1 October 2007
Brent from Mumble
know by heart, the ALP is sixteen seats short of a House of Representatives
majority, and you can plot those sixteen most marginal government-held seats
along the pendulum.
Of course, it doesn't really work that way, and no-one expects the seats to fall
in the order they appear on the Mackerras horseshoe. But how might they fall?
I've made my own list of the first seats to fall, or "the seats to
I've started with current margins and performed a few twiddles to reflect my
version of reality. For example, everyone would agree that Malcolm Turnbull's on
paper 2.5 percent margin overstates his vulnerability because of Peter King's
2004 candidacy. So I gave Malcolm a couple of percent breathing space.
Seats in which members are retiring lose a percent to account for loss of
personal vote. However, not Makin, because the retiring member Trish Draper's
rorting problems contributed to the 2004 pro-Labor swing, which was the biggest
in South Australia. That is, Makin's margin is already deflated.
Seats that changed hands in 2004 get an extra percent to account for the
personal vote of the new incumbent, and if it was a sitting member they defeated
they get another point because the other party has lost that old personal vote.
Another exception: Labor's Julie Owens gets the one percent for her own personal
vote, but not for the loss of defeated member Ross Cameron. Like Draper, his
naughtiness produced much of the 2004 swing.
And so on. I factor in a greater than average swing to Labor in Queensland
because they all love Kevin Rudd up there, and a lower one in Western Australia
because they're rolling in money. (Am I generalising?)
Unlike most observers, I reckon the classic mortgage belt demographic will swing
to Labor less than average, rather than more, so I've factored in that, plus a
bunch of other vague assumptions, to come up with this list of seats to watch.
[click seat for Antony Green at ABC]
has seventeen seats instead of sixteen, because manipulations on the WA seat of
Cowan - including the retirement of a popular sitting member - moved it to the
government side of the ledger.
The order of my list is very different to the pendulum's, and missing from it
are Bennelong (NSW), Dobell (NSW) and Hasluck (WA). That is, I don't think they
will be in the first 16 to fall.
In their place are three regional seats Corangamite, in Victoria and the NSW
North Coast National seats of Cowper and Page
You might like to try this at home. After the election we can see how close ours
the next day, the following appeared in Crikey "letters' section:
Mackerras writes: Re.
"Which 16 seats should we be watching?" (yesterday, item 7). I refer
to Peter Brent's item in Crikey yesterday. As he notes he has excluded Bennelong,
Dobell and Hasluck from his list of the first 16 Coalition seats to fall. Since
Peter has invited us to vary that list let me give my alternative.
is well known I am predicting Labor wins in all of Bennelong, Dobell and Hasluck.
Therefore, I would need to exclude three other seats in Peter's list. The seats
I would exclude are Cowper, Stirling and Wentworth. Actually I expect Stirling
to go to Labor, but I think it is a less likely Labor win than Hasluck."
Strictly speaking, Malcolm is non-sequiturising a little, as predicting wins in
those seats does not necessitate their inclusion in the top 16. (Eg he may be
anticipating a big Labor win. [Update: that
I must admit that the inclusion of Cowper is a stretch. Ok, it shouldn't be
there, and I had to try hard to push it into the top 16. It used to be my
nomination for the dark horse "ohmigod" seat that goes if Labor
wins big time, but these days it is generally recognised as being in play in a
I took Cowper out, I'd put Hasluck in.]