Three Issues in the Law of Contractual Discretion

Abstract

The new law of contractual discretion is developing apace. This article addresses three major issues in this dynamic field. First, the article propounds an analytical framework for understanding the nature and practice of reasonableness review in the contractual setting, based on doctrinal exegesis of the full run of cases on contractual discretion. Significantly, the analysis demonstrates that review of contractual discretion is characterised by a ‘variable intensity’ approach: the intensity with which courts scrutinise exercises of discretion is dependent on a series of contextual factors. Second, the article analyses the genus of the implied term, which imposes legal constraints on contractual decision-makers, arguing that the term is properly conceptualised as a term implied in law. Third, the article addresses the remedial consequences of non-compliance with implied fetters, identifying three different remedial models in the case law. The article challenges the common assertion that damages are the invariable remedy, arguing that an impugned exercise of discretion may be void or voidable.

1. Introduction

Contracts often confer discretionary powers on one party which, when exercised, affect not only their own interests but also those of the other party. Such powers arise across myriad contractual contexts. For example, loan agreements may confer discretion on the lender to amend interest rates, and employment contracts may empower the employer to grant a bonus and decide quantum. This article takes ‘discretion’ to denote ‘decisional choice’. In contrast to application of legal rules, the decision-maker is not working towards a notional right answer; rather, they may choose amongst outcomes.

It has been recognised for some time that contractual discretions are not unfettered. Courts generally imply a term to limit such powers. However, since the Supreme Court’s landmark 2015 decision in Braganza, the case law on the implied term has gained new momentum, resulting in a growing and fast-evolving body of jurisprudence.¹ One of the most significant features of that decision was importation of the Wednesbury ground, developed in administrative law, into the contractual setting.

This article addresses three major issues in this developing field: the nature of reasonableness review under the implied term; the genus of the term; and remedies. These issues have in common that they are critical to full understanding of the new law of contractual discretion, but have not been the subject of detailed examination.

The article first proffers an account of the nature and practice of Wednesbury as applied to contractual discretion. Doctrinal exegesis reveals that reasonableness—a seemingly unitary standard—is applied by courts with ‘variable intensity’. The article identifies and rationalises contextual factors which affect whether a court will apply greater or lesser scrutiny to a given exercise of contractual discretion. The article also demonstrates that, contrary to common claims that Wednesbury is an amorphous standard, the case law is characterised by specific subprinciples which ground intervention on the Wednesbury basis. Together, these features form an analytical framework for understanding the Wednesbury doctrine in the contractual setting.

Second, we currently lack a clear understanding of the term’s basic nature: is it a term implied in law or fact? The matter has not been seriously analysed in case law, while most commentators assume the term is implied in fact. This article challenges that view: as the term has evolved, all of its significant features indicate it is a legal implication.

Third, it is often assumed that non-compliance with the term is to be treated as an ‘ordinary’ breach of contract, necessarily sounding in damages. This article demonstrates that damages are not the invariable remedy: several remedial models are evident in the cases. Specifically, an impugned decision may be void or voidable. In such cases, damages may not be available.

Section 2 introduces the implied term. The article goes on to examine the Wednesbury limit (section 3), the genus of the implied term (section 4) and remedies (section 5).

2. Introducing the Implied Term

Over the last 40 years there has been a steady rise in cases in which courts have implied terms to fetter contractual discretions.² For much of this time, the case law developed in drips and drabs, was somewhat fragmented, characterised by a generally restrained judicial approach, and the term was formulated slightly differently from case to case, albeit with little consequence; there was limited consideration of the term’s normative foundations, and interrelationship with other fields.³ However, Braganza marked a turning point in the case law, and its maturation. The Supreme Court analysed prior cases as forming a distinct jurisprudence, authoritatively formulated the term, and addressed its rationale and connection with administrative law. The case law has since gained new momentum.

Turning first to the rationale, prior cases stated that the concerns motivating the implication of constraints were that discretionary power should not be abused⁴ and that one party should not be subject to the other’s whim.⁵ In Braganza, Lady Hale endorsed these rationales, articulating two reasons why a risk of abuse arises.⁶ The first may be described as ‘legal’ inequality. Because one party has legal power to unilaterally affect the other’s interests, the latter is vulnerable to abuse of that power. The risk is real because the decision-maker typically has a ‘conflict of interest’; for example, a supplier with power to vary the contract price has incentives to set an exorbitant price, and the other party shall be bound. Second, in addition to the basic inequality that goes with all legal powers to bind another, certain contractual relationships are characterised by a more general ‘factual’ inequality, such as economic or social disparity, which may heighten the risk of abuse. In Braganza, Lady Hale invoked the example of employment. It is increasingly recognised that employment relationships are lopsided, with the employee beholden to the employer for their livelihood.⁷

In some jurisdictions, such as Canada, legal constraints on discretion are conceptualised as instantiations of a general doctrine of good faith.⁸ But English law has not recognised, nor grounded the Braganza term in, such a general doctrine.⁹ Rather, Braganza is a tailored solution to a specific problem: abuse of discretion. This reflects the English courts’ general preference for incrementalism—developing ‘piecemeal solutions in response to demonstrated problems of unfairness’—ahead of general organising principles.¹⁰ This difference in rationale likely explains differences between the English and Canadian case law. Specifically, that English courts are relatively¹¹ more open to drawing on administrative law in fashioning the implied term likely follows from the rationale of protection from abuse of power having clear synergies with the basic mission of judicial review.

This rationale manifests in the legal requirements stated authoritatively in Braganza, as comprising the implied term: powers must be exercised honestly, genuinely, in good faith, and consistently with contract terms, and not arbitrarily, capriciously or unreasonably, nor for improper purpose. The status of other grounds, such as procedural fairness, remains to be authoritatively decided.¹²

In Braganza, the court acknowledged the growing alignment of these requirements with administrative law principles. Prior decisions had observed similarities between principles applied in each setting, but counselled that any analogy be treated with caution and that principles should not be equated.¹³ Other decisions were more hostile to public law analogies.¹⁴ However, as the law developed through the 1990s and 2000s, it became increasingly difficult to see daylight between certain principles. Most prominently, an ‘unreasonable’ decision, for the purposes of the implied term, was typically defined as one so unreasonable (or ‘extreme’ or ‘perverse’) that no reasonable decision-maker could reach it.¹⁵ Unsurprisingly, experienced public law judges began to more directly analogise with the Wednesbury standard.¹⁶ Reflecting these synergies, even in cases where courts disavowed public law analogies, they ended up drawing on, and even incorporating, administrative law jurisprudence,¹⁷ or adopting approaches indistinguishable from public law.¹⁸ These synergies were ultimately recognised in Braganza, where it was observed that lower courts had been understandably reluctant to deploy the fully-developed rigour of judicial review, but had struggled to articulate what the difference was—or, one may add, the reasons for resistance.¹⁹ Importantly for future legal development, Lady Hale observed that there are obvious parallels between review of contractual and statutory discretions, and signs the duties cast by the implied term are drawing ever closer to judicial review principles.²⁰

But the court went further, directly reading administrative law principles into the implied term. Prior to Braganza, there had been stuttering recognition of unreasonableness ‘analogous to Wednesbury’.²¹Braganza dropped the analogy: Lady Hale simply stated Wednesbury should now be applied. She recorded that unreasonableness, as applied in contract, had not been ‘a precise rendition’ of Lord Greene MR’s test.²² But following the ‘transplant’ effected by Braganza, both limbs of Wednesbury now apply.²³ A decision-outcome must not be ‘so unreasonable’. But in addition, a decision-maker, in the process of deciding, must exclude irrelevant matters and consider relevant matters.²⁴ Thus, Wednesbury has been ‘imported’:²⁵ ‘exercise of contractual discretion is to be judged by the same principles as the exercise of public law discretions’.²⁶

