We all despair for the predicament poor old Malcolm Turnbull finds himself in (well, some of us do – you heartless Un-Australian buggers that are wallowing in his misery should feel ashamed of yourselves!), wondering why he doesn’t do this or that thing to improve his standings and provide some real political opposition. But maybe there is little Turnbull can do – maybe there’s little Opposition Leaders, in the general case, can ever do to improve their standings. Maybe Opposition leaders are mere gimps to the actions and public perceptions of the Prime Minister of the day?
If we breakdown every Newspoll from January 1986 through to April 2009, and convert the Newspoll metrics into monthly averages (which not only gives us a time consistent data set to work with, but also has the fortuitous side effect of knocking out a large amount of polling noise from the data) – we can run a few charts that highlight how the Prime Minister of the day is really in charge of his own political destiny.
First up, we know that satisfaction ratings are an important indicator of political support – voters that are satisfied with a political leader are more likely to vote for them. But are all satisfaction ratings equal? Does the relationship between the Prime Ministers satisfaction rating and his party’s vote estimates have the same strength as the one between the Leader of the Opposition’s satisfaction rating and their party’s vote estimates?
The answer is definitely no.
To see this, we can take all 281 months worth of our Newspoll data and turn it into a series of scatter plots – where for each month we plot a satisfaction rating against a vote estimate of the polling.
If we look at how the satisfaction ratings of the government and the opposition play out against their respective vote estimates, we get these:
The red lines in the charts are a simple regression line to show the linear relationship.
As we can see, the PM’s satisfaction rating has a much stronger relationship to the PM’s vote than does the Leader of the Oppositions satisfaction rating to their primary vote. This means that improvements in satisfaction levels for the PM lead to much larger gains in primary vote levels than the Opposition can achieve by improving their satisfaction levels.
But what also needs to be mentioned here is that there isn’t a fixed piece of satisfaction pie at play in Australian politics. The level of satisfaction for a Prime Minister has no relationship to the level of satisfaction for an Opposition Leader, which we can see by plotting one satisfaction rating against the other:
That linear line isn’t statistically significant, so when the PM’s satisfaction rating goes up (or down), the public doesn’t automatically become less (or more) satisfied with the Leader of the Opposition as a result.
If we know move on to how the PM satisfaction level plays out against the Oppositions primary vote, another strong relationship pops up. We’ll also do the same for the Opposition leader Satisfaction rating against the government primary vote (just click to expand the charts)
Again, PM Satisfaction as a very strong relationship to the vote estimates for the Opposition, to the point where PM Satisfaction historically explains just over 50% of the variation of the Opposition’s primary vote – as the PM gains higher levels of satisfaction, the Oppositions primary vote starts dropping. Yet on the other side of the equation, the satisfaction rating of the Opposition has a pretty weak relationship to the government primary vote, explaining only 8% of the variation in the last 23 years worth of data.
In political terms, this suggests that while Turnbull could be expected to make marginal gains in the polling for the Opposition if he started acting in ways that increased his personal satisfaction ratings, it wouldn’t be expected to be large gains – certainly not enough to win an election.
Turnbull – like nearly all Opposition Leaders before him – is effectively a slave to the Prime Ministers own personal standing with the electorate. This can also be seen with what is actually the strongest satisfaction relationship to the Opposition’s primary vote – not the Prime Ministers satisfaction rating above, but the Prime Minister’s dissatisfaction rating.
Based on the last 23 years worth of data, with a PM Dissatisfaction rating of 23, the expected primary vote for the Opposition should be 36.2 according to that regression line. It’s currently sitting on 36.5 – bang on where it historically ought to be.
In fact, Turnbull is currently so wedded to the historical relationship between the Opposition’s primary vote and PM Satisfaction dynamics, that we can also use the regression line of the Opposition Primary Vote vs. PM Satisfaction chart shown back up the post a bit, to forecast an almost identical result.
With a PM Satisfaction average for April 2009 of 67%, we would expect the Oppositions primary vote to currently be 35.8. Again, extremely close to the 36.5 where it actually sits at the moment.
