Reading the pendulum

At the last Victorian State Election in November 2010 the Coalition secured a narrow majority in the Legislative Assembly, winning 45 seats to the ALP’s 43.

The first electoral redistribution for 12 years in 2012-2013 has delivered a very different electoral landscape or pendulum. Some electorates were abolished, several others were radically redrawn or given new names.

While two Government-held electorates, Doncaster and Rodney, were abolished, the overall changes superficially appeared to favour the Coalition, as the redistribution turned five marginal ALP electorates into marginal Liberal seats. So, if you look at the electoral pendulum going into this month’s election there are 48 seats on the Coalition side of the pendulum (including independent-held Frankston) to Labor’s 40.

However, with Labor incumbents re-contesting four of the five notionally Liberal marginals, and with one of their marginal MPs having become an independent, the Coalition will miss out on the normal benefits of incumbency.

History shows that for local members incumbency is often at its most potent at the end of their first term as an MP, a phenomenon dubbed the ‘sophomore surge’ effect. With so many Labor incumbents on their side of the pendulum the Liberals desperately need a sophomore surge in seats like Bentleigh, Mordialloc and Carrum so their first term MPs hold on.

These three seats, plus the electorate of Frankston, all have margins of under 2 per cent, making the Frankston train line corridor a decisive battleground again in 2014, just as it was in 2010. Other key areas will be Geelong and Ballarat.

The fact that Labor can get to 44 seats just by getting its incumbents re-elected, and to 45 by picking up Frankston from Geoff Shaw, must have Coalition strategists concerned.

When the swing’s not on

Other factors come into play when there is not a strong across the board swing likely to generate a uniform outcome at an election. The following is as much ‘ínsider’ conjecture as it is science, but worth a thought nevertheless.

The benefit of incumbency

Much argued about by political aficionados, it is generally accepted that incumbency bestows some electoral advantage to a sitting MP at election time. Name and personal recognition, established community networks, understanding of local issues and greater resources are all thought to generate some degree of personal vote for the incumbent member.

A noted factor in the USA’s political system, there is a great deal of conjecture – and little research – as to the quantifiable benefit of incumbency in Australian politics.

The few studies available, combined with conventional wisdom, suggest incumbency may be worth around 3% of the vote, on average, though longer serving members and those in regional seats may have more incumbent value than a first-term, city electorate MP.
As an exercise and in theory, this means that in an open race, where the sitting MP has retired, a fairer reflection of the actual party voting status is calculated by subtracting 3% from the incumbent party’s vote.

So applied in terms of open races in marginal seats in Victoria, we see:


incumbency pendulum

The donkey vote

Just as with the benefit of incumbency, the value of the donkey vote has long been debated. A donkey vote is a ballot cast in a preferential voting election, where the voter ranks candidates in the order they appear on the ballot paper: first preference for the first-listed candidate, second preference for the second-listed candidate, etc. As far back as 1968, Malcolm Mackerras suggested it was worth between 1 and 3%. More recent estimates place it as 1 in every 15 votes cast, or just over 1%.

The Victorian Electoral Commission will oversee the process to decide the order of each local ballot on November 14.

In terms of the donkey vote, the issue isn’t just who has it at the upcoming 2014 election. We must look to see who had it last time at the 2010 election.

If the same party or candidate had the donkey vote last election as this election, then it’s already factored into their two-party preferred (2PP) vote and so they gain no advantage – status quo remains.

However, if a party or candidate didn’t have the advantage of the donkey vote at the last election and get it this time around, then it’s a potential gain. On the flip side, if a party or candidate had the donkey vote at the last election and loses it to their main opponent at this election, they are at a disadvantage.

Revisit the following table after the draw of the ballot on November 14 to assess the implications of who got, held, lost the donkey vote in each of the marginal seats.

donkey vote pendulum

A word on the polls

At the last Victorian State Election the Coalition won 51.6% of the two-party preferred vote. The latest Newspoll has them on 46% which represents a 5.6% swing.

If this were the result on 29 November, and there were a uniform swing, the pendulum indicates that the ALP would pick up 12 seats on the Coalition side of the pendulum taking them to a total of 52 and leaving the Coalition with 36.

Still, a lot can change in three and a half weeks.

Call of the polling card

call of voting card


What the bookies think

Going into the election the overall odds on a Coalition win varied between 3.50 and 4.75 depending which betting agency website you looked at. What was more interesting was the individual seat markets which Sportsbet helpfully provides.

Given the polls, the ALP unsurprisingly started the campaign as the favourite in all 40 seats on its side of the pendulum. Interestingly, the most vulnerable according to the odds, is the sixth-most marginal, Macedon, where the margin is 2.3% and Labor’s odds are 1.70. The next longest price for Labor, in one of its own seats, is 1.40 in both its most marginal seat, Eltham, and its tenth most marginal, Geelong.

The ALP also starts as favourite in seven seats on the Coalition side of the pendulum, plus as equal favourite in another. Thus, if these market prove correct, Labor will finish with a small but comfortable majority, holding 47 or 48 seats.

The Coalition-held seats are generally predicted to fall in pendulum order, with the exception of Ripon, which despite its margin of just 1.6%, remains a 3.30 chance for Labor. Another Liberal seat where the odds defy the pendulum is Prahran, where a Labor win is at 4.50, much longer than the odds in several other seats where the margin is much greater than Prahran’s 4.7%. There are two other seats with a margin below the Newspoll swing of 5.6% where Labor is odds against – Forest Hill (2.10) and South Barwon (2.30).