States of Emergency

Introduction

The fifty days of the ‘COVID-19 and States of Emergency’ Symposium covered the height of the global legal reaction to the pandemic, offering a snapshot of countries in collective crisis. It began with a call for a global conversation on the kind of legal norms which should govern the situation of worldwide pandemic. This final contribution aims to trace the central themes, questions and issues raised by the Symposium. It considers constitutional safeguards on a ‘state of emergency’, and whether this is preferable to the use of ordinary legislation in managing a crisis. It examines the dangers of executive action, and whether countries have been successful in limiting the potential for abuse, as well as preventing or sanctioning it. It examines how states have struggled to maintain some degree of legislative and judicial normality, while other states have given it up entirely. Finally, it identifies the most successful approaches adopted, and the most detrimental. In doing so, it aims to form part of that global conversation which seeks to identify the most concerning legal developments in a global emergency, but also to advocate for the best practices emerging worldwide.

States in Emergency

Extraordinary Powers in Ordinary Law

Debate has raged around the degree to which countries have been prepared for the threat of a pandemic, let alone one unparalleled in its current scale and impact. In the initial phase of the crisis, most states responded with government decrees or administrative decisions – sometimes with questionable legal basis. A question is whether countries would benefit from stronger legal basis for action.

A number of states have relied on ordinary legislation to manage the crisis without resorting to emergency powers, even where such powers are available in the legal or constitutional framework. The reasons have varied: either the crisis did not constitute an ‘emergency’ within constitutional provisions; or there were sufficient powers and/or mechanisms within ordinary legislation, and thus no need to resort to emergency powers. Nevertheless, it should be underlined that declaring a state of emergency does not indicate a potential abuse of power, any more than the exclusive use of ordinary legislation means that there is no abuse. Many states that have used ordinary legislation avoided the scrutiny and conditionality which normally attaches to the use of emergency powers. Others by contrast are concerned with the negative historical connotations in light of past use of emergency powers, and so have used ordinary provisions that might be even less suitable.

An important point of consideration is whether the use of pre-existing legislation reveals ordinary powers to be extraordinary in their scope. Many countries have relied on Health Acts to provide the legal basis for sweeping powers for detention, quarantine, and even lockdown. A tangible concern is the interpretation of ordinary legislation to the effect that it allows action that ought to have been held ultra vires. For example the UK’s Public Health (Control of Disease) Act 1984 allows a Minister to make a ‘special restriction or requirement’ ‘on where [a person] goes or with whom [a person] has contact’. To interpret these sections as authorising a nationwide lockdown of the entire population is questionable at best.

In many cases, the Health Acts initially relied upon were promulgated in very different times (eg Nepal’s 1964 Infectious Diseases Act; and India’s Epidemic Diseases Act 1897). To provide legal basis for the unprecedented nature of restrictions, many states amended existing public health legislation. The speed of amendments in some countries however afforded extraordinarily little time for meaningful review (from 4 days in the UK, to only 12 hours in Denmark) and the quality of the law can suffer. Following a storm of protest from legal scholars, lawyers and judges for the secrecy and lack of accountable input, the Norwegian government, radically revised its initial draft law on action concerning the coronavirus. Finland offers an example of best practice for the pluralistic review of constitutionality and rights-compliance of executive decrees through standing committees and engagement with external legal and constitutional experts. Finland also deserves special mention for inviting public scrutiny of decrees through real-time posting on a legal blog.

While emergency necessitates urgent action, there is always capacity for subsequent review and reform. On this point, Italy is exemplary. The country suffered one of the highest mortality rates in Europe. It was among the first to introduce restrictive measures, and the second globally behind China to introduce a national lockdown. The initial measures diverged at local, regional and national level, and were introduced so quickly and haphazardly as to create ‘regulatory and legal chaos’. However, this changed: responding to the criticism from academics, lawyers and the media directed at earlier provisions the Italian government reformed the legal measures to include clear constitutional safeguards and protections for the rule of law. This trend echoed across the EU (with the notable exception of Hungary and Poland): initial legal shortcomings were subsequently rectified.

It is highly probable that this decade will witness numerous constitutional amendments and legislation governing health emergencies to take account of global crises, and some key insights may be gleaned from collective experience. Ongoing stakeholder engagement – involving external constitutional and legal experts – and readiness to reform creates better-quality law.

Wherefore a ‘State of Emergency’?

