image

A Brief History of Federal Polling

(Warning: this article is very long and contains lots of number crunching relating to opinion polls and election history, as well as hopefully informed speculation about the fate of the current government and the party leaders.  I’ve thrown in section headings to break the monster up a bit for casual skimmers, but if this is sort of stuff is really not your thing at all, read something else.)


In September last year, this column examined something odd about the state of Australian federal opinion polling.  Although the Opposition enjoyed massive two-party preferred leads over Labor, the Opposition Leader, Tony Abbott, was quite unpopular.  Of all federal opposition leaders whose parties had enjoyed such leads in the last 25 years, only John Hewson’s ratings had been similarly bad, and that had been after he had lost the “unloseable” 1993 election. 


Almost a year on and the state of play reported in that article has turned out to be only the beginning.  Labor’s polling position is not much better than it was then, and the party has not for any length of time managed to get its head consistently above 46% two-party preferred.  Yet Tony Abbott, who at the time of that article had recorded his worst ever Newspoll net satisfaction rating (approval minus disapproval) of minus 19%, went on to record still worse ratings in over 80% of the Newspolls since, reaching a new worst of -30% in the late July Newspoll.  (In the current poll he has recovered to -20, but it would be no surprise if he went backwards again.)


Still, with the Opposition riding so high in the polls, there is little pressure for Abbott’s removal.  Any suspicion that Abbott might be dumped for unpopularity alone has proved to have nothing in it so far, partly because there is no obvious replacement and partly because parties see no point in altering what looks like a winning package.  But it’s been easy to think (with the necessary adjustment for hyperbole) that, as Jonathan Green put it, “Tony Abbott is a total dud that everyone hates but he’s going to be prime minister because the other lot are just such an incompetent rabble.”


Attempting to refute this perception, Gerard Henderson said that Abbott is not, like Billy Snedden (or, in my article, Hewson), just another flawed Opposition Leader whose party rides high just because the Government is worse.  Henderson had another analogy in mind:

“But Snedden was not replaced as Liberal leader in 1975 on account of his approval rating. He was dumped because he was a lightweight who did not enjoy the confidence of his parliamentary colleagues. The real comparison is not between Abbott and Snedden but between Abbott and Fraser. In the last opinion poll taken when he was opposition leader, Fraser had an approval rating of 33 per cent and a disapproval rating of 54 per cent - in other words, a net approval rating of minus 21.

Unlike Snedden, Fraser had the ability and authority to keep his fellow Liberal MPs on side as he pursued Whitlam.

Likewise, Abbott has been able to get both Liberal and National MPs behind him as he set about the destruction of a first-term government and, since the election, a minority government. Many commentators thought this could not be done.”

Yes, in the last opinion poll taken when he was Opposition Leader, Fraser indeed had a netsat (net satisfaction) of -21% (here I am using Morgan/Gallup Polls as Newspoll doesn’t go back that far).  But what Henderson fails to mention is that Fraser’s unpopularity was short-lived.  Replacing the hapless Snedden in March 1975, Fraser generally enjoyed netsats in the +20s for his first 6-7 months.  Only in mid-October, with the blocking of supply and a major crisis unfolding, did Fraser’s ratings go into negative territory for the first time.  Fraser recorded netsats of -12, -24, -23 and -21 over four weeks, and that was his entire history of unpopularity in Opposition, since then Gough Whitlam was sacked.  After this, as caretaker PM, Fraser recorded netsats in successive weeks of -19, -14, -2, -2 and then won the election.  An Opposition Leader who was unpopular for just a month during a crisis that paid off handsomely for him is hardly comparable to one who has been in the doldrums of negative double figures for over a year, and who was last more liked than disliked over a year and a half ago.  Henderson was wrong; Tony Abbott is not Malcolm Fraser.

Adding in Morgan/Gallup polling for the years 1946-1985 to supplement the Newspoll sample I used for 1985-2011 in my previous article, I have several comments about the history of polling as it relates to the current situation.  Note that because polls were less frequent in the early days, I am more willing to use single polls to describe a government’s position when dealing with polls taken months apart from each other.  For modern-day polling, a rolling average of several successive polls is more likely to be accurate.  Looking at the Gallup polls from the early days (the days of Dewey Defeats Truman and other such fabled disasters) it is impressive how accurate they actually are, at least in the sense of the matching between polls prior to an election and the election results.

