Written evidence submitted by Vernon Bogdanor, Professor of Government, King’s College, London (FTP0001)



          The government is right to propose repeal of the Fixed Term Parliaments Act for three reasons. The first is that an early dissolution is more often desirable than undesirable. So repeal does not, as is sometimes said, increase the power of government. Instead, it increases the power of the electorate to decide on whether it wants a particular government to continue in power. The second reason is that the Act lacks clarity as to what is to happen during the 14 days after a successful no confidence vote. The third is that that the Act requires a vote of no confidence to be a specific vote. But it makes for better government if there is provision to attach confidence to an important substantive matter, rather than requiring a specific vote.

           This note discusses these three weaknesses, the first in much greater detail than the other two. In the fourth part it suggests that the government’s proposal does not, as it suggests, return the position to the status quo ante since dissolution in the past was not a matter for advice by the Prime Minister but for a request to the Queen, who could, in theory refuse it. A decision by the Queen cannot, in my view, be justiciable in the courts.


          The fundamental purpose behind the Fixed-term Parliaments Act was to take away the alleged advantage held by an incumbent Prime Minister to choose the date of dissolution. Of the 18 general elections in Britain since the war before the Act, it can be argued that there are five where Prime Ministers benefited from being able to choose an early date for dissolution – the general election of 1959, called by Harold Macmillan as Prime Minister, the general elections of 1983 and 1987, called by Margaret Thatcher, and the general elections of 2001 and 2005, called by Tony Blair. On the first of these occasions, it was alleged that Macmillan had taken advantage of a pre-election boom, which could not be sustained. In 1970, Harold Wilson sought to do the same, but was unexpectedly defeated. It does not of course follow that, in the other cases, the government of the day would have been defeated if the parliament had been allowed to run its full term. Indeed, it could be argued that the Macmillan, Thatcher and Blair governments would have won elections whenever they had been held. Perhaps, therefore, the advantage which a Prime Minister can gain from an early dissolution has been exaggerated. It would in any case be difficult to regard the outcome as in any sense undemocratic simply because the people were asked to pronounce on their government one year earlier than expected. Even so, a coalition government could always guard against a snap election by the sort of measure contained in the 2010 coalition agreement. It does not need the cumbrous legislation which is the Fixed-Term Parliaments Act.

           An early dissolution, it was argued, by supporters of the Act, represented an illegitimate attempt by the Prime Minister to gain an advantage from the temporary popularity of his government. But an early dissolution was not always such an illegitimate attempt. There are a number of perfectly respectable reasons for an early dissolution. The first is when a new Prime Minister seeks a personal mandate. That was Sir Anthony Eden’s motive in going to the country in 1955 in a Parliament which still had 18 months to run. (Critics alleged that he too was seeking to take advantage of a temporary economic boom).  Admittedly, Eden’s precedent has not been followed, even though five of the other 12 changes of Prime Minister since the war have occurred between rather than after general elections. Neither Harold Macmillan in 1957, Sir Alec Douglas Home in 1963, James Callaghan in 1976 nor John Major in 1990 considered going to the country immediately after their appointment, and there was, perhaps, little public demand for them to do so. But in 2007, there was some public feeling that Gordon Brown should go to the country to seek a personal mandate. An Ipsos-MORI poll on 10th October 2007, after Brown announced that he would not go to the country, found that 47% believed he was right not to call an election, while 42% felt that he was wrong. Perhaps Brown’s position would have been stronger if he had in fact gone to the country in 2007. The opposition certainly took up the cry that a new Prime Minister needed a new mandate, and in April 2010, shortly before the general election, David Cameron proposed that a new Prime Minister should go to the country within six months of being appointed. [1]  This was perhaps ironic since the Fixed-Term Parliaments bill which he was to propose as Prime Minister would have made it difficult, if not impossible, for a new Prime Minister to follow this course. But perhaps Cameron’s position in April 2010 was more in accordance with public feeling than the Fixed-Term Parliaments Act. Perhaps indeed there is greater public pressure now than in the past for a general election following a change of Prime Minister because, partly due to the role of the media, general elections seem to have become more presidential than they were. Politics has become more personalised. In May 2009, speaking during the expenses crisis, Cameron, although declaring himself sympathetic to fixed term parliaments, accepted that there were `strong political and moral arguments’ against them – `Political – because there’s nothing worse than a lame-duck government with a tiny majority limping on for years. And moral – because when a Prime Minister has gone into an election, and won it promising to serve a full term, but hands over to an unelected leader half-way through, the people deserve an election as soon as possible’.