Many countries, though not all, have declared a state of emergency in response to COVID-19. There are broadly four common elements to provisions on a ‘state of emergency’ (including (1) conditions for its declaration; (2) a delegation of power; (3) limitations on its use; (4) provisions for legislative or judicial oversight). The situations justifying the declaration of a state of emergency, however, are intricately linked with national historical experience. Most constitutions stipulate war, external aggression, or armed rebellion as a condition for declaration. Only a few refer to a natural disaster, and less still refer to an epidemic or health emergency. Some constitutions contain more open-ended and interpretable conditions. For example in Malaysia, a state of emergency can be declared where the King believes there to be serious threat to ‘security’, ‘economic life’, or ‘public order’.

There is no strong indication of whether constitutional safeguards can limit the potential for abuse of emergency powers. Following 2011 constitutional reform, the 1917 Constitution of Mexico envisions the oversight by both judicial and legislative branches, and these political and legal safeguards against its misuse cannot be overridden by the executive. Despite numerous situations which may have called for a state of emergency: it has only been declared once in 1942 during World War II and has not been declared in the current crisis. In contrast, some states have existed in a near perpetual state of emergency despite safeguards. The 2014 Egyptian Constitution was drafted with intention of bringing an end to the near-perpetual state of emergency since 1967. The safeguards introduced, including two-thirds approval of the House of Representatives, have been however sidestepped through formalistic proceduralism. Egypt did not declare a state of emergency through the coronavirus crisis because it has never left such a state for more than a few days since 2017.

Some states avoid the term entirely: having experienced a two-year state of emergency following terrorist attacks in 2015, France declared a new state of a ‘health emergency’ which mimics the pre-existing provisions for a state of emergency, though only provides a more limited role for parliament and did not derogate from the European Convention on Human Rights [ECHR] and the International Covenant on Civil and Political Rights [ICCPR]. An ‘unofficial’ state of emergency can create equal cause for concern: the call of a  ‘general holiday’ in Bangladesh to avoid the negative associations of a state of emergency, belies the gravity of the situation and misleads the population into high-risk behaviours including mass-migration. A negative experience of the abuse of emergency power led Japan to omit an emergency clause in the 1947 Constitution. This has led to recent debate divided between those arguing that a clause would limit the potential for abuse; and those who contend that the existence of such a clause would be open to abuse.

Many states are highly prescriptive in the conditions attaching to a state of emergency. The Constitutions of Estonia, and Chile define varying levels of emergency, each with corresponding powers and conditionality over their use. The Constitutions reserve the most serious levels to parliamentary approval. Both declared a state of emergency: in Chile the estado de catástrofe; and in Estonia the eriolukord, which are declared by the executive, and do not require parliamentary approval. Both countries notified derogations under the American Convention on Human Rights [ACHR] and ECHR respectively, and also under the ICCPR. On paper this is compliant with the rule of law and individual rights. In practice, however, both states reveal worrying trends in their use of emergency powers.

Ultimately, even where constitutional provisions on states of emergency are robust from the perspective of democratic oversight, individual rights and the rule of law; it is clear that compliance also lies in executive commitment to constitutional and legal order, and whether or not there is sufficient separation of powers to ensure it. Without an independent judiciary or parliamentary oversight to enforce constitutional norms, a constitution is little more than words on paper.

The Limits of Executive Action

Due to the urgency of action needed in emergency conditions, it is understood that executive measures do not undergo the same level of scrutiny or stages of approval. This does not mean, nor should it, that the use of emergency power is without limits or conditions, or that state agents should be allowed to act with impunity. The underlying danger of the use of emergency power, is that it may be used to introduce government policy without legislative debate or to consolidate power in the executive, and not to mitigate the negative consequences of an emergency. The almost unlimited legislative power given to the Hungarian government has since been used to suspend the operation of the GDPR in Hungary, and to transfer the most profitable revenue-sources for local governments to county governments in areas controlled by opposition parties – both of which have no plausible connection with COVID-19.

Executive action without legal justification or adequate legal basis, exemplified in unpublished decisions and government circulars, can create legal chaos in a situation that calls for clear communication, and legal certainty. The dangers of vague provisions make executive exploitation possible. One of the most pressing concerns is temporality: although the current crisis will not be permanent, some countries are permanently shifting the balance of power resulting in executive decision-making that is all but unaccountable. Executive action should always be subject to important restraints including political (eg parliamentary approval) as well as substantive (eg the inviolability of absolute rights) and procedural safeguards (eg judicial review).

