I am a policy analyst. My job is to provide expert information to decision makers and the public to help improve public policy. This job, always hard, has become harder.
Seven decades of global economic growth and relative peace have provided great wealth, liberation and enhanced well-being but have also overrun the Earth’s environmental boundaries, created vast inequalities and spawned disruptive new technologies. Transformational change, for good or bad, is now inevitable.
This creates a paradox for expert policy advice: We need experts more than ever. Unaided intuition will fail us as society leaps into the greater unknown. At the same time, transformational change makes communication among experts, decision makers and the public harder. Experts will be caught by surprise by novel shifts, just like everyone else, and consequential policy decisions will often pose questions they can’t easily answer.
Let me offer an illustrative example. About 15 years ago, I participated in a meeting of the United Nations Intergovernmental Panel on Climate Change (IPCC). The meeting I attended brought together scientific experts from many different countries and academic disciplines to grapple with the coming surge of more extreme weather events. The meeting’s location near Brisbane, Australia, was poignant. Just weeks before, the city had flooded. Several years of unprecedented extreme drought had suddenly given way to days of unprecedented extreme rain.
Extreme events often have the most consequential impacts, so these were what most interested experts like me, who study prevention of and response to climate disasters. But experts who study the physical climate were reluctant to opine on extreme events, which in the past were rare and thus difficult to predict into the future with confidence. The physical climate experts were more comfortable focusing on future changes in averages — increases in average temperature, increases in the intensity of the average storm — for which much more data is available. This was a frustration for me and other experts interested in the best ways to respond to climate risk. From our point of view, the physical climate experts know more about these outlier events than anyone else. While their information is imperfect, it’s better than what we have from any other source.
At the core of this dilemma, ubiquitous when experts seek to inform policy, is a focus on predict-then-act analyses. Predict-then-act policy analysis asks experts to generate a prediction of the future on which all the parties to a decision can agree, such as the likely intensity of future storms. These consensus predictions then inform debate on the pros and cons of alternative decisions one could make. Such analysis is a major channel through which experts communicate and collaborate with decision makers and the public.
Predict-then-act works well when experts seek to inform small changes in systems that behave in expected ways — for instance, how many people will lose medical care if premiums increase a few percentage points. But when we face and seek large changes in complex systems, like Earth’s climate or a society undergoing transformational change, prediction-based analysis can disrupt conversations among experts, policymakers and the public. Policies based on predictions can fail when experts are surprised by novel conditions. Insisting on consensus around expert predictions can promote gridlock as parties to a decision attack the predictions, which are in fact likely to be wrong, rather than search together for robust, evidence-based policies. Finally, uncertainty makes people uncomfortable.
Decision Making under Deep Uncertainty (DMDU) is a rigorous participatory approach to giving expert policy advice that aims to resolve these tensions by focusing on informing good decisions rather than making good predictions. DMDU embraces the idea that experts have vital information but are not all-knowing. DMDU rests on two pillars: collaborative co-design by experts and stakeholders, and analysis that focuses not on prediction but on using multiple models to understand the world, learning and robust solutions.
In 2015, the small town of Sitka, Alaska, suffered a fatal landslide during an intense storm, leaving the community suddenly anxious about its safety during the region’s frequent rainy days. Experts suggested a landslide warning system, but it soon became apparent that landslide science was insufficient to support reliable predictions of when the town should evacuate. The community thus used DMDU to co-design a novel landslide warning system aligned with both the science and community values. Sitka’s online risk dashboard now provides hourly and three-day forecasts of landslide risk. Rather than attempting to direct residents to evacuate on command, the system reduces anxiety by giving people the information they need to make the best possible decisions about whether they should stay or go, even in the face of scientific uncertainty.
DMDU has helped the IPCC to resolve the type of tension that surfaced in the previous meeting on climate extremes. Scientists now clearly distinguish between those processes they understand well and those they understand only poorly. For instance, several factors contribute to rising sea levels in a warming climate. Well-understood processes include how much water expands when it warms. Poorly understood processes include how fast the miles-thick ice sheets covering Greenland and Antarctica will slide into the sea. The poorly understood factors are most important for the worst cases. Thus, the IPCC now offers probabilities for the former (the likelihood that sea level rise will reach one metre by century’s end) but signposts for the latter (the possibility that sea levels can rise close to 2 metres, but if and only if specific future events occur). This information helps decision-makers and disaster prevention experts develop adaptive contingency plans.
Embracing co-design and deep uncertainty can help repair today’s broken conversations among experts and those who might benefit from acting on their information. Practising DMDU encourages humility and relevance, enabling experts to acknowledge a non-dominant role in policy discussions, while giving useful information, without overconfidence.
By encouraging thoughtful action, rather than paralysis or willful ignorance in the face of deep uncertainty, DMDU empowers — aligning actions with science, however imperfect, and with community values. By balancing dynamism and learning with democratic guidance and accountability, DMDU can help make good use of expertise as the world transforms. Zócalo Public Square