On the proper modelling of multioutput port cargo handling costs

Cargo handling activities involve various heterogeneous outputs, e.g. general cargo, containers, dry and liquid bulk and so on. These activities in ports, however, have been usually analysed using aggregate descriptions of output such as total tons moved. The main purpose of this article is to show that ignoring this heterogeneity may lead to two types of problems: (i) the underestimation of the relevance of key dimensions (i.e. marginal costs per product and economies of scope) and (ii) a bias in the estimates of the relevance of other dimensions (economies of scale). To do so, we rely on a unique new dataset on three cargo handling firms operating in a Spanish port between 1991 and 1999. We use it to estimate both a multi-output cost for these three operators, as well as an aggregate cost function. The policy conclusions are derived from an explicit and detailed comparison of these two sets of estimates.