When seeking to identify best practice to protect against unintentional misuse or downright abuse of power, there are some simple, and universal principles. The delegation of power must be time-limited, and clear as to legitimate scope for its use. The use of power must be legally prescribed and proportionate to that legitimate aim. There must be meaningful oversight by an independent body, as well as the possibility for review for those subject to the law.

The impact of COVID-19 on Democracy, Justice and Human Rights

#Democracy

Ensuring the continuity of the core constitutional functions of the legislature has been a challenge for all countries. Some states adopted provisions to allow for virtual assembly and voting, or enacted modifications to procedures, in order to allow institutions to function as normally as possible. In some cases, however, the parliament was entirely left out, suspended or threatened with dissolution by the executive. States in the middle of election cycles have been faced with the challenges of governing a pandemic without a government. This led to interim governments adopting sweeping measures, and also to power struggles when minority government have tried to handle the crisis. The separation of powers is designed to ensure constitutional checks on the use of power. It is all the more important in a situation where urgency and extreme measures can be justified, but becomes difficult to sustain where the executive is at odds with the Parliamentary majority.

Elections and referenda have been cancelled or rescheduled in many countries. The disruption of which has caused significant concern. However, of equal concern are states that stick to their election schedules, but ignore constitutional and international standards regarding free and fair elections while also failing to introduce any protective measures, forcing the electorate to choose between their democratic right to vote, and the health of themselves, their family and their community. Poland’s surreal capitulation on the presidential election exposed constitutional tragedy amid political farce, when the leader of the executive announced the Supreme Court was going to annual an election that had not happened, before the Supreme Court has become seized of the matter.

Where elections are suspended, unfree and/or unfair, public expression of dissatisfaction with government is reduced to political campaigning, and public protest both of which are difficult, if not impossible, under current lockdown or restricted movement regimes. In Chile, the government has been able to capitalise on the pandemic to curb or limit civil unrest and public protest against the government. Where ‘chaos is method’, the opportunity to consolidate autocratic regimes is enhanced in emergency: Venzeuela, in a state of exception since 2016, has been able halt public protests and to hide a widespread lack of resources, through measures designed to limit the spread of the virus. However, where such capacity for public demonstrations is limited, alternative means including digital protest have risen. Online activism can shape public policy. In Indonesia, the ‘#LockDownOrDie’ movement on social media led to changes in government policy amid widespread distrust of the executive’s handling of the situation.

While strategies to combat COVID-19 have limited democratic input, and parliamentary oversight, it is inevitable that the decisions made by governments now will have lasting impact on their electoral systems (and prospects) in the years to come. 

The Interdependence of Rights

There has been a broad variance in the degree to which states have engaged with international institutions by derogating from international human rights instruments including the ACHR, ECHR and ICCPR. For many states derogation appears a mere extension, or pure formality, of the declaration of a state of emergency. For example, Chile is one of the few states which actively distinguished constitutional and international rights. Thus, there is debate whether derogation can actually limit potential rights violations, particularly where enforcement of the ACHR, ECHR or ICCPR is comparatively weak.

On a national level, questions have been raised regarding the balance of individual rights and public health, as well as the question of what legal and moral duties are owed by the state to its citizens. Some have argued for the (re)conceptualization of the ‘right to life’ to be interpreted as the positive obligation to protect the life and health of the public, and even as a constitutional duty to prevent the spread of coronavirus. The central contention is that ‘rights-respecting’ public authority should be about more than avoiding interference with individual rights, but instead be understood as the responsibility of government to ensure the welfare of its citizens. One common understanding, however, is that the ‘rights versus public health’ paradigm is fundamentally flawed. Rights-respecting measures which secure public confidence are ‘more likely to be more effective and sustainable over time than arbitrary or repressive ones’.

Countries with well-functioning healthcare systems, particularly those considered ‘sacrosanct’ in the political culture, are inevitably better placed to tackle a major health crisis than those which have weak or dysfunctional health systems, or whose systems are already in a state of total collapse prior to the crisis. Chronic shortages in medical supplies, coupled with limited availability of tests, and compounded by wide-scale and systemic corruption where sudden influx of emergency funds incentivises opportunism, create the worst panoply of conditions for the spread of infection.

Measures implemented to prevent or slow the spread of the virus have a disproportionately negative impact on vulnerable categories of people including the elderly, prisoners, those with physical or mental disabilities, migrants, ethnic minorities, and refugees. The sudden onset of mass unemployment of part-time, low-income, and informal workers, coinciding with the shut-down of childcare and schools, has also disproportionately affected women. Worse, under stay-at-home orders, women and children also face escalating rates of domestic violence, and a majority of governments have failed to take preventive measures. For the millions living with poverty, malnutrition, or with high rates of potential comorbidities including TB and HIV, in cramped conditions and with limited access to water, the most prevalent political and medical messaging of ‘stay home and wash your hands’ reveals ignorance of endemic socio-economic disparity and reveals the endemic structural inequalities. For the most vulnerable populations, the threat of COVID-19 is a choice between staying home or risking infection, but must instead be weighed against losing access to the most basic needs of food, adequate housing, and sanitation.

Virtual Justice

A central concern of legal analysts has been the functioning of the judicial system. This is necessary not only for the ordinary administration of justice, but even more important in respect of judicial review (where provided) of actions taken in an emergency. As could be expected, the COVID-19 crisis has brought delays to sometimes already back-logged justice systems. Many states have nevertheless endeavoured to ensure the continuity of the courts, for example through suspension of oral arguments and their replacement with written submissions. The most notable innovation, however, has been the increase of legislative measures to introduce ‘e-justice’ through audio-visual conferencing, or measures to make remote access more widely available. Nonetheless, putting defendants, and particularly vulnerable defendants, on a ‘Skype trial’ raises a number of concerns about how to guarantee a fair trial online. Practical concerns of access to a computer, and a stable internet connection are at the forefront, while – almost inevitably – the neutral background of a courtroom is replaced with the intimate (and potentially prejudicial) setting of a home, hotel, prison, or refuge.

These are, however, practical and not principled challenges which can be overcome. It is far more concerning when the justice system is suspended altogether. Relying on the justification of preventing the spread of the virus, executive restrictions on access to justice have been the greatest worry: measures have included freezing courts, limiting access to ‘extremely urgent’ or critical cases; or giving sole access to violations of coronavirus measures. Israel closed courts to all except urgent matters when, coincidentally, a senior political figure was on trial for corruption. This endangers the protection of rights through the ordinary administration of justice. While there are practical (and even sartorial) issues associated with allowing remote trials, virtual justice is always better than none.

On the question of the review of executive use of emergency powers, and the intensity of review it should be afforded, proportionality must be decisive. However, many states may not employ this reasoning at all, as the exclude emergency measures from the scope of review. Hungary closed ordinary courts altogether and thus cut off the possibility for review of the proportionality of measures introduced under emergency conditions. In Thailand, an ouster clause precludes administrative review of regulations made under emergency legislation. In Czechia, even while courts have annulled some restrictive measures by the Ministry of Health, they still refused (on a split decision) to review the declaration of a state of emergency for lack of competence. The Romanian Constitutional Court refused jurisdiction but, in doing so, implicitly held that presidential decrees are outside the reach of both the Parliament and the Constitutional Court. Conversely, the Spanish Supreme Court refused to review the declaration of a state of emergency as this was properly a matter for the Constitutional Court.

The exclusion of executive action entirely from the scope of review is a profound concern, especially where courts lack any degree of independence, and particularly where there is little scope for either political or popular objection through elections or protests. Court action can be a positive agent for change, and the role of an independent judiciary is essential, particularly where emergency measures might unconstitutionally limit rights.

Controlling COVID-19

Trust and Transparency

Measures on the scale witnessed across the world, from social distancing to national lockdown, rely to the greatest extent on public acceptance and commitment. A common factor amongst the most successful states in epidemiological terms has been high levels of transparency in the decision-making process. Where transparency underlies the government action, there is a strong correlation with public trust in the actions that are taken. Related to this is a nuanced question of expertise: in an epidemic, who should make the rules to regulate risk? Across the world, there have been instances of a radical reformulation of roles: medical practitioners as detention officers; military as police; police and soldiers as doctors; doctors as politicians; and the public as epidemiologists. The answer most likely lies in a co-ordinated effort of diverse and relevant expertise.

Sweden’s measures in response to COVID-19 have been reported internationally, by some to exemplify a preferable alternative to highly restrictive measures. Premised on high levels of public trust, the Swedish approach was to collectivise responsibility, both at government level and among individuals. Public health recommendations on social distancing were introduced, accompanied by measures (albeit comparatively minimal) to limit gatherings of groups, as well as by closures of businesses and schools. The efficacy of this approach, advocating primarily social responsibility over lockdown might however be vindicated in the future with a lower rate of resurgence. However, Sweden has currently a higher mortality rate than its neighbours Denmark, Norway and Finland.