1. Governments Can Recover From Almost Anything

In my previous article I mentioned the escapes pulled off by Keating in 1993, Howard in 2001 and others.  The longer history shows that if a government can poll an appalling rating, a government can, if it is lucky, recover from it as well.  A most remarkable case of this was in early 1953 when the Menzies government, burdened by a recession and the legacy of a vicious budget, polled 39% to Labor’s 60%.  While myth has it that Menzies won the 1954 election because of the Petrov affair, the Coalition’s standing had recovered to equality even before the Petrov business happened.  Fifteen months from the election they eventually won, the Liberals under Menzies were polling even worse than Gillard’s Labor is doing now.  About year out they were polling about the same. 

image

I have graphed this quickly, in a rather crude fashion (note especially that the time series on the x-axis is not actually uniformly spaced).

The other pre-1985 case of an opposition polling a primary of 60 was in July 1975, with Whitlam’s Labor trailing 33:60 on primaries.  Yet, later that year, the landscape changed radically.  In early November, not only had Fraser briefly become unpopular in the heat of battle, but also Labor had come back to a slight primary lead, and Whitlam’s netsat had bounced back from -30 in July and -20 in early October, to a point where he was no more disliked than liked.  Probably that resurgence was unsustainable, but it does show how events can catapault even the most disliked government or leader back into some sort of contention.

There is nothing about Labor’s current polling position by itself that says that Labor can’t win the next election.  Those wanting to write the party’s chances off need to look elsewhere for reasons, but they are not very difficult to find, and here is one of them:

2. Liberals Survive Bad Polling Better

Looking at the history of incumbent governments getting into strife in the polls, which most of them do to some degree, it turns out that both Liberal and Labor governments are equally prone to slumps (the average worst polling position per term for each side is about seven points behind).  Yet the Coalition has won sixteen of the past 25 elections to Labor’s nine.  The idea that this 16-9 imbalance is because Labor was rendered uncompetitive by the infamous DLP split is not very convincing, as Labor often polled competitively despite the split, only to lose the election anyway. 

It turns out that an incumbent government’s election performance (in terms of the percentage of seats won), relative to the low point of polling that incumbent reached in that term, has been much weaker if that government is a Labor government than a Liberal one.  On a simple regression of percentage of seats won against worst polling position,  Keating’s win in 1993 is a rare case of Labor punching above their weight.  1949 (Chifley losing - quite heavily in seat terms too - despite never being far behind), 1996 (Keating thrashed from a no worse position than in 1993), 2010 (Gillard almost losing an election after the party had never seriously trailed during its whole term) are all examples of Labor suffering much worse than the polls appeared to be suggesting.  On the other side, we have Menzies’ win in 1954, Fraser’s huge win from an indifferent position in 1977, Fraser’s win from serious trouble in 1980 and the Holt blowout in 1966 (not to mention 2001), all as examples of the Liberals doing much better than it seemed they would. 

When all these results are considered together, it turns out that the size of the difference in results compared to the party’s worst polling in that term is rather large on a party basis.  Supposing that a given Labor government and a given Coalition government have identical polling histories during a given term, on average the Labor government has performed worse at the election, at a rate equivalent to 13 seats in the current parliament.  (If this pattern continues, Labor is currently on course for about 61 seats at the next election, with only a very slender chance of winning, but please don’t treat this as a serious prediction!)

It is commonly claimed that polls taken well out from an election have no predictive value.  This is an overrated view; they are useful, but adjustment needs to be made for the way that struggling governments (especially Liberal ones) bounce back.  Where polls a year out from an election (about where we supposedly are now) have shown one party clearly ahead of the other, that party has won in 13 of 18 cases.  The misfires are all down to government resilience, and in all cases bar one, that resilience has involved Coalition governments. 

It’s hard to say that there’s something about Labor - apart from a propensity for doing self-destructive things like rolling an incumbent PM who would have won easily - that has caused them to perform worse relative to their polling than the Liberals do.  Perhaps there is not really anything and the pattern I’ve observed just describes the past without predicting the future.

