As the general election approaches, there is considerable interest in the Fixed-term Parliaments Act 2011. Unfortunately, not all of the discussion to date has been legally accurate. For example, Professor Colin Talbot has written that:
[T]here’s been big speculation about what happens if a minority Labour or Conservative government gets defeated on their Queen’s Speech. Which rather misses the point – nothing happens. The Government doesn’t fall. There is no dissolution of Parliament. Under the FTPA the only circumstances in which a Government falls would be if (a) they resigned – unlikely but not impossible or (b) the following is passed by a majority in the House of Commons: “That this House has no confidence in Her Majesty’s Government.” Nothing else forces a Government out of office – not defeat on a Queens Speech, a Budget, a key piece of legislation, a vote of no confidence in the Prime Minister, nothing. Of course there may be political circumstances where a Government chooses to take such a defeat seriously and resigns. Or, more likely, they could put down the above motion themselves and dare their opponents to kick them out.
There are two difficulties with this analysis. First, it conflates two distinct questions. One concerns the circumstances in which Parliament is dissolved so as to trigger a general election; the other concerns who forms the Government. Second, this conflation is then compounded by a failure to distinguish between the operation of constitutional law — such as the 2011 Act — and constitutional convention.
The 2011 Act regulates the circumstances in which general elections occur. It is not directly concerned with, although it may incidentally influence, which party or parties form the Government. Questions pertaining to government formation remain the preserve of constitutional convention; the Cabinet Secretary’s understanding of the relevant conventions is recorded in chapter 2 of the Cabinet Manual.
The Act provides that elections take place every five years, and that an early election — i.e. one that would occur before the end of any given five-year period — can occur only in the circumstances specified in the Act: that is, (a) if a motion in favour of an early election is supported by MPs accounting for at least two-thirds of the total number of seats in the Commons, or (b) if a motion (passed on a simple majority of MPs voting) that “this House has no confidence in Her Majesty’s Government” is not followed within 14 days by a further motion (passed on a simple majority) that “this House has confidence in Her Majesty’s Government”. Talbot suggests that nothing other than a motion passed under the Act that provided that the Commons “has no confidence in” the Government can bring a Government down. But this is wrong on two counts.
First, such a motion does not bring down the Government. Rather, it starts the clock ticking such that if there is no subsequent motion of confidence with 14 days, an election will occur. If such circumstances eventuate, then the absence of any other viable Government means that the incumbent remains in office throughout the election-campaign period, ceasing to be the Government only if and when a new viable Government is formed after the election. It follows that a no-confidence motion under the Act is not the only way of bringing down a Government because it is not, in the first place, a way of doing that at all, albeit that such a motion may trigger distinct, political consequences — such as an adverse outcome at an early election or the formation of a different viable Government within the 14-day period — that would result in the original Government’s fall.
Second, however, there are (other) circumstances in which a Government can be brought down that have nothing to do with the Act. As Lord Norton has explained, there are several ways in which an absence of confidence in the Government can be manifested. The Act does not purport to regulate or change this. It does not define what a motion of no confidence is, and it does not provide that a motion of no confidence will always trigger the 14-day period. Rather, it provides that a specifically worded motion that expresses an absence of confidence will have that effect. It leaves open the possibility of an absence of confidence in the Government being manifested in other ways, including by means of a differently worded motion of no confidence. Such non-statutory no-confidence motions continue to produce effects that sound in constitutional convention rather than constitutional law. In particular, convention continues to require the resignation of a Government in which an absence of confidence is expressed. The difference under the Act is that a non-statutory no-confidence motion only requires resignation, and can no longer trigger an early election — constitutional law, in the form of the Act, having displaced the convention that used to facilitate the dissolution of Parliament whenever no confidence was expressed in the Government.
It follows from all of this that an accurate analysis of the implications of the Fixed-term Parliaments Act demands a proper understanding of the distinctions between constitutional law and constitutional convention (and of the ways in which they interact), and between the legal and conventional rules governing the holding of elections and dissolution of Parliament on the one hand, and the formation and resignation of Governments on the other.