Can Justice Be Fair?

The High Sheriff’s Law Lecture
given by The Honourable Sir Paul Walker, Judge of the Queen’s Bench Division of the High Court
at The Examination Schools, Oxford
on Tuesday, 9 October 2007

The Honourable Mr Justice Walker

Paul Walker was born in Wellington, New Zealand, and brought up in Adelaide, South Australia, where he started but failed to complete a university degree. After working as a waiter in France and a cocktail barman in London, he had the good fortune to study law at Magdalen College, Oxford.

From 1980 to 2004 he practised at the Bar in London, taking time out from 1994 to 1996 to teach law at Victoria University of Wellington, where he was also the director of the New Zealand Institute of Public Law. At the Bar his practice combined shipping and general commercial work with arbitration and a wide range of administrative law cases.

From 1998 to 2000 he was counsel to the BSE Inquiry into “mad cow disease”. He was appointed Queen’s Counsel in 1999, a Recorder in 2001, and a Judge of the Queen’s Bench Division of the High Court in 2004.

Can Justice Be Fair?

This is the third lecture in a series generously sponsored by Ian and Caroline Laing. It was started by Ian Laing when he was High Sheriff of Oxfordshire, and I am particularly grateful to the current High Sheriff, Thomas Loyd, and his team, who have done an enormous amount to make the arrangements associated with today’s lecture.

The first lecture was given by Lord Bingham, the senior Law Lord and a former Lord Chief Justice. The second was given by Lord Phillips, the current Lord Chief Justice. Today you have a speaker who is not anything like as grand. My predecessors together have nearly fifty years of judicial experience. By contrast, I am still relatively new to the job of being a judge - I was appointed in November 2004.

All I can say is that I have had a couple of years working at the coalface, and that being asked to give this lecture has challenged me to think about the justice and fairness of what I have been doing. I welcome that challenge, and I particularly welcome the opportunity to talk about it to you, my audience today.

You come from many different backgrounds. Many of you are at a stage when you are about to finish secondary education. Most of the rest of us are, I fear, long past secondary education. Whatever your age and experience, I am working on the basis that you are here because you have an interest in how our society provides justice. I hope that by the end of this lecture I will have encouraged you to think more about this.

The title to this lecture poses a question: “Can justice be fair?” An immediate response might be that justice must be fair. If you look in the dictionary you will see that one of the definitions of “just” is “fair”, and conversely one of the definitions of “fair” is “just”. The two go hand in hand.

It seemed to me that if I were to adopt that definition the lecture would have to come to an end now, which is probably not what the High Sheriff would want. So I looked for other definitions.

One definition of justice was given by the 19th-century writer Samuel Butler. He said, “Justice is my being allowed to do whatever I like.” No doubt Samuel Butler was deliberately seeking both to shock and to amuse. Perhaps he was also seeking to make us think about our approach to justice. If justice is being allowed to do whatever I like, where is fairness? Samuel Butler’s definition is consistent with being fair to one person only - the speaker.

I do not think that either justice or fairness can or should be viewed from the perspective of one person only. So I disagree with Samuel Butler’s definition. Instead, I shall define “justice” as “what the courts do.”

I am not going to give a philosophical discussion of what the courts do. Throughout this lecture what I am concerned with is what happens in practice when people find themselves in court. I shall take examples from three areas - criminal law, commercial law and administrative law. In these three areas the courts are asked to do very different things, but there are also many common features where questions of fairness arise.¹

First, anyone who becomes a judge must take the judicial oath:

“… I will do right to all manner of people after the laws and usages of this realm, without fear or favour, affection or ill will.”

Holding true to the judicial oath will in itself go a long way towards achieving fairness.

Second, there are plenty of people who think that what the courts do is not fair. Sometimes the only explanation for this is that they are taking Samuel Butler’s approach. But often there is something else which underlies the complaint of unfairness. Anyone who spends any period of time in the courts will come across disappointed litigants, some of whom could be straight out of the pages of Bleak House. As lawyers and judges we like to think that the terrible injustices described by Dickens have been ended. It would be unreal, however, to pretend that no unfairness remains.

Third, courts can get things wrong. They may be misled by a witness who lies when giving evidence, or persuaded by a witness who mistakenly, but confidently, identified the wrong person. They may misunderstand the arguments, or interpret the law in a way which is later found to be incorrect. Our legal system provides an appellate structure, a system of appeals, so that first-instance judges - and sometimes appeal judges - can be put right if they get things wrong. Our legal system also provides a way of remedying things where what went wrong emerges many years later. The civil courts long ago recognised that a judgment obtained by fraud could be overturned. Criminal cases can now be referred to the Court of Appeal by the Criminal Cases Review Commission.² These are just some of the ways in which, after the court’s decision, it may become clear that the court was wrong.

