The Effects of Nudging Summer Youth Employment Program Applicants to Apply to More Jobs

How to cite

Modestino, Alicia; Marks, Mindy; Hoover, Hanna. 2023. "The Effects of Nudging Summer Youth Employment Program Applicants to Apply to More Jobs." AEA RCT Registry. AEARCTR-0011180. https://doi.org/10.1257/rct.11180-1.0

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

Alicia Modestino* , Mindy Marks , Hanna Hoover

* Primary investigator

Registry ID
AEARCTR-0011180
DOI
https://doi.org/10.1257/rct.11180-1.0
First registered
Last updated
Status
In development
Trial dates
2023-04-01 – 2023-12-31
Countries
Private
JEL codes
I25 Education and Economic Development; I38 Government Policy • Provision and Effects of Welfare Programs; J24 Human Capital • Skills • Occupational Choice • Labor Productivity; J08 Labor Economics Policies; C93 Field Experiments
Registry record
http://www.socialscienceregistry.org/trials/11180

Abstract

Summer Youth Employment Programs (SYEP) have been shown to have significant impacts on youth outcomes such as reducing violent crime, increasing high school graduation, and boosting subsequent employment and wages. Much of this research is based on lotteries from oversubscribed programs. However, many cities find that it is not feasible to allocate youth to jobs using simple random assignment due to heterogeneous preferences from employers and youth participants. This is because the matching process is more complex so that the program needs to balance both youth and employer interests to ensure participation. Even when random assignment is used, two issues arise that may lead to inequitable outcomes. First some SYEPs conduct random assignment at the employer level as opposed to the program level and our prior research shows that this can harm BIPOC applicants when the distribution of applications to position is imbalanced by race. In addition, this racial inequity may be exacerbated if there is an imbalance between the number of applicants and the number of openings across employers (e.g., some jobs are over- versus under-subscribed), such that youth and/or employers may not get their first choice. Our prior research documents that in the extreme, where over half of youth apply to only one job, this labor market mismatch across applicants and jobs means that no match can be made for a significant fraction of youth.

To help address this problem, we propose an RCT where we test the effectiveness of an informational treatment to improve job application behavior among youth by giving them salient information about the successful applicants and helping them to reflect about their job search. Each week, we will randomize youth who have applied to less than three positions into a treatment and a control group, stratified by availability of text and parent email. The treatment group will receive an email and/or text message nudging them to apply to more jobs along with salient information about the success rate of applicants from the prior year. The control group will not receive any nudging communication. If the sample size is sufficient, we will further subdivide the treatment group with one arm receiving a subsequent 'reflection' treatment in the form of an optional Qualtrics survey to help youth identify which job positions meet their specified criteria from the remaining positions which are still seeking job candidates.

Intervention

We propose an RCT where we test the effectiveness of an informational treatment to improve job application behavior among youth by giving them salient information about the successful applicants and helping them to reflect about their job search. Each week, we will randomize youth who have applied to less than three positions into a treatment and a control group, stratified by availability of text and parent email. The treatment group will receive an email and/or text message nudging them to apply to more jobs along with salient information about the success rate of applicants from the prior year. The message will encourage youth in the treatment group to apply to at least 3 positions and to positions which have numerous openings, along with a bar chart showing the hiring rate of youth from the prior year who applied to 2 or less jobs versus those who applied to 3 or more jobs. The control group will not receive any nudging communication.

If the sample size is sufficient, we will further subdivide the treatment group with one arm receiving a subsequent 'reflection' treatment. This reflection message will only be sent to those in the treatment group who still have not applied to at least three jobs one week prior to the close of the application portal. The reflection message will encourage youth to take a Qualtrics survey to find jobs that match their interests and still have openings.

We will send informational emails to the treatment group. The treatment group will be youth who applied to the Boston SYEP within the prior week and have applied to 2 or less jobs. We will stratify across treatment and control groups by if the youth included a phone number, if the youth included their parents email address, and if the youth included their parent's phone number. If there are sufficient numbers within the treatment group, we will send a second email nudge to half of the initial treatment group. The other half of the treatment group will not receive the second nudge. Again, we still stratify selection into this second treatment group by having a phone number, parent email address, and parent's phone number. In this second informational experiment, we will tell youth to navigate to a Qualtrics survey. This survey will ask youth about which industries, occupations, and neighborhoods they prefer to work in. Once selected, we will suggest employer-partners to apply to who still have employment slots still available.

Experimental design

Each week, we will collect data from the application portal and randomize youth who have applied to less than three positions into a treatment and a control group, stratified by availability of text and parent email. If sufficient sample size allows, we will also use the same randomization procedures on April 24th to randomize those in the treatment group who still have not applied to at least three jobs into whether they will receive the follow-up reflection message and Qualtrics survey.

Design details

Randomization method
Randomization will be done using stata statistical software.
Randomization unit
Randomization will occur at the individual level each wekk.
Clusters
0
Observations
2000
Arms
Control group=1,000 Treatment group = 1,000 Arm 1 of Reflection Treatment = 300 Arm 2 of Reflection Treatment = 300
Primary outcomes
Our primary outcomes include Number of job applications youth submitted (total and whether >=3) Number of job applications youth submitted by 2 weeks after initial email (total and if greater than three) Whether youth applied to jobs with multiple openings (applied to any, how many applied to, average number of openings)
Secondary outcomes
Our secondary outcomes include: Whether participated in Qualtrics reflection Whether applied to under-subscribed jobs identified by the Qualtrics reflection Whether the youth was ever selected for employment

Cited by

No citation data available for this registration yet.

Keywords