Post-Budget Polls: Big Labor Leads, But Message Is Mixed

AS UNIVERSALLY EXPECTED, the Abbott government has taken a hefty hit in major opinion polls conducted since Tuesday’s federal budget, with leads after preferences to the ALP of as much as 12%; even so, both Newspoll and Nielsen carry mixed messages, with below the line findings that place qualifications on their headline results. These are bad numbers for the government, yet with no irony at all might have been far worse.

As I said in my article yesterday afternoon, the Left will revel in the numbers this week’s opinion polls record, perhaps justifiably so; yet when an election is eventually held the odds must favour the re-election of the Abbott government, and this contention remains very much the case.

That said, findings published overnight by Newspoll, appearing in today’s edition of The Australian, along with material from Nielsen — both of which correlate with figures released a day earlier by Galaxy — are extremely bad for the Coalition whichever way they are sifted and presented, and whilst they do not signal the death knell of the government, they inarguably declare that Treasurer Joe Hockey’s first budget has been very poorly received indeed.

There are three things I will say upfront: one, and broadly speaking, Bill Shorten is likelier to have been found to be more popular this week than he ever will be again in his political career. Two, despite its leads after preferences, the Labor primary vote in all surveys should be far stronger for it to represent anything truly meaningful. And three, just as respondents to these polls are reacting against measures that clearly affect them adversely, there are indicators that suggest today’s stated voting intention may not translate into tomorrow’s vote.

Newspoll finds Labor ahead after preferences, by 55% (+2% since last fortnight) to 45%; on primary votes, this breaks to 38% (+4%) for the ALP, 36% (-2%) for the Coalition, 11% (-3%) for the Greens and 15% (+1%) for “Others.”

In other words, fully one-third of that 55% two-party lead is totally dependent on attracting preferences from other parties; preferences from the Communist Party Greens might well be a given in 75-80% of cases, but beyond that it becomes a much tougher ask. As a rough rule of thumb, the “Others” typically split 50-50 at elections, so the glitch in the headline finding of 55% Labor support after preferences is a glaring one.

It’s a similar story with findings from a Nielsen poll, which found primary support for the Coalition at 35% (-5% from its poll a month ago) and 40% (+6%) for the ALP; it also found the Greens sitting on 14% (-3%), the Palmer United Party on 6% (-2%), and “Others” on  5%, for a two-party lead to Labor of 56-44. Preferences in Nielsen’s research increase the overall ALP outcome by almost half as much again as its primary vote, and whilst anything is possible such a scenario sits at the upper outer reaches of anything that could be regarded as plausible.

For comparison, a Galaxy poll released late on Saturday night had Labor at 53% (+1%) after preferences to the Coalition’s 47% (-1%); it found primary Coalition support at 38% (-1%), Labor 38% (+1%), Greens 10% (-1%), the Palmer United Party 8% (+2%) and “Others” on 6% (-1%).

The point I would make is that whilst Labor support has certainly risen across the board (in contrast to the round of polls leading up to the budget) there is no decisive shift of support directly to the ALP that suggests an election win is locked in, or at the very least an outcome motivating the sentiment being expressed by respondents of the respective surveys. Its 40% in Newspoll is Labor’s best federal primary vote, anywhere, in years. The fact it is not mirrored elsewhere at a time of favourable political circumstances should be very worrying for the ALP indeed.

During the last term of Parliament, the Coalition primary vote in these surveys typically fluctuated between about 45% and 50%, and usually at the upper end of the range, and whilst the ALP has the “Green factor” hiving off a considerable portion of the Left-leaning vote, a similar phenomenon seems to have appeared in recent times with the emergence of parties linked to Clive Palmer and Bob Katter Jr.

And a primary vote of 37.8% in 2010 was only strong enough to win Julia Gillard government in minority; of course every election is different, but the fact is that even with the sizeable overall movements these polls are picking up, Labor’s actual support is no better on average than it was in 2010 or for much of the period it spent in opposition during the Howard years.

I’m not trying to piss on Labor’s parade, mind; the point is that in context and as these polls are shaped by political circumstance, the reality beneath the headline result is far less rosy for Labor than it will no doubt seek to claim.

