Most political commentary and punditry on elections take as their starting point the notion that every election is winnable by either major party. John Black's recent article in online opinion is written with this premise. Unfortunately history and statistics do not support this. Since Whitlam's dismissal in 1975, the party in federal government has only changed twice. It has only been "drovers dog" elections that have seen an electoral change in federal government. The states have greater rates of change in government than the federal government - so what makes federal government so stagnant?

Poll : Preferred Churn Rate for Federal Government

Stagnant Government

Since World War II, the party in government has changed five times. This does not include when John McEwan held government as part of the Coalition for a month in 1967. As this was contiguous with the coalition holding power. Five times in fifty years is a low turn over. In Australian Politics and Government which was edited by Jeremy Moon and Campbell Sharman, a table in the appendix only records four changes in government being due to elections. This is since 1901. Of all the elections that have been held, only four have led to a change in government. Just four, in forty one elections .

The states are no better;

  • NSW has had five changes in government since 1939.
  • Queensland has had five changes in government since 1942.
  • South Australia has had six changes in government since 1938.
  • Tasmania has had six changes in government since 1937.
  • Victoria has had thirteen changes in government since 1945.
  • Western Australia has had eight changes in government since 1945.

Victoria and South Australia both had minority governments for many years which lead to great volatility. For Victoria this period was between 1945 and 1958. Once the minority government period was over, the process fell back to approximately ten year for a party to be in power.

If we compare the states to the federal government in changes of parties in power since 1975.

  • The Commonwealth has had two changes in government.
  • NSW has had four changes in government.
  • Queensland has had five changes in government.
  • South Australia has had three changes in government.
  • Tasmania has had four changes in government.
  • Victoria has had three changes in government.
  • Western Australia has had three changes in government.
  • ACT has had four changes in government.
  • Northern Territory has had one change in government.

It becomes obvious, that other than the Northern Territory, the rate of change in federal government is less than the states. The question is; Why is incumbency such a big advantage at the federal level.

We should compare the number of elections held since 1975 at the federal and state levels.

  • 12 Commonwealth elections
  • 10 Queensland elections
  • 9 NSW elections
  • 9 Victorian elections
  • 9 South Australian elections
  • 9 Tasmanian elections
  • 8 Western Australian elections
  • 8 Northern Territory elections.
  • 6 ACT elections.

Finally the electoral changes in government as a percentage of the number of elections. Rate of change (churn rate) since 1975 in increasing order;

  • 13% Northern Territory
  • 17% Commonwealth
  • 33% Victoria
  • 33% South Australia
  • 37% Western Australia
  • 44% NSW
  • 44% Tasmania
  • 50% Queensland
  • 67% ACT

The Northern Territory has the lowest rate of change, with the federal government a close second. The most competitive elections are the Australian Capital Territory's, with Queensland coming in second.

Incumbency

All the Australian forms of government model the Westminster system of responsible government, and its inevitable embedding of the executive arm of government in the legislative. A hack from the sixteenth century to route around the executive power of the monarch. But since this is comparing the federal government to the state governments, this is not an advantage unique to the federal form of government.

Nor can the bicameral or unicameral forms of government been seen as forms of differentiation. The Northern Territory which has the lowest rate of change since 1975 is unicameral, but so is the ACT which has the highest rate of change. The Federal governments low rate of change is in a bicameral system that has had a third party holding the balance of power in the upper house.

So where does the federal government differ from the states? Unfortunately I don't know enough to speculate.

I know that calling early elections gives a huge advantage to the incumbent, this also allows the government to control the initial narrative of the election. Other incumbency advantages include controlling the public purse and manipulating government departments and information for electoral gain. But again, the states also do this, but maybe not to the same dramatic effect as the federal government.

Churn, Churn, Churn

Either way, the churn rate at the Federal level is too low . Fixed term elections need to be introduced, as well as term limits on the Prime Minister position. The term limit on the Prime Minister needs to become constitutional legislation, rather than statutory, so a future Prime Minister cannot remove it when it becomes inconvenient.

The risk is we may fall into the same entropy of the British system, where the advantage is so huge that a Prime Minster remains in power for almost a generation. Australia has had one such Prime Minister, and it is not healthy for democracy. Term limits save politicians from themselves. They are a necessary limitation in government.

cam
More reading: Tags
Cam Riley: South Sea Republic. Freedom, liberty, equity and an Australian Republic.

