Signaling and Migrant Labor Market Integration: A Corrrespondence Study

How to cite

Werf, Cynthia Van Der; Busso, Matias; Muñoz-Morales, Juan Sebastián; Gonzáles-Velosa, Carolina; Valencia, Juan Diego. 2025. "Signaling and Migrant Labor Market Integration: A Corrrespondence Study." AEA RCT Registry. AEARCTR-0017413. https://doi.org/10.1257/rct.17413-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Cynthia Van Der Werf* , Matias Busso , Juan Sebastián Muñoz-Morales , Carolina Gonzáles-Velosa , Juan Diego Valencia

* Primary investigator

Registry ID
AEARCTR-0017413
DOI
https://doi.org/10.1257/rct.17413-1.0
First registered
Last updated
Status
In development
Trial dates
2025-12-09 – 2026-11-30
Countries
Private
JEL codes
J24, J61, J15, C93, O15
Registry record
http://www.socialscienceregistry.org/trials/17413

Abstract

We study how much employers value labor market signaling in the context of Colombia's recent inflow of migrants with persistent labor-market frictions and informality. Our focus is on two scalable, policy-relevant signals. The first is a job-specific skills certificate issued by the Colombia National Training Service (SENA), which validates experience-based competencies through standardized practical and theoretical tests. The second is a secondary-education validation certificate obtained through an exam. We implement a large-scale correspondence study on a major online job platform in Colombia, sending fictitious applications, and randomly assigning each of these signals. We will measure impacts on callback rates and analyze heterogeneity by migrants and locals. The results will provide evidence on whether signaling can improve the access of migrants’ and locals to the labor market.

Intervention

We will conduct a correspondence study by sending out resumes in which we randomly assign a signal to the labor market, which could be either a skills certification or a high school completion certificate. We will assess the impact of each signal separately by comparing the differences in callback rates between resumes with the signal and those without it. This difference will be analyzed for migrants and locals separately and a heterogeneity analysis based on occupation and skill level will be conducted.

Experimental design

Our study aims to estimate the causal effects of two information signals in labor-market screening: (i) skills certifications and (ii) secondary-education validation certificates.

Skills Certifications:
We will randomize job vacancies to one of three certification levels—basic, intermediate, or advanced—which reflect increasing degrees of proficiency in the certified skill. For each vacancy, four résumés will be submitted: one migrant and one local applicant with a skills certification, and one migrant and one local applicant without it. This structure enables us to measure employer responses to both the presence and the level of the certification.

Secondary-Education Validation Certificate:
To study the value of education validation, we will send three résumés per job posting: one from a local applicant who completed secondary education in Colombia and two from migrant applicants who completed secondary education in Venezuela. Among the latter, one résumé will include a Colombian validation certificate, while the other will not, allowing us to isolate the effect of validation on employer callback rates.

In our design, one migrant résumé will include a Colombian validation certificate, whereas the other will not. All remaining résumé attributes—such as name, age, sex, education, work history, and the PDF CV layout—will be randomly assigned. Each attribute will be sampled from a database compiled using actual profiles of applicants who have previously obtained skill certifications.

Design details

Randomization method
All resume characteristics will be randomly assigned by a computer program.
Randomization unit
Signals will be randomly allocated to résumés, with nationality stratified according to the requirements of each treatment. In the skills-certification arm, we will enforce a 50–50 distribution between migrant and local applicants to accurately estimate differential employer responses by nationality. In the high-school validation arm, we will ensure that one-third of applicants are locals and two-thirds are migrants, reflecting the structure needed to identify the effect of the validation certificate. All remaining résumé attributes will be randomly drawn from the predefined pool.
Clusters
We will send 30,000 resumes to approximately 8,334 job offers. We may increase the sample size depending on budget availability
Observations
30,000 resumes
Arms
Skills Certifications: We will submit 20,000 résumés to study employer responses to skills certifications. Job vacancies will be randomly assigned to one of three certification levels—basic, intermediate, or advanced—yielding approximately 6,666 résumés per level. Randomization will be stratified such that, within each certification level, 50% of applicants are migrants and 50% are locals. Secondary Education Validation Certificate: To estimate the effect of secondary education validation, we will send a separate pool of 10,000 résumés. Approximately 3,334 will correspond to local applicants and 6,666 to migrant applicants. All local applicants will possess a Colombian high school diploma. Among migrant applicants, half (≈3,333) will present a Venezuelan high school diploma only, while the other half (≈3,333) will present both a Venezuelan diploma and the Colombian Secondary Education Validation Certificate.
Primary outcomes
The key outcomes are the callback rates from employers to resumes.

Cited by

No citation data available for this registration yet.

Keywords