          The average length of a parliament since the Great Reform Act of 1832 has been 3 years and 8 months. Very few parliaments have lasted for five years, the most recent being the 1959-64, 1987-92, 1992-7, 2005-10 parliaments, and the 2010-15 parliament after the Act was passed. In those cases, there was pressure for the voters for an earlier general election, not for parliament to sit out its full term.

          There would be an even stronger argument for a government to go to the country early were there to be a second change of prime minister within a single parliament. That has occurred only once in the 20th century, in 1940, in the unusual circumstances of war, when Winston Churchill replaced Neville Chamberlain who had in turn replaced Stanley Baldwin in 1937. It might possibly have happened after 2007 when there was some pressure to replace Gordon Brown, who appeared so unpopular that some Labour MPs believed that they would have a better chance of winning the election with another leader. But these critics faced the argument that it would be constitutionally illegitimate for a second unelected prime minister to take office within the same parliament, and that the government would therefore be morally obliged to go to the country fairly rapidly at a time when opinion poll evidence indicated that it would suffer a heavy election defeat. It is by no means clear that a constitutional change designed to obviate the need for a government to go to the country if it twice replaced its prime minister was in the public interest.

          A second reason for an early dissolution might be to seek a mandate for a new policy. That was Asquith’s reason for seeking two dissolutions in 1910. The first was to secure a mandate in a general election in January 1910 for Lloyd George’s `People’s Budget’, which had been rejected by the House of Lords; the second was to secure a mandate in a general election in December 1910 for the Parliament Act, limiting the powers of the Lords. Similarly, in 1923, Stanley Baldwin sought a dissolution in a Parliament that was just one year old. He had come to the view that unemployment could be cured only by a protective tariff, but felt bound by the pledge made by his predecessor, Bonar Law, in 1922, that a Conservative government would not introduce a tariff.

        In September 1931, the National Government, although it had a majority in Parliament of 309-250, on what was in effect a vote of confidence, decided to seek a dissolution to secure a `doctor’s mandate’ for its policies, and also, no doubt, approval for the formation of the coalition. In February 1974, Edward Heath sought an early dissolution sixteen months before the end of the Parliament to obtain a mandate to deal with the miners’ strike and the changed economic conditions resulting from the oil crisis which followed the 1973 Yom Kippur war. Under the Fixed-Term Parliaments Act, the Labour opposition might well have opposed a general election in these circumstances since Labour feared that an issue fought on the issue of government versus the unions would yield a Conservative landslide. In 2017, Theresa May was able, despite the Act, to secure an early dissolution to secure a mandate for her negotiating position with the EU, and in 2019 Boris Johnson was able, despite the Act, to secure an early dissolution to secure a mandate for the Withdrawal Agreement providing for Britain’s exit from the EU.

          A third reason for an early dissolution might be that the existing Parliament is unviable. That was the reason why Attlee went to the country in 1951 in a Parliament that was just 18 months old. In the preceding general election of 1950, the Labour government had been returned with a majority of only five, and the government had been further weakened when two Cabinet ministers, Aneurin Bevan and Harold Wilson, resigned in 1951 because they could not support the budget. Attlee no doubt took the view that the country had to make up its mind whether it wanted Labour to continue or to put the Conservatives in power. Harold Wilson went to the country in 1966 for a similar reason in a Parliament that was just 18 months old. Labour had been returned in the general election of 1964 with a majority of 4, reduced to 2 by a by-election defeat. Similarly, in October 1974, Wilson went to the country in a Parliament that was just 7 months old, since he no doubt felt that a minority government could not take the difficult economic decisions necessary to combat inflation. Under the Fixed-Term Parliaments Act, it is possible, but, by no means certain, that Attlee and Wilson would have secured opposition support for dissolutions in these circumstances. If so, the Act makes no difference; if not, the Act would prove harmful. But it would be difficult for the Act to resolve a situation such as that of 2019 in which the government was unpopular in the Commons but popular in the country. It was only through a stroke of good fortune, or perhaps bad judgment by the opposition parties that Johnson was enabled to secure his dissolution.

