Working Paper: NBER ID: w24012
Authors: Juan Saavedra; Dario Maldonado; Lucrecia Santibanez; Luis Omar Herrera Prada
Abstract: It is unclear whether public sector teachers are under or overpaid relative to other occupations due to lack of knowledge about teachers’ outside labor market options and other unobserved attributes related to compensation. We estimate causal labor market returns to novice public teachers in Colombia. Our approach takes advantage of a national, standardized, teacher-screening exam, scores on which determine eligibility for public teaching jobs. We use four nationwide administrative data sources in a regression discontinuity approach to show that applicants who marginally pass the teacher screening test have greater annual earnings during the first three years of tenure than applicants below the passing cutoff. The total earnings effect is a combination of higher daily wages and greater labor supply, part of which is in outside, predominantly non-teaching jobs for a substantial fraction of public teachers. For infra-marginal high-scoring applicants, we show that being a public teacher in Colombia is as attractive, if not more, as for those at the margin. On the whole, rather than a penalty, public teachers in Colombia across all ability levels earn a substantial labor market premium early in their careers.
Keywords: Labor market returns; Public sector teachers; Colombia; Teacher screening exam
JEL Codes: I26; J22; J24; O15; O38
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| annual earnings (J31) | higher daily wages (J31) |
| annual earnings (J31) | increased labor supply (J20) |
| being a public teacher (J45) | labor market premium (J39) |
| earnings advantage diminishes over time (J31) | primary earnings source shifts (E25) |
| test's screening value (C52) | deter high-quality applicants (D29) |
| applicants who marginally pass the teacher screening test (A00) | annual earnings (J31) |