3. Is There A Blueprint For Defeat?

Looking at the six most recent cases of change of government, it is hard to find any general description for their polling histories.  The Chifley government was never more than a few points ahead or behind its opponents, and was doing quite well early in 1949 then slipped behind late in the year and lost, the downturn coinciding with the coal miners’ strike.  (By current standards, the magnitude of Chifley’s defeat, let alone that he lost at all, is quite unusual, since a sitting government beaten 51-49 two-party preferred will quite often retain office, but in those days the distribution of support bases was very unfriendly to Labor).  McMahon’s government was competitive in his first year in office after taking over from Gorton, falling behind badly in his second, through which the Liberals trailed.  Whitlam’s roller-coaster ride in 1975 - way behind then level then being thrashed then briefly level then sacked - has been mentioned.  Fraser’s government trailed, often badly, for almost the whole of his losing term, and Keating’s did likewise through his except during two periods when the Liberal leader of the time was unviable.  In the early part of Howard’s final term, the Coalition and Labor swapped leads back and forth, but in 2006 Labor started to more consistently hold small leads, which turned into huge leads when Kim Beazley was replaced by Kevin Rudd. 

Through the rear-vision mirror, most of the cases in which a change of government occurred, have had in common that once the leaders for both sides for the election were settled, the government did not lead for long if it led at all, beyond “honeymoon effects” in the first few months after an election or change of Prime Minister.  1983 was an extreme case in which the opposition even changed leader just as the election was called and won, although they had been leading for almost the whole term anyway. The exception, 1949, was unusual for geographic reasons.  But finding such a qualified pattern in five cases doesn’t prove this is how it will always happen, and the example of 1954 shows that the pattern can only be seen in retrospect - a government seemingly doomed for all but the last few months can still win.


4. Unpopular Opposition Leaders Have Never Won

The sample that I based my previous article on included just two changes of government.  Extending the sample back to the serious start of approval ratings in 1969, the twenty opposition leaderships (counting different tenures by the same leader as distinct) can be divided into those where the leader was unpopular (netsat worse than -10) for a substantial period (not just for a few weeks during an election campaign or crisis) and those where this was not the case. 

The first category includes Snedden, Whitlam (2nd time), Peacock (both times), Howard (1st time), Hewson, Downer, Crean,  Beazley (2nd time), Turnbull and Abbott.  Four of these leaderships did not make it to the election, and those that did are batting 0 from 7 (though perhaps that’s harsh on Abbott, who could almost count his as a draw).  Of these, before the current situation, only Snedden and Hewson enjoyed large two-party leads. 

The second category includes Whitlam (1st time), Fraser, Hayden, Hawke, Howard (2nd time), Beazley (1st time), Latham, Rudd and arguably Nelson (Nelson was rolled after two -17s in a row, and would have joined the list above had he not been removed).  These leaderships together had a good success rate; five of them resulted in victory.  They also lost five elections between them, but three of those involved rebuilding a party after it had been thrashed the previous time. 

At state level over the last few decades there are some cases of severely unpopular opposition leaders (repeated netsats of -20 or worse) who did win, but remarkably few.  Jeff Kennett and Richard Court are the most prominent examples, but both of those defeated long-incumbent Labor governments that were polling abysmally.  Also, as I noted in my previous EMRS review, recent graphs by Antony Green show that in state elections, the party in power in Canberra usually suffers; Kennett and Court won in 1992-3, when Keating’s government (while soon to be re-elected) was hardly popular.  Mike Rann and Rob Borbidge were also at times rather unpopular as opposition leaders but eventually just managed to fall over the line, again both with the assistance of the Canberra factor. 

5. Most PMs Become Unpopular

There seems to be something addictive about being Prime Minister, strangely enough. Only two non-caretaker tenures have ended with resignation on the PM’s own terms, with the remainder ending by election defeat (8), mid-term change of government (6), rolling by own party (4), death or disappearance (3), resignation under pressure (2) and dismissal (1).  Thus it’s no surprise that of the nine PMs for which approval ratings are well documented, most have had some time in which they were unpopular.  Gorton was removed while he was still more liked than disliked, and Rudd was allowed less than two months in negative-land before being booted.  I mentioned last time that Hawke, Keating and Howard (twice) won elections at the end of terms in which their approval ratings had been awful. Fraser also did this in 1980 (he had been as low as -31 the year before) and spent most of 1977 hovering around -10 before smashing the Whitlam-led ALP. 

 
Gillard’s trajectory most closely resembles McMahon’s, in that both were popular during their honeymoon phase (in Gillard’s case extended by an election win) but not for any time other than that.  This is still a “better” record than Keating’s, however, and he won an election decisively.  While Keating was much disliked for his “bastard” personality, that dislike was often mixed with grudging respect, and he was still seen as an authoritative leader, even if (for many voters) an annoying one.  Gillard has widely been said to lack such authority, though it will be retrospectively conferred on her from all directions should she actually manage to win.

