doctrine of judicial precedent uk essay

Hall Ellis Solicitors

Doctrine of Precedent: Court Cases and Binding Judicial Decisions

The doctrine of precedent.

The common law doctrine of precedent is applied by courts to promote predictability and consistency of judicial decision-making.

The application of precedents is the main factor in the way the law is applied equally to all.

It means that the law is applied with a greater degree of certainty. The law also remains capable of organic change.

The doctrine of precedent is the basis of the common law judicial decision-making process. It's how precedents in law:

• become established as a result of a broadly consistent series of judicial decisions
• are followed by judges in later comparable cases which come before the courts, and
• binding precedents are replaced when they are no longer good law.

The doctrine of precedent achieves this because precedents are applied to later cases with a similar or comparable factual background.

The doctrine of precedent is also known as stare decisis: "to stand by things decided". It stems from the legal maxim stare decisis at non quieta movere which translates as “to stand by things decided and not disturb settled points”.

It means that the same legal issue or principle doesn’t need to be decided again. It is authoritative precedents and persuasive precedents which are used to decide later cases.

Stare Decisis and Judicial Precedent

Stare decisis means - in theory - that any lower court in a judicial hierarchy is bound by the point of law decided in a comparable case by a higher court.    

The law as stated will be applied to other comparable cases which come before courts in the future.

But then it is a relatively simple task for a court to distinguish its decision and not follow a precedent when it wishes to do so.

What is a Precedent in Law?

A judicial decision or a case of itself is not the precedent.

It is the reasons for reaching the decision in the case which constitutes the precedent.

Those reasons are known as the ratio decidendi . The ratio is the essential legal reasoning applied to arrive at the particular legal result in the case, with the legal consequences which follow.

It is the ratio decidendi that is followed, distinguished, applied or overruled in later judicial decisions.

The Making of a Legal Precedent

Legal principles in the common law legal systems become established as a result of the outcomes of broadly consistent or comparable cases.

A rule or principle of law may become established and binding:

• by consensus when judges at the same level of the judiciary frequently cite a previous decision in their own decision. The decision is cited by judges at the same level of court in different cases with similar facts. The frequency of the citation of the previous decision makes it a persuasive authority.
• authoritatively when a higher court in the judicial hierarchy - an appellate court - states what the law is, by reference to the ratio decidendi of their own decision in the case or the ratio a previous decision of a lower court.

In concept, it’s straightforward.

It can be tough to appreciate how the doctrine of precedent works without knowing how the legal process works.

Really tough. First and foremost, without deciding the facts of the case first, there can be no (correct) application of the correct precedent.

Facts of the Case

Courts are there to decide disputes. Those disputes are based on the facts of the case: what happened in the real world.

The law is not applied to alleged facts. Facts need to be proved.

Relevant facts need to be proved on the balance of probabilities . And that is done by with witness statements and where appropriate, affidavits.

Which Facts?

But it’s not any facts that must be proved to win a case to successfully defend a case, or:

• what a party might think the court should know about;
• how the party thinks they have been aggrieved or have been offended.

Much of that is usually irrelevant and at most peripheral to the facts that must be proved.

The Cause of Action

The centre point for and focus for winning a case are the facts required to be proved by the cause of action . A successful defence depends on proving that at least one of those required facts has not been proved by the claimant.

Without knowing what the decided facts are, there is not enough information to decide whether a specific precedent applies or not.

Deciding the Facts

So judges decide the facts of the case first, and then decide the legal principles.

Lord Diplock in Bahamas International Trust Co Ltd v Threadgold [1974] 3 All ER 881 said it like this:

It is for the judge to decide for himself what the law is, not to accept it from any or even all of the parties to the suit; having so decided it is his duty to apply it to the facts of the case . He would be acting contrary to his judicial oath if he were to determine the case by applying what the parties conceived to be the law, if in his own opinion it were erroneous.

Findings of fact are based on evidence. Another judge said that:

a judge is not bound always to make a finding one way or the other with regard to the facts averred by the parties. He has open to him the third alternative of saying that the party on whom the burden of proof lies in relation to any averment made by him has failed to discharge that burden.

Once the facts are decided on the evidence, the court is in a position to apply to apply the doctrine of precedent.

The application of the doctrine of precedent operates where the proved facts of current case are similar those decided in a previous case. That is, the instant case is on all fours with the facts in the precedent: ie the facts are strictly analogous. 

The doctrine of precedent works on two levels.

Application of Judicial Precedent at First Instance

The first is by the trial judge.

Having decided the facts of the case, the trial judge might:

• apply a case, which equates to following a binding precedent.
• follow a precedent (aka an authority). The judge follows the ratio decidendi of a previous case in the current case being decided. The judge might also decide to not follow a previous case.
• consider an authority. The case is considered for its ratio decidendi and/or obiter dicta , but is it not followed. When an authority is considered, the case reference is likely to be cited, so that readers of the judgment are able to follow the reasoning in the decision.
• distinguish an authority. The judge decides that a precedent does not apply to the current case.

There are many ways of distinguishing an authority. Some of the ways that might be done is covered further down. Then, simply because a decision is distinguished does not mean that it is bad in law. But then a bad case which is not consistent with the communis opinion doctorum (Latin: “learned common opinion.”) is likely to be distinguished out of existence. See what happened to Foakes v Beer below.

Decision-making process

Once the facts of the case are decided, trial judges usually look for authorities where the facts of the case are substantially the same on a conceptual level (ie are on all fours). If the facts are not sufficiently similar, the judge then finds previous cases with comparable facts and decides whether the legal reasoning in those cases can be compared and then applied to the case being decided.

If gaps exist in the decided law, legal reasoning is applied to arrive at a just result in the case.

If the law is going to be applied equally to all that come before courts, there really is no other sensible and coherent way to apply and develop the law.

Application of Judicial Precedent on Appeal

The second is on the appellate level.

Appellate courts have a different role to trial judges.

Appeal courts don’t rerun the trial and hear the case all over again. They don’t re-hear the cross-examination of witnesses, decide the facts all over again and hear the same arguments. That would be a hearing de novo .

That’s not what appeals are about.

Rather, appellate courts review the correctness of the decision of the court below.

In appeal courts, the function is not so much to ensure that the case decided below was decided in the way that the appeal court would have decided it.  It certainly isn’t to arrive at the same decision as the trial judge.

It’s more about whether the decision falls within a range of possible outcomes for the facts decided in the case in the court below.

It’s a review, and if necessary, a corrective process is applied. That may involve:

• applying the same law in the same precedent
• deciding that a different precedent should have been applied in the case
• applying a precedent from a higher court, or
• developing the law, and deciding the case on a different basis.

The more the superior the court, the wider they can make the ratio , so that it applies in a wider variety of circumstances. 

Powers of Appellate Courts

To enable the supervisory and review functions of appellate courts, courts of appeal have powers in addition to those of trial judges.

They have the power to:

• to consider, apply, follow or not follow and distinguish binding precedents, which are the same as a trial judge.
• approve a precedent (aka “an authority”)
• overrule a point of law in a previous decision made by a lower court. The overruled point of law will no longer be binding on lower courts. The appeal court might decide that the lower court was bound by an authority of a higher court, such as the Supreme Court, and should have applied that.
• affirm the judgment of the case being appealed, and the outcome of the case
• reverse a decision, in that the decision of the lower court is set aside by the court of appeal, and/or
• disapprove a previous precedent, but not overrule it in the current case.

Appeal courts do not exercise these powers readily or even easily. What’s more, permission is needed to have an appeal heard.

Appellate courts have other roles. They include:

• to provide guidance to lower courts
• provide discussion in decisions so that lower courts are better able to fit the ratio decidendi into a context as part of deciding future cases. Factors might be raised and considered, but then ruled out of forming part of the ratio decidendi .

Multiple Appellate Courts

In civil justice systems, there is usually more than one appellate court that litigants can appeal the decision made at trial. Decisions of appeal courts may be able to be appealed further to a higher court.

The supremacy of the court increases the further you work up the court hierarchy.

For instance, in a case which starts in the High Court:

• the trial judge is the High Court Judge sitting alone and first hears and decides the case at the trial
• the Court of Appeal hears the appeal; and
• the Supreme Court hears the appeal from the Court of Appeal.

An appeal from the decision of a Master of the High Court usually lies to a High Court Judge.

If a trial takes place in the County Court, then the appeal path would usually be:

• an appeal is to a single High Court Judge, who sits as the court of appeal;
• the next appeal court would be the Court of Appeal, and then
• the Supreme Court .

At each stage, permission is required to appeal a decision before a court will review the decision. In England, it’s known as permission to appeal.

Then there are leapfrog cases.

Leapfrog cases

After a trial in civil proceedings before a single High Court Judge, an appeal may be made permitted directly to the Supreme Court.  It’s known as a leapfrog appeal, or a leapfrog case.

Sections 12 to 19 of The Administration of Justice Act 1969 enable leapfrog appeals where “a point of law of general public importance is involved in that decision and that that point of law” has been fully argued at the trial and involves:

• the legal interpretation of a statute or of a statutory instrument, or
• a decision where the trial judge is bound by a decision of the Court of Appeal or the Supreme Court

It may also be a decision of general public importance and the result of the case is so significant that consideration by the Supreme Court outweigh the benefits of consideration by the Court of Appeal. The leapfrog appeal must be justified in the opinion of the High Court judge and the Supreme Court.

That’s quite a high bar to satisfy. Similarly, under section 57 of the Access to Justice Act 1999 appeals may be available from decisions made in the County Court directly to the Court of Appeal.

Binding Precedents

Binding precedents apply to a case when the facts of the case are on all fours with those in a precedent from a superior court. 

A binding precedent is not a persuasive authority. That is something different. It’s an authoritative precedent which is the binding precedent.

When the facts are similar in the case being heard and decided, there need to be unusual features to enable the case to fall outside the existing legal precedent. If it doesn’t, the judge has no choice but to apply the binding precedent.

Binding precedents require the judge to decide a case in a particular way, consistent with the authority. That may be contrary to the judge’s own belief; they may not agree with it on a personal level.  But they have no choice but to apply the authority.

When that happens, the judge in their reasons for the decision might say that the law is somehow unsuitable or out of date. That may be accompanied by a desire for the decision to be reversed on appeal.

Stare decisis means - theoretically - that any inferior court is bound by the point of law decided.

But then a decision might be distinguished.

Persuasive Authorities

A persuasive precedent has no true binding force.

It’s a case which the court might rely on to reach its own decision, but is not bound to do so.

That is especially the case when the facts of the instant case differ from those in a cited precedent.

Another persuasive precedent (or an authoritative precedent) may apply when it is more appropriate for the facts of the case.

A precedent may be persuasive for a number of reasons.

• The case receives frequent judicial endorsement by judges at the same level of court in the judicial hierarchy. After the precedent is made, other judges rely on it in subsequent decision. As those instances increase, the decision becomes entrenched. That’s communis opinion doctorum in action. For example, a High Court Judge sitting alone is not bound by other first instance decisions of High Court Judges. Those judgments however will be persuasive when the factual matrix is close enough to the current case. And then decisions of County Court Judges don’t have precedential value. Their decisions carry no weight for the purposes of making binding precedents or persuasive precedents
• Similar cases have been decided by courts in foreign countries. That’s one of the strengths of the common law system. English Courts might cite decisions from Australia, New Zealand, the United States or elsewhere. The decided principles in those countries are adopted to English law and incorporated into English law. Decisions of the Court of Appeal in Northern Ireland are especially persuasive, and given a special status. That’s because inconsistencies between the law of England and Northern Ireland is seen as undesirable
• Decisions of the Privy Council are persuasive but not binding on courts in England. The Privy Council is the final court of appeal from countries that have retained the right of appeal to His Majesty in Council or to the Judicial Committee of the Privy Counsel, UK overseas territories and Crown dependencies
• The content of textbooks authored by academics also may form the basis of a decision. For instance, Lord Burrows (formerly Professor Burrows) was an academic prior to his appointment to the Supreme Court in 2020. Professor Wadlow is often cited in respect of the law of passing off in English Courts.
• Judges may write or give speeches ex-judicially. As an example, see the speech of Lord Sales of the Supreme Court below.
• Obiter dicta in a decision made in a higher court may guide the decision for the current case. For example, dicta in cases in an appellate court might provide guidance for the case being heard in the lower court in the future

Persuasive precedents are also able to be considered and rejected, such as in Marks and Spencer plc v BNP Paribas Securities Services Trust Company (Jersey) Ltd [2015] UKSC 72 at paragraph 31.  

Persuasive Weight

Then, persuasive precedents vary in the weight (or persuasive value) that other courts will attribute to them. This is especially important when assessing how a court might evaluate comments made obiter .

Qualitative factors to that assessment include:

• The status level of the court in the judicial hierarchy.
• The seniority of the judge (or judges) which decided the case. For instance, the decisions of Lord Hoffman almost always carry significant weight. When he was sitting, he was one of the best judges of the last 100 years
• Whether the decision was made as part of an application or at the trial. Even decisions made after an interim application can carry weight. For example, when a party is all the more vindicated when they successfully resist an application for summary judgment and then succeed at the trial

Distinguishing a Case

The law is not static. It develops over time.

Some legal principles are so fundamental that they are deeply entrenched in English law.

It’s well known than judges have a good deal of manoeuvrability in the way case law is applied.

Prior cases may be found to lead to undesirable results or injustice. When that happens, previous cases are prone to be distinguished.

The common law evolves by calling out material factual differences with prior cases and adjudicating whether the legal rule in question should apply in the case. When that happens, the reasons should be explained in the judgment. If not, it’s more likely to be appealed.

Then, new cases raise new factual circumstances. Those factual circumstances will not always be dealt with adequately by existing authorities.

When new technologies are developed and used as a new method of doing business, there may be no cases decided that match the type of facts or events which take place in a new business model.

Unsuitable Law

It may be that the law is unsuitable for the findings of fact that come before a court.

When cases are Distinguished

A judge might distinguish the current case being considered and avoid applying a precedent in a variety of ways.

• are sufficiently different to the previous case. The authorities are distinguishable because the material facts contained in the authority are missing or incomplete
• mean the earlier decision inapplicable to the case being decided
• would lead to an unjust result. That’s because courts of law are also courts of justice.
• relies on standards in society which have changed since the authority was originally decided:
• is properly thought to be a persuasive authority only
• places significant weight on the existence of one type of fact or a combination of facts, and that weight does not exist in the current case

When a binding precedent case is consistently distinguished and not followed, over time the legal principle or decision is likely to be distinguished out of existence. It may never be formally overruled. Its authoritative weight ends up being consigned to history.

Overruling Precedents

Over time, most case law is superseded and replaced new precedents.

When a precedent is overruled, it’s no longer good law. It can also be said to be bad in law.

A common forerunner to that is that judges disapprove, distinguish and/or no longer follow the authority.