I do not think that this necessarily shows that the court was unfair. Let me come back to the part of the judicial oath which refers to the usages of the realm. Our system of justice is adversarial, and it involves constraints which vary for different courts.

The adversarial system means that the court does not have the task of investigating things for itself. Instead, we proceed on the basis that the parties put their evidence before the court and the court determines the outcome on the basis of that evidence. In order to determine the outcome the court must make findings of fact and apply what it considers to be the relevant legal principles to those facts.

Different courts go about that task in different ways. Some courts use simpler and cheaper procedures than others, and so may be open to the criticism that justice has been unfairly sacrificed on the altar of economy. Whatever the procedure, however, justice is not priceless. There have to be restrictions. The court has always been able to insist that evidence must be relevant. Under modern procedures the court can and should limit both evidence and argument where the time which litigants want to use clearly goes beyond that which is necessary fairly to put their case.³

It is vital that the court gives both sides a fair opportunity to assemble their evidence and present their case, and assesses the evidence and arguments fairly, without bias or the appearance of bias - in the words of the oath, “without fear or favour, affection or ill will.”

I believe that it is equally vital that the court arrives at a result which, allowing for the constraints in which the court works, is reasonable. I think most people would agree that if a court meets these criteria then the mere fact that the court’s decision is later shown to be wrong does not mean that the court itself acted unfairly. Fairness does not require perfection.

A fourth common feature is that many complaints of unfairness are complaints about the law. The judicial oath is to do right “after the laws and usages of this realm”. The laws and usages of the realm are found historically in the common law - decisions of the courts made in the past - and in statutes - laws made by Parliament. The court has the role of interpreter. It must decide what a statute actually means, and how the common law applies to the case. In both these areas judges cannot just impose their own views of what is fair.⁴ Judges must be particularly wary of allowing their own conceptions of fairness to interfere with policy decided by Parliament.

A fifth common feature is that criticisms of unfairness are often classified into two categories: substantive unfairness and procedural unfairness. Substantive unfairness focuses on the outcome, and can be said to arise where the court has reached an unfair result. Procedural unfairness focuses on what happened as the case proceeded, and can be said to arise where a party has not had a fair opportunity to present that party’s case. The dividing line between substance and procedure is not always clear.

In this lecture I take an example from each of my three areas in order to examine what happens when judges are more or less explicitly required to consider the substantive fairness of a particular outcome.

Criminal Justice and Substantive Fairness

I start with criminal justice. Criminal courts punish people who are convicted of crime. Last year’s lecture was about punishment, and I could not hope to give you any better insight into that topic. Today I shall focus on an aspect of the procedures leading to conviction where substantive fairness comes into play. To do this I need to give a sketch of what happens at a criminal trial, and to ask those of you who are familiar with such trials to bear with me while I do so.

In many cases - in fact the vast majority - defendants in criminal proceedings plead guilty. If there is a plea of not guilty, however, then a defendant can only be convicted if found guilty after trial. The most serious offences are tried in the Crown Court where the trial takes place before judge and jury. The judge directs the jury about the law, and it is the jury who decide guilt or innocence. Less serious offences are tried in the Magistrates’ Court. Here Justices of the Peace or a District Judge take on the roles both of judge and jury, deciding questions of law as well as deciding whether the defendant is guilty or not guilty. For convenience, whatever the court, questions of fact are often referred to as “jury questions” and I will adopt that terminology today.

In all criminal trials, unless statute provides otherwise, three basic principles come into play. The first is the burden of proof: it is for the prosecution to prove that the defendant is guilty. The second is the standard of proof: the old-fashioned terminology was that the prosecution must prove its case beyond reasonable doubt. Nowadays we say - which is the same thing in different words - that the prosecution must make the jury sure of guilt. Nothing less will do. The third is that the jury must decide guilt or innocence by reference to the evidence at trial.

All these principles come together when the jury retire to consider their verdict. They may, however, come together at an earlier stage, the stage when the prosecution evidence is complete.

Criminal trials in this country begin with a speech by the prosecution summarising the case and the evidence that they are going to rely on to prove that the defendant is guilty. If there is a jury, the prosecution tell the jury that they can only convict if the prosecution have made them sure of the defendant’s guilt. The prosecution then proceed to call witnesses. They also tell the jury about any admissions or written statements which are evidence in the case. Eventually their evidence is complete, and the prosecution formally say, “we close our case.” It is a pivotal stage in the trial. The precise evidence against the defendant has been identified and presented.

