2.2.1.Coronavirus Legislation

During the coronavirus pandemic, the requirement for legislation to be published well in advance of its coming into force has been breached many times. There have been constant complaints on the lack of prior publication of statutory instruments regulating lockdown.⁶⁰ For example, the All Tiers Regulations were made at 6am and came into force at 7am on 20 December 2020.⁶¹ They weren’t even laid before Parliament until the next day. Even worse, the very first set of coronavirus regulations were made at 1pm on 26 March 2020 and came into force at 1pm on 26 March 2020.⁶² This does not just relate to legislation at the start of the pandemic. The Health Protection (Coronavirus, International Travel, Operator Liability and Information to Passengers) (Amendment) Regulations (Northern Ireland) 2022 were made on 7 January 2022, sometime around 1 or 2am. Most of the provisions contained within them came into force at 4am of the same day.⁶³ Virtually every single Report on coronavirus legislation issued by the Bingham Centre for the Rule of Law raised the point of basic unfairness of a law being applied against a citizen when the citizen had no way of knowing what that law was.⁶⁴ Reliance on Government press releases is no substitute for the actual legislation being published in advance.

2.2.2.Prerogative Legislation

Prerogative legislation is a very peculiar type of UK legislation made by the Crown (via Ministers) without any recourse to the legislature. I have previously considered accessibility of legislation made under the Royal Prerogative.⁶⁵ In investigating UK prerogative legislation relating to British Indian Ocean Territory (BIOT), I concluded that, despite the views set out above on the importance of access, this legislation was not accessible. Tomkins criticised it as ‘unaccountable, indeed wholly secret rule-making’.⁶⁶

It could be thought that the lack of official promulgation of BIOT legislation is accidental. However, it is clear that official government policy is to make it inaccessible. For example, the Secretary of State sent a confidential telegram to the Governor of the Seychelles on the subject of some of this legislation. In paragraph 3 he stated ‘our concern is that publication of Ordinance and any regulations thereunder should be limited to minimum by law so as to attract as little attention as possible’.⁶⁷ The Governor responded ‘Ordinance would be published in BIOT Gazette which has only very limited circulation both here and overseas after signature by Commissioner. Publicity will therefore be minimal’.⁶⁸ It is inarguable that, in limited cases in the UK, inaccessibility is a deliberate policy.

Even with less controversial prerogative legislation, for example the legislation regulating the civil service in Northern Ireland, it is very difficult to actually find those laws. I speak from personal experience of drafting non-prerogative legislation on this subject, and trying to find out what the prerogative legislation actually contained. There was no easy way to find it.

2.3.1.Cost

Hicks gave a presentation designed to assist drafting offices from around the world with purchasing electronic legislation systems.⁶⁹ The most basic system would cost between US $60,000 and US $120,000 (at 2011 prices) and would not include a website. In his view, a system that included a website and a way of updating legislation on that website would cost between US $120,000 and US $3,000,000. In Northern Ireland, the tender documentation for a system for drafting, printing and amending primary legislation suggested a contract price of £410,000.⁷⁰ It is possible to have a cheaper system. According to Marsh-Smith, the Isle of Man developed a system with a target budget of £60,000.⁷¹ Cost-based objections to a new legislation system can be short-sighted. In my view, a modern system can provide large financial benefits. These financial benefits ought to quickly recoup the cost of the financial outlay. For instance, time savings for civil servants researching legislation, avoidance of mistakes arising out of inability to find the law and attractiveness of the jurisdiction to businesses and investors.

2.3.2.Volume

There is no irrefutable logic that there must be a large volume of statutes, but in practice the year on year increase in statutes does seem inexorable. Dhavan argued that part of the problem was a prolific Parliament producing too many laws.⁷² Margaret Thatcher stated that the opportunities for judges criticising prolix legislation were ‘increasingly rare’ due to the fulfilment of Renton’s recommendations.⁷³ Regrettably this has turned out to be a Panglossian prediction. The 2011 annual volumes of UK primary legislation run to 2,789 pages. The 2008 volumes of secondary legislation run to 10,818 pages. The House of Commons Library prepared a briefing note entitled ‘Acts and Statutory Instruments: the Volume of UK Legislation 1950 to 2012’.⁷⁴ The results from this detailed statistical analysis show that the figures cited above are, if anything, below the average annual page output of the last 10 years. The government’s own drafters agree, citing volume as a cause of complexity in legislation.⁷⁵ The vast tidal wave of legislation enacted every year makes it difficult to access the specific legislative provision required. Spencer refers to this as ‘binge law-making’,⁷⁶ and for Karpen one goal of legisprudence is to reduce the quantity of legislation.⁷⁷

2.3.3.Interconnectedness

As a former First Parliamentary Counsel put it, we have ‘an intricate web of laws’.⁷⁸ It is rare to find a statute that exists in a vacuum, unconnected to any other piece of legislation. If a new criminal offence is enacted, it presupposes the existence of a raft of other pieces of legislation. It requires other legislation to set out the powers of arrest, rules of court indicating the manner in which the offence can be prosecuted, evidential rules on what can be relied upon to prove the offence, sentencing guidelines, prison rules, judicial appointment procedures etc. Each of these are a degree removed from the original provision, but they do have some sort of connection to it. The more interconnected legislation is, the harder it is to make it fully accessible.

2.3.4.Dynamic Legislation

There is not one tidal wave of legislation; there are fresh tidal waves every year. New law is constantly changing old law. So a system of accessible laws doesn’t simply have to deal with all the laws in force at a particular date, it has to deal with all the new laws as and when they are made.

At the micro level, although there is a single date upon which a statute is enacted, it regularly comes into force on different days. Take for example the Criminal Justice (Northern Ireland) Order 1999:

•Articles 1, 2, 33, 35, 44 to 47, 53 to 56, 57 (partially), 58 (partially), Schedule 5 (partially), Schedule 6 (partially), Schedule 7 (partially) came into force on the 25th of July 1997.⁷⁹

•Articles 3 to 22, 24 to 32, 34, 36, 37, 48 to 52, Schedules 1, 2, 3 and 4 came into force on the 1st of January 1998.⁸⁰

•Part III came into force on 1 June 1999.⁸¹

Different commencement dates make it more difficult to meaningfully access legislation.

2.3.5.Authority

Legislation needs to be authentic and authoritative. Historically this has meant that laws must be published in paper form, or even on vellum as described by Bennion.⁸² It should be noted that ‘paper’ is not synonymous with ‘authentic’. The use of correction slips to remedy defects in published legislation shows that paper is no guarantee of accuracy. In some ways, a fully auditable electronic system provides more guarantees against interference with the content of legislation. Some jurisdictions have put electronic systems onto a statutory footing. The EU adopted Regulation 216/2013 on electronic publication of the EU Official Journal. Article 1.2 provides that the Official Journal be published electronically and that this electronic publication is authentic and shall produce legal effects. As another example, the long title of the Legislation Publication Ordinance in Hong Kong reads:

An Ordinance to provide for the establishment of an electronic database of legislation and approval of a website on which the information in the database may be published and accessed; to give legal status to copies of the legislation published on an approved website.

2.3.6.Paper-Based Process

The only thing that Parliament votes upon is the two-dimensional written page before it. This rules out many accessibility improvements that relate to the medium: no hypertext links, no pop-up windows, no ‘click here’ suggestions. Most of the modern technological tools developed to promote accessibility are not strictly part of the actual statute passed by Parliament. These are editorial additions that, although helpful, have not been authorised by the legislature, and are therefore not technically part of the legislation. If this position was reversed, it would be interesting how Parliament and the law would react – does a broken link mean that a statute is wrong? Despite such theoretical concerns, in a digital age, reliance upon a printed Act of Parliament limits the use of additional digital tools to enhance accessibility.

2.3.7.Bilingualism

There is an additional factor on availability when we consider the UK as a whole – the need for bilingual legislation. Acts of Senedd Cymru are required to be in both English and Welsh, and both texts are to be treated as being of equal standing.⁸³ Scotland has a Gaelic Language (Scotland) Act 2005, although there is no legislative obligation to have Scottish legislation in Gaelic language as used in Scotland. It has also been a long-standing policy goal of some political parties in Northern Ireland that there be an Irish Language Act that would make similar provision for Northern Ireland in respect of the Irish language. There is now a policy commitment contained in the New Decade, New Approach agreement to bring forward legislation on the Irish language in Northern Ireland.⁸⁴

There are far reaching consequences of a requirement for bilingual legislation. The most obvious one is the practical one of resources – the time and personnel available to enact legislation in two languages. Then there are matters of process, whether to simply translate one language into the other at various points in the legislative process, or whether instead to co-draft (following the Canadian model for French / English legislation). In Ireland, there is an obligation that ‘As soon as may be after the enactment of any Act of the Oireachtas, the text thereof shall be printed and published in each of the official languages simultaneously’.⁸⁵ Then there are the difficulties arising from words having different shades of meaning in different languages, meaning that a direct translation of an English word into a Welsh (or Irish or Scottish Gaelic) word may not always make sense, or convey the same meaning.⁸⁶ For example, if a statute in Northern Ireland was to translate the English word ‘field’ into Irish, what Irish word would it use: geamhar – a field of corn-grass, tuar – a field for cattle at night, reidhlean – a field for games or dancing, or any of the other 29 separate words for field in Irish as described by Manchán Magan.⁸⁷

It is beyond the scope of this book to advise on the best way to draft and enact bilingual legislation, see instead the analysis of Watkins.⁸⁸ Suffice to say that bilingualism can both improve and cloud access to legislation – more people can potentially now read the law, but preparing it in both languages takes time, and has the potential to create ambiguity when there are now two competing authoritative versions of the text.

