The Permanent Emergency: How Temporary Powers Become Tools of Domination

Every constitutional democracy eventually confronts a moment when the ordinary machinery of law seems inadequate to the crisis at hand. A pandemic outstrips existing public health statutes. A terrorist atrocity exposes gaps in surveillance and detention powers. An economic collapse demands executive action faster than parliament can legislate. In each case, the instinct of government is the same: draft an emergency provision, grant the executive extraordinary and temporary authority, and promise that the powers will lapse once the danger has passed. This instinct is not unreasonable. Genuine emergencies do sometimes require flexibility that ordinary legislative process cannot supply quickly enough. But the historical record offers a sobering counter-lesson: emergency powers rarely behave as promised. They tend to outlive the emergencies that justified them, and in the hands of a government less scrupulous than the one that enacted them, they become instruments not of protection but of domination. The danger is structural, not incidental – it inheres in the very nature of law as a durable text that survives the intentions of its drafters.

The Logic of the Trap

The trap operates through a predictable sequence. First, a crisis creates political consensus that ordinary constraints on power must be loosened, if only briefly. Second, the legislation drafted to meet that crisis is written broadly and vaguely, because urgency leaves little time for precision and because broad language is easier to agree upon than narrow language that might exclude some contingency the drafters have not foreseen. Third, the law is passed with an implicit or explicit understanding that it is exceptional – a deviation from the constitutional baseline rather than a new baseline itself. Fourth, and most dangerously, the law is rarely accompanied by mechanisms robust enough to guarantee its own expiry. Sunset clauses, where they exist, are often extended rather than allowed to lapse. Judicial review is frequently deferential precisely because courts, too, are reluctant to second-guess executive judgment during a declared emergency. Parliamentary oversight, meanwhile, depends on a legislature willing to reassert itself against an executive that has just been granted extraordinary authority – a dynamic that consistently favours the executive, since incumbency, information asymmetry, and the political cost of appearing “soft” on the crisis all work against reversal.

Once these conditions are in place, the statute becomes available to any future government, regardless of that government’s fidelity to the values that produced the original law. This is the deepest danger: the emergency provision does not merely persist as dead text. It persists as a live grant of power, fully legitimated by its provenance in ordinary legislative process, waiting to be picked up by hands very different from those that forged it. A later government does not need to overthrow the constitutional order to acquire dictatorial capacity; it need only reinterpret and intensify an instrument that democratic process has already sanctioned. This is what makes the danger so much harder to see coming than an outright coup. There is no dramatic rupture, no visible seizure of power – only the quiet redeployment of an existing legal tool toward purposes its authors never intended and might have found abhorrent.

Weimar Germany: The Archetype

No case illustrates this mechanism more starkly than the collapse of the Weimar Republic, and it is worth dwelling on because it supplies the template against which every subsequent instance of emergency-power abuse can usefully be measured. Article 48 of the 1919 Weimar Constitution empowered the Reich President, whenever public security and order were “seriously disturbed or endangered,” to take whatever measures were necessary to restore them, including the suspension of core civil liberties – personal freedom, freedom of expression, the right of assembly, and the inviolability of the home and of correspondence. This was not a cynical provision. It reflected the genuine anxieties of a constitutional convention meeting in the aftermath of a lost war, a revolutionary uprising, and the collapse of the old imperial order. Its drafters wanted a democracy resilient enough to survive its own birth pangs, and they trusted that a president elected by the whole nation would use such power sparingly and in good faith. Friedrich Ebert, the Republic’s first president, invoked Article 48 more than a hundred times during the 1920s, mostly to address currency crises, regional uprisings, and administrative deadlock – disruptive, but recognisably within the spirit of the provision.

That good faith was not transferable. On 28 February 1933, one day after the Reichstag fire, President Paul von Hindenburg – by then an elderly figure increasingly detached from the machinations around him – signed the Decree for the Protection of People and State under the authority of Article 48. The decree suspended, without any fixed end date, precisely the liberties the article had been designed to protect in the ordinary run of politics: freedom of the press, freedom of assembly, the secrecy of the post and telephone, the inviolability of property and the home. It authorised “protective custody,” a euphemism for indefinite detention outside any judicial process. Nothing about the constitutional text had changed. The words of Article 48 in March 1933 were identical to the words that Ebert had once used to stabilise the Republic. What had changed was who held the pen, and what they intended to do with it. Within weeks the decree had become the legal foundation for the mass arrest of Communist and Social Democratic politicians, the closure of opposition newspapers, and the establishment of the first concentration camps at Dachau and elsewhere – all of this accomplished, notably, without a single line of new emergency legislation. The Enabling Act that followed in March 1933 compounded the damage by allowing the cabinet to legislate without the Reichstag altogether, but it is worth stressing that the initial and most consequential seizure of extra-constitutional power required no new statute at all. It required only an old one, applied without restraint by people who did not share the assumptions of those who wrote it.

