Policy Pulse 19: Fixed Term Parliaments
Why was the late and unlamented Fixed Term Parliaments Act such a failure?
Hello, and welcome to your weekly pulse of public policy. This time out, I’m making a rare foray into constitutional issues. Often I’d say this is a bit close to politics for this newsletter, but the recent repeal of the Fixed Term Parliaments Act (by the Dissolution and Calling of Parliament Act 2022) means that we have a policy failure to discuss.
Accordingly, I’m not going to get into the meaty legal and constitutional issues thrown up during the passage of the new legislation, such as questions around how you recreate a prerogative power using primary legislation. Rather, I want to explore the ways in which the FTPA was (and wasn’t) a policy failure, and try to explain them. When a piece of legislation this bad hits the statute book, and is subsequently operated so badly, it’s almost tempting simply to conclude that the people who devised it and operated it were just stupid. And while the architects of the FTPA do not emerge with any credit, that’s probably not a sufficient explanation on its own.
A starting point for a more meaningful explanation is that our political class are unusual because they like politics. They are interested in it and enthusiastic about it: for some, it is their life. For that reason alone, there will never be any such thing as a House of Commons that is representative of the general population. It can also mean that professional politicians are well versed in political lore – not necessarily political history as such, but in understandings and retellings of past politics, often skewed and shaped by the particular partisan bubble they inhabit. Politicians are interested in politics as a game: that is not to say they don’t take its real-world consequences seriously (some do, some don’t), but on at least one important level they engage with politics as a tactical and competitive exercise.
This approach, of always having one eye on partisan advantage, and viewing developments through a well-developed knowledge (though not necessarily understanding) of past political events, can certainly have consequences for regular policy decisions such as energy policy, to pick an example not at all at random. But those consequences are probably rather greater for constitutional issues, as there are no immediate real-world manifestations of things going wrong if the policy proves bad: constituents will not flood their MP’s mailboxes with complaints about how a change to the electoral system has affected them, for instance. And when there are real-world consequences, they are overtly political in nature, and so viewed again through that same political lens.
Experience with the Fixed Term Parliaments Act suggests this may be a particular blind spot for policy-making when it comes to constitutional reform. In particular, the Act seems to have been framed on the basis of a folk memory of the calling of the 1979 general election (I’ll outline the relevance of this later), which hardly any of the leading politicians of 2010 had adult memory of. As a result, it didn’t very comprehensively solve any of the problems it might have been aimed at, and which certainly exist: the ability of the incumbent government to time general elections certainly hands it a potential advantage. But while it didn’t solve these problems the FTPA did create unintended consequences of far greater constitutional importance than the short-lived effects of the Act itself.
But first things first. A regular criticism of the Act is that it failed on its own terms: after it supposedly regulated general elections at five year intervals, unscheduled elections took place in both 2017 and 2019. Actually I think this is the least convincing of all the possible criticisms of the FTPA, and that it’s fair to say that on its own terms it was at least a partial success. It did not restrict elections to a fixed schedule come-what-may, but rather it contained mechanisms to enable earlier elections if political circumstances demanded them. So the 2017 election happened through the operation of the Fixed Term Parliaments Act itself, with a two thirds majority of MPs voting for it. The same route might have been used in 2019, as Labour’s view eventually shifted to favour an election, but by the time that happened the wheels were in motion to pass the Early Parliamentary General Election Act, by-passing the FTPA. What’s more, earlier in 2019 the provisions of the FTPA had succeeded in preventing Boris Johnson calling an early election as his preferred method of busting the Brexit logjam.
However, the two early elections certainly did show up shortcomings with the Act and its overall approach. The option of using a separate bill to call an election notwithstanding the Act’s provisions was mooted in 2017, and shows a crucial difference between a codified constitution and a non-codified one: in the UK, there is no separate form of constitutional law that requires extra efforts to change; if there was, it would mean that by-passing the rules on election timings would have required an amendment to the constitution, which in most countries requires clearing a much higher hurdle than just passing regular laws. With the UK’s form of parliamentary sovereignty, there is no higher form of law than primary legislation, and any piece of primary legislation can be amended or removed by another piece of primary legislation through normal procedure. So fixing parliaments in any meaningful sense is basically impossibly anyway.
But although it was in 2019 that the Act was by-passed, the 2017 election was probably a worse policy failure. Theresa May called that election for reasons of naked partisan advantage. Any argument that she required her own electoral mandate was of course entirely spurious: there is no constitutional precedent for it, and if she wanted an election for that reason she should have gone to the country in a timely way after becoming PM, most plausibly in the early autumn of 2016. Instead, she was blatantly aiming to kick the Labour Party under Jeremy Corbyn into oblivion (and also reinforce her own hand within her party in dealing with Brexit – although that situation had not yet developed into a crisis on the scale it would later become). In other words, she was trying to secure exactly the sort of partisan advantage that the prerogative power previously made available (and now makes available again) to the government of the day. And we found that in practice, the Leader of the Opposition can’t really afford to be seen to be running scared of the electorate (barring a very clear justification, as Corbyn could plead to an extent with Brexit in the balance in 2019), so the political dynamic was such that Mrs May got her election, and with it a lesson about being careful what she wished for.