This point is borne out to some extent by some of the questions each of the surveys asked its respondents about the budget.

For instance, Newspoll asked whether the ALP would have produced a better budget: only 39% thought it would, with 46% saying Labor would have done no better. This was despite 69% of its respondents saying it would leave them worse off personally.

Nielsen found that 33% of its respondents thought the budget was fair; 63% thought it was unfair. Yet its respondents split 50-37 in favour of the so-called “deficit tax,” and 49-46 in favour of the abolition of the carbon tax. (They also split 30-66 against raising the GST — a measure not included in the budget at all).

And Galaxy — 11% of whose respondents said the budget would leave them better off, and 75% saying they would be worse off — nonetheless split 41-46 on the question of whether they felt the budget would be good or bad for the economy, with “bad” attaining only a narrow plurality over “good” despite Labor’s doom and gloom rhetoric in recent weeks.

For these polls to give Labor real succour, they ought to be so overwhelmingly one-sided as to enable us to close off the betting and start paying out. They are nothing of the sort. And whilst it is very clear the government has a hell of a lot of work to do, it is equally clear the ALP remains well short of the compelling movement it will need to stand a serious chance of winning the next election.

Ironically — given they’re probably the most meaningless aspect of these polls — the best numbers for Labor come on the personal approval measures. Yet the caution on these is perhaps even stronger than the flimsy nature of the voting intention figures: Shorten is not a popular or particularly well-liked leader, and the big gains he has achieved this week are likely to prove illusory at best.

Newspoll has Shorten (44%, +6%) leading Abbott (34%, -6%) as preferred Prime Minister; over at Nielsen, it’s Shorten on 51% (+7%) over Abbott on 40% (-5%). The movements record both a swing against the government and reflect the obvious hostility of respondents across all the surveys to the budget. Yet they pit two relatively unpopular men against each other, and like virtually every similar finding since Shorten became ALP “leader” late last year, neither man has scored a runaway win over the other on any of these personal preference measures.

Similarly, Shorten records some improvement on the personal approval front, with Newspoll finding 42% of its respondents (+7%) approve of his handling of the opposition leader’s role, and 39% (-2%) don’t; at Nielsen, 47% (+4%) approved of him and 39% (-2%) didn’t. Yet again, these are hardly Earth-shattering numbers coming off an extremely low base, and reflect the fact Shorten is a fundamentally unpopular individual receiving the rub-off effects of the sentiment being expressed against the budget.

And Abbott? With disapproval in both polls now topping 60% and his approval numbers only barely in the 30s, it’s a case of “enough said.” Yet it must be remembered that Abbott has made a career as leader out of surviving these kinds of approval ratings. He won an election with them last year in a canter.

A lengthy post, yes, but there are three sets of numbers — plus dedicated questions on the budget — to sift through. What can we take out of them?

At the risk of being blase, there is nothing in any of this that shouldn’t have been expected.

Yet just as the government shouldn’t panic today, neither should the opposition crow: the results are bad for the government, but not dire; they are good for the Labor Party, but not convincing or solid.

One thing that occurred to me after my post yesterday is that the ALP has basically front-loaded its attack; in going hell-for-broke on the budget before it was even delivered based on the aspects of it that were leaked, it could be that it’s turned what might have otherwise been a slow burn into a single shot that may yet misfire. As I said yesterday afternoon, if the government can legislate the bulk of its measures and the sky doesn’t subsequently fall in, this budget mightn’t even be an issue when the next election rolls around — arguments over broken promises from the Coalition or not.

It is still very much the case that the government has followed the textbook: get into office, “discover” things are far worse than its predecessor claimed, drop the bomb of a horror budget and ride the storm out knowing there is ample time to turn public opinion back in its favour before it next faces the people.

This strategy has worked too many times in recent political history — 1983 and 1996 specifically — for Coalition strategists not to tread a similar path with some confidence. Labor failed to do so in 2007 (and tacitly acknowledged the strength of the budget it inherited) and was almost booted out after a single term in office.