Alternately, the decision may be simply overruled by a court of appeal. That creates a new binding authority for inferior courts, overnight.

However, the judiciary has a general reluctance to overrule longstanding existing precedents. The older it is, the more likely a greater part of relevant parts of society have relied on it.

Age and Weight

That reluctance can diminish with the age and weight of the decision:

• The decision is only a few years old and is not consistent with the current orthodoxy
• The decision has only been applied at first instance (at a trial) and not been decided by the Court of Appeal
• It would cause an upheaval in the operation of the law. Courts recognise that businesses manage their affairs by reference to the law. Shifts in legal principles can cause serious - serious - problems. But it happens. For example, litigation funders did not see the decision of the Supreme Court in the Paccar Case It upended (and shook) practically the entire litigation funding industry.

Per Incuriam Decisions

The case was decided per incuriam (Latin: through a want of care), when there was some oversight in the law result in an error by the court.

• a binding authority was not cited, or failed to be taken into account
• a statute which would have changed the result of the case was not cited.
• the judge was misinformed of the law.

A judgment per incuriam is not binding and has no authority under the principle of stare decisis.

Law has moved on

Long standing decisions are more likely to be overturned when it produces unjust results or work unsatisfactorily in the marketplace or society generally. It has fallen out of date.

Speaking ex-judicially, Lord Sales, a Supreme Court Judge made these points in Default Rules in the Common Law: Substantive Rules and Precedent (2023):

• A previous decision is more likely to be overruled if it impedes the development of the law, causes uncertainty, administrative difficulties or individual injustice
• A court is more likely to not to overrule a previous decision where it “risks retrospectively disturbing the basis on which contracts and other commercial transactions have been entered into”. That would cause a serious disturbance to an otherwise stable legal environment
• The change in law is complex or has wider social consequences. Those sorts of changes are more of the role for the legislature - Parliament - rather than courts

Courts Overturning their own Decisions

Courts generally have power to overturn their own decisions, provided they’re not bound by a decision of a higher court.

Court of Appeal

The Court of Appeal bound its own previous decisions in any case with substantially similar facts.

There are cases where the Court of Appeal is not bound to its own previous decisions, which include when:

• the Court of Appeal has made conflicting decisions . The Court of Appeal decides which of its previous conflicting decisions it will follow
• the Supreme Court as decided a case which conflicts with a previous decision of the Court of Appeal, even when it is not expressly overruled
• where the previous decision was per incuriam .

Supreme Court

The Supreme Court has power to disapprove its own decisions where:

too rigid adherence to precedent may lead to injustice in a particular case and also unduly restrict the proper development of the law: Practice Statement (Judicial Precedent) [1966] 1 WLR 1234 .

The default rule however favours the status quo . The court will be “very circumspect” before accepting an invitation to invoke the Practice Statement.

General Principles of the Doctrine of Precedent

The sometimes not so obvious background to the operation of the doctrine of precedent includes:

• A case is only authority for what it actually decides: Lord Halsbury in LC Quinn v Leathem [1901] UKHL 2 A case is not authority for corollaries which may seem apparent or logical
• The application of a precedent is limited to the particular facts of the case on which it was decided
• If the decision made at trial is wrong and never appealed or can’t be appealed, it remains wrongly decided. Courts hearing subsequent cases will realise that the case has no value and not follow it
• The age of an authority will tend to weaken its authority, unless it is affirmed and followed in more recent cases
• Legal principles do not necessarily need to be stated in the decision. It is enough that the rule or principle underlies the decision and forms part of the ratio decidendi . For example, see the decision in Topalsson GmbH v Rolls-Royce Motor Cars Ltd [2023] EWHC 1765 . In that case, Mrs Justice O’Farrell DBE cited authorities in 8 paragraphs of a 400 paragraph judgment. Paragraph 170 cited the relevant law and moves on, quickly, because the relevant law is so well known.
• The law as stated in binding precedents are applied to cases at the time judgment is handed down by the court. For example, the law that applies to a contract is the law as it exists when judgment is handed down. Courts are not restricted to what the law was thought to be at the date of the contract if it has been subsequently developed. Likewise, the law that applies in copyright cases is the law as it exists when judgment is handed down.

The doctrine of precedent does not apply to obiter dicta , but dicta can have persuasive weight.

Obiter Dicta and Precedents

Obiter dicta stands in contrast to ratio decidendi . It is sometimes abbreviated to obiter or dictum (the singular), and is “a statement by the way”.

Comments made obiter do not form part of the reasoning for the decision. It will not constitute part of the precedent and is not binding.

Obiter dicta includes:

• observations by a judge on a legal question in case before them, but not arising in a so as to require the point of law to be decided. The comments and observations may however carry great weight. But they are not conclusive authority.
• set the context for the current case being decided
• set out the movement of the law to explain the rationale for a new authority
• explain the scope of the ratio , and set out its boundaries of application
• comments which may set the law in a new direction in future cases

An opinion not necessary to a judgment and therefore not binding as a precedent. it does form part of the ratio . They’re comments which are not necessary for the purpose of deciding the case on its facts.

Advantages of Doctrine of Precedent

The advantages of the doctrine of precedent include:

• the desirability of promoting certainty in the application of the law
• consistency of judicial decision-making. It establishes a common and stable approach to solving legal problems
• the stable and predictable system of law which results from consistency of outcomes from similar cases. It results in stable societies. The antithesis to the doctrine of precedent results in what is sometimes referred to as palm-tree justice and banana republics
• appellate courts provide practical guidance to judges in courts lower in the legal system for future cases. That guidance can smooth the road in advance of future decisions by other judges
• judges are more able to assign legal weight to incommensurable values.
• the predictability accommodates foreign investment and the certainty incentivises international trade and commerce
• members of society have a better idea of where they stand viz-a-viz the law

It also means that members of society are able to take legal advice and obtain a view on whether it is likely to be worthwhile (or not) going to court to enforce a legal right.

More on that below.

Disadvantages and Limitations of Stare Decisis

The disadvantages and limitations of the doctrine of precedent include:

• The benefit of predictability of the law comes at the price of speed. The law moves slowly forward. Changes to the common law are incremental. It may take several cases for the law to move to a satisfactory state. That can be mitigated by enactment of legislation by Parliament.  But that does not cure the problem.
• It takes cases to come before courts for judge-made law to develop. Firstly, cases need to come before courts. Cases attract legal fees for the skillsets of the solicitors and barristers to conduct those cases. It’s expensive. That means those disenfranchised by the law do not have access to justice. That unsatisfactory position has been publicly recognised and acknowledged for over well over 30 years.

That leads to the proposition that legal rights are more in the nature of privileges to be paid for.

• When areas of common law mature, the state of the law on any particular factual matrix can breakdown through the complexity of the law. It becomes more difficult to ascertain the state of the law which would apply to the specific set of facts. Predictability of the likely outcome becomes more uncertain. Also, those cases tend to be more time consuming and as a result, more expensive.
• Over time, the greater the detail in the relevant area of law, the more specialist the legal advice, which leads to greater costs for legal advice. The law moves out of reach of the majority of member of the public (and it has). That in turn leads to an absence of access to justice. The courts themselves since at least 2001 imposed a limit to the number of citations and authorities which may be referred to courts at hearings. See Practice Direction (Judgments: Form and Citation) 2001 1 WLR 194. The current requirements are set out in Practice Direction: Citation of Authorities (2012).
• requires dedicated and specialist legal information service providers
• leads to expense for the legal professionals to stay up to date with the law
• leads to greater legal costs for litigants
• leads to a lessening of access to justice
• ultimately means, in terms of its legal effect, and
• whether it was lawfully made.

Judicial Precedent Examples

Judicial decisions state substantive rules of law.  

The judicial reasoning is usually drawn out in the original precedent. Then the precedent is cited in later cases by name, usually with a paragraph reference, an extract, direct quotation, or summary of the legal principle.

What follows are the well-established principles of English law.

Law of Damages

One of the oldest frequently cited cases in the law of damages is Robinson v Harmen [1848] EngR 135 .  It is one of the seminal authorities on damages.

It sets up the underlying basis for the calculation of awards of damages in contract and tort cases.

The ratio decidendi runs like this:

The rule of the common law is, that where a party sustains a loss by reason of a breach of contract, he is, so far as money can do it, to be placed in the same situation, with respect of damages, as if the contract had been performed.

The facts of the case are no longer relevant.

It has been reaffirmed frequently since it was made and constantly cited in cases in damages cases. The House of Lords reaffirmed it in The Golden Strait Corporation v Nippon Yusen Kubishika Kaisha [2007] UKHL 12 (paragraph 29) and it "remains as valid now as it was then” .

Likewise, the Supreme Court reaffirmed the principle in Morris-Garner v One Step (Support) Ltd [2018] UKSC 20 at paragraph 35.

Precedent: The Rule in Foakes v Beer

And then there’s Foakes v Beer from 1884. It is the judicial precedent for the proposition that payment of a lesser sum of money (of itself) is not accord and satisfaction for discharge a larger debt. Some further consideration is required, such as “ the gift of a horse, hawk, or robe” to do so.

In the case, the House of Lords approved an even older case, the original precedent, Pinnel's Case (1602) 5 Co Rep 117.

In Rock Advertising Ltd v MWB Business Exchange Centres Ltd [2018] UKSC 24 (paragraph 18), the Supreme Court refused to overrule Foakes v Beer . That was despite it having a history of being distinguished.  Despite its age and criticism, it is still applied. See for example, in Integral Petroleum SA & Ors v Dun [2022] EWHC 659 .

Contract Law

And then for an account of how different judges in superior courts influence the judicial precedents (and how contracts are interpreted), see paragraph 17 of Contractual Interpretation: Do judges sometimes say one thing and do another? by the Chancellor of the High Court, Sir Geoffrey Vos.

The Doctrine the Precedent and Legal Advice

Another advantage of the doctrine is precedent is that members of society are able to obtain legal advice to forecast the outcome of their own cases. They can get an idea whether they’re likely to be successful, or not, in the context of legal claims.

If you’ve read what appears above, you’ll know that there is no basis for the doctrine of precedent to operate or be applied without the facts of the case.

The process of providing legal advice roughly follows the same process that courts follow.

The facts come first, and then the law is applied to those facts.

The Process of Litigation

In disputes, parties make allegations of fact which conflict with one another. Both versions of the facts can’t right.

Also, a defendant might say that on an agreed set of facts, they are not liable.

That’s part and parcel of the process of resolution of disputes.

Judges decide the facts, and they do in in a specific way.

How Facts are Decided

To apply the doctrine of precedent, judges must first decide the facts of the case, and then apply the law to those facts to arrive at the result of the case.

Facts are decided by judges after the trial.

That is, after:

• disclosure has been given
• witnesses in the case have filed and served their witness statements
• the witnesses have been cross-examined on their witness statements (ie the evidence) in open court
• each party has told the judge why their version of the facts - based on the evidence - should be preferred and not their opponents.

After the trial, the judge will assess the arguments, all of the evidence presented against the credibility of the witnesses.

Then, the judge will decide which facts are proven, and facts which are not proven. The evidence will prove – or disprove – each alleged fact.

And with that, the judge has arrived at the factual matrix of the case.

It is at this point that the judge applies relevant precedents (ie the law) to the facts decided.

Personal Guarantee Cases

Contracts of guarantee are designed to establish liability for repayment of a loan by a guarantor.

Here’s a typical case where a creditor looks to enforce a personal guarantee:

A business enters into a loan agreement. The lender/creditor obtains a personal guarantee for security to repay the loan. Money is advanced by the lender to the borrower. The loan is not repaid.

If the contract is going to say anything, it will say that the loan needs to be repaid. It’s unlikely that a contract will release a guarantor of liability without the debt being paid.

That would defeat the purpose of having a personal guarantee.

Terms of the Personal Guarantee

One can’t tell whether the guarantor can escape liability from the contract alone.

That would be looking that the legally binding obligations in a vacuum. The law does not operate in a vacuum. The law applies to facts.

One of the ways to get out of a contract of guarantee depends on the behaviour of the lender and its agents.

The fact that a guarantor was misled prior to agreeing to the guarantee will not usually appear in a contract. When a guarantor is misled, it happens in the real world: where the background facts happened.

If the client say they were misled, the facts (and the evidence to prove it) surrounding that event need to be identified.

Once they are identified, a series of questions can be raised in the context of those facts and the terms of the contract of guarantee to see if it might be set aside.

They might include questions being asked, such as:

On the facts of this case, has there been a misrepresentation by the creditor and/or its agent before the guarantee was signed? If so, what is the evidence avaiable to prove the misrepresentation was made before the guarantee was signed? Was the misrepresentation relied on by the guarantor? What is the evidence available to prove that the misrepresentation was made at that point in time, to substantiate the claim?

Without the facts, and the evidence to prove those alleged facts, it is anyone’s guess whether the personal guarantee can be set aside.

If the alleged facts establish and match a legal principle in a precedent which justifies a guarantor getting out of the contract, the precedent is a candidate to be relied on to get out of the contract.

Passing Off Cases

The law of passing off is also known as the law of unregistered trade marks.

Here’s a typical case of passing off:

A business creates its logo and branding. It starts trading. It uses the logo and branding to promote its business. It trades for 3 months. It then discovers that another business has copied its business name, and wants to stop it.

The authoritative precedent usually cited by judges to recite the requirements of passing off is Reckitt & Colman Products Ltd v Borden Inc [1990] 1 WLR 491 . That precedent says the 3 elements which must be proved to win a passing off case are:

goodwill; misrepresentation; and a likelihood of damage.

If the claimant proves facts which show those elements, they win. If the claimant can’t prove those facts, it can’t succeed.

So the claimant must prove the facts which establish goodwill: that’s relevant trading activity.

If the business cannot show a solicitor evidence that goodwill exists, the solicitor will almost certainly advise the business that a passing off case can’t succeed.

The Legal Advice

That’s an application of the law of precedent as part of legal advice. One of the elements of the tort of passing off can’t be satisfied. The legal advice would be that the client cannot succeed to prevent the competing business using the law of passing off.

Contracts: Can I get out of this contract?

Contracts are legally binding agreements.

Whether a party can get out of a contract early depends on what the contract says, and what has happened since the contract was agreed.

Just by looking at the contract, you can’t tell what happened in the real world. That would be reading the contract in a factual vacuum.

Common Questions

Common examples of questions involved in breach of contract cases include:

Does the contract remain in force or has it already ended? What has the other party done that might justify terminating the contract? If they have done something, what did my client do in response, if anything? What is the state of affairs now? That might include asking questions such as: Have the parties varied the terms of the contract since it was made? Has either party waived their rights or affirmed the contract?

It’s the events that happened in the real world that will tell you whether there are grounds to terminate a contract. Without knowing what happened in the real world, you can’t tell whether the party can escape the contract. 

That’s an application of the process to obtain facts which are relevant to establish the prerequisites of a precedent which could apply in the case.

