Working Paper: NBER ID: w21230
Authors: W. Bentley MacLeod; Evan Riehl; Juan E. Saavedra; Miguel Urquiola
Abstract: Spence (1973) noted that individuals’ choice of educational quantity—measured by years of schooling—may stem partially from a desire to signal their ability to the labor market. This paper asks if individuals’ choice of educational quality—measured by college reputation—may likewise signal their ability. We use data on the admission scores of all Colombian college graduates to define a measure of reputation that gives clear predictions in a signaling framework. We find that college reputation, unlike years of schooling, is correlated with graduates’ earnings growth. We also show that Colombia’s staggered rollout of a new signal of skill—a college exit exam—reduced the earnings return to reputation and increased the return to individual admission scores. These results are consistent with the hypothesis that a college’s reputation provides information about the ability of its student body and about its value added, broadly understood.
Keywords: College reputation; Labor market outcomes; Signaling; Earnings growth; Exit exams
JEL Codes: I21; I23; I24; I25; I26; J01; J24; J3
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| college reputation (I23) | initial earnings (J31) |
| college reputation (I23) | earnings growth (O49) |
| staggered introduction of exit exam (I21) | earnings return to reputation (I26) |
| staggered introduction of exit exam (I21) | return to individual admission scores (C29) |
| college reputation (I23) | correlation with individual admission scores (C29) |