In the longer run, these polls are meaningless, although we have to take the time to pull them apart.

Readers know I believe there are measures in the budget that should not have been included and which are certain losers of Coalition votes, but notwithstanding my criticisms there is an awful lot more that will influence the next election and the degree — if any — to which the budget shapes its outcome.

How much of Hockey’s budget even gets implemented is a very big question mark as at today’s date; how much is dropped, blocked, modified beyond recognition or enacted all promises to create ripples and knock-on effects whose outcomes — and consequences — cannot be foreseen.

And the budget aside, the electoral cards still remain very much in the Coalition’s hands.

After all, it faces a Labor Party booted from office in a landslide eight months ago; “led” by an odious and unpopular individual, that same Labor Party shows no inclination to either acknowledge the errors it made in office or even to be honest with the electorate when its own dishonest way of doing things was such a contributor to its fall from grace.

Abbott — to the extent a case can be made about broken promises — at least has the mess he inherited from the ALP to point to as an excuse. The ALP doesn’t even have that.

And smouldering away in the background are the Royal Commissions into Pink Batts and corruption in the union movement; various criminal investigations that may well see a throng of senior ALP figures charged and, potentially, jailed; and the Labor Party making noises about reform and inclusion and democratisation at the same time it refuses to dilute the level of union control of its conferences, which risks alienating even more of its rank and file than it already has when the betrayal it is wilfully committing becomes apparent.

Let’s call this for what it is: the best week of polling Labor has enjoyed in five years, and a bloodied nose for the government.

And yes, the budget process will have cost the Coalition some of its political capital, and its effort is in no way an emulation of either the fabled Costello budget of 1996 or the stellar level of acceptance it met with in the electorate.

But nobody — on either side — should get carried away by any of this. To do so is premature, baseless, and a dangerous political miscalculation to make.



Final Polling: Big Coalition Win; 53.5%-46.5% Lead

SIX YEARS of Labor government is set to end in landslide defeat today, if the final opinion polls across all mainstream pollsters are accurate; with an average lead of seven points after preferences across the polls, Tony Abbott is set to be elected in a win that will rival John Howard’s triumph in 1996.

Over the past 24 hours, I have been watching the special election eve survey results from each of the major polling outfits filter through; there is some variation between them, but nothing beyond the margin of sampling error — or anything even approaching it.

I am only going to focus in detail on the two-party figures, because — after all — every seat in the House of Representatives will be determined on a two-candidate final count.

At the business end of the process it’s votes that ultimately matter: approval ratings and “preferred Prime Minister” contests are useful in the middle of the cycle, but tomorrow Australians will vote for who they will — and with neither leader recording messianic ratings, those questions now seem redundant.

I will however make the observation that across all of the latest polls, on average, Tony Abbott ends up with a modest lead over Kevin Rudd in the “preferred PM” stakes.

Indeed, Abbott ends this campaign more popular than Rudd.

And that’s an absolute indictment on Rudd and the ALP, with Rudd purportedly the most popular politician in the country, and Labor supposedly having executed an electoral masterstroke by restoring him to its leadership to contest today’s election.

Instead, it may well have worsened the inevitable defeat.

In this final round of special polls, the two-party preferred breaks (all in favour of the Coalition) are: Newspoll, 54-46; Essential, 52-48; Galaxy, 53-47; and Nielsen, 54-46.

I saw a poll yesterday from industry newcomer Lonergan Research, effectively finding 51-49 for the Coalition; this, however, was a mobile phone only poll, and it is reasonable to expect its findings were disproportionately drawn from the younger age quintiles where ALP/Greens support is higher (even if the data was weighted to compensate).

And even the often-maligned Morgan poll — whose results are often inclined to wild and inexplicable fluctuations, quite literally, from one week to the next — has posted final figures showing a 53.5-46.5 result in the Coalition’s favour.

And that, dear readers — 53.5% to 46.5% — is the exact average of the Coalition lead over Labor, after preferences, across the four usual polls we follow, plus Morgan.