Crucial to understanding reasonableness review is the courts’ conception of the judicial role in applying the implied term. The court performs a ‘reviewing function’.²⁷ Its focus is generally upon the integrity of the process by which the decision is made, rather than the decision’s substance²⁸—language which echoes judicial review. As such, the court is generally involved in examining the defendant’s conduct in exercising the power or mental processes leading to the decision.²⁹ The court acts in a supervisory capacity.³⁰ It must not substitute itself for the decision-maker, as it was the parties’ intention to bestow the decision-power on them.³¹ The court’s only function ‘is to decide on the legal limits to the [decision-maker’s] discretion and whether the [decision-maker] has acted within or outwith the limits’.³²

Importantly, in applying the term, the courts seek to balance the repository’s decision-making autonomy with legal control;³³ the decision-maker must have room to exercise choice. To this end, courts invoke a distinction between Wednesbury and objective reasonableness.³⁴ They illustrate the distinction, as public lawyers often have, through comparison with tort. Courts distinguish: (i) the supervisory conception of reasonableness, where the repository as primary decision-maker may choose from a range of options, with courts policing the outer limits of that range; from (ii) objective reasonableness applied to assess breach in negligence, where the court, as primary decision-maker, forms its own determinative view of what would have been a proper precaution.³⁵

Conceptually, the distinction is clear-cut. But it may be blurred in application, leading some judges to prefer the idea of a ‘spectrum’ to a binary distinction between objective reasonableness and Wednesbury.³⁶ On the one hand, objective reasonableness may be applied in a manner deferential to another’s assessment, or courts may be cautious to reach a conclusion of objective unreasonableness where they lack expertise.³⁷ Further, reasonableness, as an objective qualifier, may incorporate examination of decision-processes, including whether relevant matters were considered.³⁸ On the other hand, as we shall see, courts have more recently applied Wednesbury stringently, reflective of an emergent variable intensity approach, emboldened by the Supreme Court’s authoritative endorsement of reasonableness. Post-Braganza, courts emphasise the ‘potency’ of constraints on discretion.³⁹ Further, on certain facts, only one outcome may be reasonably open,⁴⁰ or it may make no difference whether objective reasonableness or Wednesbury is applied.⁴¹

3. Wednesbury

Thus, the Supreme Court has authoritatively confirmed Wednesbury as a core principle of the implied term. This public law transplant has generated much interest. Yet, we have so far lacked a full understanding of the nature and practice of reasonableness review in the contractual setting. This section addresses this gap by offering an account of the Wednebsury doctrine, as applied to contractual discretion. The discussion references administrative law at points, not least given the increasing influence of public law thinking in this field, but the principal concern is to understand the contract jurisprudence on its own terms.

The section demonstrates that an intelligible order is immanent in the law. Reasonableness is applied with ‘variable intensity’. The intensity with which courts scrutinise exercises of discretion depends on several contextual factors, which are identified and analysed herein. The section also distils from the jurisprudence a series of reasonableness subprinciples, which ground judicial intervention on the Wednesbury basis. Together, these doctrinal features form a framework of principle which governs application of Wednesbury to contractual discretion.

A. Variable Intensity

It is accepted today that the reasonableness standard in administrative law is not unitary, but applied variably.⁴² Given the increasing influence of public law concepts and thinking in contract, but also the generality and malleability of reasonableness as a standard, it would be surprising not to find variable application in contract too.⁴³ Courts in other jurisdictions have explicitly contemplated a variable approach. In Australia, Edelman J, in Minerology, having observed developments towards a variable approach in administrative law, doubted whether the implied term provided for an invariable standard.⁴⁴ He thought reasonableness may not necessarily be tethered to adjectives such as ‘outrageousness’ or ‘so unreasonable’.⁴⁵ Rather, intensity of scrutiny may be context-dependent. Building on these dicta, Australian courts observe it is one thing to recognise a reasonableness control; the more difficult issue is determining the applicable standard of reasonableness, which depends on several factors.⁴⁶

English courts have less directly addressed the variability of reasonableness, albeit express judicial statements recognising variable intensity are increasingly common, not least in Braganza. Yet, it is possible, by analysing the full run of discretion cases, to identify an emerging framework of variable review. Specifically, one can isolate contextual factors which colour the intensity of judicial scrutiny applied on given facts. These factors are thus crucial to explaining the case law and understanding how judges apply reasonableness. There are four key factors: contractual context; nature and scope of the discretion; decision-maker’s capacities; and interests of the party subject to the discretion.

(i) Contractual context

Perhaps the most important factor conditioning the intensity of reasonableness review is contractual context, which comprises the terms and purpose of the contract and, most fundamentally, the type of contract.

Building on those totemic decisions recognising employment contracts as a discrete genus,⁴⁷ Lord Hodge in Braganza recalled: they are ‘relational’, involve a personal relationship and are characterised by an implied obligation of trust and confidence.⁴⁸ In turn, the nature of employment ‘may justify a more intense scrutiny of the employer’s decision-making process’ than applied in commercial contexts.⁴⁹ On the facts, the special nature of employment reinforced that a high threshold, in terms of proof, was required to be met before BP (the employer) could conclude, pursuant to a contractual fact-finding power, that Mr Braganza (the employee) had committed suicide while on duty, given the stigma and broader social and economic consequences such finding would carry for his family.⁵⁰ Lady Hale similarly observed the importance of contractual context, agreeing that employment and commercial contexts are distinct; the overarching obligation of trust and confidence shapes the employer’s approach to their decision-making task.⁵¹

That different contractual relationships are characterised by discrete logics, which colour reasonableness review, is a recurring theme of discretion cases. As illustrated by Braganza, a cluster of concerns reinforce stricter scrutiny in employment contexts. In the employment setting, the ‘legal’ inequality associated with all discretions arises within a relationship characterised by a more general ‘factual’ power disparity, which exacerbates the risk of abuse. Put another way, greater scrutiny is warranted the greater the extent to which the implied term’s rationale is engaged. Further, the employment relationship is defined by overarching duties of considerateness,⁵² meaning comparatively more may be expected of decision-makers, including that power be exercised in a manner sensitive to employee interests, such considerate decision-making helping to sustain the valuable employment relationship.

As recognised in Braganza, other types of contracts implicate distinct concerns, leading to a different approach. In commercial settings, involving economically powerful and well-advised parties such as Barclays or ExxonMobil, the reasonableness standard ‘is not … rigorous’, decision-makers may have ‘considerable latitude’, and the circumstances in which intervention may be warranted are ‘extremely limited’ and would have to be ‘extreme’.⁵³

Several reasons underpin a lighter touch approach in commercial settings. First, such relationships are not characterised by the sort of factual power disparity that characterises employment. Nor are they defined by overarching duties of trust and considerateness, but rather by a market logic of self-interest. Courts readily conceptualise commercial contracts as instruments of wealth-maximisation; in turn, they are reluctant to saddle parties with onerous obligations, lest they impede efficient market activity.

Second, courts shall be less inclined to intervene where commercially-savvy, well-advised parties are well-placed to negotiate terms protective of their interests.⁵⁴ In contrast, in the employment context, it is well-established that the traditional conception of contract as an agreement between free and equal parties is inapplicable.⁵⁵ Similarly, there shall be greater cause for scrutiny of discretion exercised under boilerplate contracts, which are not the product of meaningful negotiation.⁵⁶

Third, the restrained approach to review in commercial settings reflects the well-worn concern for certainty in market dealing. Courts will not readily engage in detailed scrutiny of the decision-process as this would encourage litigation, which could ‘cut across the desire for speed and commercial certainty’.⁵⁷ Rather, by maintaining a high threshold for intervention, courts provide a high degree of certainty, as successful challenges will be uncommon. For opponents of the implied term, even the modest uncertainty associated with policing plainly oppressive exercises of power is intolerable.⁵⁸ But certainty is not the only game in town. Implicit in the judicial willingness to apply Wednesbury in commercial settings is a normative calculation that whatever uncertainty is generated, it is justified by those concerns underpinning the implied term.

