White entrepreneurs in central cities face the novel experience of working in a social context in which they are racial minorities, while at the same time they are a part of the dominant coalition of firm owners and are members of the majority within the larger society.

Entrepreneurs in urban contexts find that they must build relationships across racial and ethnic boundaries. But “tokens” – numerical minorities in organizations or contexts dominated by the majority – face considerable challenges in doing so. That’s because cross-race relationships, within and outside of organizations, remain relatively unusual. In his 1987 study of core discussion networks, Harvard sociologist Peter Marsden found that only eight percent of Americans reported any racial or ethnic diversity in their networks, with white Americans having the greatest homogeneity.

In the inner-city context, then, those with the least experience in forging cross-racial relationships have the greatest need to do so. White entrepreneurs in central cities usually cannot leverage their personal knowledge of co-ethnic customer tastes and appeal to bounded solidarity to build protected markets.

While these firm owners may choose to focus on building cross-race, central city relationships personally, they may also establish relationships with other institutions or individuals – “social brokers” – that can provide links to immigrant and ethnic groups. Government agencies, non-profit and service organizations, religious institutions, and even current customers or employees can serve as social brokers, yet not be explicitly dedicated to this practice. We set out to determine the role and significance of social brokers in helping white entrepreneurs in central-city locations forge cross-racial and cross-ethnic links with employees and customers.

Then and Now

In his classic 1973 study of discrimination in hiring, sociologist Howard Aldrich examined patterns of firm ownership in the predominantly black neighborhoods of Roxbury (Boston), Fillmore (Chicago) and Northern Washington, DC. The majority of the employers (55 percent) in these areas were white, and whites were minorities in the residential population (ranging from 10 percent of the population in Fillmore to 28 percent in Roxbury). Aldrich also found that 80 percent of the white firm owners were “absentee owners.” White firm owners were more likely to hire people who lived outside of the neighborhood and were more likely to hire white employees than non-white central city firm owners. Other studies at the time found similar ownership patterns in other cities.

Do these conditions still persist? In 1970, when Aldrich’s data was collected, each of the neighborhoods studied had only a decade earlier been predominantly white. Aldrich tied the pattern of white firm ownership to an inability of white firm owners to leave as rapidly as white residents.

But the “white flight” context of the early 1970’s no longer exists in central cities. White residents have long been gone from these neighborhoods, and the absolute number of businesses has declined significantly. So today’s central city firm owners are more likely to be located there by choice. A second difference is the considerable influx of non-white immigrants from Asia, Central America and the Caribbean. Previous studies have shown these groups to have high incidence of entrepreneurship.

To update Aldrich’s study, we analyzed a subsection of employer respondents from the Multi-City Study of Urban Inequality (MCSUI). This data was collected by researchers in Atlanta, Boston, Detroit and Los Angeles between 1992 and 1995 to examine labor market dynamics, with a particular focus on jobs requiring no more than a high school education. Table 1 presents the incidence of white firm ownership by metropolitan area subsection, based on 510 respondents.

As found in studies from the 1970s, the dominant coalition of firm owners are white (84.9 percent), and seven of every 10 firm owners in predominantly non-white central city areas are white. The percentage of white firm ownership in central city areas is even greater than in studies from the early 1970’s. Why? It has long been argued that whites have greater access to critical capital stocks, making them better able to start firms and to weather economic hardships than their black and Hispanic counterparts. Second, black central city neighborhoods have been especially hard hit by the exit of the middle class, who had options to move after segregation declined in the 1970’s.

We then analyzed responses of firm owners regarding the incidence of white customers and employees by city subsection. In central city areas, where the majority of the residents are non-white, the white/nonwhite composition of the customer base and employee base is evenly split (50.1 and 47.5 percent, respectively). However, there is considerable variation across firms in terms of their customer and employee demography. The standard deviation was 35.1 percent for white customers, and 40.5 percent for white employees.

Hiring Patterns

Next, we set out to determine the influence of owner race on racial composition of the employment base. Because Aldrich found that differences in firm type (e.g., retail, service or manufacturing) accounted for some of the differences in hiring patterns, we controlled for sector of employment in our analysis.

In line with the findings of studies from a generation ago, we found that the race of the firm owner influences hiring patterns, even when adjusted for firm location and industrial sector. White firm ownership increased the percentage of white employees by an average of 40 percent.

