"Laws are like sausages: it is better not to see them being made"
We talk of an ideal parliament being one which scrutinises and improves laws before passing them, but is that its main function, and should it do it so often?
Your starter for ten, or perhaps your answer: yes, I know the quotation in the title is attributed to German chancellor Otto Prince von Bismarck but there is no record of him actually saying it (he would have presumably said it in German anyway, although he also spoke English, French, Italian, Polish and Russian). Nor would it have been a new observation, since a similar thought had been attributed to American poet John Godfrey Saxe, who as early as 1869 was supposed to have said that “laws, like sausages, cease to inspire respect in proportion as we know how they are made” (which is more insightful but less pithy). The more epigrammatic version was not placed in the mouth of Bismarck until the 1930s. But whoever said it, there is an element of truth to it.
It is worth asking “What is Parliament for?” The parliamentary website (which, parenthetically, is dramatically better in terms of content and operation that when I was in the House) outlines four major functions:
To “check and challenge” the government and its policies and actions;
To scrutinise, amend and approve draft legislation;
To “debate the important issues of the day”;
To “check and approve” Government spending.
I apologise slightly for the arch quotation marks, but I don’t like the language which has obviously been deliberately chosen to describe parliamentary procedure. “Checking” and “challenging” sound like a hellish mash-up of fifth-form maths lessons and mediocre corporate leadership, as if MPs were comparing bills and estimates against a “correct” version and making sure the two tally. However, what the explanation does do, which is useful, is reinforce the idea that government and parliament are separate entities, although in our system the former is drawn from the membership of the latter.
Composing the list is not an easy task, and I think the four functions the website has chosen are not a bad attempt. A purist might query the separation of 2. and 4., since the government derives its revenue from finance bills (to collect taxes) and supply bills (to authorise government spending). But there is extra-legislative financial scrutiny too, most obviously in the regular work of the Public Accounts Committee (PAC), the “queen of the select committees”, which since 1862 has scrutinised “the Accounts showing the appropriation of sums granted by Parliament to meet the Public Expenditure”. PAC is supported by the National Audit Office, which examines government accounts in intricate detail. As PAC is traditionally chaired by a senior opposition MP, often a former Treasury minister, it offers a high-profile platform for robust, even confrontational, oversight of spending. In recent years, Dame Margaret Hodge (2010-15) and John Boyd-Carpenter (1964-70) were particularly prominent figures in the House of Commons, but one should also recall that Harold Wilson served as chairman from 1959 to 1963, until he became leader of the opposition.
What I want to think about a little is Parliament’s legislative function. I won’t dwell here on the procedural aspect of how bills become law, as many readers will already be more than familiar with the process, but anyone who is uncertain can find the outline again on the parliamentary website. I want to examine at this point the volume of legislation, how it is handled and what effects it has.
The idea of a parliament which is effectively always sitting is, of course, a product of the English Civil War, or rather, as we more inclusively say now, the Wars of the Three Kingdoms (as Charles I was king of England, Scotland and Ireland separately, and each had its own legislature and executive, although the English government in London was dominant). Although by the end of the 16th century each parliament had a more-or-less permanent home (St Stephen’s Chapel in the Palace of Westminster, the Tolbooth then Parliament House in Edinburgh and Dublin Castle for the Parliament of Ireland), none was by any means an ever-present institution. In Elizabeth I’s 45-year reign, for example, the English Parliament only met 10 times, never for more than three sessions at a time. In Scotland, the unicameral parliament began the 16th century waning to such an extent that it was beginning to be of questionable value; a comparison might be made with France, where the États généraux were dissolved in 1484 and was not recalled until 1560, then met only a handful of times more before the Revolution in 1789.
For the past three centuries or so, then, we have had a legislature which, with periodic adjournments, is always sitting, albeit divided into parliaments and then into sessions. In that time, the scope of the state has expanded vastly, and with it the responsibilities of Parliament, especially since the middle of the 19th century. By 1900, the Empire in its entirety was governed and administered, in the broadest sense, by about 40,000 employees. Today, the public sector employs around 5.8 million people. In addition, the electorate grew beyond all recognition: in 1831, there were a little over half a million voters, one in 20 adults being entitled to vote. At the last count, in 2021, it was 46.5 million.
