The Voter Choice Project is over, but it remains the largest, most detailed data set on how Australians make complex decisions. In a twitter conversation today it was suggested a blog post covering the insights from the VCP that might be of relevance to the pandemic might be of value: so here it is. Technical and method matters are addressed at the end.
A note before we begin: I freely admit to being our of my depth on how the virus itself should be managed, and will not be makng any comments on that here. These are also just thoughts, drawing on what I know from the last 3 years of research and my unique data set. There are general insights available from the Key Findings document if you are interested in things like the vote decision heirarchy or what happened in the election. In this post I will muse about what is going on now, and some of the things I can draw out of the study to apply to the current situation.
AUSTRALIA IS AN ANXIOUS NATION
If the hoarding didn’t clue you in already, Australia is a very anxious nation. It was before the pandemic hit.
One of the clear themes of the Voter Choice Project is that people were searching for certainty. Between endless scandals, Labor’s big change agenda, and yet another change of prime minister, the average voter felt they were walking in a field of political land mines. At the same time, they were generally concerned for the future: whether it be that they were really worried about the climate, or the drought, or what many believed was an impending recession. Financial security was high on the list of wants, but also system security. They were scared by the Liberal’s ideological commitment to robodebt. They were scared by the Franking Credits policy of Labor. They were scared by the power Get Up seemed to wield, and that big business were getting into a similar game. And the Adani Convoy scared the absolute bejesus out of many, including many not living in Queensland.
There was no reassuring voice in the leadership offerings, so the voters remained scared. The more scared they were, the more they voted for their incumbent member, or conservative (depending on the psychology of the voter). When they past the point of scared to paranoid and were willing to entertain conspiracy theories is the point where they became UAP/PHON/other far right voters, not necessarily because they supported that party’s policy positions, but because those parties were selling a ‘she’ll be right with us’ message which they needed to hear.
Where we are at in the current coronavirus response is still lacking certainty. There’s no ‘she’ll be right’ overtone – indeed that aspect of Australian culture has been blamed for getting us here. Right now, the Government(s) looks scared and unprepared. Which makes the electorate even more anxious.
It’s also creating an upside down social structure: those at the welfare-dependent, extremely vulnerable end of society are relatively calm – they might be able to pay the bills for once with all the extra money floating about, and they know how to navigate the system. The wealthy and educated are being hit hard, and finding they are far too over-extended to be able to reshape their lives into something more modest, certainly not at speed. The disparity is easily identifiable in who is clamouring loudly for assistance: the $140b university sector and their staff on really good money bleating that they’ll be ruined (all the while still working and still getting paid) is a stark contrast to aged pensioners, whose normally very active lobbyists have barely said boo. Leaders of big business want the shutdown ended so they can go back to earning their megabucks, while hairdressers beg to be shut down. Hospitality workers and other gig workers have been well screwed by current events, and I want to be clear here that I’m not dismissing anyone’s pain, but they’re still not as loud or seemingly as panicked as big business in their demands for help.
The flip in social structure is however not unforecastable – a similar inverse sentiment was notable on the drought survey where those who were freaked out by it were those most detached from it, but educated and connected enough to read constant media about it; and in the questions about social connectivity, where long term home owners were more likely to fear their neighbours than the transient renter. Knowledge and wealth tends to bring isolation and fear, while suffering brings togetherness.
PSYCHOLOGY AND SOCIOLOGY IS MORE IMPORTANT THAN DEMOGRAPHICS
I cannot emphasise this enough. The demographics don’t matter. Absolutely, older men are more vulnerable to a bad outcome if they get the virus, but the response of people to the pandemic cannot be predicted or mapped with demographics.
There is no factor – age, gender, education, postcode, or anything else – that can definitively point to what an individual will decide to do in a particular situation, nor how they will decide to do it.
Voter types, as defined by psychology, engagement with influencers (especially friends and family) and media use, can be of value in understanding how people are dealing and processing, and that can be used to develop effective messages and drive behaviour change.
Note I said messages, plural. Absolutely have one clear overriding message, but how you deliver and frame that message must be tailored to something each person will recieve and accept for any campaign to be effective. Think of it like a steak house: everyone is getting the same steak, but each person is getting their steak prepared the way they want it, in the sauce of their choice, and with the sides they want, so they all eat their dinner.