Nevertheless, trust-based policies so far deliver the best outcomes. New Zealand has all but eliminated COVID-19 within its borders doing so by a combination of ordinary legal powers, and some emergency provisions, centrally driven by social nudges communicated through clear, consistent and constant government messaging. As an ‘effective rationalist’, Prime Minister Jacinda Arden repeatedly underlined the central message of social responsibility and thus met ‘rule-of-law expectations about clarity, certainty, accessibility and congruence in application’. Iceland’s ‘rule of common sense’ has similarly been driven by clear government guidelines, recommendations and daily expert advice, with promising outcomes.

Where there is an emphasis on individual responsibility, guidance must be clear, consistent and accessible. Incoherent or inconsistent government messaging with frequently changing rules and restrictions leads to critical uncertainty and low compliance. In the worst case, this can ‘[drown] the inhabitants in a hypertrophy of legal rules, exceptions, and exceptions from the exceptions’. No one should be expected to have a legal education to understand what rules apply to them. Equally, elected or government officials should never ignore the rules that they have themselves designed, at the risk of critically undermining their own legitimacy and the rules’ efficacy.

Clear, accessible, consistent, correct, and constant guidance, with officials leading by example, is essential. Ireland sets a good example here with providing clear and accessible information. This not only has the effect of tackling the spread of misinformation on the virus but is also critical to ensure legal certainty and guarantee the transparency of government action. Based on current trends, the Nordic, Irish and Kiwi experience evidences that high trust in transparent government action, paired with a strong sense of personal social responsibility based on clear and consistent expert advice, can lead to the most positive outcomes on average.

Sanction and Surveillance

One of the most concerning and widely reported developments has been the criminalisation of breaking COVID-19 measures, often with penalties that are disproportionate to the country’s median wage: for example, large fines, and extended prison sentences for non-compliance with lockdown orders. States which have introduced provisions criminalising acts likely to spread the disease has led in some cases to mass arrests. While not only caused by the pandemic, extra-judicial killings and police brutality in the enforcement of government mandates have also occurred in some states. Some sanctions are not only criminal: for example, Poland has introduced disproportionate administrative fines for breach of lockdown orders. The recourse to administrative rather than criminal measures avoids the obligation of a court hearing and the opportunity for defence. For that reason, it is also potentially unconstitutional.

Paired with criminalisation of COVID-related behaviour are state censorship and restrictions on freedom of expression. A number of states have introduced criminal offences related to ‘miscommunication’ or the publication of false or misleading assertions related to the epidemic and the measures introduced by governments to tackle it. Such offences are often drafted in terms general enough to ensure that political opponents fall within their scope, and have been introduced as permanent changes to criminal codes. In Belarus, otherwise in a state of denial concerning the reality of coronavirus, a chilling effect on reporting the virus has been created through arrests of journalists and doctors, thereby masking the true extent of infection and mortality.

A related concern is the right to privacy as countries worldwide introduce provisions to allow the collection of personal data in an effort to track the virus. The forms of tracking have varied: from the requirement to send ‘selfies’ as proof of staying at home, to the deployment of tracking software, designed for national security or military purposes. The key concerns are that the use of this technology by government or the police is without either the consent or knowledge of individuals; and without any degree of judicial or political oversight. While tracking of personal data has wide potential for misuse, unless the population are strongly following norms already, it may have little potential for limiting the spread of COVID-19. Concerns over privacy and the potential for misuse have produced widespread criticism and public protest. Israel was one of the first states to moot the possibility of the security services’ use of software to track the virus, but later backed down following protests. Similar proposals were introduced, and rejected after public outcry in Slovenia. In contrast, China has tracked its citizens prior to COVID; and Muscovite citizens on public or private transport are now required to carry a digital pass issued by government as part of a system which is now expanding throughout Russia.

Higher rates of criminalisation and surveillance powers strongly correlate with the militarisation of response to a civilian and public health crisis. An authoritarian rhetoric of ‘war’ with the virus has been used to justify extreme or disproportionate measures and/or restrictions on personal liberties. This has much more concerning dimension in the assignment of civil functions to the military: for example, placing military personnel in positions of power in government, hospitals and private enterprises; the assignment of policing functions to the military, and even the prosecution of ordinary offences through military courts. A militarisation of response correlates with less leniency (for example, targeting the homeless) the higher use of force as a means of control. The militarisation, criminalisation and severe sanctioning of coronavirus-related behaviours has not, however, significantly correlated with better outcomes based on current trends. It does, however, strongly correlate with autocratic and repressive governance, and a weaker separation of powers.