Fourth, market disciplines, such as market players’ concerns for reputation, fostering profitable relationships and avoiding tit-for-tat retaliation, may reduce risks of opportunism.⁵⁹ In turn, incentive effects, at least in markets not characterised by significant market failure, reinforce that courts need not scrutinise decision-making as closely in commercial settings. Of course, that courts have impugned exercises of discretion under commercial contracts, finding the high threshold of intervention met, reinforces that market dynamics are not invariably sufficient to overcome incentives for gratuitous exercises of power.

Thus, courts have adjusted their approach to reasonableness given the peculiarities of contractual relationships. These developments are consonant with trends in other contractual contexts, where courts perform a ‘supervisory’ role.⁶⁰ Courts have held that their approach to the penalties doctrine may depend on the type of contractual relationship. Where the contract is one negotiated by properly-advised parties of comparable bargaining power, the ‘strong initial presumption’ is that parties are the best judges of their interests, and courts will be slow to impugn a clause as penal.⁶¹ But, where one party can dominate the other in choice of contract terms, and there is consequently a greater risk of oppression, courts may more closely scrutinise a challenged clause.⁶² Thus, the type of relationship does not affect whether the penalties doctrine applies, but colours ‘application of the rule’.⁶³

(ii) Nature and scope of power

The nature and scope of contractual powers has proven significant. In Braganza, Lord Hodge said, ‘[t]he scope for scrutiny differs according to the nature of the decision’.⁶⁴

First, if the power is one to exclude a prima facie entitlement, the burden is reversed: the defendant must prove their decision was not unreasonable.⁶⁵ It also seems exercise of such power will be subject to closer scrutiny,⁶⁶ especially where vested rights are involved. If a right could be defeated with little or no justification, it would be empty—‘a sort of mirage’.⁶⁷

Second, if the power calls for broad value-judgements, ‘there is little scope for intensive scrutiny’.⁶⁸ Thus, political parties have ‘very wide’ contractual discretion to select candidates, exercise of which involves weighing ‘a wide range of considerations’ and ‘subtle judgments’; as such, only ‘something exceptional’ could render a decision irrational.⁶⁹ Review being ‘context-specific’, greater scrutiny could be expected where political parties exercise disciplinary powers or purport to exclude accrued rights.⁷⁰

Caution in reviewing qualitative judgements is also evident in the reticent judicial approach to scrutinising broad employer powers to grant discretionary bonuses; the threshold for intervention is ‘very high’, requiring an ‘overwhelming case’.⁷¹

Several considerations support restraint. The refusal to grant a discretionary bonus does not involve denial of an accrued entitlement. Such benefits are putative.⁷²

A broadly-framed power evinces an intent that the repository have maximal decisional autonomy to make a judgement call. If courts too readily intervene, they risk frustrating that intent.⁷³ With broadly-framed powers, there may be no objective standards or textual cues against which to judge reasonableness,⁷⁴ so there is a risk judicial intervention will reflect a mere difference of opinion. As Lord Reed has observed of judicial review, much depends on ‘the extent to which the powers … have limits or purposes which the courts can identify and adjudicate upon’.⁷⁵ For a broadly-framed power, identification of such limits may prove elusive, so the range of permissible decisions is correspondingly widened.⁷⁶ In the same vein, contract cases emphasise reasonableness must be judged in light of the discretion’s terms and purpose.⁷⁷

Further reinforcing restraint, courts may lack the institutional competence to assess an employer’s evaluation of employee performance in a specialist industry,⁷⁸ where market conditions are fluctuating,⁷⁹ while there is typically a fixed pool earmarked for bonuses, so impugning one allocation could distort the overall allocation. In Braganza, Lord Neuberger stated as a general principle that greater respect should be afforded to decision-makers with relevant expertise or experience,⁸⁰ while the concept of ‘due deference’ is recognised in the contract cases.⁸¹ Equally, the greater the court’s familiarity with the given inquiry, the more confident it may be in identifying error.⁸²

Third, an important feature of Braganza and its progeny is a greater judicial willingness to intervene where a power solely involves fact-finding. Lord Hodge considered courts are much better placed to review a decision over whether a state of affairs existed, compared to a broadly-framed discretion.⁸³ Indeed, the Braganza majority came close to forming their own opinion of whether there was sufficient proof to conclude Mr Braganza had committed suicide.⁸⁴ Unsurprisingly, experienced administrative lawyers have queried this approach,⁸⁵ given courts on judicial review are generally reluctant to question a decision-maker’s fact-finding.

But there are reasons that explain the Braganza majority’s less restrained approach. In judicial review proceedings, cross-examination, expert evidence and disclosure are uncommon. Whereas they are commonplace in ordinary proceedings. Therefore, courts are well-placed to probe the factual bases of contractual decisions in a way that is not possible in review proceedings—while, in terms of competence, judges are expert fact-finders. Importantly, a power to determine a state of affairs is not a true discretion: the power does not denote choice. Rather, there is a truth of the matter. As such, there is not the same rationale for affording decision-makers leeway on the basis that they enjoy scope to choose among options.⁸⁶

A further reason in administrative law for caution in reviewing fact-finding is that it may inappropriately draw judges into the merits. This is an understandable concern when fact-finding is one aspect of exercise of a discretion. Taking an example from contract, in deciding whether to award a bonus, the employer will conduct fact-finding, but that exercise is shaped by the evaluative/normative criteria the employer considers relevant to granting a bonus, and the employer’s decision flows from applying those decisional criteria to the facts. This renders it difficult to disentangle the normative/evaluative from the factual aspects. But where the power is straightforwardly one to determine the existence of a state of affairs, there is no evaluative aspect, beyond establishing the fact of the matter: there are no merits to be pre-empted, as such. Thus, in Braganza, BP solely had power to establish whether Mr Braganza’s death was due to suicide; if so, the consequence, that his widow would be deprived of a death benefit, followed by operation of law.

In contrast to the majority, Lord Neuberger suggested a more cautious approach to reviewing fact-finding. He stressed the court was not performing an originating fact-finding function.⁸⁷ Rather, he saw the court’s role as analogous to that of an appellate court reviewing a lower court’s factual findings, affording deference given the lower court’s formal findings will be necessarily incomplete and will not convey the judge’s overall impression. With respect, the analogy is questionable. First, lay contractual decision-makers do not typically issue formal findings or have a duty to provide reasons,⁸⁸ and if findings are incomplete they can be supplemented by the defendant’s evidence. Second, a judge reviewing a contractual fact-finder’s decision is not in the same position as an appellate court reviewing a lower court’s findings. A judge reviewing a decision-maker conducts a full trial and hears witnesses, whereas appellate courts do not. Indeed, a judge may be better placed to find facts than the lay decision-maker, given procedural powers, assistance of counsel, their experience, training and neutrality. Thus, consonant with the majority approach in Braganza, there are good reasons for courts not to approach fact-finding powers in the same way as true discretions.

