Employer Learning, Statistical Discrimination and University Prestige

Abstract

This paper investigates whether firms use university prestige to statistically discriminate among college graduates. The test is based on the employer learning literature which suggests that if firms use a characteristic for statistical discrimination, this variable should become less important for earnings determination as a worker gains labor market experience. In this framework, we use a regression discontinuity design to estimate a 19% wage premium for recent graduates of two of the most selective universities in Chile. However, we find that this premium decreases by 3 percentage points per year of labor market experience. These results suggest that employers use college selectivity as a signal of workers' quality when they leave school. However, as workers reveal their productivity throughout their careers, they become rewarded based on their true quality rather than the prestige of their college.

Does employer learning vary by schooling attainment? The answer depends on how career start dates are defined

“…Our results contradict ABH, who find evidence of employer learning for the S = 12 workers but not for the S = 16 workers. However, our results are consistent with Lang and Siniver (2011) and Bordón and Braga (2013) who, to our knowledge, are the only other analysts to test for employer learning using samples of college-educated workers. These studies report clear-cut evidence of employer learning for college graduates in Israel and Chile, respectively.…”

Employer Learning and the Dynamics of Returns to Universities: Evidence from Chinese Elite Education during University Expansion

“…We create an indicator variable promotion that equals 1 if an individual is currently in at least a mid-level manager position or professional rank and equals 0 otherwise. Elite and ordinary university graduates with less than five years of potential experience have similarly low and insignificantly different probabilities of having been 29 This interpretation is closer to that of Bordon and Braga (2017) who look at the impact of elite universities in Chile along with university admission scores. 30 Specifically, we include interactive terms between the elite dummy and a series of indicators for 1-2 years, 3-5 years, 6-10 years, and 11-16 years of potential experience.…”

“…More recently, Castex and Dechter (2014) find higher returns over time to educational attainment and lower returns to cognitive skills in the U.S., a pattern they attribute to slower technological growth in recent periods. A few international papers have also documented that college prestige serves as an entry signal that becomes less important throughout workers' careers (Lang and Siniver (2011) for Israel; Bordon and Braga (2017) for Chile).…”

University prestige, performance evaluation, and promotion: Estimating the employer learning model using personnel datasets

“…Moreover, it is widely known that individuals with higher ability tend to attend elite schools (Dale and Krueger (2002), Hoekstra (2009) and Dale and Krueger (2011)) and that employers are likely to form different expectations about workers' ability based on the prestige level of the university from which the workers graduated. Indeed, using a Chilean administrative data set, a recent study by Bordón and Braga (2013) reports that employers infer workers' ability from the prestige of the universities from which workers graduate at the early stage of workers' career, but its importance fades quickly as workers accumulate labor-market experience. 1 Kahn and Lange (2013)'s contribution is the most closely related to our paper, in that they estimate an employer learning model with a personnel panel data set.…”

“…Bordón and Braga (2013) use the test score and the admission cutoff to implement a regression discontinuity analysis. They do not structurally recover the parameters that govern the speed of employer learning.…”