Personas can be of use here – for example, Sally is a 34 year old teacher who is pretty calm, but confused about what she should be doing and how she can help; Dave is a 59 year old banking professional who is very worried that he is going to wind up broke and destitute, and just wants to protect his family. Develop up a persona for each type, and information resources for each type. Sally needs information – check lists, advice of the day alerts, more detailed FAQs on her specific situation and what she should tell her parents and students. Dave needs a hug – reassurance, business leaders saying it will be ok, updates from his financial planner on how they’re taking care of him, constant support from his family, and probably a more pragmatic member of the family to tell him what to do because he will struggle to process external information. Neither of these two types (which are just brief examples, I would think about 20 detailed personas would be required) need alerts on their phone every time someone dies.
Highly targeted messaging, particularly utilising closed or private social media groups and direct messaging (both of which have much greater influence than posts or ads in public feeds, or any kind of broadcast or print advertising or news), and activating people to become advocates within their friends and families for maximum influence, will get the job done.
The other thing that would help enormously for both the lack of tailored messaging and reducing the nation’s anxiety is an end to the constant and very long press conferences. You’d think these politicians had nothing to do sitting around briefing the media for hours on end… Sure, I understand they do not want to lose this leadership moment. So don’t. Appoint a national briefer who can issue a written statement at 6AM embargoed until they face the press pack to answer questions at 7.30AM every day. Appoint a Czar to oversee response and recovery (think something like Elizabeth Warren’s role in the US response to the GFC, or Peter Cosgrove rebuilding the north after Cyclone Larry). Experts can be used, but sparingly, certainly much less than we are seeing all of the CMOs, who quite frankly have better things to do. The rolling and sometimes competing press conferences, with detailed information about every single infection, helps no one: it just feeds anxiety and creates an overload of inappropriate, untargeted information. One briefing a day – this is everything that happened in the last 24 hours, and this is our message for Australians today. The routine of it alone will have an enormously calming effect. Don’t worry, ABC and Sky News will still manage to fill their broadcast day.
ATTITUDES TOWARDS THE HEALTH AND WELFARE SYSTEMS
Gleaning what I can from the health and welfare specific questions from the study, noting that obviously things have changed considerably in the intervening time and the specifics we are facing now were not asked about, it is unlikely attitude towards the health system will change a great deal, while the attitude towards welfare will never be the same again.
The vast majority (82%) already believed our health system was significantly underfunded, I’d say that’s higher now. They questioned the value of the private health system, and in particular the private health rebate. The attitudes towards the PBS and medicinal cannabis, namely the suspicion that something is wrong there but they weren’t sure what murkiness it was, is unlikely to change given the time it will take to get a vaccine. There was strong support for assisted suicide – those numbers might have changed, but in which direction it is hard to tell. Assisted suicide was an ideological issue for many, they wore it as a badge of honour of being progressive/conservative without actually engaging with the issue (this is common, on all sides of politics, and on many issues). As a significant number of Australians have faced their own mortality, and the constant broadasting of the death toll wears on the psyche, a re-evaluation of the value of life and dignity in death may occur for many people.
The welfare system has clearly failed, and its weaknesses are on full display. Welfare is always a tricky issue politically and there is an entire library of reading you can do on welfare heuristics and concepts of deservingness. Both sides deploy this weapon to define who ‘should’ have or have not, implicitly valuing some as more worthy than others in the process. The perception of being on Job Keeper as being better than Job Seeker is most interesting here; I for one would rather be sure the money is in my bank account, but then I am fluent in dealing with Centrelink (back to that upside down social structure again). You will note Scott Morrison regularly uses phrases like “through no fault of their own” when discussing the hundreds of thousands of Australians who will need support during this time. The word ‘fair’ gets thrown around a lot too.
In the welfare survey we were looking specifically at the concept of ‘entitlement’, and the Robodebt issue, and found very mixed and murky feelings. Many were conflicted between wanting to provide an adequate safety net, but also seeing the reason for compliance and the need to stop people ripping the Government off. I suspect that equivocation has gone, particularly for those who have had to deal with the system. I also supect it has gone for those who were slightly more inclined to feel people in reciept of welfare are just not trying hard enough, for example the farmer who said he couldn’t get workers because they’d all gone home to get the coronavirus money… a month before it was being paid.
Some will see the failure of the welfare system as a cause for great alarm, and more heightened anxiety. Others will see it as a temporary glitch, or something to be expected. I personally see it as a fantastic opportunity to completely redesign the whole system, but I won’t get on the UBI soap box for now.
