This thesis deals with the explanation of why European Distribution Centers (EDCs) cluster around Amsterdam Airport Schiphol. The Netherlands is an attractive location for EDCs. Approximately onehalf of all EDCs in Europe are located in The Netherlands. In 2002, the EDC sector contributed about 95,000 jobs (full time equivalents) and 1% of the gross national product (GNP) to the Dutch economy. Schiphol represents an important concentration of EDCs. The amount and nature of distribution centers located in the surrounding areas of Schiphol is a prime concern for Dutch policymakers because of two reasons. First, attracting EDCs means attracting all kinds of international goods flows, investments and employment. Second, EDCs are space demanding while the amount of open space available outside the airport is limited and earmarked for airportdependent (or airport-related) firms. However, it is unknown whether the right locations are allocated to the right distribution centers and whether a bad location policy has long-lasting bad effects. The explanation of why economic activities cluster around airports is important for location theory and location policy regarding airport regions. The traditional answer to the question why distribution centers cluster around Schiphol is that they are attracted to the airport due to the importance of having air transport services at their disposal. However, we show that this is only a partial answer. In this thesis we present a new model of the clustering of EDCs around airport Schiphol. In the model, the following location forces interplay: location endowments; agglomeration economies; locked-in logistics. We assessed the empirical relevance and explanatory power of the model. Our model shows that the clustering of EDCs near Schiphol needs to be accompanied by new insights concerning location policy. We uncovered that there are major implications of threshold effects in the surrounding areas of the airport. Threshold effects are often neglected in standard neoclassical location models. Our model explains that marginal policy interventions may have no impact on the growth rate of the EDC cluster as long as the level of the policy instrument remains below a threshold value but that the same policy intervention can have a large effect when the threshold is crossed. The possibility of such non-linear responses makes it much more difficult to forecast the effect of a given policy change. Moreover, the model shows that the effects of the EDC location policy also exhibit phenomena of irreversibility. (Author/publisher)
Abstract