         There might also be a fourth motive for an early dissolution, which has not yet occurred in Britain, but could do so were we, in consequence of the development of a multi-party system, to move to a situation of regular coalition or minority governments. This fourth motive would be to validate a change of coalition partners. Suppose that, in the middle of the 2010-2015 Parliament, the Liberal Democrats had decided to desert the Conservatives, and to form a coalition with Labour. It might be argued that the electorate should be given the chance to pronounce on this change of government, and that it would be deprived of its rights if it had no say on the formation of the second coalition in the 2010 Parliament which had not been endorsed by the voters.

          Something of this sort happened in Germany in 1982, following the constructive vote of no confidence. In 1972, when the Christian Democrats had sought to remove the Social Democrat government of Willy Brandt, the leader of the Free Democrats, who were in coalition with the Social Democrats, Walter Scheel, declared, `Today we are faced with an attempt to alter the political majority without allowing the electorate to participate. Whether or not this is technically legitimate, it is an act which strikes at the nerve of our democracy’. [2]  In 1982, when the Free Democrats sought themselves `to alter the political majority’ by switching from Helmut Schmidt’s Social Democrats, to Helmut Kohl’s Christian Democrats, these words were quoted with glee by the Social Democrats. The new Chancellor, Helmut Kohl, accepted that he ought to seek electoral endorsement for the new coalition. But since  the German constitution makes no provision for a government, even a majority government, to call a dissolution, Kohl, like Brandt in 1972, had to engineer the defeat of his government in a vote of confidence. This manipulation was to occur on a third occasion, in 2005, when Chancellor, Gerhard Schroeder also engineered a vote of no confidence in his government in order to go the country at what he believed was a favourable moment. Unlike Brandt and Kohl, however, Schroeder miscalculated, and lost the ensuing election. The Fixed-Term Parliaments Act did not, of course, prevent a government from similarly seeking to manipulate the provisions for a confidence vote to secure an early dissolution. But it is surely unwise to encourage a government to engage in such unseemly tactics to secure something which is in the public good. The Act, therefore, allowed a lame duck government to continue in office if it did not wish to manipulate the constitution by securing a vote of no-confidence in itself, or if, perhaps because it feared defeat at the polls, the opposition in a hung parliament or a parliament with a narrow government majority did not wish to vote for an election.

          The four motives for early dissolution outlined above are by no means unworthy. Indeed, some might argue that it is desirable that early dissolution should occur in these circumstances. Few, for example, suggested in 1951, 1966 or October 1974 that it would be better to continue with lame-duck parliaments rather than go to the country. It is possible, then, to argue that the value of an early dissolution in any of these four circumstances outweighs the possible disadvantage of a prime minister securing an early dissolution for purposes of party advantage. Voters are perfectly capable of making up their own minds on whether the reasons given by a government for an early dissolution are satisfactory.

          Advocates of the Fixed-term Parliaments Act claimed that it would give MPs more power, the power to dissolve Parliament, a power which had become in effect the prerogative of the Prime Minister. Sir George Young, Leader of the House, told Conservative Home Comment on 14th May, `the mechanism for a no confidence vote in the government is unchanged but what our proposals would do is give parliament a new power to dissolve itself, a power currently only exercised by the prime minister’.  Parliament could control Government on the matter of dissolution, instead of the Government controlling Parliament. But that was perhaps hardly the case since, of course, back-bench MPs normally obey a party whip. The supposed power of `parliament’ to dissolve itself was the power of the party leaders to decide to dissolve Parliament and whip their followers accordingly. The Act altered the conditions under which party leaders can seek a dissolution, but hardly gave more power to backbench MPs nor did it noticeably strengthen Parliament.