The kind of legislation with which Parliament now deals is much simpler. Essentially, laws are overwhelmingly the result of public bills, sponsored either by the government or by private Members. Private and hybrid bills are very rare, although a great deal of effort is going into the management of the hybrid High Speed Rail (Crewe - Manchester) Bill, which will allow for the next stage of High Speed Rail 2 and is currently being considered by a select committee to allow people affected to petition against it. But it was not always this way. Parliament used to be the last recourse for those who sought a benefit or redress unavailable under common law or existing statute. That meant that Parliament had to consider petitions for divorce, legal changes of name and the naturalisation or granting of citizenship to foreigners, all known as private acts (personal acts before 1948). Parliamentary authority was also required for local acts, which could authorise major projects like railways or canals, create corporations and grant monopolies.
Although personal and local acts could be considered and passed more easily than public acts, the volume was substantial. To take a year at random, in 1870, there were 177 local acts and three private acts. The sheer bureaucracy required to manage that level of activity must have been exhausting. By contrast, in 2004, for example, there were six local acts, and in 2005 there were none at all.
The modern process of consideration of legislation is a key part of Parliament’s functions, but it is by no means the most visible. The most popular part of proceedings in the House of Commons with the BBC Parliament audience is, inevitably, Prime Minister’s Questions, which attracts around 120,000 viewers. PMQs, after all, is the rowdiest and most political significant event (in the short term) of the parliamentary week. Equally, significant ministerial statements, on which the minister then answers questions, are easy to follow without any prior knowledge of the subject or of procedure. But following legislative scrutiny for the layman is much more difficult.
This is not a criticism of the process, of the nature of bills themselves, or the way in which MPs conduct themselves. Some things are just hard, and need a lot of knowledge and concentration, and consideration of a bill falls into that category. Admittedly the most complicated and impenetrable stage, line-by-line scrutiny, generally takes place in a public bill committee, “upstairs” on the committee corridor, but even so it can be difficult to dip into legislative proceedings.
Over the past century, the number of public acts being passed has actually fallen, apart from a temporary upswing between 1950 and 1970. In the first third of the 20th century, Parliament passed an average of 112 acts each session. By the 2010s, it was 52 per session: the number has halved. However, there has been a steady increase in secondary legislation, instruments and regulations which are passed with a much lower level of scrutiny under the authority of a statute. From the 1950s to the 1990s, around 2,000 statutory instruments (SIs) were approved each year. In the 1990s, that soared to 3,200 and 4,200 in the 2000s, before falling back to about 3,000 per year in the 2010s (although far more were passed between 2010 and 2015 than thereafter). This blossoming of SIs has been criticised because of the much lower level of scrutiny, ministers standing accused of trying to keep more and more powers off the face of bills and hidden in secondary legislation.
Perhaps the most telling statistic is a slightly simplistic one: pages. In 1930, the average at was between 15 and 20 pages long. By 2016, it was over 80 pages. That sheer weight of legislative detail is a significant sign that laws are getting more detailed, more complicated and, perhaps, more far-reaching. Interestingly, SIs have remained if a similar length since the 1960s, only their number increasing.
What does all of this mean? Essentially, we are making more laws. But most of the growth is coming from secondary legislation, which tends to relate to finer detail and smaller issues. This brings me to the central part of my argument. Simply, we legislate too much. In our legal system, once a measure is on the statute books, there it remains until it is specifically repealed. The oldest extant piece of legislation is the Statute of Marlborough, the remaining parts of which are cited as the Distress Act 1267 and the Waste Act 1267. Think about that for a moment, think about what the world was like when that piece of legislation was drafted (in Latin, although within 100 years it would be required to conduct all legal pleadings in English). Michael VIII, the founder of the Palaiologos dynasty, had recently recaptured Constantinople from Latin crusaders and been crowned emperor; the term Hanse, which would give its name to the powerful commercial league of merchant guilds and market towns in northern Europe, is first found in an official document; and barely a decade had passed since the Mongols had sacked Baghdad, killing perhaps a million Arabs. Still this law applies.