Comments
  • Alan . # .
    New Zealand and Canada: I agree the churn rate is too low, especially at the federal level. I\'d be interested to look at changes (if any) in the churn rate where states (NSW is an example) shifted from elections at the premier\'s discretion to fixed terms and at comparisons with New Zealand and Canada.

    I suspect the problem is as much the comparative resources available to government and opposition for research and policy development as the floating term.
  • siento . # .
    The churn rate is fine and careful number choices: By picking a year when the government changed and not including that changed you reduce the number of changes. Pick 1974 as your start date and you get one change of government every 10 years or so.

    Also, the problems of introducing term limits are considerable. Too much churn and you get Italy or France in the 20 or so years after WWII.

    Or alternatively you get the Californian problem where you have endlessly inexperienced members who are easily manipulated by experienced lobbyists.

    Incumbency is an advantage because you have proved you can do the job. People tend to choose the devil they know.

    Menzies was in too long. His government did out stay it\'s welcome and what we wound up with was the inexperienced, incompetent, sloganeering Whitlam government that almost bankrupted the country but other than that Australia\'s rate of change of government is good. Provided the government is competent they are given 5 years to get in and bring in the changes they want and then another 5 or so to oversee the implementation of those policies.

    What throws a spanner in the works is incompetent opposition. Drover\'s Dog\'s elections also need to be won, just ask John Hewson.

     
  • cam . # .
    Limits and Elections:
    Pick 1974 as your start date and you get one change of government every 10 years or so.

    Yeh but include Whitlam and then Menzies has to be included. Both those skew the data, one for length, one for brevity. I chose 1975, as post-dismissal politics are the modern manner. It also gives a thirty year span of data.

    Also, the problems of introducing term limits are considerable.

    The maximum a party has stayed in power in the ACT is six years. The PM is the one that needs to be term limited. Six years is enough to have an effect and the next one in the party take over.

    Virginia in the US term limits the Governor to one term, four years. The previous Governor to Warner, fellow by the name of Gilmore, did all manner of repugnant tax cutting and borrowing. He didnt have to face a vote over it.

    That would be the downside, the upside would be that we would have seen Keating and Costello much earlier than we are at the federal level. Another upside is that Carr, who has worn out his welcome, would have retired gracefully and new blood injected into the state parliament.

    I dont see a limit of two terms as affecting government. If anything it protects government. California was rife with incompetence and by other legislators that didnt want to do work. In my opinion Parliamentarians need to be term limited to a generation so the geriatrics like Thurman and Byrd in the US stay in Congress till they die.

    Twenty four years is long enough. In that period they can become specialists, without become entrenched. That would protect against incompetence through turn over.

    Drover\'s Dog\'s elections also need to be won, just ask John Hewson.

    By the same token, Whitlam was also re-elected. I think Hewson\'s example also points to the over-riding advantage that the incumbency gets. The 1996 election was definately a drovers dog election.

    If Hewson was a good politician, he would have just held on to the position of leader of the opposition until the next election. If Latham wants to be PM he just keep hanging on and on. Statistically elections are against the Liberal Party from now on.

    cam
  • cam . # .
    Comparing Westminsters: Yes, the next step would be to collate which westminster systems have fixed term elections and any sort of term limits on the PM. It would be interesting to see if they have any effect on data. I still reckon fixed term elections and term limiting are the right thing to do.

    cam
  • Alan . # .
    Early elections and term limits: Fixed term elections are an excellent idea. My favourite variant is Sweden which combines a fixed term with a rule that allows the government to call as many extra elections as it wants but only gives an extra election Riksdag the same term as the Riksdag it replaces. That would neatly cut most early election ideas off at the knees.

    I don\'t think term limits are an especially good idea, except for the executive. Even executive limits should only apply to immediate re-election.

    Clinton\'s made the point that US presidents now leave office at an age where once they would have been readt for retirement. Instead you now have people with long vigorous term-limited exclusions from office for life.

    We already have a severe imbalance between the bureaucracy and the parliament in terms of experience and information. Rotating MPs out of office will only exacerbate that.