The Efficiency of Public Forestry Organizations: A Comparison of Different Weight Restriction Approaches.

Data Envelopment Analysis (DEA) has become an increasingly popular approach to analyze the efficiency of public sector organizations. However, the underlying assumption of the method is that it is equally acceptable to specialize in producing any output or consuming any input. In many cases this kind of free specialization is not acceptable or desirable. In this paper, we use three extensions developed to the DEA method to incorporate additional judgment into the DEA models and study the sensitivity of the efficiency scores. the results indicate that even moderate restrictions on output weights in the optimization problem may lead to substantial changes in efficiency. It also seems that by analyzing only the technical efficiency, e.g., using only standard DEA without placing some kind of constraints on input and output weights, may lead to highly unreliable conclusions.