Separation of powers

Separation of powers is not a British doctrine in the same sense as it is an American doctrine. There is no single constitutional text allocating legislative, executive and judicial power to three sealed institutions, and no general rule that a person who exercises one kind of public power may never exercise another. The United Kingdom constitution has always contained overlaps: ministers sit in Parliament; delegated legislation is made by the executive; judges develop the common law; Parliament may legislate on matters arising from litigation. The first task in a Durham answer is therefore to avoid the false premise that the UK either has, or lacks, separation of powers in a crude all-or-nothing sense.

The better formulation is institutional and functional. The constitution distinguishes between law-making, law-executing and law-adjudicating functions; it distributes them among Parliament, ministers and public authorities, and courts; and it supplies doctrines which police abusive encroachment. Those doctrines include parliamentary sovereignty, responsible government, judicial independence, access to courts, legality, justiciability, parliamentary privilege, and the supervisory jurisdiction of the High Court. Separation of powers is thus not a free-standing trump card. It is a structural principle which explains why certain legal rules exist and why some exercises of power are constitutionally suspect.

This week builds directly on the preceding four weeks. From Week 1, the point is that separation of powers in the UK is part of an uncodified constitution: it appears through statutes, conventions, common law principles and institutional practice. From Week 2, parliamentary sovereignty explains why courts cannot ordinarily invalidate Acts of Parliament, while also raising the question whether legislative supremacy is compatible with judicially enforced limits on the executive. From Week 3, constitutional statutes complicate the hierarchy: the Constitutional Reform Act 2005 and Human Rights Act 1998 do not entrench a US-style separation, but they alter institutional relationships in ways that ordinary implied repeal analysis cannot fully capture. From Week 4, Bingham’s account of the rule of law supplies the normative centre: public power must be exercised lawfully, independently adjudicated, and not left to executive discretion alone.

In assessment, the doctrine is usually examined in one of three forms. First, an essay may ask whether there is a separation of powers in the UK constitution. Such a question requires rejection of the pure Montesquieu model, followed by a careful account of partial separation, checks and balances, and fusion under responsible government. Secondly, a question may focus on the courts: whether judicial review, declarations of incompatibility, or prorogation review amount to legitimate constitutional adjudication or judicial overreach. Thirdly, a problem question may concern government action said to frustrate Parliament, evade judicial scrutiny, or interfere with the courts. The answer should identify the institutional actor, the source of power, the constitutional principle engaged, and the remedy. The strongest answers treat separation of powers as an organising principle: not a slogan, not a substitute for authority, but a way of connecting cases which otherwise appear doctrinally scattered.

The standard starting point is Montesquieu, but a British constitutional lawyer must use him cautiously. Montesquieu admired what he took to be the English constitution because liberty was protected by distributing power. His influence is real, but the UK constitution did not adopt a strict tripartite scheme. Its history is instead the gradual subjection of prerogative, administration and adjudication to law, Parliament and institutional convention.

The seventeenth century is decisive. The constitutional struggles of the Stuart period were not abstract debates about institutional neatness; they concerned taxation, military power, imprisonment, dispensing powers and the relation between royal government and law. The Petition of Right 1628, the Civil War, the Restoration settlement, the Glorious Revolution, the Bill of Rights 1688 and the Act of Settlement 1700 all reduced the idea that government could stand above law. Article 9 of the Bill of Rights protected parliamentary proceedings from external judicial or executive interference. The Act of Settlement secured judicial tenure during good behaviour, a foundation of judicial independence. These instruments did not create democratic modernity, but they altered the constitutional grammar: the Crown governed through law; Parliament controlled supply and legislation; judges were not to be dismissed merely for displeasing government.

The eighteenth-century settlement retained executive power but gave it a parliamentary form. Cabinet government emerged from royal administration. Ministers became politically answerable to Parliament, especially the House of Commons. This is the great British fusion: the executive is not separated from the legislature in personnel, because ministers are drawn from Parliament and depend on Commons confidence. That fusion is often misunderstood. It is not the absence of separation of powers; it is a different means of controlling executive power, by political responsibility rather than by formal institutional isolation. A Durham answer should not describe fusion as a defect unless it also asks what substitute mechanism is being assumed.

At the same time, the common law courts developed a durable conception of legality. Entick v Carrington is the classic illustration: executive officials who trespassed on private property could not justify their action by reference to state necessity or general warrants unsupported by law. The case is often cited for the rule of law, but it is equally separationist: executive power requires legal authority, and courts determine whether that authority exists. The legality of government action is not finally determined by the government itself.

The nineteenth and twentieth centuries added the administrative state. Parliament legislated at scale; ministers and departments acquired wide statutory powers; delegated legislation became indispensable; tribunals and regulators proliferated. A pure separation model became still less plausible. The constitutional problem changed: how can courts police legal limits without displacing administration, and how can Parliament authorise administration without surrendering meaningful control? Judicial review became the principal legal answer, while parliamentary questions, select committees, ministerial responsibility and elections remained the principal political answers.

The modern period is marked by renewed attention to judicial independence and constitutional adjudication. The Human Rights Act 1998 required courts to interpret legislation compatibly with Convention rights so far as possible, and to issue declarations of incompatibility where primary legislation could not be read compatibly. The Constitutional Reform Act 2005 recast the office of Lord Chancellor, created the Supreme Court of the United Kingdom, and placed a statutory duty on ministers to uphold judicial independence. These reforms did not abolish parliamentary sovereignty; nor did they create a constitutional court able to annul Acts of Parliament. They did, however, make visible a separation that had previously been obscured by the mixed office of the Lord Chancellor and by the Appellate Committee of the House of Lords.

