Introduction
This publication provides an assessment of the political impact for the
final determined New South Wales electoral boundaries, proclaimed on 18
September 2013. The final boundaries contained only minor adjustments to the
draft boundaries released on 17 June 2013. The new boundaries will apply from
the 2015 election.
It must be stressed that this publication does not provide predictions for
the 2015 election. The estimated margins provided in this publication are best
thought of as possible results of the 2011 election had that election been
fought on the new boundaries. The estimates cannot take into account the
difference in candidate and campaign effort that would have occurred if the
2011 election had been fought on the new boundaries. The difficulty of
splitting polling place catchment areas to match new electoral boundaries, as
well as allocating declaration votes to new electorates, means the estimated
margins should be treated as approximations.
Calculations and Methodology
To calculate the political impact of the new boundaries, the results of the
2011 election have been re-assigned to match the new electoral boundaries. This
requires a series of assumptions to be made as part of the calculations. These
assumptions concern whether votes are cast for candidates or parties, how to
transfer polling places between old and new electorates, and how to deal with
the declaration vote from old electorates.
Candidate versus Party Vote
An assumption has been made that votes are cast for parties rather than
candidates. This is a reasonable assumption, but ignores the importance of
popular sitting members. A well-known sitting member may poll higher than his
or her party would have with a different candidate. New electoral boundaries
can see areas added to an electorate where the party vote was not influenced at
the last election by the presence of the sitting member.
The role of sitting Independent members is even more difficult to take into
account. None of the Independents elected at the 2011 election contested the
areas added to their electorates by the redistribution, making it difficult to
calculate precise two-candidate preferred margins.
Transferring Booths to Match the New Boundaries
The Electoral Districts Commission released lists of all transfers of
enrolment between old and new electorates. The transfer of polling place
results from old to new electorates has been guided by the enrolment transfers
as well as the published maps. Where the transfers and maps suggest the
catchment area of a polling place should be divided, an estimate has been made
to divide the results of a polling place between new electorates.
Dealing with the Declaration Vote
For calculation purposes, all declaration votes (pre-poll, postal, etc) and
votes cast at Sydney Town Hall (except those in the Sydney electorate) have
been accumulated into a single total and transferred to new electorates.
These accumulated totals have been transferred to new electorates in
proportion to the transfer of enrolment from old to new electorates. However,
the vote by party in the declaration vote has been weighted to reflect the vote
by party in polling places transferred between electorates.
As an example, the old electorate of Marrickville has been abolished and
divided between the new electorates of Newtown and Summer Hill. The polling
places in those parts of Marrickville transferred to Summer Hill voted more
strongly for the Labor Party in 2011 than polling places transferred to the new
electorate of Newtown. As a result, in dividing the declaration vote between
Summer Hill and Newtown, the patterns of polling place results is reflected in
the declaration vote transferred. Summer Hill has been allocated a higher
proportion of Labor vote and lower proportion of Green vote from the divided
declaration vote than has been transferred to Newtown.
Variations from Draft Boundaries
Compared to the draft boundaries, changes were minor and confined to changes
in five areas between the following electorates.
· Goulburn and Monaro
· Blue Mountains and Hawkesbury
· Kiama and South Coast
· Oxley and Myall Lakes
· Charlestown, Lake Macquarie, Newcastle and Wallsend.
None of these changes had significant impact on estimated margins.
By-elections and Changes in Party Membership
Calculations in this publication are based on 2011 election results. They do
not use the results of the five by-elections held since the 2011 election in
the National-held electorate of Clarence (November 2011), Labor-held Heffron
(August 2012), Independent held Sydney (October 2012) and Northern Tablelands
(May 2013), and Liberal-held Miranda (October 2013).
The Northern Tablelands and Miranda by-elections resulted in changes in
party status. The calculations for both seats are based on the 2011 results,
but the margins shown in the electoral pendulum on page 5 have been calculated
in the following manner.
The Northern Tablelands by-election was won by the National Party. The
electoral pendulum has been adjusted to list the seat as National rather than
Independent held. The margin shown is a National two-party preferred margin
versus Labor based on the results of the 2011 state election.
