We need a safe space for policy failure

Who remembers Google Schemer, the Apple Pippin, or Microsoft Zune?

No one — and yet such no-go ideas didn’t hold back these prominent companies. In IT, such high profile failures are simply steps on the path to future success. When a start-up or major corporate puts a product onto the market they identify the kinks in their invention immediately, design a fix, and release a new version. If the whole idea falls flat — and who ever listened to music on a Zune instead of an iPod? — the next big thing is just around the corner. Learning from failure is celebrated as a key feature of innovation.

But in the world of public policy, this approach is only now creeping into our collective consciousness. We tread ever so lightly.

Drug policy, childcare reform, or information technology initiatives are areas where innovation could provide policy improvements, but who is going to be a first-mover innovator in this policy area without fearing potential retribution should anything go wrong?

How should government be innovating?

In many cases, the meaning of innovation is unclear. Are we trying to mirror private sector innovation practice? Or do we have an entirely different approach to innovation in the public sector?

South Australia’s fling with Elon Musk is a case in point. The announcement of a giant renewable energy-powered battery is an unexpected circuit-breaker for a longstanding policy issue. It is an ambitious model for energy storage in Australia. This is an innovation both with respect to private-public partnerships and in terms of a new take on energy policy. It fuses private and public sector innovation traditions but also showcases a unique aspect of the public sector: where technology is emerging, governments must often take a risk on large-scale infrastructure projects armed only with a logic model and unscaled evidence. Is this a responsible risk for a state that — like many others — faces financial challenges?

Policymakers don’t necessarily put a product to market. Instead, they take a problem, implement their (sometimes completely untested) ideas about how to solve it, and afterwards find out all their mistakes after sorting through a whole range of unintended consequences.

Public servants don’t have the luxury of ‘making a new version’ without fear of blame or retribution. Critically, their process often lacks the ability to test assumptions before delivery. Silicon Valley entrepreneur and tech intellectual Tim O’Reilly recounts a fitting story from the UK Government Digital Service. After a period collaborating with the service on an evidence-based proposal, a policymaker admitted, “now that I’ve been working with you guys I realise that every policy I’ve put out has been 500 pages of untested assumptions”. It’s an all-too-familiar tale.

In recent years we have come some way in addressing this dilemma: pilots, design thinking principles and co-production of policy all point to a proactive approach. Randomised controlled trials are likewise becoming more prevalent. And yet despite these encouraging signs, the problem remains: there are very few safe spaces for policy failure.

The blame game

This is where the contest over public sector innovation heats up. Interestingly, the key figures who most desire innovation — politicians and the media — are those most likely to blame the public service and policymakers when innovation attempts go wrong. Public servants rightly feel that their room for error is negligible. Periodic elections and diminishing advertising revenues unfortunately provide compelling reasons for harsh and instant critique of any policy misadventure.

Adding to the problem, the current political environment in Australia and indeed globally questions the validity of well-established facts and policy evidence. The most persuasive or entertaining narrative often trumps the painstaking work — and potential missteps — required to build an evidence base to support political and policy decisions. American academics Elizabeth Shanahan, Mark McBeth and Paul Hathaway make a remarkable claim regarding the power of narrative in the policy world: “Research in the field of psychology shows that narratives have a stronger ability to persuade individuals and influence their beliefs than scientific evidence does.” If narrative and stories overtake what we normally accept as evidence, then surely we ought to be taking more notice of what the narratives are, which we choose and how we use them.

But we should not assume, in our Trump-reality climate, that narrative is simply bad. Policy instruments can be used for good or evil. Narrative without evidence is fiction. Narrative informed by evidence plays a different purpose in policymaking. It’s less sexy but more important. The challenge for policymakers is to marry rigour with persuasion. Armed with the right details, they must have the confidence to break new ground.

Of course, failure should only be allowed in certain situations. As Amy Edmonsonargues in the Harvard Business Review: “Not all failure is created equal”. It is not acceptable to have serious failures in policy arenas that touch on the fundamentals of human dignity. Death, human suffering, or threats to vulnerable people are not acceptable risks for the type of policy failure we are discussing here. Instead, safe spaces for policy failure and learning in the innovation space should be possible in controlled environments and with prescribed conditions.

Failing the right way

Essential policy spheres such as health, education and social services should benefit from innovative thinking and theory testing. What is necessary in these areas is even more robust attention to carefully calibrated and well-thought through experimentation. Rewards need to outweigh risks, and risks need to be properly managed. This has always been the case in clinical trials in medicine. Incredible breakthroughs in medical practice made throughout the 20th century speak to the success of this model. Why should policymaking suffer from a timid inertia given the potential for similar success?

An innovative approach, focused on learning while failing right, will certainly require a shift in thinking. Every new initiative will need to be designed in a holistic way, to not just solve an issue but learn from every stage of the design and delivery process. Evaluation doesn’t follow implementation but instead becomes part of the entire cycle. A small-scale, iterative approach can then lead to bigger successes down the track.

Change must be encouraged at an individual level as well as through structural initiatives. Department hierarchies can inhibit innovation and promote risk aversion among those responsible for design and implementation. Naming a new centre or branch may do nothing to foster actual behaviour change. The transition to a proactive, rather than reactive, innovation culture within the public service calls for collaboration and knowledge sharing, and the latitude to think outside the standard brief. Tolerance from the general public for this type of testing requires courage on the part of politicians and the media as well as public servants themselves.

Positive risk-taking and creativity can serve good policymaking when undertaken with integrity and informed by evidence. With encouragement and care successes will outweigh failures. We just need to be understanding when programs trip up a little on the way.

Catherine Althaus is associate professor in the School of Social and Political Sciences at the University of Melbourne and is associate dean (academic) at ANZSOG. David Threlfall is a research officer to the vice chancellor at the University of Melbourne.