Paragraph 175: The Ordinary Statute Repurposed

A second, quieter example from the same period shows that the danger is not confined to grand constitutional emergency clauses; it extends equally to ordinary criminal law drafted for entirely different purposes. Paragraph 175 of the German Criminal Code, dating from the unification of the Reich in 1871, criminalised sexual acts between men. It was a product of nineteenth-century moral regulation, enforced inconsistently and, by Weimar standards, relatively rarely – a statute more often invoked to police conventional morality than as an instrument of mass persecution. When the Nazi regime came to power, it did not need to invent a new category of criminal offence in order to persecute homosexual men; the category already existed, sanctified by six decades of legal continuity. In 1935 the regime simply broadened Paragraph 175’s language, widened the scope of prohibited conduct, and directed the police and courts to enforce it with unprecedented severity. Convictions increased more than fivefold. Tens of thousands of men were imprisoned, and a significant number were subsequently transferred to concentration camps, where they were marked with the pink triangle and subjected to conditions designed to kill.

The lesson here is distinct from, though complementary to, the lesson of Article 48. Emergency clauses invite abuse because their language is deliberately elastic – “public security,” “order,” “necessary measures” are terms built to flex under pressure. Ordinary criminal statutes invite a different kind of abuse: not reinterpretation of ambiguous language so much as escalation of an already-settled category. Once society has accepted that a certain class of conduct may legitimately be criminalised, an unscrupulous regime need only redefine the boundaries of that class and intensify enforcement to convert a statute of modest, targeted application into an engine of systematic persecution. The mere fact that a prohibition already exists lowers the political and rhetorical cost of expanding it. Legislators inherit not a blank page but a foundation, and foundations are far easier to build upon than to lay from scratch.

The Pattern Repeats: Comparative Cases

Germany’s experience is extreme, but it is not unique, and the recurrence of the pattern across very different political systems and eras is what should trouble us most, because it suggests a general vulnerability in the architecture of emergency and quasi-emergency legislation rather than a peculiarity of Weimar or of Nazism.

India’s Emergency of 1975–77 offers a democratic-era parallel. Prime Minister Indira Gandhi, invoking Article 352 of the Indian Constitution on grounds of “internal disturbance,” suspended fundamental rights, jailed opposition leaders and journalists without trial, imposed press censorship, and ruled by decree for nearly two years. The provision she invoked had been drafted by the framers of India’s constitution with external aggression and genuine internal breakdown in mind; it was not designed as a mechanism for a sitting prime minister to entrench her own rule against an adverse court judgment and a mobilising opposition. Nonetheless, the constitutional text supplied exactly the tool required, and India’s judiciary – with a few honourable dissents – largely deferred to executive discretion during the crisis, most infamously in the Supreme Court’s 1976 ruling that even the right to life could be suspended under the Emergency.

Egypt supplies an example of emergency law becoming not an episode but a permanent condition of governance. A state of emergency, first declared in 1958 and renewed almost continuously – with only brief interruptions – for over six decades, provided successive Egyptian governments with sweeping powers of arrest, detention without charge, and restriction on public assembly and the press, all justified initially by external threat and later by domestic insurgency. What began as a response to specific security conditions became, through uninterrupted renewal, the ordinary operating framework of the Egyptian state, such that emergency governance and governance itself became difficult to distinguish.