In turn, a 53.5% result for the Coalition today would represent a 3.6% swing away from the ALP and easily elect Tony Abbott Prime Minister.

Applying a 3.6% movement to the electoral pendulum sees the Coalition win 14 additional seats from Labor (plus those of Peter Slipper in Queensland and the two independents in NSW) for a total of 89 of the 150 House of Representatives seats, and a majority of 28.

It isn’t that simple, of course, and I expect Abbott to do a bit better than that.

The polling figures we are looking at today come at the end of a campaign in which the overall trend and movement has been back to the Coalition following the end of Kevin Rudd’s “honeymoon” as a restored PM.

That movement, however, can be broken into three phases: a marked movement to the Coalition immediately after the calling of the election; a very slight drift back to Labor about two-thirds the way through the campaign; and more movement to the Coalition this week to round out the run to the polling stations.

Based on these last survey results, that movement appears to be continuing, even now; this is the first reason 53.5% probably understates the current level of Coalition support.

A second reason is that the most recent findings — Newspoll, Nielsen and Morgan — all fall closer to 54% (and in the case of Nielsen and Newspoll, may well have been rounded down to that level for publication).

It’s an especially valid point in the case of Newspoll, which in 2010 published its findings to one decimal place (which I seem to recall was a 50.3% ALP lead). It hasn’t done so this time, so it becomes a matter of speculation.

But that speculation becomes a little less…er, speculative…when it’s remembered that at most of the elections won by John Howard, conservative support as measured by election eve polls was typically understated when compared to the actual results (which, of course, are determined by actual preference allocations rather than the statistical distribution of same based on the patterns at the preceding election).

And a third factor is that no election swing is uniform: the pendulum may move x seats on y swing, but with fluctuations from seat to seat, an efficient swing can yield more seats than the pendulum suggests — and that also applies in reverse, although today I doubt it.

My guess, therefore, is that the Coalition result will look more like 54%, or even 54.5%.

(And regular readers will know that I have long expected the actual election result to come in at around 54% for the Liberals, after preferences — even when the Coalition was pulling in 57-58% numbers in the polls early this year, and again prior to that).

Simply stated, if these numbers are broadly accurate — and there is little reason to believe they are not — the Liberal win today will be on the same scale as Howard’s in 1996, and may even edge toward that gold standard of election beltings: Malcolm Fraser’s in 1975.

Whichever way you cut it, though, even the polls we are talking about are already out of date; there are factors influencing people’s voting intentions even now, and if there is a continuing drift one way or the other there is still a full day of voting for more votes to shift from one column to the other.

But with all that considered, my personal prediction is that the Liberals and Nationals will collectively win 95 seats; the ALP 52; Adam Bandt will retain the seat of Melbourne for the Greens; and there will be one Independent (Andrew Wilkie), plus Bob Katter in Kennedy.

This adds up to a 40 seat Coalition majority: the same buffer secured by Howard in 1996.

If all of this comes to pass, the ALP will have many wounds to lick; and with Kevin Rudd defeated in Griffith (yes, I see that happening), its humiliation — and its punishment for six self-indulgent, dysfunctional and chaotically misspent years — will be absolute.

Election Polls, Week 1: Galaxy, Nielsen Spell Trouble For Rudd

TWO NEW POLLS signal trouble for Kevin Rudd and the ALP, with movement in the first week of the campaign to the Coalition; they by no means spell the end of Rudd’s valiant — if fundamentally dishonest — attempt to pinch the election, but the ongoing trend leaves him with no margin for error.

I intend to make this a very succinct post; I will be back later today to look at other issues, and broadly, readers can see some of the figures from relevant tables.

(And at the time of writing — 2.30am, Melbourne time — I’m about done for the “day.”)

Nielsen’s poll — its first in four weeks — shows a national split of 52-48 in the Coalition’s favour; a movement toward the Coalition of two points over that period.

This period, of course, is entirely marked by Kevin Rudd in the Prime Minister’s office, and — like Newspoll and Galaxy’s last national findings — suggests the upward bounce experienced by Labor under Rudd is now well and truly finished.

Indeed, a week into an election campaign, it shows the Labor vote in reverse.

