Industrial applications of cross-impact analysis

The implicit or explicit ceteris paribus assumption usually made in the planning of economic systems (e.g. industries) is acceptable and accepted during periods of regular economic development; in this context, sensitivity analysis around the central projection is often used in order to measure effective elasticities to unexpected small perturbations of the system.In periods of structural change, unpredictable major events greatly reduce the value of any trend projection, and the ceteris paribus assumption has to be replaced by ‘scenarios’ of the future environment.The causal structure which generates these events is unknown; in such a situation, it seems reasonable to rely upon ‘opinions’ in order to assess the a priori probabilities of these events. Cross-impact analysis is one of the methods developed in order to improve human probability judgments of single and pairs of events and in order to deal explicitly with higher order interactions; in particular, cross-impact allows for the computation of the most probable combination of the events considered, or ‘most probable scenario’. In the industrial planning process, this ‘most probable scenario’ can be used as the set of exogeneous conditions for the central projection.