Technological diversity and Jacobs' externality hypothesis revisited

Recent empirical evidence strongly supports Jacobs's (1969) externality hypothesis that urban diversity provides a more favorable environment for economic development than urban specialization. In order to correctly gauge Jacobs's hypothesis, economic development should be understood as a result of innovations. Furthermore, a relevant diversity measure should take into account the degree of diversity of the inherent classes (e.g. pharmaceuticals are closer to chemicals than to forestry). These ideas are tested using regionally classified Swedish patent application data as a measure of innovativeness. Patent data are also used to reflect technological diversity. The results show that the number of patent applications in Swedish regions is highly and positively dependent on regional technological specialization, quite the opposite of Jacobs's prediction. This paper raises general questions about earlier empirical results. It is concluded that the size of regions is an important factor to consider, since this in itself may affect patenting intensity and technological diversity.