Post-truth politics makes evidence-based policy harder, but it's no reason to give up

By Stephen Easton

August 30, 2017

Last week’s Power to Persuade symposium led to fascinating discussion about how evidence feeds into public policy and the impact of post-truth political culture. Policymakers have always seen multiple truths, and not everyone believes the widely-understood term describes a genuinely new phenomenon.

The truth has always been a slippery concept – especially in politics — but there is growing concern that Australia has developed a post-truth political culture, and reality is now of only minor importance to public policy.

While public servants are supposed to base their advice on an objective reading of the best available evidence, the policy it eventually informs is always influenced by a range of other very subjective factors.

The most senior bureaucrats always have to “keep one eye on the politics” and craft their agency’s advice so it has a realistic chance of being taken up, as last week’s Power to Persuade symposium heard from Tim Johnston, a mid-level manager in the Department of Employment.

Infrastructure secretary Mike Mrdak recently remarked that a position offered by a department could arouse little interest from the minister, but would be readily accepted from an external stakeholder, such as an industry group.

This, he suggested, was a reason to collaborate on shared positions with external stakeholders. Involving stakeholders in the process of gathering evidence also means they are more likely to support the eventual policy.

As Johnston explained,“the further you get up the chain, the more the evidence is tempered by other factors” that governments take into account in decision-making, like budget limitations, their political fortunes, and the competing interests of different groups of people and organisations.

This is an important point about how evidence fits into the policy process. When ministers look out at the world, one can argue they do not see a single objective reality, but lots of different perspectives and claims of fact that appear as truth to different actors but may be contradictory.

Johnston informed delegates from academia and the non-government sector who are keen to inject particular evidence directly into departments that, at best, it would be “incorporated among a range of voices” – it’s unlikely they will they be “in the right place at the right time with the right submission” to catalyse a major change in the agency’s thinking.

The conference enabled exploration of whether there is hope for evidence-based policy amid all the concern about post-truth politics, as well as different types of evidence produced through different research methodologies, and how they are valued by policymakers.

Representatives from NGOs that try to influence policymaking based on their work at the coalface, including the symposium’s co-director Tanya Corrie and Consumer Action Law Centre CEO Gerard Brody, said departmental officials typically prefer robust large-scale quantitative data, but may marginalise or ignore qualitative information like case studies and anecdotes.

These types of evidence are usually more successful for lobbying ministers, on the other hand, and in trying build public support for action through the media, as they generally illustrate how an issue affects real people. But this kind of emotive evidence can also be used by lobbyists to distort the truth, as public health physician Michael Bret Hart argues in relation to iron ore miner Andrew Forrest’s lobbying for cashless welfare in Indigenous communities.

Family Care CEO David Tennant also criticised how influential the Minderoo Foundation has been in social policy at the conference. “If you see Andrew Forrest walking toward the department, lock the doors and pretend no-one’s there because he hasn’t got a lot to offer,” he said.

Are we post-truth? Or have there always been multiple truths?

Opening the conference, former Australian Human Rights Commission president Gillian Triggs expressed concern that truth was being “manipulated in the interests of party politics” more and more each day.

She criticised the recent growth of unchecked executive power and said a symptom of post-truth culture was the government increasingly rejecting evidence-based reports from agencies like the AHRC’s Forgotten Children. She said the situation was getting “closer and closer” to George Orwell’s 1984, where the ruling party of a totalitarian state decides what the truth is and argues that “reality exists in the human mind and nowhere else”.

But another speaker, ANU academic and self-described “policy entrepreneur” Daniel Reeders, questioned whether the post-truth era was a genuinely new phenomenon: “When have we ever worked in a truth-based environment in policy formulation?” he asked rhetorically.

Reeders referred to a famously chilling quote from former White House aide Karl Rove which, coincidentally, showed Rove held a very similar view to the fictional party official quoted by Triggs from Orwell’s dark satire.

“When we act, we create our own reality,” Rove said of the USA in 2004. He argued the journalists and academics in the “reality-based community” were wrong to assume solutions to policy problems emerged from empirical study, arguing “that’s not the way the world really works anymore” in a bombastic assertion of US hegemony.

Daniel Reeders

Reeders believes people have always lived in a world of multiple truths, and gave some practical examples from his own experience in a not-for-profit health sector organisation.

“People – a health promotion practitioner versus a clinician – could believe diametrically opposite things about the system, and they were true within their own local context, within the knowledge practices they were using to engage with the problems they were facing,” he said.

“And that’s how we end up with multiple truths that contradict each other. My argument today is that we have always been working in a multiple-truth environment and unless we engage with that, we are always going to fail to make policy that responds effectively to the problems.”

This is a valuable perspective for everyone involved in the messy process by which public policy is developed. At the same time, however, the worry about post-truth politics is not that the truth is more contested, but increasingly irrelevant.

While the roots of the post-truth era stretch back a long way, the rapid rise of the widely-understood term also reflects a realisation that all sorts of beliefs, from the merely unproven to the utterly irrational, thrive and grow in isolated online communities.