I’m a great supporter of our legal system(s). I think the common law was one of England’s greatest gifts to the world, a system at once flexible according to circumstance but also predictable, consistent and stable, which allows a (stereotypically English) process of gentle but constant iteration and reiteration. I cherish the leveller of the jury trial (though I am sceptical of Scotland’s 15-member panels), and, however much I may labour under (slightly performative) misanthropy, I would rather take my chances with 12 ordinary people than a judge or judges raised and schooled within the legal establishment and thinking of his or her professional standing. And I am more than content with our uncodified—not unwritten—constitution. But that doesn’t mean I’m not conscious of its limitations, or maybe even just its distinctive features, and one of those is legislative accretion.
There have been several attempts to address this. The Conservative/Liberal Democrat coalition had established a policy for regulation of ‘one in, two out’, so that any new impositions on business had to be more than offset by a larger reduction in the regulatory burden. To apply this more widely to the statute book would be theoretically advantageous but perhaps too daunting a scheme, though as a general principle it is a good one. So the simplest and most effective way of tackling legislative accretion is to pass fewer laws.
That is not quite the self-imposed stricture it might at first seem. Governments of every political persuasion tend to succumb to some of the same faults and failings, and a common mistake is for ministers to think that new legislation is their only mechanism for effecting change. It’s an easy misapprehension to fall into, especially if you are blindsided by a sudden issue to which you had thus far paid little attention: solve a problem by passing a law which stops something happening or causes something to happen. Give it a snappy title or—an increasingly common and irritating habit—a nickname, ideally based on a sad-looking or deceased child, and you have all the ingredients of a presentational success. It leads the unwary to following what Antony Jay and Jonathan Lynn in Yes, Prime Minister described as “the politician’s syllogism”.
Sir Humphrey Appleby, the cabinet secretary, discusses this habit of ministers with his predecessor, Sir Arnold Robinson, and they use it to illustrate the contrary and frequent value of doing nothing. The syllogism and its explanation of impulsive political behaviour runs like this:
We must do something.
This is something.
Therefore, we must do this.
Yes, Prime Minister is often praised for its accuracy in capturing the quiddity of the British civil service, and it is a popular source of quotation and inspiration for political historians and analysts. Many practitioners found it deeply resonant—not wholly surprising as Jay and Lynn had researched extensively, drawing very heavily on the diaries of Richard Crossman in particular—and Lord Donoghue, head of the Number 10 Policy Unit from 1974 to 1979 and sufficiently long-serving to be a minister for the first two years of Sir Tony Blair’s premiership, was a great admirer but also noticed in 1997 that its content and style influenced several new ministers who had no experience of government office. There is an opposite view that it was a caricature of Whitehall (and I had a colleague who hated it with a passion for its distorting depiction of the machinery of government), but the real-life existence of the politician’s syllogism is certainly supported by evidence.
Legislate in haste, amend at leisure. Governments are prone to what is called “declaratory legislation”, which is the practice of making new law to demonstrate that ministers are responsive to public pressure or sudden events, or to underline attachment to certain principles or policies. The dangers of this are several: hastily drafted legislation is almost always to be deprecated, as it merely increases the chances of mistakes, lacunae and unintended consequences, it risks misidentifying the problem you are trying to solve, and it gives little time to understand whether the powers you seek are already available in statute, and how the measures you are introducing will interact in the long term with existing legislation.
This is not a revelation. Government bills and instruments are drafted by the Office of Parliamentary Counsel, a department of the Cabinet Office which contains lawyers with a particular expertise in legislative drafting. It is headed by First Parliamentary Counsel, the wonderful Thurso-born Dame Elizabeth Gardiner KC, one of the most straightforward, friendly and reliable civil servants I worked with in my time as a parliamentary official, and it consists of very gifted, very intelligent and high specialised lawyers whose only fault, I would gently suggest, is a tendency towards mild eccentricity as their careers progress. The most singular but electrifyingly able of them with whom I worked, Daniel Greenberg, is now the Parliamentary Commissioner for Standards. Parliamentary Counsel are very much alive to the dangers of declaratory legislation, and its potential shortcomings are explained in their 2013 review of legal tangles, When Laws Become Too Complex. But they are civil servants like any Whitehall warriors, and they must follow ministerial instructions.
