The Value of U.S. College Education in Global Labor Markets: Experimental Evidence from China

How to cite

Chen, Mingyu. 2019. "The Value of U.S. College Education in Global Labor Markets: Experimental Evidence from China." AEA RCT Registry. AEARCTR-0004037. https://doi.org/10.1257/rct.4037-1.0

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

Mingyu Chen*

* Primary investigator

Registry ID
AEARCTR-0004037
DOI
https://doi.org/10.1257/rct.4037-1.0
First registered
Last updated
Status
Completed
Trial dates
2017-06-30 – 2018-05-31
Countries
China ()
JEL codes
I23, I26, J24, J61, J63, M51
Registry record
http://www.socialscienceregistry.org/trials/4037

Abstract

One million international students study in the U.S. each year and the majority of them compete in global labor markets after graduation. I conduct a large-scale field experiment to study how employers in China value U.S. college education. I sent over 27,000 fictitious online applications to business and computer science jobs in China, randomizing the country of college education. I find that U.S.-educated applicants are on average 18 percent less likely to receive a callback than applicants educated in China, with applicants from very selective U.S. institutions underperforming those from the least selective Chinese institutions. The U.S.-China callback gap is smaller at high-wage jobs, consistent with employers fearing U.S.-educated applicants have better outside options and will be harder to hire and retain. The gap is also smaller at foreign-owned firms, consistent with Chinese-owned firms knowing less about American education. Controlling for high school quality, test scores, or U.S. work experiences does not attenuate the gap, suggesting that it is not driven by employer perceptions of negative selection. A companion employer survey of 260 hiring managers finds consistent and additional supporting evidence for the experimental findings.

Experimental design

I send 4 applications to each job. Two of them are assigned U.S. institutions ("treatment group") and the other two are assigned Chinese institutions ("control group"). Conditional on the same institutional country, I randomly vary college selectivity. I also randomize other resume attributes in order to test some of the mechanisms. There is also a companion survey that runs after the experiment. The survey started in fall of 2018 and will end in spring of 2019. See the paper for details.

Design details

Randomization method
Randomization by computer.
Randomization unit
Individual job application
Clusters
7,000 job vacancies
Observations
28,000 job applications or resumes
Arms
7,000 in each of 4 treatment arms
Primary outcomes
Whether an employer responses to a job application via phone call, text message, or email.

Cited by

No citation data available for this registration yet.

Keywords