Mid Sweden University

The aim of this paper is to estimate the relative efficiency of Iranian forest management units withboth deterministic and stochastic Data Envelopment Analysis (DEA). Towards this end, the required data were collected from 14 Iranian forest management units included 2 inputs (growing stock, plantation costs) and 2 outputs (harvesting revenues, forest protection tasks). In deterministic scenario, the most frequently applied DEA model-BCC is used in which results showed that most of the forest management units are operating at low efficiency levels (just 35.71% efficient units). However, in forestry, some external uncertain factors such as socio-economic and climate factors influence the magnitude of harvesting revenues of forest management units. Thus, in uncertain scenario, an output-oriented Chance Constrained Data Envelopment Analysis (CCDEA) is used. Results were significantly different according to different risk criterion levels (α). Nevertheless, using the Kendall's tau correlation test showed by increasing the probability levels (1-α) of CCDEA their result are closed to deterministic DEA, it meansthat there were 5 efficient units (2, 3, 8, 9, and 10) in the all scenarios. Therefore, according to the output-oriented nature of DEA models, the managers of the inefficient units should increase their output while their inputs proportions remain unchanged; otherwise they will not be able to promote their overall productivity.