Turnbull isn’t performing badly, nor is he performing well – he’s simply a slave, not only to Rudd’s satisfaction levels, but a slave to polling history.
This brings us to the question of just how big a task is Turnbull facing to get the Opposition primary vote back up to a level where they are seriously in contention to win the next election? For the Coalition to win, they need a primary vote higher than Labor’s – which would mean somewhere around the 43 to 44 mark to even begin to be in contention. Looking at those charts above that track the Oppositions primary vote against both the PM Satisfaction and Dissatisfaction ratings, Turnbull needs to knock about 25 points of satisfaction from Rudd and send those people into Rudd’s dissatisfaction column. He has to change the perceptions of 1 in 4 voters – which is no mean feat.
If we change those PM satisfaction dynamics into net satisfaction ratings (satisfaction minus dissatisfaction) and plot it against the Opposition primary vote – it tells the story perfectly.
To be in with a chance of winning the election, Turnbull needs to bring Rudd’s net satisfaction level down into the negatives, about minus 10 – which, as it so happens, is about where Turnbull’s current net satisfaction rating sits.
But while doing that, he also needs to improve his own satisfaction level while reducing his dissatisfaction level. Can he achieve that through the constant negative harping that’s become a hallmark of his leadership so far?
All it’s achieved so far is to drive his own approvals into the ground without affecting Rudds approvals at all – it’s time for Turnbull to try something different. If he doesn’t, he will likely go down in history as the man that took the conservatives to their largest ever defeat.
Einstein has been attributed with describing insanity as “doing the same thing over and over again and expecting different results”. It’s hard to argue with.
The only more disturbing thought for conservatives than their leader engaging in deliberate electoral insanity is the possibility that Turnbull’s polling position actually has little if anything to do with his own actions – for if that’s the case, the future of conservative politics in Australia is wholly in Rudd’s hands, which really would drive them insane.
UPDATE: Andrew Norton takes a slightly different angle on where Turnbull should go from here, while Mark at Larvatus Prodeo argues for the general non viability of the Turnbull approach.
21 Comments
All this raises the question of what an opposition should do. It seems that to really get somewhere you have to do something that will result in the PM/Premier’s approval declining, such is the strength of the relationship between the leader’s approval and the leader’s party’s primary vote.
So the idea of simply being positive and talking about what we should be doing, announcing alternative policies, etc might do something for your own approval ratings but that’s just not enough. You have to talk about the PM / premier, and why they should be disapproved of. (Of course this assumes that voters won’t disapprove of someone simply because they think the alternative is doing a terrific job.)
And yet negative harping is, in my opinion, hopelessly ineffective. It seems to me that when a PM/Premier loses approval it is because they have done something policy wise that people dislike, eg Workchoices. What other options are open to an opposition that must get the PM/Premier’s approval down?
Good article Poss, it suggest some interesting things about the electoral psyche.
To summarise coarsely, it seems to me that the average voter either likes the current PM, or dislike him so much that they want to vote him out. They don’t really care about the opposition, or what they’re doing.
And I guess Malcolm knows this, and that he must lower the PM’s satisfaction rating, rather than even try and lift his own. The media and people with an interest (like us) get frustrated by it and pull out our hair (fur), where in fact we don’t decide elections, so they’re not really talking to us. I’ve already decided my opinion on Malcolm though I guess, so maybe they’re right.
The results seem to be a corollary of the adage that Australian’s are reluctant to change Governments. PM’s seem to be an integral ingredient of this electoral passivity.
The electorate seems to be quite tolerant of Governments and leaders (at least initially).
Opposition Leaders are an interesting point of contrast to the incumbant. But, too often the answer is that, “The job is taken at the moment”.
I suppose it highlights the need for a succession plan. It also proves once and for all that Opposition Leader sucks as a job in Australian politics.
John
I think that trying to talk the leader down is not the strategy – it’s what the Libs have been doing, and it’s not working, is it?
