White hiring agents’ organizational practices and out-group hiring

Abstract

Prevailing accounts of variation in race–sex hiring largely emphasize applicants’ characteristics and, to some extent, employers’ preferences. In contrast, this study considers how organizational practices, workplace formality, and job characteristics influence the race and sex of white hiring agents’ most recent hires. Analyses using data from roughly 2000 work establishments confirm that the relationship between organizational practices and out-group hiring, the hiring of applicants not like a hiring agent in terms of race and/or sex, is not simple. Findings imply that the mechanisms linking race and sex out-groups to jobs differ, so certain formal organizational practices will only partly reduce differences in hiring outcomes across race–sex groups. The effects of employee referral use, soft skill screens, skill requirements, and a job’s starting wage on out-group hires differ for female and male hiring agents. These results suggest that sex differences in social networks and women’s and men’s different location on employers’ labor queues influence the hiring process.

Introduction

Most current research views race–sex variation in hiring outcomes as a result of supply-side attributes of job applicants (Park, 1999, Reskin, 2003). For example, researchers have attributed race–sex differences in hiring outcomes to group differences in human capital characteristics (Browne, 1997, Corcoran, 1999, Smith and Welch, 1989) or group differences in job search techniques and job seekers’ social network composition (Campbell, 1988, Elliot, 1999, Elliot, 2001, Mouw, 2002a, Mouw, 2002b, Mouw, 2003, Moore, 1990). These studies often conclude that group differences in education, job referral networks, and job search methods partly account for race–sex variation in job attainment.

On the other hand, several researchers have used data on the employer side of the labor market to explain race–sex variation in hiring outcomes (e.g., Browne and Kennelly, 1999, Kirschenman et al., 1995, Kirschenman and Neckerman, 1991; Leiter et al., unpublished manuscript; Moss and Tilly, 1996, Moss and Tilly, 2001a, Moss and Tilly, 2001b, Neckerman and Kirschenman, 1991, Park, 1999, Shih, 2002, Thomas, 2000, Turner, 1997, Wilson, 1996). Their research results suggest that hiring decisions are not race–sex neutral; some employers take applicants’ race and sex into account when making hiring decisions. With few exceptions, these studies did not account for the processes that link applicants to jobs or, more generally, the organizational practices associated with hiring (Reskin, 2003, pp. 4, 7–8). For example, employers in Shih’s (2002, p. 102) Los Angeles sample characterized black workers, particularly black men, as “unmanageable” and authority-resistant, yet she did not explain what employers did—or failed to do—to block black men’s job access. A sample of North Carolina employers reported that they avoided hiring Latinos because to do so would have decreased the status of their workplace (Leiter et al., unpublished manuscript). However, the researchers could not identify what employers did to avoid Latino hiring.

Although these accounts of hiring have merit, they are incomplete because the hiring process is not only related to supply-side characteristics or linked to employers’ preferences; recruitment practices that do not seek a broad range of potential workers or screening practices with a disparate impact on members of one race or sex group also affect who gains access to jobs and, ultimately, who an employer hires. Accordingly, an alternative approach that focuses on hiring agents’ organizational practices is crucial for understanding race–sex variation in hiring (Sørensen, 2004, p. 10).

Building on this understanding, the present study investigates race–sex variation in hiring outcomes in four urban areas with an eye toward how organizational recruitment and screening practices and, to a lesser extent, workplace formality and job characteristics affect the process. Analyses using data from roughly 2000 white employers in charge of hiring, hereafter hiring agents, seek to answer three questions: (1) How do organizational practices, workplace formality, and job characteristics influence the odds a white hiring agent hires an in-group applicant (of his or her same race and sex) versus an out-group applicant (of a different race and/or sex)? (2) How does the process differ for female and male hiring agents? and (3) To what extent do organizational practices and workplace characteristics have different effects on hires involving race versus sex out-group applicants?

I categorize hires into one of four types: (1) a “cross-race” hire occurs when an applicant shares the hiring agent’s sex but not his or her race, (2) a “cross-sex” hire takes place when an applicant shares the hiring agent’s race but not his or her sex, (3) a “cross race–sex hire” occurs when an applicant and hiring agent do not share a race or sex, and (4) a “race–sex specific in-group hire” takes place when an applicant shares the hiring agent’s race and sex.¹ Categorizing hires based on a hiring agent’s and applicant’s race and sex is fundamental because racial and gender stratification are not mutually exclusive systems; rather, reward systems tend to value the qualities of white men (Elliot and Smith, 2001, p. 366; Konrad and Linnehan, 1995, p. 789).

The underlying mechanisms linking applicants to jobs may, at the same time, operate differently when an out-group applicant is a different race versus a different sex than the agent. As a result, certain practices may have a negative disparate impact on members of some but not all out-groups. For example, individuals’ social networks tend to be more race than sex homogeneous (McPherson et al., 2001, p. 423). Women and men have the opportunity to form sex-integrated contacts because their distribution in the US is relatively even, they work in the same establishments, share households, and are linked by kinship networks. By comparison, residential and workplace race segregation limit cross-race contact. Consequently, when social networks or the evaluation of social interactions play a central role in the hiring process, we might expect them to operate differently when an applicant’s race versus sex is different from the hiring agent’s.

I focus on the hiring decisions of white hiring agents because they comprise a majority of employers in charge of hiring in the US (Smith, 2002, Stoll et al., 2004) (for an analysis of black hiring agents’ hiring outcomes, see Stoll et al., 2004).² Understanding the organizational practice sources of variation in out-group hiring outcomes are fundamental if we are to develop an accurate picture of the creation and maintenance of labor market race and sex segregation and its concomitant inequalities.

The remainder of the paper is arranged as follows. I first briefly discuss three accounts of race–sex hiring discrimination. The first two accounts address the cognitive and social mechanisms motivating hiring agents’ decisions while the last account addresses how organizational practices affect hiring decisions. Although analyses focus on the latter, we cannot ignore the former processes because organizational practices can only minimize, not eliminate, them. Cognitive processes shape organizational practices while organizational practices can influence who gets hired, regardless of hiring agents’ motives. Next, I describe the advantages of using an organizational practice approach to explain race and sex specific out-group variation in hiring outcomes. I follow by hypothesizing how formal and informal organizational practices and additional workplace and job factors might affect out-group hiring and why we might expect the process to differ somewhat for male and female hiring agents. A description of the data along with a discussion of the methods of analysis precedes presentation of the analytic results and discussion of the results.