Signal Extraction in Centralized Assignment: Evidence from Teacher Recruiting

  • Authors: Matteo Bobba, Tim Ederer, Gianmarco León-Ciliotta, Christopher A. Neilson and Marco Nieddu
  • BSE Working Paper: 1618 | November 2026
  • Keywords: signal extraction, teacher hiring, centralized assignment
  • JEL codes: D47, D82, I21
Download PDF Download pdf Icon

Abstract

We study the design of decentralized signal extraction in centralized assignment. Using data on Peru’s nationwide teacher allocation mechanism, we estimate a rich model of applicants’ behavior, teacher effectiveness and schools’ discretionary screening accounting for heterogeneity in schools’ preferences and signal-extraction ability. Decentralized screening improves the effectiveness of selected teachers on average relative to test-based hiring, but its returns vary substantially across schools. We find that efficiently allocating screening capacity according to schools’ signal extraction ability is unfeasible. Instead, we show that using teacher value-added measures for centralized prescreening can generate large gains and outperforms disclosing that information directly to schools. However, prescreening too aggressively leaves more vacancies unfilled as schools increasingly compete for the same candidates.

Subscribe to our newsletter
Want to receive the latest news and updates from the BSE? Share your details below.
Founding Institutions
Distinctions
Logo BSE
© Barcelona Graduate School of
Economics. All rights reserved.
FacebookInstagramLinkedinXYoutube