Diversity, Entrepreneurship, and the Urban Environment

ABSTRACT:

In recent years, the field of regional economic analysis has focused on the social and cultural environment of a place to explain variations in innovation activity, entrepreneurial dynamics, and economic growth. Prominent among these studies is Richard Florida’s creative class theory. He argues that urban economies grow because they are tolerant, diverse, and open to creativity, which in turn attracts certain groups of people, the so-called creative class. Lee, Florida, and Acs expand the theory into the realm of entrepreneurship. They argue that new firm formation is positively associated with a creative and diverse social environment. In other words, those regions that are alluring to creative talent, open to newcomers, and tolerant of those who are different, will also have more people taking the risk of founding a firm, leading to increased economic growth. The creative class theories, however, contrast with explanations that consider structural factors such as access to financial resources and markets, among others, as important markers of entrepreneurial success. In this article, we are interested in examining the ways in which a region’s creative milieu and its opportunity structures may help or hinder different kinds of entrepreneurs, in particular nontraditional entrepreneurs. We examine the effect of regional opportunity structures and creative milieu on women, Black, and Hispanic business ownership for the 50 largest metropolitan statistical areas (MSAs) in the United States in 2002. We find that opportunity structures, whether opportunity or barrier, better explain the dynamics for these entrepreneurs. They benefit from a regional environment that builds human capital and skill base, enables access to a variety of financial resources, and facilitates market access. Given the growth of women and minority businesses and their potential effect on regions, policymakers are well advised to tailor their policies to these groups.

Notes

1 The Current Population Survey is another data set that provides information on self-employed individuals. In addition, The Kauffman Foundation and Mathematica recently released firm level survey data that allow for a longitudinal view of the development of new businesses. In its complete form, the survey will include information on approximately 5,000 new businesses formed in 2004, with follow-ups in 2006, 2007, and 2008.

2 We have to note that some metropolitan areas such as San Antonio have suppressed data because they do not have a sufficient number of Black-owned firms.

3 We excluded Asian-owned businesses from the analysis because this group of entrepreneurs has been very successful outperforming their Hispanic and Black counterparts (Basu, 2006). They have also received more scholarly attention in the past than the other groups.

4 Our analysis does not focus on specific industries because the Survey of Business Owners provides the county or metropolitan level data only at the two-digit NAICS level. Thus, a meaningful analysis of these industry sectors that minorities specialize in is not feasible.

5 Census 2000 PHC-T-29. Ranking Tables for Population of Metropolitan Statistical Areas, Micropolitan Statistical Areas, Combined Statistical Areas, New England City and Town Areas, and Combined New England City and Town Areas: 1990 and 2000.

6 The 2002 Survey of Business Owners (SBO) was conducted by mail. One of two census forms was mailed to a random sample of businesses selected from a list of all firms operating during 2002 with receipts of $1,000 or more, except those classified in the following NAICS industries: crop and animal production (NAICS 111, 112); scheduled air transportation (NAICS 4811, part); rail transportation (NAICS 482); postal service (NAICS 491); funds, trusts, and other financial vehicles (NAICS 525), except real estate investment trusts (NAICS 525930); religious, grantmaking, civic, professional, and similar organizations (NAICS 813); private households (NAICS 814); and public administration (NAICS 92).

7 This is the final, updated version of the 2002 Guide to the Economic Census, Women-, Black-, and Hispanic-owned Businesses, http://www.census.gov/prod/ec02/sb0200cswmn.pdf. One other data caveat is important to address. Because the 2000 MSA definitions (and their corresponding counties) in most cases did not match the county by MSA listing in the 2002 SBO, we collected individual county data from the Census and aggregated these to obtain MSA level demographic data that corresponds with the 2002 SBO MSA definition.

8 Ladies Who Launch is designed to provide women with the tools and networking opportunities to help develop their businesses and entrepreneurial goals. It has incubator programs designed to help women expand and clarify their business visions and initiate participants to the concepts of incubating. WITI helps women advance by providing access to and support from other professional women working in all sectors of technology.

9 White’s correction adjusts the standard errors for the coefficients of the independent variables enabling proper interpretation of t-stats, but it does not change the coefficients themselves. We use the correction in all of the models examining the creative class theory as well as the model assessing women opportunity structures. However, there was no evidence of heteroskedasticity in the model of opportunity structures for Black business ownership. For the model of Hispanic business ownership under the opportunity structures theory, we utilize a weighted least squares correction for the heteroskedastic effect of percentage of Hispanic managers on the standard errors and resulting t-stats for the coefficients of the independent variables. The error variance is proportional to the percentage of Hispanic managers; therefore, the model is transformed by dividing the dependent and independent variables and constant by the square root of this variable. This transformation corrects the standard errors for the coefficients on the independent variables, making the error variance homoskedastic, but it does not change the coefficients themselves. In fact, the transformation of the data to reflect this type of heteroskedasticity improves the efficiency of the estimates (Gujarati, 1995, p. 383).

10 The rule of thumb is that if the condition index is between 10 and 30, there is moderate to strong multicollinearity. However, none of the condition indices for the models are greater than this threshold (Gujarati, 1995, p. 338).

11 Analysis is part of the authors’ research on human capital and entrepreneurship, funded by the Small Business Administration (Hackler, Harpel, & Mayer, 2008).

12 For more information, see http://www.coolcities.com/.

13 For more information about score, see http://www.score.org/explore_score.html.

14 For more information about the Rhode Island program, see http://www.everycompanycounts.com/.

15 See endnote 1 above. For more information, see The Kauffman Firm Survey at http://www.kauffman.org/kfs/ and http://www.mathematica-mpr.com/surveys/kauffmanfirm.asp.