Morgan face-to-face: 52-48 to Coalition

Morgan’s latest face-to-face poll, conducted last weekend from a sample of 878, shows no change in the two-party support from poll conducted a week earlier in the two days before the Labor leadership spill: the Coalition leads 52-48 on respondent-allocated preferences and 50-50 with preferences distributed as per the result of the 2010 election. However, both major parties are up on the primary vote, Labor by 1.5 per cent to 39 per cent and the Coalition by 1 per cent to 43.5 per cent. The Greens are down one point to 10 per cent with “others” down 1.5 per cent to 7.5 per cent. One hesitates to read too much into Morgan face-to-face polls, but I’m tempted to read this as more evidence of opinion polling’s remarkable imperviousness to recent political turmoil (though judgement should be reserved until more post-spill polling evidence becomes available). Morgan also treats us to state-level breakdowns derived from the last month of regular weekend polling, thereby producing useable samples for the states individually. This convincingly shows Labor struggling in NSW and performing best in South Australia, but eyebrows may be raised at the result from Queensland: Labor trails only 51-49, quite a lot better for them than the 54.5-45.5 New South Wales result, and has a higher primary vote than in Victoria (39.5 per cent compared with 38 per cent).

Newspoll: 57-43 to Coalition

The Australian reports the latest Newspoll (the first in three weeks, following a break for the long weekend) has Labor recovering three points from their record low primary vote last time, but continuing to languish on 29 per cent. The Coalition also picked up a point on the primary vote, to 49 per cent, and maintains a two-party preferred lead of 57-43, down from 58-42 last time. The Greens have dropped a point to 12 per cent, with “others” taking most of the damage from the higher major party vote. The Prime Minister’s personal ratings remain dismally low, with approval up a point to 28 per cent and disapproval down one to 60 per cent. Tony Abbott is up slightly, by two points on approval to 36 per cent with disapproval down a point to 53 per cent. The preferred prime minister is unchanged with Abbott leading 40 per cent to 35 per cent. Newspoll has also has responses for best party to handle various issues: these have Labor going back on all measures since the question was last asked before the election, which is entirely predictable given the normal pattern of these responses following in the direction of voting intention.

This follows today’s Essential Research poll which had the Coalition lead steady at 55-45, from primary votes of 33 per cent for Labor and 48 per cent for the Coalition (both steady), and 10 per cent for the Greens (down one). Further questions suggest the public has trouble distinguishing between the four independents: those who back the government, Rob Oakeshott, Tony Windsor and Andrew Wilkie, all have approval ratings of 23 per cent or 24 per cent and disapproval ratings of between 32 per cent to 34 per cent. Bob Katter performs slightly better, with 27 per cent approval and 36 per cent disapproval. The broad hostility to the independents individually is reflected by the unpopularity of the balance of power arrangement overall. Only 22 per cent consider it to have been good for Australia – a substantial worsening since polls in the early part of the year, the more recent of which (on June 6) had it at 28 per cent. The bad rating is up from 39 per cent to 50 per cent.

Questions on poker machine reform suggest that while Clubs Australia’s grand finals advertising blitz may have had some impact, the public remains strongly in favour of mandatory pre-commitment on poker machines. The level of support is down to 61 per cent from 67 per cent four weeks ago, which opposition up five points to 30 per cent. Respondents were also asked to nominate a figure which “reflects the social cost of problem gamblers in Australia”, and opponents seemed reluctant to do so: 42 per cent opted for don’t know compared with 25 per cent among supporters. Those that did name a figure tended to come in at well below the $4.7 billion indicated by the Productivity Commission, with options of $1 billion or lower chosen by 44 per cent ($100 million being the most favoured), compared with 9 per cent for $5 billion and 5 per cent for $10 billion. Once appraised of the Productivity Commission result, support for pokies reform returned roughly to the level it was at four weeks ago. Respondents were also advised that 2.7 per cent of poker machine revenue was invested into the community, and it seems that for some this was enough: support for reform then came down to 57 per cent, with opposition at 31 per cent.

