Does cross-impact analysis have a future?

Abstract

where event 3 is the incidence of grave social troubles in Spain in 1974 and event 6 is the end of net emigration of Spanish worker in 1975. Thus we are forced to the conclusion that when the grave social troubles of 1974 have been taken into consideration the workers of Spain are more likely than they would otherwise have been to stay in Spain. Presumably for the riots. Again, various contentions are arguable and it is fair to say that the Battelle exercise must properly be assessed in terms of all the events considered. The fact remains that such procedures are being used to endow a quite spurious mathematical respectability to the consideration of events which are grossly aggregated for simplicity and are causally related to each other in varying degrees which, because of the absence of any coherent theory of the mechanics of social systems, are not even quantifiable in relative terms. Cross-impact analysis was born of conditional probability theory which in turn was derived from the study of physical systems containing random elements. When applied to social systems it is at best a game as it must be if the events considered are sufficiently precise to allow accurate assessment of probabilities, for any such set of events fully describing even the simplest of social developments must be exceedingly numerous; at worst cross-impact analysis is an ineffective tool tempting its users to the delusion that it allows the accurate assessment of the probability of future events and scenarios. The application of inappropriate scientific techniques to the analysis of social systems has happened before and will again; it is time that this particular bag of chicken bones was laid to rest.