Aldrich generated four hypotheses regarding the possible role of discrimination in hiring patterns. First, white employers may simply prefer associating with whites over blacks. Second, white employers might practice statistical discrimination, in which negative beliefs about the work fitness of blacks cause employers to prefer not to hire black employees. Third, white employers might avoid hiring blacks because of negative reactions of other employees or the firm’s customers. Fourth, white employees might be over-represented because whites who worked in the firms prior to the wholesale white exodus from the neighborhood hung on to their jobs in these neighborhoods. Aldrich was ultimately unable to determine whether discrimination accounted for the overrepresentation of white employees in white-owned firms located in black neighborhoods.

Today, two of these hypotheses are less useful. The customers of firms in today’s central city areas are as likely to be white as non-white. And because white flight is no longer a recent phenomenon – as it was in the 1970s – there is a low potential that the current set of white employees were unable to find work elsewhere. The second hypothesis, that white employers have developed a “distaste” for non-white labor, has been examined by other researchers. In interviews with white employers in central city areas, employers expressed their tendency to practice statistical discrimination with black applicants because of past experiences with negative workplace attitudes and behaviors. Sociologist William Julius Wilson in 1996 examined black employers from the same neighborhood, and found that they expressed similar views of the attitudes and work ethic of central city black employees.

But employer distaste probably doesn’t explain the differences in hiring patterns by race of owner observed above. Perhaps white employers, like other tokens, face barriers in establishing cross-race relationships that might assist them in locating the most qualified employees from the local pool of labor. Given the generally low opinion employers appear to have of central city labor, reference-based hiring may be one of the prime means of ensuring labor quality.

Weak Ties

In his classic 1973 study of personal contacts in job-seeking, sociologist Mark Granovetter found that the overwhelming majority (83 percent) of managerial and professional job seekers found their jobs through acquaintances with whom they spoke occasionally or rarely. This finding of the “strength of weak ties” is one of the more influential ideas in the social sciences.

But Granovetter’s reanalysis of Stanley Milgram’s data on interracial acquaintance chains has been less discussed. Granovetter reanalyzed the success rate of white senders who attempted to deliver a booklet to black targets through acquaintance chains, if the first connection between a white sender and a black recipient described the black person as a “friend” or an “acquaintance.” Granovetter found that the weak tie instances – those where the first black connection was described as an acquaintance – were twice as likely to result in a successful completion to the eventual target. Weak acquaintance ties were more successful than strong friendship ties in reaching cross-race targets.

Given this, we hypothesize that cross-race weak ties might also assist in the recruitment of employees. And institutions or individuals that bridge socially segregated groups are a form of weak tie relationship that employers can use to mediate their token status. Connections through community service organizations, religious institutions, civic leaders, and current employees might assist employers in locating qualified minority employees and result in larger numbers of minority employees.

Using Social Brokers

Hiring proper employees is a critically important task for a firm owner. But when the employer is white and the employees are generally non-white, the hiring challenge may be especially difficult. A racial “outsider” may find it tough to accurately screen an applicant during the hiring process and reveal potential behavioral or attitudinal mis-hires. Once employees are hired, white employers may worry that negative on-the-job feedback will result in accusations of racial prejudice. Given the distrust, doubt and accusations that can sometimes accompany cross-race interactions in central cities, some entrepreneurs may choose to avoid central city locations or minority employees altogether.

The MCSUI contained a series of questions regarding the methods used by employers in hiring for their last employment vacancy. The positions were those that did not require the applicant to hold a college degree. We investigated the influence of hiring methods that involved social brokers on minority hiring rates in central cities. Table 2 presents the results of three linear regression analyses using dummy variables to determine the influence of the race of owner, city subsector, industry sector and hiring methods on the percentage of non-white employees in the firm.

The analysis of the full set of firms shows that manufacturing firms are more likely to hire non-white employees. This is likely due to the greater need for unskilled labor in these firms. The first evidence of social brokerage is found in the strong influence of employee recommendations on the percentage of non-white employees. Both the magnitude of this coefficient and its high level of significance is persuasive evidence of the use of this practice among entrepreneurs. The use of help wanted signs was also shown to increase the percentage of non-white employees. Help wanted signs are a strategy for employers seeking to attract employees that happen to pass the firm location, and may be a means to hire from the local community without aid of brokerage.

Private-service temporary agencies appear to serve as brokers for firms seeking to hire white employees, while community agencies serve firms seeking non-white employees. These differences are likely generated by the divergent customer needs that each agency serves.

