Every schoolboy knows there will be a federal general election towards
the end of 2007. It will be preceded by state general elections in NSW
in March 2007, Victoria in November 2006 and Queensland at a time
chosen by Peter Beattie.

These are, of course, our three most
populous states. Probably all four incumbent governments will be
re-elected. However, it is worthwhile to speculate on this question:
which one of the four is the most likely to be defeated?

Let me
give some statistics from which I am making my assessment. In my
opinion the probability of defeat is directly related to the low nature
of the figure of the two-party preferred percentage at the most recent
general election.

Thus the Labor Government of Steve Bracks is
the least likely to be defeated. Labor in Victoria in November 2002
secured 57.8% of the two-party preferred vote.

The Labor
Government of Morris Iemma is the second least likely to be defeated.
Labor in New South Wales in March 2003 secured 56.2% of the
two-party preferred vote.

The Labor Government of Peter Beattie
is the third least likely to be defeated. In February 2004 Queensland
Labor secured 55.5% of the two-party preferred vote.

The
government most likely to be defeated is the Howard federal Coalition
government. In October 2004 the Coalition secured only 52.7% of
the two-party preferred vote.

Following the release of the
federal redistribution proposals for New South Wales and Queensland the
Liberal Party’s Federal Director, Brian Loughnane, has been in
propaganda mode. He has repeatedly asserted that the federal Coalition
has “a perilously small margin for error” in respect of this next
election. I agree with his assessment, more or less. However, where I
reject his line is the way in which he is blaming the redistribution
for the Coalition’s problem.

There is no serious dispute about the statistics which can be learnt by visiting my website.

My
next claim will sound boastful but, since it is true, I shall make it.
For nearly forty years I have been the analyst generating this swing
figure being the uniform swing needed to defeat a government. That
figure comes from the swing needed by the Opposition party to win the
median seat on the Mackerras Pendulum.

On my pendulum the median
seat is the one which lies exactly half way between the strongest
Coalition seat (currently Mallee in rural
Victoria)
and the
strongest Labor seat (currently Batman in suburban Melbourne). The old
median seat of Bennelong needed a swing of 4.4% to fall to
Labor. The new median seat is Eden-Monaro needing a swing of 3.3% . So
it is true that these redistributions weaken Howard’s grip on
power.

However, that does not mean they are unfair. Nor does it
mean the Liberal Party has any ground to complain. Given a Labor share
of 47.3% of the two-party preferred vote in October 2004 we can
say that an overall swing of 2.8% would make Labor the majority
party in terms of the aggregate vote. Surely it is fairer that, in
terms of seats, the uniform swing needed for a Labor government should
be 3.3% rather than 4.4%.

So my recommendation
to Brian Loughnane is this: when asserting that the Coalition has “a
perilously small margin for error” I suggest you should always refer to
the low Coalition share of the two-party preferred vote. You know the
figure is 52.7%. Therefore you should say that.

Peter Fray

A lot can happen in 3 months.

3 months is a long time in 2020. Join us to make sense of it all.

Get you first 12 weeks of Crikey for just $12. Cancel anytime.

Peter Fray
Editor-in-chief of Crikey

12 weeks for $12