Beyond Signaling and Human Capital: Education and the Revelation of Ability

Abstract

In traditional models of ability signaling (A. Michael Spence 1973;Andrew Weiss 1995), education provides a way for individuals to sort into groups (education levels) that are correlated with ability. Employers use education to statistically discriminate, paying wages that depend in part on the average ability of the individuals with the same level of education. Building on these models, Henry S. Farber and Robert Gibbons (1996) and Joseph G. Altonji and Charles R. Pierret (2001) develop a framework in which employers do not initially observe the ability of a worker, but learn about it over time. As employers gather more information about the ability of a worker, they rely less on education and more on the new information in determining the wages. In these dynamic learning models, education serves as a tool for workers to signal their unobserved ability, although its role in determining wages decreases with experience.In this paper, we argue that education (specifically attending college) plays a much more direct role in revealing ability to the labor market. Rather than simply sorting individuals into broad ability groups, our results suggest that college allows individuals to directly reveal key aspects of their own ability to the labor market. Following in the tradition of the employer learning literature, the evidence that we provide is based on an examination of the returns to ability over the first 12 years

The effect of university attended on graduates’ labour market prospects: A field study of Great Britain

“…Students decide whether they will make a significant investment in terms of effort and time in order to enter a higher ranked university, and these investments should result in increased rewards from firms (Hanushek and Welch, 2006;Kjelland, 2008;Arcidiacono et al, 2010). Higher-ability students might find it easier to cope with the requirements of competitive universities, and the university system might simply provide employers with cheap sorting of students by demonstrating their productivity (Hartog et al, 2010).…”

Graduate returns, degree class premia and higher education expansion in the UK

“…In this context, we note evidence (Smith and Naylor, 2001;Crawford, 2014) that university performance differs by social class of family background. Arcidiacono et al (2010) provide evidence for the US that while signalling in an employer learning/statistical discrimination (EL-SD) approach might be relevant for understanding returns to high school graduates, it is less applicable for college graduates as employers have considerable information about the latter, for whom the grade point average together with transcripts and other information, such as standardised test scores, reveal rather than merely signal ability. This contrasts with traditional institutional features in the UK, where secondary school leavers receive certificates showing detailed information about performance in national examinations at ages 16 and 18, while university graduates receive relatively less finely calibrated measures.…”

Do Starting Salaries for Graduates Measure the Quality of Education? A Review of Studies by Russian and Foreign Authors

“…Some researchers find actual effects of academic achievements [Ireland 2009], while others believe academic performance only "takes on" the effects of other interrelated unobservable variables, such as abilities [Arcidiacono, 2008]. Followers of the direct effects hypothesis go by the employer learning and statistical discrimination (El-SD) model [Altonji, Pierret,…”

“…Researchers who deny a direct influence of academic performance on salary argue that employers don't need to discriminate statistically among workers, since more and more other ways to assess potential employee performance appear: CVs, tests, personal interviews, recommendations, etc. In this context, academic performance is only a proxy for abilities [Arcidiacono, Bayer, Hizmo, 2008].…”

“…• analysis of education received (quality of education, academic performance, field of study); • CV analysis (most CVs include results of various examinations and standardized test scores); • personal interviews and recommendations [Arcidiacono, Bayer, Hizmo, 2008]; • diverse tests, including standardized exams (like SHL, for instance).…”