The doctrine of precedent applies the law to the facts of the case. It applies to the terms of the contract which was agreed, as much as how the parties have performed since it was agreed.

The application of the law of precedent serves to minimise instances of courts deciding the same legal points again..

The application of the law in subsequent decisions are less likely to vary with judges’ proclivities, private sentiments or subjective views from case to case.

It is also a peer review system. Legal decisions are published. Other judges review them for their own cases. Decisions of courts are reviewable by higher courts.

Later decisions identify relevant cases and are applied when they need to be. They may be distinguish when they are unsuitable to be applied to the facts of the case or lead to unfair result.

The value of the doctrine of precedent is that the rule of law is more dependable in its certainty and predictability.

Coherent judicial reasoning would not be possible without it. The law is applied more predictably than what would other be the case.

Courts higher in the court hierarchy are not tied to precedents in the same way as trial judges.

Judges at all levels may not follow precedents in a particular case where injustice would be caused. Courts at first instance have far less discretion to do so.

Business Solicitors

Finding and interpreting the law is a fundamental part of the justice systems. The facts and the evidence that support it make or break every case. Findings of fact rely on the court accepting the evidence presented by parties to prove the version of events claimed to win court cases.

For business legal advice on applicable precedents in business disputes , making out your case and make the most of the evidence to prove your case, reach us at [email protected] .

• Terms of Use
• Privacy Statement
• Cookie Policy

Legal Advice

• Software Legal Advice
• Business Contracts
• Intellectual Property Advice

Contact Information

89 Fleet Street London EC4Y 1DH

[email protected]

+44 20 7036 9282

• Help and information
• Comparative
• Constitutional & Administrative
• Criminal Justice
• Criminology
• Environment
• Equity & Trusts
• Competition
• Human Rights & Immigration
• Intellectual Property
• International Criminal
• International Environmental
• Private International
• Public International
• IT & Communications
• Jurisprudence & Philosophy of Law
• Legal Practice Course
• English Legal System (ELS)
• Legal Skills & Practice
• Medical & Healthcare
• Study & Revision
• Business and Government
• Share Facebook LinkedIn Twitter

English Legal System (5th edn)

• New to the edition
• Guide to the features in English Legal System
• Table of cases
• Table of legislation
• 1. Studying the English legal system
• 2. An overview of the English legal system
• 3. Legislation and the law-making process
• 4. The interpretation of statutes
• 5. The doctrine of judicial precedent
• 6. The law and institutions of the European Union (EU)
• 7. Human rights in the United Kingdom
• 8. The judiciary
• 9. The legal profession
• 10. Access to justice
• 11. The criminal process: The suspect and the police
• 12. The criminal process: Pre-trial and trial
• 13. The jury
• 14. Sentencing
• 15. The civil process
• 16. Alternative dispute resolution
• 17. Tribunals
• 18. Criminal and civil appeals

p. 165 5. The doctrine of judicial precedent

• Helen Rutherford , Helen Rutherford Senior Lecturer, Northumbria University
• Birju Kotecha Birju Kotecha Senior Lecturer, Northumbria University
• , and  Angela MacFarlane Angela MacFarlane Senior Lecturer, Northumbria University
• https://doi.org/10.1093/he/9780192858856.003.0005
• Published in print: 13 May 2022
• Published online: September 2022

This chapter considers an essential source of law in the English legal system: judicial precedent (or ‘case law’). The rules and principles of the doctrine of judicial precedent are explored, including how precedents are created, developed, and followed. The chapter analyses the rule that forms the precedent—the ratio decidendi , or the reason for the decision—as well as the importance of other judicial statements that do not form part of those reasons—the obiter dicta . The principle of binding precedent is captured by the expression ‘ stare decisis ’ (stand by what is decided) and binding precedent relies on a hierarchy of courts. The hierarchy can help to establish whether a particular ratio decidendi binds a particular court and whether an appellate court is bound by its own previous precedents. The chapter is packed with case law examples and highlights the role of non-binding precedent which may still be deemed persuasive for a particular court. The relationship between the English courts and the Court of Justice of the European Union (EU) and the European Court of Human Rights (ECtHR) is considered. Finally, the chapter considers how a court may avoid following a particular precedent by the process of overruling, distinguishing, or reversing.

• judge-made law
• judicial precedent
• ratio decidendi
• obiter dicta
• binding precedent
• stare decisis
• distinguishing

You do not currently have access to this chapter

Please sign in to access the full content.

Access to the full content requires a subscription

Printed from Oxford Law Trove. Under the terms of the licence agreement, an individual user may print out a single article for personal use (for details see Privacy Policy and Legal Notice).

date: 28 September 2024

• Cookie Policy
• Privacy Policy
• Legal Notice
• Accessibility
• [185.66.14.133]
• 185.66.14.133

Characters remaining 500 /500

Judicial Precedent in Judicial Practice

Books Every Lawyer Should Have in Their Library

Sexual Offences Prevention Orders

The doctrine of judicial precedent has been at the heart of the English legal system, being a fashion fiesta for the judges to follow long-standing precedents which, in fact, only slowly evolved and matured from the nineteenth century.[1] Fundamentally, it refers to the hierarchical structure of the English courts within which a decision of a higher court will be binding on a court lower in the hierarchy. The decision contains two types of statements: first, there are ‘ratio decidendi’ statements which are binding and are statements of law applied, and second, there ‘obiter dictum’ statements which are statements made ‘by the way’. Nevertheless, some judges have been criticised for making law via the doctrine of precedent. Subsequently, the question of judges being unconstitutional arises. However, as we shall see, sometimes judges sitting in the House of Lords (the Supreme Court), Privy Council, Court of Appeal and two separate European Courts do indeed make law and state the law.

The Supreme Court, formerly the House of Lords, is the highest court in the United Kingdom and is bound by the European Court of Justice for European Union law matters and also by the European Court of Human Rights for issues related to human rights. The House of Lords was also bound by its own decisions as per London Tramways v London County Council [1898] AC 375 [2] until Lord Gardiner in the House of Lords issued a Practice Statement [3] in 1966 which allowed only them to depart from the previous decisions as a last resort to do justice. The Practice Statement was used successfully in Conway v Rimmer [1968] AC 910 [4] when they overruled Duncan v Cammell Laird and company [1942] AC 624. [5] The circumstances of the case were such that it would be an injustice to apply the 1942 precedent despite the fact that the earlier case had very similar material facts. However, crucially, the decision taken in 1942 was affected by the war period but in 1968 there was no war. The Practice Statement 1966 highlights our topic by illustrating how judicial precedent was refined thorough practice. However, the House of Lords made it clear by stating, ‘apart altogether from our recent liberation from some of the chains of precedent, which for my part I think should only be exercised rarely and sparingly’.[6]

The Privy Council, being the Commonwealth court, could also provide a good illustration of judicial precedent being changed by the courts as they wish. In 1997, the Privy Council received a case from Hong Kong cited Luc Thiet Thuan v R [1997] AC 131 (PC) [7] in which the old precedent set in R v Camplin [1978] AC 705 [8] was followed. This concerned the need for the jury to be convinced that provocation was enough to make a reasonable man do what the defendant did.  However, later the House of Lords in R v Smith (Morgan) [2001] 1 AC 146 [9] overruled Luc Thiet Thuan because they felt that the jury should take into account all the characteristics of the defendant. Then in 2005, the Privy Council received a case from Jersey cited Jersey v Holley [2005] 3 AER 371 [10] and, despite being bound by the House of Lord’s precedent set in Morgan-Smith , they in fact they followed their own precedent which was Luc Thiet Thuan . An interesting fact was the Law Lords which sat in the House of Lords also sat in the Privy Council.[11] The area of law on provocation and murder needed clarification and subsequently, the precedent was changed back to the original.[12] Thus, this illustrates that judges change precedent as they wish, like in Jersey v Holley , without waiting for a case to come up in the House of Lords and then rectify the mistake made in Morgan-Smith .

The Court of Appeal, being the second most senior court in the UK, has also played an important role in shaping the doctrine of precedent.  Originally, it was bound by its own decisions and also of the House of Lords and the two European courts. However, in the Civil Division of the court, the lordships ruled that the court may depart from their own precedents if the situation complies with any of the three exceptions laid down in Young v Bristol Aeroplane Company limited [1944] 1 KB 718 [13]. The three exceptions [14] in which precedents might be departed from at the discretion of the court are when an earlier decision given was in ignorance of some authority like a statue or even an old precedent, [15] and we saw application of this exception in Williams v Fawcett [1985] 1 All ER 787. [16] The second exception is where there is a conflict between two earlier decisions of the Court of Appeal. In this instance, the later court must decide which to follow; Tiverton Estates Limited v Wearwell [1974] All E.R. 209. [17] The last exception is where a previous decision of the Court of Appeal is in conflict with a later House of Lords decision. Here, the decision of the House of Lords must be followed; Family Housing Association v Jones [1990] 1 W.L.R. 779, C.A. [18]

The Criminal Division also follows the same principles but keeps the liberty of the subjects as a key priority. A great number of new legislation on criminal justice has prompted the Court to create new precedents because they cannot apply old precedents if they are inconsistent with the new legislation, as parliament is sovereign.  This example is consistent with our topic of courts creating precedent through practice.

The European Court of Justice (ECJ) was created by the Treaty of Rome and has been in force since 1952 in the UK. Its decisions in some instances have affected the English courts when trying to follow their establish precedents. Article 267 of the Treaty of Lisbon [19] states that if there is ambiguity with relation to European Union law, they should refer back to their community court in Luxembourg. When the ECJ rules on a specific EU law question then the national courts apply the ruling to the case before them. A popular authority here would be R v Sectretary of State for Transport (ex parte Factortame) No 2 [1991] 1 AC 603 [20] in which the House of Lords changed their decision and gave judgment in light of the ECJ’s ruling. The question of when to refer to the ECJ was raised which concluded with Lord Denning’s implication [21] which was replaced by Lord Bingham,[22] stating that not only when the question referred was conclusive of the case but whenever there is doubt over a legal issue concerning European Union law.

The preliminary ruling procedure was used in a long case of M H Marshall v Southampton and South-West Hampshire Area Health Authority (Teaching) [1986] EUECJ R-152/84 [23] where a lady was discriminated against when terminating her contract. She claimed damages, but the national law had set a limit on the amount of damages claimable which was contrary to European Union law. The ECJ ruled that there should be no upper limit on compensation for damages through preliminary ruling with the help of Article 267.[24]

Rights were brought home through the Human Rights Act (HRA) 1998 [25] and the rights also brought with them the articulation of the European Convention on Human Rights (ECHR) law. According to Section 2(1)(a) of the HRA, national courts must ensure that their decisions take into account any judgments, decisions, declarations or advisory opinions of the ECHR. The phrase ‘take into account’ lead to a series of arguments questioning whether it may or may not mean that the domestic courts always have to follow the ECHR. Subsequently, judges had to ensure that any precedent relied upon is compatible with the rights gifted under the Convention. A landmark authority which changed the old principle that the Court of Appeal is bound by the House of Lords is Kay and others v Lambeth Council [2006] UKHL 10.[26]

Lord Bingham offered some guidelines as to when a court could depart from the House of Lords and should follow a case of the ECHR: when an ECHR case has been decided after the House of Lords case, or when there was an incompatibility between convention law and domestic law and that this is not based on the Act of Parliament. Recent cases like Secretary of State for the Home Department v AF & Another and one other action [2009] UKHL 28 have illustrated that the doctrine of precedent is being taken over by the court in Strasbourg.  In this case, Lord Rodger’s obiter dicta contained comments like ‘in reality, we have no choice: Argentoratum locutum, iudicium finitum – Strasbourg has spoken, the case is closed’.[27] From his words, we can see that judicial precedents may have exceptions and they are best discovered through practice especially when judges face situations like these.

We have seen with reference to the above Courts that judges are creating precedents in the course of practice by saying that they are simply following the doctrine of precedent. Judges usually avoid the doctrine impliedly by using three methods. Overruling is used by the higher courts in the hierarchy which sets aside a legal ruling established in an earlier different case in the lower court. Higher courts, when hearing an appeal case, can used the reversing method. This is where the superior court reverses the decision of the case ruled in the lower court. Last but not the least, distinguishing is one of the judges’ favourite methods. This allows them to they simply distinguish on the facts of the case and say that the precedent cannot be applied because the material facts differ.[28] Sometimes judges use multiple precedents to create a single new precedent.

Nonetheless, understanding judicial precedent by referring to how judges follow in the course of practice is the key way to master it. Lord Denning tried to change judicial precedent by asking to amend the rules through his campaign rather than doing it tactfully through practice. Lord Denning wanted to apply the House of Lords’ Practice Statement 1966 to the Court of Appeal. However, Lord Diplock counter-argued by saying that certainty and predictability must prevail. This research essay could not end without a famous quote from Lord Denning in Davis v Johnson :

A rule as to precedent (which any court lays down for itself) is not a rule of law at all. It is simply a practice or usage laid down by the court itself for its own guidance: and, as such, the successors of that court can alter that practice or amend it or set up other guide lines, just as the House of Lords did in 1966.[29]

In a nutshell, judicial precedent has always been a core principle in our English legal system. Judges have been following it, and for one to understand judicial precedent, merely looking at the set of rules will not help. As we have seen with reference to the above courts, where a judge wants to refine or change the precedent; there are means of doing so, but not openly like Lord Denning’s campaign but rather through practice either by distinguishing on facts or in exceptional situations where above authorities allow them to override the precedent. Nevertheless, judges refine or change precedent which is not adverse at all because, ultimately, they want justice to be provided via the doctrine of precedent.

Mahesh Nanwani is a penultimate year student reading Law at the University of London.

