Designing scenarios to guide robust decisions

Community member post by Bonnie McBain

Bonnie McBain (biography)

What makes scenarios useful to decision makers in effectively planning for the future? Here I discuss three aspects of scenarios:

  • goals;
  • design; and,
  • use and defensibility.

Goals of scenarios

Since predicting the future is not possible, it’s important to know that scenarios are not predictions. Instead, scenarios stimulate thinking and conversations about possible futures.

Key goals and purposes of scenarios can be any of the following:

  • inform and educate;
  • allow us to determine what our goals are;
  • help us investigate our assumptions;
  • highlight important processes and decision points;
  • engage different stakeholders;
  • provide insight into what is possible;
  • provide visions of the future which motivate actions toward a desirable goal or away from an undesirable one;
  • show where differences between stakeholder priorities or worldviews lie in order to analyse potential areas of conflict between them;
  • communicate complex information to non-scientific audiences;
  • make infinite potential options for the future more manageable; and,
  • explore the adaptability of policy.

While scenarios will not resolve uncertainties, exploring uncertainty is a pivotal aspect of the exercise, as is helping decision-makers make better decisions in the face of uncertainty.

Designing scenarios

Scenarios can have qualitative or quantitative aspects or both. A common approach is to combine a descriptive story line with numerical modelling.

Story lines are particularly useful for aspects of the scenario for which something is not quantifiable or there is insufficient data to quantify it with the required accuracy. For instance, story lines allow social, cultural and institutional factors to be addressed explicitly, even if current knowledge does not allow these uncertain factors to be treated in a quantitative way.

Development of qualitative storylines typically involves stakeholder engagement to negotiate plausible futures that are coherent and internally consistent. This process could involve a formal or informal dialogue (workshops, interviews, surveys, etc) involving both experts and stakeholders.

Quantitative scientific modelling is often used to examine different assumptions or actions, such as the consequences of a single course of action under different key assumed uncertainties or the outcomes resulting from different policy actions.

Development of the qualitative and quantitative aspects of scenarios can happen concurrently and will then be iteratively refined in response to each other.

In general, the following iterative development process is used:

  • identifying the main driving forces affecting the state of a system;
  • determining the current state;
  • identifying the critical uncertainties;
  • making assumption about how uncertainties will evolve;
  • identifying options for mitigation; and,
  • analysing the implications.

Scenario exercises can be exploratory or normative. Exploratory scenarios contrast one or more baseline scenarios with one or more policy scenarios that consider the role of deliberate human actions and choices in shaping the future. Baseline scenarios reflect uncertainty (for example of driving forces or other parameters), so that there can be more than one baseline covering different aspects of uncertainty. Normative scenarios, on the other hand, develop stories about preferred futures. They provide a vision of a transition to a desired or alternate future.

Despite the almost infinite range of possible futures that they could consider, scenarios should be limited in number. For exploratory scenarios, an even number of baseline scenarios is better than an uneven number, in order to prevent the decision-maker from settling for a ‘middle ground’. Four baseline scenarios are better than two, in order to avoid the decision-maker interpreting two scenarios as ‘extremes’. The appropriate use of scenarios refrains from ‘picking’ any particular chain of events, but rather focuses on how a range of scenarios describes the most important uncertainties at stake.

Although limited in number, scenarios should also be diverse and represent a range of future visions, values and world-views. There are, in general, more policy than baseline scenarios, however, the number of policy scenarios must be kept manageable in order to avoid scenario fatigue.

It is also an advantage if scenarios can incorporate surprises, although this may not be fully possible as baseline scenarios are ideally of limited number. Surprises can be positive or negative and include events such as a world war, ‘miracle’ technologies, an extreme natural disaster, a pandemic or a breakdown of the climate system. Plausible, yet unexpected, occurrences incorporated into scenario building exercises can help decision makers recognise the need for adaptive management strategies that can flexibly deal with surprises.

Scenarios should also span long time horizons of at least several decades to allow adequate consideration of slow, incremental change, the full consequences of which are only felt in a distant future.

