The Senate: part three (Tasmania)

The Tasmanian Senate result must be considered a big disappointment for the Greens, who were widely expected to win a quota on the primary vote. Instead their primary vote was down to 12.8 per cent from 13.8 per cent in 2001 and they are grappling with Family First in a struggle for the final place. Antony Green offers a set of calculations based on the assumption that all votes are above the line, although Tasmania is the one state in which this is not more-or-less accurate. From a starting position of just 2.31 per cent, Family First would absorb the hefty Liberal surplus remaining after the election of their third and final candidate, and receive handy additions upon the elimination of the Christian Democrats and Labor renegade Shayne Murphy. That would put them ahead of Labor’s third place-holder, whose elimination would decide the result. If all Labor votes were above the line, Family First would emerge with 1.0579 of a quota compared with 0.9421 for the Greens. Green notes that "analysis based on treating all party votes as ticket votes almost always provides an accurate estimate of the final outcome&quot, but if ever there might be an exception to the rule this could be it. No fewer than 24 per cent of Tasmanian voters went below the line in 2001, and this would include a disproportionate number of Greens voters and Labor voters making an effort to ensure their vote does not end up with Family First as per the party’s preference deal. One who has looked at all this very closely is Geoff Lambert, noted election obsessive and frequent correspondent. His number crunching runs as follows:

(In 2001) 24% of all votes were cast below the line. It ranged from about 13% for the big parties up to 57% for The Greens (one candidate, Bob Brown). But this figure for The Greens is biased by Brown in 2001. In other years, it’s been about 25%. These figures may or may not be repeated in 2004 – probably they will, especially since the ballot paper was so small this time around.

75% of below-the line votes show ‘1’ for the #1 candidate (i.e. for candidates on top of the party lists). The latter is important, because the ‘1’ for #1 votes from the booths are ALREADY in the count … only the below-the-lines for postal/prepoll and the below-the-lines for #2,3,4 candidates have yet to show up on the system. On this basis, I would expect that the Greens #2 and #3 are carrying about 25% of 25% of 13% of the vote, i.e. about 0.8% of the state total. That will bring (Greens candidate) Christine Milne up to about 13.6%, which is still not enough. And 0.8% is probably an overestimate because it will be dragged down by the same sorts of things happening in other parties, although to a lesser extent.

On the other hand, there could be leakage from the #1 to #4 candidates of other groups. This is known to happen in the Tasmanian House of Assembly, where a lot of people voted (e.g.) 1 Robin Gray, 2 Christine Milne (and vice-versa). It is hard to know how many, but at least in this category it is potentially enough. The below-the-line for (ALP+LIB) will be about 15% of 80% = 12% of the total votes cast. Milne will need to pick up a leakage of only about 15% of these to creep up to a quota. It is at least arguable that a number of these people were peeved with the decision by everyone to preference Family First ahead of the Greens and so would hop over to the Greens early on in their papers (they have to go to Greens by preference 2 or 3 of the majors of course). That this sort of thing can swing a result was seen here in my local Council election, where the leading group lost almost half a quota through such a process.

There is little hope for Milne to gain on the postals and pre-polls. What has been counted so far shows her going backwards. The bottom line of this analysis is, I think, that Christine Milne MIGHT be elected, but it won’t be on the strength of postal/pre-polls, nor on the strength of Greens below-the-line, but it CAN be done on the strength of below-the-line leakage from Lib and ALP. The scrutineers should be able to tell whether this is likely from tomorrow (i.e. today).

ACT Election Overview

The Australian Capital Territory goes to the polls for the second week in a row this Saturday to elect a new 17-member Legislative Assembly. Chief Minister Jon Stanhope’s Labor Party currently governs with eight members and the consent of one Green and one Democrat. The balance consists of six Liberals and one ex-Liberal independent. The election will be the sixth since self-government was established in 1989 and the fourth since the original 17-member single electorate system made way for three electorates, two (Brindabella and Ginninderra) with five members and one (Molonglo) with seven, elected under the Hare-Clark sytem. The system’s closest Australian relative is that for the Tasmanian lower house, sharing with it rotating ordering of candidates on the ballot paper so that candidates must compete for votes with party colleagues, in contrast to the otherwise similar Senate system where the parties determine the order of their lists.

