Anchors Away! A Large-Scale Trial Testing Approaches to Reducing Credit Card Debt

How to cite

Hiscox, Michael; Buell, Ryan. 2018. "Anchors Away! A Large-Scale Trial Testing Approaches to Reducing Credit Card Debt." AEA RCT Registry. AEARCTR-0003414. https://doi.org/10.1257/rct.3414-1.0

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

Michael Hiscox* , Ryan Buell

* Primary investigator

Registry ID
AEARCTR-0003414
DOI
https://doi.org/10.1257/rct.3414-1.0
First registered
Last updated
Status
In development
Trial dates
2018-10-31 – 2019-12-31
Countries
Private
JEL codes
D03 (Behavioural Microeconomics: underlying principles); D14 (Household Saving; Personal Finance); D31 (Personal Income, Wealth, and Their Distributions); G02 (Behavioral Finance: Underlying Principles)
Registry record
http://www.socialscienceregistry.org/trials/3414

Abstract

The aim of the study is to test whether changes to the choice architecture of credit card repayment platforms can encourage customers who have persistently been making very low repayments to increase their average credit card repayments, pay off outstanding balances more quickly, and reduce the interest charges they incur. Specifically, we assess whether adding a "Pay 5%" option to the digital platform provided by a leading bank increases average credit card repayments among these customers. We also assess the effect of removing the "Pay Minimum (2%)" option from the menu provided in the platform. In the first case, the intervention is designed to provide an alternative anchor for customers who cannot pay their full balance, but could pay more than the minimum due. In the second case, the intervention is designed to remove the low anchor altogether and allow customers to make a choice with reference only to the high anchor (the full balance).

Intervention

Customers using the banks smartphone app to make payments on their credit card currently see a choice menu with the following options:Pay closing, Pay minimum, Pay other amount (entered manually). The interventions provide customers with two alternative choice menus:
1. Pay closing, Pay 5%, Pay minimum, Pay other amount
2. Pay closing, Pay other amount

Experimental design

Eligible customers will be randomly allocated to view either the current repayment choice menu or one the two new payment choice menus (the interventions) in their smartphone app when they receive the notification that their monthly payment is due.

Design details

Randomization method
Eligible customers will be randomly assigned by computer algorithm to one of the three experimental groups after blocking on the following baseline covariates: segment (e.g. Persistent Debt, Revolver, All Others), age, product tenure, May 2018 payment, average balance Mar-May 2018, credit limit
Randomization unit
Individual
Clusters
N/A
Observations
The total potential sample size is 1,270,481. There will be 74,967 customers in treatment 1, 75,011 customers in treatment 2, and 1,120,503 customers in the control. The intervention will only be seen by customers who choose to pay their credit card via the “Account details” page or the “Statements” page of the bank’s app. This requires the customer to have a transaction account with the bank, in addition to their credit card. Over the previous 12 months approximately 10% of all credit card payments done via the app were done through this channel. We thus expect an effective sample size of approximately 127,000.
Arms
Based on the above we expect effect sample sizes of approximately 7,500 customers in each treatment group and 112,000 in the control group.
Primary outcomes
Aggregated monthly dollar repayment ($)
Secondary outcomes
• Aggregated monthly repayment as a percentage of monthly balance (%) • Dollar repayment at transaction level ($) • Dollar repayment as a percentage of balance at transaction level (%) • Monthly fees and interest paid ($) • Whether or not customers pay more than minimum (binary variable, monthly) • Number of “persistent debt” customers (monthly) • Whether or not account falls in arrears (binary variable, monthly)

Cited by

No citation data available for this registration yet.

Keywords