Even mature, stable democracies are not immune. The USA PATRIOT Act, passed six weeks after the attacks of 11 September 2001 with almost no dissent and minimal debate, dramatically expanded surveillance authority, permitted indefinite detention of non-citizens, and lowered the evidentiary threshold for search warrants under the banner of counter-terrorism. Several of its provisions carried sunset clauses – a safeguard the Weimar and Egyptian examples lacked – yet those clauses were repeatedly reauthorised by successive Congresses for close to two decades, and the surveillance infrastructure the Act helped legitimate was later revealed, through Edward Snowden’s disclosures, to have been extended by intelligence agencies to bulk data collection on ordinary citizens far beyond anything the original counter-terrorism rationale could justify. The United Kingdom’s post-2000 counter-terrorism legislation followed a comparable trajectory, and here the figures are precise enough to be worth stating. Control orders, introduced under the Prevention of Terrorism Act 2005 after the House of Lords struck down the preceding regime of indefinite detention without trial, were imposed on fifty-two individuals between 2005 and 2011; by the time the regime was replaced, only nine orders remained in force, a gap the Independent Reviewer of Terrorism Legislation cited as evidence the power had been significantly over-drawn relative to its actual operational need. Their 2011 replacement, Terrorism Prevention and Investigation Measures, was deliberately capped at a maximum of two years’ duration for any individual – a genuine, hard-edged sunset provision built into the statute itself. That cap held for a decade. In 2021, the Counter-Terrorism and Sentencing Act removed it, allowing TPIMs to be renewed indefinitely provided the original statutory conditions continued to be met. The episode compresses the entire argument of this essay into a single decade: a hard sunset clause was written, it worked exactly as intended for ten years, and it was then quietly dismantled by ordinary legislative amendment, with the power that had once been temporary by design becoming permanent by amendment.

Australia’s own experience, though less dramatic than these examples, follows the identical structural logic and deserves attention precisely because it demonstrates that the danger is not confined to authoritarian regimes or societies in acute crisis. The tranche of counter-terrorism legislation passed in the years following 2001 and 2005 – preventative detention orders, control orders, expanded ASIO questioning and detention powers, and later the broadened metadata retention regime – was justified in each case by reference to a specific and pressing security concern. Successive Independent National Security Legislation Monitors have since documented how elements of that framework have been retained, and in some cases extended, well beyond the period in which the originating threat assessment applied, with oversight and review mechanisms frequently criticised as inadequate to the scope of the powers granted. None of this amounts to a Weimar-scale catastrophe. But the mechanism is the same mechanism, and again the detail is instructive. Control orders under Division 104 of the Criminal Code Act 1995 carry a sunset clause of their own – five years, renewable by ordinary parliamentary vote rather than supermajority. Following the most recent review in 2023, the regime was extended to December 2026, the latest in a sequence of routine renewals stretching back to the provision’s original enactment in 2005. A sunset clause of this kind imposes almost no real constraint: renewal requires nothing more than the same simple majority that created the power in the first place, and the political cost of voting against renewal – appearing to relax counter-terrorism settings – is far higher than the cost of a routine, largely unscrutinised re-tick. The mechanism is broad drafting, weak sunset provisions, deferential review, and a political incentive structure in which no government wishes to be the one that “weakened” national security by allowing a power to lapse – even when the specific threat that justified it has receded or transformed.

Hungary’s response to the COVID-19 pandemic offers a more recent and more compressed illustration of the same dynamic. In March 2020, the Hungarian parliament passed the Authorisation Act, permitting Prime Minister Viktor Orbán’s government to rule by decree, suspend the operation of existing legislation, and postpone elections for the duration of the declared state of danger – with no fixed end date written into the law itself. The stated justification was the pandemic, a genuine public health emergency shared by every country in the world. Yet the government simultaneously used its decree powers to criminalise the publication of “false” information about the epidemic in terms broad enough to chill ordinary journalism, and it retained significant elements of the expanded executive authority well after the immediate public health justification had receded, illustrating within the span of a single parliamentary term the entire arc that took Germany fourteen years to complete: emergency declaration, executive rule by decree, and quiet retention of powers whose original rationale had already lapsed.

When the Safeguards Held

The pattern is not universal, and it is worth pausing on the exceptions, because they show that the containment this essay calls for is not a purely theoretical proposition – it has happened, under real political pressure, and it is useful to see precisely what made it possible.

The origin of the UK’s control order regime is itself one such case. The indefinite detention without trial of foreign terrorism suspects at Belmarsh prison, authorised under Part 4 of the Anti-Terrorism, Crime and Security Act 2001, was struck down by the House of Lords in A and Others v Secretary of State for the Home Department (2004), which held that indefinite detention limited to foreign nationals was a disproportionate and unlawfully discriminatory response that could not be justified even under a declared public emergency. The ruling did not end executive counter-terrorism power in the UK – it forced its reformulation into the control order regime discussed above – but it stands as a genuine instance of a senior court refusing to defer to the executive’s own assessment of necessity, and it is precisely the kind of independent judicial check this essay argues is chronically underused.

