Possums Pollytics

Politics, elections and piffle plinking

Brendan Nelson – Perpetrator or Scapegoat?

Posted by Possum Comitatus on June 22, 2008

Let me tell you a secret – Brendan Nelson is unpopular.

Are you shocked? Probably not.

Let me tell you another secret – The Coalition’s poor standing in the polls is all Nelsons fault.

Are you shocked at that?  Probably not either, since we hear it everyday. But you ought to be shocked because it probably isn’t as true as its being made out to be.

Undoubtedly the leadership of a party will  impact on the polling results, particularly the voting intention results that a party receives – but to blame all of that failure, or even most of it on the leader is probably a bit of a rough call.

We can see why everything is currently being blamed on Nelson though – the MSM is filled with simple creatures that prefer simple things to talk about, like each other for instance, or each others opinions even more so.

But in terms of the really simple things, there is nothing simpler than very small numbers.

And when we talk about very small numbers in politics, the Nightwatchman’s preferred PM ratings are the most obvious thing to grab on to.

In the politics of polling, there are three variables that take up the mindspace – voting intention, satisfaction ratings and preferred PM ratings.

But unfortunately, the historical, long term, consistent relationships between these variables are rarely paid attention to, so it’s probably worth recapping the statistically significant things we discovered about these variables from last years blogging and analysis.

1. Opposition satisfaction ratings are both covariant (meaning it moves together) with the primary and TPP vote, as well as being a leading indicator of the primary and TPP vote. When Opposition satisfaction ratings go up, on average, the primary and TPP vote moves up with it, as well as moving up in the next period  – not by much, but by a little bit.

2. We also know that Opposition satisfaction ratings are covariant with the Preferred PM rating as well as being an ever so slight leading indicator of the Preferred PM rating.

3.The other thing we know is that the Primary and TPP vote is both covariant with, and a lagging indicator of the Preferred PM rating (meaning that the two measures either move together and/or the Preferred PM rating follows the primary and TPP votes with a slight lag.)

The latter being what started the Poll Wars between The Oz and the better informed blogging community.

The Preferred PM rating is essentially a meaningless beauty contest which has no statistical bearing on the vote. It either moves with changes in voting intention and satisfaction ratings, or lags behind them, and the relationship between the vote and the PPM is pretty tight as far as polling relationships go.

To demonstrate this, let’s make a spiffy chart. We’ll chart the Opposition Leaders Preferred Prime Minister rating against the size of the difference between the governments TPP vote and that of the Opposition. We’ll do it as a scatter plot with a regression line running though it, we’ll use monthly Newspoll averages over the period of the Howard government, and we’ll also mark on the chart where Nelson currently sits in this broad historical relationship (it’s a thumbnail – click it).

Nelsons current position is exactly where we would expect it to be in terms of the size of the governments lead in the TPP vote.

PPM is a function of voting intention, voting intention is covariant with, or a leading indicator of the oppositions PPM rating, and Nelsons current PPM rating sits smack bang on the regression line of the historical relationship.

So all up, there is no useful information here – zip, zilch, nadder. Nelsons PPM is where we would expect it to be with him leading a party currently experiencing a 19 point TPP vote gap.

The satirical coverage that the MSM gives this number is fair enough – it’s at record lows and good for some humour. But to give this number serious coverage is to completely miss the point of what PPM ratings actually are in practice, as well as what their relationship to voting intention really is.

The only marginally important thing about PPM ratings for the Opposition is in terms of it’s compositional make up – things like the the proportion of Coalition voters that prefer Nelson as PM – and even then, it’s simply a reflection of how well Nelson is resonating in his own party and anchored well and truly to the Coalitions vote share.

So moving on from fluff like PPM, let’s look instead at Opposition satisfaction ratings and their relationship to voting intention. If we run a chart the same as above, but this time substitute the Oppositions PPM rating with their satisfaction rating we get (it’s a thumbnail – click it):

As we can see, the relationship isn’t exactly tight between opposition leader satisfaction ratings and the TPP vote difference over the last 12 years, but it’s still statistically significant. The reason for this is pretty simple – new leaders change the satisfaction dynamics in different ways. We can see how that plays out with the following self-explanatory chart (it’s a thumbnail – click on it):

But what is interesting in terms of the relationship between the Opposition leaders performance and the Oppositions vote share is that either Nelson has a higher satisfaction rating than the Coalition vote share would suggest ought to be the case, or alternatively, the Coalition vote share is lower than it should be considering the Opposition leaders current satisfaction rating.

The Newspoll satisfaction ratings come from asking the question:


It is specifically a question on the performance of the Leader of the Opposition.

So the satisfaction/voting intention chart is effectively measuring the performance of the Opposition against the performance of the leader of the opposition.