The Coronavirus Act 2020 was an example of the government, understandably, scrambling to look decisive in addressing the onset of the global Covid-19 pandemic but legislating in extreme haste and arguably indulging in declaratory legislation. This should be seen in the light of the extraordinary scale of the pandemic; equally, however, the government is supposed to prepare for major catastrophes and should be able to rely on some degree of planning and preparation. The passage of the bill was very swift: the health and social care secretary, Matt Hancock, introduced the bill on 19 March 2020, and it went through all its stages in the House of Commons on 23 March without a division. It was then introduced in the House of Lords by health minister Lord Bethell on 24 March and given a Second Reading the same day, being referred to a committee, and the remaining stages took place on 25 March. Royal Assent was signified late the same afternoon and the bill became law, six days after it had first been presented to Parliament. As a safeguard, and under pressure from former cabinet minister David Davis, the government had amended the bill in the Commons so that its provisions had to be renewed by Parliament every six months.
There were concerns about the speed of the bill’s passage. Davis, demanding a renewal mechanism, remarked reasonably “There is nobody who can scrutinise a 300-page bill in one day. The government itself will have made mistakes, it’s a fact of life”. But there was a feeling that time had been of the essence, and even an opposition spokesman in the House of Lords, Baroness Thornton, praised it as a “perfect exercise in consultation and work across the House”. There were, however, objections which went deeper than purely procedural matters.
The case against the government is essentially this: that ministers sought extremely wide powers for authorities to act the control the virus, and, partly due to the great haste involved, not only were some of those powers excessively authoritarian, but some were simply unnecessary as they were already provided for in existing statutes. One issue was the ability to detain “potentially infected persons”. Sections 129 and 130 of the Health and Social Care Act 2008 already gave police officers, subject to the authority of a magistrate, the power to require the “medical examination, detention, isolation or quarantine” of those suspected of being infectious to address a “significant risk to human health” and allowed police to “require persons to provide information or answer questions” about their health under threat of detention. It can be argued that these existing powers made schedule 21 of the Coronavirus Act unnecessary.
The Coronavirus Act also allowed the prevention of mass gatherings to enforce social distancing and reduce the risk of infection. These were broad powers and would not be subject to review for six months. However, section 19 of the Civil Contingencies Act 2004 made provision for making emergency regulations to prohibit “an event or situation which threatens serious damage to human welfare in a place in the United Kingdom”. This surely already gave the government the necessary authority to prevent mass gatherings.
It is especially worrying that the Civil Contingencies Act was overlooked as it was the culmination of a long process to prepare for this kind of emergency situation which began with the passage of the Emergency Powers Act 1920. The act was amended in 1964 and the institutional framework to support it was strengthened in 1973 with the creation in the Cabinet Office of the Civil Contingencies Unit (which became the Civil Contingencies Secretariat in 2001). The 2004 Act contains a ‘triple lock’ procedure to ensure that an emergency can only be declared if there is a serious threat; that the regulations are necessary; and that any measures are proportionate. So there was a pre-existing superstructure of legislation, organisations and procedures to deal with exactly the sort of emergency which Covid-19 represented, but this was almost wholly ignored by ministers in favour of drafting expansive and draconian new laws de novo. The very purpose of structures and institutional memory is to avoid the time-wasting and ad hoc measures to which the government in fact resorted at the beginning of 2020.
In evidence to a joint meeting of the Health and Social Care and Science and Technology Committees in December 2020, Lord Sedwill, recently having stepped down as both secretary to the cabinet and national security adviser, gave a glimpse of the government’s thought processes. On the use of the Civil Contingencies Act, he argued that it would have caused political difficulties with the devolved administrations, as it is UK-wide legislation, whereas section 100 of the Coronavirus Act applied different parts of the law to different jurisdictions, allowing for some variation and initiative on the part of the devolved administrations. The judgement between political sensitivities and the prompt, effective and proportionate control of the virus is an arguable one.
Sedwill’s second point was hardly more convincing. He noted that the Civil Contingencies Act was drafted so that it was to be used “only if there isn’t a legislative alternative” (which in March 2020 there was not), and that the preference was to introduce powers “that are bespoke to the circumstances at the time, and, frankly, are not too broad brush and could not be abused”. In the event, he concluded, “we reckoned we had time to do this in the case of the pandemic”. Arguing that the 2020 Act was neither broad brush nor open to abuse, and further that the government had sufficient time is putting the very best, most Pollyanna-ish civil service gloss on what happened. The act went from introduction to Royal Assent in six days, and produced what the two select committees described in their eventual report as “lower levels of public trust and understanding of the regulations” which “created a gap into which misinformation was able to spread”.