But Nielsen’s poll brings the recent average of all of the reputable polls to an even 52-48 for the Coalition, or perhaps even a sliver better than that; only a couple of weeks ago we were talking of this average sitting at 51-49 for the Liberals.

As readers will have heard me say many times, it’s the overall trends identified in polling that make it useful, not individual polls themselves; even so, Nielsen’s findings are of particular interest on account of the fact they are the first to be sampled entirely within the formal election campaign period.

Nielsen’s full findings can be accessed here.

The key take-outs from this poll all mirror and validate the trends that have shown up — almost uniformly — in the other reputable published polls over the past couple of weeks. The actual numbers may vary a bit (for example, Nielsen has far fewer “undecideds” on questions of, say,  leadership approval), but the trends are unmistakable.

These are:

  • The Coalition remains on track to win the 7 September election;
  • Kevin Rudd’s overall “popularity” is fading quickly;
  • Tony Abbott is increasingly perceived as more popular; and
  • Rudd’s lead over Abbott as “preferred Prime Minister” is shrinking.

The other poll out today is a Galaxy poll, which appears to have delivered results for Queensland only at this stage, and which shows Labor lagging behind the Coalition in the Sunshine State by a 56-44 margin.

Significantly, Galaxy’s researchers were in the field on Wednesday and Thursday, so at least a portion of these results will have been recorded after word spread on Thursday morning that ex-Premier Peter Beattie was throwing his hat into the election ring.

On this account, the findings should alarm Labor; Queensland — on paper — is the only state in which it can plausibly make gains (aside from perhaps a solitary seat in WA, and at a stretch) to offset expected heavy losses down the rest of the east coast and in Tasmania.

Even so — and as I opined yesterday — I’ve never been a convert to the theory that either Rudd’s resuscitation as Prime Minister or the performance of the Newman government in Queensland were at all likely to translate into solid electoral gains for the ALP in that state.

And if we take the average of the Nielsen and Galaxy numbers for Queensland (Nielsen found a 53-47 advantage for the LNP there), the resulting 54.5-44.5 split still represents a paltry swing of 1.4% to Labor, and if applied uniformly on the pendulum would net the party a single additional seat: ironically, not the one Beattie has opted to contest.

Of course, swings are rarely (if ever) uniform, so it’s dangerous to read too much into such a scenario. But I make the point simply to illustrate just how difficult it will be for Rudd to extract any electoral joy — even in his beloved Queensland.

The short synopsis of these new numbers is that Labor is doing it tough; barring a colossal (and highly improbable) gaffe by Tony Abbott in tomorrow night’s leaders’ debate, I don’t see anything in the immediate term that is likely to change that.

My sense is that — on a points call — the Coalition has won the first week of this campaign, and the numbers we’re talking about here underline the contention.

As ever, we’ll look at further results during the campaign as they are published.

Dead Cat Bounce: Kev, Albo And Opinion Polls

DAY THREE of the brave new Kevin Rudd regime yesterday saw the first  poll since his “resumption” of the Prime Ministership showing an Abbott election win, 51-49; the significant spike in Labor’s numbers will be short-lived and, with “Kev” and “Albo” in charge, no more than a dead cat bounce for the ALP.

Prior and subsequent to Wednesday night’s leadership change over at the ALP I’ve taken numerous calls from Liberal voters, deeply worried that a switch to Kevin Rudd may allow Labor to steal an election win that by rights it has no entitlement to even contemplate.

What I have told anyone who has asked — and time will tell if I’m right — is that the first polls after the leadership change are likely to produce an immediate and significant bounce in support for the ALP, solid numbers for Rudd as PM, and possibly holding out the prospect of Labor in an election winning position.

However, such a bounce won’t last, and Liberal supporters and MPs need not panic.

And so it has come to pass, with the first of the major polls — Galaxy — yesterday showing Labor now trailing the Coalition by a narrow 49-51 margin after preferences.

The only surprise as far as I’m concerned is that it didn’t show Labor in front.