Scenarios must have the ability to communicate options and outcomes clearly to a range of different stakeholders affected by them. The story line aspects of scenarios means that they can be used to communicate complex information to non-scientific audiences in an exciting and clear way.

Producing useful and defensible scenarios

In order to be useful for decision makers, scenarios must be:

  • plausible, integrated, coherent and internally consistent;
  • analytically sound with regard to use of data and scientific theory;
  • able to incorporate the global scale as well as to be disaggregated to a regional and, ultimately, sub-regional scale;
  • able to consider environmental drivers along with the socio-ecological system. This makes scenario development more complicated and requires a broader base of expertise to ensure that scenarios have an internally consistent set of assumptions about key relationships and driving forces from a social, economic and environmental point of view.

For scenarios to be defensible, they must also be transparent and well documented. This ensures:

  • understanding of the reasoning behind the scenarios and the assumptions made;
  • informed criticism and further improvement by identifying any bias in scenario production and focusing subsequent argument on underlying uncertainties; and,
  • informing potential users of appropriate conditions under which scenarios might be used, not used or modified.

Conclusion

What has your experience been in using and producing scenarios? Is it in line with the suggestions above? Are there points you disagree with or do you have additional issues to add?

This blog post is a modified version of three previously published blog posts, which contain multiple links and references:

  1. Scenarios – creating alternate futures when we just don’t know: https://uonblogs.newcastle.edu.au/herdingthegreenchicken/2017/05/16/scenarios-creating-alternate-futures-just-dont-know/
  2. Scenarios – what form do they take? https://uonblogs.newcastle.edu.au/herdingthegreenchicken/2017/06/05/scenarios-form-take/
  3. Scenarios – guiding robust decisions if they’re well designed: https://uonblogs.newcastle.edu.au/herdingthegreenchicken/2017/06/05/scenarios-guiding-robust-decisions-theyre-well-designed/

Biography: Bonnie McBain PhD is a senior lecturer at the University of Newcastle, Australia. She has research and teaching interests that focus on sustainability. In particular, she is interested to discover how the theories of complexity and resilience can inform action towards a more sustainable future. She has extensive experience in community engagement both within the university sector and in state government. She researches ecological footprint policy development having developed a global ecological footprint model that is used to explore the effectiveness of future policy scenarios. She also has a research interest in learning for sustainability. She is extensively engaged in community change projects. Bonnie shares her expertise in a blog called Herding the Green Chicken, in The Conversation, via Twitter @TheGreenChook, via LinkedIn, and on YouTube.

15 thoughts on “Designing scenarios to guide robust decisions

  1. My initial impression of the subject matter of this post was probably coloured by my interpretation of the word scenario being based on the way I have used it. As the discussion has developed, I think I see the term being used in different ways but also, and more importantly, for different purposes.

    I regularly use scenarios to map out a range of possibilities on a fairly solid single dimension, the range of values that something such a the cost of a piece of equipment or the productivity of a team of developers might take, described in terms of a pessimistic, an optimistic and the most likely scenario. The case I outlined in my initial post was a way of clarifying options by using scenarios that captured a higher level statement of policy, in the mining sector these concepts might be high economic returns, environmental sustainability, social benefit and so on, within each of which many technical details and implementation approaches will be defined. These systematic approaches are useful in a predominantly linear constrained environment.

    Where one seeks to be more adventurous and explore what might be possible as opposed to understanding the differences between options we already know are possible, emergent methods are more useful. Here I am thinking of the work of David Snowden and Cognitive Edge. These are methods that make no assumptions about what is possible in the way they are implemented – there is no template scenario description and no presumption about what could happen before the exercise begins. They allow insights to arise from the interaction of people through narrative techniques or techniques that open up very high level metaphors and allow participants to fill them out. For example, the Future Backwards method starts from an impossibly wonderful and an unbelievably terrible future and recursively develops the prerequisites for each of the steps that could lead to that future state. As it works back, these prerequisites for the good or bad future begin to overlap with actual current conditions and the path by which that arose, providing insights into how a better future can be sought and what to avoid if the undesirable future is to be prevented.