Now, in the ordinary course of the trial, it is time for the defence evidence to be called. Not uncommonly, however, defence counsel will ask that the jury withdraw, possibly for twenty minutes or so, possibly for much longer. The reason is that at this pivotal stage the defence can submit to the judge that the prosecution evidence constitutes “no case to answer.”

What does this mean? It means that in certain circumstances the judge must stop the case and direct the jury to return a verdict of not guilty. In R v Galbraith⁵ the Court of Appeal identified two sets of circumstances when this should happen.

First, the case was to be stopped when there was no evidence that the person charged had committed the crime alleged. Second, the case was also to be stopped if the evidence was tenuous and the judge concluded that the prosecution’s evidence taken at its highest was such that a properly directed jury could not properly convict on it.

The Court of Appeal plainly thought that the first of these was straightforward. All I can say is that the straightforward cases never seem to come before me. In practice what are said to be cases of no evidence tend to involve arguments similar to those arising in the Court of Appeal’s second category - evidence so tenuous that a jury could not properly convict.

Take, for example, the discovery by the police of a man running a large cannabis-growing operation in a private house. The defendant is a woman who is the owner of the house. She is charged with permitting the production of cannabis, but the man has escaped from police custody and the police cannot call anyone who actually saw her involved in the cannabis production. The evidence against her is circumstantial.

For many years now judges have said to juries that circumstantial evidence can be powerful evidence, but it is important to consider carefully whether the evidence upon which the prosecution relies in proof of its case is reliable and whether it does prove guilt. Furthermore, before convicting on circumstantial evidence the jury should consider whether it reveals any other circumstances which are or may be of sufficient reliability and strength to weaken or destroy the prosecution case.

In my example it is easy enough to say that mere ownership of the house is not evidence of the offence of permitting the production of cannabis. Add to that, however, evidence that the owner let the house to the man and visited him in the house regularly. Could a jury be sure that she must have known and permitted what was happening? Would it make any difference if the prosecution showed that she appeared to be living well beyond her means?⁶

What we also say to juries is that they are allowed to use their common sense, which means they can draw inferences from established facts. Somehow the judge has to draw the line between those inferences which are reasonably open to the jury and those which are not. Here the judge has a procedural task with a strong substantive element. The judge must assess what substantive outcome could reasonably be reached on the basis of the prosecution evidence alone. Doing this without unfairness to either side will often not be easy.

The Court of Appeal in Galbraith gave some comfort to judges faced with a submission of no case to answer. After pointing out that views as to the reliability of a witness were generally within the province of a jury, it added that borderline cases were in the judge’s discretion. This at least suggests that provided the judge’s decision on whether there is a case to answer was reached on a proper understanding of the law, fairly and reasonably, then in borderline cases the Court of Appeal will not interfere.⁷

A successful submission of no case to answer can be very harrowing for victims of crime. It is particularly harrowing for the families of victims in murder cases. Sometimes a murder charge has only been brought after a lengthy campaign by the family and the press. In such circumstances a finding of no case to answer commonly leads to criticisms that there has not been a fair trial.

There are competing considerations here. On the one hand there is the perceived injustice of requiring someone to defend a charge which has not been made out. On the other hand, however, common sense tells us that if required to defend the charge it is at least possible that the defendant may conduct the defence in a way which convinces the jury of guilt. I do not think it surprising to find that when cases are stopped in this way, victims and their families often complain of unfairness.

The cases which caused the most concern, and led to widespread media publicity, were those where children died after dreadful abuse which must have been inflicted by at least one of two people. Suppose a baby has been left in the care of an elder sister and her boyfriend. When the parents return they find the baby covered in bruises from injuries which lead to death. Here the common law was and is clear. Murder or manslaughter is only proven if the prosecution can show that the defendant caused the death of the victim.

For a charge of murder or manslaughter to succeed against a particular defendant, on that essential element of causation the prosecution must be able to say either that the evidence points conclusively to that particular defendant having inflicted the injuries in question, or else that it demonstrates that the two defendants must have been acting in concert. If the state of the evidence was such that one of the two must have caused the injuries, but it could have been either and need not have been both, then the only possible verdict for each defendant would be not guilty.⁸

Defendants in this kind of case would often run what are known as cut-throat defences - “It wasn’t me, it was the other defendant.” If one of the defendants were to give evidence then the jury would at least have something to go on. The common law said, however, that submissions of no case to answer succeeded and defendants did not have to give evidence. The result gave rise to such an extreme sense of unfairness that Parliament intervened.