2.4.1.Electronic or Non-electronic Availability

The primary focus in this chapter is on electronic availability of legislation. Hard copies of legislation remain useful, but electronic availability is more important for several reasons. First, the hard copy can swiftly become out of date as the law is amended. Second, in the information age, information is increasingly consumed not just electronically, but online. Having said this, there is no guarantee that electronic versions are more accurate than hard copies, or will last as long as hard copies.

According to Gee, the top ambition for university law libraries in the UK is to develop and enhance their electronic and digital resources.⁹¹ His research found that the top free website in UK law libraries was the website of the British and Irish Legal Information Institute at www.baili.org and the second most popular was the official government website on legislation at www.legislation.gov.uk.⁹² According to Bertlin⁹³ legislation.gov.uk has two million separate visitors per month, and provides more than 400,000,000 page impressions per year. Oksanen goes even further, arguing not just that electronic legislation must be available for human consumption, but must also be available for machine processing.⁹⁴

Having said this, there are dangers if legislation is only available electronically. There are those who, for economic reasons, may not be able to afford the technology necessary to look up laws online. There will also be those who live in places with limited electronic infrastructure to allow for online access. During Covid lockdowns, those who rely on internet cafes or public libraries for their internet access may find that route cut off. Finally, some people may have a disability that makes electronic access more challenging.

2.4.2.What Must be Available?

I said above that the principle is that legislation must be available – but what is the exact content of this, what is it that must be made available in order to comply with the Rule of Law? In my view, there are six aspects of availability.⁹⁵ The first four aspects are what Fuller would call the morality of duty – minimum requirements that must be fulfilled. The final two are morality of aspiration – a summit to aim at, although the legislation will still be valid even if it doesn’t crest the summit.

2.4.3.Aspect 1: Availability of Law as Enacted

Legislation must be placed online as soon as it is made. All that is required for this aspect is a snapshot of the law as enacted, not how it changes over time. Drilling down into this, it becomes apparent that there is an important but subtle distinction between the date a statute is made and the date it comes into effect. Although it is important for law to be available as soon as it is made, it is even more important for it to be available before it comes into force.

There is a quantitative aspect to availability. If a specific statute or the official legislation website is online but doesn’t appear near the top of search rankings, it is less visible to users. A great website that is little known or doesn’t reach the top of search algorithms isn’t truly accessible.

2.4.4.Aspect 2: Availability of Law as Amended

A legislative system needs to keep all legislation up to date, what Benedetto refers to as interventions of compilation and consolidation.⁹⁶ This is one goal of the New Zealand Public Access to Legislation project.⁹⁷ This means that where new legislation amends old legislation, the old legislation on screen should fully reflect all those amendments. The system needs to tell users what the law is today, as well as what it was when it was enacted. This is the key aspect of a successful system. If the system cannot tell the citizen what the law is, it has failed.

There is one specific tool that has been used in this regard in the UK – the Keeling schedule. I deal with the Keeling schedule in chapter five under the heading of intelligibility.

Aspect 2 is complicated by commencement provisions, or more specifically, by the practice of different provisions commencing at different times. In the past, an Act was made and then a reasonably short period afterwards it came into force. Now it is quite common for the commencement of an Act to spread out over a long period of time, with different provisions coming into force at different times. Sometimes the different commencement dates are set out in the legislation itself, sometimes there is a power of a Minister to commence different provisions at different times. This creates an extra level of complexity for users. In addition to tracking what legislation has been enacted over time, they must also track what legislation has been brought into force over time. For a bad example, see the Public Services Ombudsman Act (Northern Ireland) 2016. In section 64 of that Act:

•Subsection (1) sets out provisions coming into force on 1 April 2016;

•subsection (2) sets out provisions coming into force on the day after Royal Assent;

•subsection (3) sets out provisions coming into force on 1 October 2016;

There are several techniques to address the difficulties posed by different commencement dates. First, the legislator can confirm that a proposed legislative provision will actually be used. There is no point in clogging up the statute book with a provision that the government are not sure will ever be brought into force. For example, as the Easter Act 1928 (introduced as a Private Members’ Bill) has never been commenced, why was it ever enacted? Second, the legislature can enact a simple commencement provision that commences the entire Act in one fell swoop, rather than having it commenced in fits and starts. Having said this, a complex set of reforms may need to be spread over time. Third, a commencement date can be inserted on the face of the Act. This way the reader will at least know, from reading the Act, when it comes into force. This tool is less necessary as there is now a power to amend legislation to include the actual commencement date on the face of it.⁹⁸ Fourth, where the government seeks a power to commence an Act by future subordinate legislation, a limited rather than extensive power can be given. For example, a power to commence the Act, rather than a power to commence different provisions of the Act on different days. Failing this, there ought to be a clear justification given why different provisions need to be commenced at different times, not just for reasons of administrative convenience.

One problematic issue here is where defective legislation is corrected judicially on the principles set out in Inco Europe v First Choice Distribution.⁹⁹ Although drafting mistakes are obviously to be avoided, it is inevitable that they will occasionally occur. Overt judicial correction of drafting mistakes (or what Bennion calls a rectifying construction)¹⁰⁰ is an established aspect of the UK legal system. The point here is not to question the limits of that power to correct, but to ask what consequences it has for the availability of legislation. Correcting a mistake may do justice in the individual case before a judge, but will not serve the population as a whole unless they know that the mistake has been corrected. The reader of the statute book will still see the incorrect words used in that statute and will not have an easy way of knowing what the judicial correction is.

If a mistake is found in legislation (by a judge or by anyone else), it is incumbent for it to be rectified in the statute book as soon as possible. Although the legislature may be embarrassed about a long title that includes the phrase ‘An Act to rectify a mistake in [previous legislation]’¹⁰¹ or ‘correcting a mistake in the repeals effected by [previous legislation]’¹⁰² it is still better to correct it than to allow the mistake to fester on the statute book.

There was an attempt in the past to pass a Bill entitled the Acts of Parliament (Correction of Mistakes) Bill. It was designed to allow for a simple procedure for Parliament to fix a mistake in an Act. The Bill was withdrawn before there was a chance to enact it, leading to the famous headline: ‘Mistake Bill was a Mistake: Minister Admits’.¹⁰³

2.4.5.Aspect 3: Law at a Point in Time

A legislative database needs to state what the law was at a particular point in time in the past. There will be many situations when past legal rights and obligations will be relevant – when a person did X, was X unlawful? For example, consider the Historical Institutional Abuse Inquiry that was obliged by statute to investigate child abuse in Northern Ireland between 1922 and 1995.¹⁰⁴ Part of its remit necessitated knowing what the legislative requirements were in respect of the care of children between those dates. Without knowing the legal architecture in place, the Inquiry could not effectively investigate breaches of the law.

It is unrealistic to expect legislation to stay the same forever. It needs to adapt and change as society changes. In some cases, there will be an ‘arms race’ as citizens take action to get around rules and then new rules emerge to stop this. However, there comes a point where frequent changes to rules make it virtually impossible to know what the rules are at any point in time.

Consider two recent poor examples of rapidly changing law. The UK Immigration Rules for ‘normal’ periods (UK) and the International Travel Regulations (Northern Ireland) for ‘emergency periods’. In both cases the rapidity of rule changes makes it extremely hard for citizens to know what the law is. Table 1 sets out the dates when immigration rules changed (between 19 January 2012 and 16 October 2014), and when the travel rules changed (between 8 June 2020 and 27 March 2021).

Table 1 Rapidly changing legislation

The immigration rules changed 27 times in three years. They changed on 9, 10 and 18 December 2013. This makes it extremely difficult for an ordinary user to know what the law is at a particular point in time. Even officials will struggle to keep up to date. The entire thrust of Yeo’s analysis of immigration legislation is that it does not comply with the Rule of Law.¹⁰⁵ Likewise, the Public Prosecution Service in Northern Ireland felt unable to bring a prosecution for breach of coronavirus rules because they had changed so frequently. Their statement on their decision not to prosecute said ‘the law as it applied to the Storey funeral was changed significantly on the evening before the funeral and further amended two days later’.¹⁰⁶ The rapidity of the change meant it would have been unfair to prosecute in respect of breach of the changing law.