           The Fixed-Term Parliaments bill was introduced as a reaction to the expenses scandal. It was one of a wide range of constitutional reforms proposed in 2009, many of them having little obvious relevance to the issues raised by the scandal. The Fixed-Term Parliaments bill seems a particularly perverse reaction. For the main problem that appears to have been raised by the expenses scandal was the need to secure greater control over MPs who, so it was alleged, had become insensitive to the reactions of their constituents. If the Fixed-Term Parliaments bill had any effect, however, it would be to insulate MPs from popular pressures by ensuring that dissolution became more difficult and general elections less frequent.

          The debate about the Fixed-term Parliaments Act indicates a conflict between two fundamental principles, the principle of parliamentary government and the principle of democratic government. The former principle provides that parliament shall choose the government, which is accountable to it, the second that the people should choose the government, and that government should be accountable to them. Normally, of course, under single-party majority government, the norm since 1945, though not before 1939, the two principles coincide. But if we do come to move into an era of multi-party politics, as is possible, then the principles will come to diverge. Under multi-party politics, the fact that a government enjoys the support of parliament does not necessarily mean that it is acting in accordance with democratic principles if, for example, it forms a coalition after an election which the voters have no chance to endorse, or if it swaps coalition partners in the middle of a parliament without securing popular endorsement for this change.

           If we are entering a world of hung parliaments, it by no means necessarily follows that dissolutions should be made more difficult. Indeed, it can be argued that they should become more frequent since changes of prime minister or changes of coalition partner are more likely to occur in a multi-party system than under a two-party system where coalitions are unnecessary. While a democracy cannot survive on an endless diet of dissolutions, making dissolution too difficult could lead to endless parliamentary manoeuvring of the sort which so discredited the Third and Fourth Republics in France. Dissolution, then, is not necessarily a threat to good parliamentary government, but can instead be one of its most important safeguards by ensuring that governments are accountable not only to parliament but also to the people. If that is so, then it is difficult to see what useful purpose was served by the Fixed -term Parliaments Act. With multi-party politics and coalition government, dissolution could be an essential weapon for the voters enabling them to prevent  inter-party manoeuvring and coalition deals organised behind closed doors after the votes have been counted. In Third Republic France, in 1902, Prime Minister, Waldeck-Rousseau declared, `the ability to dissolve --- is not a menace to universal suffrage, but its safeguard. It is the essential counterbalance to excessive parliamentarism, and for this reason it affirms the democratic character of our institutions.’


The second weakness of the Act is that it is left unclear what is to happen in the 14 days after a successful no confidence vote. Three possibilities suggest themselves.

  1. The Prime Minister could resign, and the leadership process could be concertinaed so as to yield a new leader within 14 days. The Prime Minister then recommends to the Palace the new leader of his or her party, as leader of the largest party in the Commons.
  2. The Prime Minister could resign and recommend to the Queen as her next Prime Minister, the Leader of the Opposition, as mover of the no confidence motion. If, as is likely, the former Leader of the Opposition, now the Prime Minister, could not survive a no confidence vote, he would lead the country into an election as Prime Minister.
  3. The Cabinet Manual (which I helped to draft!) declares –para. 14 – that, after a successful no confidence vote,  `a Government or Prime Minister --- is required by constitutional convention to resign’, and in para. 16, the Prime Minister `is expected to tender the Government’s resignation immediately’. But a Prime Minister could ignore the convention and seek to play out time by remaining in office for 14 days after which an election would be due. Or, perhaps he or she might seek to circumvent the 14 days by seeking an immediate election through the two-thirds mechanism, assuming that the official opposition would support this. The position is very unclear.

Whichever of these three possibilities occurred, the Queen could easily be put in an embarrassing position.


The perhaps unintended consequence of the requirement that the vote of confidence must be a specific one is to restrict a Prime Minister’s options for dealing with a political crisis. It means that a vote on a substantive issue can no longer be turned into a matter of confidence by the government. In 1972, Edward Heath declared the Second Reading of the European Communities bill to be a matter of confidence and told MPs that if it were defeated, he would seek a dissolution. Second Reading was passed by eight votes. Potential rebels had been brought into line. This possibility no longer existed following passage of the Act. Had the Act not been on the statute book, Theresa May might have been able to resolve the parliamentary deadlock in 2019. She could have made the Withdrawal Agreement a matter of confidence. Then, either the rebels would have come to heel, or she would have sought a dissolution. As it was, the Commons refused to endorse the deal and it also refused to vote no confidence in the government. This led to the position predicted in my book `The Coalition and the Constitution’ (2011) whereby a government was able to continue in office but without being able to secure its major policy.

