Humility is in the air.

A couple of weeks ago, Reserve Bank deputy governor Andrew Hauser upset some economists and commentators when he suggested that everyone should try being less confident and more humble.

Then a week later, US Federal Reserve chairman Jerome Powell repeated a similar message.

"The limits of our knowledge — so clearly evident during the pandemic — demand humility and a questioning spirit focused on learning lessons from the past and applying them flexibly to our current challenges," Mr Powell said.

And last week, Australia's Treasury secretary Steven Kennedy also spoke of the need for more humility in policymaking circles.

What's going on?

Dr Kennedy delivered the keynote address at the University of Adelaide's South Australian Centre for Economic Studies Corporate Members Luncheon on Friday.

The title of his speech was "Evidence-informed policymaking," and here's a summary of what he said.

We need to strengthen democratic governance in Australia

Dr Kennedy had a warning for his audience.

He said our "democratic capitalist system" relied on market mechanisms and trust. But globally, over the past few decades, we've seen trust in government falling.

"And as Martin Wolf, economic editor for the Financial Times, has noted, democratic capitalism and our core Western values of freedom, democracy, and the Enlightenment are at risk," he said.

"As many, including the OECD, have highlighted, we need to strengthen democratic governance to meet our communities' increasing expectations.

"Peoples' voices need to be heard, and we need to rebuild trust in expertise through humility, transparency, and better communication," he said.

He listed the steps being taken by the Australian public service to ensure that more policies will be evidence-driven in the future.

And he explained why it was so important that voters had a reason to trust that policy decisions were backed by evidence.

"There are a lot of policy disagreements in Australia, some of them acrimonious," Dr Kennedy said.

"Evidence provides a constructive path forward. Evidence is a place where debate is useful and resolution possible.

"Evidence-informed policymaking is a deliberate and rigorous approach that seeks to support policy with the best available evidence, rather than intuition, ideology, anecdote, or short-term political expediency.

"Being open to evidence is a subtle way of saying that you are open to persuasion and that you respect those you disagree with enough to hear their views," he said of policymakers.

Can you think of any policies in recent decades that governments have pursued in spite of the evidence?

Computers, large datasets, and the impact on social policy

To that end, Dr Kennedy said data analysis had been transformed in recent years and it was helping researchers to rethink past policy approaches.

He said improved computing power and improved access to data had already revolutionised our ability to derive insights from large and complex datasets.

"By way of example of this transformative power, in 1993 when I joined the Australian Bureau of Statistics, I could have easily spent a day generating a table of data from the census, and I would have had to sit in person at the ABS building to run the program on a mainframe," he said.

"And sadly, I would likely have made small errors in the code and been required to run the request a number of times. Today, the same table can be generated in seconds sitting at a computer anywhere in Australia.

"Reflecting that, the public release of data from the 1991 census had a few hundred tables, whereas the 2021 census release had more than 2.5 million."

He said these datasets were revealing how so many areas of life were connected.

"For example, economic research has increasingly shifted from a focus on the work disincentives of means-tested transfers to consider their broader benefits," he said.

"A clear example of this is Anna Aizer et al (2022), who focus on the major safety net programs for children in the United States.

"That work highlighted how economic analysis of social safety nets has traditionally focused on behavioural responses — such as, work disincentives — which in turn became the focus of policymakers.

"However, better access to linked data has opened the door to examine the long-run benefits of such programs to children — such as educational attainment, and earnings in adulthood.

"This research highlights that once these longer-term benefits are considered, many safety net programs are cost-effective, even if there are shorter-term costs and issues."

He said access to such data, and improved analysis of the costs and longer-term benefits of social programs, was likely to change the way policymakers viewed and designed a range of programs, including health, disability and support for people experiencing unemployment.

If people aren't being heard, their trust in government declines

Dr Kennedy then talked about the importance of listening to voters.

"The OECD recently conducted a survey on trust in public institutions," he said.

"It found Australians who believe the current political system does not let them have a say trust the government 52 per cent less than those who feel they have a political voice."

He said it was crucial for policymakers to engage with a diverse group of stakeholders when formulating policy to ensure that policies were grounded in real-world experiences.

That included working with people who will be impacted by a policy, through each stage of the policy formation process.

"This process not only enriches the evidence base but also builds trust in the policy process," he said.

He then provided examples of policy areas in Australia that were drawing on better evidence bases, such as climate change, merger reform, restorative justice, and the airline industry.

"Over the past 18 months, reflecting an investment from the Australian government, Treasury has built up new capability in climate modelling to better understand both transition risks and opportunities and the physical impacts of climate change, which are becoming more prevalent," he said.

"Another example of data analysis enhancing competition analysis is the work undertaken by Treasury's Competition Taskforce in collaboration with the Department of Infrastructure, Transport, Regional Development, Communications and the Arts on the Aviation White Paper.

"This work has developed a detailed evidence base on how aviation competition has evolved over time and its impact on consumer welfare," he said.

And then he returned to humility.

There's a lot we don't know, let's be humble

"To rebuild trust, we must also be humble," Dr Kennedy warned.

"We need to be humble about what we know and what we don't know, acknowledge the limitations of our evidence, and acknowledge that evidence evolves and we learn more as we go.

"It is worth remembering that it was conventional wisdom that a higher minimum wage resulted in fewer jobs — then in 1994, economists David Card and Alan Krueger used a natural experiment to show that, in the real world, this does not always happen.

He said it was difficult to improve policymaking and it took a "concerted effort" to promote a culture of continuous improvement and learning.

"Established practices and procedures, entrenched interests and risk aversion contribute to the persistence of outdated or ineffective policies," he said.

But collecting and accessing the right data is critical, too.

"Where we focus on the wrong metrics, or don't acknowledge their limitations, they can lead us down the wrong path — what we measure, affects what we do," he said.

"The 2009 Stiglitz-Sen-Fitoussi Report stressed that progress is more than increases in income, wealth, or production. However, a broader focus on defining and measuring progress has remained a challenge.

"Traditional indicators of progress such as GDP have long been the focus of public debate and remain a vital part of measuring progress.

"However, these metrics are not holistic, and they fail to capture the many ways through which society and living standards progress over time," he said.

He said governments and public institutions would always have to make decisions, and their decisions must be informed by the best available evidence.

"When they are, citizens are more likely to trust the policy changes needed for a prosperous and inclusive society," he argued.

"Even more importantly, they are more likely to trust the policy process if it is undertaken in an open and evidence-driven manner, because when it turns out we were wrong, they will accept the reasonableness of being wrong on some occasions," he said.

And in a footnote, he cautioned against becoming too enamoured with sophisticated economic models.

"It is worth remembering what the British statistician George Box said: 'The most that can be expected from any model is that it can supply a useful approximation to reality: All models are wrong; some models are useful'," Dr Kennedy said.