The Supreme Court’s judgment in Finucane — II: Three unanswered questions concerning the doctrine of legitimate expectation

In my first post on the Supreme Court’s recent judgment in In the matter of an application by Geraldine Finucane for Judicial Review [2019] UKSC 7, I considered the Court’s approach to the relevance (or otherwise) of reliance in legitimate expectation cases, and looked at the way in which the case highlights the less than straightforward nature of the distinction between the procedural and substantive (aspects of the) legitimate expectation doctrine(s). I concluded that post by saying that Finucane gives rise to, or at least brings to mind, three sets of unanswered questions concerning legitimate expectation doctrine. In this second (and final) post on Finucane, I identify (without trying to provide detailed answers to) those sets of questions, and argue that their unanswered nature can be accounted for by an unfortunate degree of judicial ambivalence concerning the place of legitimate expectation within the broader doctrinal landscape of administrative law.

A distinction between procedural and substantive legitimate expectations?

The first question concerns the usefulness, or otherwise, of the distinction between procedural and substantive legitimate expectations. That distinction was adhered to in Coughlan, and is reflected in the three categories that were set out in that case, but its appropriateness was later questioned by Laws LJ in Nadarajah v Secretary of State for the Home Department [2005] EWCA Civ 1363:

The dichotomy between procedure and substance has nothing to say about the reach of the duty of good administration. Of course there will be cases where the public body in question justifiably concludes that its statutory duty (it will be statutory in nearly every case) requires it to override an expectation of substantive benefit which it has itself generated. So also there will be cases where a procedural benefit may justifiably be overridden. The difference between the two is not a difference of principle.

Whether we should see procedural and substantive legitimate expectations in the way that Laws LJ does — i.e. as closely related aspects of a single doctrine — is debateable. One way of looking at this would be to see procedural legitimate expectations as, in essence, an extension or offshoot of the common law duty to act fairly.  If this is how procedural legitimate expectations are understood, then the place of substantive legitimate expectations needs to be considered separately, given that it cannot itself cogently be characterised as nothing more than a further extension of the common law duty to act fairly, or of the normative considerations that animate it. On this view, Laws LJ’s attempt to dismantle the distinction between procedural and substantive legitimate expectation looks decidedly dubious. (The position may be different if we were prepared to view the concept of fairness that underpins the duty to act fairly in terms so expansive as to encompass both procedural and substantive conceptions of fairness; but such an analytical step would rob the concept of fairness itself of intellectual coherence. I return to this point below.)  

An alternative approach would be — as Laws LJ suggests — to view procedural and substantive legitimate expectations merely as differential manifestations of a doctrine that rests on a unified normative foundation. One candidate for that unified normative foundation is — as Laws LJ suggests in Nadarajah, and as Lord Kerr agrees in Finucane — good administration. But this notion is very vague. What, in this context, does ‘good administration’ really mean? Indeed, we might go as far as to question whether good administration can properly be regarded as a norm in itself, as distinct from a shorthand term that describes modes of administrative conduct that align with underlying normative values. In this regard, a stronger candidate for the normative value that forms a unified foundation for legitimate expectation doctrine might be the rule of law principle of legal certainty.

My aim here is not to argue decisively for a particular answer to the question of whether a distinction can meaningfully — and ought to — be drawn between procedural and substantive legitimate expectation doctrine. Rather, my aim is simply to highlight the fact that Lord Kerr’s judgment in Finucane appears unsympathetic to any clear distinction between the two, and in that way seems to incline towards the view put forward more explicitly by Laws LJ in Nadarajah. The crucial point, however, is that which of these views we prefer turns upon some fundamental questions about the normative basis or bases upon which the doctrine(s) of legitimate expectation is or are founded. This is not to suggest that a single normative foundation for legitimate expectation cannot be articulated, thereby paving the way for the sort of unified approach championed by Laws LJ in Nadarajah and implicitly supported by Lord Kerr in Finucane. But if that is the direction in which the law is to be developed, it will need to be underpinned by a more satisfactory normative foundation than a judicial intuition that it all comes down to ‘good administration’.  

What does fairness mean in this context?

