American Rights Act

The American Rights Act, 1866 is the statutory bill of rights of the United Kingdom of America. It states the liberties of the person in enforceable statutory form, binds the Crown and every public authority to them, prescribes the conditions on which they may be limited, and provides the remedies a court may grant when they are infringed. Where the Charter of Union of the American Colonies declares rights in the language of settlement, the Act operates them.

It is a supreme statute: an ordinary Act of Parliament in form, repealable in form by an ordinary majority, and treated by the Realm’s courts and by Parliament itself as constitutional in weight. It belongs to the small body of legislation the Realm regards as foundational. The Charter is entrenched, as are the core provisions of the Parliamentary Reform Act 1859 under the Eighth Amendment of 1924. The Rights Act remains an ordinary statute; emancipation was first enacted by the Emancipation Measure and then entrenched by the Fourth Amendment of 1873.

As enacted in 1866 the Act bound federal authority alone. It did not reach a polity legislature, a polity court, a county registrar or a sheriff, and it did not reach them for ninety-six years. The limitation was not written as a proviso or a saving clause. It was written into the definition of public authority at s.4.2, in the interpretation section, where the most consequential clause in the Realm’s domestic history sat for a century looking like drafting. It was removed by the American Rights (Extension) Act 1962, and everything the Realm now understands the Rights Act to mean dates from that Act rather than from 1866.

The American Rights Consolidation Act 2008 is the most recent layer, folding the security powers taken after the May Day Bombings into the Act’s framework of warrant, record, proportionality and review. The text reproduced below is the Act as it stands after that consolidation, with the wording of 1866 and 1962 given alongside wherever it has changed.

Constitutional character

A supreme statute

The Act is constitutional in weight and ordinary in form, and the Realm has never resolved the tension between those two facts because it has never had to.

Nothing entrenches it. Parliament may repeal the American Rights Act tomorrow by the same majority that passes a finance bill, and no court could prevent it. What the Realm has instead is a set of conventions and interpretive rules, some of them written into the Act itself, which together make repeal by inadvertence impossible and repeal by design politically unthinkable. The Act cannot be amended by implication (s.1.4): a later statute inconsistent with it is read, so far as the words permit, to conform to it, and where the words do not permit, the later statute must have said so expressly and in terms. No ministry since 1962 has been willing to write those terms.

The Realm’s constitutional writers describe this as protection by embarrassment and note that it has held for a hundred and sixty years, which is longer than a good many entrenched constitutions have lasted.

The perennial proposal is to move the Act into the Charter, so that a court could void an Act of Parliament that breached it instead of only declaring it incompatible. A private member’s bill for a Fifteenth Amendment to that effect has been introduced by Rosa Delgado, a Labour and Social Democratic member from Illinois, and members of four other parties have signed it. The ministry, whose business the coalition’s Heads of Agreement confine, has said that it cannot take the bill up and will support a motion sending the question to a select committee. Like every amendment, it would need both Houses, a double-majority referendum and Assent.

Relationship to the Charter

The Charter declares; the Act operates. Article IX of the Charter of 1783 sets out the ancient liberties of the subject in the language of a settlement between a King, a Parliament and a Congress, and it provides no mechanism whatever for enforcing them against an officer. The Act supplies the mechanism.

The relationship is stated in the Act (s.1.2) and is not one of subordination in the ordinary sense. The Act does not derive its authority from the Charter and does not claim to. It is an Act of Parliament, and Parliament’s authority to pass it comes from the Charter’s grant of legislative power rather than from Article IX. What the Act does is name Article IX as the source of the liberties it secures, which makes the two instruments read together and gives the Act’s interpretation a fixed point older than itself.

Where the two conflict, the Charter prevails and the question has never arisen.

The Fourth Amendment

The Fourth Amendment of 1873 entrenched the abolition of slavery and a guarantee of equal protection in the Charter itself, seven years after the Act, and the relationship between the two instruments is more awkward than the Realm’s schoolbooks suggest.

The Amendment is entrenched and the Act is not, so on any conflict the Amendment governs. What the Amendment does not do is supply machinery: it declares equal protection in a sentence and provides no definition, no test, no remedy and no procedure, which is precisely the defect the Act had been passed to cure in respect of Article IX. The consequence is that the Realm has an entrenched equality guarantee with nothing attached to it and an unentrenched equality Part with everything attached to it, and litigants plead both.

Until 1962 the Amendment guaranteed equal protection against the Union alone, and the Act likewise bound federal authority alone. Neither reached a polity’s own authorities. The Eleventh Amendment of 1962 extended Article IX against the polities, and the American Rights (Extension) Act of the same year supplied the Act’s machinery against them. The distinction between entrenching a guarantee and supplying its remedies remained.

How the Act is cited

The Act is cited by section and subsection, so that s.2.2 or s.9.4 is an ordinary phrase in a courtroom, a leading article or a school examination. The Parts are used for navigation and are not usually cited. Sections are numbered continuously through the Act from 1 to 68 rather than restarting within each Part, an arrangement adopted in 1866 and preserved through both later Acts specifically so that a section number would mean the same thing for ever.

The consequence is that the numbering carries its own history. Section 4.2 is a live section with a spent subsection, section 41 to section 45 did not exist before 1962, and the sections of Part V arrived one at a time across the twentieth century, which is why their language is not uniform.

History

The Everett memorandum

Shortly before his death on 15 January 1865, Governor-General Edward Everett sent Lincoln a private memorandum endorsing equal civil personhood, rights of contract and testimony, protection in the courts, and federal remedies against local denial.

It was published after his death. It was not a draft, not a Crown programme and not an enactment in advance, and Everett was careful that it should be none of those things, having spent his whole vice-regal term refusing to make the Crown a weapon of either sectional faction. What it did was identify, in four pages and before anybody else had set it out in one place, the specific problems that a statute would have to solve: that emancipation had made four million people free without making them persons in law, that a freedman who could not contract, testify, sue or hold property was not free in any sense a court could enforce, and that the denial would come from local institutions rather than from federal ones.