Those charged with making legislation have a role to play here. Parliamentarians should not simply acquiesce in enacting legislation that is so volatile that it makes a mockery out of any attempt by the citizen (or immigrant) to have meaningful access to the legislation. There comes a point when incontinent legislative changes undermine the Rule of Law and legislators have to at the very least counsel that there needs to be more stability. Lord Justice Beatson expressed concern over the Immigration Rules on this very point. He stated

The detail, the number of documents that have to be consulted, the number of changes in rules and policy guidance, and the difficulty advisers face in ascertaining which previous version of the rule or guidance applies and obtaining it are real obstacles to achieving predictable consistency and restoring public trust in the system, particularly in an area of law that lay people and people whose first language is not English need to understand.¹⁰⁷

2.4.6.Aspect 4: Law in a Particular Geographical Area

The legal jurisdictions in the UK¹⁰⁸ are:

•England and Wales

•Scotland

•Northern Ireland

Even though England and Wales are a single jurisdiction, it is possible to have different laws for England and Wales.¹⁰⁹ So a more accurate description of the different laws applying in the different parts of the UK is:

•Law of England and Wales as it applies in England and Wales

•Law of England and Wales as it applies in Wales

•Law of England and Wales as it applies in England

•Law of Scotland

•Law of Northern Ireland

The rise of devolution has increased the degree of divergence of legislation between these areas. This is not simply in terms of where they apply, but also the way in which they apply. As Watkin points out, the devolution statutes are ‘inevitably silent as to how the style of the legislation made at Holyrood and Stormont or in Cardiff Bay is to relate to that enacted at Westminster and as to how the approach to the drafting of legislation for the devolved legislatures is to develop’.¹¹⁰ This means that we have three jurisdictions and five geographical areas in which law can apply. This does not consider experimental/trial legislation. It also does not consider the expired jurisdictions such as legislation that applied throughout the island of Ireland before Northern Ireland was established.¹¹¹

When the procedures for enacting legislation are considered, this becomes even more confusing. The Westminster Parliament can pass legislation for all, or any part of the UK. The Northern Ireland Assembly can only pass legislation for Northern Ireland.¹¹² As part of this, it can amend any UK legislation in so far as it applies in Northern Ireland. Senedd Cyrmu can pass legislation for Wales.¹¹³ As part of this, it can amend any UK legislation in so far as it applies in Wales. It can also amend the law of England and Wales as it applies in England, if that is incidental to amending the law of England and Wales as it applies in Wales.¹¹⁴ The Scottish Parliament can only pass legislation for Scotland.¹¹⁵ As part of this, it can amend any UK legislation in so far as it applies in Scotland.

There is a further level of complication. The Secretary of State (for Wales) can make secondary legislation, in consequence of legislation made by Senedd Cymru, which amends the law of England, Wales and Northern Ireland (but not Scotland).¹¹⁶ The monarch, by order in council (in effect the Secretary of State for Northern Ireland), can make secondary legislation, in consequence of legislation made by the Northern Ireland Assembly, which amends the law of any part of the UK.¹¹⁷ The monarch, by order in council or a Minister of the Crown, can make secondary legislation in consequence of legislation passed by the Scottish Parliament (presumably amending the law of any part of the UK although this isn’t explicit).¹¹⁸

The net result of this is that it is hard to ascertain the geographical extent of UK laws. They may apply (or extend) to one or more parts of the UK. Sometimes they will, on their face, amend the law of the UK, but the effect is only to amend the law of the UK in so far as it applies to one particular part of the UK. The law may be repealed in one jurisdiction, but not another. For example, the Autism Act (NI) 2011 amends the Disability Discrimination Act 1995 by inserting new provisions relating to autism. But even though the 1995 Act is UK-wide, the provisions relating to autism only apply to Northern Ireland. This is further complicated by the fact that the 1995 Act has now been repealed, in relation to England, Wales and Scotland, leading to the following footnotes in the text of the online version:

E1 In its application to Northern Ireland, this Schedule has effect subject to the modifications set out in Sch. 8; see s. 70(6)

E4 Following the repeal of this Act for E.W.S. by the Equality Act 2010, this provision now extends to Northern Ireland only

F5 Act repealed (E.W.S.) (1.10.2010 (except for the repeal of ss. 49A-49D) and 5.4.2011 in so far as not already in force) by Equality Act 2010 (c. 15), 216(3), Sch. 27 Pt. 1 (as substituted (1.10.2010) by S.I. 2010/2279, Sch. 2) (with ss. 6(4), 205, and with amendments and savings in the said S.I. 2010/2279, art. 16); S.I. 2010/2317, art. 2(15)(f) (with arts. 4-25, Schs. 1-16); S.I. 2011/1066, art. 2(h)

F6 Sch. 1 para. 4(1)(i)(j) inserted (N.I.) (9.8.2011) by Autism Act (Northern Ireland) 2011 (c. 27 (N.I.)), ss. 1(2), 5

For someone who has spent years reading legislation, this makes sense. For the non-specialist reader, it will look like a foreign language.

What does this all mean? It means that there is no universal text of legislation. Unlike the speed of light, which is constant to all observers, legislation varies depending upon where the observer is standing. This complication is exacerbated by a particular focus on Westminster legislation that leads to consideration of devolved legislation being marginalised.¹¹⁹

A good system of legislation needs to specify to users what the law is in a particular area. As well as the temporal division in aspects 1 to 3, there is also the spatial division in aspect 4. When it comes to the form of legislation, there are two main tools used to demarcate the geographical boundaries of legislation: the extent section and the territorial limits of devolved legislatures.

The extent section is the section in a (Westminster) Act setting out to what parts of the UK the Act applies.

This Act extends to England and Wales, Scotland and Northern Ireland.¹²⁰

A more complicated one is the Botulinum Toxin and Cosmetic Fillers (Children) Act 2021.

(1)Subject to subsection (2), this Act extends to England and Wales only.

(2)Sections 4(3) and 5 extend to England and Wales, Scotland and Northern Ireland.¹²¹

It can then get increasingly complex. The Modern Slavery Act 2015 has the following provision

(1)Parts 1, 2 and 5 (except for section 53) and section 54A, and Schedule 4A, in Part 7 extend to England and Wales only, subject to subsection (4).

(2)Part 3 extends as follows—

(a)section 35 extends to England and Wales only;

(b)section 36 extends to Scotland only;

(c)section 37 extends to Northern Ireland only;

(d)sections 38 and 39, and Schedule 2, extend to England and Wales, Scotland and Northern Ireland.

(3)Part 4, section 53 in Part 5 and Parts 6 and 7 (except for section 54A and Schedule 4A) extend to England and Wales, Scotland and Northern Ireland, subject to subsections (4) and (5).

(4)An amendment or repeal made by this Act has the same extent as the provision amended or repealed.

(5)But the amendments and repeals made by the following provisions of Schedule 5 extend to England and Wales only—

(a)paragraph 2,

(b)paragraph 5(2),

(c)paragraph 6,

(d)paragraph 8,

(e)paragraph 21.¹²²

Each devolved jurisdiction has the territorial limit on its legislative competence set out above. For example, the Northern Ireland Assembly can only enact legislation for Northern Ireland. It has no competence to enact legislation for any other part of the UK. This territorial limitation is contained in the legislation establishing the devolved legislatures, but it is not stated in the legislation that they enact. So, legislation of devolved legislatures will include ‘Northern Ireland’, ‘Scotland’ or ‘Wales’ in the title, but will not contain an express statement anywhere ‘this Act only extends to Northern Ireland/Scotland/Wales’. It could therefore be assumed that if the geographical name appears in the title, the reader will know that the legislation only extends to that area. Unfortunately, this is not so. An Act may be UK-wide even if a particular name appears in the title. For example, the Corporation Tax (Northern Ireland) Act 2015 forms part of the law of England and Wales, Scotland and Northern Ireland (it was passed by the Westminster Parliament). But the Air Passenger Duty (Setting of Rate) Act (Northern Ireland) 2012 only forms part of the law of Northern Ireland (it was passed by the Northern Ireland Assembly).

The second legislative drafting solution is more radical. Each Act of a devolved jurisdiction should include an extent clause at the end setting out the jurisdiction to which it applies. This is most definitely not the practice in any of the devolved jurisdictions, no legislation passed by them contains an extent clause. This is for the simple reason that it is unnecessary as, constitutionally, their laws can only extend to their own jurisdiction. The only exception I have seen is in a Private Members’ Bill currently before the Northern Ireland Assembly that includes an extent clause stating that it extends to Northern Ireland.¹²³ An extent clause is constitutionally unnecessary. However, it would greatly assist the reader of legislation who is unfamiliar with all the complications around geographical application of laws set out in this section.

2.4.7.Aspect 5: Legislation in Context

A legislative system needs to be able to cite legal provisions in context. If another piece of information (legislative or otherwise) would help the reader to understand a law, it would be helpful if the reader were made aware of that information. The Interim Woolf Report expressed this aspect as follows:

Eventually, fully integrated legal information systems will combine legislation, case law and specialist commentary, all structured in such a way that users will be able to find their way more easily through these materials.¹²⁴

The form of legislation, in the broadest sense of how legislation is presented, can address this ambition in several ways. First, a statute could contain internal (hypertext) links to related parts of the statute. For instance, if section 2 refers to section 50, as well as a textual reference there could be a hypertext link. The National Archives and the Office of the Parliamentary Counsel carried out research in 2012 /2013 on how users interacted with The National Archives website.¹²⁵ One finding was that many users are not familiar with the structure of legislation, for example division into sections and schedules. Internal links help overcome lack of familiarity with legislative structure. An easy way to help readers unfamiliar with the relationship between sections and schedules is to have a hypertext link to the schedule in the section.

Second, the statute could contain hypertext links to related external statutory provisions. The most obvious one would be if Act X refers to Act Y, that it also includes a clickable link to Act Y. Less obvious, but potentially even more helpful, would be a link to secondary legislation made under the power of primary legislation. A factory manager reading the Climate Change (Emissions Reduction Targets) (Scotland) Act 2019 will be less interested in knowing that there is a power to make regulations changing emission targets, and more interested in clicking through to the actual text of those regulations.

External legislative links are easy to imagine when there is a direct textual reference to the provisions of another Act. They are much more difficult if there is no textual reference to the other statute, but that other statute is still connected. Take the example of section 33 of the Justice Act (Northern Ireland) 2016. That section provides for the establishment of the office of Prison Ombudsman for Northern Ireland. The reader of section 33 is unlikely to know that, under section 16 of the Interpretation Act (NI) 1954, members of a body can be appointed in advance of an enactment coming into force. The way the legislation is presented makes it nearly impossible for the casual reader to know of these interconnections. Creating this type of external link requires a certain amount of subjective judgement and is less easy to satisfy by way of algorithms.

