We live in a time when the ability for constant learning is considered crucial to the welfare of individuals, organisations and nations. This is the age of the ‘learning society’. However, in environmental impact assessment (EIA), which is the main methodology used by decision makers to predict environmental effects of megaprojects, surprisingly little learning is taking place. Or, to put the matter more positively, learning is only now beginning. This is true for megaprojects as well as for other types of project. The reason for the lack of learning is that projects and their environmental impacts are rarely audited ex post, and without post-auditing learning is impossible.

A recent study describes this situation as an unfortunate ‘stalemate’ and concludes: ‘there is much scope for raising the profile of post-auditing in EIA world-wide’. One consequence of the current state of affairs is a lack of knowledge about the actual environmental risks involved in infrastructure and other development. Although many studies on environmental impact analysis have been elaborated, and many environmental impact statements made, there is still a general sense among professionals and decision makers that the state of the art is not satisfactory. In what follows we will elaborate on three reasons for the presumed deficiencies of environmental impact assessment:

1. (i) a lack of accuracy in impact predictions;

2. (ii) the narrow scope of impacts and their time horizon; and

3. (iii) an inadequate organisation, scheduling and institutional integration of the environmental impact assessment process in the overall decision-making process.