Scenarios do not predict the future: they prepare the decision for review

Forecast, option and scenario are three different objects

A forecast estimates a future value from a model, data and assumptions. An option describes an action the organization could take. A scenario describes a coherent external context in which that option might need to work. Confusing the three can turn a preferred plan into an apparently established future.

In the case study published in Sustainability, the authors define scenarios as possible futures rather than predictions. They use them to compare the implications of several transformation pathways with participants from thirteen organizations. The distinction appears on page 6 of the full scientific PDF.

A scenario does not remove uncertainty. It makes uncertainty open to examination by separating what the organization controls, what it assumes and what depends on external actors or conditions.

What the research actually allows us to conclude

The Blekinge study describes a one-day workshop involving 43 participants from 13 organizations. The format enabled the group to produce and assess regional transformation scenarios quickly. However, “rapid” refers to participant time: the authors report substantial preparation before the workshop.

The evaluation remains limited. Post-workshop feedback rests on a 37% response rate, or 16 people, plus one interview with a manager. The authors also state that this is a single case and that transferability to other contexts has not been demonstrated. They call for more quantitative modelling. These limits appear on pages 9 and 21-22 of the PDF.

A 2024 synthesis, limited to participatory scenario planning for climate adaptation and management, examines four systematic reviews published between 2015 and 2021. It finds that post-workshop monitoring and evaluation remain limited. The authors conclude that the literature assessed does not establish lasting effects, including effects on public policy. This finding narrows the claim instead of generalizing the method.

The authors nevertheless propose a mechanism compatible with decision review: repeating the process may force participants to reconsider their planning as uncertainty evolves. They present this as a possible mechanism, not a demonstrated effect. The synthesis and its caveats appear on pages 8 and 12 of the full PDF.

A comparison that genuinely prepares the trade-off

The proposed synthesis treats every scenario as a common test bench. The same options should be compared in the same futures, using the same criteria, rather than designing a favourable scenario for each option. This discipline does not replace judgment. It gives judgment an explicit basis.

1. Define the question and time horizon.

   State the decision before producing scenarios. The horizon should match the period during which the commitment is difficult to reverse.

2. Separate external uncertainties.

   Demand, costs, regulation, finance, timing and actor behaviour are external conditions, not internal options.

3. Apply constant criteria.

   Expected value, financial exposure, feasibility, dependencies and reversibility must retain the same meaning across scenarios.

4. Identify breaking points.

   An option becomes fragile when critical assumptions cross a defined threshold, not merely because a scenario is labelled pessimistic.

5. Prepare the review.

   Each critical assumption receives an observable signal, an owner and a stated consequence for the decision.

This reading is not presented as an experimental conclusion of the study. It is a structuring proposal consistent with the use of scenarios as possible futures and with the purpose of a traceable decision file.

From scenario to decision file

A comparison supports governance only if the conditions of the choice survive the meeting that produced it. The file should connect each conclusion to its supporting assumptions and to the signals that could invalidate it.

Within a decision file, context, criteria, scenarios, dependencies, risks and financial trajectory remain more useful when they stay distinct. Bringing them together prepares the trade-off without implying that an automated chain decides on behalf of people.

What this analysis does and does not establish

• It clearly distinguishes scenario, forecast and option.
• It supports retaining assumptions and review signals in the decision file.
• It does not show that more scenarios automatically improve a decision.
• It does not provide a reliable probability for every future described.
• It does not turn Hornsea 4 into evidence that an undeclared method was used.

The value of a scenario lies less in telling a future story than in making the choice reviewable. An option may remain preferable without winning in every scenario, provided that its vulnerabilities, thresholds and fallback paths are known.