Third, the legislation could have external links to non-legislative information that is relevant to the statute in question. Explanatory Notes are official documents, created by the government and designed to assist the reader in understanding legislation. However, technically, they do not form part of the legislation. Jenkins, former First Parliamentary Counsel in the UK, recommended that ‘to make [Explanatory Notes] as widely accessible as possible, they would be made available on the internet alongside the Bill and subsequently the Act’.¹²⁶ This is current practice for UK legislation, where the official website contains both the legislation and the explanatory notes. Another type of official document is an official Impact Assessment prepared by the government to assist in the consideration of the merits of a piece of legislation. These are also now included on the official legislation website. However, the website does not as a matter of course include links to the reports of the Regulatory Policy Committee that assesses the quality of these Impact Assessments. This is an additional document would assist readers in determining the meaning, merits and purposes of legislation.

In addition to the non-legislative information suggested by Woolf above, there could also be links to Parliamentary debates, House of Commons Library Briefing Papers on Bills, etc. Irresberger states that if legislation allows for ‘acceptable means of compliance’, then those means must also be included or linked in some way to the legislation, otherwise accessibility is incomplete.¹²⁷ Watkin lauded the inclusion of the Convention on the Rights of the Child within a Welsh statute as ‘a reader of the Measure would be able to discover what had been incorporated without having to seek out the text of the Convention and Protocols from another source’.¹²⁸ Occasionally, during a debate on a Bill, a minister won’t be able to answer a point raised by a legislator and undertakes to answer the point in a letter at a later date. Greenberg points out that although the promise to write will be contained in the Hansard report, the actual letter won’t. He argues that this letter should also be linked, in some way, to the proceedings in Parliament.¹²⁹ There is the potential for different levels of access to this information, that is, some items restricted for internal government use.

As well as external links to information that provides helpful background on legislation, or that may act as an aid to understanding it, there is also external information that is critical to the operation of the legislation. These could be things like official forms, Departmental guidance, codes of practice that must be complied with, documents or maps referred to in the legislation. If these things are not directly contained in the legislation, they must be referenced in the legislation, along with directions on how to access them. The Joint Committee on Statutory Instruments have expressly made this point, stating that

Information on the availability of documents referred to in legislation should be included in the instrument itself (either in the footnotes or the Explanatory Note) rather than in the Explanatory Memorandum.¹³⁰

It is perhaps when we come to consider criminal law that the need for this type of external information is most pressing. A criminal statute will tell the reader what the definition of an offence is, but it does not set out when a person will be prosecuted for that offence. That decision is a function of the Crown Prosecution Service, or the prosecuting authorities in the various devolved jurisdictions in the UK. Those prosecutors will often have key information that affects how offences will be prosecuted in practice. One of the most contentious of these types of decisions is the decision to prosecute for the offence of assisted suicide. Section 2 of the Suicide Act 1961 makes it an offence to encourage or assist in the suicide of another. It is perhaps too much to expect the Act to set out the details of when a prosecution will be launched. The factors to be considered have however been set out by the Crown Prosecution Service.¹³¹ These include sensible factors such as the age of the victim, whether they had mental capacity etc. The difficulty is not the existence of these guidelines, but the fact that there is no reference to them in the statute law database entry for the Suicide Act 1961. This is useful contextual information and ideally should be visible and accessible alongside the legislation itself.

There are two methods for generating these links. The traditional way is the top-down approach where civil servants think through what they consider relevant to an Act and then create the links. The more innovative (and riskier) way is to use some Big Data concepts and allow links to be influenced by the choices previously made by users. For example, if people who look at the Companies Act 2006 also usually look at the Insolvency Act 1986, then the system could automatically suggest that a reader of one also may find the other useful.

This desirability of making contextual information available is increasingly important with the rise of quasi-legislation.¹³² Quasi-legislation is difficult to define, but it is essentially something that can have a legal effect without being itself legislation. Codes of practice, administrative rules, official guidance, memorandums of understanding, these are all examples of quasi legislation. Sometimes these may have statutory authority or be referred to in legislation. They can have a critical effect. For example, the codes of practice referred to in the Human Transplantation (Wales) Act 2013 set out guidance on how to handle family disputes in cases of organ donation after death. Accessibility of legislation extends to access to this type of relevant material also.

The court has recognised the importance of additional contextually relevant material. It stated in R (Salih) v Secretary of State for the Home Department [2003] EWHC 2273 (Admin) that

it is in general inconsistent with the constitutional imperative that statute law be made known for the government to withhold information about its policy relating to the exercise of a power conferred by statute.

How far this extends is a question of degree. Official guidance drawn up by government in accordance with a statute does seem like precisely the kind of information that should be accessible alongside the legislation. However, what about guidance drawn up by private bodies that is referred to in the legislation? This issue arose in the Dutch Copyright Act. The Dutch Supreme Court ruled that this guidance did not constitute ‘legislation’ and was therefore not subject to the copyright provisions that ordinarily applied to legislation.¹³³ Although this decision may be correct in the narrow sense that guidance produced by a private body is not legislation, it is submitted that it is incorrect if it leads to such guidance being inaccessible. If it is directly relevant to the law, if it colours the way the law is applied, then it ought to be accessible alongside the law.

Another example is the memorandum of understanding on the role of regulators of financial services industry. Section 3E of the Financial Services and Markets Act 2000 obliges regulators to prepare a memorandum of understanding setting out their role. This must be given to the Treasury and laid before Parliament. It is clearly an important document, referred to in primary legislation, but it is not available alongside that legislation. This goes against the Rule of Law requirement of accessibility.

2.4.8.Aspect 6: Findability

•The degree to which a system or environment supports navigation and retrieval.¹³⁴

Lavery used the word in a legal context as long ago as 1943, considering a lawyer with a legal question, ‘somewhere in that vast storehouse of the law is the answer to his problem, but he cannot find it’.¹³⁵ For Lavery the problem was the volume of law piling up, making it harder to find the desired nugget of legal information.

This requirement is the hardest of the accessibility principles to satisfy – the user ought to be able to retrieve the legislative answer to their question from a legislative database, even if the user does not know which piece of legislation contains the answer. If we start from the proposition that the user knows which section in which Act contains the answer to the question, it is relatively easy to design a legislation database that can quickly point to that section in that Act. However, it will be quite rare that the user knows the precise chapter and verse of the provision they are looking for. Much more common will be that they have an open question to which they seek an answer. The electronic system needs to facilitate easy access to help find the answer. For example, if the police arrest a child in Northern Ireland and decide not to release the child on bail, where must the child be placed in detention? The answer is contained in Article 39(8) of the Police and Criminal Evidence (NI) Order 1989. For the system to be effective, it must be able to display an accurate and up to date version of Article 39(8). But it must also be able to indicate to the user that the answer is contained within Article 39(8).

Findability flows into the form of legislation principle of navigability. Findability means the ability to find a particular piece of information whereas navigability relates to the ease of moving around the statute book. The basic concept for the legislator to bear in mind is to make it as easy as possible for the reader to find the information they are looking for.

2.4.9.Availability of Retained EU Law in the UK

Before Brexit (the United Kingdom leaving the European Union) EU legislation ran in a separate stream from domestic UK law in terms of accessibility. If an EU legislative measure wasn’t transposed (directly written into domestic law), then it was only available on the EUR-lex servers hosted by the EU. In terms of availability, this was not ideal for UK citizens. There was an entire corpus of law, directly applicable in the UK, which nevertheless did not show up on the UK’s domestic legislation website.

Paradoxically, post-Brexit there is now greater availability of EU law that is applicable in the UK.¹³⁶ The newly established category of ‘retained EU law’ is EU law that applied in the UK just before Brexit and that continues to apply until it is repealed.¹³⁷ The National Archives now keep a copy of retained EU law on its own website, which can be accessed alongside its records of domestic legislation. According to them

Legislation.gov.uk is where you can find legislation originating from the EU as it now applies in the UK. This legislation will be kept up to date with any amendments made by the UK Parliaments, Governments and Assemblies.¹³⁸

In terms of availability of legislation, this is very much to be welcomed.

However, there is now a caveat to this. Clause 56 of the Nationalities and Borders Bill reads (at present) as

Section 4 of the European Union (Withdrawal) Act 2018 (saving for rights etc under section 2(1) of the European Communities Act 1972) ceases to apply to rights, powers, liabilities, obligations, restrictions, remedies and procedures derived from the Trafficking Directive so far as their continued existence would otherwise be incompatible with provision made by or under this Act.

This is quite a mouthful, but what it means is that if there is something in the Trafficking Directive that is incompatible with a provision in the Bill, or a provision in secondary legislation made under the Bill, then that part of the Trafficking Directive no longer has effect. This is very unsatisfactory in terms of availability of legislation. The Trafficking Directive is part of retained EU law, and therefore part of the legislation applying in the UK. If the new policy is to repeal or modify part of it, then that is a normal part of the ebb and flow of the statute book, and it should be repealed or amended in the normal way. But the provision in the Bill doesn’t make a formal repeal or amendment to the Directive. There is no change to the text of the Directive, but nevertheless that text is no longer authoritative. If there is something in the Directive that is incompatible with the Bill, then that part of the Directive is no longer law. How is the citizen to make sense of this? If they look on the official government website, they will see the Directive there with no amendments. How can anyone be expected to know if a provision of it is incompatible with the Bill, if not even the government can point out any incompatibilities?¹³⁹

The principles set out by Locke over 300 years ago still apply – laws must be promulgated. If the government or Parliament wants to get rid of bits of an old law, it is fully entitled to do so. But at a very minimum it must point out the bits of the old law that it is getting rid of.

