Both Denmark and the UK experienced a period of election frenzy among the media in recent weeks and Canada may be headed for (yet another) early election. In all three countries, the prime minister have the power to dissolve parliament and call a general election at nearly any time which begs the question if this is a good thing for the political process or if these countries would be better off with fixed election dates as in the US, Norway and, for all practical purposes, Sweden. The present Canadian government under Stephen Harper has presented such a proposal for amending the Canadian constitution.
Exactly why the election frenzies broke out in Denmark and the UK and what the rumours were based on is a little hard to say: Gordon Brown had taken over as PM after Tony Blair in June and the media speculation was that Brown would seek his “own mandate” at a time when the Conservative Party was performing badly in opinion polls.Historical evidence would suggest otherwise. Since 1945, Macmillan, Douglas-Home, Callaghan and Major became UK prime ministers without seeking an immediate electoral mandate – the one exception to the rule was the ill-fated Anthony Eden who called an election immediately after taking office in 1955, 3 1/2 years after the 1951 election.
In Canada, the election called (and lost) by John Turner in 1984 is the only case of an incoming prime minister seeking an electoral mandate while the closest Danish case would be the November 1960 election called by Viggo Kampmann who had taken over as prime minister after H.C. Hansen’s death in February. Still, elections were due within a year in both of these cases.
For the moment I will leave aside the interesting question about the mandate and formal theories of parliamentary representation and look at under what circumstances elections have been called in these three countries since the introduction of general suffrage and if we can say that prime ministers have abused the right to dissolve parliament.
In this document (pdf-format) I have listed elections for the Canadian House of Commons from 1922 onward, the UK House of Commons from 1923 onward and for the Danish Folketing from 1920 onward. The first column shows the year and month of the election, the second column the motivation given for the election, the third column the status of the outgoing government and the fourth column the length of the term of the dissolved parliament. The fifth column contains notes about some of the elections.
Two technical notes: 1) For the sake of speed, I have generally relied on data from the English version of Wikipedia (which in general is not recommendable for this kind of work) and 2) I have had to make informed guesses about the motivation for the elections. We should also note that Canada and the UK applies the first-past-the-post system while Denmark introduced PR for the Folketing in 1918.
Canada: Alternating between Stability and Unstability
If we turn to the Canadian case first, we see a pattern of short-lived parliaments alternating with series of longer-lived parliaments. The parliaments elected in 1925, 1957, 1962 and 1979 all lasted for less than a year while parliaments elected in 1972 and 2004 lasted less than two years. Finally, the parliaments elected in 1963 and 1965 lasted less than three years. This points to a very simple rule in Canadian politics: No parliament without a clear one-party majority lasts longer than three years (Two years and ten months to be precise) – and conversely: No parliament with a clear one-party majority has lasted for shorter than three years (Three years and three months).
Another interesting pattern is that between 1918 and 1993 only three parliaments with clear one-party majorities have lasted for shorter than four years – but the parliaments elected in 1993, 1997 and 2000 all lasted for less than four years despite giving the Liberal Party working majorities. Something is clearly going on in Canadian politics (Hint: The fragmentation and realignment of the conservative parties since the early 1990s).
UK: Moving Towards Stability
When we include the inter-war era in an analysis of the UK House of Commons, we note that elections yielding “hung” parliaments or very slim majorities have not been that unusual up until 1979. The parliaments elected in 1923, 1929, 1950, 1964, 1974 (I) and 1974 (II) all failed to yield a clear working majority – and sure enough: With the exception of the House elected in October 1974 none survived for more than a little over two years. The 1970s are usually described as a chaotic period in British politics which makes the longevity of the 1974(II) House all the more spectacular.
Parliaments with a clear working majority have on the other hand generally lasted between four and five years. Here the exception is the parliament elected in 1922 which only lasted for a little over a year as incoming Prime Minister Stanley Baldwin decided to call an election over trade policy. He lost and was forced to resign following a vote of no confidence in the House of Commons, opening the way for the first Labour government. The hung 1923 House only lasted for another ten months.Some other exceptions to the four-year rule are the Houses elected in 1918, 1951, 1966 and 2001 which all lasted a little under four years. The 1955 and 1970 elections were cases of a prime minister deliberately seeking a new mandate – in 1955 Anthony Eden succeeded (only to be destroyed by the Suez Crisis) while Harold Wilson lost his bet in 1970.
