Professor David Howarth, Professor of Law and Public Policy, University of Cambridge – written evidence (FPA0007)

I am the Professor of Law and Public Policy at the University of Cambridge and a former MP for Cambridge. In the latter capacity I introduced the Fixed-term Parliaments Bill 2007, whose proposals were similar in intent to the 2011 Act, albeit very different in the detail of how they worked.


1. The Fixed-term Parliaments Act should not be repealed.

2. It should, however, be amended to remove the ‘no confidence’ route to an early election, clearly separating the issues of confidence and dissolution.

3. Further legislation is required to deal with other potential abuses of the Prime Minister’s power to ‘advise’ the Monarch, including legislation on the process of government formation, prorogation and Royal Assent.

Responses to the Committee’s Questions:

To what extent has the Act led to a meaningful transfer of power from the Prime Minister to Parliament?

Conventional wisdom had it that the Fixed-term Parliaments Act had produced no meaningful transfer of power from the Prime Minister to the House of Commons because no Opposition would ever turn down the opportunity to call a general election. Conventional wisdom was decisively refuted by the events of 4 and 9 September 2019, when the Commons twice refused to pass an early election motion by the required two-thirds majority.

The result is that it is now clear that the Act does diminish Prime Ministerial power not only over the Opposition but perhaps more importantly over Government backbenchers.

As long as an Opposition can think strategically and not be caught up in the macho rhetoric of ‘being up for a fight’ and accusations of ‘cowardice’, it will usually be in a position to stop any early election motion. Under a minority government, the Opposition will also be able to stop the dubious device of a government giving time to a motion of no confidence in itself, if necessary by moving the procedural motion that the question be not now put.

As for backbenchers the change is more subtle, but still significant. Prime Ministers could never guarantee that an election would be called whenever they wanted one, because the Monarch reserved the right to refuse a dissolution in line with the principles laid down in the Lascelles letter. But the likelihood of an election if the Prime Minister requested one under the previous arrangements was very high. That meant that, as long as Prime Ministers, in their other capacity as party leaders, could get their own way over candidate selection (some could not, of course), they could in effect threaten to deprive recalcitrant backbenchers of their seats, at least subject to the proviso that a Prime Minister a long way behind in the opinion polls could not make a particularly credible threat to call an election in the first place.

Under the Act, in contrast, those same recalcitrant backbenchers have votes in the decision whether to call an election. As long as they see some chance that the Opposition might not co-operate in calling an election, the existence of the FTPA might at least to a limited degree embolden them. It is significant that on 4 and 9 September 2019, none of the MPs deprived of the Conservative whip on 3 September voted for dissolution.

What impact has the Act had on the notion of the House of Commons having "confidence" in Government?

The Act does not affect the ability of a government to threaten to resign or to remove the whip from recalcitrant backbenchers or to ask for a dissolution if it loses a specific vote, and so it is difficult to see how the Act could have replaced the doctrine of confidence or even affected the content of the doctrine. For example, the Johnson government treated the vote in the Commons on 3 September 2019 on the Standing Order 24 motion as a matter of confidence and deprived those who rebelled of the whip. It then proceeded to ask for a dissolution by putting down an early election motion. That is all entirely in line with previous practice, except that the request for a dissolution is made to the Commons rather than to the Monarch. The only difference the Act makes is that the likelihood of an election if the Prime Minister loses a vote of confidence is lower because it is more likely that the Commons will reject an early election motion than that the Monarch would refuse a dissolution under the Lascelles letter. But that change in no way alters what counts as having or losing the confidence of the House.

The very fact that the Committee has asked the question, however, shows that some confusion has broken out. The difficulty is that many commentators seem to think that the previous situation was that after being defeated in a confidence vote the government had an untrammelled power to call an election, so that somehow the FTPA, having removed the alleged power of the government to call an election, allows the government to ignore the confidence vote. This is not the case. The Prime Minister could never call an election without asking someone else’s permission. The Act has merely changed the identity of those to be asked. The choice facing a government after the loss of a confidence vote was always and still is asking for an election or making way for a new government.

