ABSTRACT
This chapter serves principally as a methodological preface to the analysis of EISs' predictive accuracy and provides details of ideal environmental audit methods. The first and still one of the best environmental audits was the University of Wisconsin study of the Columbia generating station. Most of the studies reported at the 1985 Banff conference on environmental audits were case studies of major Canadian projects' impacts. Two other studies audited four cases apiece: the Aberdeen University evaluation of four British environmental assessments and Hunsaker and Lee's analysis of four EISs' "worst case" forecasts. Of these evaluations, the Columbia station case study best illustrates the trade-offs involved in environmental audits. Several aspects of environmental assessments preclude the use of a pooled cross-sectional time-series measurement model. A few gaps in agency record keeping are difficult to reconcile with any notion of a properly functioning public bureaucracy.