Lord Neuberger was more generally sceptical of the extent to which contextual variables should affect the intensity of review. For example, he considered a sound decision-process is a sound decision-process whether in a commercial or an employment context.⁸⁹ His view seems tied to the power at issue in Braganza: good fact-finding is good fact-finding. However, even in relation to such a decision-power, Lord Neuberger identified institutional expertise as a relevant factor;⁹⁰ that factor varies with the decision-maker. Further, the majority judgments suggest, contra Lord Neuberger, that contextual variables are relevant to reviewing fact-finding. Lord Hodge considered the overarching duty of trust and confidence meant employers should adopt a more rigorous approach to fact-finding where the outcome would have serious implications for the employee’s family—a point echoed by Lady Hale.⁹¹ Whereas in commercial contexts courts are more accepting of decisions based on ‘imperfect’ information or ‘mistakes’,⁹² recognising there may be good reasons for affording leeway, such as where decisions must be made quickly to enable transactions in a fast-moving market.⁹³

(iii) Capacities

A further variable affecting depth of judicial scrutiny is the decision-maker’s capacities: what one might realistically expect, in terms of the thoroughness of decision-making, may depend on the defendant’s resources. Underpinning this factor is arguably a concern to maintain fair balance between parties—specifically, not to overburden decision-makers by imposing standards that are difficult or unduly costly to abide.⁹⁴

Consonant with prior jurisprudence,⁹⁵ all Justices in Braganza considered it would be wrong to hold a lay decision-maker to the standards of factual decision-making expected of a trial judge.⁹⁶ Nonetheless, the majority were unwilling to equate the capacities of the defendant, BP, with those of a natural person. Both Lady Hale and Lord Hodge, in elaborating an intensive approach to review, emphasised that BP is a large company with ready access to legal advice, and ‘should be able to face scrutiny’.⁹⁷ Other courts have refused to water down standards where the decision-maker is ‘a large sophisticated organisation’,⁹⁸ whereas less may be expected of small voluntary organisations.⁹⁹ But where a breach is found, a lack of training or resources shall be no defence.¹⁰⁰

(iv) Interests

Courts have, in calibrating reasonableness review, considered how an exercise of discretion impacts the interests of the party subject to the decision. Early on, courts abstained from considering such impacts, even if draconian.¹⁰¹ The law has moved on. In Braganza, Lord Hodge observed that one factor reinforcing a strict approach was the ramifications a finding of suicide would carry for Braganza’s family, including deprivation of the death benefit; Lady Hale similarly emphasised these ‘serious consequences’.¹⁰² Lord Hodge linked this factor’s salience to the special character of employment, reinforcing that the significance of the other party’s interests may depend on other contextual factors, particularly type of contract.

Thus, in contexts characterised by a paradigm of considerateness, such as employment and tenancy,¹⁰³ courts readily accept the decision-maker should have regard for the other party’s interests. In contrast, within relationships characterised by a logic of commercial self-interest, the factor is less prominent, and decision-makers may have primary regard to, and prioritise, their own interests, especially where the discretion’s very purpose is to enable the decision-maker to safeguard their interests.¹⁰⁴ However, even in commercial settings, there is an outer limit beyond which decision-makers cannot press their interests: ‘abusive self-interest’ is impermissible,¹⁰⁵ and one of the indicia of such is a decision that seriously burdens the other party.¹⁰⁶

In specific contexts, such as insurance and charterparty, courts have variously held the decision-maker must act ‘fairly as between both parties’ not ‘disregarding the [other party’s] interests’,¹⁰⁷ and in the ‘common interest of themselves and the other party’.¹⁰⁸ But in no context have courts mandated selflessness; the implied term does not impose fiduciary obligations.¹⁰⁹ Though, the more a decision-maker must consider the other party, the more the distinction between law and equity may be blurred.

How the relative importance of interests, and specifically the presence of basic interests, should affect reasonableness review is an issue which requires further judicial consideration. But as the law stands, it seems the presence of important interests will affect application of reasonableness, at least in certain contexts. Various important interests may be prejudiced by contractual decisions, including reputation,¹¹⁰ earned benefits,¹¹¹ one’s livelihood,¹¹² pensions,¹¹³ access to injury insurance,¹¹⁴ freedom of expression or religion,¹¹⁵ or the ability to vote in party elections.¹¹⁶ There are different ways the presence of such interests could affect review, and different approaches are evident in the cases.

On one approach, the potential for the decision to impact important interests is something a reasonable decision-maker must pay ‘due regard’¹¹⁷ to, and ask themselves whether the prejudice is justifiable. This approach is synonymous with the ‘anxious scrutiny’ variant of Wednesbury.¹¹⁸ Such approach is consonant with a supervisory conception of review focused on decision-process:¹¹⁹ the decision-maker is required to consider and weigh the relevant interests, but if they have done so, the ultimate balance struck may only be impugned on the ‘so unreasonable’ standard.¹²⁰ However, according to this approach, courts may require more to be convinced the balance struck is reasonable, the more important the affected interests. Thus, in employment settings, courts require stronger justifications for decisions denying accrued rights than those denying inchoate benefits.¹²¹ Significantly, authority tells against application of anxious scrutiny in commercial contexts, reinforcing the importance of type of contract. Thus, in banking contexts, where ‘[b]ankers, as commercial men, have a keen instinct for where their own interests lie’ and the capacity to protect those interests, and affected interests are merely those in profit-making, courts have resiled from requiring decision-makers to balance their interests against the other party’s.¹²²

A more intrusive approach, beyond anxious scrutiny, would be for the court itself to strike the appropriate balance between competing interests, according to structured proportionality. But, as courts observe, this approach is contentious, as courts would encroach far more extensively on the repository’s decisional autonomy, to the point that this approach, involving objective balancing, may be irreconcilable with a supervisory role.¹²³

Critically, an objective approach risks undermining the parties’ intention that it should be the decision-maker who strikes the balance—not the court. However, in certain contractual contexts, not characterised by a pure paradigm of freedom of contract, courts may not attach as much weight to party intention as a constraint on their role. The normative force of appeals to party intention is diminished where discretion arises within an unequal relationship where the weaker party is effectively a rule-taker. For instance, courts acknowledge employment is not a paradigm of ‘party autonomy’ in which ‘free and equal’ parties are able to ‘negotiate whatever terms they liked’.¹²⁴

Thus, we return to a core theme: context matters. Courts are most reluctant to apply an objective approach, which strongly protects affected interests, in commercial settings, observing such approach would be too ‘onerous’, and that ‘[t]his is the world of sophisticated investors not that of consumer protection’.¹²⁵

However, beyond commercial contexts, there are hints of objective balancing. In the IBM litigation concerning pensions, Warren J, at first instance, found the employer had engendered in scheme beneficiaries reasonable expectations that policies would not change.¹²⁶ He held, under the guise of Wednesbury, that these expectations could not be reneged on unless there was a significant change in financial circumstances.¹²⁷ He effectively determined for himself whether there was sufficient justification for reneging, applying a ‘necessity’ principle synonymous with proportionality: any departure from the expectation must be the least necessary to protect the decision-maker’s legitimate interests.¹²⁸

Ultimately, the Court of Appeal rejected this approach: it went beyond a supervisory approach, according to which the decision-maker, not the court, conducts balancing, and verged on the judge substituting his view for that of the decision-maker.¹²⁹ Instead, the Court of Appeal effectively endorsed anxious scrutiny. The reasonable expectation should principally be protected procedurally, as a mandatory consideration,¹³⁰ and the decision-maker must have some justification for reneging, to be scrutinised according to the ‘so unreasonable’ Wednesbury standard (rather than objectively).¹³¹ Nonetheless, where the expectation has a strong basis, such as a direct promise (which was not the case in IBM), Wednesbury may be applied more strictly,¹³² and intriguingly the court contemplated the possibility that substantive protection, akin to Warren J’s approach, might apply.¹³³ Similarly, it is in promise cases—where the normative foundation of a legitimate expectation is strongest—that courts on judicial review have been most open to applying proportionality to determine whether an expectation can be resiled from lawfully.¹³⁴