The 49-51 deficit Galaxy reports represents a swing to Labor of 4% on its previous survey taken a little over a fortnight ago, which showed that under Julia Gillard the ALP was on course for a 55-45 landslide loss.

This result is to some extent reflected in an automated ReachTel poll taken immediately after the leadership switch that showed Labor behind the Coalition by a slightly wider 48-52 margin under Rudd.

ReachTel typically overstates the size of the Labor vote, as we saw during the Queensland state election campaign last year. So it’s early days, but the portents for the government aren’t good, or at least certainly not as good as they might have hoped.

Galaxy finds primary vote support for the ALP up six percentage points to 38%; the Coalition down three points to 44%; the Greens down a point to 10%; and support for “Others” falling two points to 8%.

On the “preferred Prime Minister” measure, Galaxy finds Rudd heading Abbott by 17 points, 51-34; this compares to a 37-33 Abbott lead over Gillard in the previous survey.

This in itself isn’t a bad result for Abbott by historical norms; he will be disappointed to fall behind on the measure, but the “preferred PM” count is the one most heavily biased toward an incumbent, and one on which Abbott can be expected to retrieve ground as the novelty of Rudd’s return starts to wear off.

And that’s about as good as it gets, if you sit in the ALP tent and/or barrack for Rudd.

(Galaxy tables here, courtesy of Twitter colleague @GhostWhoVotes).

Galaxy asked its respondents whether they expected the ALP to unite behind Rudd or to continue to be divided; just 36% said they expected Rudd to bring Labor unity, as opposed to 53% who said they expect the divisions of the past three years to continue.

Unsurprisingly, this was split along party lines, with 68% of those identifying as Labor voters saying Rudd would unite the ALP compared to 22% suggesting the division would continue; of Coalition supporters, 79% expected Labor to remain divided with just 15% expecting it to unite behind Rudd.

But stripped down, this suggests almost a third of those intending to vote Labor either expect it to remain disunited or, at the minimum, aren’t confident the party will pull together, with almost two in three respondents overall being of the same opinion.

And when we step away from polling altogether, there are ominous signs for the ALP that its rough ride will continue apace under Rudd.

Stories suggesting Rudd is finding it difficult to field a full team of ministers are surfacing; apparently — and unbelievably — he has been begging Gillard loyalist and former Communications minister Stephen Conroy to reconsider his decision to refuse to serve.

It seems obvious that whatever team Rudd eventually puts up is going to be largely inexperienced; already there are names such as Fremantle MP Melissa Parkes and Ballarat MP Catherine King being bandied about — names that otherwise wouldn’t rate a mention in terms of ministerial office so close to a critical election campaign.

Rudd himself is showing no sign of being any different to the fractious, unruly Prime Minister he was between 2007 and 2010; his rhetoric that Coalition policy on turning asylum seeker boats back could start “a war” with Indonesia is — plainly put — mad, bad, and dangerous.

And with no pun intended, it seems that in new deputy PM Anthony Albanese, Rudd has an unexpected albatross around his neck; “Albo” might be a fine tactician and headkicker, but he is no salesman and is a PR disaster whose time in the open has come.

I watched “Albo” on Sky News this morning as he battled and stammered his way through murky half-sentences and barely articulated responses to questions (those he bothered to attempt to answer, that is) as he left viewers in little doubt that when it comes to clarity around the objectives of a Rudd government, he will not be the man to provide it.

Rightly or wrongly, the barely coherent “Albo” looks and sounds like a thug, and it’s a point that will take little time to register with voters as the level of his public exposure balloons.

Appearing with Rudd late on Wednesday night, he struggled to even complete his sentences; it may have been nerves in the euphoria of the moment, but his performances since then have been every bit as bad.

And with perception a potent killer of political careers, “Albo” is likely to prove a liability to Rudd in short order.

Of course, it’s early days yet, and there are other polls to come; we will discuss them as they present, and especially in the coming few weeks.

But my sense is that all of the major polls will record hefty surges in Labor support in the short-term; as I said at the outset, the only surprise in the Galaxy numbers is that they don’t show Labor in a winning position.

