The impact of university prestige in the employment process. A field experiment

• In Event: Impact of the global university

Proposal

Introduction and relevance

Does university prestige lead to discrimination? This research aims to understand if employers prioritize the signal associated with the name of the university someone graduated from above an applicant’s skills in the employment process. I conduct a field experiment of the labor market (Bertrand & Mullainathan, 2004; Daniel, 1968; Riach & Rich, 2002) in three countries: the United States, the United Kingdom, and Australia. The design requires the submission of fake resumes to job openings where the experimental variables are randomly assigned to each application, and measuring the call-back rate. The resumes belong to fictitious citizens with full working rights, both female and male, that have attended universities of varying prestige in the United States, the United Kingdom, and Australia respectively.

I submitted job applications to positions in two skill intensive sectors of the labor market: accounting and information and communication technology. To each job opening, I submitted two resumes. One resume has a high skill match with the generic requirements of entry level jobs in each sector. The high skill match IT resume belongs to a computer science graduate, it includes one summer internship and a year-long part time job as a software developer, additional IT certifications, as well knowledge in a variety of programming languages. The high skill match accounting resume belongs to an applicant with a degree in accounting, internship and part time experience in accountancy, as well as knowledge in key software used in the field.

A second resume has a low skill match with these requirements. The same low-skill match resume is used for both accounting and IT. The low-match resumes belong to an applicant with a degree in biology or ecology (country variation), with one listed course in accounting or IT. The low match applicant has work experience as a customer service representative and a tutor in biology and mathematics, as well as knowledge in a number of research related software. The resumes have been validated through interviews with career counselors and human resource experts in all the experimental countries, as well as IT and accountancy professionals. Altogether, I submitted 2,400 job applications.

This design uniquely allows for the control of social capital effects—absent for fictitious persons—in securing employment (Bayer, Ross, & Topa, 2005). More importantly, this design is able to differentiate between the effects of human capital, as measured through the high and low skill match resumes (Becker, 1975; Mincer, 1974; Schultz, 1959; 1961) and the signaling effect of university prestige (Spence, 1973) in the labor market.

This research is timely as the stratification among higher education institutions increases as a result of massification (Altbach, 2017). The signaling effects of university prestige in a stratified higher education system may undermine the social mobility effects of college attainment. A better understanding of the effects of university prestige in the labor market is particularly important for minority students who are less likely to be admitted, attend, and complete an elite university degree around the world (Cook & Frank, 1993; Mok & Neubauer, 2016; Posselt, 2016; Seider, 2008).

Theoretical framework

Academic literature suggests that graduating from a prestigious university has a wide array of pecuniary and nonpecuniary benefits for individuals (Black & Smith 2004; 2006; Brewer, Eide, & Ehrenber, 1999; Long, 2008; Long, Allison, & McGinnis, 1979; Morley & Aynsley, 2007; Rivera, 2015).

There are three generic mechanisms that have been recognized to significantly impact the employment process. First, it is inferred that employers seek to maximize productivity, and thus are looking to hire individuals with the skills and abilities that best contribute towards this very productivity. This mechanism, better known under the label of human capital theory (Becker, 1975; Mincer, 1974; Schultz, 1959; 1961), explains why attending a university yields better employment returns for graduates: universities equip individuals with additional skills recognized and rewarded by the labor market.

The second mechanism suggests that under conditions of information asymmetry between employers and employees, employers are incentivized to identify and use signals that would indirectly indicate the level of productivity of an individual (Spence, 1973). In the signaling model, a visible, observable feature, such as the presence or the absence of a postsecondary degree, or the prestige associated with this degree, is interpreted as an indicator for a broader non-visible feature, in this case potential productivity. Signaling theory, as introduced in social sciences by Michel Spence (1973), refers to the transfer of information in a market system in the form of signals. In the job market, an employee is perceived as a possible asset identifiable by analyzing observable characteristics and attributes rather than, and instead of testing actual productivity. These characteristics can be alterable or non-alterable. Alterable or changeable characteristics are defined as signals. Non-alterable or unchangeable characteristics are defined as indices. Education is an example of a signal, while sex and race are perceived as indices.

Last, the third mechanism, best explained by social capital theory, entails that individuals obtain jobs based on their social network and the trust they have established with people they know directly or indirectly (Bayer, Ross, & Topa, 2005; Petersen, Saporta, & Seidel, 2000) through practices such as sharing job specific information, occupation specific norms, and job referrals (Montgomery, 1991).
The experimental design I propose is able to differentiate between the effects of human capital, signaling, and social capital of prestigious universities in the labor market—albeit solely in the recruitment process.

Results

The results of the study suggest that university prestige does not predict callbacks across experimental countries and across field of the labor market. Similarly, no evidence of sex based discrimination is found in this study.

Scientific significance of the study or work

Discrimination studies traditionally focus on understanding the use of unchangeable individual and group features such as sex, race, and disability status. This dissertation expands the concept of discrimination to understand the negative effects drawn by signals, or changeable and observable features taken as proxies for unobservable characteristics, such as university prestige.

Author

• Georgiana Mihut, Boston College