Search

We evaluate consistency and patterns among a 20-year sample of EISs for road projects in North Carolina, USA. Applying multivariate statistical methods, we find substantial variation in reporting of project and site characteristics, the EIS process itself, and types and definitions of impacts. We find a relationship between environmental impacts and bio-physical or geographic factors. The relationship between predicted environmental impacts and economic factors suggests more environmental impacts occur, or are identified, in high-income counties. Variables describing public involvement and completeness are associated mainly with the number of residential relocations required. Farmlands potentially impacted are reported inconsistently and are not reliable for comparison, so rural environmental impacts may be seriously underestimated. [ABSTRACT FROM AUTHOR]