Four years, more or less, is the standard term of office for a parliament. To go up to the full five years, as in 1996-97, 1991-92, 1978-79 and 1963-64, is usually because the government’s position is too weak to risk it when the four-year span comes up. Every now and then, however, the electorate is summoned to the polls before four years are up, and these cases cast some light on the feverish weeks in which a November 2007 general election seemed at first a possibility, then a probability, and then a very silly idea indeed.
An indecisive result in the first election is the usual reason for a short parliamentary term. This was the case in the most recent short term – the parliament of February to October 1974. The February election resulted in a hung parliament with a precarious 301-297 Labour lead in seats, and insufficient Liberals to install a majority government with either main party. The 1974 parliament could not last and Harold Wilson ran a minority government always with an eye on another election. Some obvious problems were solved, such as the miners’ strike, and some popular policy options, such as devolution, were brought forward, but the difficult decisions over matters like inflation and Europe were put off for another day. The ‘Wilson two-step’ was successful, but only just, as Labour won the October election with an overall majority of three and a sufficient margin over the Conservatives to sustain the parliament into a disastrous fifth year in 1979.
There were two cases in which small majorities in one election led to subsequent elections after a parliament of only a year and a half in duration – in March 1966 and October 1951. In both cases, a Labour government with a single-figure majority called an early election to end the draining experience of parliamentary deadlock, but the circumstances were otherwise quite different. In 1966 Harold Wilson timed it to perfection and gained a massively increased majority. Government from 1964-66 had been run as a confidence-building, agenda-setting exercise and the economy was stimulated to provide a rise in real wages at the election. It was in 1966, rather than 1964, that Labour most used the ‘Thirteen Wasted Years’ slogan against the Tories, contrasting this with the ‘Go-Ahead Year’ of Labour. In 1951, by contrast, the election was an expression of sheer exhaustion on the part of Clement Attlee and his ministers, and was widely expected to result in a considerable Conservative victory. In the end, Labour actually got more votes and the Conservative margin in seats was itself rather narrow – a majority of only 17.
It is of course not at all necessary or even customary for a new prime minister to ask for an immediate ‘mandate’ to govern. Britain’s system of government is parliamentary, and there is no real discontinuity when a party that has been elected with one leader continues in office under a new leader. Since 1900 there have been seven occasions (1902, 1908, 1937, 1957, 1963, 1976 and 1990) when a change in prime minister has not led to a general election. The only apparent instances of this were in 1935 and 1955. In the former, Baldwin called an election a few months after taking office but the normal four-year span was up by that time anyway. In 1955 Sir Anthony Eden went directly to the polls. But even in that case the Conservatives were three-and-a-half years into a parliament where they had a small majority, and were confident of winning.
Even when there is a discontinuity in the nature of the government, there is not always a general election. On three occasions there have been big changes in wartime (1915, 1916 and 1940) when an election would not have been appropriate or even possible. An election did happen in 1922, when the Lloyd George coalition was dissolved and an election called as soon as Andrew Bonar Law formed a Conservative government, but this parliament was nearing its end anyway. An early election followed the formation of the coalition government in 1931, but only after a delay of a few weeks and much uncertainty. There is probably a reasonable expectation that such big changes in peacetime should be ratified in a general election.
The record of governments calling early elections because of avoidable political situations is not a happy one, and should have given Labour caution before election speculation ran out of control in September 2007. In 1923, after a change in prime minister and a radical shift in policy in favour of protective tariffs, the Conservative government called an election after only a year in office. The Tories lost a comfortable majority and ended up the largest party in a hung parliament, only to be ejected from power in 1924 and replaced by a minority Labour government. That government, in turn, lost an unnecessary vote of confidence later in 1924 and lost power – although the 1924 election saw Labour achieve the strategic objective of wiping out the Liberals as a competitor for power.
The next rushed election was in June 1970, although this was actually after a full four-year term. Harold Wilson was misled into calling the poll after a dramatic – but, as it proved, shallow – Labour recovery in the polls after a nightmarish period of unpopularity. Most people had previously anticipated an October 1970 election, to allow more time to elapse following the disasters of 1967-69. Wilson lost.
Perhaps the nearest comparison to 2007 was the sudden build-up to an election in February 1974. Edward Heath’s government had stumbled into a confrontation with the miners. The idea of having a ‘Who Governs?’ election to strengthen the Conservatives’ mandate was at first favoured only by the most hawkish Tories but won increasing party and media support in the winter of 1973-74. Had the original idea of an election in early February gone ahead they might have won, but that date was allowed to slip and by the time of the actual election at the end of February, the Tories had lost control of the agenda and lost the election.
The record, therefore, of rushing into an election is not a terribly good one. In 1923 and February 1974, governments hoped to capitalise on a divided and less than credible opposition, only for that opposition to pull together and fight a surprisingly effective campaign. In February 1974 and 1970, sudden recoveries in the opinion polls did not last through to election day, and in October 1974 the rewards ended up being very disappointing. In February 1950, the government had an election in a winter month, shortly after an economic crisis, out of sheer rectitude – a later election would have made the more relaxed 1950 budget look like a piece of electioneering. Labour’s austere chancellor Stafford Cripps could not allow that, and the penalty for doing the right thing was an unworkably small majority rather than the comfortable margin that the government arguably deserved. However, in 1955 and 1966 the early election gambit worked perfectly.
Is it wise, therefore, to resist a clamour for an election? Not inevitably, as the classic case of Jim Callaghan’s refusal to call an election in October 1978 indicates. There are some grounds for thinking that John Major might have got a better majority in early 1991 than he did in April 1992, and as a result had a less rocky ride than he did. But in general, despite the embarrassment, it is probably safer to slink away and fight another day than be provoked into a reckless dissolution of parliament and the sacrifice of a couple of years of power.