    I guess that can all be summed up by saying that interesting concepts such as scenarios are an example of a complex concept. There are many ways that they can be interpreted and, as always, meaning depends on context.

    • Yes, Steve. Its great to know more about how you use scenario development in your job. The application of scenario planning is highly diverse and can be customised to fit the purpose. Another way that you can think about varied scenario applications is by the degree of uncertainty. Gabriele’s post in this blog ‘Six types of unknowns in interdisciplinary research’ (url: https://i2insights.org/2016/05/10/six-unknowns-in-interdisciplinarity/) is another way of defining different scenario approaches. Each of the six different uncertainties requires varying considerations such as – who to invite to collaborations for designing scenarios, the balance between the use of qualitative and quantitative data and the normativity of the scenario purpose.

  2. This is a great post! Excellent summary of the uses of scenarios and criteria for an effective scenario exercise. In my experience, another use of scenarios is to get people out of entrenched positions in the here-and-now (which can devolve into the same tired talking points on different sides of an issue, with no learning or progress) and engage their imaginations around what is possible, potentially leading to novel solutions. I also love scenario exercises that engage the arts, like what was done in Madison, Wisconsin (see https://wsc.limnology.wisc.edu/yahara2070).

    • Thanks, Laura. Yes, this is a very exciting role that scenarios can take. In this day and age, we need novel solutions to complex problems that have resisted solution and creativity that can come with scenario building can be a very powerful source of new ideas. I agree with you. If we are able, I think it’s critical that we engage not only the narrative elements of scenarios but also our other senses if we are to seek engagement beyond the stakeholder group who designed the scenarios. Visual and auditory elements just add to the richness of the futures that are described.

  3. What I find fascinating about scenarios is the power of creating stories as a way of integrating perspectives and combining qualitative and quantitative information, and the diversity of ways to approach creating and sharing these stories, depending on the purpose.

    This post to some extent emphasises a particular, dominant, view of how to design and use scenarios. I’d be interested in your perspective on a few questions:
    – How do you feel about quantitative methods that involve sorting through large ensembles of scenarios looking for specific insights (e.g. in exploratory modelling or “scenario discovery”), rather than focusing on building a small set of scenarios?
    – The development process you describe is a top-down/forward approach. Have you had any experience with bottom-up approaches, working back from outcomes rather than forward from drivers, e.g. when developing normative scenarios?

    I was involved in a review paper with a strong focus on scenarios, here:
    Maier HR, Guillaume JHA, van Delden H, Riddell GA, Haasnoot M, Kwakkel JH (2016) An Uncertain Future, Deep Uncertainty, Scenarios, Robustness and Adaptation: How Do They Fit Together?. Environmental Modelling & Software 81 (July): 154–64. doi:10.1016/j.envsoft.2016.03.014

    • Hi Joseph,

      Thanks for sharing your paper. I’ll read that with interest.

      I’ve not undertaken a quantitative analysis of available scenario sets but I did, a long time ago, undertake an informal qualitative assessment of scenario themes to find out if there were some ‘archetypal’ scenario themes. Some of the common themes were future population, affluence, technological trajectories (which seem to underly all environmental scenarios I was interested in at the time) but also level of globalisation; international cooperation of lack of cooperation (uni/multilateralism); income distribution; nationalism and other isms; success of governance; global power configurations; and surprises.

      I have used the bottom-up approach in higher education teaching (i.e. imagine a positive (normative) future and work backwards to see what needs to be put in place to work towards such a future. I’ve also used the same approach in catchment management workshops. Its a very interesting process to see what range of futures different people design and it helps us understand the assumptions and values behind decision making.

  4. My experience of using scenarios relates to developing options for large mining developments with social, environmental, commercial and other forces at work. The approach used by one of my colleagues was to: generate a large set of what, collectively, might be termed topics; condense these by clustering similar items and replacing them with a single summary item, which still left about 50 topics to address. A divergent process followed by simplification.