Misha Schubert of the Sydney Morning Herald has also brought tidings of a Galaxy poll of the electorate of Melbourne which shows Greens incumbent Adam Bandt headed for an easy victory regardless of what the Liberals do with their preference recommendation. Bandt’s primary vote is at 44 per cent against 29 per cent for Labor and 23 per cent for the Liberals, which compares with respective results at last year’s election of 36.2 per cent, 38.1 per cent and 21.0 per cent. This would translate into a 65-35 win for Bandt if Liberal and other preferences were allocated as per the 2010 election result: an anti-Labor swing of 9 per cent in Labor-versus-Greens. We are told that if the Liberals put Labor ahead of the Greens on their preference recommendation, as they did to such devastating effect at the Victorian state election, Bandt would still emerge 56-44 in front – exactly the result he achieved at the election. This result appears to have been arrived at by splitting Liberal preferences 60-40 in Labor’s favour rather than the usual 80-20, which seems soundly based on results from the state election. The poll was conducted two weeks ago from an unspecified sample size, and I’m guessing was conducted for a corporate or peak body client (UPDATE: It’s been pointed out to me that the article notes it was conducted for the Greens).

Final 2PP: 50.12-49.88 to Labor

The Australian Electoral Commission has finalised the last of its two-party preferred Labor-versus Coalition counts, and it confirms Labor has won a narrow victory on the national total of 6,216,439 (50.12 per cent) to 6,185,949 (49.88 per cent), a margin of 30,490. If distinctions to the second decimal place are what matters to you, Labor did about 0.05 per cent worse than last time due to the arbitrary fact of the Nationals finishing ahead of Wilson Tuckey in O’Connor, meaning the AEC finalised a two-party result on a Nationals-versus-Labor basis where the 2007 Liberal-versus-Labor result was more favourable to them. So while I think it reasonable to cite the published figure as the definitive national result, a slight discount should be factored in when considering the matter of the swing, which should properly be rounded to 2.5 per cent rather than 2.6 per cent.

Whatever the specifics, the result leaves quite a few people looking foolish:

Barnaby Joyce: “We’d won the two-party preferred vote by the time the independents made their decision.” (Lateline, 7/9).

Andrew Bolt: “Labor won fewer votes, fewer seats of its own and less of the two-party preferred vote.” (Herald Sun, 8/9).

Alan Jones: “Is it a healthy democracy when a party wins the majority of the two party preferred, wins the majority of the primary vote and wins more seats in the Parliament than the other party but the other party forms government?” (2GB, 8/9).

Sarah Martin: “Yesterday, Julia Gillard’s Labor Party won government despite losing the primary vote and the two-party-preferred vote, or securing a majority of seats.” (The Advertiser, 7/9).

Kerry Chikarovski: “The Coalition won the primary vote, they won the two-party preferred …” (The Drum, 7/9).

Lateline: “Labor loses two-party preferred vote” (report headline, 30/8).

Kenneth Wiltshire: “It is probable that the Coalition will win more third-party preferences.” (NB: This of course is absurd – Labor got 65 per cent of third party preferences, much as they always do – but I think we know what he’s trying to say.) (The Australian 6/9).

Lisa Wilkinson (to Wayne Swan): “Now, you won fewer primary votes, fewer two-party preferred votes and fewer seats.”
(Swan explains to her that she’s wrong.)
Wilkinson: “But in the end you got 49.9 per cent of the vote and the Opposition got 50.1.”
Swan: “No, I don’t think that’s … Lisa, that is not a final count.”
Wilkinson: “Well, that’s what the AEC is saying and that’s what Australia said at the polls.” (The Today Show, Nine Network, 9/9).

No doubt there were others.

Our troubles here began on August 30, when the AEC removed three electorates from the national total on the basis that the Labor-versus-Liberal counts there had been discontinued after election night, as it became apparent the Greens (in the case of Batman and Grayndler) or Andrew Wilkie (in the case of Denison) rather than the Liberals would face Labor at the final count. As three of the weakest seats in the land for the Liberals, these were by extension among the strongest seats for Labor in two-party terms. The resulting adjustment in Labor’s two-party vote from 50.4 per cent 50.0 per cent led to a great many uncomprehending reports of a “surge” to the Coalition, which had an added edge due to Julia Gillard’s post-election claim that Labor had, apparently, won the two-party vote. Those who wanted a clear and accurate exposition of the news had to ignore, say, The Australian, and look to an evidently more reliable source of information in Bob Brown, who explained the absence of eight electorates from the published result and correctly concluded: “If you look at the whole of Australia and you treat every seat equally, when you do that Labor’s ahead and is likely to keep that lead right the way through to the finishing pole.”

Antony Green defends journalists on the basis that they were within their rights to take an official AEC figure at face value, but I’m not so kind. Even if awareness of the missing electorates was too much to ask, those quoted above should at least have been aware that the count was incomplete. As it stands, we have a result that leaves those of us who had done the sums with exactly what we were expecting, and a lot of dopey pundits and dishonest politicians with egg on their faces.