And on this count, Nelson is certainly performing better personally than the Coalition is performing in terms of their vote share, making me question just how much of the Coalitions poor polling performance is actually Nelsons fault, compared to how much of it is a result of the undisciplined rabble rousing that the rest of the front bench and the back bench pork chops have been carrying on with.

Also worth considering is what would happen to the satisfaction ratings and voting intention if the leadership changed. When we look at the ALP experience over the period in the above chart, if Turnbull or someone else were to become leader, would the consequences resemble Crean, Latham or Beazley Mk 2? – with neither 3 being particularly good.

Crean dragged everything down, Latham peaked and then crashed (and I’ve long had this feeling that Malcolm Turnbull as leader would just be Mark Latham in a Fioravanti suit – but minus the suburban mum bounce) and Beazley had a small boost then a larger decline in the satisfaction ratings, but unusually running hand in hand with a small growth in the primary vote support – but essentially nowhere near enough.

Yet whatever happens, whilst we can all navel gaze over whether Nelson is holding back the party vote share and by how much, the historical polling stats suggest that regardless of what Nelson is doing, the party itself is doing even worse and needs to shoulder a fair amount of responsibility for their own failure rather than blaming it all on one person.

But then, considering the historical way that the Coalition treats their leaders – that’s probably asking for too much. For a party that prides itself on waxing lyrical about the importance of personal responsibility being shown by the community, they seem to have distinct aversity to ever taking any themselves.


And on another thing – I am still doing my age attrition model for projecting the vote share hit the Coalition is facing in about ten years – but I stuffed it up and had to start from scratch. But it is coming!

AddThis Social Bookmark Button

add to kwoff


10 Responses to “Brendan Nelson – Perpetrator or Scapegoat?”

  1. Clear and concise explanation as usual.
    You certainly settle any arguments we have (behind the scenes)at our blog.

  2. bilko said

    Your view of turnbull is spot on rather like Al Grasby and his ties and they are still waffling on about petrol excise on insiders today the pity is the public has moved on

  3. Toby Ziegler said

    Only a minor matter, but why are you so disinclined to use the possessive apostrophe?

  4. Possum Comitatus said

    Toby, I’d like to say that I’m emancipating apostrophes everywhere, but I’d just be pulling your leg.

    I’ll try and get my act together… honest!

  5. JP said

    Possum, this analysis leaves me cold, I’m afraid (nicest way I could think of to say that I think it’s utter bollocks).

    Australian politics has become increasingly presidential: The ALP didn’t run an ALP07 campaign, and Howard positioned himself as the face of his party (compare how often the Libs used the term “the Howard Government” vs “the Coalition Government” or even just “the Government”). Even the Dems and Greens are/were seen as extensions of their leaders personal styles and policies.

    Whether or not the PPM polling moves with a slight lag, it’s clear that PPM and TPP and just both indicators of who people want in government. TPP is more reliable for mine, because PPM has the added noise of “did the leader act like a goose/announce a hard decision/offer a tax cut this week?”.

    From your chart it’s clear that when a party changes to a leader considered better than their predecessor (Crean -> Latham, Beazley -> Rudd) then along with the higher PPM rating, the party (which has usually not simultaneously announced new policy) gets a significant boost in their TPP.

    If Nelson (Mr 15%) was replaced with Turnbull (who would probably rate about 30-35%) Then the Libs could reasonably expect their TPP to jump up at least 5%. If it did, it wouldn’t be because of policy, it would purely be attributable to perceptions of the leader as a capable potential PM.

    And even if Turnbull’s PPM and the Libs new TPP lay right on your regression line, and even if the rise in PPM rating slightly lagged their TPP (ie, if all your stats were completely accurately predictive) then the rise would still be entirely due to ditching Nelson and replacing him with someone more palatable and credible.

    Your analysis above for not replacing Nelson with Turnbull could just as well have been used to justify not replacing Beazley with Rudd. Do you really believe that when that change happened there was something else that dragged up the ALP TPP, and the PPM of their leader just shot up to 60+% because of a lag effect?

  6. Possum Comitatus said

    JP – what I’m saying is that PPM is irrelevant and tells us nothing that the primary and TPP vote estimates dont already tell us. Let us say that PPM was never invented – would we actually know less about the leadership impact on the vote estimate (the only thing that elections are won and lost on) because it was never invented?

    No – we’d know exactly no more or no less.

    It’s a superfluous metric, especially when satisfaction ratings actually do have a leading indicator relationship with the vote estimate.

    At the beginning of the article I said “Undoubtedly the leadership of a party will impact on the polling results, particularly the voting intention results that a party receives”. Leadership matters – but leadership doesn’t seem to be everything, and it’s a huge leap too far to effectively say that the party doesn’t matter at all and that it’s all about the leadership.

    So we have the reality being somewhere in between – that the party and the leadership both matter.

