Opening the Black Box of the Matching Function: The Power of Words

Working Paper: NBER ID: w22508

Authors: Ioana Marinescu; Ronald Wolthoff

Abstract: How do employers attract the right workers? How important are posted wages vs. other job characteristics? Using data from the leading job board CareerBuilder.com, we show that most vacancies do not post wages, and, for those that do, job titles explain more than 90% of the wage variance. Job titles also explain more than 80% of the across-vacancies variance in the education and experience of applicants. Finally, failing to control for job titles leads to a spurious negative elasticity of labor supply. Thus, our results uncover the previously undocumented power of words in the job matching process.

Keywords: Job Titles; Labor Market; Wage Variance; Job Matching; Employer Search

JEL Codes: J31; J63; J64


Causal Claims Network Graph

Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.


Causal Claims

CauseEffect
job titles (M54)posted wages (J31)
job titles (M54)education and experience levels of applicants (I24)
wages (J31)number of applicants (C00)
job titles (M54)spurious relationship between wages and number of applicants (J79)

Back to index