Questions and answers about
the economy.

How does uncertainty about the future affect climate change policy?

Uncertainties about the future mean that the design of climate policy is a challenging task. Being able to make adequate predictions of different potential scenarios is vital for building effective tools for both climate mitigation and climate adaptation.

Uncertainty is all around us. Tsunamis; the systems-wide repercussions of a global pandemic; ecosystem collapse following the loss of a keystone species; or even the surprise win of a presidential candidate. These are all examples of unexpected events – and we are usually poorly prepared for how such events will ripple through society, the environment, the economy and our everyday lives.

This is also true for climate change. It is certain that climate change is happening and is driven by human factors (Intergovernmental Panel on Climate Change, IPCC, 2014). But its inherently complex nature makes it less clear what the impacts will be – including when and where all of them will happen, and to what degree.

The uncertainty of future climate policies, greenhouse gas emissions, complex climate and socio-economic feedback loops, and unknown tipping points all further complicate our projections. For example, it is unclear to what extent warming oceans will affect global fish supplies, and how these changes may affect the broader food system and national economies. Similarly, the impact that heat waves may have on human health and labour productivity is uncertain.

We cannot predict the future with precision, but that does not mean we cannot or should not prepare for it. Acknowledging that uncertainty is inherently present in the world and affects our decision-making is a crucial first step. Understanding that different types of uncertainty exist, and how to approach them, comes next.

These include scientific uncertainty, which is the lack of exact knowledge, regardless of why this knowledge deficiency exists. Deficit uncertainty, for example, arises from a lack of accurate models, ignorance, biases and measurement errors. It can be minimised through technology and learning.

Complex uncertainty involves different interdependent factors and patterns of which we are both aware and unaware, and we are likely not to understand fully how they affect one another (Tye and Altamirano 2017). It is therefore much harder to account for.

When both deficit and complex uncertainty are at work – as they are when dealing with the climate crisis – another layer of unknowns is added, resulting in deep or cascading uncertainty.

Figure 1: Areas of climate uncertainty

Source: World Resources Institute

Climate adaptation and mitigation

Approaches to tackling the climate crisis are often split into two camps, with blurred lines between the two. The first – mitigation – focuses on reducing greenhouse gas emissions that are the root cause of the problem. The second is the process of helping societies prepare for and adapt to current and future climate change impacts – adaptation.

A third camp – climate resilience – is intended to be a bridge, acknowledging the many synergies and co-benefits of addressing the two together, but a full merger has not yet been achieved in practice.

Uncertainties abound with both mitigation and adaptation. But these are more salient in decision-making for climate adaptation because it is much less quantitative in nature. For mitigation policies, scientists can calculate the quantity of emissions in the atmosphere, know what is being emitted by different countries and industries, and come up with informed projections of future emissions paths. They can also develop recommendations on how to reduce emissions more effectively for each sector and estimated costs on how to reach targets.

On the other hand, it is much harder to understand and attribute the miniscule and systemic ways in which the climate crisis affects people, ecosystems and sectors like agriculture. For example, how much of an incoming drought is worsened by the climate crisis, and how much can be attributed to other factors? Or what will the pattern of infectious, vector-borne diseases look like in 20 or 50 years, and what will be the full health and social consequences of these phenomena? These are questions with fuzzy and incomplete variables and answers.

Yet despite these challenges, approaches have been developed to navigate uncertainty in climate mitigation and adaptation.

How can policy-makers navigate uncertainty to mitigate carbon emissions?

Often simulations (or models) help researchers and policy-makers to analyse quantitatively the climate policy and actions needed to understand uncertainty. A key aspect of these models is that they explore potential scenarios.

They can identify a range of options for reaching long-term climate mitigation targets and indicate what kinds of investments, actions and policies can enable the transition. Models do not predict precise future outcomes. Instead, they analyse possible futures states of the world (or scenarios) from which policy implications can be drawn.

Building a scenario analysis to explore how the climate crisis affects people and places over time is a challenging task. One of the biggest difficulties lies in selecting and building plausible future scenarios.

There is still great uncertainty over the future course of climate policies, climate impacts and other factors that materially influence businesses and investments. Uncertainty permeates scenarios and models. Often, researchers and planners analyse past and present emissions and try to forecast future ones. From this, they identify the optimal plan or strategy for the perceived circumstances.

This can be risky where uncertainty is significant, as with climate change. If the analyses and forecasts turn out to be seriously wrong, the outcome could be devastating and entail irreversible losses. For example, many insurance companies became insolvent after Hurricane Andrew – which struck the Bahamas, Florida and Louisiana in 1992 – largely because the risks had been underestimated.

Uncertainty can be accounted for in modelling studies in many ways. The most frequent method is to look into uncertainty about input indicators (for example, GDP growth, impacts from climate, and discount rate – the interest rate used to determine the value of future cash flows) that can be addressed by including a sensitivity analysis of the key inputs for the scenarios modelled.

