I’ve been doing a lot of thinking about how techniques from decision making under deep uncertainty (DMDU) can help during COVID-19 recovery planning. DMDU approaches and tools aim to improve the robustness of decisions when it is difficult to predict what the future might look like and to build in flexibility so that plans can adapt as external conditions change. This seems rather apt in the current times when it’s almost impossible to predict how our economy and society might change over the coming weeks and months.
Many organisations are producing scenarios of the many different ways that this change might develop. These scenarios are useful for thinking through what futures we might need to work within but don’t directly help us to make decisions to become robust to, or even influence travel towards, these varied futures. One technique from DMDU, Robust Decision Making (RDM), is specifically designed to improve the robustness of decisions across a range of scenarios and to build in flexibility to plans. It is a detailed and quantitative process but I’ve set out some qualitative steps below that could be taken that draw on the principles of RDM. Note that examples related to infrastructure are given throughout, because that’s the field I work in but this is definitely applicable beyond infrastructure. It’s also based on the premise that you have scenarios already. Firstly, some definitions that might be useful:
Plan – a group of activities that together will result in a desirable change in some part of the system of interest.
Uncertain factor – a variable for which the future value cannot be predicted but which strongly affects how activities in a plan might perform against its objective.
Vulnerability – a specific weakness in a plan, caused by a change in an uncertain factor or external conditions, which means it is no longer able to meet its objective.
Modification – a change made to a plan to amend activities occurring now or in the future to reduce its vulnerability to uncertain factors.
Trigger points – the point at which a plan is sufficiently vulnerable to a particular uncertain factor to require a change in activities.
Step 1 – Define the performance of the system you are hoping to achieve during the recovery from Covid – what are the objectives you want your plan to achieve? For example:
- Ensuring the survival of public transport;
- Locking in travel or energy consuming behaviours that have evolved during the crisis; or
- Achieving an existing vision under the new post-covid starting point.
It is also worth considering how these recovery-related objectives interact with or link to other organisational objectives, for example those related to health? This can help to ensure that plans implemented during covid-recovery are as aligned with these longer-term objectives as possible and do not encourage investment in projects that would lock you in to activities or assets that are at odds with this long-term vision.
Step 2 – Identify the uncertain factors that might affect your plan. This might be the main variables that define the COVID recovery narratives (e.g. economic recovery, public attitudes). However, there may be others specific to your plan e.g. climate change, land use change that could affect the success of a plan in the medium term or prevent it from achieving longer-term movement towards your objective. Therefore, it may be worth considering a wider range of uncertain factors.
Step 3 – Test the performance of options relevant to your plan in each of the your covid recovery scenarios. This will need to be done qualitatively using expert judgement at this stage but quantitative analysis using models could be used when issues are less urgent. Identify the vulnerabilities that arise in each of the scenarios – how does each of the uncertain factors in that scenario affect each options’ ability to achieve the performance of the system you wanted? Map vulnerabilities to the uncertain factors for each option you have.
Step 3 – Identity preferred option. If you have several options, identify which of your options is most likely to achieve the objectives you have across most scenarios (i.e has fewer and less significant vulnerabilities). At this stage you could assess the options against the broader range of objectives identified in step 1. The option which has fewest vulnerabilities and is more likely to achieve broader objectives, is your preferred option but you can still build in adaptability in case any of the uncertain factors changes.
Step 4 – Building in flexibility. For your preferred option take each of the vulnerabilities you identified in step 2 and explore how you could modify your plan to reduce that vulnerability. Map these modifications onto uncertain factors so that you can implement these should this factor change. Then identify trigger points, which tell you when an uncertain factor has changed sufficiently that a specific vulnerability might occur and your plan is less likely to meet the desirable performance. This might be a measurable change in public attitudes or a change in regional economic performance. When a trigger point is reached, you will need to change your plan and bring into play the modifications you identified for that uncertain factor.
This is a high level summary of stages but I do hope it’s helpful. Please do get in touch if you’d like to know more about how you might better manage uncertainty during recovery from COVID-19.