The Miranda by-election resulted in Labor gaining the seat from the Liberal
Party. The Labor margin based on the by-election was 5.1%, but the
redistribution calculations indicate that Miranda is 2% stronger for the
Liberal Party on the new boundaries. Applying this difference to the
by-election result, Miranda is shown as being Labor held with a margin of 3.1%.
Two-Party Preferred versus Two-Candidate Preferred Results
The two-candidate preferred count for an electorate is defined as the
final count of votes between the two remaining candidates after the exclusion
and distribution as preferences of votes for all other candidates. At each
stage of the distribution of preferences, the candidate with the fewest votes
is excluded and his or her ballot papers re-examined to determine the candidate
with the next available preference. To achieve victory, a candidate must
receive more than 50% of the total two-candidate preferred vote. Where a
candidate achieves more than 50% of the vote before the final distribution, the
two-candidate preferred count is continued to completion for information
purposes.
The two-party preferred count represents a distribution of
preferences between candidates representing the Labor Party and the Coalition.
Candidates are again excluded based on lowest progressive vote, though at some
point the second lowest candidate is excluded to retain candidates representing
Labor and the Coalition. The two-party preferred count plays no part in
determining the winning candidate but is used for information purposes.
In most cases, the two-party preferred result will be the same as the
two-candidate preferred result. At the 2011 election, 66 of the electorates
finished as two-party preferred contests between Labor and Coalition
candidates. A separate 2-party preferred count was carried out in the other 27
electorates, allowing a 2-party preferred margin to be calculated for the 93
new electorates. The estimated new 2-party preferred totals for are shown on
the listing for each new electorate in this publication.
Calculating new 2-candidate preferred totals was more complex. The most
difficult example was Newtown, which includes transfers from three existing
electorates, Heffron, Marrickville and Sydney. At the 2011 election, Heffron
finished as a Labor-Liberal contest, Marrickville as a Labor-Green contest, and
Sydney as a Liberal-Independent contest. The transferred first preference votes
produce a contest between Labor and the Greens. The estimated new margin has
been calculated by assuming the preferences of Clover Moore had been
distributed 25% to the Greens and 10% to Labor, and distributing all other
votes using the actual preference count for Marrickville.
In Balmain and Lake Macquarie, new two-candidate preferred totals have been
estimated by applying the 2011 preference distribution to the new first
preference totals. Other electorates, such as Davidson and Sydney, did not
produce amalgamation problems and sensible two-candidate preferred totals have
been provided. Some new electorates, Goulburn being the best example, produce
totals with separate Liberal and National entries in the two-candidate and
two-party preferred totals. These cases can be resolved by adding the Liberal
and National votes together.
Polling place two-party and two-candidate preferred results can be found in
my publication “NSW Legislative Assembly Election 2011: Two-Party
Preferred Results by Polling Place”, Background Paper No 1/2012, NSW
Parliamentary Library Research Service.
Electorate Name Changes
Of the 93 Assembly electorates, 86 retain the same name, though some of
those electorates have undergone extensive boundary changes. Seven electorate
names have disappeared, and seven new names appear or re-appear on the
state’s electoral map.
For the purposes of this publication, I have treated two electorates as
being abolished. These are the rural electorate of Murrumbidge, and I have also
treated the inner city seat of Marrickville as being abolished, with
Marrickville forming the basis of two new inner-city seats, Newtown and Summer
Hill. The Electoral Districts Commission chose to treat Summer Hill as a
successor seat to Marrickville. Both seats are listed as abolished electorates
on page 107.
The other name changes have come about to reflect changed geography.
Burrinjuck becomes Cootamundra, Menai becomes Holsworthy, Murray-Darling
becomes Murray, Smithfield is now known as Prospect, and Toongabbie is re-named
Seven Hills.
Throughout this publication the new names for electorates will generally be
used.
Acknowledgements and Disclaimer
All estimates have been calculated based on the actual results of the 2011
election and with reference to data and maps provided by the Electoral
Boundaries Commission. My thanks go to the staff the NSW Electoral Commission
and the Electoral Districts Commission. My particular thanks go to Paul Beeren
for his work on calculating the enrolment transfers between old and new
electorates.
However, responsibility for all calculations, and the assumption underlying
them, must rest with the author.
Antony Green
October 2013
