A comment first

Well, that was interesting.

Obviously, our prediction of hung parliament or close 1-2 seat victory was bang on the money, and the Incumbent v Challenger (IvC) measure wins two elections in a row. That’ll do pig.

The rest of it was pretty much useless – although the Effective Vote measure is going to be saved from execution for nailing a cross bench of 16 (I thought that was ridiculous when we published it).

I regret not doing a poll in the formal election period. Although we knew where the numbers would have moved to – there is always a significant move after the candidates are announced, especially from the other minors into UAP, PHON and the Coalition when they realise they can’t actually vote for the Australia United First Patriots Rights Forever Great Freedom Party*, and from the undecided and Independents column to (usually) major parties. And we did say in the commentary that the numbers would move, but I have learned that some are too impatient to read, y’know, words, and only look at the numbers.

(*This is not a real party name, I’m trying to be funny. Although it wouldn’t surprise me if such a party did exist in some corner of Parler.)

And the numbers weren’t great. The insights of value came from the qualitative side of the research – the quant helped, and fused is always better than either by themselves. But the numbers were just off, I knew they were off, but I didn’t believe in myself or the model enough to do what I should have been doing. And for this I need to apologise.

There was 6 years of work to develop this poll and try and find something better than 2PP. Aside from the 22 unique question features, one of the really critical things that made the KORE Poll different was the multi-source sampling. This includes:

• intentional over-recruiting of low information, undecided and non-partisan voters into the KORE Panel (which yes, does kick up a higher minor party and independent vote, and thus by definition this poll should be a point or two off other polls and move around a bit more – but this reveals momentum which was to me is more important);
• one public stream which is more about brand building and rewarding participation than recruitment (in English: people get warm fuzzys from being able to share the survey they just participated in, so let them. Very few people actually do the survey from the publicly shared links but the brand awareness makes them more likely to respond to an ad or email should they receive one); and,
• three (or more) semi-public targeted recruitment streams. These involve highly targeted advertising, mostly on social media, to reach all kinds of tightly curated demographics and ensure a really balanced sample across age, gender, every electorate, and targeting any specifics that were being under-represented; as well as giving us the power to tinker with those targeted ads on the fly in response to who was responding to the poll.

Purchased samples and email lists were specifically discarded as not being good sources of respondents for the poll, and incentivising (offering prizes or gift cards) was also found to result in poor quality responses.

The relentless attacks and criticism – including personal harassment of myself and the people I managed to get to work for me, especially the long-suffering Lia who looked after the KORE email – eventually led to me cowering in a corner in a ball of anxiety. While I had done the work to know that there was absolutely no difference between someone receiving a highly targeted social media ad inviting them to participate in a survey, and say an email inviting them to participate in a survey (other than a significantly better response rate and a much lower cost of acquisition), I let the blatant lies and assorted allegations get to me. Midway through January I caved to that pressure and abandoned the science in favour of what the bully boys told me I ‘must’ do, paying an awful lot of money to get brain dead respondents provided by survey companies and purchased email lists, most of whom provided no insight and just ticked boxes.

And, shock, when we used the same recruitment methods as other pollsters, we got the same Labor bias as they do in their polls. By February, I’d stopped doing the momentum tracker, which was the nerdy bit I really loved and had the most potential to really upend the whole polling game, and had abandoned a number of the KORE Poll’s survey design features. By March, I had to be coerced into continuing with the poll at all. I did what they demanded, the attacks didn’t stop, and I’d compromised the numbers.

I am so sorry.

A session with Maniferra reminded me why I was doing this in the first place, and the importance of diversity in polling for a healthy democracy, making it clear what I had to do next.

The below is my last KORE Poll and will be the last KORE Poll until appropriate funding and a new research director can be found. I believe in the power of Lazarsfeldian methodology and Columbia voter behaviour theory, and the importance of continuing to find something better than the 2PP, it just can’t be me to take this important piece of work home.

I want to take this opportunity to thank Lia Edwards, who finished up with KORE CSR last week, all of the members of the KORE Panel, and everyone else who has helped KORE along the way.

~ Raphaella Kathryn Crosby

---

KORE Poll 8 – Post Election Survey
Collection Dates	May 21-24, 2022
Format	Online
Sources	KORE Panel, Snowball, Open social link
Total sample	1486
Weighting	age, gender, 2022 vote
Effective sample after weighting	1257
Confidence/Margin of error	95% / 3%

The important bit.

I think it is critical that we start with the most important bit. 45% of Australians are being denied access to a Democracy Sausage at the polling booth. Some did explain that they were at pre-poll booths, but generally this is a travesty that must be addressed. Big celebrations for those who were able to get the vegetarian or halal sausages at the polling booth this year!