A cleaner case of outright lapse comes from the United States. Section 215 of the PATRIOT Act, the provision underpinning the NSA’s bulk telephone metadata collection exposed by Edward Snowden in 2013, was first narrowed by the USA FREEDOM Act of 2015 and then, when Congress could not agree on reauthorisation terms, allowed to expire altogether on 15 March 2020. No replacement was enacted. The authority simply lapsed, and the underlying bulk-collection program had, by the government’s own later admission, already been quietly shut down for operational reasons some months earlier. It is the rare example in this essay of an emergency-derived surveillance power ending not through political drama but through the ordinary, unglamorous failure of Congress to agree on its extension – precisely the outcome a hard sunset clause is designed to make possible.

A third example brings the pattern into the pandemic era and answers the case for a genuinely contemporary, digital-age instance. In March 2020, Israel’s government authorised its domestic intelligence service, Shin Bet – an agency whose ordinary legal mandate is national security, not public health – to track the mobile phones of COVID-19 patients and those who had crossed their paths, doing so by emergency regulation rather than primary legislation. Civil rights groups petitioned immediately, and on 26 April 2020 Israel’s High Court of Justice ruled that the General Security Service Law provided no lawful basis for the Shin Bet to be turned to a purpose so far removed from its statutory function, and that continued surveillance of this kind required the government to go to the Knesset and legislate openly rather than govern by emergency decree. The court’s language is worth noting: it warned explicitly of a “slippery slope” in the use of “extraordinary and harmful tools” against ordinary citizens. The episode is not an unqualified success – the Knesset went on to authorise a narrower, time-limited version of the same tracking power, repeatedly renewed in three-week increments through subsequent case surges – but it demonstrates that even a powerful security agency’s emergency mandate can be halted by a court applying ordinary principles of legality, and it shows that the digital character of a twenty-first-century emergency power does not place it beyond the reach of the same safeguards that constrained Article 48 or Section 215.

What unites these three episodes is not that the underlying power disappeared for good – in the UK and Israeli cases it did not, and even in the American case adjacent surveillance authorities under other statutes remained available. What they share is that some part of the machinery this essay recommends – a court willing to test proportionality against a declared emergency, a sunset clause with a genuine possibility of lapsing, a judiciary insisting that novel powers be legislated openly rather than governed by decree – actually functioned as designed, at least once. The lesson is not that containment is easy. It plainly is not, given how quickly the UK Parliament removed the TPIM cap once it had proved inconvenient. The lesson is that containment is possible, that it depends on specific, identifiable mechanisms rather than good intentions, and that those mechanisms only work when something or someone is prepared to use them against a government that would rather they went unused.

Why the Danger Persists

Four structural features recur across every one of these cases, and understanding them is essential to understanding why the danger is so difficult to legislate away.

First, urgency and precision are in permanent tension. Legislation drafted in the shadow of a crisis is, almost by definition, drafted quickly, and speed is the enemy of careful boundary-setting. Terms like “public security,” “national emergency,” “unnatural conduct,” or “reasonably necessary” are chosen because they command consensus across a legislature that may otherwise disagree about almost everything else – but that same elasticity is precisely what permits a later government, operating in entirely different political circumstances, to pour new content into old words.

Second, the institutional mechanisms that might constrain misuse – sunset clauses, mandatory reauthorisation, independent judicial review – are chronically underused or insufficiently robust, and where they exist, they are vulnerable to the same political dynamics that produced the original law. A sunset clause that requires only a simple parliamentary majority to renew provides little more protection than no sunset clause at all, particularly when opposing renewal can be characterised, however unfairly, as indifference to public safety.

Third, all of these arrangements rest on an unspoken assumption of institutional continuity and good faith: that whoever next occupies the relevant office will share, more or less, the values and restraint of those who wrote the law. History supplies no basis for that assumption. Constitutions and statutes are inherited by people the original drafters never met and could not have vetted, and the text does not carry the drafters’ intentions forward with it – only their words, available to be read however a new government chooses to read them.

