When to Apply?

How to cite

Coffman, Katherine; Collis, Manuela. 2020. "When to Apply?." AEA RCT Registry. AEARCTR-0006522. https://doi.org/10.1257/rct.6522-1.0

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

Katherine Coffman* , Manuela Collis

* Primary investigator

Registry ID
AEARCTR-0006522
DOI
https://doi.org/10.1257/rct.6522-1.0
First registered
Last updated
Status
Completed
Trial dates
2017-08-25 – 2017-11-16
Countries
United States of America ()
Registry record
http://www.socialscienceregistry.org/trials/6522

Abstract

Labor market outcomes depend, in part, upon an individual’s willingness to put herself forward for different career opportunities. In this paper, we use both laboratory and field experiments to better understand decisions around willingness to apply for higher return, more challenging work, with a focus on gender differences. We find that, in male-typed domains, women are less likely to view themselves as well-qualified for a given opening, both because of differences in forecasts about own ability, but also, and importantly, because of differences in how high men and women believe “the bar” is. We show that these beliefs matter for application decisions. Finally, using a simple field experiment, we provide evidence that a soft-touch intervention, in which required qualifications are stated more precisely and objectively, helps to close the gender gap in willingness to apply, increasing both the average talent and the diversity of the applicant pool.

Experimental design

This is a field experiment on Upwork which is an online labor market platform.

We posted three versions of a job ad onto the Upwork platform. Each job ad asked participants with [management/analytical] skills expertise to apply for a short-term (1-hour) job answering brief essay questions about [management/analytical] skills. Each job ad contains two open positions (an intermediate-level and an expert-level job) and the candidate is invited to choose which job to apply to .

The three versions of the ad are very similar. All that differ are the stated job qualifications. The first ad (control version) consists of a generic sentence about who should apply to the expert level job rather than the intermediate level job. The second ad (“positive” treatment version) consists of the same generic sentence. It further includes a specific stated qualification about the expected test score of a successful applicant to the expert level job. The third ad (“normative” treatment version) consists of the generic sentence. It further includes a specific stated qualification encouraging applicants with at least a given test score to apply for the expert level job.

We invited a sample of participants on Upwork to view the job ad and consider applying. Each participant was randomly assigned to be invited to one of the three jobs postings (control, positive, normative). We recorded their application decision. We followed up with participants who did not apply before the specified deadline or who applied to both jobs with the request to specify which job they wanted to apply to. We followed up one time. We hired the four most qualified applications in each treatment to complete the job. We did not retain their completed work as part of our dataset.

Design details

Randomization method
By computer in Excel
Randomization unit
Individual
Clusters
1,200
Observations
1,200
Arms
evenly split across three conditions (control, positive, normative)
Primary outcomes
Application decision

Cited by

No citation data available for this registration yet.

Keywords