The propriety of what the Johnson government did in putting down an early election motion after losing a confidence vote is thus not in doubt. The request for a dissolution having been rejected, the government should then perhaps have resigned. But given the uncertainty over the position of the principal opposition party, it was arguably acceptable to make a second request for a dissolution on 9 September. At that point, however, when it became completely clear that the Commons would not grant a dissolution, the government should have resigned, becoming a caretaker administration while a search began to find a government that could command the confidence of the House.

As we have seen in September 2019, it is possible for a situation to arise in which the government becomes a zombie government. It has lost the confidence of the House, failed to obtain a dissolution but refuses to resign. But the problem is not one of the meaning of confidence or of the effect of the FTPA but of enforcement of the convention that governments that have lost the confidence of the House and failed to secure a dissolution must leave office.

It is crucial to understand that precisely the same position could have been reached under the Lascelles letter. A prime minister denied a dissolution under the Lascelles letter was under the same obligation to resign and the same problem would have arisen if that Prime Minister had refused to resign. All that the FTPA changes is the likelihood that a dissolution will be refused, but the zombie government comes into being in both cases because a Prime Minister refuses to resign.

The confidence convention depends on Prime Ministers acting honourably and decently. It seems, however, that those days are over. The Commons can raise the stakes – refusing supply, finding the Prime Minister in contempt, even expelling the Prime Minister from the House or reviving the process of impeachment – but ultimately the legal power to remove the Prime Minister currently rests with the Monarch alone. And even if Parliament were to pass a Bill specifically removing the Prime Minister from office, we now face threats that the Prime Minister might advise the Monarch to veto such a Bill, even though that advice would itself demonstrate that the Prime Minister lacked the confidence of the House and would arguably be unlawful on the ground the Prime Minister would be using the power to advise for an improper purpose.

If the Monarch never uses the power to dismiss the ministry but always submits to the sitting Prime Minister’s advice, even the advice of a Prime Minister who has lost the confidence of the House, the doctrine of confidence will have been destroyed. But that will have happened not because of the Fixed-term Parliaments Act but because of a deep flaw in the system, that in those circumstances it would lack ultimate guardians able to protect the constitution from the actions of dishonourable holders of the office of Prime Minister.

Even so, it was almost certainly a mistake for the Act, through section 2(4), to link its main function, namely reducing the power of the Prime Minister to secure an election, with the issue of confidence. Making that link allowed the confusion to start. It would be better to break the link and to repeal the no confidence parts of section 2.

One subsidiary issue is whether the Act somehow makes ineffective votes of no confidence using wordings other than that laid down in the Act, and thus not triggering the 14-day period and the possibility of an election. It is difficult to see how the Act can have this effect. If the House of Commons resolves, for example, ‘That This House has lost confidence in Her Majesty’s Government and requires it to resign’, it seems absurd to say that the House nevertheless has confidence in the government. But removing the statutory link between dissolution and confidence would helpfully eliminate the issue completely.

Removing the no confidence route would also eliminate the power the Act creates for an Opposition under a minority government to call an election, by winning a vote of no confidence and then either refusing to take office in the 14 days or taking office and refusing to put a motion of confidence to the House. But just as opportunism by a government is undesirable, so also should be opportunism by an opposition. Calling an early election should be a matter of broad consensus, not party advantage. The two-thirds requirement achieves that consensus, as does, in a different way, the option of a one clause ‘notwithstanding’ bill, since that route requires the consent of the House of Lords.

What obligation is the Government under to make time available in the House of Commons for a motion of no confidence?

The issues here are (i) whether the Government is under an obligation to make time in the Commons not only for motions of no confidence expressed in the terms laid down in the Act but also for motions of no confidence expressed in other terms; and (ii) how is the obligation enforced, given that technically motions of no confidence are subject to the government’s control of the order paper under Standing Order 14.

Neither issue, however, really matters now because of the Speaker’s reinterpretation of the Commons’ Standing Order 24. Under the new interpretation (foreshadowed in the new edition of Erskine May at paragraph 19.22) the Speaker can allow emergency debates on substantive motions. The result is that a motion of no confidence expressed in any terms can now be brought forward for debate in the Commons as long as it has the support of the Speaker and 40 members rising in their places.