Even after controlling for city subsection, industry and hiring method, the strongest influence on percentage of non-white employees is still the race of firm owner. This suggests that our analysis has failed to account for other factors influencing the hiring choices of white owners, and that these results do not rule out preferences for homophily.

Help Wanted

Splitting the files by city subsection allowed us to compare the incidence of brokerage strategies by firm location. We found that outside central city areas, manufacturing firms have a greater tendency to hire non-white employees, and help wanted signs increase the percentage of non-white employees. An alternative perspective is that employers located in suburban areas are familiar with available local labor (predominantly white), and use help wanted signs as an “affirmative action” strategy, designed to attract potential employees that are not in their current social network.

Employers may find that non-white applicants that learn of their openings by passing their location are more likely to be “acculturated” or familiar with the workplace behaviors necessary to work in suburban contexts. The positive finding for all firms appears to be driven by the use of this brokerage strategy in central cities. Outside of central cities, employers utilize current employees as brokers for non-white employees, though to a lesser degree. Private-service temporary agencies play a strong role in bringing white employees into firms. Finally, referrals from educational institutions enhance non-white hiring outside central cities. It appears that for firm owners in these areas, educational institutions play a brokerage role in assisting in the hire of non-white employees.

Customer Relations

Entrepreneurs must also manage another critical constituent group on the demand-side of the equation: customers. Customers are not only a firm’s source of revenue, they are a prime means of attracting new customers through word-of-mouth. But for white entrepreneurs operating in central city areas, building relationships with customers from the local community may present many of the same challenges found in locating employees. White firm owners are less likely to be personally familiar with community members and are less likely to be personally aware of emerging customer tastes and needs. Non-white customers may resent the presence of white firm owners, and customer dissatisfactions may take on an accusatory tone generally not experienced in contexts where the customers are predominantly white. There is a historical legacy of mistreatment of minority customers in businesses owned by white proprietors. White central city entrepreneurs may therefore attempt to use employees as brokers to manage potentially fractious relations with a substantial base of non-white customers.

We set out to determine the influence of customer demography on the makeup of a firm’s labor pool. If increasing percentages of non-white customers positively influences the percentage of non-white employees, it would suggest that employers use employees as social brokers to manage relationships with customers. To investigate, we ran a linear regression analysis of race of firm owner, industry sector, firm location, and percentage of non-white customers on the mean percentage of non-white employees.

The results of this analysis are consistent with the hypothesis that customer demography explains some of the variance in employee demography even after controlling for race of firm owner. Subsequent analyses of these influences by firm location showed the same pattern of results throughout. However, the magnitude changes of coefficients provided some interesting findings. First, when compared with the prior analysis of employer race influence on employee demography, the coefficient for white firm owners decreased when the predictor variable for customer demography was entered. Some of the variance explained by employer race in the earlier analysis is now shown to result from customer demography. Second, white customers have a positive influence on the number of white employees in all locations, although the relationship became stronger in suburban areas. Third, white firm owners have an even greater positive influence on the percentage of white employees in central city areas. This suggests one of two alternative hypotheses: a) white firm owners in central city locations have an even greater preference for white employees than in suburban areas; b) white firm owners face even greater challenges in locating non-white labor in central city areas. Given our theoretical framing of white firm owners as tokens we suspect the latter.

Emerging Markets

Taken together, our findings suggest that relationships play a critical role in job seeking, especially when operating cross-racially. And understanding this dynamic is becoming more important. For over the past several decades, patterns of social and racial segregation have created structural holes, which in turn have created economic opportunities in central cities – America’s emerging domestic markets.

Entrepreneurs of all races and ethnicities are figuring out how to build wealth while providing jobs and leadership that diminish many of the social problems we’ve come to associate with inner city communities. In May 2003, Inc. magazine released its annual list of the most rapidly growing inner-city firms. The characteristics of the members of the Inner City 100 may seem surprising: average sales of over $25 million, and five-year growth rates over 600 percent.

Social brokers will play an important role in developing these markets further. America’s inner- city neighborhoods will increasingly show promise as sites for investment, and many of the entrepreneurs pursuing these opportunities will not be ethnic and racial minorities. On Inc.’s list, 62 percent of the firm owners were white. Locating high quality employees in a cross race situation requires the recognition that relationships matter and that relationships tend to stay within the same race. Without building social brokerage relationships, employers run the risk of missing the most qualified members of the labor pool.

Gregory B. Fairchild is assistant professor of management at Darden Graduate School of Business Administration at the University of Virginia

Jeffrey A. Robinson is assistant professor of management at NYU Stern.