Controlling for exogenous environmental variables when using data envelopment analysis for regional environmental assessments.

Researchers are increasingly using data envelopment analysis (DEA) to examine the efficiency of environmental policies and resource allocations. An assumption of the basic DEA model is that decisionmakers operate within homogeneous environments. But, this assumption is not valid when environmental performance is influenced by variables beyond managerial control. Understanding the influence of these variables is important to distinguish between characterizing environmental conditions and identifying opportunities to improve environmental performance. While environmental assessments often focus on characterizing conditions, the point of using DEA is to identify opportunities to improve environmental performance and thereby prevent (or rectify) an inefficient allocation of resources. We examine the role of exogenous variables such as climate, hydrology, and topography in producing environmental impacts such as deposition, runoff, invasive species, and forest fragmentation within the United States Mid-Atlantic region. We apply a four-stage procedure to adjust environmental impacts in a DEA model that seeks to minimize environmental impacts while obtaining given levels of socioeconomic outcomes. The approach creates a performance index that bundles multiple indicators while adjusting for variables that are outside management control, offering numerous advantages for environmental assessment.