 |
Research Note no. 25 2004–05
Electoral matrix: 2004 federal election
Gerard
Newman
Statistics Section
29 November 2004
Introduction
At the 2004 federal election the Liberal/National Coalition Government
was returned with an increased majority. The Coalition now holds 87 seats
in the House of Representatives (Liberal Party 74, The Nationals 12 and
the Country Liberal Party 1)—a majority of 24 over the Labor Party (60
seats) and the three independents.The Coalition would lose its absolute
majority if 12 seats were to be lost at the next election, compared with
only 8 seats before the election. In contrast, the task for the Labor
Party has become much harder. From a situation before the election of
requiring an additional 13 seats for an absolute majority, the Labor Party
now requires an additional 16 seats to win government.
Electoral ‘pendulum’
Traditionally, seat margins have been presented in a pendulum shape,
with seats listed from the highest to the lowest swing for each seat to
change hands on either sides of the pendulum.1 The pendulum
construct implies that election swings are uniform across seats and that
seats will change hands in order of their respective two-party margins.
Unfortunately, electoral swings are not necessarily uniform across seats
and the pendulum is not a necessarily reliable predictor of election outcomes
in particular seats. At the 2004 election, there was a 1.8 per cent two-party
swing to the Coalition. Based on the pre-election pendulum, a uniform
swing of this nature would have resulted in the loss of five Labor seats
(Brisbane, Wakefield, Kingston, Stirling, and Hasluck) to the Coalition.
At the 2004 election, the Coalition won four of the above seats (not Brisbane)
and four other seats held by a margin greater than 1.8 per cent (Bonner,
Bass, Greenway and Braddon). However, the Labor Party won four seats from
the Coalition (Adelaide, Hindmarsh, Parramatta and Richmond) and one seat
from the Greens (Cunningham) for a net loss of three seats. While the
pendulum was reasonably accurate in predicting the number of seats to
change hands, it was highly inaccurate as to the actual seats that changed.An
additional problem with the pendulum style is how to account for seats
held by independents.
Electoral ‘matrix’
This Research Note attempts to overcome the uniform
swing assumption implied in the pendulum construct by presenting seat
margin data in a different format. Rather that arranging seats in descending
order of safeness within two columns, the electoral matrix arranges seats
in columns of safeness and in rows of party holding each seat.
The electoral matrix shows the two-party preferred
swing required for each party-held seat to change hands at the next election,
and the two-candidate preferred swing for the three seats held by independents.
Seats are arranged in safeness categories according to the Australia Electoral
Commission’s classification of safeness. ‘Safe’ seats require a swing
of over 10 per cent to change, ‘fairly safe’ seats require a swing of
between 6 and 10 per cent, while ‘marginal’ seats require a swing
of less than 6 per cent.
A notable feature of the matrix is the large number
of Coalition-held seats that are classified as safe. Over half of these
(45 out of 87) are now classified as safe, compared with a third of the
Coalition-held seats before the election. In contrast the number of Labor-held
seats classified as safe has dropped from 25 before the election to 19
after the election.
Both parties have an equal number of seats classified as marginal (23).
However, the proportion of Labor seats classified as marginal is much greater
(38 per cent as opposed to 26 per cent). The number of fairly safe seats
held by each party is also similar (19 for the Coalition and 18 for Labor).
The three seats held by independents are all classified as safe.
- Often referred to as the Mackerras Pendulum after its creator Malcolm
Mackerras.
Electoral matrix, 2004 federal election
For copyright reasons some linked items are only available to
members of Parliament.
|
 |