As Chris Uhlmann of the ABC notes, "it is hard to conceive of any set of results that does not add up to a return of the Stanhope Labor Government". The main source of interest is thus whether Labor can secure the one extra seat they need to become the ACT’s first-ever majority government, with Ginninderra looking their best chance. One observer who is not backing such an outcome is Greens Senator Bob Brown, who has never been shy of making implausible predictions on his party’s behalf. Brown has claimed that the federal election results suggest the Greens will increase their representation from one seat to four by winning one each in Brindabella and Ginninderra and two in Molonglo. Crispin Hull of the Canberra Times also assesses that the Greens will now "have left-field on their own" and are thus assured of a seat in each electorate, although to the Poll Bludger’s mind a lesson of the federal election is that the Greens cannot count on making all the hay from the decline of the Democrats, who received considerable support from those making what they perceived to be an ideologically neutral anti-major party vote. Hull also notes that a repeat of last week’s voting patterns in Canberra would give Labor 10 of 17 seats, and provides the following instructive analysis of why there is no chance of this happening:

Firstly, a lot of Canberrans vote for federal Labor for the purely selfish reason that they are federal public servants whose power and pay, they think, is improved by a federal Labor Government. But when it comes to local government they get as fretful as the rest of Australia, especially over rates – the local equivalent of interest rates. Secondly, in the ACT we have multi-member electorates … So we have a name-recognition factor plus a ‘my-local-member-helped-me’ factor. It means quite a few people will have both Labor and Liberal candidates in their first five or seven preferences. In Molonglo in 2001, 8 per cent of the preferences of the first excluded Liberal went to Labor.

All of which makes the process very convoluted and difficult to pick. Over the coming days the Poll Bludger will review the contests for each of the three electorates in turn. For the time being readers can amuse themselves with the following table, outlining votes and seats won by the various parties in each of the three electorates since their establishment at the 1995 election.

2001 41.7 (8) 31.6 (7) 9.1 (1) 8.1 (1) 9.5 (0)
1998 27.6 (6) 37.8 (7) 9.1 (1) 6.0 (0) 14.4 (3)
1995 31.6 (6) 40.5 (7) 9.1 (2) 3.9 (0) 14.9 (2)
Molonglo (7) ALP LIB GRN DEM OTH
2001 39.3 (3) 34.0 (3) 12.6 (1) 7.7 (0) 6.4 (0)
1998 25.6 (2) 41.5 (3) 10.1 (1) 5.1 (0) 17.7 (1)
1995 30.9 (3) 42.8 (2) 10.1 (1) 3.2 (0) 13.0 (1)
Brindabella (5) ALP LIB GRN DEM OTH
2001 44.0 (3) 31.8 (2) 5.5 (0) 6.9 (0) 11.8 (0)
1998 28.5 (2) 37.1 (2) 8.1 (0) 6.1 (0) 20.2 (1)
1995 31.6 (2) 37.1 (2) 8.0 (0) 3.8 (0) 11.5 (1)
Ginninderra (5) ALP LIB GRN DEM OTH
2001 42.8 (2) 28.0 (2) 7.9 (0) 9.7 (1) 11.6 (0)
1998 29.6 (2) 33.2 (2) 8.7 (0) 7.2 (0) 21.3 (1)
1995 32.7 (2) 40.7 (2) 8.7 (1) 5.0 (0) 12.9 (0)

Tasmania and Queensland Senate latest

Apologies for continued heel-dragging on a full review of the state of the Tasmanian Senate contest, one remarkable feature of which has been a Liberal Party bungle in fielding only three candidates when their primary vote was sufficient to have raised at least the possibility that they could have won a fourth. Some idea of why they might not have thought this possible can be gleaned by tracking the astonishing growth in their vote over the past three elections – from 33.8 per cent in 1998 to 38.8 per cent in 2001 to 46.6 per cent in 2004. At present the contest for the final seat (the other five splitting three Coalition and two Labor) is between Family First and the Greens – while Family First appears to have the edge, Antony Green says he is "doubtful" this will be the outcome since 20 per cent of Tasmanian voters and as much as 50 per cent of Tasmanian Greens voters go below the line, and counting on these votes will not begin until tomorrow.

The most interesting contest from every perspective is that in Queensland, in which the Coalition is well in contention to secure a fourth seat and hence an historic Senate majority (notwithstanding National Party candidate Barnaby Joyce’s assertion that if elected he will behave more-or-less as an independent). The key point in the count will be when either Family First or the Nationals are eliminated – here the Nationals’ lead has narrowed from about 3500 on Sunday to about 2900 today. If the Nationals survive at this point, Family First preferences will put them ahead of One Nation by a similarly narrow margin (and here their prospective lead has increased from about 3500 to about 3700, Pauline Hanson and One Nation predictably doing less well out of pre-polls and postals) and then over the Greens and into the Senate on their preferences.

