ABSTRACT
Introduction Richard Florida’s Rise of the Creative Class (2002) makes a compelling argument that recent urban economic development has depended largely on novel combinations of knowledge and ideas, that certain occupations specialize in this task, that people in these occupations are drawn to areas providing a high quality of life, and thus the essential urban development strategy is to attract and retain these workers. Florida bases much of his argument on interviews and focus groups, which he backs up with geographic data on where employment in occupations identified as part of the creative class is most prevalent. The high rankings of cities such as San Francisco, California, Austin, Texas, and Seattle, Washington, demonstrate the statistical similarities of these places that seemingly capture an essential dimension in differentiating creative centre exemplars from creative backwaters. This chapter addresses three major limitations in Florida’s analysis. First, while Florida focuses attention on metropolitan areas, particularly the largest areas, this chapter asks whether the creative class explanation applies to regional development more generally. In particular, it is asked whether rural economic development is also dependent on the novel combination of ideas that puts a premium on the attraction and retention of creative individuals. In addition to testing the potency of the creative class construct for understanding development in rural (non-metropolitan) US counties, this chapter also identifies outdoor amenities that tend to attract such workers to rural counties. To learn by contrast, the same model developed for rural counties is applied to urban (metropolitan) counties. These results also show a link between the presence of creative class and subsequent growth and they suggest a process in line with Garreau’s (1992) ‘edge city’ model of urban development. Because the present unit of urban analysis is the county rather than the metropolitan area, the analysis does not pertain directly to Florida’s main concern: metropolitan area development. The second limitation addressed in this chapter relates to the measurement of creative class. The research strategy puts a premium on having a valid construct for testing the creative class hypothesis, as a weak construct may fail to identify a true contribution to rural growth. Florida’s creative class measure has two
problems that are especially evident in a rural context. First, he relies on the 22 summary occupations in the 1999 Occupational Employment Survey to define employment that generally requires a high level of creativity. Using the Occupational Information Network (O*NET; US Department of Labor), which estimates functional requirements for more than 1000 detailed occupations, and the more detailed occupational categories available from the 1990 and 2000 US Censuses of Population, several detailed occupations that apparently require relatively little creativity are screened from Florida’s summary categories. Including these occupations would tend to blur the creative class measure particularly in rural areas, where empirical evidence of a spatial division of labour suggests that these lower skilled occupations are more concentrated (Massey, 1984; Wojan, 2000). In addition, although a premise of Florida’s work is that the creative class is relatively footloose, some occupations included in the definition, most notably ‘education, training, and library occupations’ and ‘healthcare practitioners and technical occupations’, are involved in economic reproduction and locate largely to provide essential services to a population. In rural areas, the perverse result is that high employment shares in these occupations can indicate a dearth of economic development. In recasting the creative class, these two broad occupational categories are dropped. Comparing results using the recast creative class measure with Florida’s original measure provides strong assurance that the present measure is a more valid construct. The third limitation addressed relates to statistical analysis. While Florida presents corroborative evidence on the relationship between creative class and growth, he never fully tests the creative class thesis in a multivariate framework. Here, growth (including the growth of the creative class) is modelled using a system of simultaneous equations, which addresses the endogeneity of population and employment growth while also controlling for influences from a number of local attributes. This approach allows a critical examination of the most cutting critique of Florida’s analysis: that he is merely substituting employment in highly skilled occupations as a proxy for the endowment of human capital (Glaeser, 2005). The present analysis confirms a strong independent influence on employment growth from both the initial share employed in the recast creative class occupations and its growth over the decade. By contrast, the statistical association with human capital variables is quite weak. The analysis also identifies possibilities for advancing rural development strategy. First, not all rural areas are likely to benefit from a strategy to attract creative workers. Rural areas most attractive to creative workers tend to have a sufficient density to provide a reasonable level of services, appealing landscapes and other natural amenities, and growth in surrounding areas. Yet, adjacency to a metropolitan area does not appear to be a prerequisite. Second, the analysis provides intriguing evidence that attracting creative workers may be influenced by local development strategies. The one missing prerequisite for making these findings actionable is the ground-truthing of this phenomenon parallel to Florida’s qualitative analysis of the urban creative class. Having supplied the
first step in identifying the rural potential of the creative class, the paper concludes with a discussion of research needed for rural strategies to realize this potential.
