Multi-stakeholder elicitation methods for resilience-building decision-making

Presupposing that resilience is a latent characteristic, whether a system is resilient or not can only be determined post hoc; until then, its resilience will remain more or less probable. Furthermore, a system may be resilient to impacts of type X and yet be utterly fragile to impacts of type Y. Hence, the success of resilience-building efforts will partly depend on the expected type of future impacts.

With input from a wide variety of stakeholders, with knowledge of local and general specifics, it is more likely that the necessary attributes, values, and functions can be elicited and subsequently validated.

Consequently, the odds that resource allocation for building societal resilience is meaningful and aligned with the local context would likely increase.

Multiagent elicitation can result in a wide range of values with a large degree of uncertainty, even when surveying experts.

Therefore, elicitation methods should reasonably account for vague and imprecise input, such as rank orderings with uncertainty intervals.

Aggregating such data is inherently difficult and complex; the underlying sets of objects can be of different sizes as well as disjoint.

This study presents a framework for objectives and value elicitation from a disparate set of experts and stakeholders, including techniques for aggregating and validating the input.

The study also suggests possible ways of applying the results to resource allocation decisions for societal resilience-building efforts.