UPDATE: Morgan has published results from a phone poll of 541 respondents conducted on Wednesday and Thursday evening which has Labor leading 52-48 on two-party preferred from primary votes of 35.5 per cent for Labor, 42.5 per cent for the Coalition and 15 per cent for the Greens. The margin of error on the poll is about 4.2 per cent.

UPDATE 2: As Peter Brent points out, the 52-48 result comes from the less reliable two-party measure based on respondent-allocated preferences – going on previous elections, which the most recent election has again vindicated as the superior method, Labor’s lead is only 50.5-49.5.

D-day plus 17 …

… and finally, the small matter of the result. The winner is Labor. Not a famous victory by any stretch of the imagination, but under the circumstances they’ll be happy with the four points. For my part, I have my doubts about members of parliament going against the obvious preference of their constituents in so fundamental a matter as government formation, barring the proverbial extraordinary and reprehensible standards. I’m not entirely sure that a bit of creative accounting from the Coalition clears the bar on this count. That said, I have no doubt – none – that conservatives tempted to echo these sentiments will find themselves constrained by their philosophy’s most eloquent champion, Edmund Burke, who famously told his constituents: “Your representative owes you, not his industry only, but his judgment; and he betrays, instead of serving you, if he sacrifices it to your opinion”. (Burke’s speech has been given a fair run around the place in recent times, including by apparently learned people who would do well to read it through to the end.)

Besides, my difference of opinion with Windsor and Oakeshott is purely a matter of degree – there were respectable arguments that could have been mounted whichever way they jumped. Ultimately they deserve our gratitude for the patient and considered fashion with which they have navigated through their delicate position. Most particularly, they should be heartily congratulated for ignoring a fortnight’s worth of quacking and bleating from a section of the media that does not at heart believe in consensus, checks and balances or even particularly in democracy. Whatever uncertainties might lie ahead, one thing is sure: these voices will spend the coming months and/or years peddling distortions and hyperbole to create a sense of crisis about a situation which in reality has every chance of serving Australia well. Long may the independents continue to turn a deaf ear.

Essential Research: 51-49 to Labor

The latest weekly Essential Research poll has Labor opening a 51-49 lead after 50-50 results over the previous two weeks, including the weekend of the election. Labor is up a point on the primary vote to 39 per cent and the Coalition down one to 43 per cent, with the Greens steady on 11 per cent. The survey also finds a majority supporting a fresh election, 52 per cent to 33 per cent, with 42 per cent thinking minority government will deliver worse government against 30 per cent better, and 70 per cent thinking an election within the year likely against 16 per cent unlikely. This is contrary to the weekend’s JWS/Telereach poll which had only 26 per cent plumping for a new election against 37 per cent for a Labor and 31 a Coalition minority government – possibly this indicates you get different responses when a respondent can choose a minority government of their favoured party over a new election. The Newspoll targeting the three rural independents’ electorates shortly after the election found opinion evenly divided on the subject of a new election, which is perhaps less appealing to those whose local members are presently holding the balance of power.

Newspoll: ALP favoured for government 47-39

The Australian has published a Newspoll survey of 1134 respondents which finds 47 per cent of respondents want the rural independents to back Labor, compared with 39 per cent for the Coalition. There is, predictably enough, “almost unanimous partisan support among voters for the party they supported” – which can only mean primary vote support for the Coalition has taken a solid hit since the election, at which they polled 43.7 per cent. Hopefully more to follow.

UPDATE: We also have another JWS/Telereach robopoll courtesy of the Fairfax broadsheets, this time of 4192 respondents, which has 37 per cent for Labor, 31 per cent for the Coalition and 26 per cent for a new election. However, on voting intention the Coalition leads 44.9 per cent to 35.4 per cent on the primary vote and 50.4-49.6 on two-party preferred, suggesting most of those in favour of a new election are Coalition supporters.

UPDATE 2: Full JWS-Telereach release here, courtesy GhostWhoVotes. I gather the poll targeted 55 seats with post-election margins of less than 6 per cent, and the vote results above extrapolate the swings on to the national results. On Coalition costings, 40 per cent of respondents professed themselves very concerned and 19 per cent somewhat concerned, with only 35 per cent showing little or no concern. People are more concerned about the Greens balance of power in the Senate (49 per cent say “bad for Australia” against 39 per cent good) than the value of the Labor-Greens alliance (opinion evenly divided). Julia Gillard only just shades Tony Abbott as preferred prime minister, 43 per cent to 41 per cent, and respondents are evenly divided on which party would prove more “stable and competent”.