What this tells us is that the idea of “fixing” parliamentary terms isn’t really a workable one in our constitution. A more sensible approach back in 2010 might have been to set a regular schedule of elections every four years, assuming nothing was done to change that, but set a maximum term of five, with a mechanism for the Commons to extend the parliament if it wished, and a further provision for early elections at parliament’s choosing. This would have taken at least some of the flexibility of the prerogative power away from the PM, although even then it would have been subject to the political dynamics of the day, as we saw in 2017. Perhaps if the issue ever gets revisited, policy makers will be able to learn from the experiences of the 2010s and take an approach more along these lines.
However, the Fixed Term Parliaments Act failed most seriously in terms of unintended consequences: without it, it’s certainly possible that the UK would have exited the European Union under the terms of Theresa May’s deal, and even a possibility that the campaign for a second referendum could at least have run for longer; and with such significantly altered outcomes a possibility, it’s also quite feasible that in this parallel universe without the FTPA, Boris Johnson would never have become Prime Minister.
How so? Well, to understand how the Act had such massive unintended consequences – indeed, had an impact on aspects of the constitution on which its terms were entirely silent – we need to get into what it said and did. Firstly, the Fixed Term Parliaments Act was about parliaments. It set out rules for fixing their terms – that is, how long they could last. Sounds obvious, but a crucial consequence of the Act was that many people thought it did other things too. It didn’t. It’s quite a short and readable piece of legislation, so check it out for yourself. As you’ll see, it fixes the terms of parliaments, and that’s it (other than tidying up some of the consequences for the devolved elections of the 2010s, which would otherwise have clashed with the newly scheduled general elections).
However, the Act had a major unintended consequence: it accidentally created a new convention about the formation of governments (which presumably has now also fallen). As you’ll have spotted, that is not about parliaments, even though that’s all the Act talks about. Under this convention, parliamentarians and others acted as if a lack of confidence in the government could only be expressed by the Commons under the terms of a motion envisaged by the Act for prematurely ending a parliament (the second of its two measures for breaking the fixed cycle). As losing the confidence of the Commons means that the government cannot remain in office, this new convention was of enormous importance: it created a new rule for when a government had to leave office, and therefore also for when it could remain in office. This remarkable development represents the most profound failure of the Fixed Term Parliaments Act: for a piece of legislation whose whole approach and guiding principle was to get away from conventions and unwritten rules to instead create an enormous new unwritten rule with massive constitutional and political implications was a failure of breathtaking proportions.
How did this happen? The Act created this problem by bringing the issue of confidence in the government into the matter of whether parliament should be dissolved. Anyone who understands the constitution generally, or the concept of separation of powers, can see the error: parliament and government (or the legislature and the executive branch, in more classic constitutional terms) are not the same thing. The two appear to have been muddled together as a result of memory and re-telling of what happened in 1979, when Callaghan responded to the loss of a confidence vote by calling a general election. However, that made perfect sense within the existing rules: if the Commons lost confidence in the government, either the government had to change or the Commons did; and as the government held the power of calling elections, it used that power in the hope of securing a new House of Commons that did have confidence in it, thus enabling it to remain in office. In short, the rule is that the government must enjoy the confidence of the Commons, NOT that a vote of no-confidence in the government triggers a general election, even though things happened to play out that way in 1979.
However, the legacy of 1979 appears now to be that MPs (and the entire commentariat) believe that confidence votes are mechanisms for regulating whether parliament is dissolved, not regulating whether the government remains in office. This shows an appalling failure of understanding of even the most basic ideas of how a constitution operates, and therefore illustrates my suggestion that politicians view things too much through the lens of their (skewed and imperfect) knowledge of past politics, not understanding of the constitution.
Where the Act brought in the issue of “confidence” was in a second mechanism for an early election: under this, after a no-confidence motion was passed a countdown of 14 days would begin; if the Commons could not pass a motion of confidence in the government – either because it was a new government, or because the Commons had changed its mind – the parliament would be dissolved automatically. The Act specified wording for both motions. Again, this poor mechanism illustrates that the Act’s provisions did not extend beyond the question of the duration of the parliament, as they did not contain any provision about what should happen during the 14 days. Was the government obliged to resign? Should the Leader of the Opposition be invited to form a government? Nothing was specified: if the provisions had ever been used, it would have been possible for the government to sit in office (at least pretending to look like it was trying to change the Commons’ mind, perhaps) and let the clock run out to force a new election. Depending on the politics, it’s possible to imaging the Commons using the Act’s two thirds mechanism to speed things up rather than wait for the clock to run down. It was a wretchedly useless process, precisely because it was restricted to the formation and dissolution of parliaments, and did not set out any rules for how governments could take office or be ejected from it.