2.4.10.Availability of EU Law Applicable in Northern Ireland by Virtue of the Northern Ireland Protocol

This is another Brexit related point. Although the UK (which includes Northern Ireland) has left the EU, special provision has been made for certain EU legislation to continue to apply in Northern Ireland. This has been done under the terms of the Withdrawal Agreement between the UK and the EU, and more specifically the Northern Ireland Protocol to that Agreement.¹⁴⁰ Annex 2 to that Protocol sets out the EU legislation that continues to apply in Northern Ireland. This is different in character from retained EU law that applies by default until it is repealed. This is EU law that has been specifically agreed between the parties to continue to apply in Northern Ireland and that cannot be unilaterally repealed by the UK. There is provision to allow additional EU legislation to be added to Annex 2 by agreement between the parties. However, there is no easy way for citizens or even legislators to know what EU law applies in Northern Ireland by virtue of the Northern Ireland Protocol. This can cause serious problems. For example, there was a proposal for a Private Member’s Bill in the Northern Ireland Assembly on single use plastics, which partly sought to align Northern Ireland with EU legislation on single use plastics.¹⁴¹ However, under the agreed procedure for adding EU legislation into Annex 2, the Joint Committee added the Single Use Plastics Directive EU2019/904 to the list of EU legislation applying in Northern Ireland.¹⁴² Without any easy way to find out what is in Annex 2 a Bill was to be proposed to enact what was already obliged to be law in Northern Ireland.

The lack of availability of EU legislation applying in Northern Ireland undermines the Rule of Law. Provision ought to be made for these laws to be as easily available as other legislation applying in Northern Ireland.

2.4.11.Correction Slips

It is inevitable that legislation will sometimes contain minor mistakes. The obvious way to correct these is to make fresh legislation that rectifies those mistakes. But sometimes if the mistake is on a technical point, such as a typographical error, or wrong cross reference, it may be possible to use the correction slip procedure. This is a UK procedure, in respect of secondary legislation only, whereby an official document is issued that ‘corrects’ the error. The decision on whether to use a correction slip is made by the Statutory Instrument Registrar at the National Archives, along with the Department responsible for the statutory instrument containing the error.¹⁴³

The use of correction slips has clear implications for democratic legitimacy – officials should not be changing the law by way of an administrative procedure, only Parliament can authorise a change in the law. The Joint Committee on Statutory Instruments have identified the following circumstances where a correction slip is justified:

•Errors are small scale

•Errors are obvious

•Text and location of error are obvious

•Corrections are small scale.¹⁴⁴

The key point is that the correction should not amount to a substantive change in legal rights or obligations, it ought to be limited to purely superficial or typological corrections.

In terms of availability under the Rule of Law, the following are required. First, if a correction slip is issued, then the corrected version of the statutory instrument ought to be sent to everyone who purchased it originally. This requirement is contained in the Government’s manual on statutory instruments.¹⁴⁵ Second, the online version of the statutory instrument ought to be corrected, along with a note or chronology that shows what the original version was, what the correction was, and how the correction was effected. Third, if there is any chance at all that someone could be prejudiced by this change, or that it could amount to a substantive change of the law, then correction slips must not be used.

Mistakes in legislation are very embarrassing. But correction slips must not be used to spare the blushes of Departments or Government lawyers.

2.5.1.Short Titles

The short title of an Act contributes to the navigability of the statute book by

•indicating the content of an Act, and

•assisting in the logical ordering of the statute book.

A title most effectively conveys meaning by being short and descriptive. It assists in the arrangement of the statute book by tessellation – fitting in harmoniously with what is already there. Duprat calls it ‘a tool indicating the main issue under regulation’.¹⁴⁹

The descriptive aspect of titling legislation is under threat. The trend is arguably a movement from pure description, to rhetoric as a way of persuading legislators to vote for a Bill, to bare-faced political sloganeering. The best short title (in terms of navigability) is the one that is politically neutral and best encapsulates the content of the Act in as pithy a way as possible. Orr charts how in Australia titles have become more and more tendentious, observing ‘a loss of descriptiveness in titling in favour of an image-conscious and rather cynical form of political rhetoric’.¹⁵⁰ The classic US example is the USA PATRIOT ACT, which is an acronym for Uniting and Strengthening America by Providing Appropriate Tools Required to Intercept and Obstruct Terrorism Act. It is a political argument masquerading as a title. Another argument against propagandist titles is that they clamp down on proper political debate – how can a true patriot vote against the USA PATRIOT Act? There are similar moves in the UK with the emotive use of the legally meaningless term ‘veteran’ in the Overseas Operations (Military Personnel and Veterans) Act 2021.¹⁵¹ Strause sees this as a problem with naming conventions in American statutes, that they bring ‘the law, and perhaps the lawmakers, into contempt by seeming too narrowly partisan’.¹⁵²

This affliction of propagandist titles is global. In Australia there was the Roads to Recovery Act 2000 that is about funding expenditure on roads. The title was chosen to align with a political slogan and is criticised by Orr for being a ‘corny pun’. In Canada there is the Standing up for Victims of White Collar Crime Act that amends the criminal law in respect of sentencing for fraud. The Roads (Funding) Act and the Fraud (Sentencing) Act may have been more appropriate. In Uganda, legislators did not like the title of the Marriage and Divorce Bill for religious and moral reasons. A leading daily paper called for it to be renamed the Marriage and Family Relations Bill.¹⁵³ Thankfully the drafters were successful in arguing that if it was about divorce, then divorce should be in the title.

But is this criticism too harsh? In the real world, legislation must pass (as razors must sell).¹⁵⁴ In order to pass, a Bill must be politically persuasive. After all, legislation also represents political vision turned into law. The Criminal Justice Act (No 2) 2003 may not inspire political interest, but the Hang Child Killers Act 2003 will. Jones has carried out some fascinating empirical research into this, and his unsurprising findings show that people are more likely to be in favour of a Bill if it has an ‘evocative’ title rather than a ‘descriptive/technical’ title.¹⁵⁵ But protecting the Rule of Law is more important than a transient political slogan.¹⁵⁶ Legislators should seek to resist government ‘spin’ in the wording of legislation.¹⁵⁷ A slogan may help an Act to pass, but this is only one instance in the life-cycle of an Act. For years to come, readers will have to struggle to navigate badly titled law.

The UK is doing quite well in this regard. Although there is always a political desire towards catchy names, by and large short titles here are relatively straightforward. In fact, in Scotland, the guidance is explicit, stating that the text of a Bill ‘including both the short and long titles – should be in neutral terms and should not contain material intended to promote or justify the policy behind the Bill, or to explain its effect’.¹⁵⁸ According to Erskine May ‘the short title must describe the content of the bill in a straightforwardly factual manner. An argumentative title or a slogan is not permitted’.¹⁵⁹

Being neutral doesn’t necessarily mean being boring. Contrast the approaches of Hong Kong and Australia to the same problem of junk emails. In Hong Kong the legislation is the Unsolicited Electronic Messages Ordinance, whereas in Australia it is the Spam Act. Both are good and accurate titles, but the Hong Kong one is more legalistic and perhaps stuffy. The Australian one is much more quotidian and instantly tells all readers what it is about. A good example of a pithy and descriptive title is the regulations granting consumers vouchers to spend in shops as part of a covid recovery plan in Northern Ireland. The regulations are entitled the High Street (Coronavirus, Financial Assistance) Scheme Regulations (Northern Ireland) 2021.

The tessellation argument is not one that is often made in the context of legislation. Tessellation means that the individual pieces of a mosaic fit together without gap or overlap. In the context of the statute book, it means that the statute book fits together as a whole, that it is coherent. To assist in the logical ordering of the statute book, titles must follow an agreed format. They must make it easy to fit a particular statute within the general scheme of the statute book. This will help any indexing or referencing system as well as making it easy for the user to see at a glance if a particular statute falls within a group of statutes. So for example, in Northern Ireland there are the:

•Housing (NI) Order 1976

•Housing (NI) Order 1978

•Housing (NI) Order 1981

•Housing (NI) Order 1986

•Housing (NI) Order 1988

•Housing (NI) Order 1992

•Housing (NI) Order 2003

•Housing (Amendment) (NI) Order 2006.

The reader can easily identify these in an alphabetical index of statutes, as O’Brien says, ‘one of the main purposes of the short title is to locate the legislation in the statute book’.¹⁶⁰ The reader also knows that any new legislation called the Housing Order will deal with the same subject matter as these existing orders. But if a new statute was called the Buildings for Residential Use Act, this would immediately make the statute book must less navigable. In this context, the keyword is consistency. Legislation dealing generally with the same subject matter must have generally the same name.

Personalised names in legislative titles cause problems with navigability. For example, in the US, there is the Lily Ledbetter Fair Pay Act, named after a famous litigant in an employment case. Legislation can also be named after victims, for example Megan’s Law deals with making public information about sex offenders, and was named after a victim of a sex offender. It can be named after politicians closely connected with it, for example the Landlord and Tenant (Amendment) Act (Ireland) 1860 is more commonly called Deasy’s Act after the Irish Attorney General of that time. Naming statutes for individuals is sometimes done posthumously, as a mark of respect, for example the Ronald W Reagan National Defence Authorization Act for Fiscal Year 2005. Jim Allister MLA in discussing a Bill before the Northern Ireland Assembly stated ‘whereas the Bill must officially be called the Civil Service (Special Advisers) Bill, I trust that, in common language, it will, if passed, become known as “Ann’s law”’.¹⁶¹ This was in tribute to a particular victims’ campaigner. Jones has written about the different approaches to what he calls ‘humanised’ titles in the USA and the UK:¹⁶² he concludes that the UK establishment zealously guard the ‘neutral names’ approach. This non-personalised approach to short titles is one that is in accordance with the Rule of Law.

