Learning in the Limit: Income Inference from Credit Extension

Last registered on October 31, 2022

View Trial History

Pre-Trial

Trial Information

General Information

Title

Learning in the Limit: Income Inference from Credit Extension

RCT ID

AEARCTR-0010315

Initial registration date

October 26, 2022

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

October 31, 2022, 3:55 PM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Country

China

Region

Primary Investigator

Name

Xiao Yin

Affiliation

UC Berkeley, Haas

Contact Primary Investigator

Other Primary Investigator(s)

Additional Trial Information

Status

Completed

Start date

2020-03-31

End date

2020-10-10

Keywords

Behavior, Finance & Microfinance

Additional Keywords

Consumption, MPC, MPB, Credit Supply, Income Expectations.

JEL code(s)

D14, D15, D91, E21, E51, G21

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

Combining a randomized controlled trial with bank account and survey data, I show that credit-limit extensions significantly increase consumer expectations about their future income. A one-dollar increase in credit limits raises consumer income expectations over the next six months by 40 cents and total consumption by 34 cents. The expectation changes explain around 35% of the total spending responses to credit limit extensions. The results show that consumers infer information from lenders' credit-supply decisions, and this learning behavior impacts consumers' economic decision-making greatly.

External Link(s)

Registration Citation

Citation

Yin, Xiao. 2022. "Learning in the Limit: Income Inference from Credit Extension." AEA RCT Registry. October 31. https://doi.org/10.1257/rct.10315-1.0

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

Intervention Start Date

2020-04-05

Intervention End Date

2020-04-12

Primary Outcomes

Primary Outcomes (end points)

The key outcome is the changes in participants' beliefs elicited from the surveys before and after the experiment.

Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

I work with a commercial bank in China to perform the RCT. At the end of March 2020, the bank selected a group of consumers and decided on a new credit limit that was higher than this group of customers’ current credit card limit, based on the bank’s own proprietary rules. A random sample of approximately 16,000 consumers was selected from the group of consumers as the potential subjects in this study. Of the selected subjects, 7,000 consumers were randomly selected to form the control group, and the rest formed the treatment group. Between April 5 and April 12, the participants in the treatment group were informed about the opportunity to increase their credit card limits to the
amount offered by the bank. They could decide to either accept or ignore the offer within one week of the extension. The control-group offers were postponed to the beginning of October 2020.

Within one week before and after the experiment, we sent out a survey to the participants to ask for their beliefs about their future income, spending, saving, etc. The participants were informed that the survey would be used to study the expectations and preferences of representative credit card holders in China and that the information would be used for only scientific research purposes.

A random 15% of the participants were shown the following information on top of the post-experiment survey:

"To test their business strategies, banks often randomly select some people to have a change in their credit card limits and see how they change their spending."

Experimental Design Details

Randomization Method

Randomization is based on a uniform random number generator.

Randomization Unit

Randomization is at the individual level.

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

1 cluster

Sample size: planned number of observations

16000 individual

Sample size (or number of clusters) by treatment arms

5950 individuals with the postponed limit offer and without information treatment
1050 individuals with the postponed limit offer and with information treatment
7650 individuals with the unpostponed limit offer and without information treatment
1350 individuals with the unpostponed limit offer and with information treatment

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

Supporting Documents and Materials

IRB