It created a new offence under section 5 of the Domestic Violence, Crime and Victims Act 2004 of causing or allowing the death of a vulnerable person. In broad summary, the new offence does not require that the defendant personally caused death or acted in concert with someone who did. It is enough that the defendant knew or ought to have been aware of a significant risk of grievous bodily harm being caused to a vulnerable person by someone in the same household. The jury do not have to be able to say who did what. If, in relation to a vulnerable victim, they are able to say, “We are sure that either this particular defendant caused death or death was caused by a member of the household in circumstances where this particular defendant knew or ought to have been aware of a significant risk of that person causing grievous bodily harm”, then the offence under section 5 is made out.

Section 6 of the 2004 Act adds that where a defendant is charged with murder or manslaughter, so long as the defendant is also charged in the same proceedings with an offence under section 5 in respect of the same death, the question whether there is a case to answer on the charge of murder or manslaughter is not to be considered before the close of the evidence.

Thus the extreme sense of unfairness I identified earlier has been removed in very particular circumstances. What is needed is evidence of a vulnerable victim and of a significant risk of grievous bodily harm from another member of the household. If that is present, then an offence under section 5 can be included in the indictment, and so long as that is the case the jury will be able to return a verdict of murder or manslaughter even if the prosecution evidence alone did not identify the culprit.

It seems to me that this is an example of a case where apparently fair common law procedures led to substantive unfairness, and also of a case where only Parliament could remedy that unfairness.

Administrative Law and Substantive Fairness

I now turn to look at issues of substantive fairness in a completely different area, that of administrative law. Administrative law is concerned with the exercise of public powers. The vast majority of cases involve complaints about something done or proposed to be done by central or local government.

Here too there are different types of court. For much of the day-to-day business of government, such as deciding applications for benefit, tribunals were set up so that the citizen could appeal against adverse decisions. Until recently tribunals would usually come under the umbrella of the department whose decisions they were reviewing. Now most tribunals are the responsibility of the Ministry of Justice and are supported by one common administrative service, known as the Tribunals Service.

The breadth of cases that they deal with can be seen by some of their names: the Asylum and Immigration Tribunal, the Gambling Appeals Tribunal, the Gender Recognition Panel, the Proscribed Organisations Appeals Commission, and the VAT and Duties Tribunal.

Cases which do not fall within the remit of tribunals, and some cases which involve challenges to tribunal decisions, come before the High Court. For the most part they are dealt with in a specialist court, the Administrative Court, which sits in London and Cardiff. A specialist procedure has been developed, known as an application for judicial review. Particularly important cases come before the Divisional Court of the Queen’s Bench Division, where two or sometimes three judges will sit.

As with criminal law, different procedural regimes can be more or less complex, and more or less expensive. A development which I personally strongly encourage has been the increasing use of ombudsmen to resolve disputes. At their best, ombudsmen involve little or no expense to the parties, and offer a speedy, efficient and informal service where the outcome is likely to be strongly influenced by considerations of fairness.

What is the court’s role in relation to government? Nicholas Ridley was a government minister under Margaret Thatcher. When a journalist questioned him about a controversial decision, he gave a reply which displayed the same cynicism as Samuel Butler. What he said was, “It’s my law, and I do what I like with it.” The answer may have been in jest. Whether jesting or not, the fact of the matter is that on at least two occasions his decisions were quashed by the High Court,⁹ and he thus had good cause to know that it was the court, and not the government, which decided the law.

The law applied by the High Court when quashing Nicholas Ridley’s decisions was administrative law. Statute and case law combine to set out the legal principles which determine who can make an administrative law claim, against whom, on what grounds, and the remedies available.

The courts have focused not on whether the defendant is, in general terms, amenable to judicial review, but on whether a particular decision is. The exercise of powers conferred by statute will in many cases be amenable to judicial review, but this will not always be so. Conversely, the exercise of powers which have no statutory basis may in an appropriate case be subject to judicial review.

There has been much debate as to how active judges should be in intervening in the business of elected government. What has been described as “judicial timidity” in the first half of the last century was overtaken by a more activist approach and major procedural reforms. The importance of the judicial function has been confirmed by the Human Rights Act 1998, enacted so that the courts in this country can, in particular circumstances, give effect to the European Convention on Human Rights.

Some 14 years before the Human Rights Act, Lord Diplock identified three grounds of review. He said¹⁰ that the first ground could be called “illegality”, the second “irrationality” and the third “procedural impropriety”. He recognised that further development on a case-by-case basis might in course of time add further grounds, and had in mind particularly the possible adoption in the future of the principle of “proportionality”. This latter ground now applies in most cases governed by the Human Rights Act, and the courts have inched closer towards holding that it can operate as a common law ground of review.

All these grounds of review can in appropriate cases bring into play questions of fairness. Review for illegality requires the court to identify the correct meaning of the relevant statute, and as I mentioned earlier, a court will be reluctant to adopt a meaning which entails unfairness. Questions of procedural propriety commonly involve examination of whether a party has had a fair opportunity to present that party’s case. I am particularly interested today in review for irrationality - commonly now referred to as “unreasonableness” - and the allied ground of proportionality.