The nature of Australian politics is that Opposition is a long game. You have to be patient and spend the time building yourself and your party up as credible, so that when the s**t does hit the fan, as it always does in the end, you’re there, the safe pair of hands.
History also tells us that the next Liberal govt front bench won’t contain any of the faces that were there under Howard.
So the Opposition should have at least a 5 year plan. Right now, they should be preselecting new members for ex Minister’s seats, and gently easing the old blood out (keeping a few around for another term for mentoring purposes) whilst educating the new guys. They should be carefully testing out potential leaders by putting them in ‘hot seats’ to see how they cope.
In the meantime, they should look for a couple of weak points which run across the government as a whole and keep gently pointing these out. Any criticisms of the government should be based on emphasising these weak points.
They should simultaneously be the voice of sweet reason themselves. All criticisms should be evidence based and sung off the same songsheet.
Oh, and they should sound/look as if they actually are in politics for some reason other than it might look good on their CV or their too lazy to make it somewhere else.
Possum you seem to have statistically proven that oppositions don’t win elections, governments lose them
Zoomster – agree with most of what you have written. The fact is that governments, esp long term ones, base their re-elecetion campaigns on fear of the unknown, ie what the other side would be like if they won. I think an important task of opposition is positioning yourself so that the inevitable scare campaign has little effect.
I like what you say about “gently” pointing things out and being the voice of “sweet reason”. How it is said is as important as what is said. I would like to see opposition front benchers acknowledge that the latest bureaucratic bungle is NOT the fault of the minister. I think the nation is crying out for opposition leaders to say things such as “X is an excellent policy and the PM/Premier should be congratulated.”
I also think that oppositions can set the agenda to some extent – loved how Mark Latham did this with politicians super entitlements. In that case he even got the reform done from the opposition leader’s chair!
The Liberal’s tactic of going negative and personal against Rudd has entrenched Rudd’s personal standing and character with a few valuable rough edges. Nobody minds a PM with a few scars and rough edges. Rudd’s too sweet and plain character needed that addition – he couldn’t have bought better PR for himself.
The same tactics that have helped Rudd and Labor have also tarred the Liberal brand and Turnbull’s standing.
If the Liberals & their media look at these charts will they be stupid and assume all they have to do is more personal attacks on Rudd?
Nobody takes notice of the village idiot and at the moment that is the standing of Turnbull and Co. The more an idiot smears the leader the better the leader appears.
They have to build their own credibility before anybody will listen to what they say or see them as credible alternative and this is what they should have done in the first 12-18 months. They are still at square one.
But the hatred within the conservative media and the Liberal party make them like moths and a candle.
I still think the Liberals and the Murdoch media will get something out of their xenophobia boat people tactics. But long term it will fade and simply leave the impression that Turnbull and co are no different than the Howard era.
Bring out some more candles.
This also implies a tactic that Rudd could use is set himself up for personal attacks from the Liberals and Media. It won’t matter how much it affects his standing it will do more harm to Turnbull and the Liberals.
> improvements in satisfaction levels for the PM _lead to_ much larger gains in primary vote levels
I think you misspelt _are correlated with_.
There’s nothing in these numbers to say what leads to what. There’s no evidence here to support causation. Intuitively, if the government is going well, the PM will be given credit, and vice-versa, and rightly so, and besides, perceptions of both are to some extent driven by the evil mejia, which we all know to be highly influenced by… opinion polls.
What I do find interesting is the almost complete lack of correlation between PM satisfaction and Opposition Leader satisfaction – that runs completely at odds with the way the press gallery usually present things as if politics were some sort of zero-sum popularity arm-wrestling competition where a win for one is always a loss for the other.
On the contrary, it strongly suggests that the public are quite capable of seeing that both leaders are simultaneously good, or that both leaders are simultaneously bad.