[1] Smith, C. (1820-1914). The English Legal System. The Oxford History of the Laws of England, XI . [2] London Tramways Co. v London City Council [1898] AC 375 [3] Practice Statement [1966] 3 All ER 77 [4] Conway v Rimmer [1968] AC 910 [5] Duncan v Cammell Laird and Co. Ltd [1942] AC 624 [6] Lord Upjohn in Conway v Rimmer [1968] AC 910 [7] Luc Thiet Thuan v R [1997] AC 131 (PC) [8] R v Camplin [1978] AC 705 [9] R v Smith (Morgan) [2001] 1 AC 146 [10] Jersey v Holley [2005]  3 AER 371 [11] Privy Council overrules Lords to put judgment back on track, Clare Dyer, legal editor, The Guardian , Monday 30 January, 2006 [12] Ibid. [13] Young v Bristol Aeroplane Company Limited , [1944] 1 KB 718 [14] Gooderson, R. N. (1950). The Rule in Young v. Bristol Aeroplane Co., Ltd. The Cambridge Law Journal, 10 (03). [15] Decision per Incuriam [16] Williams v Fawcett [1985] 1 All ER 787 [17] Tiverton Estates Limited v Wearwell [1974] All E.R. 209 [18] Family Housing Association v Jones [1990] 1 W.L.R. 779, C.A [19] Treaty of Lisbon, European Community (Now Referred as European Union) [20] R v Sectretary of State for Transport (ex parte Factortame) No 2 [1991] 1 AC 603 [21] Lord Denning in Bulmer v Bollinger [1974] [22] Lord Bingham in International Stock Exchange Limited Ex parte Else [1982] [23] M. H. Marshall v Southampton and South-West Hampshire Area Health Authority (Teaching) . [1986] EUECJ R-152/84 [24] Law Report: No limit to discrimination award: Marshall v Southampton and South West Hampshire Area Health Authority. European Court of Justice, Luxembourg, 2 August 1993. London: The independent . (Found on The Independent ’s website Archives) [25] Human Rights Act 1998 [26] Kay and others v Lambeth London Borough Council [2006] UKHL 10 [27] Secretary of State for the Home Department v AF & Another and one other action   [2009] UKHL 28 [28] Distinguishing facts to depart from Precedent, Quin v Leatham [1901] AC 495 [29] Lord Denning in Davis v Johnson (1979) AC 264

Related posts

Does Artificial Intelligence have greater benefits than limitations in Law?

How to avoid allegations of ‘washing’ one’s image.

Leave a reply cancel reply.

Your email address will not be published. Required fields are marked *

Save my name, email, and website in this browser for the next time I comment.

The Student Lawyer has assembled a team of qualified lawyers to provide careers assistance to law students applying for training contracts and vacation schemes.     

CV & Cover Letter Reviews

Application Form Reviews

Day Workshops

• Search Menu
• Sign in through your institution
• Advance articles
• Author Guidelines
• Submission Site
• Open Access
• Why Submit?
• About Oxford Journal of Legal Studies
• About University of Oxford Faculty of Law
• Editorial Board
• Advertising and Corporate Services
• Journals Career Network
• Self-Archiving Policy
• Dispatch Dates
• Journals on Oxford Academic
• Books on Oxford Academic

Article Contents

1. introduction, 2. the possibility of the null model, 3. the rule of law, 4. a fresh start, 5. two modes of precedential reasoning, 6. evaluating the modes, 7. conclusion, precedent and the rule of law.

Somerville College, University of Oxford. Email: [email protected] . I am very grateful to Timothy Endicott, Katharina Stevens, Joseph Raz, Crescente Molina, Donald Bello, Joshua Pike, Hafsteinn Dan Kristjánsson, Benjamin Shoemaker, Nick Grant, Manuel González, Michele Boggiani, Tom Kohavi, Gehan Gunatilleke, Mikolaj Barczentewicz, Li-kung Chen and two anonymous OJLS reviewers for comments on previous drafts of this article. An earlier version of the article was presented to the Edinburgh Legal Theory Discussion Group. Many thanks to the participants in that seminar, in particular to Claudio Michelon, Joaquín Reyes, Maggie O’Brien, Euan Macdonald, David Cabrelli, Amalia Amaya, María Ignacia Besomi, JP Fassnidge, George Dick and Ismael Martínez.

• Article contents
• Figures & tables
• Supplementary Data

Sebastian Lewis, Precedent and the Rule of Law, Oxford Journal of Legal Studies , Volume 41, Issue 4, Winter 2021, Pages 873–898, https://doi.org/10.1093/ojls/gqab007

• Permissions Icon Permissions

Courts may reason using precedents in various ways, but not all of them satisfy the rule of law. This article provides two ways that are compatible with this ideal and one which is not. In doing so, the article aims to explain the practice of following precedent in law and to offer criteria for evaluating its value. Two claims are defended. First, courts always have a reason to decide precedent-governed disputes by following precedent. This reason is a minimum requirement of the rule of law, and in some cases this reason may be reinforced in the form of an obligation. Secondly, depending on whether courts have a reason or an obligation to follow precedent, two modes of precedential reasoning may be identified. The article explains them in detail. The modes, together with the considerations that are reasons in favour of them or against them, provide a valuable philosophical foundation of precedent-following in law.

Many of our normative practices are influenced by the force that past decisions exert on similar, future situations, in favour of reaching the same result as before. We normally call these past decisions precedents , and when it comes to the law we often say that precedents influence the decisions of courts when relevantly similar disputes arise.

But the practice of following precedent is not something we should take for granted. To begin with, one can imagine legal systems where courts do not pay attention, in a normatively significant way, to the ways in which similar disputes have been decided in the past. Similarly, there may be legal systems that prohibit courts from relying on precedents, at least for the purpose of using them as a legal basis for reaching justified decisions. On the other hand, it is not clear whether following precedent has value in its own right, or whether we follow precedent in order to advance other values. Finally, precedents can influence the decisions of later courts in various ways. In the civil law tradition, precedents are often used in order to tip the balance in favour of particular outcomes, but are also used as a means to illustrate how a legal point has been dealt with before. In the common law, by contrast, precedents typically play a more decisive role. In many cases, precedents are authoritative sources of law, in the sense that if the facts in a later case are legally the same as those of a precedent, the later court is often required to deliver the same decision.

My aim is to explain the practice of following precedent in law and provide criteria for evaluating its value. Since there is significant literature on the topic, 1 I will do this by defending two connected and novel claims. The first argument is that, contrary to what may happen in practice, whenever a precedent-governed dispute arises, the later court always has a reason to decide that dispute by following precedent. This reason is grounded on the idea that, by doing so, a legal system has a distinctive way to advance at least the following values: maintaining legal stability, allowing people to rely on reasonable expectations and providing equality in the judicial application of the law. 2 I will call these values respectively ‘stability’, ‘reliability’ and ‘equality’. Since I take them to be part and parcel of the rule of law ideal, I will refer to them collectively as ‘the rule of law’. 3 I argue that the principles of judicial transparency and good government strongly militate in favour of courts having to give reasons why they will not follow relevant precedent.

The reason in favour of following precedent is only a minimum requirement of the rule of law. On top of this requirement, a legal system may reinforce its commitment to this ideal by requiring courts to treat the same reason as an obligation . Accordingly, the second claim is that we can identify two modes of precedential reasoning precisely by whether courts have a reason or an obligation to decide precedent-governed disputes by following precedent. Consistent with familiar terminology, I will call them ‘persuasive’ and ‘authoritative’. 4 These modes are not necessarily depictions of what particular courts do, but they provide us with a valuable philosophical foundation for assessing whether what these courts do is desirable or not.

The article is structured as follows. In section 2, I argue that the practice of following precedent should not be taken for granted. To do this, I posit what I will call the ‘null model’, under which courts are authorised to decide precedent-governed disputes without paying attention, in a normatively significant way, to relevant precedent. By showing that the null model is both conceptually possible and we may find arguments supporting it, this offers a useful baseline against which precedent-following can better be assessed. In section 3, I show the extent to which the null model fails to live up to the rule of law. Section 4 argues that one distinctive way to advance the rule of law is by requiring courts to always put in the balance the reason in favour of following precedent. I will argue that, based on a legal system’s commitment to the rule of law, this reason is both non-contingent and content-independent. By showing their main features and differences, section 5 articulates the persuasive and authoritative modes of precedential reasoning. Finally, section 6 offers an assessment of how both modes deal with two evaluative questions: (i) How does each mode advance the rule of law? and (ii) How does each mode avoid replicating substantively incorrect decisions? The purpose of this section is to present some of the tools that are vital for determining whether there is a mode that should be preferred, and for eventually criticising a specific practice of precedent held by courts.

With respect to terminology, I will speak of the ‘earlier court’ to refer to the court that decided a case in the past that is relevantly similar to the one being decided by the ‘later court’. I will also ignore alternative ways of speaking of precedents, such as the erga omnes force of judicial decisions, 5 the constant jurisprudence of courts and the common law. Here, a precedent should be taken to mean a past decision or past case that is relevantly similar to the case at hand. Finally, I will use the expressions ‘relevantly similar’ and ‘legally the same’ as equivalents. The latter stresses the idea that the similarities between the two cases should be relevant in the eyes of the law. 6

There is one caveat to bear in mind. My account, like that of many others, 7 does not depend on a strict separation between vertical and horizontal stare decisis . 8 As I will explain, I think this distinction is useful to capture those specific precedents that, for various considerations, certain courts must follow. But that question is secondary to determining whether later courts always have a reason to follow precedent—the question that motivates this article.

Justice and courts have always had a complex relationship. John Gardner wrote that ‘judges should first and foremost administer justice’, 9 and before him HLA Hart held that we naturally think of justice as being administered according to law. 10 This complex relation has contributed to the way we have historically called courts, shifting between courts of justice and courts of law . 11 Today it is not controversial to say that, when courts decide disputes, and thereby attempt to deliver justice, they do so by applying the law to the case at hand.

But what is the specific law that would allow judges to reach legally justified decisions? The answer to this question inevitably varies across different legal systems, as it depends on the sources of law in the particular legal system under consideration. Legislation is typically a source of law, meaning judges can reach legally justified decisions by applying statutory law. But things are more complex when it comes to ‘case law’. In the common law, judicial decisions are generally given legal status, 12 meaning courts can reach legally justified decisions by applying case law to the dispute at hand.

The same cannot be said, or at least not as simply, 13 with respect to various countries in the civil law tradition. In many of these countries, it is generally the case that judicial decisions are authoritative with respect to the parties to the dispute only , and this effect is commonly known as the ‘relative force’ of judicial decisions. 14 This means, first, that as a general rule, 15 only the parties to the dispute may benefit from the authoritativeness of the decision—for example, to demand conformity with its terms. (To this extent, the relative force of judicial decisions is also present in the common law.) But, secondly, if a relevantly similar dispute arises between other parties in the future, the later court—in the civil law—may be prevented from reaching a legally justified decision solely by following precedent. As Eva Steiner writes concerning French law: ‘Explicit reference by a court to its own jurisprudence when giving a decision and, more generally, citation of previous cases is not allowed when these are meant to serve as a legal basis for the court’s decision.’ 16

Of course, French law does not exhaust the civil law tradition. The point is that, often, the relative force of judicial decisions entails that precedents are not generally given legal status. As Merryman and Pérez-Perdomo observe: ‘[in the civil law] prior judicial decisions are not “law”’. 17 Therefore, in order to reach decisions justified by law, civil law courts usually rely on undisputed legal sources. Legislation, again, presents itself as a good example of an undisputed source of law, but in some jurisdictions the constitution also serves as a legal basis—as well as ratified international treaties. For this reason, if a court relies on a legal source, but in doing so it ignores or contravenes relevant precedent, the decision may nonetheless be legally justified. 18

But if precedents may not provide justificatory basis whatsoever for reaching legally justified decisions, what normative role do they play? Drawing from a comparative analysis including both civil law and common law jurisdictions, Aleksander Peczenik concluded that precedents typically perform the following roles: (i) they bind formally; (ii) they have normative force but do not bind formally; (iii) they are neither formally binding nor have any normative force, but provide ‘further support’ (see note 19); and (iv) they serve to illustrate points of law. 19 Peczenik’s conclusion makes room for the proposition that the existence of a relevant precedent may make no normative difference in the decision of a later court. This conclusion is supported by the fact that, as we saw before, many civil law courts may reach legally justified decisions even if they have contravened or otherwise ignored relevant precedent.

It is true that the civil law is changing towards giving precedents a more decisive role (see note 13). That has been because of the long and arduous work of courts, the legal community and the legislatures. But this change reveals a point of departure: the possibility that the existence of a precedent may make no normative difference in the decision of a later court. This possibility provides the conceptual basis for thinking that a particular model of precedent may exist. I will call it the ‘null model’, because for a later court the existence of a precedent regulating the issue at hand makes no normative difference to the vital question of how such an issue should be decided. This type of case should be distinguished from those in which the later court offers reasons for not following relevant precedent. In this situation, the precedent makes a normative difference, since the later court has to meet a justificatory burden for failing to conform to it. The null model, by contrast, seeks to capture those situations in which the existence of a relevant precedent is another non-normative fact of the world: the precedent does not even tip the balance of reasons in favour of reaching the same outcome as before. These could be, for instance, cases where the later court knows about the existence of a precedent, but it does not show why the precedent will not be followed.

Various reasons may support having the null model. With some adjustments, an advocate of this model could rely on the following. First, legal systems have, all else being equal, an interest in preventing judges from making creative innovations in the law. In order to keep the law pure, the influence of judicial decisions on future cases should be limited as much as possible, and the null model would be one way of doing this. Secondly, judges have a duty to treat parties impartially and objectively. Since the practice of following precedent is normally affected by factors external to the law, such as the prestige of the earlier court, these factors may introduce bias in the court’s decision. In order to avoid this, judges should not follow precedent. Finally, the law should have a democratic character. Precedents are often given legal status, but they are not created by a democratic process. One way to avoid this non-democratic upshot is by giving judicial decisions relative force only, thus restricting the influence of precedents on later courts.

The possibility of the null model is valuable for two reasons. It reminds us, first, that the values brought by judicial conformity to precedent, soon to be studied, should not be taken for granted. The null model, as such, is a way to appreciate these values by conceiving of a legal system where the same practice does not exist. Secondly, the null model confronts us with the following challenge. If nothing in law’s nature prevents courts from lawfully adjudicating disputes under the null model, 20 then why doesn’t any legal system have it—at least officially? Put differently, what causes many later courts to pay attention, in a normatively significant way, to the ways in which earlier courts have decided relevantly similar cases?

Many ideas are usually associated with the expression ‘the rule of law’. It is often said, for example, that the rule of law is first and foremost a requirement of good governance. People exercising positions of political authority should be subject to mandatory rules aiming to minimise the potential for arbitrary government and even tyranny. 21

The rule of law is also identified with a set of formal demands that seek to allow individuals to better anticipate what the law may require from them. Accordingly, the standards by which the law will guide conduct should be clear, public, general, prospective and stable, so that people may know in advance how to plan their lives under the law. 22

Similarly, the rule of law requires minimum procedural conditions that courts should observe when they administer justice, in order to ensure parties a ‘fair hearing’. 23

Finally, the rule of law is also associated with substantive conditions for the existence of the law. Some say, for example, that the law should meet basic demands of fundamental rights. 24 Others claim that the rule of law should go one step further and secure basic social rights. 25

These various ways of conceiving of the rule of law have led many scholars to distinguish between ‘formal’ and ‘substantive’ versions of the rule of law. 26 As Brian Tamanaha notes, ‘formal theories focus on the proper sources and form of legality, while substantive theories also include requirements about the content of the law’. 27 However, as Tamanaha also writes, though the distinction is informative, it should not be taken at face value. The reason is that ‘formal versions have substantive implications and the substantive versions incorporate formal requirements’. 28 Other scholars have shared similar worries. 29

Importantly, for the purposes of this paper, I will adopt a modest understanding of the rule of law—one usually associated with formal versions. The reason is not because I reject substantive versions of the rule of law—a point we need not settle here. The point, rather, is that this modest version is sufficient to ground precedential constraint, while avoiding the profound disagreement entailed by many of the substantive versions. In effect, the more substantive one’s conception of the rule of law becomes, the more disagreement it seems to produce. 30 On the other hand, ‘All substantive versions of the rule of law incorporate elements of the formal rule of law’. 31 In sum, certain formal demands are necessary conditions of the rule of law, while also being sufficient to ground precedential constraint. This is the moderate strategy I shall pursue. 32

In particular, there are three rule-of-law values the advancement of which gives us reasons to reject the null model, and prefer instead one of the modes that I will propose. 33 These are the values of stability, reliability and equality in the application of the law. 34 I will briefly explain these values, and will then show the extent to which the null model runs counter to them.