Other legal requirements, from within and without contract law, may require judicial balancing. The penalties doctrine incorporates a type of proportionality test,¹³⁵ which could potentially apply as an independent fetter on discretions, to proscribe the discretionary setting of sums disproportionate to the defendant’s legitimate interests. Conversely, the implied term, by imposing a reasonableness limit on a discretion to set a sum, may save the empowering clause from being characterised as penal.¹³⁶ Aspects of the restraint of trade doctrine also resemble proportionality. The doctrine potentially applies directly to limit discretions, such as powers to place staff on gardening leave, or could possibly inform application of Wednesbury.¹³⁷

Human rights law provides the prime example of structured proportionality being applied to protect basic interests. The Human Rights Act 1998 is unlikely to directly bind contractual decision-makers given courts have interpreted the Act’s scope restrictively where contract is in play.¹³⁸ However, positive obligations may require courts to apply proportionality as a substantive limit on contractual discretion; that is, human rights might have ‘horizontal effect’. For example, the rights to privacy¹³⁹ and freedom of religion¹⁴⁰ in the European Convention on Human Rights impose positive duties on the state, including courts, to protect rights from third-party interference. If a contractual decision-maker (the third party) exercises discretion in a way that infringes the other party’s privacy, the courts would be required by Convention requirements to ensure effective rights-protection, which includes protection from disproportionate interferences.¹⁴¹ The malleable reasonableness doctrine provides an obvious vehicle for courts to give effect to these requirements. There is also the possibility of a more general ‘spill-over’ effect: human rights values may inform common law development, including application of Wednesbury. The impetus for cross-fertilisation will be greater within contractual relationships, such as employment, that lend themselves to analysis through a human rights lens.¹⁴²

B. Reasonableness Subprinciples

Thus, variable application has emerged as a fundamental feature of reasonableness review. But there is a further dimension of Wednesbury critical to full understanding of the doctrine as applied in contract. The Wednesbury ‘so unreasonable’ formulation is often criticised as open-ended.¹⁴³ But, as in public law,¹⁴⁴ the ‘headline’ formulation is not the be all and end all of Wednesbury. More specific subprinciples are evident in the jurisprudence, which guide judicial intervention and promote predictability.¹⁴⁵ As explained below, these principles interact with the contextual factors discussed above, so one cannot have a full understanding of either variable intensity or the Wednesbury subprinciples without understanding the other.

I have identified eight subprinciples.

First, the claimant bears the burden of proving unreasonableness.¹⁴⁶ But if the claimant shows a prima facie case of unreasonableness, for example because they were treated inconsistently, the defendant comes under an evidential burden to explain the decision and how it was reached; if they fail to do so, the claimant shall succeed.¹⁴⁷

Second, a decision based on one irrational reason and one sound reason will not be impugned if the same decision would have been reached absent the irrational reason.¹⁴⁸

Third, lack of genuine intellectual engagement with the decision-making exercise will likely render a decision unreasonable.¹⁴⁹ This requires a matter that calls for exercise of discretion to be considered¹⁵⁰ and a decision taken after consideration of and on the basis of facts giving rise to the claim.¹⁵¹ Some cases hold that inquiries are required to ascertain salient facts,¹⁵² while an ill-informed decision or one based on misplaced assumptions will likely be impugned,¹⁵³ although slight misdirections are forgivable.¹⁵⁴ If the defendant cannot provide evidence of how the decision was reached, or reasons, this may favour a conclusion of unreasonableness.¹⁵⁵ Reasoning provided ex post may trigger a suspicion that there was no good reason for the decision when made.¹⁵⁶

Fourth, there is an emergent consistency principle. If a person receives a bonus far lower than those similarly placed, the defendant must justify the disparate treatment.¹⁵⁷

Fifth, a decision that appears drastic or inexplicable, such as reducing a benefit to nil, is likely unreasonable, unless good reasons are proffered.¹⁵⁸

Sixth, in the employment and pensions contexts, a decision-maker’s representations as to how discretion will be exercised may generate ‘reasonable expectations’, which must be considered when exercising the discretion, and there must be some justification for resiling.¹⁵⁹ Emergence of this principle in employment and pensions contexts reflects that such relationships are characterised by an overarching legal concern for trust and confidence, which is fostered by parties keeping their word.

Whether an expectation is ‘reasonable’, and thus legally salient, may depend on: whether the expectation was positively engendered by the defendant, and whether any representation by the defendant was clear and unequivocal, and took the form of a direct promise;¹⁶⁰ contractual provisions;¹⁶¹ surrounding circumstances, including whether it was known the claimant would rely on a given representation;¹⁶² and market practices.¹⁶³ It remains unclear whether the claimant must know of the representation.¹⁶⁴

Seventh, if the defendant adopts a policy to govern exercise of discretion but deviates from it without reason, this may be indicative of unreasonableness.¹⁶⁵

Eighth, decisions seriously prejudicing the claimant accompanied by no real explanation are likely unreasonable.¹⁶⁶ Even in contexts where courts afford decision-makers significant latitude, decisions must have ‘some basis’.¹⁶⁷ Thus, a commercial decision-maker may adopt a decision that significantly prejudices the claimant if made for genuine commercial reasons.¹⁶⁸ But even a commercial decision-maker, who may legitimately press their interests, cannot adopt extreme or seriously lopsided decisions, such as setting ‘exorbitant’ prices ‘way above’ reasonable rates of return.¹⁶⁹

The foregoing principles are bases on which courts may find a decision unreasonable, whereas the contextual factors identified in section 3A affect how strictly these principles are applied. Thus, where there is a prima facie case of unreasonableness, for example because a decision is seriously prejudicial, how closely courts scrutinise purported justifications may vary according to the type of contract, the breadth of the discretion and the importance of affected interests. Similarly, the expected degree of decision-making rigour, including whether the decision-maker must actively investigate salient facts, and any permitted margin of inaccuracy in fact-finding may depend on the type of contract, the decision-maker’s capacities and what is at stake for the claimant.

4. Conceptualising the Implied Term: Fact or Law?

Despite the growing body of case law on the Braganza term, we continue to lack an understanding of the basic nature or genus of the implied term. Specifically, is it a term implied in fact or law? And are there reasons to favour one conceptualisation over the other?

A term implied in fact or ‘ad hoc’ term is an individualised term implied into a particular contract to give effect to the inferred intentions of the parties, in light of that contract’s express terms and context. Such terms are only implied where strictly necessary for business efficacy, or ‘obvious’. In contrast, a term implied in law is a standardised term imposed by general law into all contracts of a class, unless excluded expressly. Such terms address general questions concerning the relationship between classes of contracting parties. They are not based on party intention, but are justified by broader concerns of fairness, reasonableness and public policy that transcend the interests of parties to any given contract.¹⁷⁰

While, in other jurisdictions, the Braganza-style term has been analysed as one implied in law,¹⁷¹ English courts have sent mixed signals and not squarely addressed the matter in depth. Prima facie, one might think authority favours the factual approach, drawing on hints in case law. In Braganza, Lady Hale appeared to contemplate the term could differ from one contract to the next,¹⁷² and in analysing a prior decision the term was explained as giving effect to the parties’ ‘reasonable expectations’.¹⁷³ Some lower court decisions have applied the tests for factual implication.¹⁷⁴ Commentators typically assume or assert the term is implied in fact,¹⁷⁵ though some have recognised the ambiguities inhering in the cases.¹⁷⁶

This section challenges the assumption that the term is a factual implication. Deeper analysis shows the term bears all the hallmarks of a legal implication. As such, there is a disconnect between what some judges say, or their reasoning suggests, and the term’s significant features. This is a source of confusion, which should be resolved by formally recognising the term for what its core features show it to be: a term implied in law.