The proper placement of the onion on the Democracy Sausage is of course a perpetual question that may never be solved. A majority of Australians clearly feel that on top is the correct way to do it, but I have to say I was swayed by a number of people hiding in the ‘other’ box that in a Democracy Sausage, the voter should be able to choose their onion placement.

Ok, ok. Enough fun.

Relief.

The overwhelming sentiment from voters about this election is that they were relived. Relieved the election was over, relieved the Morrison Government was done, relieved to have had their vent.

There was a great deal of strategic and protest voting by the looks, either to ‘get rid’ of the Morrison Government, or to “send a message” – both to the Labor Party that they have not been Progressive enough or to the Liberals that they have gone too far to the right. This protest element should put both the newly elected Government, Independents and Greens on notice that their wins are not that wide and barely an inch deep. Any win on the back of protest votes will require a great deal of work to convince electors you are worthy to stay in the gig. Also: the next election could be equally as chaotic as this one.

“I do not think the Labor Party has been noisy nor progressive enough. I knew enough to know that my preference would flow to Labor.”
75 and over male Greens voter in Bendigo

“In a Liberal safe seat this was the only realistic way to get rid of a member of the corrupt, amoral, bullying Liberal/Coalition govt. and Falinski was useless.”
64-75 year old female Independent voter in Mackellar

“To send a message that I didn’t appreciate how far to the right the Liberal Party had moved. Also to relieve Scomo of his responsibilities that he was so reluctant to accept.”
45-54 year old female Labor voter in Whitlam

“Tactics. Labor was never a chance of winning the electorate, but by voting Independent, preferencing Labor, the votes go in the right direction.”
45-54 Female Independent voter in Monash

“Coalition has moved too far to the ‘right’, that started under Howard. Also their anti-social and unethical (if not corrupt) behaviour towards less fortunate Australians (e.g. Robodebt, Indue Card).”
55-64 year old male Independent voter in North Sydney

“I prefer Labor but I know I live in a safe Labor seat. I was going Hoping that if more votes put the greens first they may notice the importance of environmental issues. The floods and fires were/are terrible.”
45-54 year old female Greens voter in Lalor

and there was a fair bit of this…

“No Teal candidate so Green was next best option to send a message to Labor and more so the LNP.”

And that, my friends, was a 65-74 year old male cattle producer in Oxley – aka Pauline Hanson’s old seat in Queensland. I kid you not. And you thought this was all about professional women…

At about this point there’s probably some people that were paying attention in Politics 202 stroking their imaginary beards of wisdom and pondering if this was a ‘Realignment‘. There is little in the data to support this. There was chaos – as predicted. Nominally Labor voters protested by voting Independent or Green. Nominally Liberal voters protested by voting Independent or Labor or PHON or UAP. Nominally Green voters strategically voted Labor or Independent. Supporters of Independent candidates voted Labor because they wanted the change of government more than they liked the Independent.

This was the electoral equivalent of a massive scream of frustration: voters were sending all kinds of messages, none of them you could write on a nice card to your grandma, and to try and infer any kind of collective message other than ‘the voters are unhappy’ would be a stretch.

Break it down.

The following is a bunch of quick graphs to compare party votes by a bunch of demographic things – gender, education, financial situation, and then how comfortable they are with who they voted for and how much they care who won.

No surprises that the Greens and ALP have a slight female advantage while UAP, PHON, other minor parties and the Coalition have a male advantage. That non-binary block in the ‘other minor’ isn’t an error by the way, it’s non-binary identifying people voting for smaller progressive and socialist parties like Victorian Socialists, Fusion, or Reason. The surprise? 52% of Independent voters being male. Unfortunately the sample isn’t big enough to do a teal/non-teal split, I’m sure that can be done with bigger data sets in the future.

Education has been one of the factors (along with home ownership and length of time in a community) that can be predictive of voter behaviour, and there are both some substantive differences between parties and between these numbers and 2019. There’s a pretty clear break now with very few left wing voters not having done some post school education and more than 60% of ALP voters have attended university. The Coalition’s postgraduate figures have dropped significantly from our results in 2019, aligning with the other indications that it was professionals who abandoned the Liberal Party for Teals in the city seats.

Self assessed financial situation is often a more reliable measure than income ore employment as it is a measure of financial stress, and thus an indicator of how much finances are impacting on emotions, which in turn impacts on voting decisions. Unsurprisingly, Coalition had the largest proportion of voters doing well and saving money, and the smaller protest parties had more people who were struggling. 

Unsurprisingly, so close after voting there is a high level of comfort with who people voter for, with Coalition and PHON/UAP voters being the most likely to be uncomfortable with their vote choice.