This analysis is done by including different assumptions for single key scenario inputs and seeing how the results change. For example, by varying the discount rate (reflecting how society value future costs and benefits), the assessment can produce different results and hence different policy implications.

More advanced techniques, such as ‘decision-making under deep uncertainty’, take a different track. Instead of seeking the optimal option under a certain future scenario or just reporting how results change when varying one parameter at a time, this approach seeks to find robust options across diverse possible future scenarios. This involves testing hundreds, thousands or even more scenarios that are constructed by identifying material uncertain factors and assigning varied combinations of values to these.

This method provides flexibility in the decisions that policy-makers make: instead of choosing one scenario that provides an optimal solution for one future state of the world, they would choose the pathway that will perform well over many different scenarios (which are uncertain).

How can policy-makers navigate uncertainty to adapt to climate change?

Two of the main frameworks used by planners, policy-makers and adaptation practitioners to plan for and identify adaptation measures are resilient and adaptive approaches.

The resilient approach works well over a variety of possible outcomes, and involves techniques such as robust decision-making (RDM). This is an analytic methodology for planners that begins with a decision to act, and then looks at climate models, socio-economic data and other relevant information to identify the best strategy over a variety of future scenarios.

For example, RDM probability calculations could determine that an extensive drought in a specific region may last four to eight months, with the latter being more likely. Planners might then undertake drought preparations for six months or more considering these forecasts and be ready for multiple scenarios as a result.

An adaptive approach is more flexible. It usually responds to triggers and can be modified in real time as events unfold. One technique under this approach includes iterative risk management (IRM), a participatory technique that allows for flexible and reversible decision-making even when risks and thresholds are unclear.

For example, planners took an IRM approach to the Thames Estuary 2100 Project – a response to the region’s 1953 flood catastrophe, which caused a heavy loss of life and property. More than a million people live within the Thames floodplain and the area comprises about £200 billion in property in and around London. So planners developed a system that allows river barriers and defences to be raised or lowered at any time, providing flexible protection against sea level rises and tidal flooding.

How do different uncertainties affect policy design?

Future reductions in greenhouse gas emission depend on what technologies and innovations become available, how quickly they are deployed and scaled, and whether they are widely accepted by society. For example, the UK’s long-term strategy sees the future roles of electrification and hydrogen fuel in the building and transport sectors as sources of uncertainty. Similarly, the United States identifies the growth of clean vehicles (for example, electric vehicles and fuel cell vehicles) as one important uncertainty.

With climate adaptation, integrating uncertainty into planning and policy design is crucial because it improves the robustness of adaptation strategies and will affect how resources are allocated and distributed over the short, medium and long term.

Governments can systematically expand the multiple scenario approach explained above by ‘stress-testing’ their climate strategies against the different scenarios. Stress-testing is a risk management tool that involves analysing the impacts of extreme scenarios that are unlikely but feasible. This approach is similar to the stress tests that banks widely introduced after the global financial crisis of 2007-09 to examine their vulnerabilities to external shocks such as a stock market crash or a severe economic downturn.

The idea is the same. Repeating the stress tests for different strategies will help countries to identify the most important uncertainties to address and robust options to include in their long-term strategies that would perform well across many different scenarios.

For example, in one analysis, we stress-tested three policy packages for a hypothetical country’s long-term strategy against 1,000 scenarios with different assumptions of future cost reductions in low-carbon technologies.

In this demonstration, the stress test may suggest that the cost reduction of electric vehicles was the most critical uncertainty across scenarios. This is because, although costs of electric vehicles have been decreasing in recent years, it is uncertain whether they will compete (in terms of price) with petrol or diesel cars.

The analysis also indicated that policies to boost electric vehicle sales would reduce uncertainty in the hypothetical country's emissions outcome in 2050. For a real country, the most important uncertainties and the most robust policy options may differ, but the stress test process would be similarly useful for identifying robust policy options.

It is clear that uncertainty affects decision-making in climate mitigation and adaptation in numerous ways, and that scientists and policy-makers rarely have perfect information or knowledge. But this does not mean that decisions and actions should be put on hold.

Reducing emissions and protecting people and ecosystems from climate impacts remains an imperative, and increasingly sophisticated tools continue to be developed to take account of uncertainty and face it head-on.

Where can I find out more?

Who are experts on this question?

  • Juan Carlos Altamirano, economist at World Resources Institute
  • Ichiro Sato, Executive Senior Research Fellow, JICA
  • Robert Lempert, Director, Frederick S. Pardee Center for Longer Range Global Policy and the Future Human Condition; RAND Corporation
  • Stéphane Hallegatte, Senior Climate Change Adviser, World Bank
Authors: Stefanie Tye and Juan Carlos Altamirano
Photo by oobqoo from iStock
Recent Questions
View all articles
Do you have a question surrounding any of these topics? Or are you an economist and have an answer?
Ask a Question
OR
Submit Evidence