The last of those observations is the one the Act failed to act on, which is the reason the memorandum is read now with a certain amount of pain.

Enactment

The Act passed in 1866, on the third of the three votes of confidence the Lincoln ministry staked on the war settlement, after conscription and emancipation. The ministry carried it narrowly.

Governor-General Salmon P. Chase regarded it as the necessary legal completion of emancipation and granted assent immediately on Lincoln’s advice, in a public constitutional ceremony. In private audience he had put technical questions to Lincoln without drafting, directing or championing the bill. Whether the Act reached a polity sheriff was a question Chase did not have to answer and his successors did.

The debate on the bill ran for eleven weeks and is the best-recorded parliamentary argument of the century, because both Houses knew what was at stake and neither pretended otherwise. The ministry’s case was made chiefly by its law officers and rested on a single proposition: that emancipation without civil standing was a fraud upon the people emancipated, and that the Realm had abolished a status without conferring one. The opposition’s case was not a defence of slavery, which was gone and which the Lords had not obstructed. It was a defence of the polities, and it was argued by men who had voted for emancipation and who held that a statute stating rights enforceable against public authority was a federal intrusion into internal government that the Charter did not permit.

Both propositions were serious and the Realm has never decided between them, because the compromise that resolved the Bill made deciding unnecessary. The ministry got a statute. The Lords got a definition. Nobody had to say out loud that the two were incompatible.

Frederick Douglass was the first Governor-General installed after enactment, and his oath of 1873 was the first to name the Crown in right of America, the Charter and the Rights Act together. During Reconstruction he treated the Act as an enforceable national guarantee, pressed ministers where local institutions denied protection, and left enforcement to responsible government.

The clause that was struck out

The bill as introduced contained a clause, immediately following the definition of public authority, which would have bound the polities. It was struck out in the Lords, and the Realm has the text because the bill as introduced was printed and survives.

The clause was the point of the bill and its removal was the price of the bill, and the men who accepted the price said so at the time. What replaced it was nothing: the clause was not amended, softened or made conditional, and no substitute was offered. It was taken out, and the definition at s.4.2 was left to carry the consequence.

Ninety-six years later the Extension Act put the substance back, in modern language and with machinery the 1866 clause did not have. It is not the same text. It does the same work.

The two are worth reading against one another. The 1866 clause is a single sentence ending in the old formula any law, usage, constitution or practice of such polity to the contrary notwithstanding, which is the language of a Parliament asserting supremacy and daring a polity to test it. The 1962 definition is a list of twenty-odd named bodies. The change of method is the change of century: the drafters of 1866 wrote a proposition and expected the courts to apply it, and the drafters of 1962 had watched a hundred years of propositions being construed narrowly and wrote something that could not be.

The federal limit

The limitation was accomplished by definition rather than by saving clause, and the choice of method mattered.

A saving clause states an exception, and an exception invites the question why. A definition states what a word means, and a reader who is not looking for the answer will not find the question. The Lords’ amendment therefore did not say that the Act should not bind the polities. It said that in the Act, public authority means the Crown in right of the federal government and the officers, ministries, courts, forces and bodies of that government. Every operative section of the Act binds public authorities. None of them was altered at all.

What the limit produced

The Act’s guarantees ran against federal authority and nowhere else. A federal officer could not do the things it forbade, and a federal court would strike down a federal statute that offended it. A polity legislature, a polity sheriff, a county registrar and a polity court were outside its scope entirely.

The former rebel polities accordingly built, across about twenty years, a complete apparatus of exclusion that broke no provision of the Act, because the Act had been drafted not to reach them. Nothing had to be defied. Registration rules, jury selection, poll administration, vagrancy and labour offences, sheriffs’ discretion and private violence a county declined to prosecute all sat squarely within a polity’s own competence. The loyal slaveholding polities, which had escaped direct rule and therefore escaped supervision, arrived at the same position more slowly and by the same route.

Where the limit ran the other way

In two places the limitation operated in the opposite direction, and both are recorded because they show precisely how narrow the Act’s reach was.

The American Postal Service is exclusively federal under the Charter, so no polity could segregate a federal counter. For most of a century the post office was the only unsegregated public building in a great many counties, and its savings bank was where a family refused by the local bank could open an account.

And no federal body could author exclusion either. The Realm never produced the mortgage-lending maps that a federal agency elsewhere might have drawn, because a federal agency directing credit away from Black neighbourhoods would have been unlawful on the face of the Act from 1866. What the Realm got instead was the same exclusion conducted privately and at polity level, which nothing reached.

The extension of 1962

The American Rights (Extension) Act 1962 repealed the words of limitation in s.4.2, substituted a definition of public authority reaching every public body in the Realm, and added Part VI on the Indigenous nations, which the 1866 Act had not mentioned at all.

It is the single most consequential piece of domestic legislation the Realm has passed since the Act it amended. It did in one statute what a century of litigation had failed to do, and everything the Realm now understands the Rights Act to mean dates from 1962. The popular belief that the Second Founding settled the question in the 1860s is among the most widespread historical errors in the country, and it is repeated in good faith by people who have read the Act, because the Act as it now stands does not obviously carry its own history.

The passage of the Extension Act

The Extension Act was not the work of a single ministry and did not pass easily.

Its immediate causes were the desegregation judgments of 1948 and 1949 and the fifteen years of enforcement litigation that followed them. The judgments had established the principle and could not deliver it, because each ran against the particular authority before the court and the next authority had to be sued afresh. By the late 1950s the Realm’s courts were processing school district cases at a rate that made plain what the litigants had been saying since Shelley, which was that the remedy was legislative and had been legislative since 1866.

The schools crisis at Little Rock in 1957 to 1959 supplied the political conditions. A polity ministry deploying a guard against nine children, a Crown representative refusing to sign, and federalised troops enforcing a court order were between them the clearest possible demonstration that the Realm’s rights settlement did not reach the places it needed to reach, and the arguments made against extension after 1959 were noticeably weaker than those made before it.

