Monitoring and auditing: after the decision - Introduction to Environmental Impact Assessment - page 202

7.4 Auditing in practice
Auditing has developed a considerable variety of types. Tomlinson & Atkinson (1987a,
b) have attempted to standardize definitions with a set of terms for seven different points
of audit in the “standard” EIA process, as follows:
• decision point audit (draft EIS)—by regulatory authority in the planning approval
process;
• decision point audit (final EIS)—also by regulatory authority in the planning approval
process;
• implementation audit—to cover start-up; it could include scrutiny by the government
and the public and focus on the proponent's compliance with mitigation and other
imposed conditions;
• performance audit—to cover full operation; it could also include government and public
scrutiny;
• predictive techniques audit—to compare actual with predicted impacts as a means of
comparing the value of different predictive techniques;
• project impact audits—also to compare actual with predicted impacts and to provide
feedback for improving project management and for future projects;
• procedures audit—external review (e.g. by the public) of the procedures used by the
government and industry during the EIA processes.
These terms can and do overlap. The focus here is on project, performance and
implementation audits. Whatever the focus, auditing faces a number of major problems as
outlined in Table 7.2.
Such problems may partly explain the dismal record of the early set of Canadian EISs
examined, from an ecological perspective, by Beanlands & Duinker (1983), for which
accurate predictions appeared to be the exception rather than the rule. There are several
examples, also from Canada, of situations where an EIA has failed to predict significant
impacts. Berkes (1988) indicated how an EIA on the James Bay mega-HEP (1971-85)
failed to pick up a sequence of interlinked impacts, which resulted in a significant
increase in the mercury contamination of fish and in the mercury poisoning of native
people. Dickman (1991) identified the failings of an EIA to pick up the impacts of
increased lead and zinc mine tailings on the fish population
Table 7.2 Problems associated with post-auditing
studies
Nature of impact predictions
Many EISs contain few testable predictions; instead, they simply identify issues of potential
concern.
Many EIS predictions are vague, imprecise and qualitative.
Testable predictions often relate to relatively minor impacts, with major impacts being referred to
only in qualitative terms.