    So looking at the performance metrics of each, the entire party (albeit clouded with leadership issues) via the vote estimate and the pure leadership performance of the Oppositions satisfaction rating – Nelson is doing personally better than the vote is suggesting ought to be the case and/or the Coalition as a whole is doing worse than the leadership satisfaction rating would suggest ought to be the case.

    Which leads to the question of whether the Coalitions poor fortunes can all be sheeted home to Nelson – which seems to be a bit of a dubious proposition.

    Under both Crean and Beazley Mk 2 – the ALP vote was holding up higher than would have been expected with the satisfaction ratings Crean and Beazley were receiving – so the question of those two holding back the vote was a good one to ask.

    But Nelson is actually in the opposite position – his satisfaction ratings are higher than both Crean and Beazley Mk2 ratings were when they were axed, and the party vote is lower than it was under Crean and Beazley Mk2.

    I’m not saying the Libs should or should not replace Nelson with someone else – I’m saying that it looks like the problem with the Coalition is larger, substantially larger than Nelsons leadership alone.

  7. Greetings Possum! I have been following your excellent adventures for some time now but I think this is the first time I have actually commented here (how slack!)

    I have a question: is there anything not forbidden by relevant nuclear-non-proliferation treaties that we psephologists can do to stop these heathens in the popular press referring to leaders’ preferred PM or premier scores as their “approval rating”?

    It’s especially important given that, as you show above, one of these figures (whether it is called “approval” or “satisfaction” rating) actually tells us something and the other is merely a symptom. I can’t remember confusion between the two being *everywhere* in the past, but with the extremely low preferred-PM scores for Nelson, a number of hacks suddenly started calling it his approval rating when it wasn’t. Here in Tassie an epidemic of similar cluelessness broke out when Paul Lennon received a 17% preferred-Premier rating from a local pollster who had previously surveyed his satisfaction rating in the low 30s. Even the advent of a new Premier (preferred Premier rating 46%) hasn’t stopped the preferred Premier rating being falsely referred to as the “approval rating”.

    What will it take to get these dunderheads (and sadly, it’s not just the commercial press) to understand that a person who prefers X as leader over Y may well approve of the performance of both, or the performance of X only, or the performance of neither (even Y only is possible, albeit quite unusual) and therefore a preferred leader rating is *not* an approval or satisfaction rating and should not be described as such? Aaaaaaargh!

    Rant concludes; I feel much better now. 😉

  8. JP said

    Fair enough, Possum. Perhaps PPM’s only real usefulness is as a comparison between two potential leaders (eg polling both Rudd vs Nelson PPM and Rudd vs Turnbull PPM) to gauge the effect of leadership on a party’s potential vote.

    I agree that the Libs have problems than run beyond Nelson’s leadership – that’s painfully obvious. But just because his PPM score is on the line in that first graph doesn’t mean he’s not doing a crap job, because part of the impact of the fact that he’s hopeless is to drag down the Libs’ TPP. If they switched to Turnbull, and *he* was on the line too, but further up and to the left, then that to me (a) shows that people think of Turnbull as a better leader, and (b) seems like exactly the sort of thing the Libs should aspire to.

    I actually think Nelsons *satisfaction* rating is above the line precisely because he’s a bad leader. He’s bland and uninspiring, which means he offends fewer people, and everyone knows he’s just a placeholder, so what’s to be dissatisfied with? Many Lib voters would be happy that he’s crap, because that only brings the gallows closer. Not to mention that Labor voters like myself are satisfied with Nelson because his tepid uselessness is dragging the Libs to places we’ve thought they deserved to go for over a decade – a bit like the “is Australia heading in the right direction?” question. I think Nelson is worse than hopeless, but I’m satisfied with his performance, and even hope he’s leader for the next ten years 😉

    The other problem with PPM is that it’s not only dependent on the person getting the rating, but their opposite number also. Clearly one of the problems for Nelson’s PPM rating is Rudd’s popularity. If Rudd suddenly became embroiled in some scandal, then Nelson could be just as hopeless as he is now, and his PPM would rise anyway – so on that I totally agree with you that satisfaction ratings are superior. As an aside to that, does anyone poll satisfaction with the broad policy platforms of each party? And if not, why not?

  9. Possum Comitatus said

    Kevin – I’ve tried everything from leaving small possum droppings in their favourite slippers through to belly flopping onto their roof about 10 minutes after they turn the lights off at night, and every 5 minutes thereafter.

    Alas – it’s to no avail!

    The nincompoopery of hacks calling PPM ratings “approval ratings” will surely be with us for as long as there are hacks to discombobulate about it.

    However, I do think you’re onto something with it being a relatively new disease – I can’t remember it being ranted about before like this either.

    JP – very occasionally stuff like that comes out.I’ll have a ferret around and see if I can find some.

  10. Hi are using WordPress for your site platform?
    I’m new to the blog world but I’m trying to get started and create my own.

    Do you require any coding expertise to make your own blog?
    Any help would be greatly appreciated!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: