The role of data in the EIS process: Evidence from the BLM wilderness review

Various propositions hove been offered about the role of the environmental impact statement (EIS) in agency decision making. These include statements that agencies are (1) using the information collected in the EIS to make rational decisions; (2) justifying decisions made a priori; (3) using the EIS to gain support or consensus for projects; or (4) simply fulfilling a legal mandate, with the EIS having no substantive impact on decisions. Previous studies regarding the role of EIS data have focused on the quality of the data in the EIS and whether or not the data are related to decisions. The role of site-specific information in the Bureau of Land Management (BLM) wilderness EIS process is analyzed and the results are used to reflect on the impact of the EIS in agency decision making. These results are compared with an earlier analysis of the Forest Service's Second Roadless Area Review and Evaluation (RARE 11). The results of the statistical analyses of three sets of BLM wilderness EISs indicate that although some of the site-specific information about resource potential is statistically related to agency wilderness recommendations, the vast majority of the information is not. In addition, in some cases, the information was related to wilderness recommendations in a counterintuitive direction. Overall, of the 190 measures of resource potential found in these documents, only 17 (9%) were statistically related to BLM recommendations in an intuitive direction. The fact that most of the information in these EISs is not statistically related to decisions lends support to the proposition that the agency was primarily fulfilling the legally mandated procedure of the National Environmental Policy Act in producing these EISs, rather than achieving the spirit of the law. Results from the analysis of the Forest Service's RARE H wilderness review are similar. Although these analyses may provide support for proposals to improve the EIS through shortening of the documents, more research is needed before it can be assumed that shorter EISs will ensure a link between the remaining information and agency decisions,