The Bill was introduced in 1961, and the argument against it was the same argument that had been made in 1866, which its opponents said openly and which did them no good at all. Its passage required a guillotine in the Commons and consumed most of a session in the Lords. The final vote in the Lords was closer than the Commons vote by a considerable margin, and a substantial number of peers who voted against it did so on the ground, sincerely held and constitutionally respectable, that a Parliament which could reach into the internal law of every polity by ordinary statute had made the polities’ autonomy conditional on federal forbearance.

They were right about that, and the Realm accepted the consequence. The Extension Act is the moment at which the Charter’s guarantee of polity autonomy stopped being a limit on what Parliament might do and became a description of what Parliament had so far chosen not to do.

The social and economic rights

Part V was not in the Act of 1866 and was not added at once. Its sections arrived one at a time across the twentieth century, each attached to a particular political settlement, which is why their language is not uniform and why the Part reads less like a bill of rights than like a sedimentary deposit.

The right to education came first, in 1919, following the war. Labour’s rights to organise, to bargain and to withdraw labour were added in 1936, in the Roosevelt ministry, and are the most heavily litigated sections in the Part. The right to necessary health care was recognised in 1946, in the Fair Deal, and the National Health Service of America was built the following year to make an existing right operational rather than to create a new one. Provision in old age, sickness and disability followed in 1948. The protection against eviction into homelessness is the newest, added in 1978, and is deliberately narrower than a general right to housing.

No right to income security has ever been added, which is the subject of the case now before the Supreme Court.

The Consolidation of 2008

The May Day Bombings of 1 May 2006 produced the largest expansion of executive security power in the Realm’s peacetime history: watchlist authorities, transport security measures, surveillance capabilities and emergency detention powers.

The American Rights Consolidation Act 2008 was the constitutional settlement of those expansions. It did not create the powers, which had been taken in 2006 and 2007. What it did was bring them inside the Act, so that each operates subject to the requirement of warrant, the duty to record, the test of proportionality, sunset provisions, parliamentary oversight and judicial review, and none may touch the non-derogable core of s.9. It also rewrote Part VII, which had been three sections of nineteenth-century generality, into the modern emergency doctrine.

The Consolidation could do this only because 1962 had made the framework general. A security settlement folded into an Act binding federal authority alone would have left every polity constabulary outside it.

The text of the Act

The Act is reproduced below in full, Part by Part, as it stands after the Consolidation of 2008. Editorial commentary follows each Part and is not part of the statute. Where a section’s wording was materially different in 1866 or before 1962, the earlier text is given.

The Act is long and its Parts are of very unequal weight. Part I does almost all of the constitutional work and is the Part that lawyers argue about; Parts II to VI state the rights themselves; Parts VII to IX are machinery, and are the reason the Act functions at all.

The Preamble

The preamble is the only part of the Act written in the manner of the Charter, and the resemblance is deliberate. Its second recital, that freedom which the law will not enforce is not freedom, is the most quoted sentence in the statute and the one carved at the entrance to the Supreme Court.

The final formula is the load-bearing one. The rights are declared and not created. The distinction had a practical purpose in 1866, which was to answer the objection that Parliament was legislating for matters within the polities’ competence: an Act which declares an existing liberty is not creating a new legal relation, and the drafters expected the point to do some work in the courts. It did not, because s.4.2 made the question moot. It has done a great deal of work since 1962 in a different direction, since a right which is declared rather than granted is a right the Act’s repeal would arguably not extinguish, and the Realm has never had to find out whether that argument is any good.

Part I: Preliminary and Constitutional Status

Part I is the constitutional Part and contains, in ten sections, everything the Act does that could not have been done by an ordinary statute.

Section 2.2 is the Act’s operating principle and is dealt with at length below. Section 5 is the proportionality test, and it is the section a government most often loses on, because a ministry defending a measure will almost always concede that it is prescribed by law and directed to a lawful object and will then have to argue about whether it impairs the right no more than necessary. Section 9 is the floor, and the Realm’s courts have treated its list as exhaustive of what cannot be derogated from and not exhaustive of what is prohibited. The distinction matters: a thing not named in section 9 may still be unlawful under another section, and what section 9 settles is only that no emergency reaches the things it names.

Section 10.2 is the extraterritorial provision, and it is drafted as an obligation on the Realm’s authorities rather than as a claim about territory. The Act follows the officer, not the flag. A federal officer acting abroad is bound by the Act in respect of what they do, which is why the Realm’s forces operate under rules of engagement written by lawyers and why its intelligence services cannot lawfully do abroad what would contravene the Act at home. The provision was inserted in 2008 and was among the most contested clauses of the Consolidation.

Section 1.5, which binds the Crown, is a single line and does a great deal. No prerogative power may be exercised contrary to the Act, no commission or warrant under the Crown may authorise what the Act forbids, and the Realm’s law on the prerogative since 1962 is largely a series of decisions working out what that means.

Section 4.2 is the section with the history. As enacted it read, in its entirety:

The second limb is the whole of it. Everything else in the 1866 Act is compatible with a bill of rights binding the Realm entire; that clause, and only that clause, confined it. The Extension Act of 1962 repealed the words from and does not include to the end, and substituted the present definition, which reaches every public body in the Realm by naming them. The present s.4.2 is drafted as a list rather than as a general phrase for a reason: the drafters of 1962 wanted a definition that could not be narrowed by construction.

Section 4.3, which provides that words importing one gender include every other, was inserted in 1928. The Act as enacted in 1866 used he and him throughout, and in three sections used men where it meant persons, and the Realm’s courts had held twice before 1928 that the sections in question did not extend to women. The correction was made without ceremony and without apology, and the sections it corrected have never been re-enacted, so the 1866 wording is still what appears in the enrolled text.

Part II: General Liberties

Part II is the oldest-feeling Part and the least amended. Sections 11 to 14 are substantially the text of 1866, and section 14 is the section the Act was actually written for: the capacities it lists are precisely the capacities a freedman did not have in January 1865, and the section was drafted directly from Everett’s memorandum.