The historical lesson is therefore twofold. First, the UK constitution has never been built on absolute separation. Secondly, the rejection of absolute separation has never meant acceptance of executive self-judgment. The executive acts under law; Parliament legislates and holds ministers to account; courts adjudicate legality and rights; and each institution is restrained by a mixture of law, convention and political consequence.

The first principle is functional differentiation. Legislative power concerns the creation, amendment and repeal of general legal norms. Executive power concerns the formulation and implementation of policy, the administration of statutory schemes, the conduct of foreign affairs, and the use of prerogative or statutory authority. Judicial power concerns the authoritative determination of legal disputes by independent courts according to law. These categories are not watertight, but they matter. A court that applies a statute is not legislating merely because interpretation involves choice. A minister who makes regulations under an enabling Act is not Parliament, but exercises delegated legislative power subject to parliamentary and judicial control. Parliament deciding the content of law is not an ordinary litigant, but it may legislate in ways that affect pending litigation, subject to political and human rights constraints.

The second principle is that UK separation is partial rather than pure. The most obvious fusion is between legislature and executive. The Prime Minister and most senior ministers sit in one of the Houses of Parliament. Government normally controls Commons time and, where it has a majority, much of the legislative programme. This creates a danger of executive dominance of Parliament. Yet it also supplies the mechanism of responsible government: ministers must explain, defend and, in extreme cases, resign for their conduct; the government must maintain the confidence of the Commons; and political scrutiny is concentrated in a representative institution. Separation of powers in the UK is therefore not merely about keeping ministers out of Parliament. It is about ensuring that ministerial power remains accountable to Parliament and reviewable by courts.

The third principle is judicial independence. This has personal, institutional and decisional dimensions. Personal independence concerns tenure, remuneration and security from improper removal. Institutional independence concerns the courts’ organisation, resourcing and protection from executive manipulation. Decisional independence concerns freedom from pressure in particular cases. Section 3 of the Constitutional Reform Act 2005 gives statutory expression to these commitments by requiring ministers and others concerned with the judiciary or administration of justice to uphold continued judicial independence, and by prohibiting ministerial attempts to influence particular decisions through special access. Judicial independence is not a privilege of judges. It is a right of the citizen to have disputes with the state determined by an impartial legal institution.

The case law is best read as a series of constitutional boundary disputes rather than as a single doctrine. Entick v Carrington supplies the classical common law limit on executive power. State officials entered and searched Entick’s premises under a warrant issued by the Secretary of State. The court held that no such power existed unless authorised by law. The lasting point is not merely protection of property; it is that executive assertion does not create legal authority. Courts decide legality.

M v Home Office applied that insight to modern government. A minister could be subject to contempt proceedings where the Home Office failed to comply with a court order. The executive’s public character did not place it above law. The case is important for separation of powers because it rejects the idea that courts may declare the law only against private persons while ministers comply as a matter of grace.

R v Secretary of State for the Home Department, ex p Fire Brigades Union concerned a statutory criminal injuries compensation scheme which had been enacted but not brought into force. The minister introduced a different tariff scheme under prerogative or administrative power. The House of Lords held the conduct unlawful. A minister with commencement discretion could not use executive power to defeat the policy of an Act which Parliament had enacted. This is a particularly strong case on executive frustration of legislative will.

A v Secretary of State for the Home Department addressed anti-terrorism detention under the Anti-terrorism, Crime and Security Act 2001. The House of Lords held that indefinite detention of foreign suspected terrorists was incompatible with Convention rights. The case illustrates the judicial role under the Human Rights Act: the courts did not invalidate the statute, but they issued a declaration of incompatibility and subjected national security legislation to legal scrutiny. The separationist significance lies in refusing both extremes: courts defer to political branches on some evaluative matters, but they do not abdicate rights adjudication.

Miller 2017 held that ministers could not use prerogative powers to notify withdrawal from the European Union where doing so would alter domestic law and remove statutory rights. The decision follows orthodox separation reasoning: foreign affairs are generally executive, but changes to domestic legal rights require Parliament. Miller/Cherry 2019 extended boundary-policing to prorogation. The Supreme Court held that advice to prorogue Parliament was justiciable and unlawful where it had the effect of frustrating or preventing Parliament’s constitutional functions without reasonable justification. The case remains controversial because it brought courts into a politically charged field, but doctrinally it rests on preserving Parliament’s ability to perform its legislative and scrutiny functions.

Privacy International and UNISON show the contemporary strength of judicial review. Privacy International concerned an ouster clause and the ability of courts to review errors of law by the Investigatory Powers Tribunal. The Supreme Court was reluctant to read legislation as excluding the supervisory jurisdiction unless Parliament used unmistakable language. UNISON invalidated employment tribunal fees because they effectively impeded access to justice. Both cases are separationist in a deeper sense: courts must remain available to determine legal rights, otherwise executive and administrative legality becomes optional.

The lesson across the cases is not judicial supremacy. It is that each branch is legally bounded. Parliament may legislate; ministers must act within legal authority; courts determine legality and protect access to adjudication; and political questions are not automatically non-legal merely because they are constitutionally important.

See practice questions section below.

See further reading section below.