Why is Tony Abbott’s unpopularity as yet failing to drag down the Coalition’ primary vote?

It depends which poll you look at, of course, but no one in Canberra took seriously Newspoll’s 50-50 poll last week. What does seem clear from Essential and Nielsen is that Labor’s primary vote has recovered from the low 30s to the mid 30s, but the Coalition still has a strong lead.

Given how unpopular Abbott is with voters, the Coalition’s continuing strong primary vote is intriguing. Abbott is now less popular than the Prime Minister — indeed, she appears to be widening the gap between them. This is an under-appreciated phenomenon: Gillard is the woman who broke faith with voters on the carbon price, who has been the target of an extended campaign of misogynist abuse and smears from major media outlets and an angry campaign from the far Right, and yet she is, at least at the moment, better regarded than Abbott, who has done nothing beyond criticise the government. Her net disapproval rating in Essential Report a fortnight ago was 19 points; Abbott’s was 23 points; hers had narrowed, his had widened, and she had reversed Abbott’s lead as preferred prime minister.

But no matter how disliked Abbott is, if even 55% of voters don’t approve of his performance, he’d be prime minister with a huge majority if an election were held this week. What appears to be happening is that, for Labor, the wrong voters are growing in their antipathy towards Abbott. The voters it needs to shift back, those who have switched to the Coalition, seem less hostile to Abbott than the rest of the electorate.

Last week Essential ran its semi-regular questions on the attributes, positive and negative, voters perceive in Gillard and Abbott. The results were noteworthy for a significant lift in sentiment towards the Prime Minister, and a slightly smaller drop in sentiment towards Abbott, from when the question was previously asked in March. For example, the number of voters who regard Gillard as “superficial” fell from 54% to 46%; the number who regard her as “a capable leader”: rose from 38% to 43%. Abbott, on the other hand, saw his rating for “understands the problems facing Australia” fall from 49% to 43% and a five-point rise in the proportion of people who regard him as narrow-minded.

The outcomes may have been influenced by Gillard’s bereavement, which might have stifled some of the long-running anger towards her, and by the focus on Abbott’s behaviour at university. But nonetheless the shift in sentiment is worth exploring. Essential’s breakdowns for March and last week show that on personal attributes, Coalition voters were unmoved between March and September on Abbott: his average score across all positive attributes from Coalition voters fell by 0.25 points i.e. basically stayed the same; similarly, his score on negative attributes from Coalition voters only rose 0.5.

Instead, it was Labor voters who turned (further) on Abbott between March and September: they already disliked him, but they scored him on average one point lower on positive attributes and 3.5 points higher on negative attributes.

In short, a lot of the deepening hostility to Abbott is coming from Labor voters, who already can’t stand Abbott, and not from the voters Labor needs to be turning on him, Liberal voters.

It was different with the PM: Liberal voters warmed to her a little: she was up 2.5 points on average with them on positive attributes and down more than six on negative attributes — quite a fall, which at least suggests the PM has some momentum with the voters she needs to lure back. Labor voters gave her a lift of nearly seven on positive attributes and reduced her average score on negative attributes by seven.

Women also stayed reasonably unmoved on Abbott. He stayed essentially the same with women on positive attributes, and only went up 2.75 points on negative attributes. But woman shifted more on the Prime Minister: she rose over four points on average on positive attributes and fell 4.5 on negative attributes. Women consistently rated Gillard more highly on positive attributes and lower on negative attributes on than men, whereas there is little difference between them on Abbott.

This isn’t to say Labor’s campaign of targeting Abbott won’t work. There are some aspects of Abbott that even Coalition voters dislike: 42% of them think he’s aggressive; 39% think he’s arrogant; 32% think he’s narrow-minded; they represent areas where voters will be receptive to Labor messages — one of the reasons Labor has been using terms such as “go the biff” about Abbott, and arguing he has a problem with women. And Liberal voters aren’t as enthusiastic about Abbott as Labor voters are about Gillard: a fortnight ago, 81% of Labor voters approved of the PM’s performance; only 63% of Coalition voters approved of Abbott’s. Coalition voters were twice as likely to say they “don’t know” whom they prefer as PM, compared to Labor voters. There is potential here for Labor; they just haven’t managed to exploit it yet.

Abbott’s difficulty on this is that voters have fairly clear views about him and have had for a long time: he’s been a high-profile politician for over a decade and he’s not the sort of man who leaves people lukewarm in their views. Dramatically shifting voters’ impression that he’s aggro or that he’s too rigid in his social views will be very difficult; the defence of him over the past fortnight has been left to his colleagues, who’ve explained what a gentle, giving soul he really is. But as the Prime Minister is showing, you can begin to recover in voters’ esteem if given enough time.