Urgency, and Competence in Crisis

The obligation of states to act can be understood as the question of competence to act, particularly where powers are divided among local, regional, and federal levels. While divergent approaches to address the pandemic are (at least conceptually) coherent across national borders, they are irrational within them. Inconsistent rules on social distancing, travel restrictions, or the obligation to wear masks, create conflict and uncertainty in practice.

Among federal systems, emergency powers are usually concentrated at federal level (the USA and Australia are notable exceptions), while healthcare is devolved to state level. Where the federal executive is at odds with state government in downplaying the threat of COVID-19 the latter must step in to fill the vacuum in leadership. While there is a convincing rationale for the localisation of emergency response based on speed and local-awareness, its ultimate success depends on being complemented by a coordinated national response. The political gain of relegating responsibility (and hence blame) to adopt measures to regional governors is not offset by lower mortality rates, or likely to ensure faster economic recovery. Denial or a reluctance by federal executive power to respond cause the high costs of underreaction, and are only now beginning to show: the USA and Brazil have quickly become the countries with the highest number of cases globally.

A lack of competence (rather than denial) can limit international intervention: the EU has few relevant legal powers in healthcare, as the area remains the exclusive competence of its Member States. There has been, as of yet, little meaningful or coordinated EU response, which has led the EU to appear, at least institutionally, to fail critically at a time when it was already in crisis. Member States, rather than the EU, closed internal borders. The lasting impact of this absence could likely be a rebalancing of power (even if it was only ever a perception that there had been an asymmetry) in favour of the Member States. However, this is an incomplete vision. As countries begin to ‘exit the emergency’, the EU could assume a critical role in the coordination of the re-opening of borders, reflecting the larger role of international institutions in underpinning international cooperation.

Ultimately, the speed of the reaction determines the outcome: the earlier the action, the lower (or even non-existent) the mortality rate. As the UK has demonstrated to its terrible loss, delayed reaction clearly correlates with the highest mortality rate in Europe, and the second highest in the world. While there are justified concerns in overreaction, the new parable of state action will warn against underreaction in a time of desperate urgency. It will also underline that while urgent action is essential – it must confirm in so far as possible in the first instance to the principles of legality, legal certainty, clarity, and transparency; or be corrected to align to these principles as soon as practicable.

Exiting States of Emergency

There is neither a cure to COVID-19 nor vaccine to the coronavirus. At the time of writing, confirmed cases have reached 5.5 million worldwide. After months of lockdown, restrictions and closures, states are now beginning conversations about how to exit a state of emergency. They must consider the how to design and implement long process of removing or reducing measures to begin social and economic recovery. They do so in ‘a geopolitically shattered world’. Notable for their absence from discussion are global and transnational institutions: they have largely been absent from decision-making. The measures adopted to combat COVID-19 have been bound by both political borders and by the reach of national budgets. This potentially heralds a new era of ultra-nationalism and protectionism. But this is not inevitable. While nationalism has characterised the initial legal and political responses to coronavirus, international cooperation on trade, movement, medicine, and research will determine the next stage as we exit states of emergency.

Once the immediate crisis has passed, there will be an opportunity for states and international bodies to examine and review their constitutional and legal architecture, as well as health and crisis response preparedness. In the reconstruction of emergency frameworks, I offer these observations based on currently available data, and the assumption of continuing trends. First, declaring a state of emergency or relying on ordinary legislation made the likelihood of abuse of power no more nor less likely. Rather, the effectiveness of legal safeguards against abuse depends on executive observance of the rules, and on the strength of the separation of powers to enforce it. Second, the militarisation of response correlates strongly with criminalisation and disproportionate sanctions, but it does not strongly correlate with better outcomes. Third, silencing journalists and doctors can lead to the spread of misinformation; while not engaging with (often justified) criticism of laws or applying double standards, can mean rules are poorly followed. Fourth, and finally, while it is too early to identify the best practice, there is emerging evidence of good practice: those state policies based on legal certainty, transparency, clear communication, and early-reaction have strongly correlated with lower mortality rates, and the sooner lifting of restrictions. These principles are simple, and universal.

Ultimately, while national isolation has reduced infection; international co-operation will ensure recovery.

With thanks to Alice Donald, Stefan Enchelmaier, Ciara Staunton, Laurent Pech, Laura Bramley, Kim Lane Scheppele, and Tom Daly