A. Stability and Reliability

Stability is a state of affairs in which the content of the law of one country is settled over a considerable amount of time. The proviso of time is important because it would not seem plausible to think that a legal system has achieved stability in relation to a specific matter if the content of the law on that matter changes too frequently. Achieving stability is a matter of degree, for there is no threshold indicating how much time is required for the content of the law to become stable. This depends on many factors, such as the area under examination, where in some cases one specific change in the content of the law may have a much more systemic impact than changes in other, less sensitive areas.

It is not clear whether stability has value in its own right. 35 It seems that we value stability not for the sake of stability, but because it is instrumental to the realisation of other values, in particular one which directly bears on the decisions we make in our daily lives: the value of reliability. It is valuable that people know what the content of the law is, what the law commands, allows and forbids, for when people have this knowledge, they can rely on it as valuable information to adopt decisions that will impact their lives. In many ways, people can shape their lives, anticipate events and be psychologically confident thanks to the structural boundaries offered by the law. But in order for the law to allow this, the law cannot suffer changes too frequently. The law, in other words, needs to be stable. Stability and reliability are thus two sides of the same coin: what may advance the former may also promote the latter, and what may undermine the former may also affect the latter.

Since stability and reliability are intimately connected, I will focus on reliability to argue against the null model. Recall that under the null model the fact that a precedent exists—regulating the matter at hand—does not even tip the balance of reasons in favour of reaching the same decision. True, if the case is governed by statute and the application of the relevant norm is straightforward enough, then people have a reason to believe that such a case should be decided in the way prescribed by the statute (assuming, of course, that courts have a duty to give effect to the content of statutes). But where the application of the relevant norm entails a ‘hard case’, this reason for belief might become weaker, or even disappear, if the court adjudicates under the null model.

Take cases of statutory vagueness, or where there are various candidates for a correct interpretation of the statute, or where applying the norm would conflict with a moral consideration. For our purposes, a hard case is one where there are good reasons to believe that the court may arrive at different conclusions concerning the statute’s application. All these outcomes, further, are justified by law. In a hard case, the null model entails a probabilistic lottery. If there are two correct ways of deciding the case, but the fact that one of them is supported by precedent does not count in the balance, then ceteris paribus parties have a 50% chance of anticipating the court’s decision rightly. This number decreases when the amount of possible justified scenarios increases. If, say, there are four possible correct answers, the chances of parties getting things right is 25%, again ceteris paribus .

Provided that there are alternative modes of adjudicating precedent-governed disputes that can offer more reliability to parties, this probabilistic lottery runs counter to the rule of law. As I will argue, one of these ways is by making it the case that courts always have a reason to decide precedent-governed disputes by following precedent. The existence of this reason increases the degree of reliability given to parties vis-à-vis the null model. Parties can rely before litigation on the fact that courts always have a reason to decide disputes in conformity with relevant precedent. But since this reason does not exist in the null model, parties are left to whatever assessment of the correct scenarios the court deciding the case makes.

B. Equality

Equality is a contested notion, and some have shown scepticism about whether there is something distinctive in equality other than the generality presupposed in every norm-application. 36 But the existence of a norm is one thing; its application is another. Thus, for reasons external to the norm itself, there can be cases that fall within the norm’s scope, but which may receive a different treatment—despite the fact that in virtue of the norm’s generality they should not. 37

This difference of treatment can be explained by the discretionary element entailed in the application of a norm. To begin with, applying a norm requires determining, and eventually justifying, whether certain facts of the world fall within the norm’s scope. This process is generally known as subsumption , and though the margin for discretion might be reduced, adjudication will often require sensible judgment. It is here where judges come into play. Judges are trusted with the power to determine, inter alia , whether facts before them can count as instances of the norm’s material scope of application. If they count, then we often say that judges have an obligation, though not necessarily a conclusive one, to decide the case according to the result provided by the norm. 38

Subsumption can be a straightforward or complex process, depending on whether it is clear or disputed that the facts of the case fall within the norm’s scope. When it is clear, we can think of the case, at least in respect to subsumption, as an easy one. But subsumption does not exhaust the adjudicatory work of judges: they still have to determine whether the norm should be applied all things considered. As Kenneth Winston observes, judges may ‘refrain from applying a law to a case that it clearly covers, for example, on the ground that there are features to the case that were not anticipated’. 39 Conversely, complexity in subsuming facts is a reason for the case to be seen as a hard one.

As it relates to precedent, equality requires courts to limit the menu of possible correct answers that, absent relevantly similar cases decided before, they would otherwise have. This means that, if the earlier court reached a decision that was initially justified, the later court is prevented from reaching a different decision without justification. For example, suppose it was an open question for an earlier court whether to treat a particular object as an instance of the statutory term ‘wheelchair’. In a relevantly similar case, a later court would thus not be able to say, at least not without justification, that the same object is not a wheelchair had the earlier court decided that it was.

But equality, in this sense, can also play a role in the equitable dispensation of legal requirements—assuming, of course, that courts have the corresponding power. If, on grounds of equity, an earlier court had made a particular exception in favour of a party, then the later court is prevented from denying the same exception without justification. 40

The null model authorises courts precisely to dispense with this justificatory requirement—namely, of having to argue why they will depart from the decision of the earlier court. The final decision, which affects equality but does not offer a justification, is legally justified. Contrary to this state of affairs, as I have argued, equality requires later courts to give reasons why they will depart from past decisions. I have considered the examples of legal interpretation and equitable dispensation, but these do not exhaust the adjudicatory work of judges.

In sum, the null model can be lawful, and supported by various reasons, but there are important reasons as well for thinking that it fails to live up to the rule of law. And the rule of law is an ideal worth pursuing. It has been said, for example, that the rule of law is ‘law’s own virtue, respect for which is needed for the law to have any other virtue’. 41 Others think of it as ‘the specific virtue of legal systems’. 42 Still others see it as ‘the morality that makes law possible’. 43 These various ways to think about the rule of law suggest that a legal system has important reasons to live up to this ideal, including stability, reliability and equality. These reasons might not outweigh all competing values, as some scholars have rightly noted, 44 but, all else being equal, they do recommend a commitment to the rule of law. In virtue of this commitment, departures from the rule of law, as Tom Bingham argued, should have a ‘clear justification’. 45

In the next section, I will lay out the structure for articulating two modes of precedential reasoning that, unlike the null model, advance the rule of law.

Having argued against the null model, it is time to propose a fresh start. Unlike before, where I proceeded bottom-up, 46 this time I will proceed top-down. I will start from the rule of law and then articulate two modes of precedential reasoning that are consistent with this ideal.

The argument in this section is as follows. The first subsection argues that one distinctive way to advance the rule of law is by requiring courts to always decide precedent-governed disputes by following precedent. Notice that the claim here is a modest one. I will not be arguing that courts should always have an obligation to follow precedent, but only a pro tanto and non-contingent reason. Precisely because of this caveat, the claim is, I think, novel. What is needed, in order to establish the non-contingency of the reason in favour of following precedent, is the normative backup of a positive second-order reason. 47 In the second subsection, I will argue that a legal system’s commitment to the rule of law can do this work. The upshot is that the non-contingent reason to follow precedent is a minimum requirement of the rule of law.

A. Advancing the Rule of Law

The task before us is to show that the reason to follow precedent is a valuable means to advance the rule of law. I do not think, based on the relevant literature, 48 that this claim is controversial. But it is important not to take it for granted—observing also, as others have done, 49 that sometimes following precedent may not necessarily advance the rule of law. This is why a ceteris paribus clause is needed: all else being equal, following precedent is a distinctive means towards a valuable end.

Consider stability and reliability. Both values demand that the content of the law remain stable over time, so that people can rely on it to adopt decisions. Since stability is instrumental to reliability, think of the problem in the following way. If the precedent and present cases are legally the same, but they receive different treatment, can people form reliable expectations concerning the ways in which these types of cases will be decided? Possibly yes, if such a decision is treated as an exception. 50 But where there is a systematic and widespread practice of giving a different treatment to two cases that are legally the same—the past and present cases—people might have a reason not to form any expectations whatsoever. When courts follow precedent, people are given a reason to believe in, and thereby rely on, the fact that if their cases are relevantly similar then they will receive the same treatment.

Sometimes, however, courts will have to depart from past decisions, and this is particularly true with respect to those decisions that are notoriously suboptimal—ie are notoriously unjust, arbitrary or otherwise unsound. Admittedly, not every suboptimal precedent ought to warrant a departure; otherwise, the contribution of the practice of following precedent to stability and reliability may be jeopardised. 51 This raises the question of whether there is a threshold for not following precedent.

It seems that, if a legal system wants to balance tailored justice (to the parties) with stability and reliability, that threshold will unavoidably be affected by what Andrei Marmor called ‘moral vagueness’. 52 The threshold, in other words, will contain an evaluative concept whose application requires a value-laden judgment by the later court.

Take the House of Lords’ 1966 Practice Statement: precedents of that court are ‘normally binding’ (on the same court), but they admit departures ‘when it appears right to do so’. 53 It seems difficult, and perhaps self-defeating, to anticipate all the necessary and sufficient conditions for establishing when it is right for that court to depart from precedent. 54 For this reason, a legal system will most likely delegate this determination on judges themselves, who will balance tailored justice with stability and reliability, often leaning towards the latter—as the same House of Lords (today the UK Supreme Court) has shown over time. 55

Concerning equality, we could ask a similar question. Could an observer affirm that courts in a particular country are being consistent with past exercises of judicial discretion when they adjudicate differently in disputes that are legally the same? Again, possibly yes, if these decisions are exceptional; but not if they count as the general rule. The practice of following precedent aims to prevent the passage of time from making a difference to the treatment of two disputes that are relevantly similar. Put differently, were a judge required to adjudicate simultaneously two disputes that are legally the same, she should decide them in a like manner. But judges seldom adjudicate disputes simultaneously; rather, they do it over a period of time. The practice, therefore, aims to prevent the passage of time from altering this kind of treatment. Accordingly, when two disputes are legally the same, but they arise at different times, one way to deliver consistent treatment is to decide the new case in conformity with the precedent.

Conformity to precedent is a distinctive means to advance the rule of law. By precedent-following, a legal system can foster this ideal in ways that can only be done through adjudication—for example, by tackling specific problems raised by litigation among individuals, resolution of which has systemic implications for the legal community as a whole. The null model, by contrast, is precisely characterised by the fact that the legal system has decided not to live up to this distinctive opportunity—of advancing the rule of law via precedent-following.

B. Commitment to the Rule of Law

I have argued that the values of stability, reliability and equality are ends towards which the practice of following precedent presents itself as a distinctive means. If my discussion is sound, it provides a pro tanto reason for judicial conformity to precedent. That is, to the extent that courts have a reason to advance the rule of law, they also have a reason to follow precedent.

But more needs to be done. In particular, it is vital that the reason to follow precedent does not depend on whether a later court agrees with the precedent’s correctness. Otherwise, we face two crucial objections: one descriptive, the other normative. The descriptive objection would argue that many courts follow precedents independently of their correctness. 56 The normative objection, by contrast, would insist that, if precedent-following is to make a contribution to the rule of law, it should not depend on a later court’s sympathy with the precedent. 57

I shall pursue the opposite strategy: the reason to follow precedent is content-independent, in the sense that later courts have it whether they agree with how the precedent was decided or not. One challenge that arises immediately is the following: how can we justify later courts that knowingly replicate incorrectly decided precedents? How can an agent be justified in performing an action she knows to be incorrect? Drawing from contributions on argumentation theory 58 and practical reasoning, 59 I will argue that a legal system’s commitment to the rule of law can justify later courts in following precedents they know to be wrong.

To illustrate the problem, it will be useful to distinguish between two versions of following precedent: ‘weak’ and ‘robust’. 60 According to the weak version, a later court has a reason to follow precedent only if the precedent was correctly decided. By contrast, in the robust version a later court has such a reason independently of the precedent’s correctness. On this view, the mere fact of a precedent’s existence—regulating the case at hand—is a reason for the later court to follow it. It is the robust version that Frederick Schauer has in mind when he argues: ‘if we are truly arguing from precedent, then the fact that something was decided before gives present value despite our current belief that the previous decision was erroneous’. 61

For Schauer, precedent-following entails the robust version. Yet the problem with this approach is that an incorrect precedent cannot become correct simply by virtue of being followed in the future. If I lied to you in the past, lying to you again will not be justified simply because I already did it once. The reason, in general, is that the mere performance of an action is not sufficient to alter its moral quality. 62

A similar situation happens in law. If an earlier court set a precedent that the later court knows to be wrong, following it will not make the later court’s decision right. Or not unless further premises are added to the normative background. 63 A typical premise of this kind is a positive second-order reason for action. 64 In general, first-order reasons are considerations in favour of performing or refraining from performing a certain action. Positive second-order reasons are considerations in favour of acting on one of these first-order reasons. Negative or ‘exclusionary’ second-order reasons are considerations supporting not acting on certain first-order reasons. 65 The combination of a first-order reason plus an exclusionary second-order reason is what Joseph Raz calls a ‘protected reason’. 66 (I will return to this idea in the next section.)

Commitments are a good example of a positive second-order reason. If, for instance, one has good reasons to pursue a particular project, a commitment can be a second-order reason to act on these reasons. More importantly, sometimes this commitment can be the sole reason for why one perseveres on the chosen path—say, because one’s initial reasons no longer exist, such as when one’s preferences, or the circumstances of life, have changed. In these cases, commitments provide, as Ruth Chang writes, ‘the grounds for new will-based reasons’. 67 Similarly, in virtue of a particular commitment, sometimes we see ourselves making decisions we would not otherwise make. Your commitment, say, to a professional tennis career can put you in a situation where every year you miss the birthday party of your daughter, because every year you must play the same tournament in another country. Without this commitment, it would be wrong to miss your daughter’s birthday. In more dramatic scenarios, as Raz writes, ‘one’s choice does make it right for one to pursue a goal which but for one’s commitment to it would have been a wrong goal to pursue’. 68

As a positive second-order reason, therefore, a commitment is a premise that can be added to the normative background to justify what would otherwise be unjustified. Likewise, when it comes to precedent-following, commitments can do a similar work. Suppose two years ago your daughter Claudia turned 15 and you allowed her to drink wine for the first time. In about a week, Peter, your youngest son, will reach the same age, and he wants to know whether you will allow him to drink wine as well. Assume, further, that with time you have come to realise that it was never right to allow Claudia to drink wine at such a young age. That decision was, in other words, a mistake. Now, if the situation of Peter is relevantly similar to that of Claudia, 69 do you have a reason to allow Peter to drink wine?

