Coronavirus and parliament: A brief history of sunset clauses

Is this kind of provision an adequate safeguard against the normalisation of emergency powers?

April 28, 2020
article header image

Sunset clauses are provisions that determine the expiry of a law or regulation within a predetermined period. Through their use, legislation or a section of legislation automatically ceases in its effect after a certain time. In theory, sunset clauses can limit the potential reach and adverse impacts of governments that adopt legislation in response to emergency situations, such as the current Covid-19 pandemic. Indeed, the Coronavirus Act 2020 contains several such provisions and this was viewed as essential in limiting the duration of radical powers. However, it is important to adopt a healthy degree of scepticism. Sunset clauses do not always serve as the safeguard people expect. As the coronavirus emergency continues it is important to reflect on how they work and where they come from, to ensure best practice in the months ahead.


The precise origins of sunset clauses are contested. Some locate their roots in Roman law but the first philosophical reference is in the laws of Plato. At the time of the Roman Republic, the empowerment of the Roman Senate to collect special taxes and to activate troops was limited in time and extent. Those empowerments ended before the expiration of an electoral office, such as the Proconsul (Antonios Kouroutakis). In the UK, sunset clauses were employed by parliaments by at least the time of the reign of Henry VII and appeared in statutes by 1500. The series of Habeas Corpus Suspension Acts enacted from 1689 all had specified expiry dates; up until 1777, these acts lasted on average for five months. In the US, Chris Mooney traces the history of sunsetting back to the writings of Thomas Jefferson. Jefferson believed that a version of sunsetting sprang directly from natural law: "Every constitution… and every law," he wrote, "naturally expires at the end of 19 years," which was considered the length of a generation in his era. Sunset clauses are also readily identifiable in some of the first post-independence legislation. For instance, enacted as part of a legislative package known as the Alien and Sedition Acts during a period of bitter partisan division between the Federalists and Republicans, the famous Sedition Act of 1798 criminalised the making of certain statements critical of the government. In accordance with its sunset clause, the Act expired on 3rd March 1801, the last day of John Adams’s presidency.

Regardless of origins, the sunset clause, as a legal device, makes the law to which it applies temporary, whereas legislation is ordinarily permanent in that it persists unless and until repealed by subsequent legislation. The reasons for the inclusion of a sunset clause differ depending on the legislation in question and the context in which that legislation is adopted. However, the importance and contested nature of sunset clauses is best exemplified in reference to legislation adopted to combat terrorism, a reference that is particularly useful when considering these devices in the context of Covid-19.

Contemporary use

Sunset clauses have mostly been associated with emergency legislation introduced in the wake of the terrorist attack of September 11. In the US, for instance, the Patriot Act was enacted in late October 2001, and made sweeping changes to many laws pertaining to counterterrorism. It included a sunset clause to trigger at the end of 2005. In Canada, the Anti-terrorism Act (ATA), enacted in December 2001, was the cornerstone of Canada’s initial legislative response to the 9/11 attacks. The ATA created new criminal offences as well as a host of new counterterrorism powers. The ATA’s sunset clause provided that certain provisions—those relating to investigative powers—expired at the end of the 15th sitting day of parliament after 31st December 2006, unless extended (for up to five years) by a resolution passed by both Houses of Parliament. In the UK, The Anti-terrorism, Crime and Security Act 2001 (ATCSA) provided for further terrorist offences and counterterrorism powers not covered under the existing Terrorism Act 2000. The operative sections were set to expire in 15 months, but the secretary of state could extend the expiry date for 12-month periods. The Terrorism Act 2006, enacted after the London bombings of July 2005, was subject to a 12-month sunset clause, but the expiry date was again extendable by parliament.

Sunsetting a piece of emergency legislation attempts to strike a particular balance. Special and extraordinary measures are enacted to respond to crises, in derogation of existing standards and rules. Sunset clauses enable the state to respond effectively while keeping the exercise of emergency powers within the rule of law. Part 10 of The Anti-terrorism, Crime and Security Act 2001, for example, allowed the police to forcefully obtain fingerprints and other identifying features from an individual to ascertain their identity and for the Ministry of Defence Police to operate outside of their normal jurisdiction in certain circumstances. It also included controversial measures which allowed for the indefinite detention of non-national terrorist suspects. Given the nature of these powers, there are significant risks associated with over-extension in ways that are disproportionate and excessive. They can have adverse effects on, for example, the right to life, a fair trial, liberty and security, and freedom of assembly and association. Moreover, there is the risk that emergency provisions become the new normal, thereby entrenching systems of power that can create widespread and systematic rights violations. Sunset clauses can, in theory, strike a balance between granting the powers necessary to respond to particular circumstances, while also ensuring that rights are not unnecessarily eroded beyond the time period required.

It is the attempt to achieve this balance that explains the inclusion of sunset clauses in Covid-19 legislation. As with responses to terrorism, emergency provisions have popped up across the globe, giving states powers that would, in normal times, be unfathomable. Sunsetting these powers can ensure that they are, in law, temporary measures. In the UK, for instance, section 89 of the Coronavirus Act provides that the majority of the provisions will expire after two years. However, this period may be extended by six months or shortened in accordance with section 90. The government also accepted an amendment, which introduced the requirement that the operation of the Act must be reviewed by parliament every six months (see section 98). In Ireland, the powers under The Health (Preservation and Protection and other Emergency Measures in the Public Interest) Act 2020 will cease to have effect after the 9th day of November 2020, unless a resolution is passed by both houses of the Oireachtas (parliament) to approve the continuation of the measures. In Scotland, the Coronavirus (Scotland) Act includes a “sunset clause” according to which most of it will automatically expire six months after it comes into force. MSPs will be able to vote to extend this for another six months if necessary, and then for another six months after that, but this is the absolute limit—so the measures in the Act have a maximum duration of 18 months.