         The government is, however, mistaken to assume that, before the Act, dissolution was a matter for Prime Ministerial advice. It was not. It was a request to the sovereign and it could have been refused. In his memoirs, Harold Macmillan insists that in 1959 he `asked’ for a dissolution, although of course it was hardly likely that one would be refused in a parliament which had only a few months to the end of its legal term. [3] It could not have been the case that any government could secure a dissolution whenever it wanted, even if it was in a minority in the House of Commons. Suppose for example that, after the 2017 election, Theresa May as Prime Minister had sought an immediate second dissolution. That would clearly have been refused by the Queen. Suppose that Harold Wilson had been defeated on the Queen’s Speech in March 1974 in a minority parliament, and had sought a second dissolution. It is not obvious that he would have been granted one just a month after a previous general election. It was, admittedly, unlikely that an alternative government could have been formed in that parliament in those circumstances since, immediately after the February election, Edward Heath as Prime Minister had sought a coalition with the Liberals but had been rebuffed. Had he not done so, however, there is a real question as to whether a dissolution would or should have been granted.

          In the past, a no confidence vote would lead to a dissolution only if there was no viable alternative government within Parliament. Otherwise, the Prime Minister would resign, and the alternative government would take office. There have been three defeats in votes of no-confidence in the 20th century. The first was in January 1924, when Stanley Baldwin was defeated six weeks after losing his overall majority in the election, on the King’s Speech. In that situation, it would clearly have been improper for Baldwin to seek a second dissolution, and he resigned.  George V then appointed Ramsay MacDonald as Prime Minister of a minority government. In March 1979 when James Callaghan’s minority government was defeated in a vote of confidence, in a parliament four and a half years old, he sought a dissolution which was granted. It was then obvious that an election was necessary, and in any case the existing parliament would reach its maximum term in another six months. The precedent of October 1924 is of greater interest. For, after MacDonald’s minority government was defeated in the Commons on what he chose to regard as a vote of confidence, George V, before granting what would be the third dissolution in two years, inquired, through his Private Secretary, of the two opposition leaders, Baldwin and Asquith, whether they were prepared to form an alternative government. Only after receiving a negative answer and it was clear that no alternative government was available did George V agree to a dissolution.  He told MacDonald `that no other Party could form a Government that could last. He would protect himself by sending me a memorandum saying that he granted the election with great reluctance, and hinted that I might say so’.[4]

          The sovereign has not refused a dissolution in modern times, but that may well be because no Prime Minister has illegitimately sought one – or it may be that a Prime Minister has refrained from seeking a dissolution after making informal inquiries to the Palace and being told that he would not be granted one. In the parliamentary systems of the Commonwealth, Governor Generals have refused requests for dissolution with mixed results. In 1926, the Governor General of Canada, mistakenly believing that there was an alternative government available within the existing parliament, refused a dissolution to a minority government with unfortunate results, since he then had to grant one to the opposition leader, who had failed to form a viable government. But in 1939 the Governor General of South Africa refused a dissolution to General Herzog’s government and appointed General Smuts as Prime Minister. In that case, admittedly, Herzog lacked a majority in the Cabinet as well as in parliament. Smuts, upon being appointed Prime Minister, was able to command a majority in the South African parliament and lead the country into war.

          Because dissolution is a decision by the Queen and not a matter of advice, I doubt if it is justiciable. Therefore the status quo ante offers a better protection for a government anxious to avoid any involvement by the courts than the ouster clause proposed in the bill, since the courts are unwilling to allow any unfettered executive discretion and are perfectly capable of ignoring an ouster clause.


Vernon Bogdanor,

December 2020






[1] Independent, 24 April, 2010.

[2] Quoted in Vernon Bogdanor, ed, Coalition Government in Western Europe, Heinemann, 1983, p. 275.

[3] Harold Macmillan, Riding the Storm 1956-1959, Macmillan 1971, p.750.

[4] From MacDonald’s diary, quoted in David Marquand, Ramsay MacDonald, Jonathan Cape f1977, pp. 377-8.