Section 12.3 is the Realm’s answer to a question it settled the hard way. The writ was never suspended anywhere during the rebellion, because the Commons would not pass the statute, and Ex parte Craddock had established in 1863 that a rebellion elsewhere did not affect the availability of the writ where the courts were sitting. Section 12.3 puts Craddock into the statute, which is why it is drafted as a proposition about geography rather than about principle.

Section 16.5 forbids retrospective punishment and was in the Act from 1866, which is unusual for a statute passed within three years of a rebellion. It was proposed by an opposition member and accepted by the ministry without a division, and the standard explanation is that both sides expected to need it.

Section 17, on unlawfully obtained evidence, is the Act’s most argued-about provision after section 5. It does not require exclusion; it requires exclusion where admission would bring the administration of justice into disrepute, which puts the question to the court rather than to a rule. Its critics say this makes the outcome unpredictable. Its defenders say that a fixed rule would produce either the exclusion of reliable evidence for trivial breaches or the admission of evidence obtained by serious ones, and that the Realm has chosen the discomfort of judgement over the comfort of arithmetic.

Section 20.2 is the sleeper. It was enacted in 1866 to prevent polities from taxing the entry of freedmen, was ignored for a century because it bound nobody, and has since 1962 become one of the most heavily litigated sections in the Act, chiefly in cases about residency requirements for polity benefits.

Part III: Expression, Assembly and Political Rights

Part III was thin in 1866 and is now the second-longest Part in the Act. Sections 21 and 24 are original; sections 22, 23.3, 26.2 and the whole of section 27 were added between 1919 and 1962.

Section 27.2 is the Extension Act’s most direct blow. Every device by which the polities had excluded Black voters is named in it: literacy, understanding, character, payment. The section does not use the word race in that list because race is already in section 3.1, and the drafters of 1962 wanted the tests named separately so that a polity could not argue it had applied one of them impartially.

Section 26.2 was inserted in 1962 and rewritten in 2008. Its present form is the reason the watchlist provisions of the Consolidation Act had to be drafted as narrowly as they were: a power to list a person by reason of lawful political opinion or association is prohibited outright, so every listing provision in Part VII had to be tied to conduct rather than to belief, and the security services regard this as the single greatest constraint on their work.

Section 22.2, the protection of journalistic sources, was added in 1962 and has been the subject of four attempts at amendment, all of them after a particular case and none successful. Its structure is characteristic of the Act: the protection is not absolute, a court may order disclosure, and the order requires the court to be satisfied both that disclosure is necessary and that no lesser measure will serve.

Section 21.3, on prior restraint, is original to 1866 and is the strictest provision in the Part. The Realm’s courts have granted prior restraint of a publication fourteen times since 1962 and have discharged the order within a week on nine of those occasions.

Part IV: Equality

Part IV did not exist in 1866. The Act as enacted had section 14 on civil standing and nothing else on equality, and the omission was not accidental: an equality Part binding federal authority alone would have been close to meaningless, and a Part capable of binding the polities was what the Lords had just struck out.

The whole Part was inserted by the Extension Act of 1962, and section 30 is the sentence the Extension Act exists for. It abolishes the separate-but-equal doctrine by refusing to hear evidence about it, which is a drafting choice rather than a doctrinal one and was made deliberately: the Realm’s courts had spent the 1950s receiving detailed evidence about the comparative quality of separate provision, and Parliament ended the practice by making the evidence inadmissible.

Sections 32 and 33 are later. Section 32 was enacted in 1974 and section 33 in 1991.

Section 29.2 is the indirect discrimination provision and is the most consequential thing in the Part after section 30. It reaches a measure neutral on its face which disadvantages a protected group, subject to justification under section 5, and it is the section under which most modern equality litigation is actually decided, because a public authority rarely discriminates in terms and frequently does so in effect. Its drafting in 1962 was contested: the ministry wanted the effects test and the Lords wanted an intention test, and the compromise was to include the effects test and subject it to the general proportionality provision rather than to a special defence.

Section 29.3, which provides that amelioration is not discrimination, was inserted at the same time and for the same reason. Without it the effects test would have made every remedial measure vulnerable to challenge by the group it did not benefit, which was pointed out in Committee by a peer opposing the whole Bill and accepted by the ministry on the spot.

Section 30 has never been amended and has been litigated four times, each time by a public authority arguing that its provision was not separate rather than that it was equal. All four failed. The section is drafted to make the second argument impossible and the Realm’s authorities long ago stopped attempting it.

Part V: Social and Economic Rights

Part V is the Act’s sedimentary Part. Nothing in it is original, its sections arrived one at a time across sixty years, and the drafting is visibly of different eras: section 34 is Edwardian in cadence, section 35 is a Roosevelt-ministry statute, section 36 is the Fair Deal, and section 38 reads like what it is, a compromise reached in 1978 between people who wanted a right to housing and people who would not have it.

Section 39.2 is the provision that gives the Part teeth, and it was inserted in 1978 with section 38. A right to be realised progressively is ordinarily unenforceable, since a court cannot compel a legislature to spend. What section 39.2 does instead is make retreat justiciable: a government may move as slowly as it likes toward a right and must justify moving backwards, which converts an aspiration into a ratchet.

Section 40 is the limit. A court may declare, may require a plan and may supervise it, and may not appropriate money or choose between lawful spending priorities. The line has held, though it is under more pressure in 2028 than at any time since it was drawn.

Section 36 preceded the National Health Service of America by a year, and the ordering is constitutionally important. The health service did not create the right it delivers. It made an existing right operational, which is why the entitlement does not depend on the statute that built the machinery and would survive that statute’s repeal.

Part VI: Indigenous Nations and Treaty Rights

Part VI did not exist before 1962. The Act of 1866 contained nothing whatever about the Indigenous nations, which is not surprising given that it bound federal authority alone and the federal government was at that moment conducting the removals, and which is nonetheless the omission the Realm finds hardest to explain.

