Blog Essay

States of Emergencies: Part II


In our earlier post, we described how the legal bases for countries’ coronavirus responses typically fall into three broad categories. First, some countries have declared a state of emergency under their constitutions, which allow them to take special measures, including restrictions of civil liberties, for the duration of the emergency. Second, some countries have relied on existing legislation relating to communicable diseases, health, and/or national disasters and used these laws as the basis for their COVID-19 response. Third, some countries have (also) passed brand new legislation that grant the governments new powers to respond to COVID-19.

When governments are granted powers to restrict rights, there is always a risk of abuse. To minimize these risks, three broad principles are important: (1) providing for legislative and judicial oversight of the executive, (2) limiting exceptional measures to those strictly necessary, and (3) ensuring that such powers endure only for the duration of the outbreak.  These principles, rather than those focused on whether the relevant authorities are constitutional or statutory, are the most essential in ensuring that liberal constitutionalism does not become a casualty of the virus. The different models, however, each have their own vulnerabilities when it comes to the risk of abuse. Each government’s invocation of legal authority, whatever form it takes, should therefore be carefully scrutinized for compliance with these principles. 

Ensuring Oversight

Taking these principles one at a time, what are the costs and benefits of the different models? Start with oversight. We distinguish judicial and legislative oversight, and they cut in opposite directions. Ensuring judicial oversight is easiest for the legislative models, whether relying on existing or new legislations. Courts do not typically scrutinize actions taken by the executive after a constitutional state of emergency has been declared, although they can scrutinize whether the state of emergency was properly invoked and whether other procedural requirements were followed. By contrast, the key feature of the legislative model is that the ordinary constitutional framework stays in place and all actions remain subject to ordinary procedures of judicial review. Of course, there is a question as to whether courts will actually exercise this power in practice. A body of literature has shown that, in times of emergency, courts widely defer to the executive, at least until the crisis is over. Certainly this has been the experience in the United States with regard to executive authority during wartime.  But in the current crisis, we are observing some courts insisting upon checks and balances in the COVID-19 response.  The Constitutional Court of Kosovo, mentioned above, demanded that the government pass new legislation that authorizes any restrictions on rights that are needed to combat the virus. The Supreme Court of Israel has likewise insisted upon parliamentary oversight of various emergency measures, while the Constitutional Chamber of the Supreme Court of El Salvador urged the government and Congress to pass a law regulating arrests for noncompliance with quarantine orders. And when Brazilian President Jair Bolsonaro ignored his public health authorities and continued to advise against social isolation, the country’s federal courts enjoined him. They also blocked his decree exempting religious services from isolation orders. These examples suggest that courts may be willing to insist on legislative oversight and to challenge government omissions in a public health context. At the same time, whether courts will be receptive to broader civil liberties claims remains to be seen.  For example, in a series of cases in Bavaria, courts held that the government can impose civil liberty restrictions only by law, but upheld the temporary restrictions as proportionate infringements on individual rights.

If courts are likely to be more involved in the legislative model, it is somewhat paradoxical that legislatures might have less of an oversight role in this model than they do under the constitutional model. When a constitutional emergency is declared, democratic constitutions typically require legislative involvement in both declaring an emergency and ending it. Well-crafted constitutional emergency provisions also provide an automatic expiration date, thus requiring legislative action to renew the emergency should the crisis continue. By contrast, legislative involvement in passing legislation is often a one-off event, and not always kept up to date.  Notably, it is unlikely that existing public health legislation was intended to authorize the kinds of measures governments have taken to combat COVID-19. Although these laws may authorize governments to order quarantines, it is hard to imagine that their drafters sought to authorize the lock-down of entire countries for an extended period of time. Typically, quarantines involve a relatively small number of citizens who have been exposed to a disease in a particular geographic location, rather than the entire population. In the case of India, there was arguably no legislative involvement at all, as the statutory basis for the nationwide lockdown dates back to colonial times.

One might think that the answer is to pass new legislation once the crisis arises, but this is no panacea: laws passed in the face of an ongoing emergency might include overly broad delegations of power,  and lack a mechanism for lifting measures passed under these laws. The Hungarian case represents an extreme example, but recent Norwegian legislation has also been widely criticized by civil society groups as being rushed and overly broad. Norway’s law permits quick layoffs, and allows the government to bypass existing laws, as well as to issue regulations without parliamentary debate. In short, actual legislative oversight tends to be minimal when government relies on legislation for its response, whether the law is new or old.

Minimal Intrusion

Turning to the second principle, which model is more likely to ensure a proportionate and minimally invasive response?  Given the extreme measures it authorizes and the potential for abuse, declaring a constitutional emergency should be a measure of last resort, to be used only if legislative tools are insufficient. States of emergency mark a distinct period, in which ordinary rules do not operate.  In some systems, when a constitutional emergency is declared, (some) rights can be suspended entirely, meaning that they are not enforced at all for the duration of the emergency. Such derogations are considered a kind of “safety valve” for governments, to relieve the significant pressures that a crisis creates.  Derogations are explicitly contemplated by international human rights instruments, such as the International Covenant on Civil and Political Rights and the European Convention of Human Rights.

One reason these international instruments address derogations is out of concern that they might generate more severe civil liberty restrictions than when rights are limited through ordinary legislation.  Most rights are not absolute, and can be restricted when doing so serves other important goals. But the ability to limit rights is not unlimited: in many constitutional systems, any limitations must be passed by law, necessary and proportional to address the threat, and subject to judicial oversight. In comparison with such limitations, a blanket derogation might invite more invasive restrictions on liberties than are strictly necessary. It might therefore be a good sign that some democratic governments have been hesitant to declare a constitutional state of emergency, since as a general rule, less restrictive measures should be preferred. At the same time, we should note that a legislative response is less invasive only when legislation is narrowly tailored and actually subject to judicial oversight.

Limited Duration

The third and in some ways most important principle to prevent a long-term deterioration of civil liberties is that the measures should be limited to the duration of the outbreak. Here, the drafting of new legislation might be the most dangerous approach.  The sense of crisis in some countries has led to legislative delegations that are overly broad, do not include a mechanism for repeal, and have no sunset provision. By contrast, while the measures taken during a constitutional state of emergency might be draconian, the state of emergency is temporary. Likewise, under the model in which existing legislation is the main basis of the response, the COVID-19 response might certainly exceed what the drafters intended to delegate, but no overly broad and hastily crafted new laws will stay on the books.


These three principles of oversight, tailoring, and limited duration can be applied to whatever legal authorities are used to tackle the COVID-19 challenge. In particular, if new legislation is necessary to deal with the crisis, it should be narrowly tailored to the particular threat, of limited duration, and subject to oversight. The UK coronavirus law is a good example. It applies to that disease only; furthermore, its measures are specifically temporary and limited to the duration of the threat, and the repeal of the act will automatically remove other related temporary measures in the statute book. Many of the laws currently being passed in other countries do not meet these basic requirements. If they do not, it might be preferable to rely on existing legislation or to invoke the constitutional framework for dealing with emergencies. Otherwise the long-term effects of the virus will go well beyond the realm of public health, leaving democratic institutions that are crippled, even as populations recover.