This paper modifies and interprets Data Envelopment Analysis (DEA) using a linear economic model. This approach is similar to the cone input/output and assurance region approaches to DEA, but it is implemented so that the multipliers are measured in the same units across all linear optimization problems. This approach allows one to interpret alternatives as profit maximizing organizations and the DEA multipliers as prices that are comparable across the alternatives. This is a useful extension of the assurance region concept, but more important, is that our approach enhances communication with decision-makers. The improved communication is illustrated by applying the model to the siting of a long-term health care facility. This application is interesting because the multiplier bounds make practical sense, and because the problem has dimensions that sometimes lead to interpretation problems with the traditional DEA model. For example, the site characteristics do not result from coordinated decisions, some sites exhibit zero values for some variables, and the problem has many variables compared with the number of potential sites. Problems with these dimensions have, at times, been deemed unsuitable for DEA, but they are handled without problem by the linear economic model.
No Reference information available - sign in for access.
No Citation information available - sign in for access.
No Supplementary Data.
Document Type: Research Article
University of Mississippi, Hearin Center for Enterprise Science, University, Mississippi USA
Department of Law, City of Chicago, Chicago, Illinois USA
George Mason University, The School of Public Policy, Fairfax, Virginia USA
Publication date: 2003-09-10
More about this publication?