Abstract
Following the using a panel of 402 commercial banks from 15 developing countries over the period between 2000-2003 period, we assess the effect of bank size on technical efficiency and its two components: pure technical and scale efficiencies. We use in this study data envelopment approach (DEA) under specifications that allow the examination of the impact on results of the choice to measure banking activities with an intermediation or a value added approach, and of course a test for the relevance of including non traditional activities. Results indicate that examined banks suffer from serious problems of technical inefficiency involving a total average waste of resources that exceeds 46% of their actually levels. This inefficiency is mainly due to pure technical inefficiency for all size of banks except the largest banks for which we found high levels of scale inefficiency. The conducted test results show that the models with and without non-traditional activities are equivalent in terms of overall technical efficiency for banks of all size classes except for those of the smallest size. However, it is proved by these tests that the choice of an intermediation or a value added approach for measuring banking activity can significantly influence the generated average levels of technical efficiency for all bank sizes, but scale efficiency estimates appeared to be less sensitive to this choice.