Again, according to weak precedent-following, you do not have such a reason if it was wrong to allow Claudia to drink wine. Now, suppose you have a commitment not to disappoint the reasonable expectations that your past decisions have created in your children. Or assume that it is really important for you not to give your children the impression that you treat them differently when they are similarly situated. Thus, unless very exceptional circumstances arise, you have a commitment to show them that you treat them equally by following your past decisions.

This type of commitment can ground robust precedent-following. It can give you now a reason for following a past decision you know is incorrect. Similarly, a legal system’s commitment to the rule of law can justify those courts that follow precedents known to be wrong. As Neil Duxbury writes, ‘Even a decision widely considered wrong might continue to be followed if people have reasonably relied on it arranging their affairs’. 70

Therefore, if a legal system has a commitment to the rule of law, and if judges live up to that commitment, 71 then robust precedent-following is a distinctive way in which they could do so. Now, we should bear in mind two things. First, there is a gap here between precedents that are wrong and those that are utterly wrong. I want to leave open the possibility that, with respect to the latter, a commitment to the rule of law may not necessarily turn right what is utterly wrong—simply by following precedent. 72 This caveat, however, does suggest that, with respect to less suboptimal precedents, or to precedents that do not necessarily reflect the later court’s preferred views, a commitment to the rule of law can do the justificatory work we are looking for.

Secondly, commitment to the rule of law does not necessarily ground a content-independent obligation to follow incorrectly decided precedents. It does, however, ground a content-independent reason —again, pro tanto . To say that courts may follow wrongly decided precedents in order to foster the rule of law is not to say that they must . It means only that we need to keep the door open for courts to reach legally justified decisions even if they follow precedents known to be wrong.

Commitment to the rule of law can make room for courts to adjudicate under robust precedent-following. Thus, the mere fact that an earlier court decided a dispute is a reason for the later court to follow this decision in a relevantly similar dispute. Under the strong version, the existence of a precedent-governed dispute triggers a non-contingent reason for action. The later court always has a reason to decide the same dispute by following precedent.

This non-contingent reason is a minimum requirement of the rule of law. Without it, we are susceptible of going back to the null model. But once this minimum requirement has been secured, there is yet another question that we need to ask ourselves. The answer to this question may divide legal systems. The question is whether the non-contingent reason to follow precedent should stay as a minimum requirement or should be reinforced. One way in which a legal system may reinforce its commitment to the rule of law is by requiring courts to treat the same reason as an obligation . For present purposes, an obligation is a requirement that derives from the combination of a first-order reason plus an exclusionary reason not to act on some conflicting reasons—a protected reason. 73

Therefore, depending on a legal system’s commitment to the rule of law, we can identify two modes of precedential reasoning. One is characterised by the fact that the non-contingent reason to follow precedent is only a reason—and, as such, it can be outweighed by a more compelling reason. The other is characterised by that reason being a protected reason—ie having an exclusionary character—meaning later courts are prevented from acting on certain excluded reasons. Unlike the null model, these two modes are consistent with the rule of law—in fact, they are derived from this ideal.

In the following two subsections, I will articulate these modes in more detail. Following familiar terminology, I will call them ‘persuasive’ and ‘authoritative’. 74 I will start with the latter, because it presents complexities worth disentangling from the outset. Before doing so, however, it is important to consider one important methodological caveat.

Whether an account is fully descriptive or may be detached, to a sensible extent, 75 from a certain practice depends on the account’s purpose. In the introduction, I suggested that my aim is not to depict specific adjudicatory practices, for example, of either the civil law or the common law, but to provide a framework within which we may understand these practices. In particular, the authoritative mode aims to explain the idea of having a protected reason to follow precedent, while the persuasive mode involves that of having a bare reason . The authoritative mode, thus, is not necessarily a depiction of the common law, nor is the persuasive mode necessarily one of the civil law. Courts in these two traditions can have both types of reasons to follow precedent. Accordingly, the civil law may share features of the authoritative mode, while the common law may share ones of the persuasive mode. 76

A. The Authoritative Mode

Under the authoritative mode, whenever a precedent-governed dispute arises, the later court has a protected reason to follow precedent. This reason entails both a reason to decide the dispute in conformity with the precedent and an exclusionary reason not to decide the same dispute on certain reasons against following the precedent. If there are no non-excluded reasons against following precedent, then the later court must follow the relevant precedent—unless, as we will see, the precedent can be distinguished. 77 Again, it is because of this combination of a first-order reason plus an exclusionary reason that we can think of the protected reason to follow precedent as giving rise to an obligation . 78

Typically, the obligation of later courts to follow precedent applies whenever the precedent and present case are legally the same: whenever their similarities are relevant for the law, while their differences are not. When, conversely, these two cases are legally different , a later court may distinguish. In other words, the obligation of later courts under the authoritative mode has disjunctive form: either to follow or distinguish the precedent. 79 ‘Distinguishing’, in particular, is the practice by which the later court shows that the present case has a novel fact not captured by the precedent, and which is legally relevant for handing down a different decision. In the common law, the majority of scholars think that distinguishing is an integral part of the practice of following precedent—and hence the obligation’s disjunctive form. 80 Some scholars, however, think that later courts should not distinguish. 81

This debate between common law scholars has one consequence for our analysis. What exactly does the obligation to follow precedent consist of? Is there an obligation for later courts to apply the legal rule provided by the ratio decidendi of the precedent? Or is such an obligation, as Grant Lamond claims, to treat the precedent as correctly decided on its facts? 82 The fact that this debate is still ongoing makes it difficult to provide a clear-cut answer. 83 My own view, which I hope to provide in detail soon, is that the obligation of a later court is to respect the decision of the earlier court to treat certain established facts as material, to give them a particular legal significance and to decide on the balance of reasons. 84 This view would lean towards that of Lamond and later developed by John Horty. 85 It can be summarised as follows: the later court must render a decision that is consistent with the authoritative treatment that the earlier court gave to the facts or factors of the precedent, as these were reported.

In any event, for now we need only to bear in mind that the specific content of the obligation to follow precedent is a matter of discussion in the common law. That obligation can take the form, inter alia , of having to apply the rule laid down by the ratio , 86 or having to decide in a way that is consistent with the ‘background case’, as Horty puts it. 87 But, to be sure, in both cases the later court has an obligation to decide the precedent-governed dispute by following precedent. In both cases, then, an authoritative mode exists. This mode, again, entails a reason for action—namely, decide the present case by following precedent—and an exclusionary reason not to act on certain conflicting reasons. 88

One reason typically excluded from being acted upon by the later court is disagreeing with the precedent’s correctness. It is often the case that the later court must follow the relevant precedent even if it thinks that, by doing so, the court will render an otherwise suboptimal decision. In some exceptional cases, the legal system may authorise some courts to act on the otherwise excluded reason of disagreeing with the precedent’s substantive merits. When these precedents originated in the same court that may act on this non -excluded reason, the authorisation is typically known as the power to overrule . Its normative effect is twofold: for the later court, a reason otherwise excluded is a non-excluded reason. For the legal system, the overruled precedent loses its erga omnes force. The upshot is that such a precedent typically ceases to be part of the law. 89

The grounds on which a court may overrule vary across jurisdictions—and also within one legal system. To illustrate: by virtue of the 1966 Practice Statement, 90 the House of Lords (today the UK Supreme Court) is authorised to overrule its own precedents when ‘it appears right do so’. Accordingly, not following precedent when ‘it appears right to do so’ is a non-excluded reason. 91 In contrast, in Young v Bristol Aeroplane Co Ltd , 92 the Court of Appeal (Civil Division) declared itself to be bound by its own precedents, with three exceptions: (i) where two of these precedents would conflict with each other; (ii) where one of these precedents is incompatible with a decision of the House of Lords/Supreme Court; and (iii) where a precedent was given per incuriam . Thus, were the Court of Appeal to overrule one of its precedents because ‘it appears right to do so’ (as in the Practice Statement), that decision would be incompatible with the authoritative mode. The court would be acting on a reason that was excluded. 93

A final remark. For various reasons, most of which have to do with effective action-guiding and better coordination between courts, a legal system may restrict the scope of the precedents that later courts must follow. Vertical and horizontal stare decisis are precisely ways to characterise this restriction. Now, the logic of grounding precedent-following in the rule of law does not necessarily preclude bottom-up stare decisis —ie higher courts being bound by the precedents of lower courts. However, by virtue of considerations such as effectiveness, this authoritative force is typically restricted to either vertical (top-down) or horizontal stare decisis . Yet one could still argue that the rule of law provides a non-contingent and pro tanto reason (albeit not a protected reason) for higher courts to consider at least some of the precedents of lower courts.

B. The Persuasive Mode

Under the persuasive mode, whenever a precedent-governed dispute arises, the later court has a bare reason to decide the same dispute in conformity with the precedent. As such, this reason can be outweighed by a more compelling reason.

As suggested before, possibly a notable feature of the persuasive mode is that it is flexible, because it allows for different degrees of precedential constraint depending on the weight to be given to the reason in favour of following precedent. This reason could be very weighty if, say, the precedent originates from the highest tribunal in the land. Even in that scenario there might be room for various alternatives, depending on, say, whether the past decision was reached by a clear majority, a prestigious judge concurred and so on. By contrast, the same reason may be normatively weaker if, say, the precedent was reached by a lower court, or a court composed by one judge only.

Another factor giving more or less weight to the reason in favour of following precedent is whether the later court thinks that the precedent was rightly or wrongly decided. But this creates a problem: how can this substantive analysis be compatible with the content- independent character of the same reason? How can this reason allow for various weights depending on whether the later court agrees with the substance of the precedent?

This dilemma, to be sure, is merely apparent. Nothing in the notion of a content-independent reason prevents this reason from gaining extra weight by reference to some of its substantive merits. 94 The idea of content independence guarantees the reason a normative minimum: a nucleus that cannot be reduced but which can become weightier. Thus, to the extent that the later court has a reason to follow precedent, that reason has a normative nucleus that will not be affected by whether the precedent was rightly or wrongly decided. On top of this nucleus, the later court may add extra weight if it considers that the dispute was, say, rightly decided, or it may add no further weight at all if the reverse holds.

As we can see, the structure of the persuasive mode allows room for various scenarios. One may argue that within this mode there exists a weak version in contrast to a strong one. Within the latter, certain courts will take very seriously the existence of a precedent regulating the case at hand, but will not see themselves as being bound by it. The former would make room for those judges for whom the reason to follow precedent can easily be outweighed by another, conflicting reason.

Finally, there are two features that distinguish the persuasive mode from the null model. First, in the persuasive mode, whenever a relevantly similar dispute arises, later courts always have a reason to follow precedent. This non-contingent reason does not exist in the null model. Secondly, the principles of judicial transparency and good government strongly militate in favour of courts adjudicating under the persuasive mode to explain why the reason to follow precedent has been outweighed. This is because, all else being equal, it is contrary to the rule of law to depart from relevant precedent without explaining why.

In this final section, I want to analyse how both modes of precedential reasoning deal with two evaluative questions. The purpose of this analysis is to offer some of the tools that are vital for determining whether there is a model that should be preferred, and eventually for criticising the particular practice of precedents held by courts.

A. How Does Each Mode Advance the Rule of Law?

Let us start with stability and reliability. The authoritative mode often, but not necessarily, entails a view where precedents are part of the law. To the extent that courts are bound by precedent, those precedents may be treated as part of the law. 95 Where this is so, every court with the power to set a precedent has also, in theory, the power to change the content of the law. 96 Yet the authoritative mode reduces this margin for potential change by requiring later courts to decide the case at hand in conformity with the precedent. Thus, any time a precedent-governed dispute arises, the content of the law will remain somewhat stable, because the later court will have an obligation to follow precedent. This situation fosters reliability: people know in advance what courts have an obligation to do when they face a precedent-governed dispute.

On the persuasive mode, precedents are not, strictly speaking, part of the law, because they lack one essential feature that most legal norms (claim to) have: to provide authoritative guidance. 97 Hence, it may seem that courts under this mode cannot change the content of the law when they hand down rulings. Although this is true, there can be grey areas, such as a legal system where the obligation of courts to follow precedent is not de jure but de facto . 98 This situation can be problematic, for though, under a persuasive mode, the content of the law cannot be changed—again strictly speaking—by how courts apply the law, courts can nevertheless lead people to think that applications of the law are themselves part of the law. Courts, in other words, can affect reliability (what people may think the content of the law is) without changing the law itself.

This situation can be problematic because that de facto obligation depends mostly on contingencies of the legal system in question. It may depend, for example, on the particular composition of a court at a given time. Judges under this contingent composition may treat the reason in favour of following precedent as a protected reason. This, in turn, may create a contingent pattern of deciding precedent-governed disputes in conformity with the precedent. But because this pattern is the result of a de facto contingency and not of a de jure obligation, the pattern may change, such as when the composition of the court changes. In this new scenario, judges may treat the reason in favour of following precedent as a weak, non-protected reason. This can affect reliability, for people rely on this de facto pattern instead of relying on the law itself. 99

Finally, in relation to equality, again the authoritative mode presents advantages vis-à-vis the persuasive mode. If the dispute at hand and the precedent are legally the same, then the later court has an obligation to hand down the same decision. In relation to the persuasive mode, there is no denying that if the dispute at hand and the precedent are legally the same, then both can be adjudicated alike. The problem, again, is that this adjudicatory treatment often follows as a matter of contingency, not obligation.

B. How Does Each Mode Avoid Replicating Substantively Incorrect Decisions?

Suppose, to simplify the argument, that in 2019 the Supreme Court of an imaginary country handed down a decision that many respected judges and scholars consider mistaken on good grounds. How would both modes treat such a decision?

The authoritative mode, governing (typically, but not necessarily) 100 the treatment that lower courts should give to the decision of the Supreme Court, would prevent these courts from rendering a decision different to that of the Supreme Court. For a lower court to hand down a substantively correct decision, the Supreme Court would have to overrule its precedent and issue a new decision that is substantively correct. This means that in those areas where the authoritative mode is in operation, lower courts can only reach substantively correct decisions contingently, as this correctness depends on whether the relevant precedents were correctly decided. If they were, then the later court would replicate a correct decision. If they were not, then the lower court would be required to replicate a substantively incorrect decision. 101

Accordingly, there are three options a lower court has when faced with a wrongly decided precedent: to distinguish; to follow the precedent and render yet another substantively incorrect decision; or to adjudicate under the persuasive mode. If a later court wants to distinguish, as we saw above, then it must provide reasons showing that the case at hand presents a novel fact that is relevant for the decision. Sometimes, however, this fact does not exist. When this happens, the later court cannot distinguish, meaning it must replicate a decision that is substantively incorrect. This is the second option the court has, and seems to be the option the legal system wants the court to adopt when the precedent cannot be distinguished. Finally, the later court could adjudicate under the persuasive mode, but there might be responsibilities for doing so, particularly if that court has a legal duty to adjudicate under the authoritative mode.

