The Knowledge Economy and Urban Economic Growth

ABSTRACT

In this paper we contribute to the longstanding discussion on the role of knowledge to economic growth in a spatial context. We observe that in adopting the European policy strategy towards a competitive knowledge economy, the Netherlands is—as most European countries—mainly oriented towards industrial, technological factors. The policy focus is on R&D specialized regions in their spatial economic strategies. We place the knowledge economy in a broader perspective. Based on the knowledge economy literature, we value complementary indicators: the successful introduction of new products and services to the market (“innovation”) and indicators of skills of employees (“knowledge workers”). Using econometric analysis, we relate the three factors “R&D”, “innovation” and “knowledge workers” to regional economic growth. We conclude that the factors “innovation” and “knowledge workers” are more profoundly related to urban employment and productivity growth than the R&D-factor. Preferably, urban research and policy-makers should therefore take all three knowledge factors into account when determining economic potentials of cities.

Acknowledgements

The authors would like to thank Martijn Burger, Koen Frenken and two anonymous referees for helpful comments on an earlier version of this paper.

Notes

1. Factor analysis is a statistical technique to identify the underlying variables (named factors) in a dataset in which multiple characteristics are included, that simultaneously show mutual correlation. This technique is often used to remove the overlap between the different indicators and reduce the characteristics to independent factors: the similarity within a factor is high while low between the factors.

2. The “innovation” and “R&D” indicators are measured in 2002, but reflect the CIS questionnaire in which is asked about the renewal in products, services and process over the previous 2 years. Data on R&D and innovation before 2002 are incomparable for analysis. Compared to the other variables used in the models, this implies a time lag in the variables of innovation and R&D, what might induce endogeneity. Jaffe (1989) shows in a knowledge production function setting in the US that spatial patterns of innovation indicators are to a large degree stable over time.

3. Combes (2000) shows that including the level of the local sectoral employment in the analysis strongly changes the interpretation of the specialization variable and leads to an overestimation of the localization economies. Actually, the impact of the share of the sectoral employment in total employment, holding the level of the sectoral employment constant, is simply the inverse of the effect of the total employment. Thus, it cannot be interpreted as intrasectoral local externalities. The correct interpretation is obtained if the level of the sectoral employment is replaced by the level of the total employment in control variables—what is done in our analysis.