The Part’s organising choice is section 41.1: the rights are preserved and not granted. They arise from treaty, prior sovereignty and the nations’ own law, and the Act’s function is to be prevented from diminishing them. Section 41.3 gives the treaty right priority over the Act itself, which is the only provision in the statute that subordinates it to anything.

Section 43 is the section that is litigated, and section 43.2 is the sentence that decides most of the cases: a consultation conducted after the decision does not discharge the duty. The Realm’s public authorities have lost a great many proceedings on it, most of them about infrastructure, and the Missouri corridor framework is the largest current example.

Section 43.3 is less quoted and does comparable work. The duty is owed to the nation and not to individual members, and it is not discharged by consulting a person the authority has itself selected. The provision was drafted against a practice the Realm’s departments had followed for a century, which was to find an amenable individual, obtain agreement, and record that the nation had been consulted.

Section 42.1 is the provision that gives the Part its shape. An Indigenous nation’s own government, legislature and courts are not a public authority for the purposes of the Act except as that nation’s law provides, which means the Act does not run inside the nations against their own institutions. This was argued about in 1962 and is argued about still, from both directions: by those who say a citizen of a nation is thereby left with fewer protections against their own government than any other American, and by the nations, who reply that a rights instrument imposed on a treaty government is the oldest form of the thing the Part exists to prevent. Section 44.1 is the answer the Act gives and is not regarded by either side as a complete one, since it secures the individual’s rights under the Act without altering who may be sued.

Part VII: Emergency

Part VII in the Act of 1866 was three sections of nineteenth-century generality which provided, in substance, that emergency did not suspend the Act and left the rest to the courts. The present Part was substituted entire in 2008, and it is the constitutional settlement of the powers taken after May Day.

Its method is to accept the powers and to surround them. Nothing in Part VII prohibits a watchlist, bulk collection or emergency detention; every one of them is contemplated and permitted. What the Part does is attach to each a warrant, a record, a period, a review and an expiry, on the reasoning that the Realm’s difficulty after 2006 was never that ministers wanted powers but that powers once taken are never given back.

Section 50.4 is the provision that does the most work in practice. A listing that has not been reviewed within ninety days simply lapses, without anyone having to apply, and the administrative effect has been to make the lists very much smaller than the authorities that maintain them originally intended, because reviewing an entry costs more than adding one.

Section 52 is the sunset, and it is drafted to prevent renewal by instrument. Every emergency power must be re-enacted by a Bill considered in the ordinary way, which means debated, amended and voted on, which is the whole of the protection.

The Part’s practical record is mixed and the article does not pretend otherwise. The declaration machinery at section 47 has worked: no emergency has been declared since 2007 and none has been renewed beyond a second period. The listing machinery at section 50 has worked better than its drafters expected. The surveillance provisions at section 51 have worked least well, because bulk collection under a court warrant for a specified purpose has proved compatible with warrants drawn so widely that the specification does little, and the Realm’s independent auditors have said so in every annual report since 2014 without Parliament acting on it.

Part VIII: Remedies

Part VIII is the machinery, and the Realm’s constitutional lawyers regard it as the best-drafted Part in the Act.

Section 58 is the compromise on which the whole arrangement rests. A court may not strike down an Act of Parliament for incompatibility with the Rights Act, because both are Acts of Parliament and the later would ordinarily prevail. What it may do is read the later Act compatibly if the words permit, and declare it incompatible if they do not, leaving the enactment standing and obliging a Minister to answer within six months. The Realm has issued forty-one declarations of incompatibility since 1962 and Parliament has amended the offending enactment in thirty-nine of them, which is the measure of how the arrangement works: it has no legal force whatever and is almost never ignored.

The two that were not amended are instructive. One concerned a polity’s licensing statute and was resolved when the polity repealed the provision itself before Parliament reached it. The other, in 1988, concerned a federal immigration provision that the ministry of the day defended, laid a response under section 58.4 declining to amend, and left on the statute book. It is still there. The Realm’s constitutional writing treats the 1988 response as the proof that the arrangement is genuinely a dialogue and not a disguised power of annulment, and treats the fact that it has happened once in sixty-six years as the proof that the dialogue usually goes one way.

Section 55.2 is the immunity provision and is drawn narrowly. No immunity of a public authority or of an officer bars an award of damages, save an immunity conferred by Act of Parliament in express terms. The Realm has very few such immunities, and each of them was voted for individually and can be found.

Section 56 is the systemic remedy and is the reason the Ferguson municipal courts settlement took the form it did. It is used sparingly, and the Realm’s judges have been careful with it, since a court that retains a proceeding for years to supervise a public authority’s plan is doing something close to administration. The longest supervision under section 56 ran for eleven years and concerned a prison.

Section 57.2 is the standing provision, and it is drawn more generously than in most comparable systems. A person not directly affected may apply where the question is of public importance, where there is no other reasonable means of bringing it before a court, and where the applicant has a genuine interest. All three conditions must be met, and the second does most of the work: an applicant who cannot show that a directly affected person could realistically have brought the case will be refused. The provision is the reason the income security case is before the Supreme Court at all.

Section 60 completes the arrangement. A court may protect a public-interest applicant from an authority’s costs, without which section 57.2 would be an invitation to be bankrupted.

Part IX: Public Duty

Part IX turns the doctrine of section 2.2 into administrative practice, and section 63.2 is the sentence that makes it operate. A public authority which cannot produce a record of a coercive act is presumed to have acted without authority, and the presumption is not rebuttable by testimony that a proper basis existed. The provision was inserted in 1962 and rewritten in 2008, and the effect of it is that the Realm’s constabularies keep records of a fullness that officers in other countries find astonishing, because an unrecorded stop is a lost case.

Section 64, the duty of candour, was added by the Consolidation Act after three security cases in which material adverse to the government had not been disclosed. It is owed by every officer personally and not merely by the authority’s lawyers, which was the specific point of failure in all three.

