# Predicting the Election

In the modern era of politics, we are lucky to have features such as routine consistent polling and the advent of an election calculator / pendulum.

In 2007, Kevin Rudd’s Australian Labor won the election with 52.7% of the Two Party Preferred (2PP) vote.  The exact same figure John Howard’s Liberal Party retained by in 2004.  This was a swing of 5.4% against the incumbents.

It saw Labor gain 23 electoral seats though a uniform swing of 5.4% only implied a gain of 21 seats.  It did however correctly predict 17 of those 23 seats.  Four of the predicted seats were retained by the Liberal National Coalition (LNP), all of the still considered fairly marginal seats.  The seven seats that were not predicted by the electoral calculator suffered much greater swings than expected but every single one of them can also be considered marginal.  This means their is a greater chance of them going to an opposing party at election time.

This result was also very close to Newspoll’s polling near the election date of 53-47 in favour of Labor.  The overall polling throughout the election year was always in favour of Labor and their numbers varied anywhere from 54  to 59 on a 2PP basis which is still above the end result of 52.7 or 53.  Electoral calculators will never be accurate but they are a useful assessment tool.

Rudd Labor has consistently been polling around 56 or 57% and the excessive predicted margin recorded by polls in the last election, this end result is unlikely.  It is not however beyond the realms of possibility however as Harold Holt won the 1966 election with 57% of the vote.

Using data from 1949 since the advent of the Liberal Party as we know it today, the average 2PP of the winning party is 52.1%.

As most polls claim a margin of error of +/-3%.  This would make Labor’s polling more like 53 or 54% on the 2PP which is much closer to the last election result on a 2PP basis.

This gives us about three scenarios for predicting the number of seats at the next election, a 2PP of 52.1, 52.7 (53) or 56%.

If the 2010 Federal election was concluded on a 52.1% 2PP basis, the LNP would arguably gain the following seats:

Herbert*
Gilmore
Macarthur
Macquarie
Robertson
Solomon
Swan*

This would give the LNP 66 seats, not nearly enough to form government and there is probably room for argument against Herbert and Swan.

If the 2010 Federal election was concluded on a 52.7/53% 2PP basis, there would be no net gain of seats for either side.  A possibility of two for Labor but that is highly debatable.  So for the purpose of this exercise we will use the 54% 2PP in favour of Labor which would see them arguably gain the following seats:

Bowman
Cowper*
Hughes
La Trobe
Macquarie
McEwen
Paterson
Ryan*
Stirling*
Sturt

This would see Labor have 97 electoral seats which is 21 in excess of what is required to retain government.  Three of those are quite debatable.

If the 2010 Federal election was concluded on a 56% 2PP basis, Labor would arguably gain the following seats:

Boothby*
Bowman
Cowan
Cowper
Fairfax*
Hinkler
Hughes
La Trobe
McEwen
Paterson
Ryan
Stirling
Sturt

This would see Labor have 101 seats in Federal parliament, 18 more than they currently have.  Two of which could be in doubt using the 2PP as a guide.

With the options of losing up to nine seats, or gaining 10 or 13 seats, it is near impossible for Kevin Rudd to lose the election.