The Coronavirus Act therefore demonstrates many of the dangers of acting hastily and with too much concentration on presentation. To seem in control and responsive, the government moved too quickly, despite what Lord Sedwill would later argue, which caused them to duplicate some statutory powers, ignore existing structures and produce a bill which was effectively unscrutinised—a 329-page draft law passed by both Houses in six days without a division—when all of these pitfalls should have been anticipated and identified and could have been avoided.
The charge of duplicating existing powers was laid at the National Security Bill which the government described as “new legislation to compel those acting for a foreign power or entity to declare political influencing activity—and criminalize those who do not”. The declaratory attraction of the bill was obvious: as tensions between western countries on the one hand and especially Russia and China on the other have risen, it suits ministers to be seen to take decisive and tough action on foreign influence, especially in the shadow of scandals like the Facebook-Cambridge Analytica data misuse. The aims set out by the government would be particularly addressed by its addition of a Foreign Influence Registration Scheme. But the Russell Group of leading UK universities argued when the scheme was unveiled in February that “current proposals will duplicate existing measures and capture many partnerships that are already scrutinized… that kind of duplication drains resources and increases the chance of genuine risks being missed”. That last point is significant: unnecessary provisions are not merely undesirable from a theological view of legislative tidiness, but they accumulate and make the interpretation and implementation of the law more difficult and more open to challenge.
In this case, the government, to its credit, backed down. When the duplicate provisions were challenged in the House of Lords, the Home Office minister, Lord Sharpe of Epsom, conceded the principle and promised that the difficulties would be minimised. The bill is now in its final parliamentary stages of agreeing amendments and will receive Royal Assent. So it stands as an example of a government using too broad a brush in terms of drafting for largely presentational purposes, but agreeing to look again when criticised by interest groups and faced with amendments in Parliament.
It is perfectly possible to make the argument of declaratory legislation in the case of the International Development (Official Development Assistance Target) Act 2015. This began in the autumn of 2014 as a private member’s bill introduced by the Liberal Democrat former cabinet minister Michael Moore, and its purpose was twofold: firstly, it would enshrine in law the target of spending 0.7 per cent of gross national income (GNI) on official development assistance, which had been set by the United Nations General Assembly as far back as 1970 but which the UK had never achieved until that year, 2014; second, it placed a duty on the secretary of state to introduce independent evaluation of that spending. Moore had been placed second in the ballot for private members’ bills and so stood a good chance of being able to steer his bill through the Commons so long as the government did not actively oppose it; and the aspiration behind it, of spending 0.7 per cent, was supported by all three major parties. Moore had been the Liberal Democrat spokesman on international development before the 2010 so he knew the policy field well, the Conservative/Liberal Democrat coalition had committed to achieving, maintaining and putting into legislation the target at its inauguration in 2010 and the Department for International Development had provided explanatory notes for the bill.
The bill made its way through Parliament will reasonable speed for a private member’s bill, being approved by the Commons in December 2014, introduced into the House of Lords (by fellow Liberal Democrat Lord Purvis of Tweed) later that month and then being passed by the Lords and receiving Royal Assent in March 2015. That was all very satisfactory, and few could doubt that the intention behind the new act was noble and philanthropic. However, it was an unusual kind of law, committing the government for as long as the statute remained extant to spend a certain amount of money on a specific area. But most people were content with this curious beast, given its benignity of intent and effect.
However, there were some discordant voices during its various stages of scrutiny in both Houses which accepted that the notion of spending 0.7 per cent of GNI was a good thing, but objected to its placing on the statute book. The reason was simple. Quite apart from being seen oddly to privilege spending on overseas aid as especially important, it was largely without effect if a future government came to dislike the mechanism. After all, it is a fundamental principle of the British constitution, part of the theory of parliamentary sovereignty which had deep roots in the early modern period but was clearly articulated by A.V. Dicey in his Introduction to the Study of the Law of the Constitution, that no parliament may bind its successor. That is, a parliament cannot ‘ring-fence’ a law as being beyond future amendment or repeal; one parliament may pass a law which the next can simply repeal with a short bill (and could, if it wished, pass another law with exactly the opposite effect). So the act placed a statutory duty on the government, but that or any subsequent government could remove that statutory duty with a simple repeal bill, and it would be as if the act had never been law.
