Measuring cost efficiency in the presence of quasi-fixed inputs using dynamic Data Envelopment Analysis: The case of port infrastructure

Port infrastructure has characteristics that are classified by quasi-fixed inputs, meaning they cannot be immediately adjusted, contributing to the production of transport services over long periods of time. Another important characteristic of port infrastructure is that its current technology is a consequence of investment decisions made in the past and whose effects resonate over various periods. A dynamic approach that acknowledges the inter-temporal relationship between the inputs used and the resulting outputs is more adequate to handle this type of inputs. In this article, the efficiency achieved for Spanish Port Authorities from 2000 to 2007 has been obtained using a dynamic inter-temporal Data Envelopment Analysis cost model, obtaining dynamic and static components of inefficiency. The result has been compared with that obtained from a static approach, showing that the static model overstates all components of cost inefficiency. Moreover, since by its very nature the static model is unable to yield the dynamic inefficiency, it has a tendency to transform efficiency into technical and allocative ones, which can lead to faulty investment decisions. If the specific characteristics of port infrastructure are not considered, the consequences could lead to investment decisions being made that seek to expand the productive capacity of the port when such an expansion is not warranted.