6. Opposition Polling/Leader Disconnect Hits Record Highs

A figure I like to keep tabs on represents the extent to which voters are willing to vote for a party while disapproving of its leader.  This can be measured through a figure which tries to estimate the minimum proportion of voters of whom this can be said - just add the leader’s dissatisfaction score to their party’s 2PP, subtract 100, and if the result is negative change it to zero.

In the late July Newspoll this figure reached its equal highest level since regular approval ratings commenced, at 17 points - meaning that over one in six Australians disapproved of Abbott but would vote Coalition anyway - before dropping back to 10 in the current poll. In the late July poll, the figure also reached its highest rolling average (over four polls with the most recent weighted double) at 14.6 points (now 10.8, which is still very high).  Abbott is no longer merely equivalent to the doomed and defeated Hewson as an unpopular leader of a popular opposition; he has been in uncharted waters on this score.  (The previous record-holder was Downer, not long before he was removed.)

It’s remarkable that this situation has been locked in for so long.  The rolling “disconnect” score has been over 5 for fourteen months now, and over 9 for a year.  The situation we have had now is one that, in the past 25 years and probably much longer, never lasted for more than three months without something breaking.  We have had more of this in a row now than we have had in the past three decades, total.

The disconnect appears on the Labor side too, but not so much of it.  It’s currently running around six points and has been around that range for the past five months, and also at times late last year.  While many Opposition Leaders have recorded high disconnect scores, it is rare for PMs to get above 5-6 points.  Keating was, of course, the big exception, holding office at one stage with disconnect scores about the same as Abbott’s are now.

When a lot of people are willing to vote for a party despite disliking its leader, there are a few different things this could be said to mean.  It could be a negative reflection on the leader, but it can also be read as a backhanded positive to the leader - that tactics which cause him to be disliked are effective in bringing votes to the party anyway.  But it is notable that the greatest leader disconnects occur at times when both party leaders are strongly disliked.  And that is certainly the case now; the combined netsat for both leaders in the late July Newspoll (-63 points) was among the worst in Newspoll history (only twice has it been below -65; it is now at -53).

7. Abbott Only Routinely Unpopular

While Abbott is recording awful ratings (given that his party leads by so much), there is not much evidence that Abbott is a historically unpopular Opposition Leader.  Snedden, Howard (first time), Peacock, Hewson, Downer, Crean and Beazley (second time) all polled worse netsats than Abbott has done so far. While some of them were the subject of successful and intensely negative scare campaigns, most by the end were more comically hopeless than frightening.  While the Left may see Abbott as uniquely scary (more so even than Howard) and seize upon his bad ratings as proof that this view is shared, many who dislike Abbott don’t feel it in such a frightened, vote-deciding way.  An instinctive repulsion towards Labor’s mess has thus far been felt more strongly, and the Opposition being led by a loose cannon with some antiquated views is just the price you pay.  An observation of this not long ago was a comment by a poster calling himself “Roy Orbison” on pollbludger (comment 1552):

“[..] the people I talk to, who range from professionals to the unemployable, would be running at aound 90/10 in Abbott’s favour. Yes, they think he is a fool and yes, they acknowledge he is dangerous but they want Labor GONE. Not just beaten but GONE. G.O.N.E. Forever. They don’t say why and I don’t think they care. [..] But you know what the most interesting thing is? Not a single one of these people ever mentions Rudd. Not. Ever.”

With all that in mind, the following are some comments on the current situation:

All The Trouble In The World

It’s not news to anyone that the government’s push for re-election is struggling.  Pundits can generally be divided into those who think Labor is at least likely to lose the next election and those who think it has already lost it.  I am certainly not in the latter category.

The Gillard government is something that is very hard for many average voters to comprehend.  It has faced a legitimacy crisis, caused mainly by a combination of three factors:

* The minority government, the first at federal level in several decades and a chaotically balanced one at that.  While minority governments have run successfully in most states, only Tasmania and the ACT have experience of minority governments with the Greens specifically as partners.  In Tasmania, that experience has always been divisive.  The perception is that such a government is hostage to the Greens.

* The breach of Gillard’s infamous “There will be no carbon tax under the government I lead”.  Governments often get away with breaking promises, but breaking a promise not to do something that was feared is particularly disliked by voters.  It is not possible to explain away this statement as a “non-core promise” as John Howard used to do.  It is possible to point out, as Tony Windsor did recently, that breaking this promise was part of the price of forming government, and that Tony Abbott may well have broken his own promises for the same reason had the crossbench let him, but it’s doubtful whether anyone is listening.