The courts have long asserted that those exercising what would now be described as public law powers must act within the confines of what is reasonable. A crucial feature has been a self-imposed injunction that it is not for the courts to determine the merits of an administrative decision. In relation to review for unreasonableness, judgments have stressed that the court is not concerned with what it regards as the appropriate decision, but rather with the quite different test of whether sensible decision-makers, properly directed in law and properly applying their minds to the matter, could have regarded the conclusion under review as a permissible one.

At first the approach of the courts was that the substantive fairness of an administrative decision was a no-go area. It was for the decision-maker to decide whether any particular outcome was a fair outcome. Provided that the procedure adopted by the decision-maker was fair, the court should not intervene.

This came under challenge when the courts developed a concept of legitimate expectations as part of the process of identifying those individuals who would be protected by procedural safeguards. As a matter of procedural fairness the courts insisted that if a decision-maker had encouraged a legitimate expectation that a benefit would be conferred, or continued, then the benefit should not be withheld without a hearing. In the case of a legitimate expectation that a particular procedure would be followed, the court could hold the decision-maker to that procedure. From this developed the suggestion that where a decision-maker could be said to have encouraged a legitimate expectation of a particular substantive result, then the court might, in appropriate circumstances, require the decision-maker to fulfil that expectation.¹¹

Over the same period the courts became increasingly concerned to protect the citizen from abuses of power. In R v IRC, ex parte Unilever plc,¹² both the taxpayer and the Inland Revenue had, for many years, shared an honest error as to the applicability of a time limit. When the Inland Revenue, without notice, sought to rely on the time limit in circumstances where it had suffered no prejudice, the Court of Appeal held that reliance on the time limit was an abuse of power.

In substance, what happened was that for more than 20 years Unilever’s practice was to put forward figures to adjust their tax several years after the events giving rise to the adjustments. The Inland Revenue made no complaint about this, until suddenly it pointed to a time limit in the relevant regulations and told Unilever they were too late. This was so inconsistent with ordinary conceptions of fair dealing that the court intervened.

Sir Thomas Bingham MR¹³ said:

“The categories of unfairness are not closed, and precedent should act as a guide not a cage. Each case must be judged on its own facts, bearing in mind the Revenue’s unqualified acceptance of a duty to act fairly and in accordance with the highest public standards.”

A claimant relying on unfair abuse of power still needs to show more than mere “unfairness” in the general sense.¹⁴ The twin approaches of the courts, both in relation to legitimate expectation and in relation to abuse of power, seem to me to fall well within the recognised ground of review for unreasonableness. The task of identifying a point which goes beyond the limits of reason requires the court to identify and apply the fundamental values of our society.¹⁵ The court is in this way adopting a case-by-case approach to identifying those cases of substantive unfairness that constitute such an abuse of power as to require intervention.

Commercial Law and Substantive Fairness

The third and last area of law I want to touch on is commercial law. The subject means different things to different people. For some it covers anything to do with business. In that sense it might extend to trade marks, patents and other ways in which business is protected. It might also cover transactions between consumers and businesses. In the High Court it can give rise to proceedings in the Chancery Division or in the Queen’s Bench Division. Within the latter division there is a specialist court, the Commercial Court.

The rules governing the Commercial Court describe a “commercial claim” as including any claim involving a business document or contract. However, the work of the Commercial Court tends to involve specialist areas such as shipping, banking and reinsurance.

Commercial law has long been familiar with alternative dispute resolution. Historically this has focused on arbitration, where the parties to a contract agree that disputes will be resolved by a tribunal of one or more arbitrators appointed by the parties or by somebody they trust. Now, however, numerous forms of mediation exist, and in many fields there are inexpensive ombudsmen schemes, often with the advantages I mentioned earlier of cheapness, speediness and ability to decide matters by concentrating on fairness. Prominent among these has been the Insurance Ombudsman, whose decisions have gone a long way towards mitigating some of the least satisfactory features of insurance law.

The vast bulk of commercial law is concerned with the law of contract - deciding when a legal agreement is made, who is bound by it, what it means and what happens when things go wrong. Historically the common law of contract concerned itself only rarely with substantive fairness. With the exception of a few special cases it took the view that adult contracting parties of sound mind could be left to themselves to determine the fairness of their agreements. That approach led to a great deal of unfairness to consumers, which in turn led to major statutory changes under the Unfair Contract Terms Act 1977.

