It has been only about 30 years since the development of data envelopment analysis (DEA), yet it has already obtained remarkable achievements in evaluating the performances of operations. It can not only measure the efficiency of each decision making unit (DMU) objectively, but also improve the inefficient DMUs by projection. With the time goes by, the CCR model, which is the first DEA model, is now not suffice for the more and more complicate problems. For example, the results of the two-stage DEA model is much better than that of the CCR model in measuring the efficiency of two-stage processes. The main task of this paper is to study a variety of properties of the two-stage additive DEA model and then test the effects of its projection. So in this paper we first construct the production possibility set and the production frontier of the two-stage additive DEA model, then define its projection. At the next part we use the case of non-life insurance companies in Taiwan to compare the effects of projection of the two-stage additive model and the CCR model. The results indicate that the projection of the two-stage additive model can provide a much better result than the CCR model. |