Finucane reminds us that procedural legitimate expectations are not absolute but are instead defeasible — hence very clear undertakings as to the holding of a public inquiry did not ultimately have to be delivered, because other considerations justified a different course of action. The possibility of procedural expectations being trumped by countervailing considerations was countenanced by Laws LJ in Nadarajah in the passage set out above, as well as by the Court of Appeal in Coughlan itself. However, while the question of defeasbility — and, in particular, the applicable standard of review — has received a good deal of attention in respect of substantive legitimate expectations, that issue has been largely neglected in relation to procedural expectations. That is likely because cases such as Finucane, in which the issue of defeasibility is front and centre, are rare. But the case does underline the need for clarity as regards the applicable legal test. And as things currently stand, such clarity is lacking. 

In recent years, a number of cases have taken the view that questions about the lawfulness of decisions that frustrate legitimate expectations fall to be answered by reference to the proportionality test. Whether or not one thinks that this is (always) the right approach, it at least has the merit of grounding the judicial analysis in a now-familiar doctrine that can readily be understood. However, the Supreme Court in Finucane — perhaps led astray by the fact that this the expectation at stake was ‘procedural’ — took a different approach. Like the Court of Appeal in Coughlan,it invoked the language of ‘fairness’. Thus, according to Lord Kerr, ‘where a clear and unambiguous undertaking has been made, the authority giving the undertaking will not be allowed to depart from it unless it is shown that it is fair to do so’. But this gets us only so far, and there is clearly work to be done as regards determining what ‘fairness’ means in this context. For one thing, Lord Kerr’s formulation appears to invite a paradoxical analysis, by requiring us to consider whether it would be fair to frustrate a legitimate expectation of procedurally fair treatment. The paradox can be resolved, of course, if it turns out that we are applying two wholly different notions of fairness: one substantive, the other procedural. Now, the question makes more sense: is it (substantively) fair to deny someone a (procedurally) fair form of treatment that they were led to expect? But this simply kicks the can down the road, because it tells us nothing about what it actually means for something to be substantively fair. Indeed, in Gallaher, Lord Carnwath’s leading judgment appears to pour cold water over the very idea of substantive fairness: while ‘procedural fairness’ is ‘well-established and well-understood’, ‘[s]ubstantive unfairness … is not a distinct legal criterion’.

Given these very recent remarks in Gallaher, it is surprising that Finucane appears to rehabilitate the notion of substantive fairness by using it as the touchstone for determining whether the frustration of a legitimate expectation is lawful. It is particularly surprising that Lord Carnwath in Finucane endorsed Lord Kerr’s judgment in that case, given Lord Carnwath’s remarks about substantive fairness in Gallaher. And it is disappointing that, having so recently expressed scepticism about the meaning and usefulness of the notion of substantive fairness, the Supreme Court in Finucane so blithely invokes that concept. In doing so, it creates uncertainty about the nature of the test being applied, about the operative standard of review, and about how a test of substantive fairness thus conceived relates to and/or differs from other (better understood) potential tests such as rationality and proportionality.

Like the questions considered in the previous section, concerning the relationship between procedural and substantive expectations, the questions that Finucane raises about the role and meaning of fairness go largely unanswered. Such failures highlight a common underlying problem — namely, an unfortunate judicial tendency to seek to avoid facing up to difficult doctrinal and normative questions by sheltering behind superficially attractive but ultimately vacuous notions such as ‘good administration’ and ‘fairness’. Such language may have a motherhood-and-apple-pie form of appeal, but it is incapable of doing the sort of analytical heavy lifting that would be required if the law in this area were to be placed on a more intellectually coherent basis.

How do procedural legitimate expectations relate to the common law principles of procedural fairness?

The judgment in Finucane places in sharp relief a fundamental question about the relationship between procedural legitimate expectations and procedural fairness. In some cases, forms of procedural fairness that are legitimately expected would be legally required anyway by application of the normal principles of procedural fairness (or natural justice). In such circumstances, framing the case in terms of legitimate expectation is unnecessary but may be convenient (e.g. because it circumvents the need for the contextual analysis of what fairness at common law requires). In other cases, however, framing the case in terms of procedural legitimate expectation is imperative because the claimant is seeking access to a form of procedure to which she would not, at common law, otherwise be entitled. Finucane itself is an example of such a case, there being no basis for arguing (independently of a legitimate expectation, and leaving aside the separate question arising under Article 2 ECHR) that the common law principles of procedural fairness require the holding of a public inquiry into a death.