It might be thought the term’s genus makes little difference. Notwithstanding formal classification, courts have tended to imply it routinely. But if implied in fact there is always room for a court to refuse to imply the term or water it down, especially given the tests for factual implications are so vague and contested, and impose a strict necessity requirement. That the tests for ad hoc terms may not be made out is one reason Australian courts have given for approaching the term as a legal implication.¹⁷⁷ Furthermore, the concept of ‘business efficacy’, relied on to imply terms in fact, does not mesh with the term’s judicially-stated rationale: protection from abuse of power. Business efficacy may be the very reason the defendant sought to include unfettered discretion. Indeed, those critical of the implied term argue it cannot be justified on the basis of business efficacy; this leads them to conclude the term should be expunged.¹⁷⁸ But there is another conclusion to be drawn, which has the strength of being consistent with the continued existence of the term: it is a legal implication. The more the term expands to include new principles such as relevant considerations or legitimate expectations, the more difficult it will be to justify as necessary for business efficacy or obvious. It is also artificial to premise the term on party intent when it is routinely implied into discretions explicitly stated to be ‘absolute’ or to depend solely on the decision-maker’s subjective assessment.¹⁷⁹

More generally, given the normative pull of party autonomy and free-market thinking within contract, there is a danger that if implication is left to case-by-case decision-making, some judges will resist implying the term. It is not difficult to find resistance:¹⁸⁰ ‘If the parties believe that the market would welcome the term proposed … they have only to persuade the market to adopt it for the future. I wonder whether they would succeed.’ Such statements raise the spectre of inconsistency, but also reinforce the uncertainty that goes with an ad hoc approach. Ironically, certainty is cherished by supporters of market-based conceptions of contract. A factual approach would not only directly undermine the judicially-stated rationale of protection from abuse of power by leaving some parties vulnerable to the untrammelled whim of decision-makers, but the uncertainty associated with an ad hoc approach creates opportunities for gaming. A more powerful, well-advised party will have incentives to assert the term does not arise, while the less powerful party will often lack resources to back Rolls-Royce litigation, and have to lump the decision-maker’s assertion, and potentially oppressive decision, even though the true position is that fetters apply.

One possible argument for conceptualising the term as implied in fact is that it is implied into a range of types of contract—employment, banking etc. A standard term may not be responsive to the peculiarities of different settings. However, there are several responses. First, the requirements imposed include principles, such as non-arbitrariness or improper purpose, that are so basic that any discretionary power should presumptively abide them.¹⁸¹ Second, the same term has been implied in myriad contexts with no doubts raised over its workability. Reflecting this consistent practice, in Braganza the term was referred to as ‘the contractual implied term’, and Wednesbury was characterised as ‘the test’ of rationality.¹⁸² Subsequent courts refer to ‘the Braganza duty’.¹⁸³ Third, to the extent different contexts warrant variation of approach, this is accommodated in the way the term is applied. Doctrinal exegesis revealed the emergence of a context-sensitive approach. Such approach is possible because concepts such as reasonableness are general enough to permit varied application. If the same principles have been applied in administrative law to thousands of statutory powers addressing myriad subject-matters, it is difficult to argue they cannot be adapted to different contractual settings.

The most important argument for conceptualising the term as implied in law, however, is that it bears all the hallmarks of such a term.

First, the term’s rationale lies in wider considerations that transcend any one contractual matrix. That rationale, stated authoritatively in Braganza, is that power not be abused; where one party has a power affecting both parties’ interests, there is a risk of abuse. From this general justification, it follows logically that the term should be implied into every such discretion, because each is characterised by the same basic risk. This is consonant with administrative law, where review principles presumptively regulate every statutory discretion unless expressly excluded.¹⁸⁴ And, indeed, this has been the practice in contract. The same or a similar term has routinely been implied into contracts concerning loans, asset valuation, employment, charterparty, insurance, landlord–tenant, share options, telecommunications, franchises and trading platforms. Consistent implication of the term across an array of contexts is suggestive of a standard approach, and courts refer to the term as ‘standard’ and imposing ‘standard limits’.¹⁸⁵ Of course, the risk of abuse is heightened in certain contexts. But this is addressed through context-sensitive application. Importantly, the more review is characterised by a general framework, as detailed in section 3, the more it shall resemble an autonomous enterprise proceeding according to general law.

A fundamental feature of standard terms is that they are a legal incident of a particular class of contract. The courts have articulated a relevant class in relation to the Braganza term: any contract under which one party is bestowed with a discretion involving the making of an assessment or choosing from a range of options, affecting both parties’ interests.¹⁸⁶ The Court of Appeal has said: ‘In any contract under which one party is permitted to exercise such a discretion, there is an implied term.’¹⁸⁷ And where courts have not implied the term, they have reasoned that the given contract does not fit the articulated class.¹⁸⁸ For legal implications, the given class is typically framed by reference to a type of contractual relationship, such as employer–employee, rather than by reference to a particular type of clause within the contract, such as a clause conferring discretion. However, a contract conferring discretion can be said to give rise to a type of contractual relationship: one where one party possesses discretionary power over the other. In any case, there is no ‘rule’ as to how a relevant class must be defined. Rather, courts have significant flexibility.¹⁸⁹ The class is logically dictated by the normative concerns underlying the Braganza term: all contracts that confer discretion to affect both parties’ interests give rise to a risk of abuse, so constraints are implied into every such contract.

Also symptomatic of the Braganza term being implied in law is that it applies unless explicitly excluded, and shall be ‘extremely difficult’ to exclude, requiring ‘very clear language’.¹⁹⁰ If the term were ad hoc, the analytical starting point would not be that the term applies unless excluded; rather, courts would apply tests of factual implication to determine whether the term should be implied into a specific contract, given its terms and context.¹⁹¹ The reality is that as the jurisprudence has matured, courts hardly ever apply tests of factual implication in any serious way.¹⁹² Rather, the term applies ‘as a result of the authorities’,¹⁹³ corresponding with the general approach to legal implications. Even if the term began life as a factual implication, its now routine implication and core features suggest it has crystallised into a legal implication.¹⁹⁴ Thus, whereas older decisions may be tied to ‘their own particular facts’, Braganza is an authority ‘of general application’, stating ‘the general law’.¹⁹⁵

A theme of the jurisprudence concerning legal implications is that such terms are often recognised to protect vulnerable parties, within relationships of power imbalance, such as employment or tenancy.¹⁹⁶ Concerns over power disparity are systemic within given types of relationship, thus terms apply to the entire class. Similarly, the concern that power may be abused is common to all contractual relationships characterised by one party having discretion to affect both parties’ interests; the Braganza term is ‘standard’ because it goes to the ‘very essence’ of the ‘relationships’ it governs.¹⁹⁷ Another recurring theme is that terms in law often reflect contemporary societal expectations with which contracting practice has not caught up.¹⁹⁸ The Braganza term can similarly be explained as giving effect to an emergent societal expectation,¹⁹⁹ that all power, whether private or public, be commensurate with ideas of fair dealing.²⁰⁰ In this regard, legal implications, by ensuring contracting keeps pace with societal expectations, help to maintain the legitimacy of the institution of contract within society. There may be valid questions over how courts determine social expectations, and whether they are equipped to do so, but these are general concerns with standard terms and are not peculiar to the Braganza term. Even those who normatively favour classifying the term as ad hoc acknowledge that the shift, evident in Braganza, towards a rationale based on broader concerns is difficult to reconcile with the term being a factual implication.²⁰¹

Lastly, if the term is implied in law, there may nonetheless be exceptions to the general class. First, some cases suggest the term may not apply if statute (or some other mechanism)²⁰² provides adequate protection.²⁰³ But care must be taken; the term confers default contractual rights, and the ‘principle of legality’ holds ordinary rights can only be ousted by express statutory words or necessary implication. Thus, that a contract arises in a regulated environment is not sufficient basis for disapplying the term, though it might colour its application.²⁰⁴