COMPLEX DECISIONS ARE NEVER MADE ONCE
A complex decision in a highly intense social construct, whether that decision be who to vote for or whether to stay at home, is not a decision made once. Like deciding to quit smoking, you need to recommit to that decision over and over and over again. Every minute of every day you remake that decision, consciously or subconsciously. Throughout the study even the most active and engaged partisans occasionally wobbled in their commitment to vote a particular way.
There are many complex decisions Australians are having to deal with right now, and they are new decisions they have to make for the first time, so are cognitively hard and often physically exhausting. The constant reassurance and reinforcement of those decisions is crucial to ensure effective behaviour change.
Sometimes there is an end point. For elections it’s when they vote, for buying a house it’s when you get the keys, and for going to uni it’s when you enrol. But it’s a deceptive end point, really only have a moment of respite, before the decision process begins again. The drive to declare a ‘restart date’ (yes you NRL) or wanting a more positive timeline (thank you Mr Farr) are understandable desires from people and systems struggling under the weight of all these new decisions. But they are deceptive desires: a restart date will not return things to normal. From now on, the ‘do I work from home today’ question will be ever present for many, as will the more simple ‘do I need to wash my hands’, and many other decisions large and small that will stay with us.
I’m sorry, I don’t have any tips for how to get through the hard part of the change curve to where all these new decisions will become automatic ones fuelled by heuristics- the short cuts that allow us to make decisions quickly. But I assure you we will get there.
That is my best insights for the moment. If you have a particular question, message me on Twitter @ktxby.
About the VCP
The Voter Choice Project (VCP) was an extensive study of Australian voter behaviour and vote decision conducted from June 2018 to June 2019. Based on the seminal Columbia Studies (named after Columbia University where they were conducted in the 1940s), the VCP was a panel study designed to capture influence and campaign dynamics in vote decision. Surveys were conducted online each month from June 2018 to March 2019, then weekly for the five weeks of the formal campaign, with a ‘reflection’ survey issued two weeks after the election. Each of the by-elections had their own 3 wave study embedded within the main study.
A total of 12617 Australian voters, most verified by their electoral roll details, participated in at least one of the 108 different surveys delivered in 16 waves, yielding 32,820 total unique survey responses and over 4,000 data points. The largest response to a single survey was the 2079 respondents to the snap poll when Turnbull was dumped and Scott Morrison became the Prime Minister. The only demographic group under-represented was women under 30; the youngest participant was 17, the oldest 89. All 151 electorates were represented.
The study asked participants about different influences and the influence of each on their political opinions: media exposure; social media activity; conversations with friends, family and co-workers; partisanship; political involvement; interaction with the campaigns; and many other potential influence sources. Issue-based surveys on subjects like Adani, drought, and asylum seekers, were conducted on a regular basis, as well as specific surveys on the bigger political parties. The study was responsive to the electoral mood and context, altering questions and issuing specific surveys on events as they happened, including the Townsville flood, Menindee fish kill, Christchurch shooting, scandals and, as noted above, the change of Prime Minister.
The study has many limitations and events that compromised the work. The Double Dissolution election in 2016 had made the timing of the election very uncertain, and required the expansion of the survey period from an ideal 12 weeks to a full 12 months; changes to method to facilitate the greatest retention of participants over such a long period also caused more gaps in each respondent’s data. (That is, we kept people in the study for longer by telling them not to worry if they missed a survey, and allowed people to join at later points, but that means there were gaps. Less than 200 completed the full set from June to June, but enough completed enough surveys for some really valuable insights). The Cambridge Analytica scandal hit just as recruitment for the study began, causing the study to be subjected to a high level of scepticism and unfounded criticism, causing multiple changes to the proposed recruitment and target panel numbers, and the abandoment of planned observation data collection from Facebook. The change of Prime Minister, and six by-elections during the survey period, both complicated the subject matter, and provided opportunities to experiment with questions and probe different aspects of vote decision.
The Voter Choice Project began under the oversight of the UTS ethics committee, but due to an unavoidable change in circumstances became a private study in August 2018. (I later resumed my PhD through JCU but did not resume ethics oversight there). Thus, there will be very little published academically on the VCP, and the data set is not available publicly or for use by other academics. As promised to our participants, the database will never be sold.