The Act in the courts

Power must show its warrant

Section 2.2 is the Act’s operating principle and the source of nearly all of its practical force. Every other provision states a right; section 2.2 states what a public authority must be able to do before it interferes with one, which is to name the law it is acting under.

The doctrine’s power lies in where it puts the burden. A person stopped, searched, listed, detained or refused does not have to prove that the authority lacked a power. The authority has to produce one, and section 2.4 makes the failure to produce it a contravention in itself, independent of whether a power existed. The Realm’s constabularies describe this as the single largest constraint on ordinary policing, and the Realm’s courts describe it as the reason ordinary policing is lawful.

Wardell

The leading authority is Wardell v the Commissioner of Police for the Columbia District (1969), decided seven years after the Extension Act, on facts of complete banality.

Thomas Wardell was a plasterer who was stopped while walking home in New Westminster at half past eleven at night, searched, and released. Nothing was found and no charge followed. He asked the constable under what power he had been searched and was told that the constable had a general power to search where he thought it necessary. Wardell brought a proceeding under Part VIII seeking a declaration and one shilling in damages, and the Realm’s law on public authority is what it is because he was stubborn enough to take it to the end.

The Supreme Court held that no such general power existed, that a belief in its existence was not a substitute for it, and that the search was therefore unlawful notwithstanding that the constable had acted in complete good faith and in accordance with what he had been taught.

The final paragraph is the most quoted passage in the Realm’s constitutional law after the Charter’s preamble, and governed rather than policed has entered ordinary speech.

Wardell is also the case that produced Part IX. The Court observed that a doctrine requiring an authority to name its power is worth little if no record is kept of what was done, and the duty to record at section 63 was drafted in direct response, with the presumption at 63.2 added because the Court had found itself unable to determine, on the evidence, how many times the constable had done the same thing before.

The limits of the doctrine

The Realm’s courts have been careful to state what section 2.2 does not do, and the leading case there is one the government won.

In Attorney-General v Brandt (1994) a public health authority ordered the closure of a food premises during an outbreak, acting under a statutory power drawn in wide terms. The proprietor argued that the power was too general to satisfy section 2.2 and that an authority relying on a broad discretion has not identified a law but only a permission. The Court rejected the argument. Section 2.2 requires an authority to identify the source of its power, not to demonstrate that the power is narrowly drawn, and the width of a statutory discretion is a question for section 5 and for Parliament rather than for the warrant doctrine.

Brandt is cited by public authorities more often than Wardell, and it is the reason the two must be read together. Section 2.2 asks where the power comes from. Section 5 asks whether its exercise was proportionate. An authority that cannot answer the first loses immediately; an authority that answers it still has to answer the second.

The other leading cases

Four further decisions are cited more often than any others and between them describe the shape of the modern Act.

Kettleman v the Realm (1966) is the first case decided under the Extension Act and is the reason the Act operates at county level at all. A registrar in a rural county had refused to enter a voter on a roll and had given no reason, on the practice of forty years. The Court held that a county registrar is a public authority within the amended s.4.2, that the giving of no reason is not the giving of a bad reason but a contravention in itself, and that a practice of forty years is precisely what s.2.3 means by usage. The registrar had not defied the Act. He had been doing his job the way it had always been done, and the judgment is unusual in saying so without softening what it cost.

In re the Detention of Ferrers (1980) settled the relationship between the emergency Part and the non-derogable core. A man had been detained under an emergency power for eleven months in aggregate under successive orders, each individually within the period then prescribed. The Court held that the core at s.9.4 is not satisfied by a series of lawful periods which together amount to detention without meaningful review, and that where a statute permits a thing to be done repeatedly the Act asks what has actually happened to the person. The aggregate limit now at s.49.3 was drafted in response.

Nation of the Ponca v the Minister of Transport (1997) is the leading authority on the duty to consult, and it is the case s.43.2 is written from. A federal department had held nine meetings with a nation over a road, all of them after the route had been fixed. The Court held that the duty is a duty to consult about a decision and not to explain one, that the test is whether the decision could still have been altered, and that an authority which has resolved the question and then talks about it has not consulted at all.

Adeyemi v the Chief Constable of the March (2011) is the first major decision under the Consolidation Act. A man had been placed on a watchlist, was refused employment as a consequence, and could not be told why. The Court held that s.50.2 requires notification unless a court is satisfied that withholding it serves a specified and current operational purpose, that the burden of establishing the purpose lies on the authority and is not discharged by assertion, and that the special counsel appointed under s.49.2 must be given the material and the time to test it. Something over four thousand listings lapsed in the eighteen months after the judgment, most of them because reviewing them was more trouble than removing them.

What the Act does not do

The Charter’s article in this Register carries a section on what the Charter excluded, and the Act deserves one for the same reason: a great deal of what the Realm believes about its rights is not in the statute.

The Act binds public authorities and not private persons. An employer, a landlord, a shopkeeper, a bank or a private club is not a public authority within s.4.2 and is not bound by the Act at all. Section 31, on public accommodation, is the single exception and was drafted narrowly in 1962 to reach services offered to the public without reaching private conduct generally. Everything else the Realm does about private discrimination it does by separate legislation, polity by polity and sector by sector, and the resulting patchwork is the largest gap in the Realm’s rights protection and the one least discussed.

The Act does not create a right to a particular level of provision. Part V’s rights are progressive and its remedies are supervisory, and s.40.2 forbids a court to appropriate money or to choose between lawful spending objects. What the Act protects is the direction of travel, by s.39.2, and not the distance travelled.

The Act contains no right to property in the American sense. Section 19 protects against deprivation without lawful authority and compensation and expressly preserves the regulation of property in the public interest, which is a European formulation rather than an American one, and it has never been read to constrain economic legislation.

The Act contains no general right of privacy against private intrusion, only against public authority under s.18. The Realm’s law of confidence and its data legislation do that work, and they are not constitutional.

The Act says nothing about the Crown’s prerogative in foreign affairs, defence, or the making of treaties, save that a treaty may supply lawful authority under s.2.2. The prerogative is bound by s.1.5 in respect of the rights the Act declares and is otherwise untouched.