The foregoing is straightforward enough. However, by underlining the fact that a procedural legitimate expectation is defeasible in the face of countervailing public policy considerations, Finucane raises a question about the normative reason for such defeasibility. That question arises because a different approach applies to forms of procedural fairness to which a claimant is entitled at common law, independently of any legitimate expectation. At common law, once it is determined that fairness demands a given form of procedure, we do not normally regard the procedural requirement as a defeasible one that is subject to being traded off against competing policy considerations. Rather, we generally see it as an entitlement that the decision-maker must respect. Why should there be a different approach in respect of procedural legitimate expectations? The answer, presumably, is that they have (or at least are now understood to have) grown from a normative root that is different from that which supports the common law principles of procedural fairness. Whereas the latter are anchored (it appears) in a combination of dignitarian and instrumental considerations, legitimate expectations – at least according to much of the case law, now including Finucane — find their normative basis in the notion of good administration. And the distinction that Finucane implies — according to which procedural legitimate expectations are defeasible in a way that common law requirements of procedural fairness generally are not — suggests that the normative considerations that animate common law procedural fairness are understood to have a degree of fundamentality that the good-administration considerations underpinning procedural legitimate expectations do not.

However, this, in itself, gives rise to important questions with which Finucane fails to grapple. One has already been canvassed: namely, whether the notion of good administration is, and even intellectually capable of serving as, the normative foundation of legitimate expectation doctrine. But further questions arise even if we are prepared to accept the courts’ view that good administration is a serviceable normative basis. For instance, are the normative values that the doctrine of procedural legitimate expectation seeks to protect really less important than those that are safeguarded by the common law principles of procedural fairness, thus accounting for the apparently readier defeasibility of procedural legitimate expectations? And is it in fact the case that procedural legitimate expectations are defeasible to an extent that the common law principles of procedural fairness are not, or is it merely the case that — at least according to Finucane — procedural legitimate expectations’ defeasibility is openly acknowledged, whereas the common law principles’ defeasibility is not? In other words, is the explicit defeasibility countenanced by Finucane in fact matched by less obvious forms of defeasibility in respect of common law procedural fairness — such as the willingness of courts, when determining what in the first place procedural fairness requires, to take account of such matters as the efficiency and resource implications that may arise if more-generous forms of procedural protection, such as oral hearings, are afforded? And if, in fact, defeasibility operates in relation to both procedural legitimate expectations and common law principles of procedural fairness, is it acceptable that they appear to operate in such different ways — in one instance explicitly, and in the other beneath the radar, through manipulation of what it is that procedural fairness is taken to require in the first place?  

Concluding remarks

The foregoing, as will be clear, is not an attempt to answer the various questions to which Finucane gives rise. Rather, it is an attempt to highlight the range and significance of those questions, all of which relate back to a fundamental question with which this post began concerning the place of legitimate expectation doctrine within the landscape of English administrative law. The courts’ failure to develop a coherent answer to that question inevitably affects the coherence of the law in this area more generally, with consequences that are apparent from judgments like Finucane (as well as others, most notably Gallaher). Courts are never likely to provide analyses of the law that tie up every loose end and supply the sort of pristine taxonomy that (some) academic commentators might wish to see.

But the difficulties that continue to beset this area of administrative law amount to more than a failure to tie up loose ends. Rather, they are attributable to continuing uncertainty about the normative foundations of the doctrine of legitimate expectation and about how it relates to cognate areas of administrative law. Until those normative and taxonomical issues are resolved, the doctrine of legitimate expectation is likely to remain in a state of uncertainty that prejudices its coherence. An odd feature of this area of administrative law is that the initial stages of its development took place only at High Court and Court of Appeal level. Following the Court of Appeal’s controversial judgment in Coughlan, it seemed that what the doctrine of legitimate expectation needed was scrutiny at apex court level that would be capable of bringing greater order to the field. But twenty years on, and a handful of Supreme Court judgments later, it seems that one should be careful what one wishes for.

I am grateful to Alison Young for her comments on an earlier draft of this post. This is the second of two posts on the Finucane case. The first post can be found here.