It seems, therefore, that the authoritative mode places more value on advancing the rule of law than on deciding cases correctly, at least where this correctness entails contravening established precedent. 102 By contrast, the weakness the persuasive mode showed in the last subsection can now be seen as virtues in relation to the value of deciding cases correctly. Under the persuasive mode, later courts do not have an obligation to follow precedent. Courts may outweigh the reason in favour of following precedent and act on a potentially more compelling reason, such as deciding the same dispute in a substantively correct way. 103 This is not to deny the possibility that sometimes courts can have reasons as well for following wrongly decided precedents—this is why I argued in favour of robust precedent-following. For example, if the issue under consideration is so sensitive, reaching a different outcome may affect reliability in ways that are more prejudicial to the legal system than deciding the case in a substantively correct manner.

There is a minimum requirement of the rule of law in regard to precedent, and I have called it the persuasive mode of precedential reasoning. The null model, however lawful, runs counter to the rule of law. Yet the question still remains open as to which of the two modes that are compatible with the rule of law should be preferred. I do not think there is a unique answer here; it depends on the history of the legal system, its legal tradition and practices, and the values people in that system hold dear. Possibly in a legal system that attaches significant value to having democratic legislation as the main source of law, the reasons for leaning towards the persuasive mode are more demanding. For a system preferring flexibility in the adjustment of the law, the authoritative mode might be the best answer.

I have posed two evaluative questions, but much more could be said. For example, there are reasons to think that the degree of adjudicatory power that later courts have to give legal significance to the facts of the case varies between the two modes. In the authoritative mode, later courts cannot alter the legal significance that earlier courts gave to the facts of the precedent. In the persuasive mode, this obligation does not exist, at least not strictly speaking. Finally, there is the question of whether the authoritative mode entails judicial law making—at least in the robust form of a law-making power. There might be situations where indeed it does, but I will have to postpone that analysis for another time.

Much of the literature has become increasingly technical, partly because of the valuable contributions of nonmonotonic logics and artificial intelligence. On the other hand, many scholars have focused exclusively on precedent in either the common law or the civil law. But even within one of these traditions we may find courts treating precedents differently. For example, an authoritative treatment in one common law country may appear as a persuasive treatment in another common law country.

I have borrowed part of this terminology from various contributions on the value of precedent-following. See eg D Lyons, ‘Formal Justice and Judicial Precedent’ (1985) 38 Vand L Rev 495; F Schauer, ‘Precedent’ (1987) 39 Stan L Rev 571, 595–602; L Alexander, ‘Constrained by Precedent’ (1989) 63 S Cal L Rev 1, 26 and 51; R Cross and JW Harris, Precedents in English Law (4th edn, Clarendon Press 1991) 11–12; N Duxbury, The Nature and Authority of Precedent (CUP 2008), ch 5; L Alexander and E Sherwin, Demystifying Legal Reasoning (CUP 2008) 35–7; J Waldron, ‘Stare Decisis and the Rule of Law: A Layered Approach’ (2012) 111 Mich L Rev 1, 31; N Varsava, ‘How to Realize the Value of Stare Decisis: Options for Following Precedent’ (2018) 30 Yale JL & Human 62, 70–3.

It goes without saying that this ad hoc terminology does not prevent other values from being part of the rule of law. The point is that, to the extent that there is considerable agreement in the literature in identifying stability, reliability and equality as rule-of-law values, we can call these values collectively—and for current purposes only—‘the rule of law’. cf text to nn 30, 31 and 32 below.

The names ‘authoritative’ and ‘persuasive’ are commonplace in the Anglo-Saxon literature on precedents, but their exact meanings vary. For ease of reference, I will follow this familiar terminology, hoping to provide a consistent understanding of what it means to treat precedents as either authoritative or persuasive. See n 74 below.

Many legal systems give erga omnes force to the decisions of specific courts. Thus, these rulings are universally authoritative—namely, they apply to all future cases that happen to be legally the same. Hence, we loosely say eg ‘this case is governed by precedent’. My notion of a precedent is prior to this universal authoritative force. A precedent may have pro tanto persuasive force, or it may have no force at all.

See K Stevens, ‘Reasoning by Precedent—Between Rules and Analogies’ (2018) 24 Legal Theory 216, 217.

The distinction between horizontal and vertical precedent does not play a crucial role in the valuable contributions of EM Wise, ‘The Doctrine of Stare Decisis’ (1975) 21 Wayne L Rev 1043; Lyons (n 2); Alexander (n 2); G Lamond, ‘Do Precedents Create Rules?’ (2005) 11 Legal Theory 1; J Horty, ‘Rules and Reasons in the Theory of Precedent’ (2011) 17 Legal Theory 1; Varasa (n 2); Stevens (n 6).

Horizontal stare decisis is the authoritative effect that precedents have on later courts of equivalent hierarchy. Vertical stare decisis , by contrast, is the authoritative effect that precedents of higher courts have on lower courts. In the literature, Frederick Schauer is one who has consistently maintained this distinction. Yet when it comes to the reasons for following precedent, he has generally focused on horizontal stare decisis —see eg Schauer (n 2) 578 fn 11; F Schauer, ‘Precedent’ in A Marmor (ed), Routledge Companion to Philosophy of Law (Routledge 2012) 130–2. In F Schauer, Thinking Like a Lawyer (Harvard UP 2012) 41–2, Schauer claims that ‘the justifications for precedential constraint [vertical stare decisis ] are fairly obvious’: ‘lower court judges are expected to follow the “instructions” of those courts above them in what the military calls the “chain of command”’. This ‘chain of command’, however, is not obvious. The relation between higher and lower courts need not be of adjudicatory hierarchy—it could be of administrative hierarchy. But if a relation of adjudicatory hierarchy exists, it is not clear whether following precedent amounts to obeying a military command . Be that as it may, nothing in Schauer’s work suggests that the reasons for having horizontal stare decisis cannot support vertical stare decisis as well. For a somewhat sceptical analysis, in US law, of the arguments in favour of vertical stare decisis , see E Caminker, ‘Why Must Inferior Courts Obey Superior Court Precedents?’ (1994) 46 Stan L Rev 817.

J Gardner, Law as a Leap of Faith (OUP 2012) 192.

HLA Hart, The Concept of Law (3rd edn, Clarendon Press 2012) 7.

cf R Pound, ‘Justice According to Law’ (1913) 13 Colum L Rev 696.

In England, J Raz, The Authority of Law (2nd edn, OUP 2009) 184 fn 8 writes: ‘In English Law [the doctrine of precedent is] in fact part of the rule of recognition.’ This claim is extended to all common law countries by S Legarre and JC Rivera, ‘Naturaleza y dimensiones del “ stare decisis ”’ (2006) 33 Revista Chilena de Derecho 109, 112.

As one anonymous referee has correctly pointed out, many civil law countries have shown a tendency to depart from the classic civilian understanding of precedent to one where courts, at least in certain matters, are required to follow precedent. Good examples are Colombia, Mexico and Italy; see, respectively, C Bernal Pulido, ‘El precedente en Colombia’ (2008) 21 Revista Derecho del Estado 81; V Fernández Fernández, ‘La justicia de los precedentes. ¿Invasión a la independencia y autonomía del juzgador?’ (2016) XXIX Revista de Derecho (Valdivia) 9; A Cadoppi, Il valore del precedente nel diritto penale (2nd edn, Giappichelli Editore 2014). This tendency has led some scholars to consider where the civil law and the common law intersect; see Cadoppi (ibid.) 199. The tendency is undeniable, at least with respect to some civil law countries. But I would hesitate to call this tendency a full-blown system of stare decisis . To begin with, in many civil law countries the binding effect of precedents is usually reserved for the decisions of specific courts only—typically, higher courts and constitutional courts. Moreover, this effect is piecemeal, ie applicable to specific areas of law, such as constitutional law and criminal law. In many countries, furthermore, there is still considerable disagreement as to whether precedents should bind later courts—for a list of scholars who, in Spain, reject this view, see T García-Berrio Hernández, ‘La Controversia sobre el Precedente Judicial: un Clásico del Derecho en Constante Renovación’ (2006) 4 Foro, Nueva Epoca 127, 145–6. Given this disagreement and variety of practices, and also bearing in mind that the argument in this part is not about what the civil law is or is not today, I will rely on a rather classic approach to precedent-following in the civil law. The argument, to be sure, is about what we may learn from such an approach, namely that the null model is conceptually possible.

In Europe, examples of this relative force are provided by: art 1355 of France’s Code Civil; s 325(1) of Germany’s Zivilprozessordnung; and art 2909 of Italy’s Codice Civile. In Latin-America, similar examples can be found in: art 3 of Chile’s Código Civil; art 218 of Uruguay’s Código General del Proceso; and art 17 of Colombia’s Código Civil.

There are exceptions that could be classified as follows: (i) non-parties that may also benefit from the authoritativeness of the decision—eg lawful successors; and (ii) courts whose decisions bind erga omnes , eg constitutional courts.

E Steiner, French Law: A Comparative Approach (OUP 2010) 91 (emphasis in original). See also D Pugsley, ‘Two Systems of Precedent’ (1981) 15 The Law Teacher 7, 7: ‘Indeed French courts cannot even simply rely on a precedent; they must justify their own decisions themselves.’

JH Merryman and R Pérez-Perdomo, The Civil Law Tradition (4th edn, Stanford UP 2019) 84. In a similar vein, see Legarre and Rivera (n 12) 112; B Spaić, ‘The Authority of Precedents in Civil Law Systems’ (2018) XXVII Studia Iuridica Lublinensia 27, 28.

In Italy, see M Taruffo and M La Torre, ‘Precedents in Italy’ in N MacCormick and R Summers (ed), Interpreting Precedents (first published 1997, Routledge 2016) 155: ‘A judgment not applying a relevant precedent may be lawful in itself, since its lawfulness depends on its consistency with the statutory and constitutional provisions, not on its coherence with precedents.’

A Peczenik, ‘The Binding Force of Precedent’ in MacCormick and Summers (n 18) 463. Based on the example of Poland, Peczenik takes ‘further support’ to be cases where eg a precedent adds weight ‘to other legally relevant arguments for the decision handed down’; ibid 462.

If courts can lawfully adjudicate disputes without paying attention, in a normatively significant way, to relevant precedent, then it is doubtful whether something in law’s nature requires the practice of following precedent. It could be objected that nothing in the nature of the civil law would require this practice, but it would be part of the common law. Against this view, see Duxbury (n 2) 34, for whom ‘the common law does not need the doctrine of precedent in order to function’. cf Wise (n 7) 1057–8.

See eg Raz (n 12) 224; B Tamanaha, On the Rule of Law. History, Politics, Theory (CUP 2004) 33–4; J Waldron, ‘The Concept and the Rule of Law’ (2008) 43 Ga L Rev 1, 6; Waldron (n 2) 3–4.

This is the view associated with the well-known accounts of L Fuller, The Morality of Law (Yale UP 1969) ch 2; Raz (n 12) 213–14; J Finnis, Natural Law and Natural Rights (2nd edn, OUP 2011) 270–3; J Rawls, A Theory of Justice (first published 1971, Harvard UP 1999) 235–43. See also Tamanaha (n 21) 34, 66–7, 71, 93–4, 96 and 119; Waldron, ‘The Concept’ (n 21) 6; J Raz, ‘The Law’s Own Virtue’ (2019) 39 OJLS 1, 3–5.

See eg Tamanaha (n 21) 119; T Bingham, The Rule of Law (Allen Lane 2010) ch 9; Waldron, ‘The Concept’ (n 21) 7–8; J Waldron, ‘The Rule of Law and the Importance of Procedure’ (2011) 50 Nomos 3, 6.

See eg Tamanaha (n 21) 102–4 and 109–11; Bingham (n 23) ch 7; J Waldron, ‘The Rule of Law’, The Stanford Encyclopedia of Philosophy (Summer edn, 2020) < https://plato.stanford.edu/archives/sum2020/entries/rule-of-law/ > accessed 20 June 2020.

See eg Tamanaha (n 21) 112–13.

See eg P Craig, ‘Formal and Substantive Conceptions of the Rule of Law: An Analytical Framework’ (1997) Public Law 467; Tamanaha (n 21) chs 7–9; Waldron, ‘The Rule of Law and the Importance of Procedure’ (n 23) 4–5; Waldron, ‘The Rule of Law’ (n 24); A Zanghellini, ‘The Foundations of the Rule of Law’ (2016) 28 Yale JL & Human 213, 214.

Tamanaha (n 21) 92.

Gardner (n 9) 198–204.

A famous example of this disagreement can be found in Raz, The Authority of Law (n 12) 210–11. On the disputed nature of the rule of law, see J Waldron, ‘Is the Rule of Law an Essentially Contested Concept (in Florida)?’ (2002) 21 Law and Philosophy 137, 159; P Burgess, ‘The Rule of Law: Beyond Contestedness’ (2017) 8 Jurisprudence 480, 481.

Tamanaha (n 21) 102; similarly, ibid 111.

There is room in the literature for grounding precedent-following in more robust conceptions of the rule of law. The legal philosophy of Ronald Dworkin is one alternative. See R Dworkin, ‘Hard Cases’ (1975) 88 Harv L Rve 1057, 1087–101; R Dworkin, Law’s Empire (Harvard UP 1986) ch 7. A Dworkinian view of stare decisis was articulated by S Hershovitz, ‘Integrity and Stare Decisis’ in S Hershovitz (ed), Exploring Law’s Empire: The Jurisprudence of Ronald Dworkin (OUP 2006) ch 5. A critical examination of Dworkin’s views on precedent can be found in D Pannick, ‘A Note on Dworkin and Precedent’ (1980) 43 MLR 36. Discussing Dworkin’s views on precedent, see also S Perry, ‘Judicial Obligation, Precedent and the Common Law’ (1987) 7 OJLS 215, 223–6.

Other considerations may support following precedent—see eg Duxbury (n 2) ch 5; Varsava (n 2). Among them is the epistemic guidance that later courts may receive from knowing how earlier courts have dealt with statutes that are vague or have an unclear semantic meaning. Similarly, we may find arguments relying on the overall efficiency of the system’s adjudicatory process. I will not explore these other approaches, but I do not think my account is incompatible with them.

One may account for stability, reliability and equality in terms of rule-of-law values by looking at some of the most influential accounts of this ideal—see eg n 22 above.

