Comments
I have a question that might sound pedantic.
Is the term “leakage” misleading? As I understand its use, it appears to presume that a primary vote is a vote for a party rather than an individual. If we presume that Tasmanians are voting for individuals then we can’t really speak of leakage (away from parties). Well, we can use the term as a description of an apparent phenomenon but it (to me at least) almost sounds like some unintentional subversion of democracy.
What do the experts say about how people mark their boxes in the Hare-Clark system? My own anecdotal evidence suggests that most people don’t know how the electoral system works here.
Are some pundits making early, misleading predictions because they presume that people will generally vote 1-5 for a party?
Posted by stan robert on 31/03/06 at 02:13 AMFor those interested in very detailed analysis of whether running six candidates cost Labor the final seat, see my comments at http://www.pollbludger.com/350.
My answer to this on much further thought is a very firm “no” - there is no evidence that loyal Labor voters vote 1-5 within the ticket then stop in any significant numbers.
What Labor does need to address is the need to push star third candidates up to significant primary vote totals rather than allowing their votes to scatter.
Posted by Dr Kevin Bonham on 31/03/06 at 02:51 AMCongratulations to Kevin for being near the money in Bass while so many others floundered.
However, while I very much enjoyed Kevin’s excellent analysis I think he skips over an important issue by selectively quoting Greens election predictions from the past.
For example, it is important to remember how many times the media wrote of Di Hollister’s chances when she came through in the end. Or as recently as the South Australia poll how little attention was given to Greens Upper House candidate Mark Parnell who now looks set to win a seat.
And remember the Cunningham by-election where Michael Organ became the first minor Party MP elected to the House of Representatives despite little analysis saying he could win. If anything, the Greens undersold their chances there.
And I think Christine Milne should be congratulated not criticised for her attempt to be positive after the 1998 Tasmanian election when the major parties colluded to rort democracy by changing the system. The one prediction I remember her making was “we’ll be back”. Of course she was proved right with 4 Greens being elected at the subsequent election against the predictions of most pundits.
Yes the Greens are always positive about their chances of winning seats. Just like the Tasmanian Liberal Opposition was positive about winning Government when in reality we all knew they couldn’t. This is accepted and understandable practice. If you stand for a seat you must say you are aiming to win even if the chances are slight - why vote for someone who says they have lost already?
Some pundits have chosen to criticise Bob Brown for saying the Greens were aiming to win a Senate seat in each state at the 2004 federal election (which would have resulted in a total of 8 Greens Senators). The Greens were indeed ‘aiming’ to win, but they were done over by Democrats and Labor who preferenced Family First above the Greens. This cost seats in SA and Vic and the party just missed out in QLD and NSW. Despite these narrow losses (if they had gone the other way it would have resulted in 8 Greens Senators), many have tried to paint it as a big mistake that the Greens ‘oversold’ their chances.
In fact it is the Greens duty to try and win more seats and votes than they currently hold or have earned. It is the media and pundits role to get the predictions accurate.
Again and again there are pundits wanting to write the Greens off (although not Kevin or pollbludger). While I can accept this from political rivals it is not acceptable from supposedly neutral observers.
Posted by Ben Oquist on 31/03/06 at 05:16 AM“many have tried to paint it as a big mistake that the Greens ‘oversold’ their chances.”
Over selling yourself leads to great disappointment when the results show that it was hype. In fact over selling and under delivering leads to people thinking you have failed when in fact that might not be the case.
Wouldnt it be far better to under sell and over deliver?
Posted by to ben on 31/03/06 at 06:41 AMSenator Abetz must have changed his mind somewhat to state that the Liberals should be the only legitimate opposition. They are, in fact, the only legitimate political force. They are not in opposition; they are merely exercising power through their proxies, the ALP.
As the party of capital, their ideology will always prevail, as long as we have the current economic system, and I can’t foresee any viable and feasible alternative economic system.
When the Greens learn to operate the system in the interests of capital, then, and only then, might they be allowed to play the game by becoming the second proxy Liberal Party.
Posted by Justa Bloke on 31/03/06 at 06:51 AMWell put together, Kevin. But what is the effect of the ‘exhaust’ vote? For instance, Booth led Reissig by about 400 votes. The last candidate ‘cut’ was David Fry, with over 2000 votes.
How many of his Liberal supporters shot themselves in the foot by not marking their ballot outside of the Liberal team?
I’d expect most of them would prefer Labor (Reissig) to Greens (Booth), but by allowing their vote to exhaust, they may have achieved the latter by default.
I haven’t seen the numbers, but hope that you may have put some thought into this.
The above example could be replicated by considering whether Cassy O’Connor’s supporters might have preferred Julie Collins to Graeme Sturgess, but failed to vote past the Greens candidates. If so, they voted for Sturgess.
And for ‘justabloke’, capitalism is OK if you take the monopoly out of it. There is another economic system - on the shelf, and waiting to go. See my website http://www.justicetheaim.net
Posted by Leo Foley on 31/03/06 at 09:47 AMStan - any time a candidate is excluded, most of their votes (usually 80-96%) will go to a candidate from the same party if there is one remaining. That suggests that most voters vote entirely or mostly for parties foremost and candidates within those parties second, so a vote that bucks that trend is seen as “leaking” from that party. However the word actually covers a range of things - people who direct their vote all over the ballot paper, people who vote for someone they know ahead of their preferred party, people who vote mostly for candidates from one party but leave out some of them, etc.
Experts are aware of leakage as an issue but sometimes underestimate it. Less experienced pundits, including many journalists, sometimes just assume votes are likely to stay within parties, when in fact it is possible to predict that certain candidates will do better off leakage than others.
Ben - you are absolutely right about Di Hollister and I distantly recall at least twice correctly determining that she was more likely than not to win in cases where she had been written off. Whether this is something the media are especially prone to do to the Greens though, I don’t know. I recall exactly the same thing happening with the late Robert Bell. I scrutineered for him at a Senate election at which he was being written off by journalists, but my calculations always showed him ahead, and he was indeed elected.
My comments about talking up one’s chances above related specifically to such bravado on counting night, rather than in advance of the poll. I think Rene Hidding was the main offender this year. I don’t have much to add to what others have said about the issue of public optimism in advance of a poll, except that if I was leading a minor party and wanted to claim that party was on track to win six seats, I would want to do so on something a hell of a lot more robust than 200-vote electorate samples. Reading too much into too little like that can make you look like a bit of a tryhard and I think Peg Putt should have been more circumspect. Something like “We are very encouraged by this poll but we realise it is based on a small sample and a lot can change between now and the election. We are going to work very hard between now and the election to try to make this level of support a reality and try to build it even higher.”
Leo - I don’t have a copy of the full cutup yet, and in any case it will be difficult to determine how many of the exhausting votes were diehard Liberals who had voted 1-5 then stopped (as opposed to other voters whose votes happened to have gone through Fry, Gutwein and Napier), but I strongly suspect you are right in the Bass case. In past elections once the last Liberal is out with only Labor and Green candidates standing, around 60% of Liberal votes have exhausted. Had those voters indeed kept going I suspect they would have heavily favoured Labor, especially in a northern electorate, and especially if they were voting for David Fry.
Where the last Green candidate goes out it is different. Green supporters are generally more clued-in about Hare-Clark than Labor or Liberal voters and so less of their votes (around a quarter to a third) exhaust. The gaps between Denison Labor candidates were large enough that I don’t believe exhaust from the Greens would have made any difference - and also Green voters stopping at 5 are more likely to be making a conscious and deliberate decision to do so rather than being uneducated about how to make their vote more powerful.
Posted by Dr Kevin Bonham on 01/04/06 at 02:34 AMCongratulations to Dr Kev on his perspicacious and thoughtful coverage and analysis of the election.
He’s made the other pundits look like amateurs and added a lot of value to Tasmanian Times.
Posted by Fred Baker on 01/04/06 at 03:00 AMLeo, I never said that capitalism wasn’t OK, and I don’t see why you have to take the monopoly out of it, as long as I’ve got the monopoly. Look at AWB. That’s a monopoly and it did brilliantly, helping out Saddam Hussein and still having lots left over.
The point is that capitalism is the system we live with and under capitalism capital always has power over mere people, whoever they are and however they vote.
A political party that favours people over profits will never be effective at running an economy whose aim is the opposite. Voters instinctively know this, and only the idealistic, altruistic, compassionate and thoughtful will vote against the interests of the system that serves them. They will always be a minority.
As little as people matter, the ecology of their physical environment matters even less, until the planet runs out of resources. Before then, huge profits can be reaped, and extraordinary lengths gone to in order to reap them.
Posted by Justa Bloke on 01/04/06 at 06:08 AMWell done on your election commentary pre and post Kevin. Great work.
Posted by Rick Pilkington on 01/04/06 at 07:51 AMKevin
You seem to be forgetting my comment more than a week and a half ago in response to your post-election article where I said that Kim Booth still had a significant chance of winning, and listed a couple of factors in his favour.
Posted by Tom N on 01/04/06 at 09:42 AMTom, far from forgetting your comment I in fact referred to it in the comments section at http://www.pollbludger.com/350 . My article above largely restricted itself to the comments of “experts” because of the false claims in the media about the view of “experts” on the likely Bass outcome.
One issue Tom raised (the same issue as flagged by Bowe) was leakage from Michelle O’Byrne. I have just obtained the full scrutiny sheet for Bass and noticed that in fact leakage from O’Byrne was not as great as I thought it had been, and only slightly greater than I expected. When I responded to Tom’s question I estimated it at 11%; it was in fact (including leaks via Cox’s surplus) 12.4%. Also, those split more to the Liberals than to the Greens, so Booth’s position after O’Byrne’s surplus was actually only 15 votes better than I expected.
This now suggests to me that the O’Byrne leakage issue wasn’t significant. What, primarily, caused Reissig to lose was the massive leak from Cripps (18.7% leaked to all sources) to Liberals, Greens and exhaust - a consequence of Labor’s failure to build enough profile for, and concentrate votes for, a clear third candidate.
The second issue that Tom raised was the potential for Labor voters to vote 1 to 5 then stop without numbering box 6. When Cripps was excluded, 246 votes exhausted. However, by comparison to cases where the Labor Party has had similar leakages when only running five candidates, it seems very likely that the exhaust problem here comes mainly from voters spraying their vote across party lines. I firmly believe the number of voters who vote 1 to 5 from a 6 candidate team and then stop is negligible - a few dozen at most. See the comments section linked above for full detail.
However in an election where many people, some of whom should know better and some of whom are merely journalists, were calling it as a definite win to Labor prematurely, anyone who saw it otherwise, whether their reasoning was close or not, deserves some credit. So well done Tom!
Posted by Dr Kevin Bonham on 02/04/06 at 08:27 AMThe sex ratio in this house also remains the same.
Posted by phill Parsons on 02/04/06 at 10:09 PMAnother question I can now answer more accurately is Leo Foley’s question about exhaust from David Fry.
Sue Napier’s surplus, worth 1107 votes, consisted entirely of votes from Fry (about 80% of these Fry’s #1s and another 8.7% via other Liberals) which went to Gutwein and then to Napier. 40.8% of this surplus (452 votes) went to exhaust. (This is actually lower than I expected, I was expecting exhaust in the high 50s).
Had these voters completed their ballots in the same proportions as the rest of the Napier surplus, Reissig would have won by 44 votes. It’s therefore quite likely that Liberal voters failing to vote beyond 5 caused the election of Kim Booth.
Posted by Dr Kevin Bonham on 03/04/06 at 05:07 AMThanks Kevin.
Well I still think it’s a bit stupid for the Labor Party (or any party) to run 6 candidates in a five seat electorate.
Labor did the same thing in Bass in 2002 and lost what could have been a third seat for them to the Liberal Party’s Peter Gutwein.
Wouldn’t overall “leakage” as you call it be greater if there are six candidates than if there are five?
Posted by Tom N on 03/04/06 at 08:50 AMOverall leakage may be greater with six candidates rather than five, but the primary vote is also likely to be greater, since the sixth candidate will always pull in at least a few hundred votes from people who vote for them because they know them.
So whether running six candidates is a benefit or a disadvantage is not clear.
In both 2002 and 2006 Labor’s failure to win the third seat in Bass was not caused by having six candidates but simply by the third candidate not polling a high enough primary vote.
In Denison Labor ran six candidates and elected three with no trouble at all because Lisa Singh polled an excellent primary vote.
Posted by Dr Kevin Bonham on 03/04/06 at 11:09 AMA comment on how Paula Wriedt won.
Paul Lennon’s surplus being almost twice the size of Will Hodgman’s could have placed Wriedt at some danger, but actually Lennon’s surplus leaked at a miserly 4.4% compared to Hodgman’s 12.3%.
This meant that after the two surpluses Labor trailed by only 272 votes.
Although the Socialist Alliance’s Matt Holloway’s votes favoured the Liberals (77 votes) compared to Labor (39 votes) - the Socialist Alliance and Liberals were the first two parties on the paper so the 77 were almost all donkey votes - Labor gained almost 100 votes compared to the Libs off independent candidate Hall and over 300 due to less leakage from minor candidates, to be practically level when the last Green (MacDonald) went out.
Of MacDonald’s preferences, 32% exhausted with the rest favouring Labor. MacDonald’s preferences favoured Labor over the Liberals 64:36. At this point there was no effective chance of Wriedt being caught and the final margin was 1032 votes.
Posted by Dr Kevin Bonham on 03/04/06 at 12:52 PMDear Kevin,
Could you please provide a view on the likely causes for the disparity between the Greens poll and pre-poll votes in Bass, Lyons and Braddon.
Could this be a result of the last week advertising?
I don’t have the results of previous elections, which might reveal a trend, however it seems strange that there is up to a 6% difference between the pre and poll in some of these seats.
I know the negative advertising was concentrated in these seats.
Do you think this might be a valid explanation?
Posted by red hot on 04/04/06 at 11:32 PMApologies for missing the final question before - have just seen this while looking for old stuff by another poster.
The main reason for the difference between the poll results and the actual results is that the poll sample sizes are too small to be all that accurate.
Even a 400-vote sample can easily be several percent out for a party with a relatively low vote.
Another is that voters sometimes pick “Green” as an alternative to “don’t know” when they really haven’t made up their mind.
Still another is undecided voters sticking very strongly with Labor.
I doubt that the blitz by the Exclusive Brethren actually had a huge impact, although I have heard, anecdotally, of some cases of people changing their vote away from the Greens because of it.
Posted by Dr Kevin Bonham on 21/04/06 at 11:12 AM