Anyway, the net result of the FTPA and the convention it created that the Commons’ confidence could only be expressed under its terms had the effect of destroying the previously clear conventions around confidence in the government. This held that any government that lost the confidence of the Commons could not remain in office (although as governments also had the power to call elections, they could seek to change the situation by getting a new Commons, as Callaghan attempted). It was also the case that the confidence of the Commons could be expressed in any suitable vote: this might be an overt motion of confidence or, crucially, any substantial vote on a piece of legislation of central importance to the government’s programme such as, for instance, a deal to re-write the UK’s constitutional and economic relationship with the European Union. To be fair, this is a convention that had itself evolved over time: in the nineteenth century, any loss of a vote at all would require the resignation of the government, but in modern politics it has clearly been understood that it applies only to substantial items in the government’s programme.
I’ve had more arguments about this online than I care to remember. Even when the measure was first proposed, back in the heyday of blogging and early days of political use of social media, I remember an online campaign sprang up to object to the supposed proposal to require a 55% majority for votes of confidence in the government (which wasn’t proposed: the proposal was a 55% threshold for dissolving parliament early, which eventually became the two thirds mechanism). The misunderstanding about votes of confidence has therefore been longstanding. Even now, university lecturers on politics can write that the, “FTPA specified what did (and therefore what did not) count as a confidence vote.” This is completely untrue: the measure in the FTPA that mentioned “confidence” was a mechanism for enabling an early dissolution of parliament, and did not define specify any rules for expressing confidence in the government at all. However, given that the politicians operating the constitution in the 2010s did so widely under this misapprehension, it became true: by convention (though not by legislation), the Commons felt itself able to express confidence in the government only through motions under the terms of the FTPA.
The consequences that flowed from this misunderstanding, and the new convention about when and how the House of Commons could express confidence in the government, were profound at a time when the government did indeed face a struggle to pass its most important legislative measure. Theresa May was uniquely vulnerable among British Prime Ministers to this “free hit” effect, where her own MPs believed they were (and therefore were indeed) free to vote down all her government’s legislation without it amounting to an expression of no confidence that would require it to resign. Those who argue that the FTPA did not have any effect on the “meaningful votes” on the Brexit deal, such as in this unsophisticated piece, are often looking at the wrong question, and asking what difference it would have made if May had been able to call an early election. That wasn’t a viable political play for Mrs May: she could not have gone into an election seeking a refreshed political mandate against her own backbenches. In the hypothetical scenario of the 2019 meaningful votes without the FTPA on the statute book, defeat would simply mean Mrs May would have to resign and, with her party obviously split, either attempt to advise the Queen to appoint effectively an interim PM while a leadership contest took place (Damian Green, most likely), or the only other available figure who might have been able to form a short-term government: the Leader of the Opposition. Able to threaten uppity MPs with this scenario, would she have stood a better chance of getting her deal through? Undoubtedly.
The are other counter-factual to consider is if the FTPA had still existed, but been framed better in the first place. Suppose it had not included the “motion of no confidence” mechanism, and therefore the old rule about confidence was still understood to have been in place, and the option of calling an early election was clearly unavailable. This is a more clear-cut version of the above, and would certainly have made the implications of the votes clear in terms of confidence in the government. Suppose then, under either scenario, that the deal got passed: this would have avoided the 2019 European Parliament elections at which the Conservatives were relegated to fifth in the popular vote, and which spooked them into turning to Boris Johnson. And if Boris Johnson’s government had not been in office when the COVID-19 pandemic struck, who knows how much more effective the UK’s response might have been. It’s tempting to wander too far down these counter-factual rabbit holes, but while it’s a rather silly and futile exercise to do so at great length, that’s not the same as saying they are not there: the role the FTPA played in dictating how events unfolded in 2019 was a substantial one, and things would very likely have been very different either in its absence or if it had been better drafted.
In closing, I want to reflect on the zeal or even mania for constitutional reform in some quarters, and in the Liberal Democrats in particular. It was of course the formation of the coalition government in 2010 that foisted the FTPA on us, and it was the second of two Lib Dem “wins” that gave rise to Brexit. The first of those was Paddy Ashdown’s one, scant crumb from Tony Blair after the mid-90s promise of closer working between the two parties was washed away in Labour’s 1997 landslide: Blair agreed to using proportional representation in elections to the European Parliament. This gave UKIP the platform it needed to rise to prominence and eventually attract enough electoral success to spook the Conservatives into offering an in-out referendum on EU membership. If ever you need a warning about where piecemeal, short-sighted constitutional tinkering can lead, just remind yourself of the Liberal Democrats’ role as midwives of Brexit.
But suppose we were to attempt constitutional change properly, looking at the constitution in the round, using a proper constitutional convention over a period of years? And suppose that exercise resulted in a codified constitution? Reflecting on the saga of the FTPA, I cannot feel optimistic that our political class, through their systemic misunderstanding of constitutional matters, would not balls it up comprehensively. And if that happens, undoing the new settlement will be much harder than repealing the Fixed Term Parliaments Act.