And the Act does not entrench itself. Section 1.4 prevents amendment by implication and does not prevent amendment. A Parliament that wished to repeal the American Rights Act could do so by a majority of one, provided it said so expressly, and the only thing standing in the way is that no ministry has been willing to draft the sentence.

The income security case

The largest constitutional question before the Realm is whether Part V includes a right to income security, and the case is presently before the Royal American Supreme Court.

The proceeding

Amalgamated Tenants and the Cooperative Congress v the Attorney-General, universally the Tenants’ Case, was brought by a tenants’ union and a cooperative federation, both affiliated to the Left and Solidarity Bloc, the smallest of the Realm’s six parties. It was commenced in 2025, heard at first instance in 2026, heard on appeal in 2027, and argued before the Supreme Court over eleven days in the spring of 2028. Judgment is reserved.

The applicants’ standing was contested and was allowed under section 57.2, on the ground that the question is of public importance and that individual applicants would be unable to sustain a proceeding of this length. The standing ruling is itself now a leading authority and is disliked by the Attorney-General’s department more than anything else in the case.

What is claimed

The claim is that Part V, read as a whole, already contains a right to a minimum income sufficient to secure the rights it expressly declares, and that the Realm has simply never noticed.

The argument runs in four steps. First, that the rights in Part V are declared and not created, by the preamble’s formula, so the question is what the Part secures rather than what Parliament intended to grant. Second, that a right to necessary health care under section 36, to education under section 34, and against eviction into homelessness under section 38 cannot be enjoyed by a person without means, so that each of them presupposes a level of subsistence the Act does not name. Third, that section 39.1 requires the rights of the Part to be realised progressively to the maximum of available resources, and that a subsistence floor is the least resource-intensive way of realising several of them at once. And fourth, that section 40.1 permits a court to declare a failure and require a plan without appropriating money, so the remedy sought is within the Act’s own limits.

The applicants ask for a declaration and a plan. They expressly do not ask the Court to fix an amount.

What is answered

The Attorney-General’s answer is section 40.2, and it is a strong one.

The Realm’s law does not permit a court to direct the appropriation of public money or to substitute its judgement for a public authority’s as to the allocation of resources between competing lawful objects, and the Attorney-General’s case is that a declaration of the kind sought does precisely that whatever its form. A plan for a subsistence floor is a plan to spend, the sum involved would be among the largest single commitments in the federal budget, and a court that requires the plan has made the decision even if it declines to name the figure.

The second answer is historical and is the one the Court appeared most interested in. Part V’s sections were enacted one at a time, each after a political fight, each naming its subject precisely. Education in 1919, labour in 1936, health in 1946, provision in old age and sickness in 1948, security of the home in 1978. On five occasions Parliament chose what to add, and on none of them did it add income. The Attorney-General says that a Part built by deliberate accretion cannot contain by implication the one thing that was never acceded to, and that the applicants are asking the Court to enact the sixth section because the fifth Parliament would not.

What is at stake

The case is the largest in the Realm because the answer governs everything Part V might become.

If the applicants succeed, the progressive-realisation machinery at section 39 and the supervisory jurisdiction at section 40.1 become general, and the Realm acquires a mechanism by which a court can require a government to plan toward a social right it has not legislated for. If they fail, Part V is fixed at what Parliament has put in it, and the route to a right of income security runs through the Commons, where the Bloc holds fewer seats than any other party.

The wider constitutional stake is the line at section 40.2, which has held since 1978 and has never been seriously tested. Both sides accept that the line exists. They disagree about which side of it a declaration falls on, and there is no authority on the point because no applicant has previously got this far.

The politics

The case has produced an alignment nobody predicted. The Labour and Social Democratic Party supports the object and is uneasy about the method, its constitutional spokesmen having spent a century arguing that social provision is the business of an elected chamber. The Conservative and Crown Union opposes both. The Liberal Federal Democrats are split, and the split runs between the party’s constitutional lawyers, who think the applicants are right about the Act, and its front bench, which does not want them to be. Alliance 90 and the Green Realm supports the claim. Alternative for the Realm has called for the Act to be amended to reverse the result in advance, which no party has supported and which the Attorney-General has pointedly declined to endorse.

The polities have intervened in unusual numbers, and not on party lines. Fourteen have intervened against the claim, on the ground that delivery under section 36.2 already falls upon them and that a subsistence obligation would follow the same route. Six have intervened for it, all of them polities with large Delta or interior poverty districts.

What the Court was asked at the end

The last day of argument produced the exchange the Realm has been quoting since, and it is quoted here because it states the difficulty better than any commentary has.

Counsel for the applicants was asked what the Court should do if it agreed that Part V presupposed a subsistence floor and also agreed that fixing one was beyond its power. He answered that the Court should say the first and decline the second, and that a declaration which the Commons then ignored would at least have told the Realm what its own statute meant. Counsel for the Attorney-General was asked what the applicants should do if they lost. He answered, without apparent irony, that they should win an election.

Both answers are constitutionally correct and neither is an answer, which is the position the Realm is in and the reason the case matters beyond its subject.

The Register makes no prediction. The question is open, the Court is reserved, and the Realm has been arguing about section 40.2 since it was drafted precisely because nobody knew where the line was.

The Act in public life

The Rights Act is the Realm’s best-known statute and the most misunderstood, and both facts have the same cause: it is taught as a founding document and it did not do the founding thing until 1962.

The popular chronology is wrong in a specific and consistent way. The Realm’s schoolchildren learn the Emancipation Measure of 1865, the Rights Act of 1866 and the Fourth Amendment of 1873 as a sequence which settles the constitutional consequences of the war, and the sequence is real. What is missing from it is that the middle term bound nobody who mattered. Survey work by the corporation and by the universities has found consistently that a substantial majority of Americans believe the Act’s guarantees ran against the polities from 1866, and that the belief is strongest among people who have read it, because the Act as it stands does not carry its own history on its face.

