Decreasing SNAP Denial Rates with a Text Campaign and Document Uploader

How to cite

Read, Blair; Brown, Julia; Permut, Stephanie; Turner, Mary Clair; Speigel, Lauren; Moynihan, Donald; Herd, Pamela; Jilke, Sebastian. 2023. "Decreasing SNAP Denial Rates with a Text Campaign and Document Uploader." AEA RCT Registry. AEARCTR-0012634. https://doi.org/10.1257/rct.12634-1.0

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

Blair Read* , Julia Brown , Stephanie Permut , Mary Clair Turner , Lauren Speigel , Donald Moynihan , Pamela Herd , Sebastian Jilke

* Primary investigator

Registry ID
AEARCTR-0012634
DOI
https://doi.org/10.1257/rct.12634-1.0
First registered
Last updated
Status
On going
Trial dates
2023-09-07 – 2024-10-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/12634

Abstract

The Office of Evaluation Sciences (OES) is collaborating with Code for America (CfA) and a state partner to evaluate the effectiveness of a text message campaign encouraging use of a new online document uploader to submit required documents for Supplemental Nutrition Assistance Program (SNAP) applications. The campaign is designed to make it easier to submit verification documents, reduce procedural denials, and increase access to benefits for eligible applicants. These findings will help identify strategies to decrease SNAP procedural denial rates, which can be scaled across other US states and possibly other public benefits programs. Conditional on data availability, this project will also build evidence on equitable outcomes by answering to what extent the intervention increased access to SNAP resources among underserved groups.

Intervention

The intervention is sending a text message with a link to a new document uploader to submit SNAP income verification documents. Applicants in both the intervention and control groups will be sent a series of text messages that prompt and remind them to upload their verification documents to complete their SNAP applications. The text message campaign between intervention and control only differ in whether applicants are sent a link to the Code for America document uploader (in the intervention group) or are not sent any link to an online document uploader. Applicants can receive a maximum of seven text messages over the course of the evaluation (an introductory message and up to six reminders).

Experimental design

The intervention is sending a text message with a link to a new document uploader to submit SNAP income verification documents. Applicants in both the intervention and control groups will be sent a series of text messages that prompt and remind them to upload their verification documents to complete their SNAP applications. The text message campaign between intervention and control only differ in whether applicants are sent a link to the Code for America document uploader (in the intervention group) or are not sent any link to an online document uploader. Applicants can receive a maximum of seven text messages over the course of the evaluation (an introductory message and up to six reminders).

Design details

Randomization method
Applicants will be randomized based on the time (tens seconds digit) they submit their application and the date treatment was assigned. Treatment is assigned a day after applicants submitted their application. Applicants are assigned to either a treatment or a control condition using the following algorithm: 1. Look first at the date of randomization. If the day field of the randomization date is an odd number (e.g. July 17, 19, 21, etc.), then proceed to step 2. If the day field of the randomization date is an even number (e.g. July 18, 20, 22, etc.), then proceed to step 3. 2. Look at the last digit of the second value of the timestamp (i.e. the bolded digit in a timestamp of the form HH:MM:SS). If this digit is an odd number, then assign the application to Treatment. If this digit is an even number, then assign the application to Control. 3. Look at the last digit of the second value of the timestamp (i.e. the bolded digit in a timestamp of the form HH:MM:SS). If this digit is an odd number, then assign the application to Control. If this digit is an even number, then assign the application to Treatment.
Randomization unit
Units are randomized at the level of individual application ID. An application ID is linked to an applicant who applied on behalf of their SNAP unit (typically, a household).
Clusters
25,000 - 35,000
Observations
25,000 - 35,000
Arms
We anticipate between 25,000 and 35,000 applicants in the final study, split roughly evenly between treatment and control.
Primary outcomes
(1) Submitted verification documents (2) No procedural denial (3) Enrolled in SNAP
Secondary outcomes
(1) Monthly dollar value of SNAP benefits (2) Procedural denial for missing verification documents (3) Procedural denial for missing an interview (4) Ineligibility for SNAP (5) Time to decision (6) Opt out of text messages

Cited by

No citation data available for this registration yet.

Keywords