A relatively recent case, however, emphasises that even where no statute is involved there can be room for argument about substantive fairness in the general law of contract. The case I have in mind is called Equitable Life Assurance Society v Hyman,¹⁶ and I put it broadly under the rubric of commercial law as it is concerned with a contract of insurance.

In the 18th century mutual societies were formed to provide life assurance. They were a special kind of club. A deed of settlement set out the contractual obligations which each member undertook to the other. Equitable Life Assurance Society, which I shall call “Equitable”, was one of these. Each member contributed to the assets of the society, and would gain benefits from the capital and revenue profits generated by those assets.

By the mid-20th century a well-recognised structure was adopted in which members took out life assurance policies and were thus known as policyholders. At regular intervals the society would declare various types of bonuses which determined the profit for that period accruing to members. Also by the 20th century Equitable Life had registered under the Companies Acts as an unlimited company, and so the contractual obligations of members were set out in Articles of Association.

One of the attractions of life assurance was that policies could make provision for pensions. A substantial part of the value of the pension would be determined by the amount of bonuses declared at the date of retirement. The annual pension that a policyholder would receive on retirement could then be calculated, and the basis for this calculation would be to multiply a capital amount, comprising all or a specified part of the total bonuses, by the current annuity rate.

The recent problems of Equitable have caused misery to many policyholders. The detail is complex, but I can outline the position in broad terms. In the late 20th century the pensions market became more and more competitive. Equitable decided that one way to attract customers was to offer a guaranteed annuity rate. I will refer to policies with these guaranteed annuity rates as “guaranteed policies”.

In the 1980s and early 1990s this cost Equitable nothing, as the current annuity rates were above the guaranteed level. However, in the mid-1990s market conditions changed. The guarantees became expensive to honour. Life assurance companies like Equitable stopped selling policies with guaranteed annuity rates. That, however, did not solve the problem of how to finance the cost of meeting the guarantees already issued.

Equitable said that bonuses were intended to represent a fair distribution to members of their shares in the value of its assets, and that it was unfair for the guaranteed policyholders to gain a disproportionate share of that value. From the end of 1993 the directors of Equitable resolved to award different bonuses to those who had guaranteed policies. Their bonuses would be reduced so that even though they were entitled to insist on the guaranteed rate, if they did so the capital amount was reduced to a level where the annual payment would be the same as if there had been no guarantee.

The power to do this, said Equitable, lay in the Articles of Association. Article 65 said that the directors could apportion bonus on such principles and by such methods as they might from time to time determine. The directors justified differential bonuses as being fair to all the membership. They added that whether this was fair was for them to decide, not the courts.

The first-instance judge agreed with them. The Court of Appeal were split. In the House of Lords all five Law Lords held that Equitable were wrong. The leading speech of Lord Steyn¹⁷ decided the case against Equitable on the basis that although the contract contained no express limitation on the directors’ powers to declare bonuses, the court should imply such a limitation:

“… the self-evident commercial object of the inclusion of guaranteed rates in the policy is to protect the policyholder against a fall in market annuity rates by ensuring that if the fall occurs he will be better off than he would have been with market rates. … The supposition of the parties must be presumed to have been that the directors would not exercise their discretion in conflict with contractual rights. These are the circumstances in which the directors of the Society resolved upon a differential policy which was designed to deprive the relevant guarantees of any substantial value. In my judgment an implication precluding the use of the directors’ discretion in this way is strictly necessary. The implication is essential to give effect to the reasonable expectations of the parties. The stringent test applicable to the implication of terms is satisfied.”

The decision of the House of Lords has caused much controversy among lawyers.¹⁸ I said earlier that it caused misery for many policyholders. Equitable’s reserves were inadequate to meet the guarantees. It had to close to new business and reduce the apparent value of its in-force policies by around £5 billion.¹⁹ One can sympathise with the view of many that the overall outcome could hardly be described as fair. To my mind, however, that view ignores two vital features of what courts do when they are interpreting contracts.

First, the court decides the interpretation of the contract by reference to the circumstances when the contract was made. In contracts between businessmen there is a strong case for seeking to ensure that the court interprets the contract in a way which meets their reasonable expectations.²⁰ Second, when laying down principles of interpretation the courts have in mind not only the parties to the present case but also the principles which are to apply for cases in the future. The Equitable Life decision is a clear recognition of certain fundamental principles of fair dealing in commercial law.

A Conundrum

Having given you examples from each of these three fields, I now go back to the distinction between procedural fairness and substantive fairness. Here I have a conundrum - a puzzle that I have found perplexing.

The decision of the House of Lords in Equitable Life was a unanimous decision. What happens, however, when a court is not unanimous? Here it seems to me that substantive fairness and procedural fairness really do merge.