Ministerial responsibility also seems to be declining and politicians can increasingly withstand accusations of misleading the public, at least among some voters.

The many different versions of the truth that live in human minds also seem to have multiplied and become more separated. Triggs pointed to people sticking to sources of information that confirm their beliefs, and the “extraordinary idea” that balance in news reporting means giving equal weight to the views held on these isolated islands of belief, which is a “matter of profound concern” to her.

“What is balance when you’re dealing with the truth?” she asked. That question is becoming harder to answer as the once well-understood distinction between facts and opinions gradually dissolves.

BETA ways of finding evidence

The symposium also featured two speakers from the Behavioural Economics Team of the Australian Government (BETA), managing director Tara Oliver and economics professor Robert Slonim, who is acting as its chief executive for six months.

Robert Slonim

“Not all evidence is created equal,” Slonim said, and this is another valuable point. Public servants need to collect various types of evidence to support government in different ways, but factors like sample sizes and research methodologies matter.

The professor said it was “quite dangerous” to give equal weight to different evidence in an effort to create artificial balance between competing interest groups, and presented a quick summary of his own research into voluntary blood and organ donation rates, which disputes the validity of a supposedly evidence-based policy position.

Richard Titmuss proposed in 1970 that rewarding people for giving blood or organs resulted in worse outcomes than relying on voluntary donations. His theory was hugely popular and remains highly influential to this day, but his conclusions were not based on strong evidence.

The policy was backed up by surveys, but refuted by stronger data about how actual donors behaved, Slonim said. He studied this behaviour for a decade and helped show that rather than volunteerism leading to a more reliable blood supply, “exactly the opposite” was true.

“The most trivial of compensation” like a $5 gift card would increase supply by about 20%, he found, and this work has been “replicated around the world” including through “at least nine massive studies” in different populations.

“There are hundreds of thousands of people today that probably will pass away because of the current system, because there is not enough kidneys, there’s not enough blood, and there’s a lot of work suggesting that actually very minor compensation can solve this,” said the professor.

“This is a policy that’s been around for 50 years and it’s been based, in my opinion, just on either no evidence or very poorly done evidence. So I hope we take evidence seriously.”

Randomised controlled trials on top

Since its inception, BETA has promoted randomised controlled trials as the “gold standard” across the APS, although Tara Oliver acknowledged they “are not always fit-for-purpose for answering all problems” and in practice, often take too much time and resources to test policy interventions before they are implemented.

“But… the great benefit here is they are overwhelmingly becoming seen as the gold standard for evaluation and for evidence, because they are able to test causality and impact … so, we can clearly look at the impact of a policy intervention and measure that against a counterfactual,” Oliver said.

As a lawyer, Gerard Brody noted that in his profession there are always competing versions of the truth and voiced some concern about RCTs being described as the “gold standard” despite Oliver’s caveats.

“When something like that is put out there and known among politicians and senior decision-makers, I worry that it’s going to deflect other types of evidence,” Brody said, and Tanya Corrie largely echoed his comments.

Daniel Reeders argued that having a hierarchy of evidence is all about excluding or giving less weight to some forms and said public health policy had suffered for decades because anecdotal and qualitative information was routinely ignored.

“It’s not to say that other evidence isn’t valuable,” Slonim countered. “My view would be: you have to be cautious of some evidence.”

For example, the professor explained that the research on pro-social behaviour like blood donation shows people are less generous and charitable in real life than they say they are on surveys.

When RCTs are feasible and can be done quickly enough to inform government decision-making, he said it should be recognised that, done properly in the right context, they can provide the most credible evidence to inform policy.

“In terms of what we need to produce evidence-based policy, we need to know what evidence is needed first – that’s not always clear – and we need to know what is good evidence,” Oliver told the symposium.

“We need to think about who creates the evidence and how that might be perceived and valued. We need to know where to find it, and how to build support for it as well.”

Oliver believes the challenges are greater than ever, with higher expectations of service delivery and policy impact in a “post-truth environment” featuring 24-7 news reporting and ubiquitous social media that is tailor-made for appeals to emotion. To “build support” for anything on the basis of evidence is becoming more difficult, but that’s no reason not to try.

“We know that if we don’t bring evidence to bear… the big risk is that we do fall prey to our gut instincts, our best intuitions and ‘the wisdom of the day’ to create policies,” she said.

About the author
1 Comment
Newest Most Voted
Inline Feedbacks
View all comments
Paul Burton
Paul Burton
4 years ago

But… the great benefit here is they are overwhelmingly becoming seen as the gold standard for evaluation and for evidence, because they are able to test causality and impact … Oliver said.
I’m not sure that the overwhelming majority of members of the ‘policy evaluation community’, in Australia or elsewhere would accept this, even if many agree that experimental methods can be very powerful evaluation tools. A review of Ray Pawson’s work and that of his former colleague Nick Tilley will also draw attention to some of the logical flaws in claiming generalised causality without paying sufficent regard to contextual factors.

The essential resource for effective
public sector professionals