The National Forum   Donate   Your Account   On Line Opinion   Forum   Blogs   Polling   About   
The Forum - On Line Opinion's article discussion area


RSS 2.0

Main Articles General

Sign In      Register

The Forum > Article Comments > The Tasmanian Liberals must climb a mountain > Comments

The Tasmanian Liberals must climb a mountain : Comments

By Peter Tucker, published 5/10/2005

Peter Tucker shows the Tasmanian Liberals may get what they don't want: a minority government.

  1. Pages:
  2. Page 1
  3. All
Generally a good analysis but I don't agree with the comments for Bass, because rather than the lion's share of Liberal preferences going to Labor, past results show that most Liberal voters just vote 1 to 5 and their votes exhaust when there are no Liberals left in the count. About 60% of Liberal preferences exhaust at this point and the remainder only favour Labor slightly (recently around 55:45) so there is nothing doubtful about Bass on the figures given in Table 3 - it would be an easy Greens win, and the distribution of seats should be 12/8/3 and 2 undecided, while the hung parliament quotient should really be .25. Also, when the Greens are excluded from the count, about a third of their preferences exhaust and the rest favour Labor about 60:40 - if this trend held the Liberals would probably be just ahead in Denison and Franklin but these seats would be genuinely close. Throw in the potential of the Liberals to improve their position through the count by being less susceptible to leakage (as seen in Bass in 2002 when they won from behind) and the real quota for minority goverment might be more like .35, so I think the article understates the chance of a hung parliament based on the results it predicts. However I agree with the basic claims that the Liberals show no signs of being able to win outright, and that a Labor outright win is now the most probable result. Indeed, those predicting a hung parliament appear to have gone to ground in Tasmania in recent months.
Posted by Kevin Bonham, Thursday, 6 October 2005 6:16:14 PM
Find out more about this user Recommend this comment for deletion Return to top of page Return to Forum Main Page Copy comment URL to clipboard
Kevin. Thanks. I am happy to defer to those with greater knowledge about preference distribution, something which I admit to knowing little.

First, we’ll give Bass to the Greens which makes it 12/8/3 “certainties”. If the preference distribution goes how you say then the Greens, being over a quarter of a quota in front on primaries, would win.

That leaves Denison and Franklin a fight between Labor and Liberal. Accepting your feeling of the way preferences might flow, what quotient would you give the parties? This is the first time I’ve used the model and I am keen to keep trying it. It will be close in these seats, but you think the Liberals have the edge: .55 to .45 about right? If so, this gives final election quotients of .6975 for a Labor majority and .3025 for a hung parliament. If you think it is more .6 to .4. the final quotients become .64 Labor, and .36 hung (which is close to your thoughts of .35).

With a small number of seats I admit it becomes a bit of silly exercise to argue over .3 or .35. Of course none of us knows to that level of certainty. Where I believe the model has “power” is in elections where there are 10 or 20 doubtful seats (certainly not in Tasmania) where you can give each party in each seat a quotient, then see how it looks for the party outcome. One problem (for me, anyway, given my limited spreadsheet skills) is the number of combinations that have to be calculated. For example, 20 seats can give over 3.5 million combinations. I welcome any comments or criticisms, not only on my Tasmanian analysis, but also on this idea of giving election outcomes probability scores. And if anyone thinks they know how to write up a spreadsheet, or other program option, I would be delighted to here from you. Either post here or email me on

(Kevin, I enjoyed your item on the Hobart City Council elections. Please let me have your email as I have a couple of questions.)
Posted by Peter Tucker, Thursday, 6 October 2005 8:06:21 PM
Find out more about this user Recommend this comment for deletion Return to top of page Return to Forum Main Page Copy comment URL to clipboard
Yes, I guessed that the Liberals' very slight projected edge on preference distribution in Denison and Franklin plus their tendency to gain through leakage would make their chances in both those seats close to 60:40 on the projected figures, hence my suggestion of about .35 as a hung parliament quotient, but I did not attempt to analyse the chances in any detail. To do so would be difficult as there are lots of quirks and uncertainties in Tasmanian Hare-Clarke preference distributions whenever the projected results are closer than about 0.1 of a quota. The ban on how-to-vote cards makes the way votes are distributed between candidates within each party relevant - both in terms of how many of its votes each party risks losing to leakage or exhaust and also in terms of the chance of the party that is "behind" winning because the last opposing candidate gets eliminated while two or more of its candidates are between that candidate and a quota. These would be very tricky things to model.
Posted by Kevin Bonham, Wednesday, 12 October 2005 1:37:33 PM
Find out more about this user Recommend this comment for deletion Return to top of page Return to Forum Main Page Copy comment URL to clipboard
  1. Pages:
  2. Page 1
  3. All

About Us :: Search :: Discuss :: Feedback :: Legals :: Privacy