Understanding Uncertainty - Decision Making for a New Era

As part of the launch of our new research cycle, The Rift, we’re publishing three articles to introduce the core research strands which make up the cycle: Collapse, Uncertainty, and Renewal. This week we’re turning our attention to our second research strand: Uncertainty.

The importance of understanding and learning to navigate deep uncertainty has never been clearer. Climate realities increasingly outpace our predictions. Geopolitical turmoil and tariffs mean that business-as-usual can feel a distant memory.

In many cases, the decision-making practices of the past are not fit for the realities we are now facing. But how can we better understand our present and make decisions about an uncertain future?

To help us answer these questions, this month’s expert interview is with Mark Workman.

Mark Workman is Director of Foresight Transitions Ltd and an Affiliate Researcher at Imperial College London where he undertakes research on resource systems, energy transitions, environmental and climate change, violent conflict and the processes of decision making under uncertainty.

There are some situations where we can work out the chance of something happening, and other situations where we can’t. Take the example of rolling a dice on a flat surface, if we rolled the dice hundreds of times, we could work out that the likelihood of rolling a 4 (or any other number) is one in six (so long as a fair, six-sided dice is being used on a flat surface on Earth).

But if we rolled the dice on the International Space Station (ISS) you wouldn’t even be able to say when the dice had actually landed, let alone which face it had landed on. How the dice would behave in space is a circumstance surrounded by deep uncertainty.

Because the dice behaves differently on the space station than it does on Earth, it wouldn’t make sense to apply the calculations we made whilst the dice was on Earth to the dice in space – in fact, using those calculations to guide our predictions for the behaviour of the dice in space would be a bad decision, and lead to misguided conclusions.

In the real world, governments and other institutions need to make really important decisions about lots of things – how we try and prevent runaway climate change, how we adapt to climate impacts. Many of these questions are surrounded by deep uncertainty, like that dice in space. And when we make decisions in ways that don’t recognise this, we are in danger of choosing actions that don’t work – or leave us off worse than when we started.

(I hope that wasn’t too complex – i'm envisaging this this five-year-old is a protégé, well versed in the language of complexity and will eventually go onto be awarded the Noble Prize for Physics. They do exist apparently...

Decision making is clearly a competency-based skill in the same way that Maths, English and Science is taught in schools from the earliest stages of our intellectual development, but it’s not treated in this way.

Right now, people who generate data and outputs for decision makers are taught analytical approaches as processes, such as an econometric or cost benefit analysis exercise. Nothing is taught about the strengths and weaknesses of those processes and the impact they have on how we frame and approach decisions.

This means we rarely pay attention to things like psychological impacts, anthropological considerations, cognitive biases, heuristics distorting inference, narrative frames, incentive impacts, decision architectures etc. In reality, these are all the important aspects of decision-making that result in a decision outcome.

On top of this, senior leaders themselves are not taught decision making – they learn it informally and by rote – in non-systematic ways which can reinforce bad practice. There is a lot of bad practice.

This predicament is insane. It would be akin to teaching surgeons all the best practice surgical procedures to undertake an organ transplant without teaching them any human anatomy and what the individual organs that make up the human body actually do. They could follow all the procedures to the letter, but the patient’s chance of survival is reduced. Instead, surgeons are taught both procedures and anatomy - so that they can adapt procedures with the behaviour of the patients’ organs when undertaking the transplant. We need to implement these kinds of practices into decision making.

I am increasingly confident that its importance will be recognised, that it will be included more thoroughly in education, and that it will become integral to future degrees.

There are three:

Firstly, getting audiences to understand what sits in the realm of science and what sits in the realm of philosophy or other disciplines. We need ways to accommodate for gaps in scientific knowledge and for acting under the inevitable uncertainty it holds. This requires humility, about both the limits of scientific knowledge, and about when to stop turning to science to solve problems. Policymakers need to focus on when it is best to look beyond science for ethical solutions, and scientific advisers need to admit that other sorts of analyses must also inform political decisions.

Secondly, there is the need to ask the appropriate questions when dealing with problems which are steeped in deep uncertainty. When science is called forwards to address policy challenges, we need to ask things like: what are the conditions that would affect how our current or leading strategies perform? Under what conditions does our strategy fail to meet different stakeholders’ goals? And are those conditions sufficiently plausible that we should improve on our strategy, rather than trying to predict the known and therefore the unknowable?

Finally, we need to upgrade thinking by undertaking more scrupulous analysis of decisions after they’ve been made, and to treat decision making as a skill we are constantly learning. To paraphrase Adam Grant, there is a need to learn and upgrade our thinking in the same way that we upgrade software on our computers. To do this, we need coaching and reflective processes in which we analyse the pros and cons of our decisions after we’ve taken them. In sport, performance reviews and coaching are commonplace, and a key means by which teams and individuals improve and grow, learn from past mistakes, and build on prior successes. Whilst this process is used for leadership, it needs to be better tailored to the intellectual complexities and specifics of decision making as a practice. Doing this could greatly enrich the quality of decision-making.

The most exciting developments are (unfortunately) a result of some economically and disastrous public policy making and the subsequent public inquires that have been called in response to them.

I am specifically referring to COVID-19 and the Bank of England’s inflation forecast. In these examples, the application of the conventional orthodox set of decision support tools and processes obscured decision makers from the enormity of the uncertainty, complexity and emergence which pervades possible futures. The lack of competency in decision making was glaringly revealed during the COVID-19 Pandemic which had uncertainty characteristics that will become increasingly pervasive in the future operating environment.

Outputs of the first report of the COVID-19 and the Bernanke review of the failures of the Bank of England to anticipate the extent of UK inflation in 2022 – who forecasted 6% when it was actually 11.9% - have come up with the need for Scenarios Focused Decision Making Approaches to better manage complexity and deep uncertainty as main recommendations. These approaches are effectively deep uncertainty approaches. This is excellent progress, and also means others in the finance sector and energy and defence policy space are increasingly willing to adopt these tools and embrace new decision-making methods.

I wouldn’t be so interested in setting up a gathering to discuss deep uncertainty. It’s more important to appear around the tables which already exist, and to help people implement deep uncertainty thinking into their decision making in a way which aligns with their specific values and perspectives.

Whatever their priorities or values, we’d encourage people to do three things:

• Consider multiple possible futures: including a collection of future scenarios – comprising different combinations of uncertainties external to investment decision making – which would be as diverse as possible to properly stress test proposed decisions being sought;
• Seek strategies that are robust rather than optimal: robust strategies perform well and are able to survive across a wide range of possible futures
• Apply novel computer visualisations: avoid the trap of using data to push one decision. Instead, use these to navigate multiple perspectives about values, options, and trade-offs.

I’ve three tips:

Firstly, embrace it; uncertainty works for the upside as much as it does for the downside. Secondly, be humble; consider how you can make sure you’re doing the right thing, as opposed to doing things right. And, finally, engage with others; in the immortal words of Wordsworth: ‘minds that have nothing to confer find little to perceive.’

Find out more about Mark's work at Foresight Transitions here.