An additional barrier to navigable statutes is the overly broad and bland Act title. We know the Crime and Courts Act 2013 is about criminal law, but we don’t know much more than this. We know the Northern Ireland (Miscellaneous Provisions) Act 2013 is about Northern Ireland, but we don’t have an idea about the details. There isn’t necessarily very much the legislator can do about this. Sometimes an Act will deal with a broad subject and make lots of different amendments to different aspects of that subject. If we were to list everything that the Act dealt with, it would be far too long a title. For example, a more descriptive short title for the Northern Ireland Act would be the Northern Ireland (Political donations, Assembly membership, Justice Minister, Excepted and Reserved Matters, Electoral Matters etc.) Act 2013 – but this is obviously far too long to be of any real use. But sometimes we can be more specific, for example most legislation amending criminal justice law is often called the Criminal Justice Act, but we also have the Criminal Justice (Evidence) (NI) Order 2004 – as that Order only deals with criminal evidence, the title can be more specific. There is no easy prescription here, the legislator has to balance three things: keeping the short title short, indicating its content and promoting indexing of the statute book.

2.5.2.Omnibus Bills

The section above discussed the dangers of bills dealing with a broad subject and having a bland title. Much more problematic are omnibus bills – ‘a massive bulk of unrelated rules modifying numerous existing statutes’.¹⁶³ In the UK, these are sometimes termed ‘Christmas tree bills’ as there is something for everyone in them.¹⁶⁴ Popelier describes this concept as ‘mosaic legislation’ and they can also be referred to as programme laws.

An omnibus Bill makes navigation of the statute book very difficult. The reader has no idea of what an omnibus Bill may contain. Its title will not convey meaning and will not assist in the logical ordering of the statute book. An omnibus Bill is enacted for the benefit of legislators, not of citizens. In Belgium they are called ‘Programme Laws’ or ‘Law holding diverse provisions’ – how on earth is the user to divine content from such a broad title? Popelier directly states the disadvantages of mosaic law as the difficulty of citizens accessing the law.¹⁶⁵ In Israel they are called ‘Arrangement Laws’ and the Israeli Supreme Court heard objections about these been railroaded through the Knesset.¹⁶⁶ Although that case (as Bar-Siman-Tov points out)¹⁶⁷ was about procedural rationality in law making, it also illustrates the difficulty with navigability of the statute book. The Supreme Court in Israel has ruled ‘Arrangements Law’ as unconstitutional, although this is more for reasons of legislative due process than for navigability of legislation.¹⁶⁸

An example of omnibus legislation in the UK is the Enterprise and Regulatory Reform Act 2013. The long title shows the diversity of subjects within it:

An Act to make provision about the UK Green Investment Bank; to make provision about employment law; to establish and make provision about the Competition and Markets Authority and to abolish the Competition Commission and the Office of Fair Trading; to amend the Competition Act 1998 and the Enterprise Act 2002; to make provision for the reduction of legislative burdens; to make provision about copyright and rights in performances; to make provision about payments to company directors; to make provision about redress schemes relating to lettings agency work and property management work; to make provision about the supply of customer data; to make provision for the protection of essential supplies in cases of insolvency; to make provision about certain bodies established by Royal Charter; to amend section 9(5) of the Equality Act 2010; and for connected purposes.

2.5.3.Section Headings

Unlike short titles, section headings can be longer and more specific. Depending upon how pithily the heading can be expressed, it can be a ‘content-indicator’ title or a ‘rule-summary’ title. A content-indicator title gives an idea of the content of the provision, whereas a rule-summary title actually summarises the rule itself. For example, when the Civil Service (Special Advisers) Bill was introduced into the Northern Ireland Assembly, clause 2 was entitled ‘special advisers not to have serious criminal conviction’. This accurately reflected the content of the rule set out in clause 2, it was an abbreviated description of the rule itself. However, as the Bill passed through the Assembly the content of clause 2 was changed and there was no longer an absolute prohibition on special advisers having a serious criminal conviction. Therefore, upon enactment, the title of that clause had changed to ‘special advisers: serious criminal convictions’. The new title didn’t summarise the actual rule, instead it gave an indication of the subject matter of the provision. Rule-summary titles are usually more difficult to set out than content-indicator titles as it is hard to condense most rules into one line. However, rule-summary titles make it easier for the user to know the thrust of the legislation. A combination of a well-organised table of contents and descriptive titles means that reading the table of contents can be like reading a condensed narrative of the legislation.

For constitutional and historical reasons, there is a certain ambiguity about the legal status of headings in legislation in the UK. In the past they have been viewed, not as part of the legislation enacted by the legislature, but as an editorial addition inserted by the clerks and administrative staff of the legislature.¹⁶⁹ The actual practice of making legislation shows this to be nonsense – every part of the Bill, including all the headings, is crafted by the legislative drafter working on behalf of the legislator. However, the rule of interpretation is that section headings may be used to assist in the interpretation of the legislation, even though they are not determinative.¹⁷⁰ This is an eminently sensible canon of construction as three or four words in a heading can’t adequately describe the entirety of the section that follows.¹⁷¹ However, this rule does give the legislator an unusual opportunity for creativity and innovation, unburdened by the usual requirements of precision and legal accuracy. The purpose of a navigable heading is to convey meaning, not to be legally precise, therefore the drafter is relatively free to select headings that best convey meaning.

An example will illustrate this. The Northern Ireland Law Commission produced a Bill to amend the law on bail.¹⁷² Clause 30 of the Bill is entitled ‘repeal of street bail’. But ‘street bail’ has no legal meaning, the correct legal term is ‘bail at any place other than a police station’. Street bail is the term used in practice, it is quicker and less cumbersome, but it is legally incorrect. However, as a title it works, the reader with a connection to the criminal justice world instantly knows what this provision is about. The detail of the provision uses the legally correct terminology, as it must, but in crafting the title, the legislator is freer to convey meaning. This practical example provides a partial answer to Heaton’s concern that titles in legislation don’t convey meaning to non-legal users.¹⁷³ The legal lexicon (words like tort, fee simple, actionable) are familiar and relevant to lawyers, but not to non-lawyers. So in section headings it is open to use the demotic word, not the legally precise word.

A contrasting EU legislative provision shows the benefit of this more flexible UK approach. Article 5 of Directive 2002/58/EC is entitled ‘confidentially of these communications’ and paragraph 3 of that Article begins

Member States shall ensure that the use of electronic communications networks to store information or to gain access to information stored in the terminal equipment of a subscriber or user is only allowed on condition that the subscriber or user concerned is provided with clear and comprehensive information in accordance with Directive 95/46/EC, inter alia about the purposes of the processing, and is offered the right to refuse such processing by the data controller.

2.5.4.Structure

An Act is more navigable if it is organised and structured¹⁷⁴ – but what does this actually mean in practice? It means that users can see at a glance the general thrust of the Act and see, again at a glance, which provisions are relevant to them. Voermans distinguished between ‘structure’ (division of legislation, referencing) and ‘superstructure’ (relation to other legislative texts, hierarchy) but I include both these here within my use of the word structure.¹⁷⁵ Voermans went on to say that ‘the structure of an Act can act as a road map for users who want to find the relevant provisions’.¹⁷⁶ Onoge also saw structure as extremely important, writing that ‘careful layout and design of legislation is as important as clear language’.¹⁷⁷

Unless an Act is extremely short, it should have a table of contents (sometimes referred to as arrangement of sections, or contents). Like a short title, a table of contents gives the reader the gist of the Act, it also sets out the scope of the Act. So, although the Crime and Courts Act 2013 is a very broad title, looking at the table of contents reveals it is about the National Crime Agency, administration of justice, self-defence etc.

Structure is helped immeasurably by divisions and sub-divisions of the legislation – as Duprat notes these make ‘the provisions easily seen by the public and to simplify the interpretation of enactments’.¹⁷⁸ Related material should be grouped together and listed under a separate heading. Related headings should also be grouped together and listed under larger headings. The UK taxonomy is of Parts, divided into Chapters, divided into cross headings. It is not necessary to have each of these tiers in a single Act – a short Act may only have cross headings, whereas a larger one may have multiple Parts, Chapters and cross headings. The more topically different provisions are, the higher up the tier the divisions should be. For example, in the Water Act 2014, Part 1 deals with the water industry. This is conceptually very different from provisions on flood insurance, so it is quite right that flood insurance is dealt with in Part 4. Therefore, the reader interested in the new law on flood insurance doesn’t have to wade (pun intended) through the other water provisions in the Act, but can dive straight to Part 4.

Different writers have suggested different ways to structure legislation. For example, Thring suggested stating the law first and then how the law was to be administered.¹⁷⁹ He suggested including the standard provisions first followed by the special or more complicated ones. It is also possible to take a chronological approach, such as setting out in order the steps with which a user must comply in order to satisfy a particular administrative hurdle. For example, in Part 2 of the Companies (Guernsey) Law, 2008, the first cross heading is ‘conditions precedent to incorporation’ and the second cross heading is ‘incorporation of a company’. Bergeron suggests having permanent provisions first and temporary provisions later.¹⁸⁰ This practice is followed nearly all the time in the UK, where transitional provisions or transitory provisions appear at the very end of a statute. Bentham wanted to separate the ‘imperative’ provisions (which would be short and come first) from the ‘expository provisions’ that would be longer, contain more explanatory material and come later.¹⁸¹

Xanthaki proposes an innovative approach to structure, what she calls the layered approach.¹⁸² Under this structure, legislation is drafted in different layers, corresponding to the different users of it. So, Part 1 is for lay people and sets out the basic regulatory message of the law. Part 2 is for non-legal professionals and sets out the detail that they need. Part 3 deals with issues of interpretation, application, amendments, repeals etc. It is for those lawyers who need to know the precise details of every last provision.