Nevertheless, given my view that the Act has in no way changed the content of the doctrine of confidence, I would have said that the government is obliged to make time for any motion of no confidence put down by the Leader of the Opposition regardless of whether it is worded in the terms laid down in the Act. As for enforcement, a way around Standing Order 14 exists in the form of motions relating to privilege and contempt of the House.

What constitutional and legal obligations is the Prime Minister under in the 14 days following a defeat in a no confidence motion?

The Explanatory Notes to the Bill are clear that the purpose of the 14-day period is to allow the formation of an alternate government if one can be found. If such an administration potentially exists, the mechanism envisaged by the Act is that the Monarch invites its leader to become Prime Minister, at which point, under Commons Standing Order 14, that person takes control of the Commons’ order paper and, if so inclined, proposes the statutory motion of confidence. The order is important. Formation of the ministry comes first. The vote of confidence comes second.

The idea that the sitting Prime Minister might try to reverse the effect of the vote of no confidence by putting forward a vote of confidence in herself/ himself is an extension of the original idea of the Bill, not its central idea.

Consequently, the most important obligation of a defeated Prime Minister under the Act is to refrain from obstructing the formation of an alternate government. Whether the courts would enforce that obligation is not entirely clear, but it is at least potentially enforceable and exists as a constitutional obligation even if not legally enforceable. The obligation is in principle enforceable by the Monarch, although the events around the 2019 prorogation have cast serious doubts on the Monarch’s ability to enforce constitutional principles when faced by a Prime Minister who advises the Monarch to allow the government to act unconstitutionally.

The defeated Prime Minister is also a caretaker and so subject to the conventions associated with that position, namely, not to take important decisions without the consent of the opposition parties and not finally to resign until a new administration is ready to take office. The first part of the caretaker convention is enforceable to a very limited extent by the Cabinet Secretary, but, as we saw in 2010, the second part is not enforceable by anyone.

Should the Act be amended or repealed?

The Act should not be repealed. The previous situation in which Prime Ministers could in nearly all circumstances decide the timings of elections on the basis of their own political advantage was intolerable. It was as if one of the runners in a 100m race controlled the starting pistol. In addition, the Act’s effect of rebalancing power in favour of the House of Commons, including in favour of government backbenchers, is, at least from a liberal point of view, one of its biggest advantages. It attacks an excessive concentration of power. Of course, from an authoritarian, Hobbesian point of view, the rebalancing might look more like a disadvantage.

Admittedly, many commentators have found one consequence of the Act very difficult to accept, namely that it envisages changes of government without an election. Such changes are not unusual in our history, especially at times of crisis. But populist rhetoric, building on decades of assertions of various versions of the mandate theory of elections, delegitimises the idea that the Commons should decide who governs. We live at a time when the idea of representative democracy is under fundamental threat from authoritarian ideas such as ‘the will of the people’ being embodied in charismatic leaders who set themselves up in opposition to ‘the political elite’ or ‘the political class’. The Fixed-term Parliaments Act is built on the foundations of and enhances representative democracy. To repeal it would contribute to undermining those foundations and would embolden the opponents of parliamentary democracy.

But the Act is not perfect. The link it creates between allowing an early election and the doctrine of confidence is confusing and should be eliminated. The lack of detail about what happens in the 14-day period provides another reason for simply eliminating the vote of no confidence route to an early election. The dismissal and formation of governments should be the subject of separate legislation. One way to do that would be to legislate to transfer the formal power to appoint and dismiss the Prime Minister to the House of Commons. A very effective option to consider including in that legislation is the German constitution’s provision for ‘constructive’ votes of no confidence (Article 67), that any vote of no confidence in the Chancellor must specify a new Chancellor to take over.

The Act does not provide a solution for the wider problem of the collapse of what Lord Hennessy has long called the ‘good chap’ theory of the constitution. The Fixed-term Parliaments Act was designed to control the Prime Minister’s ability to abuse the power to ‘advise’ the Monarch in one a very specific respect, advice about dissolution. Dealing with potential (and actual) abuse of the ‘advice’ power in other circumstances – prorogation, Royal Assent and determining who has the confidence of the House, for example – will require a much broader and deeper reforms. It is a fair criticism of the Act that it was an example of piecemeal reform. But the right response is not to repeal it but to carry out a comprehensive reform.


September 2019