Thus, sunset clauses, when included as part of emergency legislation, can be seen as a mechanism by which democracies devise ways to accommodate governmental powers within a pre-established legal framework, rather than leave it to governments to use raw power and untrammelled discretion to deal with emergencies in an unregulated way.

A healthy scepticism

At the same time and as I have argued elsewhere (see here), there are reasons to be sceptical about sunset clauses. Firstly, their inclusion can be used to garner cross-party approval to push through contentious pieces of legislation only for them to be later removed. By including sunset clauses, opponents of particular bills (largely because of the wide powers that are afforded under them) are reassured that any measures are time-bound. But this does not necessarily safeguard sunset clauses from political lobbying and subsequent amendments post adoption. For instance, the first Bush tax cut was passed in 2001 to terminate at the close of 2010. No sooner had the laws been passed than their Republican backers launched a pre-emptive strike, criticising the sunsets and attempting to undo them. The Republican-led House of Representatives subsequently voted to make permanent the repeal of the estate tax contained in Bush's first tax cut. Thus, in the context of taxation, Manoj Viswanathan assesses that “sunset provisions… are the product of political manoeuvring designed to bypass budgetary constraints and are exploited as a means of enacting permanent legislation under the guise of an ostensible expiration date.” In short, a sunset is no guarantee that come the proposed end, the clause will remain in place.

Secondly, the addition of sunset clauses notwithstanding, pieces of emergency legislation can remain in force long after the proposed sunset. It will be remembered that part of the logic of sunsetting legislation is to ensure that emergency powers do not normalise. But because most sunset clauses allow (and indeed require) subsequent extensions, there are no guarantees that the legislation in question will cease. This is more likely when, with terrorism as with Covid-19, the enemy is invisible and thus unknown. In the US, for instance, the 2001 Patriot Act adopted in the aftermath of the September 11th attacks included 16 sections originally meant to sunset on 31st December, 2005. The Act was reauthorised several times in the following years following very limited evaluation. Similarly, in the UK, the Prevention of Terrorism Act 2005 (PTA)—another piece of emergency legislation—was renewed annually from 2005 until its repeal in December 2011 by the Terrorism Prevention and Investigation Measures Act 2011. This is not to say, of course, that sunset clauses are always extended. For instance, in Canada in February 2007, a motion to extend the expiry date by three years was introduced by the government (by this time the Conservatives). This motion was opposed by the other political parties, including the Liberals, the party that had been in power at the time of the ATA’s enactment. The government’s motion was ultimately defeated on 27th February 2007, and the provisions expired on 1st March 2007. Nevertheless, the history of sunset clauses illustrates a tendency for the legislation that they are supposed to terminate extending beyond the original time limits set.

Thirdly, when sunset clauses provide for further debate and scrutiny, the efficacy of the review process is of central importance. The mere provision of future scrutiny is no guarantee for the effectiveness of that process. For instance, the Counter-Terrorism Review Project highlights that in the 2003 debate in the House of Lords on whether to renew the Part 4 powers of the Anti-Terrorism, Crime and Security Act 2001, just four Lords spoke. This included the minister who had introduced the renewal order. Only 13 MPs attended the first debate in 2006 on whether to renew the Prevention of Terrorism Act 2005—the legislation which established the control order regime. In addition, the time allotted for debates on sunset clauses is also very short, often limited by parliamentary procedure to only an hour and a half (see here for discussion). In addition, there are questions regarding the most effective form of review. If parliamentary post-legislative review is the chosen approach, there may be problems associated with politicisation of the legislation in question. Should, then, the review be undertaken by an independent expert, committees of the House of Commons or Lords, or an independent group? If so, how democratic would the process be?

Fourthly, the necessary period of time between adoption and review and between different review processes raises additional issues. Although the UK’s Coronavirus Act allows for review after a period of six months, this may still be too infrequent. During the House of Lords review of Fast Track Legislation in 2009, for instance, The Better Government Initiative argued that “post-legislative scrutiny is all the more necessary” in cases of fast-track legislation, and that “it should perhaps be more frequent.” Such is the nature of the pandemic and such is the extent and wide-ranging nature of powers afforded under the Coronavirus Act (and similar pieces of legislation adopted globally), that more review processes might be required. But how might this be achieved in light of social distancing? Of course, many of these are issues that arise in the context of any review process, but they nevertheless demonstrate that there is a range of considerations that flow from sunset clauses, which require ongoing scrutiny themselves.

So sunset clauses have a rich pedigree and will continue to be included in emergency legislation adopted in response to the Covid-19 pandemic. They are unquestionably a useful mechanism by which to ensure that emergency provisions do not normalise, thereby entrenching powers that can adversely affect the enjoyment of individual rights and freedoms. However, there are limitations. They can exist on paper but have little impact in practice. They can be renewed on an ongoing basis, often with little or insufficient scrutiny. Adherence to sunset clauses must be scrutinised.