    The topics were classified into: Facts that might be palatable or disturbing but are not subject to appreciable uncertainty; Decisions that will alter the path forward but are at the discretion of the proponent so they too are not subject to uncertainty, and; Uncertainties that cannot be controlled, or at least not completely.

    Then the focus switched to a small number of strategic themes or policy directions that the development could take: clean and green, cheap and dirty, small and low key, large and imposing. The topics were linked to the Facts and Decisions and a handful of options that could actually take concrete form were worked up. A convergent process.

    These options then formed the basis of detailed study to determine which would be feasible and of those which would be the preferred option. Uncertainties were assessed for each option in a comparative analysis, exploring one type of uncertainty across all the options.

    • Hi Steven,

      It sounds like that scenario development process worked well to open up possibilities for the future and clarify the level of uncertainty associated with various scenario elements.

      If you do identify elements that are highly uncertain in your scenarios you can also use it retrospectively to check how robust your decision-making is. For instance, if you select a recommended or preferred policy options then you can go back to the original range of possibilities and check how successful your policy is under some of the more uncertain variables. If you policy approach is not likely to succeed under a number of surprising futures then you can 1) make adjustments to the policy and/or 2) ensure your policy has mechanisms which enable it to adapt to change, should it be required.

  5. Hi Steven,

    Great to hear from you. I would suggest that it is very difficult to get that sustained level of engagement for most of the decision-making around complex issues, not just scenario development. It’s not until participants understand the need for longer-term collaboration that we can get to the underlying drivers of issues you are trying to manage. We need those higher level managers involved in these sorts of decision-making forums because they are the ones that can influence the direction of their own organisations.

    Just playing around with ideas here. What about the following:
    – use examples of other scenarios (e.g. IPCC) to demonstrate how longer-term collaborations can provide more powerful scenarios
    – if you do have a longer-term collaborative decision-making group, refer to the scenarios often especially when they do not represent the direction that discussions are going, therefore justifying a revision
    – develop scenarios face-to-face initially but then use digital collaboration to continue refining them (thereby not requiring further meetings that take high-level managers away from their roles)

    If anyone else has suggestions, feel free to contribute them to the discussion.

    • We do generally have the benefit of clients who have been tasked to prepare concrete investment proposals with a delivery date and clear expectations of the quality of the proposals. There is no shortage of commitment, sometimes with bonuses on the line simply to complete the concept study by a certain date.

      • Thanks, Steve. Yes, I suppose this is the benefit of working within the private sector that can use bonuses as a lever for engagement.

        Steven Cork, it looks like you’ll have to pay your participants a bonus to get longer-term collaboration in scenario building 😉

    • Thanks Bonnie. I have used each of the approaches you suggest with different projects and all help us make progress. In a project with the Academy of Science in 2013 (Australia2050) we jumped in with some archetype scenarios (generalised narrative outlines typical of what usually emerges from scenario planning – catastrophe, growth, restraint and transformation) and asked the participants to consider how Australia’s future might evolve along those different paths. This generated very good conversations – in fact the main aim of this project was to get people listening to, and understanding, one another’s thinking about possible futures.

      • I think you’ve highlighted a very important part of scenario building, Steven, which doesn’t focus so much on the end product but rather on the process. The social learning that can take place when people collaboratively build scenarios together can be an end in itself. The use of narrative when you build scenarios is very powerful for bringing together people from different backgrounds because it doesn’t necessarily require technical expertise to take part.

  6. A great article Bonnie. I support everything you have said and hope that the article will help readers see the many opportunities for using scenarios to support decision making. One key issue that I have faced when working with clients as diverse as public administrators, private businesses and community groups is that everyone wants to do their scenarios in a very short time. The rule of thumb seems to be that “important people” can’t spare more than half a day. A full day is seen as extravagant. Clearly this is not long enough to get the benefits of scenarios. Do you have any thoughts on how to deal with this trend?

Leave a Reply to Bonnie McBain Cancel reply