Second, there are doubts whether the term should apply to decisions to terminate.²⁰⁵ A firm conclusion cannot be reached here. However, one concern is that there are pre-existing tests governing when the power to terminate arises, which balance parties’ interests; superimposing the Braganza principles could cut across that balance. On the other hand, given the repeated concern in discretion cases that no party should be subject to another’s whim, and given termination can have drastic consequences and be used oppressively, there may be a good case for further controls, especially as there is presently little regulation of exercise of the power to terminate.²⁰⁶ Legitimate concerns over protecting the innocent party’s freedom to choose could be accommodated through light-touch review, which would allow for significant decisional autonomy while imposing very basic standards. In administrative law, the ‘super-Wednesbury’ standard has been developed for such sensitive contexts, restricting intervention to cases of mala fides or improper motive.²⁰⁷

A third possible exception arises from an enigmatic line of cases in which courts have refused to imply the term as the parent clause confers an ‘absolute contractual right’.²⁰⁸ This label is often invoked in conclusory fashion and it is unclear how powers so characterised differ from other discretions, or why the term’s rationale—protection from abuse of power—is inapplicable. Suggested differences include that such powers confer binary choice rather than a range of choice, and are framed broadly.²⁰⁹ The arguments are flimsy. An employer may have a broadly-framed binary choice whether to confer a defined bonus, yet the term applies, while fact-finding powers implicate no choice, only one truth of the matter, yet the term applies. Another suggested difference is that absolute rights are designed to serve the decision-maker’s interests.²¹⁰ But it is well-recognised, as in the case of lenders’ powers to vary interest rates, that even where discretion is included for the decision-maker’s benefit, and may legitimately be exercised for self-interest, it may not be exercised abusively.²¹¹ Indeed, where discretion is included to serve the decision-maker’s interests, that is precisely where the temptation for abuse may be greatest. It is also said that absolute rights, when exercised, do not affect the other party’s interests, and thus do not fall within the class of contract to which the Braganza term applies.²¹² Where deployed, this argument has been unconvincing, as where a decision-maker decided to suspend payments under a settlement agreement; the court’s view that this did not affect the payee’s interests is perplexing.²¹³

Given the distinction between ‘absolute rights’ and other discretions is one without difference, it is unsurprising that recent appellate decisions have rejected arguments based on ‘absolute rights’, cautioning against the concept’s blunt invocation.²¹⁴ Significantly, in PAG, the Court of Appeal rejected the argument specifically in relation to a broadly-framed power which conferred binary choice and was inserted for the decision-maker’s benefit.²¹⁵ While the Court acknowledged a commercial decision-maker may act for self-interest, it refused to accept a decision-maker is free to act vexatiously or maliciously. The decision would appear to signal the end for the ‘absolute rights’ concept. Yet despite PAG some lower court²¹⁶ decisions continue to rely on the concept, specifically to afford commercial decision-makers untrammelled discretion. It is difficult not to see continued invocation of ‘absolute rights’ so as to afford commercial decision-makers a free hand as subversion of the policy against abuse of discretion enunciated in Braganza, in favour of absolutist commitments to free market and party autonomy concerns.²¹⁷

5. Remedial Consequences

What are the remedial consequences where a decision-maker fails to comply with the implied term? Neither courts nor commentators have addressed the question in depth. There are assertions in the literature that breach of the term sounds in damages, like any other breach of contract, so ‘little needs to be said’ of remedies,²¹⁸ and that there is no mechanism for quashing decisions in contract.²¹⁹ However, it is difficult to reconcile these statements with the fact judges have voided exercises of discretion. The reality is that there is an open question as to the proper remedial approach to non-compliance with the implied term. There are at least three possible remedial models that could be adopted, each of which finds some reflection in the jurisprudence and each of which raises discrete issues.

A. Model 1: Damages

Courts have often treated non-compliance with the implied term as a standard contractual breach sounding in damages. However, this approach suffers several problems.

First, such approach often requires the court to stand in the decision-maker’s shoes and determine how the power should rationally have been exercised. For example, where a discretionary bonus is denied unreasonably, at the damages stage the court decides whether a bonus should have been awarded and quantum. This is difficult to reconcile with the supervisory conception of the court’s role and a normative concern with good decision-making processes, rather than pre-empting outcomes. Courts state repeatedly that it is ‘not open’ to them to ‘retake’ the decision, which is for the repository and ‘no one else’.²²⁰ Further, decisions whether to award discretionary bonuses, and how much, are paradigm examples of decisions courts feel ill-equipped to assess given they implicate broad value-judgements.²²¹ Yet, where proceedings result in the court ordering the defendant to pay a specific sum, the decision-power is effectively overridden, as is the parties’ intention that the repository exercise the discretion. This disjuncture was brought squarely into focus in Horkulak: Potter LJ said that determining damages ‘will involve the court in assessing the employee’s bonus … thus putting itself in the position of the employer’,²²² despite earlier stressing the court must not supplant the decision-maker.

A second problem with damages is that such remedy does not affect the validity of the original decision. In law, the decision continues to exist, and could logically continue to have legal effects.

A third problem concerns application of the basic measure of damages. Contract damages are generally assessed on the expectation measure; that is, the court awards the monetary equivalent of what the claimant expected to receive under the contract. However, the Braganza-style term only imposes requirements about certain qualities a proper decision should have. The term creates no entitlement to or expectation of a particular outcome; there are any number of possible outcomes a reasonable decision-maker could reach. It is difficult therefore to see what benchmark expectation damages could be assessed against. Indeed, in bonus cases, lower-court judges have struggled to assess quantum,²²³ while appellate courts acknowledge serious difficulties where the discretion is ‘at large’, and where there is no ‘signpost’ or ‘point of reference’ for quantification or past bonuses to analogise to.²²⁴ In other cases, damages have been denied as the claimant was unable to prove their loss, it being impossible to settle on one figure rather than any other.²²⁵

Where the expectation measure proves difficult, other measures may be applied. One possible measure is loss of a chance.²²⁶ But where the amount of a bonus is discretionary, similar difficulties arise: loss of a chance of which sum? Alternatively, the reliance measure could be applied. Damages would correlate to monies the claimant expended in anticipation of a lawfully-made decision, which are rendered wasted when an unlawful decision is made. But the claimant must prove they would not have expended the monies but for the implied promise of a reasonably-made decision, rather than the expectation of a specific decision-outcome—which may prove difficult. And in many contexts, including bonus cases, the claimant is unlikely to have outlaid monies in anticipation of performance. Even if these issues are overcome, it is open to the defendant to prove the monies would have been lost even if the implied term had been complied with.²²⁷

B. Model 2: Voidness

An alternative approach is that where a court finds non-compliance with the implied term, the decision would be void ab initio; the purported decision would not constitute a valid decision, as it was not made in accordance with requirements necessary for validity. The decision-maker could then be enjoined to remake the decision lawfully.²²⁸ This approach meets concerns with treating damages as the primary remedy: it avoids the court supplanting the decision-maker’s role, and the flawed decision is deprived of legal effect.