Stability seems valuable only when we have awareness of something remaining stable and can thus be relied upon for future decision making. Now, because some may think that stability has value in its own right, I have decided to keep stability and reliability as separate values. Nothing in the main argument turns on this point. That stability is instrumentally valuable to reliability is a view present in T Benditt, ‘The Rule of Precedent’ in L Goldstein (ed), Precedents in Law (Clarendon Press 1987) 91; Duxbury (n 2) 160–1. Traces of this view can also be found from Lord Mansfield in Bingham (n 23) 38 and Hayek in Tamanaha (n 21) 66.

See eg C Perelman, The Idea of Justice and the Problem of Argument (J Petrie tr, Routledge & Kegan Paul 1963) 38; H Kelsen, Pure Theory of Law (M Knight tr, 2nd edn, University of California Pres 1967) 141–2; P Westen, ‘The Empty Idea of Equality’ (1982) 95 Harv L Rev 537, 551; J Raz, The Morality of Freedom (Clarendon Press 1986) 220–1.

On this point, see eg K Winston, ‘On Treating Like Cases Alike’ (1974) 62 CLR 1, 22; E Chemerinsky, ‘In Defense of Equality: A Reply to Professor Westen’ (1983) 81 Mich L Rev 575, 579–80.

On the various duties of judges, see eg L Green, ‘Law and the Role of a Judge’ in K Ferzan and S Morse (eds), Legal, Moral, and Metaphysical Truths: The Philosophy of Michael S Moore (OUP 2016) ch 22.

Winston (n 37) 18. Winston seems to have in mind epieikeia —equity, a corrective to law’s generality. See Aristotle, ‘Nicomachean Ethics’ in J Barnes (ed), The Complete Works of Aristotle: The Revised Oxford Translation vol II (Princeton UP 1984) 1796 (1137 b 17–18).

The requirement of justification is particularly stringent when the relevant party has showed to the court that there is precedent regulating the matter at hand, but also if it happens that the later court is aware of this precedent—or others that are relevant.

Raz, ‘The Law’s Own Virtue’ (n 22) 15.

Finnis (n 22) 270.

Fuller (n 22) 33.

Raz, The Authority of Law (n 12) 229; Gardner (n 9) 192; L Green, ‘Should Law Improve Morality?’ (2013) 7 Criminal Law and Philosophy 473, 487.

Bingham (n 23) 8.

That is, I first articulated the null model by looking at specific features of the civil law tradition and then criticised it on the basis of the rule of law.

On the idea of a second-order reason, see text to nn 64, 65 and 66 below.

Some of which is provided in n 2 above; but see in particular Duxbury (n 2) 159: ‘The point … is not that precedent-following is supportable because stability and closure are intrinsically good objectives, but that if judges consider these objectives desirable then precedent-following is one way by which they can pursue them.’

See Lyons (n 2) 511–12; Schauer, ‘Precedent’ (n 2) 597–8; Duxbury (n 2) 163.

Or part of an exceptional set of decisions. This suggests that fostering the rule of law is more a matter of degree than of kind, insofar as a legal system may tolerate exceptions to the various rule of law desiderata . cf T Endicott, Vagueness in Law (OUP 2000) 191.

As Schauer, ‘Precedent’ (n 2) 598 puts it: ‘maintaining a serious regime of precedential constraint entails some number of suboptimal decisions’. In a similar vein, see R Kozel, Settled Versus Right: A Theory of Precedent (CUP 2017) 6.

A Marmor, ‘Should Like Cases Be Treated Alike?’ (2005) 11 Legal Theory 27, 29.

[1966] 1 WLR 1234 (emphasis added).

cf Hart (n 10) 128; L Duarte d’Almeida, Allowing for Exceptions (OUP 2015) 184.

According to Duxbury (n 2) 128, ‘only rarely since 1966 has the House of Lords used the power that it created for itself’. On the House of Lord’s Practice Statement, see ibid 123–49; W Twining and D Miers, How to Do Things with Rules (5th edn, CUP 2010) 283–5.

Such as a significant number of courts in the common law tradition.

cf n 51 above; Stevens (n 6) 219.

K Stevens, ‘Case-to-Case Arguments’ (2018) 32 Argumentation 431.

Raz, The Morality of Freedom (n 36) 388; R Chang, ‘Hard Choices’ (2017) 3 Journal of the American Philosophical Association 1, 17.

This distinction is based on Stevens, ‘Case-to-Case Arguments’ (n 58) 435–7, who distinguishes between weak and strong versions of the principle ‘treating like cases alike’. Now, one should be careful not to conflate precedent-following with treating like cases alike or to ground the former in the latter. Both have been a debated topic in the literature. See eg Lyons (n 2); F Schauer, ‘On Treating Unlike Cases Alike’ (2018) 33 Constitutional Commentary 437.

Schauer, ‘Precedent’ (n 2) 575; cf Schauer, Thinking Like a Lawyer (n 8) 90; Varsava (n 2) 76; Stevens (n 6) 219; Kozel (n 51) 6.

Stevens, ‘Case-to-Case Arguments’ (n 58) 436–7.

See J Raz, Practical Reasons and Norms (first published 1975, OUP 1999) 39.

ibid; Raz, The Authority of Law (n 12) 17.

Raz, The Authority of Law (n 12) 18.

Chang (n 59) 17 (emphasis added).

Raz, The Morality of Freedom (n 36) 388.

This proviso is important. It means that there are no relevant differences between the two cases that would warrant a different treatment.

Duxbury (n 2) 128.

This commitment may well constitute part of the duty that judges assume when they are sworn in. I thank Crescente Molina for this suggestion.

cf L Green, ‘Law as a Means’ in P Cane (ed), The Hart–Fuller Debate in the Twenty-First Century (Hart Publishing, 2010) 22; Green, ‘Law and the Role of a Judge’ (n 38) 335.

See text to n 66 above.

See n 4 above. In my view, what is either authoritative or persuasive is not the precedent itself—a precedent is a non-normative fact. Only by virtue of a moral consideration can a precedent be a pro tanto normative fact. That is why I prefer to speak of the modes of precedential reasoning as being either authoritative or persuasive.

As a minimum methodological requirement, any sound account needs to pick central features of the practice it wishes to analyse; see Lamond (n 7) 15; A Rigoni, ‘Common-Law Judicial Reasoning and Analogy’ (2014) 20 Legal Theory 133, 135–6; Stevens, ‘Reasoning by Precedent’ (n 6) 218–19. But once these central features are picked, there is room for a sensible detachment from practice—see n 76 below.

The methodological approach I have adopted is vital for sorting out possible factual counterexamples. To be sure, any descriptive account of what courts do—eg how they reason using precedents—is potentially subject to counterexamples. Consider the following case: Perry (n 32) 230–4 attributed to Raz a ‘positivistic’ account of precedent-following—one where later courts have exclusionary reasons to follow precedent. He then criticised this positivistic framework on grounds of descriptive inaccuracy—many common law courts distinguish in ways that are incompatible with Raz’s proposal. For example, they drop conditions for the ratio ’s application—ibid 239. In a nutshell, according to Perry, legal positivism (at least as presented by Raz) fails to account for what many common law courts do when they reason using precedents. Perry then offered another possible understanding of the practice . But Perry’s objection would misfire if there are some common law courts in the world that treat precedents as giving rise to exclusionary reasons. His objection would backfire if his own account is affected by a set of plausible counterexamples. Both options are always possible, given the large field of ongoing adjudication. A sensible detachment from practice is beneficial because it allows us to identify modes of precedential reasoning that can exist independently of a particular legal tradition.

Stevens, ‘Reasoning by Precedent’ (n 6) 218 characterises the situation of a later court that has not found a non-excluded reason against following precedent as having a binding reason to follow precedent. When, by contrast, such a court has an authoritative reason, the court can either follow or distinguish the precedent by acting on a non-excluded reason.

For current purposes, we can follow Raz on this point; see eg J Raz, ‘Promises and Obligations’ in PMS Hacker and J Raz (ed), Law, Morality, and Society: Essays in Honour of HLA Hart (Clarendon Press 1977) 223, arguing that ‘An action is obligatory if it is required by a categorical rule, ie a mandatory rule which applies to its subjects not merely because adherence to it facilitates achievement of their goals’. Later, in Raz, The Authority of Law (n 12) 234–5, Raz seemed to refer to such a categorical rule in terms of a protected reason: ‘I have argued elsewhere [ie the work just cited] that “an action is obligatory only if it is required by a protected reason”.’

Raz, The Authority of Law (n 12) 185; Lamond (n 7) 3; Schauer, Thinking Like a Lawyer (n 8) 58.

See eg Lamond (n 7) 15, arguing that ‘Any satisfactory view of precedent must see distinguishing as an integral part of it rather than as an add-on or quirk of the common law’.

Notably Alexander and Sherwin (n 2) 122: ‘The rule model of judicial decision making has no conceptual room for the practice of distinguishing rules’; ibid 123: ‘distinguishing rules is logically equivalent to repealing rules at will’.

Lamond (n 7) 1, 3, 16, 23 and 25.

See Stevens, ‘Reasoning by Precedent’ (n 6) for the current state of this debate.

By treating certain established facts as material, I mean making them relevant for deciding the present case. By giving a particular legal significance, I mean deciding whether a specific fact, previously treated as material by the court, favours one of the parties. By deciding on the balance of reasons, I mean establishing the necessary and sufficient conditions for reaching a justified decision.

Horty (n 7).

There is significant literature on the various tests for discovering the ratio decidendi of a case. Much of this literature dates back to the first half of the twentieth century—and even before. By then, as is well known, legal realists like H Oliphant and JF Francis offered revolutionary views about the ratio of a precedent—many of them attacking its supposed determinacy. Authors such as Arthur Goodhart offered tests where the purpose was to make the ratio more determinate, so that the distinction between the ratio and the obiter of a case may be useful in explaining precedential constraint. I unfortunately cannot address this vast literature here. For a helpful study, see Cross and Harris (n 2) ch II; Duxbury (n 2) 67–92; Twining and Miers (n 55) 304–12.

Horty (n 7) 13.

In relation to the common law, Lamond (n 7) 19 attempts more precision by saying that ‘the ratio would provide the basis for the first-order part of the protected reason. The other facts of the case would provide the basis for the exclusionary reason; later courts are excluded from relying on reasons provided by features that were present in the precedent case to defeat the first-order reason for the result.’ Recently, Robert Mullins, ‘Protected Reasons and Precedential Constraint’ (2020) 26 Legal Theory 40 has defended an account of what he calls ‘the protected reason model of precedent’. See eg ibid 52–3: ‘On the protected reason model developed in the previous section, a past decision binds a subsequent decision-maker when the rule relied on in the previous decision is applicable to the factual scenario that applies in the subsequent case. The decision-maker must take the ratio of the previous decision as a reason for deciding in the same way and a reason for excluding otherwise applicable reasons present in these past cases from acting as the basis for an inconsistent judgment.’

With the proviso of n 95 below. There is also the question of whether a precedent could lose its erga omnes force prospectively only . See Twining and Miers (n 55) 285 fn 48, discussing Re Spectrum Plus Ltd [2005] UKHL 41, where the House of Lords concluded that, in order to do so, it would have to be an exceptional case.

See text to n 53 above.

See Twining and Miers (n 55) 284 for some instances in which the House of Lords has acted on this non-excluded reason.

[1944] KB 718.

The argument needs to be qualified. According to Twining and Miers (n 55) 286–7, the Court of Appeals has restated aspects of the doctrine laid down in Young v Bristol Aeroplane , allowing for departures from relevant precedent when ‘the previous decision is manifestly wrong’.

I think G Lamond, ‘Persuasive Authority in the Law’ (2010) XVII Harvard Review of Philosophy 16, 17 would agree with me: ‘Obviously the second approach [ie that certain sources have content-independent normative weight] does not rule out considering the persuasiveness of the view. The view could carry even more weight where it is found to be convincing ’ (emphasis added).

Strictly speaking, however, authority and legal validity are very different ideas. Whether a norm is authoritative depends—at least for current purposes—on whether it provides a protected reason for action; see eg Raz, The Authority of Law (n 12) 21–5. By contrast, whether a norm is legal depends on whether it meets the criteria for legal validity given by the rule of recognition. This makes room for the possibility that a precedent could be ‘authoritative’ (in a loose sense—see n 74 above) without being part of the law. But we need not make too much out of this conceptual possibility. Raz, The Authority of Law (n 12) has argued that, at least in England, the rule of recognition gives legal validity to authoritative precedents. There is no reason to think that the same could happen with respect to other legal systems—cf Legarre and Rivera (n 12).

cf Raz, The Authority of Law (n 12) 185; Lamond, ‘Do Precedents Create Rules?’ (n 7) 17; Gardner (n 9) 80.

See eg Raz, The Authority of Law (n 12) ch 2. An excellent elaboration of law’s claim to authority can be found in Gardner (n 9) ch 5. The qualification ‘strictly speaking’ (in the text) is important because there could be a grey area, which I cannot explore here, where norms that provide a reason only, as opposed to a protected reason, could nonetheless become legal norms. This will depend on other, possibly complementary jurisprudential views about law’s authoritativeness. See eg A Marmor, ‘Soft Law, Authoritative Advice and Non-binding Agreements’ (2019) 39 OJLS 507.

In relation to the civil law, see Merryman and Pérez-Perdomo (n 17) 47; Peczenik (n 19) 461 and 465–6. I believe a real tension may exist between de jure and de facto precedent-following when a statutory provision explicitly provides that judicial decisions are binding on the parties to the dispute only —as in Chile; see n 14 above.

There is no denying that the law may offer less guidance, and thus affect reliability, without authoritative applications of it binding later courts.

See the final remark in section 5A.

As correctly noted by Duxbury (n 2) 24, Waldron, ‘Stare Decisis’ (n 2) 8 and Stevens, ‘Reasoning by Precedent’ (n 6) 219, all writing in relation to the common law.

cf G Postema, ‘Some Roots of our Notion of Precedent’ in Goldstein (n 35) 31.

In relation to the civil law, see Merryman and Pérez-Perdomo (n 17) 85.

Author notes

Email alerts

Citing articles via.

• Recommend to your Library

Affiliations

• Online ISSN 1464-3820
• Print ISSN 0143-6503
• Copyright © 2024 Oxford University Press
• About Oxford Academic
• Publish journals with us
• University press partners
• What we publish
• New features  
• Open access
• Institutional account management
• Rights and permissions
• Get help with access
• Accessibility
• Advertising
• Media enquiries
• Oxford University Press
• Oxford Languages
• University of Oxford

Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide

• Copyright © 2024 Oxford University Press
• Cookie settings
• Cookie policy
• Privacy policy
• Legal notice

This Feature Is Available To Subscribers Only

Sign In or Create an Account

This PDF is available to Subscribers Only

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.