Abstract: | The authors advocate the use of cross-impact models for scenario generation and describe a calibration technique which reduces problems of scaling. In a study of the Netherlands construction sector to 1990, aggressive interviewing of experts produced quantified trends. After cross-impact analysis, an input-output table for 1990 was estimated from that for 1975. A number of scenarios were also developed, by adding events to the matrix. The authors review the strengths and weaknesses of the methods used and summarise the results of the study. |