Göran Lindqvist

The connection between intra-industry proximity effects and industry localisation has been considered unproblematic. Industry localisation has been treated as an indicator of Marshallian proximity effects. However, this paper argues that localisation is an effect also of urbanisation effects. To identify intra- industry agglomeration, localisation therefore must be separated into two components of concentration and urbanisation. The paper proposes the Q function, a measure based on Ripley’s K function, as a way of achieving this. The Q function has an intuitive interpretation directly relevant for proximity effects. When applied to 30 industries in Sweden, the Q function reveals patterns of concentration and urbanisation, which with traditional localisation measures will be confounded and indistinguishable. The analysis also suggests that using the whole manufacturing sector as the reference distribution will give biased results when measuring localisation.

It is a general fact of economic life that regions differ markedly in their innovative capacity and economic prosperity. This is certainly true for nations across Europe, but is similarly true across regions within each nation. Successful regions have economic profiles different from those of less successful ones. This paper compares differences in both regional specialization giving rise to Marshallian (or MAR) externalities and urban diversity giving rise to Jacobian externalities. The analysis is based on a structural equation model with data from 211 regions in Europe. It shows that both types of externalities play important roles, but in different ways. First, our results show that MAR externalities are important for economic prosperity, but only indirectly through innovation. Specialized regions in Europe perform much better in terms of innovation input and output, which in turn leads to improved GDP/capita. Second, urbanization plays a direct positive role for economic prosperity. It also plays a role in explaining public R&D, but is not directly linked with business R&D or innovation output. Third, public R&D only leads to innovation output through business R&D, and thus the notion that more investments in public R&D should lead directly to more innovation does not agree with the empirical results for Europe.

Studies comparing industry concentration in Europe and USA have faced three challenges: obtaining data on sub-national level in Europe and sub-state level in the USA, disaggregating data into relevant and comparable industry groups, and devising a method for a consolidated comparison between the two continents. This paper reviews five previous studies, and presents a new comparison. It is based on considerably more granular data than previously used, a new grouping of industries, and applies the polarisation measure as a consolidated indicator of total industry concentration. The results confirm the previous conclusion that concentration is higher in the USA than in Europe, and show that this result is robust for different measures.

This paper examines cluster policy from the perspective of cluster organisations. Cluster organisations are dedicated to enhancing the growth and competitiveness of selected clusters, and do so by engaging in a wide range of activities. Using survey data from a large number of cluster organisations in many different countries, these activities are shown to fall into seven distinct groups. The effect of each upon self-reported performance is assessed and found to vary between activity groups. In addition, the performance effect of cluster organisations initiated by government or undertaken in collaboration with other cluster organisations is tested, but no evidence is found to support the assumption that these factors have an effect in general.