One of the most dispiriting things I learnt at university came from Sir Rupert Cross’s book on precedent.²¹ He was writing about appeal courts. In fact, the principle concerns what happens whenever you have a tribunal comprising more than one person, and there is more than one legal question to be resolved. How do you decide who wins?

To explain the problem, consider a notional case where three points of law arise, and the claimant appeals to the Court of Appeal after losing at the trial.

Suppose a statute says:

“Any resident of England and Wales who is hit by any motorised vehicle and suffers damage shall be entitled to compensation from the owner of the vehicle.”

The claimant is an illegal immigrant who has lived in England for many years. A car starts to roll downhill with its motor turned off. It hits the claimant, who suffers no physical injury but develops a phobia of cars, making it impossible to lead a normal outdoor life.

The defendant owner of the vehicle raises three points by way of defence.

The first is a contention that to be a “resident of England and Wales” a claimant must be lawfully resident.

Second, the owner says that a “motorised vehicle” does not include a vehicle with its motor turned off.

The third point is a contention that “damage” means physical damage and does not include a mental phobia.

I shall call the point about lawful residence “Issue 1”, the point about the motor being turned off “Issue 2”, and the point about lack of physical injury “Issue 3”.

In order to win the case the claimant must succeed on all three issues. Conversely, if the defendant succeeds on any one of these issues then there will be no liability under the statute. My puzzle does not ask you to decide what the law should be on any of these issues. It asks you to consider what happens if there are three judges in the Court of Appeal and they reach different conclusions on each issue.

Here is the position:

Issue 1Issue 2Issue 3Who wins the case according to each Lord Justice?Lord Justice A✓✓✗DefendantLord Justice B✓✗✓DefendantLord Justice C✗✓✓DefendantWho wins on each issue?ClaimantClaimantClaimant

✓ = agrees with claimant
✗ = agrees with defendant

Conundrum: who has won after the appeal?

What the diagram shows is that A thinks that the claimant is right to say that an illegal immigrant can claim, and is right to say that the motor need not be turned on. However, A thinks the claimant is wrong on the question whether mental damage is enough.

B thinks that the claimant is right to say that an illegal immigrant can claim, and is right to say that mental damage is enough. However, B thinks the claimant is wrong on the question whether the motor needs to be turned on.

C thinks that the claimant is right to say that the motor need not be turned on, and is right to say that mental damage is enough. However, C thinks the claimant is wrong on the question whether an illegal immigrant can claim.

If you ask the question, “who wins?” separately on each issue, the claimant wins by a majority of two judges to one on each issue. A and B both think the claimant is right to say that an illegal immigrant can claim. A and C both think the claimant is right to say that the motor need not be turned on. B and C both think the claimant is right to say that mental damage is enough. So the claimant has won on all three issues, and surely must be entitled to compensation under the statute.

Or must that be the case?

If you ask the question, “who wins?” by asking the three judges to vote on whether the claimant wins, imagine that you are sitting in court and you listen to them give their judgments in turn.

The upshot of A’s judgment is that the claimant fails because the statute only applies if there is physical damage, and so the appeal should be dismissed.

The upshot of B’s judgment is that the claimant fails because the statute only applies if the motor is turned on, and so the appeal should be dismissed.

The upshot of C’s judgment is that the claimant fails because the statute only applies to those lawfully resident, and so the appeal should be dismissed.

Thus all three judges are of the same view - the claimant loses.

Everything here turns on procedure. Sometimes cases are decided issue by issue. Usually, however, the assumption is that they will be decided by a majority of votes. In our example, that procedure would mean that the defendant has had a resounding victory, while deciding the case by issue would give the claimant victory by a margin of two to one.

Is one result any more or less fair than the other? I said earlier that at university I found this conundrum dispiriting. The reason was that it seemed to make fairness an even more elusive concept than justice. Now I am less perturbed by it. Cases where the problem arises are likely to be rare.²² If the problem were to arise, I am confident at least of this: the court would strive to find a fair answer to it.

Concluding Thoughts

Lord Bingham, when giving the first of this series of lectures, said that justice was “the most elusive of concepts”. Using my definition of “what the courts do”, I have given you various examples of how courts try to be fair.

When judges say that courts at least try to be fair, a sceptic can be forgiven for replying, “He would say that, wouldn’t he?” When seeking to answer the sceptic, the feature that I would stress most strongly is public involvement in justice.

Public understanding of justice is promoted by active discussion in all our educational institutions of what courts do. It is essential to proper public involvement in justice that there is a free, responsible, independent and fearless press.

Turning to fairness, it seems to me that in the most difficult cases fairness involves striking a balance between conflicting virtues. In striking that balance courts must be constantly alert to the dangers of prejudice and assumption. It is vitally important that their actions are monitored and criticised.