I think Ben Aveling might be under estimating what the data is telling us. Sure, causation is a moot point but…First, there is a fairly strong relationship between PM satisfaction and government share of the vote. As one goes up, so does the other. Second, the relationship between oppoosition leader satisfaction and opposition share of vote is weak. The line shown is a bit misleading. Yes, it’s the line of best fit (to use jargon) but the points are scattered all around it and not in an esepcially linear fashion. That is, satisfaction with the opposition leader doesn’t appear to be related to whether the opposition wins an election. The point is that somethng does seem to be going on which may well have a causal explanation. This may well be evidence for the addage that oppositions don’t win elections, governments lose them.
Ben’s right in that this data only shows correlation rather than causation, but if we test the satisfaction ratings of a PM as a leading indicator of the government primary, its statistically solid and explains between 8 and 20% of the variation, depending on how the lags are specced out.
PM satisfactions are a leading indicator, but Opposition leader satisfactions aren’t. This might be worth a nerdy Sunday post sometime to run through the numbers.
Poss, your graph shows that PM satisfaction against Government primary vote regresses to a straight line going up.
Doesn’t that mean that the PM satisfaction against Opposition primary vote is a straight line going down automatically?
In other words, with the first graph couldn’t you determine the other?
Not a criticism, just a question, this post is a keeper. Great stuff.
Don went:
It should be mostly automatic, but the magnitude of the two relationships (the angles of those respective regression lines) should be slightly different since the minor party vote comes into play. Part of the electorate will move from a major party to a minor party (and vice versa) rather than move from one major party straight to the other. So the way the minor party vote starts interacting with the satisfaction ratings of the major party leaders slightly changes the magnitude of the two relationships over time.
This time you do have an apostrophe missing – early on – opposition’s.. and a bit later as should be has.
That aside, could the PM’s satisfaction also be tied up with the public perceptions of the party as a whole.. ie if Ministers are generally well liked and seen as competent, this reflects well on the PM? Conversely, an opposition leader surrounded by bozos and demonstrated incompetent detritus from a recently defeated government is adversely affected by those perceptions?
Kind of how people liked Gough – but he got caned because of the incompetency of his Ministers. Could a rather nondescript leader surrounded by well liked and respected Ministers or Shadows receive a bonis in their approval?
Anyway – the evidence is in – unless things change drastically at Lib HQ – they better get used to opposition for at least 3 terms.. or hope that Rudd does something as disastrous to his approval as WorkChoices was for Howard’s. Perhaps after gaining control of the Senate in 2013/2014, Rudd increases taxes 100% across the board. It would have to be a mistake that affects the majority adversely… bashing the top 10% he could probably get away with with impunity.
David@15:
This is tempting fate. I’ve given up (mostly!) criticising spelling and punctuation because of Don’s Law:
Those who point out spelling or punctuation mistakes are bound to make a mistake of their own in that post, or one soon after.
You have served one up in that same post:
Bonis should be bonus.
I’ve checked and rechecked this post, and can’t find a mistake, but next time there is bound to be one…
Let’s try some smileys:
:innocent:
Nope, :innocent: doesn’t work on this board. evil with colons does. Read what you will into that…
Poss, I can’t find a list of smileys that work on these boards. Is there a list somewhere? I’ve copied and pasted my own list as I have seen them used on this and other blogs in the crikey.com area to a document, but surely somewhere there is a help page?
Poss@14
Sorry Poss, I was including the minor parties as part of the opposition, as opposed (!) to just the Coalition.
17
You missed one.
The greatest problem Mr Turnbull has – and it’s one about which he can do almost nothing – is that the times suddenly ceased to suit him in about October last year. The very fact that the government here can be foreshadowing multi-billion dollar deficits without there being so much as a blip in their poll support confirms this. Such a paradigm shift has only happened once before in my lifetime, when the 70s stagflation blew away the post-war Keynesian consensus and paved the way for the revival of market-oriented Austrian school economics. Before that you have to go back to the Wall Street crash and the New Deal. As time passes the importance of this will diminish, and the normal political cycle will start to cut in again, but for the time being Mr Turnbull has about the same chance of beating Mr Rudd in 2010 as Alf Landon had of beating President Roosevelt in 1936, or Michael Foot had of beating Mrs Thatcher in 1983.