The time of his life: reflections from the founder of Australia’s national nudge unit


Taking two years off from his professorial post at Harvard to start up the Behavioural Economics Team of the Australian Government has been a major career highlight for Michael Hiscox.

“It was the happiest professional time of my life,” BETA’s inaugural research director told The Mandarin during a spare moment at this week’s Behavioural Exchange conference. “It was a wonderful experience… the team is remarkable.”

University of Sydney economics professor Bob Slonim took over the role last year for six months, which turned into 12. His contract is up at the end of this month and a new structure involving a four-member academic advisory panel has been announced.

The new advisors are professors Jolanda Jetten of the University of Queensland, Uwe Dulleck from the Queensland University of Technology and Ben Newell from the University of New South Wales, along with Australian National University associate professor Nick Biddle.

“We have to think about where we go next with BETA, but we were doing something exciting and new, which was really trying to set up a team that was bringing new, best-practice research into discussions of difficult areas of policy, with the new behavioural insights bundled in,” added Hiscox.

“We’re doing a few new things all at once – behavioural insights, randomised controlled trials – in partnership across agencies, and sometimes with the private sector, so we did a lot of innovative things really quickly with a small team.”

The excitement was clearly shared by a lot of other delegates at the conference, academics and public servants alike.

BETA, part of the Department of the Prime Minister and Cabinet, staged the event this year in Sydney, where it was first held in 2014 by the New South Wales Behavioural Insights Unit, before the baton was passed to like-minded groups in London, Boston and Singapore. Several eminent international experts commented favourably on how well Australian governments have picked up the idea and run with it responsibly and effectively.

It was also a good opportunity for BETA to launch some new reports on what it’s been doing: assisting the Tax Office to increase compliance with the Deferred GST scheme; working with the Department of Social Services on boosting survey participation to improve program evaluation; and helping the Department of Health nudge GPs who prescribe a lot of antibiotics, compared to their peers.

Full transparency

One standard feature of nudge units like BETA is transparency about their work, and Hiscox says the PM&C team has taken this concept further than any other with the support of APS mandarins.

“One of the things I’m proudest of is that we did something that no other nudge unit, I think even to date, has done — which is that we insisted on full transparency. So we pre-register all our trials; we have a commitment from all our partners that we publish the results.

“Nothing is put in a file drawer; it’s all going to come out, and in fact even before the trial runs, it’s published as ‘We’re doing work in this area.’ We don’t say exactly what we’re doing and when, typically; we don’t want it to mess up the trial. But we say what we’re working on in fairly clear terms.”

Despite BETA providing copious information on upcoming and completed projects, the one exploring ways to get GPs thinking about their personal contribution to breeding antibiotic-resistant superbugs still became a “secret government trial” in the pages of The Australian this week.

Still, this is not academia or the media. Government publishing involves a lot more consideration of how the information could be interpreted — and the interpretation of BETA’s findings can be influenced by the simple fact they involve public servants and carry the PM’s imprimatur.

Well, that escalated quickly…

Most of BETA’s work attracts little mainstream media attention. Last year’s report on a trial of blind recruiting, however, was swept up in the news cycle and fired up debate around the merit principle, APS workforce diversity targets, and the concept of affirmative action.

Hiscox didn’t expect the report to energise strident critics of affirmative action — those who decry so-called reverse racism and supposed discrimination against Anglo-Saxon men — in quite the way it did. Suddenly he was being interrogated about APS diversity policies and fielding lots of questions the study did not even try to answer.

“It wasn’t a trial trying to assess the benefits of affirmative action,” he reiterated. “It’s not making any claim about whether affirmative action is good or bad, or whether diversity is good or bad – we [weren’t] doing that trial.

“We were looking at what happens with introducing de-identification at this one part of the recruitment process, we were not even looking at all [the other stages of recruitment].”

In hindsight, he thinks the report could have emphasised this point more strongly but isn’t sure it would have made any difference. And on the other hand, there’s also a problem with second-guessing.

“The downside for me as an academic is just the caution that is induced then, because we all then have to think about every trial [in terms of how it might be interpreted].”

The randomised controlled trial was looking at one simple intervention that had been proposed as a way to help departments achieve their workforce diversity targets. It didn’t appear to help in that way.

It did find small biases in favour of historically disadvantaged groups, perhaps indicating public servants trying to follow the policies handed down by their secretaries. A year on, the Liberal Party’s federal executive voted to adopt blind recruiting all over the APS, possibly due to its reported efficacy in stopping affirmative action in its tracks.

Either way, it was only one study, so it doesn’t allow any definite conclusions to be drawn. Discrimination or unconscious bias in favour of Caucasian men could well be present elsewhere in the many other steps to getting an APS job or a promotion, too. Looking at the current senior executive service, this would not be a surprise.

On the meaning of merit

A longer and more detailed research paper on the blind recruiting study will soon be published in an academic journal. Despite a lot of hype about the fad around the world, BETA’s RCT is one of very few that have ever tested it.

Others, from Europe, have also found disadvantage to women and minorities appeared to increase with anonymisation. Their authors have proposed a few theories as to why, as BETA did, but overall the body of evidence is small and inconclusive.

The ideal of objective recruitment decisions giving the best candidate the job remains extremely popular, but illusive. Leaving aside unconscious bias and measures designed to reduce it, Hiscox points out that merit can be interpreted in different ways.

This simple view is that the contents of a CV are objective facts, so the best person for the job can be determined based on the applications. End of story. But there is another view, more common in the United States than Australia, that the same qualifications count for more if you had to struggle harder to obtain them.

One could argue it is more impressive for an Indigenous woman who grew up in relative poverty to reach the same level of skills and experience as a white man from a wealthy family. Many would agree there must be something special about a 45-year-old brickie’s labourer who goes to university and graduates with honours.

“Do we call them equally meritorious?” ponders the professor. “Or do we acknowledge the fact that that woman must have overcome tremendous barriers to get to the same point in terms of qualifications on paper?”

Hiscox thinks the idea that qualifications on paper are all equal is flawed: “They signify different things based upon background and structure and society and barriers.”

Technology could end the argument

In search of merit, some big employers have abandoned the focus on qualifications and work experience altogether, in favour of aptitude testing.

A return to the public service entrance exams of the past is not on the cards, but there are a range of new digital tools emerging with a similar kind of aim, such as Pymetrics, Mapped and Applied.

The latter is a spin-off of the Behavioural Insights Team, a consultancy that started life in the United Kingdom’s Cabinet Office, and Applied CEO Kate Glazebrook was among the speakers at Behavioural Exchange.

Typically, these kinds of apps promise to help improve workforce diversity, reduce human biases, cut through statistical noise and allow employers to plug in various other priorities.

Hiscox believes it is a worthy aim to select higher-performing staff by building “a truer test” of merit and actual potential than can be achieved through interviews and selection criteria. This, he reminded us, was demonstrated by the true story told in one of the most entertaining movies ever made about economics, Moneyball.

Not only did the baseball team get an edge over its rivals by using objective data analysis to overcome the biases of the old-fashioned talent scouts — one values the attractiveness of a player’s girlfriend, for example — it also earned the first mover’s advantage.

More to come from this year’s Behavioural Exchange conference in due course. 

About the author
Premium

The essential resource for effective public sector leaders

Can you afford to miss the next briefing from Mandarin Premium? Sign up today.

Get Premium Today