the Chagossians before having issued the 2004 Orders speak to both the Government’s exercise of goodwill and the competence of its decision-making process. It seems reasonable that a public authority who takes their promise seriously and makes best efforts to fulfil it would consult with those affected by the decision to resile from the promise. And simultaneously, the Chagossians’ views were relevant to the decision before the Government, thus reflecting appropriate knowledge. Therefore, much like in Bibi, the House of Lords demanded that the Government act trustworthily, although the Law Lords disagreed on what trustworthiness included and how it applied in the case.
It should be recognised, however, that this balancing exercise raises separation of powers issues. The courts, as always, cannot overstep their institutional limits. Hence, while the courts may seek to ensure that public authorities act trustworthily, exercising goodwill and competence in seeking to fulfil a promise or representation, that exercise is limited by the courts’ relative legitimacy and competence. It is for this reason that the courts have shown deference to public authorities where, for example, the promise or representation pertains to a matter in which the executive has more competence,Footnote 157 concerns macro-political issuesFootnote 158 and affects a large number of people.Footnote 159 In Bibi, for instance, the court recognised that the “invidious choices” involved in the provision of housing at public expense are “essentially political rather than judicial”, making Newham “the appropriate body” to make decisions about to whom to provide accommodation.Footnote 160 These various factors do not speak to trustworthiness; they reflect, rather, recognition by the courts that their evaluations of the official’s reason for not fulfilling their promise or representation and the relevance of information to the official’s attempt to fulfil are subject to institutional limits.
Additionally, because trustworthiness requires the public official to exercise goodwill and competence in making their promises or representations, it also explains principles that fall outside the above balancing exercise. For instance, trustworthiness supports the idea advocated by some writers that there should not be a “blanket rule of non-enforcement of ultra vires” promises or representations.Footnote 161 Because goodwill requires the official to make best efforts not to over-promise or over-represent, an official who knowingly makes an ultra vires promise or representation does not exercise goodwill. And an official who does not know that their promise or representation was ultra vires but could have determined it to be such by making reasonable inquiries, does not exercise competence, because they cannot reasonably be said to have exercised appropriate knowledge and skills. Nevertheless, trustworthiness also supports a court’s decision not to enforce an ultra vires promise or representation where so enforcing would not be in the public interest as this speaks to the public official’s goodwill and competence in seeking to fulfil the promise or representation.
To be fair, given the above conceptualisation of trust, encompassing elements of goodwill and competence, this reimagined trust conception arguably aligns with the pluralist approaches advocated by scholars like Bell and Daly. However, a full consideration of this alignment is well beyond this article’s scope. That said, I think that such pluralist approaches, to the extent that they overlap with trust as conceptualised, reflect support for – rather than a critique of – the trust conception. Put simply, depending on the plurality of values said to be protected by the legitimate expectations doctrine, those values may come under the overarching umbrella of the trust concept.
V. Conclusion
While the trust conception of legitimate expectations has become a leading conception to explain the theoretical underpinning for the doctrine, as critics have stressed, it has several issues. I have argued that its fundamental weakness is its lack of conceptual clarity, with trust, the conception’s central concept, under-conceptualised. This has yielded ambiguity in the conception and prevented trust from being usefully deployed in the law of legitimate expectations. To remedy this, I have drawn on scholarship on trust from disciplines outside law to clarify what it means for the citizen to trust a public official and to elucidate the relationship between legitimate expectations and trust.
Under this article’s reimagined conception, trust can explain the legitimate expectations doctrine. I have reimagined the conception, in particular, as focusing on the trustworthiness of the official, rather than the citizen’s trust, with legitimate expectations creating the conditions under which the citizen’s expectations of goodwill and competence are likely to be fulfilled. And so reimagined, the conception explains how courts have used the doctrine, including their attempts to balance the citizen’s interests in having the promise or representation made to them fulfilled with the wider public interest. In Coughlan, for instance, trustworthiness explains the court’s conclusion that the health authority’s decision to move Coughlan from Mardon House was unlawful. In not fulfilling its predecessor’s promise that Coughlan could remain in Mardon House “for as long as [she] chose”, the health authority failed to exercise goodwill towards her, since the authority did not have sufficient justification for not fulfilling the promise. As Lord Woolf noted in his judgment, there was “no overriding public interest which justified” the authority’s decision.Footnote 162 Thus, in not fulfilling the promise made to Coughlan, the health authority did not act trustworthily.
Given its explanatory power, the reimagined trust conception presented in this article has the potential to offer some much-need certainty to this confused area of administrative law. Rather than rely on the overly abstract and over-inclusive principles of fairness, abuse of power and good administration, the courts can rely on the more concrete concept of trustworthiness to determine what the legitimate expectations doctrine protects. Granted, further work needs to be done to carve out precisely what government trustworthiness requires; and this work can be done on a case-by-case basis by the courts. However, by offering a more detailed conceptualisation of trust and in turn reimagining the trust conception, this article lays the required foundation for that work.
Open access
