Agglomeration Economics Beyond the Specialisation-Diversity Controversy

In this paper it is argued that the regional specialization-diversity debate, although very rich in the number of empirical studies testing for the dominant type of agglomeration economies, appears to be too diverse in methodology and operationalization to make sound conclusions. There are two arguments why the focus in this research field should shift. Conceptually, studies should focus more on the transfer mechanisms of knowledge and knowledge spillovers. Such a focus is necessary to capture the heterogeneity in space, in firms and in sectors that actually cause urban growth to differentiate. By accepting that spatial and firm level heterogeneity are much larger than the present models capture, the phenomena of related variety, path-dependency and heterogeneous sectoral development trajectories emphasized by evolutionary economic arguments do not necessarily contradict the analytical outcomes of the present growth models – instead, they will probably open some of the black box of agglomeration economies. Related to this conceptual shift, new methodologies that capture this heterogeneity are needed. Multi-level modeling, continuous space modeling and survival and selection models are good examples that complement the present analysis of agglomeration economies. The focus on the firm level, networks and heterogeneity of spaces and sectors also has implications for policy. If the recognition sustains that it is in transfer mechanisms, networks, skill- and transaction relatedness and individual firms’ capacities that growth opportunities occurs, policy needs to be place and target group specific.