Regional service clusters and networks. Two approaches to empirical identification and development: the case of logistics in the German port city-states Hamburg and Bremen

This article discusses two approaches to the identification and measurement of regional clusters and its networks in 'cross-sectoral' services which are not available through official industrial statistics. The first approach is a 'secondary-statistical' one consisting of a firm-based blending of two separate official statistical data-sets, industrial and 'functional' (that is, the professions practised within firms). Thus, a service 'cross-sector' is identified across manufacturing and service industries. In the matrices resulting, weights are attached in an expert survey to the numbers of employees to aggregate the 'real' logistics 'cross-sector'. This is applied to the two German port city-states, Hamburg and Bremen. The second approach is 'primary-statistical', based on a small firms survey which generated data on 'functional' supplier relations (the cluster) and on project-based 'strategic' cooperations (the networks within that cluster). This follows a two-stage model of emerging clusters and 'its' networks. This data-set is combined with the firms' affiliations to branches, firm size, age and sales growth classes, in order to connect information with the industry statistics. Also, the net densities and centrality structures are calculated. The combined information provides indications of the relevance of the service cluster and its networks as factors of future regional development. The latter approach is applied to the State of Bremen only. Two results appear to be transferable beyond the German cases: first, the two approaches improve the knowledge about policy-relevant 'cross-sectors', clusters and networks; and second our knowledge about service, namely logistics, clusters and networks (for which port regions are prominent nodes) is improved. Finally, some implications for regional cluster strategies are discussed.