If what courts do affects any one group in a disproportionately adverse way, then there must be a rethink so as to excise anything not essential to those actions and decisions. Judges must be cautious before concluding that their conception of fairness is one that commands such broad assent in the community that it can be adopted by the court.

This is, I think, one of the most important aspects of what judges do, and it can only be done well with your help.

¹ These common features are found in many other areas of law as well.
² Criminal Appeal Act 1995, section 9.
³ See Lord Hoffmann in Sutradhar v Natural Environment Research Council [2006] UKHL 33 at para 42. The Civil Procedure Rules and the Criminal Procedure Rules begin by setting out an “overriding objective” of dealing with cases “justly”. The Civil Procedure Rules define “justly” to include saving expense, and dealing with the case in ways which are proportionate to the amount of money involved, to the importance of the case, to the complexity of the issues, and to the financial position of each party. The definition of “justly” in the Criminal Procedure Rules makes no express reference to cost, but includes dealing with the case efficiently, and taking into account the gravity of the offence alleged, the complexity of what is in issue, and the severity of the consequences for the defendant and others affected. Both sets of rules require that when deciding procedural issues in a particular case the court must have regard not only to the circumstances of that case but also to the needs of other cases.
⁴ In both these areas, however, fairness has a limited role to play. When interpreting statutes, if one meaning of the words used would entail obvious unfairness and the other would not, then a court will naturally tend to adopt the meaning which avoids unfairness. Similarly, if one interpretation of the common law would entail obvious unfairness and the other would not, then a court will naturally tend to adopt the interpretation which avoids unfairness.
⁵ [1981] 1 WLR 1039.
⁶ The Court of Appeal has said that a judge should exclude evidence of this kind if it is of marginal relevance: R v Gordon [1995] 2 Cr App R 61. If such evidence is adduced, the judge should normally give the jury a special direction designed to warn against jumping to conclusions: see R v Grant [1996] 1 Cr App R 73 and R v Malik [2000] Crim LR 567.
⁷ One such borderline case was R v Lesley [1996] 1 Cr App R 39.
⁸ R v Lane and Lane 82 Cr App R 5.
⁹ See R v Secretary of State for Transport, ex parte Greater London Council [1986] QB 556, decision on the financing of the London Transport Executive held to be irrational; R v Secretary of State for Transport, ex parte Greater London Council [1986] JPL 513, decision prohibiting the GLC from banning heavy lorries held to have been made without jurisdiction. There were, nevertheless, many other legal challenges from which he emerged unscathed.
¹⁰ In Council of Civil Service Unions v Minister for the Civil Service [1985] AC 374 at 410.
¹¹ R v Secretary of State for the Home Department, ex parte Khan [1984] 1 WLR 1337. In R v North and East Devon Health Authority, ex parte Coughlan [2001] QB 213, the Court of Appeal recognised that a breach of promise may be so unfair as to constitute a reviewable abuse of power. For more recent cases see articles by Mark Elliott, Chris Hilson, Ashley Underwood QC and David Anderson QC in [2006] JR at 281, 289, 294 and 348.
¹² [1996] STC 681.
¹³ [1996] STC 681, 692. Note also the observation of Simon Brown LJ at [1996] STC 695 about unfair abuse of power being illogical or immoral or both - an observation which links with Lord Diplock’s explanation in Council of Civil Service Unions of irrationality as a ground of review.
¹⁴ See Judge J in R v IRC, ex parte MFK Underwriting Ltd [1990] 1 WLR 1545 at 1573.
¹⁵ See Supperstone, Goudie and Walker, Judicial Review, 3rd edition, 2005, para 8.16.2.
¹⁶ [2002] 1 AC 408.
¹⁷ [2002] 1 AC 408 at 459.
¹⁸ For example, see the case comment by Alan Berg at [2002] JBL 570.
¹⁹ Lord Penrose’s report to the House of Commons, Foreword, para 6.
²⁰ See David McLauchlan, “The ‘New’ Law of Contract in New Zealand” (1992) NZ Recent LR 436. A valuable recent discussion is found in Adam Kramer, “Implication in Fact as an Instance of Contractual Interpretation” (2004) 63 Camb LJ 384.
²¹ Precedent in English Law, Rupert Cross and J W Harris, 4th edition, 1991, p 92, citing 66 LQR 298, which attributes the point in question to Lord Simonds.
²² Cross and Harris refer at p 93 to Boys v Chaplin, reported as Chaplin v Boys [1971] AC 356. The question of what the case actually decided is discussed in the 12th edition of Dicey and Morris on the Conflict of Laws, 1993, pp 1494-1498. Whether it is a true example of my conundrum may depend on how you formulate the issues.