* The rolling of an incumbent Prime Minister during the previous term on pretexts that were unconvincingly explained and largely insincere.  This has been compounded by the view that his rolling was for nothing.  It has been further compounded by not even admitting an error in the first place, but doubling down on it when given the opportunity to go back.

A government would struggle under the weight of even one of these.  It scarcely helps to reassure those who like their politics much too predictable that the PM ticks none of the common identity boxes for the job (male, married, with multiple children etc).


The government has other problems too.  A seeming inability to sell its successes to the public, a leader who has repeatedly botched her attempts to establish a likeable and consistent non-fake public persona, hostile media, policy failures, scandals and a frequent (though perhaps reducing now) sense of shambles - all these have been covered well enough elsewhere.  A less well recognised problem is the electorate itself - an over-aspiring voter base craving simplicity and certainty in times in which neither is achievable, wanting to live beyond its means and expecting politics to help it do so.  All up, a bizarre government exists in the times that seem to suit it least.  Some major points along the aspirational line are covered in Possum Comitatus’ (aka Scott Steel’s) The Great Unhinging - Revisited, and an important point is that making such a government suffer in the public eye has not been all that difficult:

Thus it was that the more Abbott complained about the minority government and the more Abbott called for a new election — the easiest and most obvious thing for him to do regardless — the more public traction Abbott inevitably gained with these calls. Abbott was seen to be siding with stability, certainty and control — even though he was actually creating most of the uncertainty that was being sheeted home to the minority government. People blamed the government — confirmation bias — because the public had the predisposition to blame the minority government simply because of the uncertain nature of its very existence.

In my view, there is still another problem.  In 1999 Australians voted in a referendum on whether to make the nation a republic.  Thanks partly to the meddling of the monarchist PM of the time, the proposal was trounced, but it was clear enough that most people supported a republic of some kind.  Even though the position of President of Australia was never created, and even though the electorate now says it no longer supports it under any model, its ghost may be heard as you pass by the ballot-box in the actions of politicians who run for it anyway.  Rudd and Turnbull are both like that - ideally suited to being the Napoleonic ego in charge of a small and compliant executive, but round pegs in the square hole of the modern party system.  Yet they are what the electorate keeps saying it wants back.  It is no good to say that the Prime Minister is determined by the party in government, which in turn is determined by the Parliament.  Many voters just don’t get it, and believe they should decide who gets the job.  Even voters who agree that Rudd was lousy still think that it was their call, not the party’s, to get rid of him.

What The Government Has On Its Side

Despite its woes, the government has significant advantages, if only it can make them work in time.  Firstly, strong economic performance (in the circumstances and compared to the rest of the world); secondly, that it has survived at all and managed to get quite a lot of legislation through in trying circumstances; and thirdly, Tony Abbott.  However, voters aren’t yet listening about the first, the trying circumstances are self-inflicted, and for Abbott see item 7 above. 

A fourth advantage, of a very reduced sort, is incumbency.  As peculiar as the nature of that incumbency is (a government that lost its majority at the last election and clings to power by a very bare margin doesn’t have much authority) the government at least has the advantage that it has rolled out some of its major projects.  When the implications of Coalition policies of dismantling the NBN, for instance, are considered under the intense scrutiny of the campaign period, it is possible that the Coalition’s populist rejection of many government schemes will not travel as well as at present. 

In A Bad Position All Moves Are Bad

Every now and then, and this is one of those times, a few fairly benign polls and the odd bad day for Abbott (and he’s had quite a few of them lately) cause Labor supporters to think their party is finally climbing out of the mire and that voters have stopped panicking about the carbon tax, or whatever else is bothering them.  Perhaps this will actually happen this time, but there have been false dawns before.  There is little cause for optimism until Labor consistently polls well into the “competitive zone”, ie well above 46% 2PP. 

If this does not happen within the next few months, the choice of whether to retain or dump the Prime Minister will become a much more acute one.  The problem here is that it is doubtful that the party has any viable alternative.  A fairly seamless return to Rudd would be likely to cause a big bounce in the party’s polling in the short term, based on the rose-coloured view seen in polling that asks respondents how they would vote if Rudd were the Labor leader.  It is possible even that a Rudd-return bounce would put Labor in a competitive enough position to quickly force the Coalition to replace Abbott with a less divisive and dislikeable leader. 