Down to the wire

Five House of Represenatatives seats are still listed as doubtful by the Australian Electoral Commission; no less than three of them are in South Australia, and four are held actually or notionally by Labor. Of these Labor is trailing in two, both being seats that sitting MPs are trying to win for the first time following redistributions. These include the new Queensland seat of Bonner, which looked the better bet for Con Sciacca after his existing seat of Bowman (where the Liberals picked up a 6.1 per cent swing) became notionally Liberal; and Wakefield in outer Adelaide, where Labor’s Martyn Evans is attempting to move after the abolition of his own seat of Bonython. Sciacca trails by 0.17 per cent with 82.7 per cent of the vote counted, while Evans is currently falling short by 0.37 per cent with 84 per cent of the vote counted. Also in Adelaide, Labor member for Kingston David Cox leads by just 0.19 per cent with 84.9 per cent counted. One wonders how far ahead the Liberals’ Andrew Murfin would be in the Perth seat of Swan if he had kept out of the news during the pre-election period. As it stands he trails Labor’s Kim Wilkie by just 0.11 per cent with 79.4 per cent of the vote counted. Labor’s remaining hope to pick up another new seat to add to its meagre haul of Adelaide, Parramatta and Richmond (where Larry Anthony, trailing by 0.65 per cent, is not giving up yet) is Hindmarsh, an Adelaide seat being vacated by retiring Liberal Chris Gallus, where Labor leads by 0.17 per cent with 80.8 per cent counted. If those currently ahead remain so, the Liberals will emerge with 73 seats compared with 68 in the old parliament, with the Nationals down one to 12 (bearing in mind that he said he would resign if any seats were lost, and that their Senate seat in Queensland is also in doubt). Labor is down from 65 to 62 and the three independents have been returned. On the latter point at least the Poll Bludger’s overall prediction was correct, but he short-changed the Liberals by six seats. The Runs on the Board tally to the left now records the actual rather than predicted result. Post-match summaries for every seat will be added to the federal election guide with all possible haste when the figures are finalised.

The Senate: part two

Tasmania will require a good hard think, so that will have to wait for tomorrow. The situation in the other states not yet covered is as follows:

South Australia: With the Liberals certain of three seats and Labor certain of two, the final place will be a contest between Labor and the Democrats. The crucial point of the count is who drops out first out of the Democrats and Family First, a finely poised contest in which Family First currently appears to have a slight edge. If the Democrats can get ahead, the Family First-Democrats preference deal will then put their candidate John McLaren ahead of the Greens and then over a full quota with their preferences. Otherwise the Greens’ preferences will be deciding the issue between Family First and Labor – predictably, they have favoured the latter.

Western Australia: A refreshingly straightfoward outcome of three Liberal, two Labor and one Greens. The Greens’ quota will be built upon a primary vote of 7.73 per cent plus Labor’s surplus (4.4 per cent) and preferences from the Democrats (1.97 per cent).

The Senate: part one, sub-section A

Antony Green has offered the astounding assessment that the Coalition could achieve the unprecedented feat of winning four seats in one state, namely Queensland, with both Nationals candidate Barnaby Joyce and the Liberals’ Russell Trood set for victory (other readers who concur include Bryan Palmer of Oz Politics and regular contributor Geoff Lambert). Since they also appear set to win three seats in each state plus one each from the territories, this suggests that the Coalition will emerge with the first absolute majority in the Senate since 1981. If so, it appears that the scale of the Coalition’s triumph at this election has yet to be fully appreciated. Green’s perspective on the Senate election was one of many the Poll Bludger received from various readers, which have collectively exposed a few wonky assessments in yesterday’s posting. Updates appear in order:

New South Wales: Here at least it looks the Poll Bludger had it right yesterday in calling it as three Labor and three Coalition. If 100 per cent of votes were above the line (the real figure being more like 90 per cent), on current figures the last three standing in the battle for sixth place would be Liberals for Forests (0.81 of a quota), Labor (0.66) and the Greens (0.53). Since the Greens have put Labor ahead of Liberals for Forests, their elimination would result in the re-election of Labor Senator Michael Forshaw.

Victoria: An apparently wrong call by the Poll Bludger in writing off Family First, which was based on the assumption the third Labor candidate would finish ahead of the Greens and then win the seat on their preferences. On present figures the Greens would be leading Labor 0.69 of a quota to 0.59. That would leave Labor preferences deciding the outcome between Family First and the Greens – and perhaps contentiously, they are delivering them to Family First. The Greens have thus been doomed in being frozen out by both Labor and Family First. Ironically, if they had done slightly less well their preferences would have given the seat to Labor, which they would surely have preferred.

Queensland: It appears that Hetty Johnston (apologies to serious Senate watchers for yesterday’s rather strange conclusion that Johnston was in with a chance), Liberals for Forests, the Fishing Party and the Democrats will progressively drop out leaving the third Liberal and Labor candidates, the Nationals, Pauline Hanson, One Nation, the Greens and Family First. Labor and Pauline Hanson will then drop out in quick succession – somewhat surprisingly, this means One Nation will have done better than its founder. The next to be eliminated will be either Family First or the Nationals – the result is very tight with the Nationals just a shade ahead. If the Nationals prevail they will aborb Family First’s preferences which will put them about even with One Nation, again just slightly ahead. Since most of One Nation’s support comes from last-minute fine-avoiding impulse voters, it seems reasonable to assume that this gap will only widen as absentee votes come into play. When One Nation and the various components of its accumulated vote scatter at this point, Antony Green calculates that the Liberals will have 1.08 of a quota and the Nationals 1.03, with the Greens stranded on 0.89.