Do we discriminate against immigrants? A random field experiment in the Argentine labor market

How to cite

Laffitte, Gabriel. 2023. "Do we discriminate against immigrants? A random field experiment in the Argentine labor market." AEA RCT Registry. AEARCTR-0012410. https://doi.org/10.1257/rct.12410-1.0

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

Gabriel Laffitte*

* Primary investigator

Registry ID
AEARCTR-0012410
DOI
https://doi.org/10.1257/rct.12410-1.0
First registered
Last updated
Status
In development
Trial dates
2023-11-06 – 2024-03-18
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/12410

Abstract

The research intends to study immigrant discrimination in the Argentinian labor market, using a Randomized Control Trial (RCT). Based on the seminal paper of Bertrand and Mullainathan (2004), fictitious but plausible resumes will be sent to active job vacancies published on a digital platform. The treatment is being from a non-Argentine nationality. I am sending resumes to four different kinds of work and different industries. Due to the origin from the immigrants in Argentina, the treatment will be the nationality from neighboring countries (Paraguay, Bolivia, Chile and Uruguay). I distinguish two different treatments, considering the various cultural similarities between those countries and the Metropolitan Area of Buenos Aires (AMBA). Particularly there will be two treatments: being from Paraguay or Bolivia, and being from Chile or Uruguay. The control group will be the resumes with argentinian nationality.

Experimental design

Fictitious resumes will be sent to active job vacancies from the Metropolitan Area of Buenos Aires (AMBA), in order to estimate the causal effect of discrimination. The job vacancies are published on a digital platform and several fictitious accounts will be made for that purpose. The resumes are built using actual resumes from the labor market, and have been altered to anonymize them. The fictitious resumes differ in two characteristics: nationality and gender. Only one resume will be sent to each job vacancy, and which resume will be sent is chosen at random.
I have made four models of resumes for each position where the resumes will be sent. This has been done to match the resume with the job search, in order to increase the chance of being called and in order to avoid being detected by the recruiting team of each company. The randomization of the treatment will be done within each industry, and I will send resumes until 18th march 2024 or until I reach an 80% of power.

Design details

Randomization method
Randomization will be done with a computer software.
Randomization unit
Job vacancy
Clusters
2500 job vacancies
Observations
2500 job vacancies
Arms
1250 resumes with argentinian nationality (control), 625 resumes with chilean/uruguayan nationality (treatment 1), 25 resumes with bolivian/ paraguayan nationality (treatment 2)
Primary outcomes
The main outcome is the rate of callbacks from the firms who show any interest in continuing the recruiting process.

Cited by

No citation data available for this registration yet.

Keywords