Also unsurprisingly, PHON, UAP and Coalition voters were the least invested in the outcome of the election, and ALP voters the most invested. 

Get on with it.

While Climate Change was a massive factor in this election in more ways than one, it would be overly simplistic to think that it was a decisive issue or that it was the only issue.

Detailed analysis of the language used by voters themselves to describe what motivated their vote revealed that it was usually a combination of issues – including the climate change and refugee combination we picked up in the March poll – and it wasn’t so much the issues, but that people were sick of the issue being an issue. They were sick of the lies, delays and excuses – they want things to be dealt with.

“ALP must adopt ‘no new coal or gas’ extraction or power plants immediately. ALP need to be dragged to do better.”
45-54 year old male Greens voter in Grayndler

“If the major parties don’t understand now that the electorate is crying out for action on climate change and integrity – they never will.”
45-54 year old male voter in Macnamara

“[my vote was influenced by] a sense that the LNP has been hijacked by men who “play” politics, rather than work for the good of Australians.”
65-74 year old female Greens voter in Higgins

When we look at the key policy issue profile, the economy still came out as number 1, and a federal ICAC, Aged Care/Health were on level pegging with climate change. Refugees and the Uluru Statement from the Heart, NDIS, women/gender equality as well as Aged Care as a separate item dominated the ‘others’ and probably should have been in the pick list – alas, you can’t have everything in the list.

Over before it began.

The analysis of influence factors showed that almost everything done in the formal campaign had little to no effect.

Barely a third of viewers watched the leaders debates, and 90% of the people who watched the first one watched all three. (The next meeting of Politics Anonymous will be held on Thursday at 10…). Only 1% of viewers switched their vote on the basis of the debates.

There were very few key moments in this campaign, particularly for the Coalition with no major TV appearances like Q&A or the National Press Club. Not to worry though, you’ll get saturation coverage for tackling a child. 

When asked to identify which factors they felt influenced their vote, a third said events before the campaign, by far the largest factor. 

The final chance measure of the How to Vote Card is an Australian tradition, but there is always much debate about whether it actually helps deliver any votes or not. The level of media fixation on HTVs and the preferences parties put on their flyers always exaggerates their influence… which is arguably pretty minimal. Over half of all voters either don’t take, or don’t have on offer, any HTV. A further 17% take them, but don’t follow them.

Of the 25% that take them and do follow them, they are mostly only Labor and Coalition voters, meaning that minor party HTVs are a waste of paper. However, it is worth noting that many of the ‘others’ noted downloading or researching online, or devising their own HTVs on apps or websites before going to the polling booths. Perhaps instead of killing unnecessary trees in a futile attempt to staff booths, minor parties could provide a service and advertise at the point of click?

Hopes and Fears.

Mixed feelings about the next government – although remember this survey was done long before the result was clear. Or at least, as clear as it is now. Which is *checks the counting in Deakin and Gilmore* is not exactly crystal…

The undertone to the comments is one of hopeful skepticism. Or more simply, ‘don’t stuff it up’.

“Labor better not f*ck it up. They need to deal with social inequality, implement a fully-fledged and retrospective ICAC, fix medicare, deal with inflation, fix housing affordability and fix supply chain issues…”
35-44 year old male voter in Blair

“Don’t stuff it up. This is a chance to realign politics in Australia permanently.”
55-64 year old male voter in Indi

There is significant concern about the media and what they will do…

I hope Albo can bring about a consensus and we will see some real meaningful change for Australia. My biggest fear is we will see a repeat of the damage caused to the Gillard government by an extremely hostile Murdoch media and a sickeningly aggressive opposition (albeit much diminished!).
35-44 year old female voter in Fenner

“We have to prove the journos wrong, delighted as I am with the result this has to be more than a protest vote, people must stay engaged, democracy works best when people are engaged.”
45-54 year old female voter in Whitlam

“My fears are about Labor being constantly attacked by media and feel they won’t get fair treatment or hearing. With Murdoch against Labor they have an uphill climb all the way and unless something is done about Murdoch’s over representation in media outlets Labor will only have one term again to try and fix the mess we are in.”
55-64 year old female voter in Spence

And even those who don’t know him or like him are willing to give Albo and his team a go, maybe…

“I don’t see Albo as a personality that I like but I think given the chance he’ll do a good job..I hope.”
55-64 year old female voter in Ballarat

“Fingers crossed the team approach will be effective versus the one man band Morrison who had no team.”
65-74 year old male voter in Blaxland

Fingers crossed indeed.

Thanks again to everyone who has participated in the KORE Poll for helping us to try and improve political polling in Australia.