Creative Class vs. Individual Creativity ? A Multi-level Approach to the Geography of Creativity

For more than ten years the topics creative class, creative industries and creative regions are on the agenda in economic geography literature. Contributions mainly focus on the distribution, mobility and economic impact of creative individuals or companies from a regional perspective. It is believed that the spatial concentration of creative individuals lead to positive regional economic effects and those creative individuals agglomerated in urban and amenity-rich regions characterised by a climate of tolerance. Above all, it is the work of Florida, who gives attention to creative individuals. His quantitative occupational approach can be seen as the prototype and dominant approach, which was adopted in a wide range of other studies. More recently, there is a growing body of literature empirically testing Florida's hypotheses for regions outside the U.S. However, previous empirical studies relied on occupation or industry based definitions as a proxy to identify creative individuals and aggregated regional numbers (e.g. share of creative class). Thus, results are potentially distorted. Instead of observing creative individual's behaviour there are occupation or industry specific characteristics in a region that might correlate with a concentration of these individuals. Moreover, using aggregated data bears the risk of ecological fallacy. Hitherto, economic geographers seem to have ignored insides from other disciplines studying creativity. In psychology it is not a dichotomy of creative and non-creative individuals, instead it is acknowledged that creativity is a matter of degree. The five-factor model ? or big five model ? is a well-recognized concept from psychology to describe an individual's personality based on five basic dimensions: extraversion, agreeableness, conscientiousness, neuroticism, and openness to experience. The dimension openness captures creative, innovative and artistic performance and interest of an individual. Hence, it should be a more direct measure to identify creative individuals than the approximation by occupations. Thus, the article builds upon insights from psychology and the data of the German Socio-Economic Panel to directly identify creative individuals based on their personality traits. Applying multilevel regression analysis, hypotheses derived from the creative class literature are tested comparing creative individuals with the rest of the workforce. The analysis controls for the individual and industry level to isolate the influence of regional characteristics. The empirical results show some evidence for Florida's hypotheses. Individual characteristics turn out to be the most significant differences, followed by characteristics at the industry level. Regional factors are less important, but urbanity and the share of bohemians are significant predictors. However, the location of creative class can be explained more precisely by their level of human capital and the location of the industries they are working in.