Fourth, mere existence breeds normalisation, and this process has a bureaucratic engine that is worth examining on its own terms rather than treating as a vague cultural drift. An emergency power is rarely administered by the politicians who voted for it; it is administered by an agency, a division, a unit of officers trained in its use, and that apparatus does not disappear when political attention moves elsewhere. Budgets are allocated, personnel are recruited and promoted on the basis of their expertise in wielding the power, and internal reporting lines and performance metrics come to assume the power’s continued existence as a baseline condition of the agency’s work. Officials whose professional standing depends on a power’s continuation have every incentive to argue, in the technical and unglamorous language of annual reviews and risk assessments, that the threat has not really receded, or that the power now serves purposes subtly broader than those for which it was first granted – the fourteen-day detention regime extended by the states beyond the Commonwealth’s original forty-eight hours, the TPIM cap quietly removed once a decade of officials had grown accustomed to indefinite renewal being merely one legislative amendment away. Each annual reauthorisation, each five-year sunset review, becomes progressively easier to wave through precisely because it has been waved through before: path dependence operates not because anyone consciously decides the power should become permanent, but because reversing course requires an affirmative act of political will against an entrenched administrative status quo, while renewal requires only inertia. What begins as an extraordinary, closely scrutinised departure from constitutional baseline gradually becomes an unremarkable feature of the administrative landscape – invoked routinely, defended reflexively by the very apparatus built to invoke it, and eventually so embedded in ordinary governance that its removal comes to look like the radical step, rather than its original enactment.

Toward Durable Safeguards

None of this counsels abandoning emergency powers altogether; genuine crises do occur, and a state entirely incapable of rapid, flexible response is not obviously safer than one that retains such capacity under strict conditions. The task is to design emergency and quasi-emergency legislation so that its temporary character is structurally enforced rather than merely asserted in a preamble.

Several safeguards recur, in varying combinations, wherever emergency powers have been successfully contained. Automatic and genuinely hard sunset clauses – requiring an affirmative supermajority vote to extend, rather than a simple majority, and lapsing by default if that vote fails – shift the political burden from those who would end a power to those who would prolong it. Precise, narrowly drawn triggering conditions, rather than open-textured phrases like “public order” or “national emergency,” reduce the interpretive room available to a later government seeking to repurpose the provision. Mandatory, meaningful parliamentary reauthorisation on a short cycle – months rather than years – keeps the political cost of renewal visible and recurring, rather than allowing a power to fade into administrative background noise. Independent judicial review with a genuine capacity and willingness to strike down or narrow executive action, rather than the reflexive deference courts have historically shown during declared emergencies, supplies a check that does not depend on the same political incentives that constrain the legislature. Sunset provisions and reauthorisation requirements should also be paired with independent, statutorily empowered monitors – bodies modelled on arrangements such as the Independent National Security Legislation Monitor – tasked specifically with assessing whether the original threat still justifies the retained power, and with public reporting obligations that make quiet extension harder to accomplish unnoticed.

Beyond formal mechanisms, something less tangible but equally important is required: a constitutional culture that treats emergency power as inherently exceptional and provisional, never as a convenient permanent addition to the ordinary toolkit of governance. This is ultimately a matter of civic vigilance as much as legal design. The drafters of Article 48 and of Paragraph 175 did not write those provisions with Hindenburg’s 1933 decree or Himmler’s 1935 police directives in mind. That is precisely the point, and precisely the danger: they did not need to. A statute, once enacted, does not require its authors’ complicity in order to be misused by their successors. It only requires that it continue to exist, unrepealed, unrefined, and available.

Conclusion

The through-line connecting the Weimar Republic’s Article 48, the expanded Paragraph 175, and their contemporary analogues in India, Egypt, the United States, the United Kingdom, Hungary, Israel, and Australia is not that emergency legislation is always drafted in bad faith. Quite the opposite – nearly every example canvassed here began as a good-faith, often genuinely necessary response to a real and pressing danger. The tragedy is structural rather than moral: legislation outlives the political context and the personnel who produced it, and its formal legitimacy – the fact that it was passed through ordinary democratic process, by people who meant well – travels forward with the text, ready to be claimed by anyone who next controls the levers of the state. A concentration camp system did not require the invention of new legal categories in 1933; it required only the redeployment of an existing one. The persecution of homosexual men under the Third Reich did not require Parliament to create a new crime in 1935; it required only the intensification of an old one. This is the enduring risk of any legislation whose original purpose is narrow but whose textual reach is broad: the gap between purpose and text is exactly the space in which tyranny finds room to work. Guarding against it demands more than good intentions at the moment of enactment. It demands structures – hard sunsets, narrow triggers, genuine reauthorisation, independent review – built with the explicit and uncomfortable assumption that the next hand on the lever may not be a trustworthy one.

– – –

Bakchos is the founder of Blak and Black, an Australian media and advocacy platform established in 2010. Bakchos writes from the intersecting perspectives of Wiradjuri heritage, Jewish identity, and humanism.

© Bakchos, August 2026

BLAK AND BLACK  |  MEDIA AND ADVOCACY  |  EST. 2010