The Act’s language has entered ordinary speech to a degree unusual for a statute. Show me your warrant is a phrase used by people with no idea that they are quoting section 2.2, and it is used in circumstances the section does not cover, which the Realm’s lawyers find irritating and its constitutional writers regard as the healthiest thing about the statute. Governed rather than policed comes from Wardell. Freedom which the law will not enforce is not freedom is from the preamble and is on the Supreme Court’s entrance.

Rights Act Day is not a public holiday and has twice been proposed as one. Both proposals failed on the same objection, made from opposite directions: that a Realm which commemorated 1866 would be commemorating the year it decided not to bind the polities, and that a Realm which commemorated 1962 would be admitting how late it was.

The Act in the constitution of ordinary life

The Act is present in the Realm’s daily administration to a degree that surprises visitors, and most of that presence is procedural rather than dramatic.

Every public authority publishes an annual report of its coercive powers under section 66, in a prescribed form, and the reports are read: by the press, by the parliamentary committees, and by the inspectors appointed under section 67, whose own reports are published under the same section. The Realm’s constabularies publish stop figures by ground, by outcome and by the characteristics named in section 3.1, and the publication is not a policy that a chief constable adopted but a statutory duty that a chief constable cannot decline.

Officers are trained in the Act by statutory obligation under section 65, and the same section forbids an authority to issue an instruction incompatible with it, which is the provision that made Wardell actionable at the level of the institution rather than the individual: the constable in that case had been taught wrongly, and section 65 now makes the teaching itself a contravention.

The oath of the Governor-General names the Act, and has since Douglass took it in 1873. So do the oaths of the Realm’s judges, of its constables and of its commissioned officers, and the wording in each case places the Act alongside the Crown and the Charter rather than beneath them.

The Act is taught in the Realm’s schools as a matter of curriculum in every polity, which is itself a consequence of section 34.2 and the requirement that education be adequate to the exercise of the rights the Act declares. What is taught is the modern Act. The 1866 limitation appears in the curriculum of most polities and is treated briefly, and the Realm’s historians have complained about this for forty years.

The Act is also, unusually for a statute, a thing people own. Printed copies of the nine Parts are sold cheaply, are given to schoolchildren in several polities, and are kept behind counters in a great many small businesses for the purpose of being produced. The corporation’s legal affairs programme has run a segment for thirty years in which a member of the public describes an encounter with authority and a lawyer identifies the section engaged, and it is among the ABC’s most durable formats.

The phrase most often misused is section 3.2. It provides that no person shall be placed outside the protection of the Act by executive designation, and it is regularly cited by people who mean that they should not have to comply with something. What it actually secures is narrow, absolute and hard-won: that there is no category of person, however designated, to whom the Act does not apply, which was the argument the Realm had after May Day and settled in 2008.

Assessment

The American Rights Act is the Realm’s most important statute and the clearest evidence that important statutes can be built to fail.

Its drafting is admirable and has aged extraordinarily well. The warrant doctrine at section 2.2, the proportionality test at section 5 and the non-derogable core at section 9 are all provisions that other constitutional systems arrived at decades later or have not arrived at yet, and they were substantially in place in 1866. The Act anticipated the emergency problem before the Realm had a modern emergency, and the machinery the Consolidation Act needed in 2008 was already there to be filled in.

And for ninety-six years none of it reached the people it had been written for. The Act was drafted to make four million newly free persons into persons in law, and it was passed in a form that left them to the mercy of exactly the institutions Everett had identified as the danger. Nothing had to be defied, which is the sentence that ought to be attached to the Act wherever it is taught, because it is the difference between a law that was broken and a law that was built not to apply.

The Realm’s usual account of this is that the Lords exacted a price and the ministry paid it to get the Act at all, and that half a loaf was better than none. The account is true and is not a defence. The half that was omitted was the half that mattered, the men who voted knew it, and the Act’s own preamble tells them so: freedom which the law will not enforce is not freedom, and from 1866 to 1962 the law did not enforce it where it needed enforcing.

What the Act has been since 1962 is a different institution wearing the same name, and it has been very good. It reached the polities, ended separate provision by refusing to hear evidence about it, named every device the polities had used to exclude voters, acquired a Part on the treaty nations, grew a body of social rights, and absorbed the largest expansion of security power in the Realm’s peacetime history without conceding the non-derogable core. The Realm is entitled to be proud of the Act it has. It is not entitled to be proud of it for as long as it has been.

There is a second reading, made by the Realm’s constitutional historians and worth recording because it is less comfortable than either the celebration or the indictment. The Act’s drafting quality and its ninety-six-year failure have the same cause. A statute written to bind federal authority alone could afford to be exacting, because the people writing it were describing constraints on a government they controlled and could predict. The warrant doctrine, the proportionality test and the non-derogable core were all easier to enact in 1866 precisely because they were not going to be applied to a county sheriff in a polity that had just been at war with the Realm. What the Lords took out is what made the rest of it safe to leave in.

If that is right, then the Act is not a good statute that was crippled by a bad amendment. It is a statute whose excellence and whose limitation were the same bargain, and the Realm got the whole of it in 1962 only because by then it was willing to pay a price the Parliament of 1866 was not. That reading does not excuse anything. It does explain why the Act reads as though it were written by people who understood exactly what they were doing, which they were.

Chronology

Trivia

See also

• Charter of Union of the American Colonies: whose Article IX declares the liberties this Act operates
• The American Civil War: the rebellion, the Emancipation Measure and the settlement the Act completes
• Edward Everett: whose memorandum identified the problem the Act failed to solve
• Salmon P. Chase and Frederick Douglass: the Governors-General of assent and of enforcement
• National Health Service of America: built to make section 36 operational
• Left and Solidarity Bloc: to whose affiliates the income security case belongs
• May Day Bombings: the emergency the Consolidation Act settled
• Treaty Nations of the United Kingdom of America: the nations whose rights Part VI preserves
• Institutions and Government

Categories

Institutions · Law · Rights · Supreme Statutes · Parliament · Constitution · Civil Liberties