Thredbo-1 The Relative Efficiency of Public and Private Bus Companies? James Odeck The Norwegian Pubic Roads Administration PO box 8142 Dep., N-0033 Oslo, Norway e-mail: james.odeck@vegvesen.no Øyvind Sunde Molde University College PO box 308, N-6401 Molde, Norway e-mail: oyvind.sunde@hsm.no By applying Data Envelopment Analysis (DEA), this study examines the efficiency of a subset of Norwegian bus companies in order to test the hypothesis that public companies are less efficient than private ones. DEA consists of constructing a piecewise linear best practice frontier enveloping the input-output combinations of the companies, efficiency being measured for each individual company in terms of its distance relative to the frontier. DEA is widely acclaimed for being flexible, letting the data reveal the unknown and possibly complex relationships between inputs and outputs. As a consequence, each company is most likely evaluated against similar companies being located in the ‘neighbourhood’ in the input-output space. If private and public companies are clustered in separate subspaces in the input-output space however, each company will most likely be evaluated against companies of its own rather than the opposite category of ownership. In order to avoid this pitfall, we identify and include in the study only those companies that are found to be either efficient or inefficient as compared to at least one company in the opposite category of ownership. The results from the study are compared to those obtained by the more conventional approach where all companies are included in the study.