As with all these drafting principles, these approaches to structure are only guides and can be over-ridden. For example, in the Telecommunications Act 1984, there is a provision vesting property belonging to British Telecommunications in a successor company. Normally this sort of transitional provision is relatively unimportant and is included as a schedule at the end of the statute. However, as it was considered of crucial economic and political importance, it appeared in the body of the statute, not at the very end. The key point for the legislator is to choose a structure and an ordering of provisions that makes it as easy as possible for the reader to find their way through the statute.

The other difficulty with maintaining a sound and coherent structure is amendments, both amendments to a Bill as it passes through the legislature,¹⁸³ and amendments to an Act after it is made. These can often wreck a carefully balanced piece of legislation by forcing new provisions into a structure that was not designed to accept them. The Aristotelian idea of the perfectly structured entity does not work when faced with the grubby realities of making law in a political legislature. One thing that those charged with preparing the law can do however, is to strongly counsel that the Bill as introduced into Parliament is the finished product, at least from the government’s perspective. Introducing a Bill that even the government knows to be incomplete is a good predictor of a Bill that will be structurally unsound, with subsequent amendments stretching it out of shape. The more pre-legislative scrutiny and consultation, the less likely that the introduced Bill’s structure will be deformed by amendments as it passes through the legislature. The other approach is to leave figurative ‘space’ for future amendments that are expected to come.

2.5.5.Parallel Structures

If an Act regulates two or more similar activities or entities in similar ways, it is made more navigable if it adopts parallel structures for that regulation. A parallel structure means that the same approach and headings (so far as possible) are used for each entity. This makes it easier for the user to follow, and they can also quickly see where the regulation differs. This is the approach taken in the Companies (Guernsey) Law, 2008. Listed in Table 2 are cross headings and section headings for Part XXVII (Protected Cell Companies) and Part XXVIII (Incorporated Cell Companies). This approach makes it easier for readers to compare and contrast these two different types of company than if the structure adopted had been wildly different.

Table 2 Protected Cell Companies v Incorporated Cell Companies

In the same vein, provisions establishing a body in legislation generally take a similar approach: setting out how people are appointed to the body, their salary, staff, reporting arrangements, budgeting, pensions, procedure, status, termination of membership etc. If a new body is set up, the reader would expect to see similar provisions in a similar order. The Office of the Parliamentary Counsel is explicit about this, setting out common provisions for the establishment of a statutory corporation.¹⁸⁴ This is not an argument for blindly following precedent – it is an argument for assisting readers by using a structure with which they will be familiar.

2.5.6.Sign Posting

A statute is a single document and the full meaning of that document is only apparent if the full document is read. In the same way that one line from a recipe doesn’t make the meal, one section from an Act doesn’t explain the Act in its entirety. In a short Act, the connection between sections is clear and obvious, and there is no need for any aids to make it more navigable. But in a longer Act, there may be a gap between two related sections – a section may make more sense if read in conjunction with another provision that may be in a different part of the Act. Take a simple example: section 10 sets out an offence and section 11 sets out the defence. It is easy for the reader to navigate from one to the other. But what if sections 10 to 15 set out the offences, and sections 30 to 32 set out the defences? Suddenly we have reduced navigability, and we have made it more difficult for readers to link together important information.

There are a number of form of legislation solutions to this. First, there is the long established adage in legislative drafting of keeping like material with like. Thring, for example, recommends keeping provisions declaring the law together, and keeping them separate from provisions relating to the administration of the law.¹⁸⁵ The difficulty with this is that a provision may be ‘like’ one provision in one way, but ‘like’ a different provision in another way. For example, we could put all offences together because they are like each other, or we could put an offence relating to a particular subject next to the other material on that subject. This is a zero-sum game – if we put two like sections together, we automatically displace another like section and push it further away. This may be an unavoidable consequence of the interconnectedness of legislation. The rules of interpretation tell us that we can’t read provisions in isolation, we need to read the entire Act.

The next solution is to try to place all relevant material together in a single section. And if material is relevant to more than one section, it can be repeated in each of those sections. There are two problems with this. First, we will dramatically extend the length of each individual provision. So, section 1 will be 10 pages long as it includes the rule, the exceptions, the defences and the definitions. A 10-page section is far too long to be easily comprehensible (see further chapter five). Second, such an approach can be repetitive and increase the chance of making copy errors. For example, if the same definitions apply to the first five rules, it would seem silly to repeat them in each of the first five rules. Even worse, we increase the risk of future errors. Say that next year we want to amend one of these definitions, we will have to amend it five times in five different places, and hope that we make exactly the same amendment in each place.

The third solution to this issue is signposting – pointing the reader to where the other relevant provisions are within an Act. So, section 1 could be set out as follows.

1.(1)This is the rule.

(2)But this is subject to –

(a)the exceptions in section 5, and

(b)the defences in section 10.

(3)The elements of the rule set out below are defined in section 50 –

(a)element A,

(b)element B.

This puts the user on notice that there is a context to the rule, and that the rule cannot be read in isolation. The reader is signposted to further relevant provisions. This is especially helpful for non-lawyers, particularly those not familiar with the standard structure of legislation, including its use of interpretation sections for definitions or schedules.¹⁸⁶ Signposting is a great aid to navigability.

However, there are at least two problems with signposting. The first is knowing where to stop. It is conceivable that every provision in an Act could have a bearing on every other provision in an Act – does that mean that each individual section must contain a signpost to every other section? Extending the example above, does section 1 also need to state:

The second problem concerns the nature of legislation. If the function of legislation is to make law, everything extraneous to that is redundant and should be excised from legislation. A signpost is not law, it only clutters up the law. As Bowman put it, ‘excess matter in Bills, as in people, tends to go septic’.¹⁸⁷

The answer to these two problems rests in the judgement of the legislator. Balance these two problems against the need to enhance navigability. So, do not signpost everything, just those provisions that are most directly relevant (or most likely to be otherwise missed). This will sometimes mean shoehorning in a provision that is not strictly necessary from a legal perspective, just to make the reader aware of it. For a good example of legislation with signposts, see the Public Services Ombudsman Act (Northern Ireland) 2016:

Power to investigate complaints made by a person aggrieved

5.(1)The Ombudsperson may investigate a complaint, made by or on behalf of a member of the public who claims to have sustained an injustice (in this Act referred to as ‘a person aggrieved’), if the requirements of this section are met.

(2)The complaint must relate to action taken by a listed authority (see sections 12 and 13).

(3)The complaint must relate to a matter which can be investigated (see sections 14 to 23).

(4)The correct procedure must have been followed (see sections 24 to 27).

In some Australian jurisdictions, navigability is taken a stage further, by making references not just to different provisions within an Act, but to provisions in different Acts. In the example below, the unauthorised version of the Children and Young People Act 2008 (as put online by the Australian Capital Territory Parliamentary Counsel’s Office) is as follows:

356 Offence – mandatory reporting of abuse

(1)A person commits an offence if –

(a)the person is a mandated reporter; and

(b)the person is an adult; and

(c)the person believes on reasonable grounds that a child or young person has experienced, or is experiencing –

2.5.7.Highlighting Defined Terms

Defined terms make legislation easier to read and understand. So for example, rather than using the phrase ‘a person under the age of 18’ throughout it, the Children Act 1989 instead uses the term ‘child’ and then defines it in section 105. This is both a useful shorthand term and a good way of indicating meaning to the user. But what of the reader who does not know of the existence of interpretation sections? They may be struggling to know whether a 17-year-old is a ‘child’ for the purposes of the Act. This is where navigability principles can assist. If a defined term is used in legislation, it can be highlighted in some way to indicate that more information is available on that term. For example, in Guernsey legislation, defined terms are in bold in the place where they are defined. This at least draws the readers’ eye to the term so that they have a chance of knowing that this term has additional levels of meaning attached to it.¹⁸⁸ Even better would be a system whereby the casual reader would immediately be alerted (for example by ‘pop-up’ text) to the fact that a particular word has a precise meaning, and an easy way of getting (for example via a hypertext link) to that meaning.

2.5.8.Numbering

In physics, a quantum is the smallest size of an entity possible. So, a photon is a single quantum of light, incapable of being divided any further. However, in legislation, there is no limit to how far the gap between two sections in a statute can be divided. So, between sections 171 and 172 of the Social Security Contributions and Benefits Act 1992, there are 54 sections. The full list is as shown in Table 3.

Table 3 Sections numbers between 171 and 172

Westlaw does a better job than the official government database of legislation, legislation.gov.uk. That database dodged the question, stating: ‘This is the original version (as it was originally enacted). This item of legislation is currently only available in its original format’.

When subsections and paragraphs are factored in, this means that we have (for example) section 171ZZ5(3)(b) of this Act. How are end users meant to navigate through this? How are they meant to know that 171ZZ1 comes after 171ZA but before 171F? Although there is reason behind it, it just looks like gibberish to the ordinary citizen.

The Immigration Rules are worse.¹⁸⁹ We can find:

•Rules with a number, eg rule 15 is about the common travel area.

•Rules with a number with a single letter as a suffix, eg rule 20A is about non-lapsing leave.

