Embracing uncertainty for better decision-making

The World Bank
Laura Tuck, Julie Rozenberg

Photo Credit: Nick Page via Flickr Creative Commons

This post was originally published on the website of the Society for Decision Making Under Deep Uncertainty.

We all face uncertainties.

What if the train’s late? What if it rains? What if traffic is bad? What if there’s a shift in government before the project starts?

Every day we’re hit by all the “what ifs” especially in our line of work at the World Bank Group, whether in the field or within our organization. But how do we best cope with this? Embracing uncertainties may be the answer.

The World Bank Group has been at the forefront of mainstreaming new methods to deal with uncertainties. In fact, you may not know this, but the World Bank is one of the founding members of the Society for Decision Making Under Deep Uncertainty.

Today’s decision makers face conditions of fast-paced, transformative, and often surprising change. Traditional decision analysis relies on point and probabilistic predictions. But under conditions of deep uncertainty, predictions are often wrong, and relying on them can prove costly and dangerous. Fortunately, new methods and processes now exist to help decision makers identify and evaluate robust and adaptive strategies, thereby making sound decisions in the face of these challenges.

You may wonder about the difference between what we call “deep uncertainty” and “risk.” The difference is whether you can characterize what you don’t know about the future through a probability distribution. For example, the likelihood of experiencing a car crash next year in the United States is easily calculable from ample historical data. But many future conditions cannot be reliably quantified, because the mechanisms involved are too complex to be modelled or because the parameters of the model keep changing unpredictably.

This distinction is critical for infrastructure projects, which need to be resilient to change over time. Risk is something engineers have managed for generations, designing roads, bridges, and other types of infrastructure to handle a wide range of water levels, wind speed, and other weather conditions. They typically base their design on decades of meteorological data that show how weather conditions vary, but remain within fairly well-identified boundaries. But things are changing.


Ch-ch-ch-changes

Climate change is also changing traditional approaches in how we do our work. With climate change, weather patterns are evolving in ways we don’t understand and can’t predict accurately. This introduces “deep uncertainty” and interferes with the familiar boundaries of conditions for which we would traditionally design projects.

Each year we invest billions of dollars in infrastructure. We expect to leverage even more private finance with the help of partnerships such as the Global Infrastructure Facility and the International Finance Corporation and the Multilateral Investment Guarantee Agency.

But how can we sustain development gains if projects can’t adjust to long-term shifts in demography, technology, or in socioeconomic or environmental climates? Most scenarios forecast a rise in sea levels after 2100. New scenarios show a possible increase by several meters in 2050. The deep uncertainty regarding the impacts of climate change is even more obvious on rainfall patterns: In many places around the world, particularly in West Africa, models disagree on whether rainfall will increase or decrease with climate change.

You can’t always get what you want

We expect experts to give us reliable projections of the future. But some variables are not predictable, no matter how sophisticated the models.

• Hydrologists can no longer predict rainfall patterns and river flows because of climate change.
• Transport specialists know their projections of traffic demands are usually incorrect.
• Economists know they can’t forecast economic growth with much accuracy.

And they should not be blamed for that. Acknowledging that some future conditions cannot be predicted helps us make better decisions. Making decisions based on one projection of the future is indeed costly if this projection turns out to be wrong.


Bridge over troubled waters

Fortunately, new analytical methods are being developed to cope with uncertainty and ever-changing conditions.

This is where the methods of “Decision Making Under Deep Uncertainty” come into play. We focus on what can be done, not on what could happen, opening the door to a myriad of possibilities.

Decision Making Under Deep Uncertainty:

• Helps us agree on decisions we have to make, rather than assuming what the future holds; and
• Explores future threats as well as opportunities possibly confronting each plan.

Somewhere, over the rainbow

With this approach, experts can explore and quantify hundreds of thousands of possibilities instead of predicting the future. Models can be used to find mechanisms and combinations of conditions that lead either to success or failure.

More importantly, “Decision Making Under Deep Uncertainty” builds consensus, even among people or groups with different value priorities and diverse views as to whether an event is likely to occur (climate change is a very good example).

We create gridlocks, however, if we always rely on experts to give us the “right answer” about the future. With our new approach, people can agree on a plan for very different reasons, as stakeholders get comfortable with the unknown and more confident in their choices.

The World Bank is already at the forefront, mainstreaming “Decision Making Under Deep Uncertainty” and applying these methods successfully to many World Bank projects:

• We helped Lima refine its water supply master plan and identified necessary investments and components to achieve water reliability by 2040, regardless of future climate and demand.
• Using the Decision Tree framework, we assisted Nepal in prioritizing hydropower investments in the Koshi Basin, identifying robust investment portfolios amid uncertainties: glacier melts, electricity prices, and export opportunities.
• We helped Peru prioritize investments to reduce road networks vulnerable to natural disasters.
• We supported Sri Lanka in assessing the costs and benefits of preserving urban wetlands in Colombo despite varying conditions in the housing market and climate change.
• Presently, we’re aiding Mexico and Lima in the water sector; Mozambique with investments in rural roads; and Tanzania (Dar es Salaam) in the transport sector.

The majority of these projects were made possible by the Korean Green Growth Partnership, which has supported us for the past three years. The KGGTF has recently renewed and increased its funding to help us apply these methods to new operations in water, hydropower and transport.

One thing is certain, you will be hearing more and more about “deep uncertainty.” Learning how to embrace uncertainty gives us all a different perspective on the future, and hopefully we can all avoid “future shock.”


The 4th Annual Workshop of the Society for “Decision Making Under Deep Uncertainty” will take place in Washington, D.C., November 16 and 17. If you are interested in learning more about this, please contact: Edoardo Borgomeo at:edoardo.borgomeo@gmail.com.