A tentative conclusion is that Gordon Brown would have broken if not the rules, then the patterns of British parliamentarism by calling an election for October or November: Incoming Prime Ministers do not call elections to seek a fresh mandate and calling an election after only two years and six months would also have been highly unusual.
Still, we should remember that UK politics has been exceptionally stable on the parliamentary level since 1974 (sic!): Parliaments and governments have consistently sat for long terms compared with the period between 1918 and 1974. And Gordon Brown might want to contemplate the Baldwin’s fate as well as Wilson’s (1970) and Heath’s (1974) failed attempts to schedule elections.
Denmark: Minority Governments and Minority Governments
The Danish case differs from the British and Canadian as minority governments have been the rule in Danish politics since 1913. Only the periods between 1929 and 1940 and again between 1957 and 1964 have seen prolonged periods of formal majority rule.
As a rule, the length of the parliamentary terms in Denmark are linked with the respective governments’ parliamentary basis. If the government has a firm parliamentary base, it – and parliament – will sit for between three and four years. If the government doesn’t have a firm parliamentary base, it will sit for less than three years before calling an election.There is one interesting exception to this rule as Thorvald Stauning controlled a stable and working majority of Social Democrats and Social Liberals in the Folketing between 1929 and 1940 but called early elections in 1932 and again in 1935. Stauning used these elections as a weapon against the majority of Liberals and Conservatives in the upper house of parliament which blocked crisis and reform legislation: The 1932 and 1935 elections were in effect referendums on the government’s general policies – “referendums” which the government won with a clear majority.
After 1945, prime ministers have used the dissolution weapon a number of times in order to gain a stronger position on the parliamentary arena: 1950, 1966, 1975, 1977 and 1990 may serve as cases of this kind of elections. These cases also suggest that the dissolution weapon can be a double-edged sword even for experienced political leaders. 1966, 1975 and 1990 definitively saw Jens Otto Krag’s, Poul Hartling’s and Poul Schlüters respective strategies backfire.
Since 1990 Danish politics has settled – as in the UK “settled” does not mean an absence of change on the electoral and party levels, on the contrary – and parliamentary terms have gotten longer although the 2005 election could be said to have been called unusually early; only three years and three months after the last election.
As in the UK, a September election in Denmark would have been a break with the patterns in electoral politics that have developed. It would have come only two years and seven months after the previous election – an interval otherwise indicative of an unstable parliamentary situation. If previous behaviour was something to go by, we should expect the next Danish election to come somewhere between three years and three months and three years and eight months after the 2005 election, i.e somewhere between May 2008 and November 2008. September 2008 would be my immediate guess.
One argument which has been presented to support the suggestion that the Danish Prime Minister will call an election soon is that it can be used preemptively to secure the government’s position before the ratification of – and possible referendum about – the new EU Treaty and the series of wage negotiations which will take place in the spring of 2008. As far as I can see there is no tradition for using parliamentary elections preemptively in the way suggested by political reporters. This is not to say that Anders Fogh Rasmussen will not break the game and take such an initiative, just that if he chooses to call a preemptive election, it will be an innovation in Danish politics.
Abuse of the Powers of Dissolution?
As far as I can see, there has been no general tendency towards an abuse of the powers of dissolution of parliament from the prime ministers in the three countries, but there are some periods and patterns that are worth noting.
First, Denmark lived through a prolonged period of parliamentary and electoral instability from 1971. We may argue if the 1984 or the 1990 election marked the end of instability and the introduction of a new equlibrium in Danish politics but in terms of general trust in democracy, electoral turn-out, stability of government and policy performance the negative consequences do not seem that obvious.
Second, on the parliamentary level UK politics has been surprisingly stable since 1974 (again: This is not a typo – I do mean nineteen hundred and seventy-four). We should note that commentators have suggested that James Callaghan waited too long in calling an election – had he chosen to call an election in mid-1978, it would have been within the patterns of parliamentary tradition and he might have been able to return a safe Labour majority in the House of Commons. Instead of a 17-year Conservative hegemony, UK politics might have been a more competitive affair in the 1980s and 1990s.
Finally, Canadian politics seems to be going through a period of marked change starting with the demolition of the Progressive Conservative Party in 1993. Viewed in the short term, there is nothing unusual about the parliamentary instability following the 2004 and 2006 elections, but in a longer perspective the short parliaments of 1993, 1997 and 2000 point to a change in (Liberal) prime ministers’ behaviour but in my opinion it is still be to jump to conclusions to speak of an abuse of the power of dissolution.