The supporters of the bill, not least Moore himself, had accepted that constitutional truth but argued that it would be deeply symbolic, arguing that “maintaining our commitment [to the 0.7 per cent target] will enable the UK to show leadership across the world”. He added that the security of the target would give the UK government a degree of stability which would make long-term planning easier—how many other departments would have welcomed that level of certainty!—and would insulate the international development secretary from the yearly brawling with the Treasury of the spending round. Andrew Mitchell, a former secretary of state and a sponsor of the bill, admitted the status of the legislation but was happy to justify it: he told the House that he did “not like declaratory legislation and fully understand why many Members believe that it is insulting”, but suggested that this bill was in a different category as it was putting on a statutory footing a target which had already (that year, for the first time) been achieved.
From my silent perch as a clerk, I didn’t think much of the argument behind the bill, thought it was ineffective and would only store up trouble for the future, and disliked the principle of elevating one discrete part of public spending above all others. But no-one asked me. It was not long before such difficulties arose. In the summer of 2020, the foreign secretary, Dominic Raab, whose department was about to re-absorb DfID after its 23 years as an independent ministry, announced that the financial pressures imposed by the coronavirus pandemic would force the government to reduce Official Development Assistance by £2.9 billion but that the 0.7 per cent target would be maintained. In November’s Spending Review, however, the chancellor, Rishi Sunak, revealed that ODA for 2021 would be reduced to 0.5 per cent of GNI. Although the acy required the government to spend the higher proportion, the only effective sanction for non-compliance was that a minister had to report to Parliament why the target had been missed, which was obvious to everyone.
This showed the hollowness of the act. Although there were objections to the cut in the aid budget, and although the government was careful to commit itself to returning to the 0.7 per cent level and set out conditions for doing so, everyone understood that the pandemic had wrecked the government’s spending plans and that the pain would need to be shared around Whitehall. So, in less than six years since the act’s passage, its effect had been reduced to its declaratory one, that for those years, the UK government had enjoyed some bragging rights and credentials for international leadership in the form of an unenforceable commitment. Had it been worth the parliamentary and administrative time and effort? And was it possible to quantify the presentational and diplomatic advantage it had conferred. The jury is very much out; and I suspect it will never come back. But it does seem, at this distance, like nugatory work.
A laudable example of the government avoiding declaratory legislation occurred only this year. After considerable public concern about a rise in the prevalence of drinks being spiked, there were calls for new legislation to address the issue. The chair of the Home Affairs Committee, Dame Diana Johnson, had endorsed these calls, and in 2022 the then-home secretary, Dame Priti Patel, had announced that she was looking into “a specific criminal offence to target spiking directly”. Sarah Jones, a shadow Home Office minister, had agreed, saying “We should call a spade a spade in this case and introduce a specific offence for spiking”. With admirable restraint, as the opportunity for good publicity was obvious, the safeguarding minister, Sarah Dines, wrote back to Dame Diana saying that the government had not found “any gap in the law”. She continued, “The existing offences cover all methods of spiking, including by drink, needle, vape, cigarette, food or any other known form”. Spiking with the intent to achieve sexual intercourse is covered by the Sexual Offences Act 2003, while administering any substance in order to cause harm is captured by the provisions of section 23 of the Offences Against The Person Act 1861. Fresh legislation would therefore only add to accretion without creating any new offences or providing additional powers to the police.
I hope this has demonstrated that there are several different ways in which Parliament, at the government’s initiative, legislates and that these have different outcomes. And I hope it has also given you pause for thought. Do we leap too quickly to legislation, despite the inevitable accretion of laws like Japanese knotwood? I think we might, and, were I advising a government, I would suggest it was worth further exploration. There can be several different routes to a particular objective, and the most obvious is not always the most effective or the most elegant. Sometimes, less is more.