There is ample precedent supporting voidness as a remedial consequence in contract. It is well established in domestic tribunal cases that decisions²²⁹ or rules²³⁰ inconsistent with procedural fairness are void. Thus, where a member of a society is deprived of membership absent due process, the decision is void: it is ‘as if no act had been done to deprive him’, so in law he never ceased being a member.²³¹ Indeed, in Ridge, which established the modern administrative law of procedural fairness, the Law Lords derived the consequence of voidness for breach of natural justice from domestic tribunal cases.²³² Further, exercises of discretion contrary to the Braganza-style term have been held void and of no effect. In Westlb and The Product Star, the courts held ‘the discretion was never exercised’,²³³ and in Hannover, a defective decision was of no effect.²³⁴ Courts routinely employ language suggestive of voidness as a consequence, such as ‘purported’, ‘valid’ and ‘invalid’.²³⁵ In Sim, the court held a teacher is under no obligation to comply with their employer’s unreasonable administrative directions.²³⁶ One way of explaining this is that an unreasonable direction is a nullity. A procedurally defective notice of dismissal is void,²³⁷ as are other types of notice given contrary to legal requirements.²³⁸ At common law, where a covenant against assignment without consent is qualified by a proviso that consent cannot be withheld unreasonably and consent is unreasonably withheld, there is no right to damages: the tenant can simply make the assignment.²³⁹ The same approach is adopted where a decision-maker, acting contrary to the Braganza term, withholds consent to stock options being taken.²⁴⁰ And a refusal to approve purely contractual sublicences was ‘set aside’ and ‘invalid’ for non-compliance with the implied term.²⁴¹

Contract therefore has the resources to recognise voidness as a remedy. However, an important drawback of this model is that, as a matter of logic, it is difficult to see how damages could be awarded. If a purported decision never existed, it cannot constitute a breach of contract, which is a prerequisite to awarding damages.²⁴²

Nonetheless, it does not follow that a money remedy could never be granted. For example, the money award in Braganza was consistent with the contractual decision being void. Mrs Braganza was contractually entitled to a prescribed benefit upon Mr Braganza’s death in the course of employment. The entitlement could be excluded if BP, Mr Braganza’s employer, established he had committed suicide. BP’s decision to this effect was found to be unsupportable. The consequent order²⁴³ that BP pay the benefit is explicable as an order for payment of an agreed sum, where there was no valid decision excluding the benefit.

Similarly, in Mallone, an employer’s decision to reduce a former employee’s stock options to nil was ‘invalid’ and ‘of no effect’ for non-compliance with the Braganza-style term.²⁴⁴ Absent valid alteration of his benefits within a given time, which had now passed, Mallone was contractually entitled to his mature options; in principle, he thus had a right to damages equivalent to the value of those stocks.²⁴⁵

The Product Star presents a further scenario where damages may be awarded despite the contractual decision being a nullity.²⁴⁶ A vessel’s owners were in prima facie breach of a charterparty for failing to discharge at a prescribed port. They sought to avoid liability by relying on a clause granting them discretion not to dock if it was too dangerous. The court held the decision not to dock lacked rational foundation. As such, the discretion was ‘never exercised’,²⁴⁷ so damages were awarded for the breach involved in not discharging as required.

In other circumstances where a decision is void, a claim for restitution might be open. For example, BHL was a successful claim for restitution of funds paid pursuant to an irrational decision setting a high fee.²⁴⁸ The unjust factor was mistake: the claimant paid the sums under a mistaken belief that they were obliged to pay the high fee. The court did not consider the legal status of the impugned decision. But the judgment only makes sense if the impugned decision were void—if legally effective, how could there be a mistake of law?

C. Model 3: Voidable

A third possibility is that a judicial finding of non-compliance with the implied term does not render the impugned decision void but voidable, meaning a court could choose between voiding the decision or leaving it intact and awarding damages for breach.

The advantages over Model 2 include that the court can tailor the remedial response to the case, maintain a balance between the parties’ interests through remedial choices and avoid the consequences of voidness where unattractive. The principal drawback is introduction of a measure of uncertainty as to remedial outcome, which may be particularly unwelcome in commercial settings. However, the decision whether to void could be structured to reduce uncertainty and facilitate consistency. For example, courts could recognise a starting presumption that non-compliance results in voidness, motivated by concerns with damages as the principal remedy (discussed above), with enumerated exceptions. This approach would promote transparency, a value not always apparent in remedial decision-making in judicial review, which is not characterised by a consistently-applied framework, resulting in an unruly jurisprudence.²⁴⁹ Open acceptance that practicalities may necessitate an outcome other than nullity would avoid the confusion that characterises judicial review, where courts maintain that unlawfulness invariably results in voidness, despite the proposition being obviously unworkable and not borne out in practice.²⁵⁰

Thus, in IBM, Warren J observed that, ‘as with all extreme propositions’, a cautious approach should be taken to the view that all unlawful exercises of discretion invariably result in invalidity.²⁵¹ He favoured flexibility: in some cases, it would not be ‘just or proportionate to strike down the exercise of power’; in others, the result may be ‘partial’ rather than ‘total’ invalidity; and in yet others, ‘monetary compensation’ would be preferable to ‘unravelling’ the decision.²⁵² Other judgments also suggest that invalidity and damages are discrete remedial possibilities.²⁵³

On this approach, one situation where invalidation may be an appropriate response is where the decision would otherwise have ongoing effects within a long-run contractual relationship, making it important to deprive the decision of legal force. Voiding a decision might also be appropriate in the context of a wide discretion that requires broad value-judgements to choose among a range of options. Ceteris paribus, it would be preferable to void the decision and refer the matter back to the repository, to exercise their discretion lawfully—rather than a court, within a damages inquiry, speculating how the discretion would have been exercised, absent objective parameters, and overriding the decision-maker’s right to choose.

On the other hand, damages may be apt in some cases. Consider a situation where a decision-maker, pursuant to a narrowly-framed discretion, has a binary choice whether to confer a defined benefit, which has been conferred on every person similarly-placed to the claimant, and the only basis for denying the claimant the benefit was an irrelevant consideration. In such circumstances, there is a prima facie case for conferring the benefit, based on consistency; no rational basis for refusal if the matter were referred back; and unlike discretionary bonus cases, assessing damages does not involve the court in the value-laden task of deciding the appropriate sum, as the benefit is contractually-defined. Other examples where courts may resist voiding, and referring back the decision, include where the contract has ended or there are ongoing reasons to doubt the decision-maker’s bona fides.

6. Conclusion

The post-Braganza case law is developing apace, giving rise to important doctrinal and conceptual issues. This article addressed three central issues.

First, the article interrogated the nature and practice of Wednesbury review. The intensity of review is context-sensitive, being dependent on four key variables: the contractual context; the nature and scope of the power; the decision-maker’s capacities; and the interests of the party subject to the discretion. A series of subprinciples, or indicia of unreasonableness, guide when judges will intervene on the Wednesbury ground. Together, these doctrinal features comprise an emergent analytical framework governing Wednesbury review in the contractual setting.

Second, the article argued that the implied term is best conceptualised as a standard term implied into every contract which confers a discretion on one party that implicates both parties’ interests. Claims that fetters on discretion are case-by-case factual implications are increasingly out of step with fundamental features of contemporary case law.

Third, the article challenged the common assertion that damages are the invariable remedy for non-compliance with the implied term. First, damages as a remedy for breach of the term raise problems of principle and pragmatism. Second, other remedial models are evident in the jurisprudence, including the remedial consequences of voidness and voidability. These models meet concerns raised by the damages remedy, but raise others. Voidness as an invariable consequence may be too blunt a remedial tool. By comparison, voidability has the advantage of finesse, but may suffer drawbacks associated with discretionary remedialism.

My thanks to Nick McBride, David McLauchlan, Andrew Robertson, Peter Turner, Nicola Varuhas, Julia Wang and the anonymous referees for valuable comments on earlier versions of this article. Drafts were presented at the Obligations IX Conference, Melbourne, July 2018, and a Melbourne Law School Obligations Group Public Lecture, May 2019, and I am grateful to participants at each event for helpful comments. The usual disclaimer applies.

Footnotes