Australians love an underdog; we are unique in this country in that new governments and/or leaders get a “honeymoon” in opinion polling. Inevitably, however, it wears off.

I think what will happen is that within the next month, solid poll numbers for Rudd and Labor will fall away; Rudd has probably already made the terminal mistake of not showing up to his swearing-in ceremony at Government House with advice of a dissolution of Parliament in hand and an election date in early August to advise.

The longer it takes Rudd to herd voters into polling booths, the greater the margin of the inevitable defeat he will suffer.

The key thing at this stage is for the Liberal Party to hold its nerve and to stand firm behind Abbott; already there is a “Malcolm For Leader” putsch getting underway in social media circles, whose only beneficiary can be the ALP.

We’ve been over this before. Turnbull’s best poll result as leader translated to a 48-52 loss to Rudd; the average was about 44-56. Turnbull as Liberal leader is an electoral red herring fuelled by supporters of the Left as much as by whatever support it retains among voters on the Right.

Even so, with a liability as deputy, his penchant for making ridiculous and pompously grandiose pronouncements intact and a frontbench comprised largely of amateur hands soon to be confirmed, Rudd has a shallow well of political fortune from which to draw.

And this in turn means that any “bounce” he receives from these early polls will be no more than a dead cat bounce; even a cat rebounds to an extent if it falls far enough but — in the end — fall it does, until it hits the bottom of the pit.

The worst thing Liberals across Australia can do, right now, is to panic.

With Polls Averaging 55-45 To Libs And Nats, Where Is Gillard’s Bounce?

THE SMOKE has cleared from last week’s federal budget, and the post-budget polls are out; the reputable ones average a 10-point Coalition lead that would see Labor lose 20 seats at an election. So why the media excitement about a “bounce” for Gillard? The truth is here; perhaps boring, it shouldn’t surprise.

It comes as little surprise that in some corners of the press, there have been loud headlines this week proclaiming recoveries, bounces and comebacks for Julia Gillard in the wake of last week’s budget and the polls that followed it.

There is virtually nothing in the relevant numbers to justify such talk; we’ll go through it, but deluded lefties getting excited about tiny movements within error margins are just that: deluded.

I’m not going to get misty-eyed about primary votes tonight, which in any case are most relevant these days in depicting the split of the vote for the Left between Labor and the Greens; it’s the two-party measure that wins and loses elections — barring oddities (1961, 1990, 1998) — and that’s what I am going to focus on.

Here are the two-party figures from this week’s polls for the four main opinion polls, all taken after the budget; movement is from the previous survey — one week ago for Essential, two for Newspoll, and four for Nielsen and Galaxy.

Essential: 55-45 to Coalition (no change)

Newspoll: 56-44 to Coalition (no change)

Nielsen: 54-46 to Coalition (+3% to ALP)

Galaxy: 54-46 to Coalition (no change)

And whilst I’m not drilling into primary vote movements — some of which involve moves against the Coalition, but not all — all the movements recorded except by Nielsen fall within the margin of statistical error, and are therefore difficult to regard as reliable.

This is especially the case with this particular Nielsen survey; not because it shows a 3% movement against the Liberals after preferences, but because it contains a state-by-state breakdown that just happens to feature a 12% swing against the Coalition in South Australia and the Northern Territory, after preferences, from its previous findings.

This movement has “rogue” written all over it — especially as its other state-by-state numbers are broadly unchanged.

Readers can view the Nielsen numbers here.

The Fairfax papers seem to have lost their rag over the Nielsen numbers altogether though; under a headline “Gillard’s Budget Boost” The Age begins by saying that Julia Gillard has arrested a three-month decline in her standing with voters, to be back level with Tony Abbott as preferred Prime Minister” (although it is at least honest enough to add that “Labor would still be beaten if an election were held now”).

We’ll come back to preferred Prime Minister numbers, because these really are the only basis for trying to portray last week as any sort of victory for the government — and even then, it’s a wide bow to draw.

But I want to stick with Fairfax for a bit, because their presentation of the claim of Gillard getting a bounce has been pretty disingenuous, to say the least.