However, Labor under Rudd would face great problems selling its story to voters at the next election.  Not only would it again be unable to run on its strengths in an uncomplicated fashion (a problem which blighted it at the last election) but there would be no way for voters who considered supporting a returned Rudd to know that if he somehow won, he would not just be rolled by the backroom hacks again. 

Also, Gillard supporters in this year’s leadership fight salted the earth against any Rudd return with insults more vicious than those they would usually reserve for even the Opposition - and ready to be used against him by the same Opposition should he recapture the job.  These were partly a release of frustration, partly a very belated attempt to come clean on why Rudd was really booted in the first place,  and partly an attempt to manage public disappointment when Rudd was defeated.  However, I believe they also had a strategic function: to make Rudd unelectable.  The less chance he would have to win an election because of the ammunition against him, the less reason for waverers in the party room to vote for him, and the wider the margin of the leadership vote in Gillard’s favour. 

It would also be too easy to make the case that Labor had rolled Rudd because he was a failure, then replaced him with another failure, and had nothing better to go back to the original failure.  Perhaps there is an advertising genius out there who would know how to mastermind a Rudd victory on nothing but the fumes of populism, justice, soap opera and relief at the removal of Gillard, but it’s something I wouldn’t believe in unless I saw it happen.  I do not think a return to Rudd, or a move to any third leader, will be strategically justifiable until it is absolutely clear that a Gillard-led Labor must lose.  That is not at all clear now, and by the time it might become clear, it will be too late.

Abbott - Not So Safe

Betting agencies currently have Abbott at odds around $1.20 to be the Leader of the Opposition at the next election.  I don’t think he should be that short.  Sure, if Labor remains uncompetitive all the way to the polls it’s unlikely that he will be rolled.  It’s also true that there is no obvious replacement at this stage.  Turnbull, like Rudd, is popular through rose-coloured glasses (his record fall from grace in the Ozcar affair now forgotten) but he is so moderate that re-appointing him would wreck the basis of most of the party’s opposition to Labor over the last few years.  That would hardly be consistent with the idea that the Coalition had the right policies but just needed a better leader.  Joe Hockey, the other former contender, has done little but sulk since losing the leadership stoush he should have won.  Now and then he shows signs of a desire to differentiate himself on policy from his leader, but few that he has recovered from his defeat and built himself up as a contender.  This is possibly because his claims were overrated in the first place. 

However, if Labor ever makes serious inroads, and we start seeing regular polls that are close to 50-50, then doubts about Abbott will quickly become enormous.  He might not be removed immediately, but a few months of that would be very hard for him to survive.  A party that has had the lead for so long can, understandably, panic if that lead should disappear, and even if a replacement leader is not obvious now, one could build up from nowhere rather quickly.  The rise of Fraser is an example of sorts - while not exactly appearing from nowhere, six months before assuming the leadership he was the preferred Liberal leader of just 5% of voters, and even a few months out he was well down the list of hopefuls.

The above was written largely before Abbott’s remarkably bad interview with Leigh Sales (see here).  It’s easy to overestimate the damage caused by any interview seen mainly on the ABC with its primarily left-leaning audience, even one with 87,000 youtube views as I write and probably many more by the time you read this.  However, if this is a sign of things to come in the commercial media as well, then this is another reason to be sceptical about Abbott’s long-term prospects. 

At the moment, many see Abbott as a brilliant politician despite his unpopularity because Labor has polled so miserably badly on his watch.  Actually if Abbott was really brilliant, a government facing all its problems might be polling even worse.  The Liberals are taking considerable risks by placing so much in the hands of a leader who is unpopular.  They would do well to quietly build up someone, even Hockey, as a more viable reserve option, in case a switch is needed.

In Conclusion

There has been a widespread view, based on a long record of dreadful polling, that Labor will certainly lose the next election.  This article has shown that while Labor’s task is difficult, polling shows a long history of government comebacks from seemingly impossible positions, and therefore polling alone provides no reason to write the government off.  There is an equally strong case that it is the Coalition that is flying in the face of electoral history by trying to win office from Opposition with a clearly unpopular leader.  However, it should not be assumed that just because some governments recover, that this one will, and we could yet see the pattern of the last sixteen months - a pattern that in the past has been unstable and short-lived - continue all the way to the ballot box.  All up the next election is looking like a repeat of the last one, in that both parties are making impressive efforts to lose.  They cannot both - well may we say alas - succeed.