The effect of eldercare on job search: Evidence from a correspondence study

How to cite

Gaulke, Amanda; Meyer, Zack. 2025. "The effect of eldercare on job search: Evidence from a correspondence study." AEA RCT Registry. AEARCTR-0016940. https://doi.org/10.1257/rct.16940-1.1

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

Amanda Gaulke* , Zack Meyer

* Primary investigator

Registry ID
AEARCTR-0016940
DOI
https://doi.org/10.1257/rct.16940-1.1
First registered
Last updated
Status
In development
Trial dates
2025-10-20 – 2027-03-31
Countries
Private
JEL codes
J14, J23, J16
Registry record
http://www.socialscienceregistry.org/trials/16940

Abstract

In the United States, about twenty percent of adults provide care to another adult and that prevalence has increased over time (NAC and AARP, 2020). Recent estimates suggest that more people are providing care to another adult than to a child (NAC and AARP, 2020) and the value of unpaid care to other adults with limitations in daily activities totals $600 billion (Reinhard et al., 2023). Given the literature documenting scarring effects from unemployment spells (for example, see Kroft et al., 2013; Eriksson and Rooth, 2014), this raises questions about whether revealing eldercare as the reason for leaving the previous job and as the activity undertaken while unemployed helps or hurts job candidates. We use a correspondence study to quantify whether and to what extent the gap in callback rates between the eldercare providers and the continuously employed group differs from the gap in callback rates between the unexplained (no reason given) group and the continuously employed group.

Intervention

We randomly assign one of the applicant triplets to have provided eldercare in the past - the reason for leaving the previous job and the time spent unemployed is revealed to be eldercare. We also have the previously used intervention of the unexplained group - the applicant in the triplet that has the same length of gap but no reason for why left the job or what the time unemployed was spent doing.

Experimental design

We use a correspondence study to determine whether revealing the eldercare spell results in a smaller gap in
callback rates compared to continuously employed applicants relative to a group that provides no explanation.
This means that we create fictitious applications and submit them to real job postings.

Implementation:
1) collect job ads that meet the study criteria
2) Once we find an eligible job ad, we apply using one of the triplets. We randomize the order and have rules for spacing out applications.
3) After submitting the applications, we measure the callback rate, which is our main outcome of interest.
4) After receiving responses from potential employers, we respond that the applicant is no longer available for the job to minimize costs for firms.

Design details

Randomization method
In office by a computer
Randomization unit
Randomization happens at the applicant level
Clusters
We plan to apply using 9,099 applicants or 3,033 jobs (given the triplet)
Observations
9,099 triplets
Arms
Each triplet will include an applicant that is continuously employed, one that is part the eldercare treatment group and one part of the explained group. This means 9,099/3=3,033 applications per arm of the experiment
Primary outcomes
Callback rate
Secondary outcomes
NA

Cited by

No citation data available for this registration yet.

Keywords