•Rules with numbers with multiple letters as a suffix, eg rules 245ZZ to 245ZZE deal with tier 4 (child) students.

•Rules with numbers with letters as prefixes and suffixes, eg rules A57A–A57H deal with persons entering the UK for short term study.

•Rules with multiple letter prefixes eg E-ECP.3.1 deals with financial requirements.

So, in Appendix FM-SE paragraph A1 reads ‘To meet the financial requirements under paragraphs E-ECP.3.1., E-LTRP.3.1., E-ECC.2.1. and E-LTRC.2.1. of Appendix FM, the applicant must meet’. The Court of Appeal in considering Immigration Rules stated that the ‘provisions have now achieved a degree of complexity which even the Byzantine emperors would have envied’.¹⁹⁰ The Supreme Court spoke of the difficulty for users ‘to navigate their way around the requirements’.¹⁹¹ These rules will most likely have been renumbered 10 times in the period between this paragraph being written and it being published.

This kind of complex numbering makes legislation extremely difficult to navigate and undermines accessibility and the Rule of Law. Although it may make things easier for government if the same numbers remain in place through many amendments, it does not help the reader. At some point the legislator has to insist that instead of the tortuous quantum fracturing of squeezing in new numbers where none ought to exist, there should instead be a fresh restatement of the legislation containing only actual numbers. And perhaps a schoolchild can advise that there is no such number as 171ZZ5.

2.5.9.‘Hidden’ Provisions

Legislation is not a weapon to be used against the people, it is a tool to be used for their benefit.¹⁹² Equally, it is not something to be ‘sneaked through’ the legislature, it is something to be enacted by the legislature. A statutory provision is not navigable if it is hidden somewhere within the legislation, in an unexpected place, or in a place that is out of context with its content, or with a misleading title. Thring stated that provisions shouldn’t be hidden in a schedule.¹⁹³ The more important, or revolutionary a provision is, the more prominence it should have in the legislation. Take as a positive example the legislation authorising the UK Prime Minister to initiate the procedure for leaving the European Union. This fundamentally important provision is not buried in a schedule of a bill on restructuring government, it is section 1 of the European Union (Notification of Withdrawal) Act 2017.

An Italian example shows the pernicious use of hidden provisions. The Italian Prime Minister Silvio Berlusconi had lost a civil case and was obliged to pay damages totalling millions of Euro. He wanted to amend articles 284 and 373 of the Civil Code so that payment of damages would be suspended pending appeal to the Court of Cassation. In 2011 he inserted this amendment to the Civil Code into a budget bill that was being fast-tracked through the Italian parliament due to the financial crisis. It was spotted and removed.¹⁹⁴ Although this amendment was clearly wrong as being inserted for the personal benefit of the legislator, it was also wrong as it was in the wrong place, seriously hindering navigability.

2.6.1.Tax Legislation and Extra-Statutory Concessions

It is a fundamental principle of UK tax law that we can only be taxed in accordance with legislation laid down by Parliament.¹⁹⁷ However, for at least 100 years, we can now be ‘untaxed’ (or excused of an obligation to pay tax) by an Extra-Statutory Concession (ESC) granted by the government.¹⁹⁸ As its name indicates, this is a situation, not provided for in legislation, where the government does not insist on applying the full rigour of the law, and thus reduces the tax liability of an individual. ESCs have been given (sometimes begrudgingly) a seal of approval by the judges.¹⁹⁹

Extra-statutory rules are anathema to the general principle of the Rule of Law that we are all subject to the law. They are also anathema to element 1 of Bingham’s definition of the Rule of Law as they are not accessible alongside the legislation. Having said this, the government do publish the list, together with details, of all ESC annually. The government do have a rolling programme to turn some ESCs into actual legislation. There is legislative power to give effect to ESCs, including turning them into statutory concessions.²⁰⁰

In discussing ESCs, Freedman’s main objections are in terms of lack of certainty and clarity over the law.²⁰¹ Tax-payers could not be sure how much they could rely upon these ESCs. Freedman suggests that ‘one way forward would be for all concessions and guidance to have a statutory basis’.²⁰² However she sees this as a counsel of perfection and not feasible for the small, transitory and complex areas in which ESCs normally operate.

Three recommendations can be made. First, so far as possible, there should be nothing outside the statute book that denudes the contents of the statute book – the principle of inclusivity requires that concessions ought to be contained within the statute book. Second, where this is not possible, the reader ought to be alerted at the very least to the existence of such things as extra-statutory concessions when they are reading the tax statutes. Third, those ESC must be easily accessible in full.

2.6.2.Broadly Drafted Legislation Constrained by Non-legislative Indications of Narrow Application

On occasion, legislation on its face has a broad remit, but the persons introducing the Bill into Parliament state that it will only be applied in narrow circumstances. If the intention is that it will only be applied in narrow circumstances, then this should be stated on the face of the Bill and should not be reliant upon political whim. There are two problems with this kind of extra-statutory statement. First, as they are not included with the legislation, they are inaccessible to the reader. Second, they are not legally binding and can be treated as effectively flim-flam.

As an example, consider the Anti-Terrorism Crime and Security Act 2001. This was enacted in response to the terrorist bombings on 11 September 2001.²⁰³ When introduced, the Home Secretary David Blunkett used the following phrases in Parliament to describe it:

•‘legislative steps necessary to counter the threat from international terrorism’;

•‘we will reinforce action against the perpetrators of organised crime, drugs and people trafficking’;

•‘the emergency legislation will build on the provisions of the Proceeds of Crime Bill to deal specifically with terrorist finance through monitoring and freezing the accounts of suspected terrorists’.²⁰⁴

From the general tenor of the debate, it was quite clear that the powers contained within this bill, in particular the power to freeze assets, were to be used for the purposes of defeating terrorism. The actual legislation contained the following as one condition for making an order freezing assets of someone outside the UK:

The first condition is that the Treasury reasonably believe that—

(a)action to the detriment of the United Kingdom’s economy (or part of it) has been or is likely to be taken by a person or persons …²⁰⁵

However, during the financial crisis, there was a risk to UK money held by the Icelandic bank Landsbanki. The government made an order under this provision freezing money belonging to the bank. There was no suggestion of terrorist activity on the part of the Icelandic banks, instead this was a measure used as part of the response to the economic crisis of the time. Although the actual action done may have been necessary and even laudable in economic terms, the fact remains that legislation with an ostensibly narrow remit (terrorism) was used for a much broader purpose. There was no requirement for there to be terrorist activity under the terms of the Act in order for the freezing powers to be used, but the idea that it be used for purely economic reasons seems out of keeping with how a reasonable person would expect the legislation be used. As Smyth stated ‘it was controversial because it was made under the Anti-terrorism, Crime and Security Act 2001’.²⁰⁶

Terrorism powers being used for non-terrorism purposes is the most egregious example of this kind of abuse. But it can happen in other circumstances where a Minister seeks a broad power in legislation, but then gives an undertaking in Parliament that it will only be used for a particular purpose. To make an analogy, once you give the solider a weapon, he is going to use it against all enemies, not simply those you tell him to use it against at the start.

When a Minister has to stand up and say, ‘Yes, this offence is really broad but don’t worry because we intend for it to be prosecuted only in a subset of places’, that is the very definition of an overbroad criminal offence, and it should be tightened up to cover the egregious cases that are on the Minister’s mind.²⁰⁷

This form of legislation principle is a matter of degree and balance. If the legislator is aware of some important limitation to the way the legislation is to be applied, then that limitation ought to be contained in the legislation itself. To do otherwise is to mislead users and subvert the legitimacy of the law-making process.

2.6.3.Broadly Drafted Legislation Due to the Difficulty of Defining Exactly What is to be Covered

Sometimes it can be very difficult to enact a law in such a way that it will cover all the cases that are intended to be covered, while excluding all the cases that are not intended to be covered. This was precisely what happened with the Sexual Offences Bill in the UK Parliament. The main thrust of the law was to protect children from sexual assault. The thing that was meant to be excluded was children involved in natural, low-level, sexual experimentation with other children. One specific difficulty that exercised Parliament greatly was whether the law really ought to be criminalising consensual kissing or petting among 14- and 15-year-olds. The entire debate in Parliament is extremely interesting. The Minister responsible for piloting the legislation through Parliament was David Blunkett, the Home Secretary, and he said that ‘it is extremely difficult to come up with a formulation’²⁰⁸ that includes and excludes the right activities. He said that the government simply could not craft a provision that protected those who ought to be protected, but would not allow for prosecutions of those who ought not to be prosecuted. In fact he went on to say that he would offer ‘a flagon of champagne for anyone who comes up with a satisfactory answer’.²⁰⁹ Dominic Grieve MP (who later went on to become the Attorney General) stated that:

I have an inherent anxiety about administrative discretion. I accept that administrative discretion may be the only remedy in cases where the CPS will not charge. Nevertheless, when such matters are put on to the statute book in such stark terms, I always fear that, at some point, something will not work properly and that we will end up with prosecutions that cause serious problems. I admit to the Secretary of State that I am not sure that I know the answers, and I suspect that, if he had known the answers, he would have already put them on the statute book.²¹⁰

One solution proposed to this problem was that the Director of Public Prosecutions would offer guidelines on what activities should be prosecuted and what not. To my mind, this breaks the principle of inclusivity – an important limitation on the legislation is placed, not within the legislation, but outside it. The person reading the legislation would come away with the misleading impression that if they did X they would be prosecuted, even though the defence of Y existed somewhere outside the statute book. I can only agree with Sandra Gidley who said in the debate that ‘I feel that we need to deal with that matter in the Bill’.²¹¹