If you look at this article you’ll see that its “2010 to Present” visual of primary and two-party voting intention is nothing of the sort; it omits all of 2011 and most of 2012 — conveniently, a period in which Nielsen found the Coalition recording two-party leads of 15% and above — commencing instead at about the time of Gillard’s “misogyny” outrage, and coinciding with the temporary narrowing of the Coalition lead in the Nielsen surveys.

The imputation is that the return to wide Coalition leads this year was a short-term blip; and to complete a point, were it not for the wild movement recorded by Nielsen in one state, the Coalition would probably have recorded a 56-44 result in this survey too.

It’s the same story with their “Prime Minister’s Performance/Preferred Prime Minister” visual, too; the worst of her ratings have been omitted.

News Ltd got it rather better.

“Labor Going Nowhere In Latest Newspoll,” headlined The Australian, before going on to say that “Labor’s electoral base of low-income earners has turned against Wayne Swan’s pre-election budget on the grounds of economic management and being personally “worse off” as a result of measures announced last week.”

(Newspoll tables can be found here).

And the Herald Sun and Daily Telegraph were even blunter, stating that “the opinions of Australian voters are now entrenched and not even a Federal Budget promising help for the disabled and education reforms will change their minds, the latest Newspoll suggests.”

So what of the widely reported “bounce” Gillard is supposed to have got from the budget?

Two of the four polls — Essential and Galaxy — didn’t even ask the leaders’ approval and preferred PM questions; Essential, in any case, only does so monthly, and its May findings the previous week showed Gillard’s position unchanged relative to her standing against Tony Abbott.

And a third — Newspoll — found both leaders with very slightly higher personal approval numbers from the previous fortnight, and a 2% narrowing in Abbott’s lead as preferred Prime Minister well within the poll’s margin of error, and with Abbott remaining ahead in any case.

That leaves just Nielsen, with what seems to be at least partly rogue results, and its SA/NT findings seemingly contaminating the rest of its numbers, showing Gillard gaining four points on the “preferred PM” measure at Abbott’s expense for the pair to tie on 46% apiece, with 8% undecided.

And that’s what the entire concocted story of Gillard’s new-found “momentum” is based on: that one survey, when the glaring hole in its findings would cause anyone to wait until the next to see if the 12% swing in SA/NT was replicated or whether it was, truly, just a rogue blip.

Somewhat surprisingly, even the News Ltd papers ran with the Nielsen numbers too, even as they were reporting the doom and gloom Newspoll had found awaiting the government at the polls, and if you click into the Herald Sun link I’ve included, you’ll see exactly that.

Then again, it’s not the sort of caution likely to be shown by those elements in the press pack desperate to pump up Gillard’s tyres, and determined to demonise Abbott at any cost: even if the price is the truth.

I don’t expect to see a 57-43 result at this year’s election or anything close to it, frankly, and whilst I am loath to make predictions (aside from the Coalition winning, and doing so reasonably well) I think 54-46 is pretty much where the votes will end up falling.

It is also probably no coincidence that across the different surveys, and with a bit of back-of-the-envelope calculation and some educated guesswork, 54-46 just happens to be the average of all the polls since the 2010 federal election.

In other words, I actually think the Coalition will do a little worse than the average of these four latest polls suggest it will. Just a little.

Politics is a changing game, and anything can happen; in the past six months we’ve seen Labor and the Prime Minister move to within striking distance of the Coalition on the back of one dodgy speech — only to see the Coalition race back ahead as the memory of that distasteful event wore off in the minds of voters.

It’s just an example, but it’s instructive; and whilst polling day is now nominally less than four months away, there is still time for the unexpected to occur.

Even today, there are rumours of a third and final leadership challenge from Kevin Rudd, on or about 3 June: who knows whether it will happen or what the result will be if it does, but it will change the dynamic irrespective of what happens should it come to pass.

I would urge readers to disregard the bullshit of the past week about a “bounce” for Gillard; no such phenomenon has occurred (which is not to say that it won’t at some point between now and the election). The whole thing has been a media beat-up.

Next week, however, you never know what might happen — that’s politics!