Can robo-advice help with debt repayment?

How to cite

Shaw, Jonathan; Burke, Chris; Croxson, Karen; Smart, Laura; Reuter, Jonathan; Rossi, Alberto; D'Acunto, Francesco. 2020. "Can robo-advice help with debt repayment?." AEA RCT Registry. AEARCTR-0006447. https://doi.org/10.1257/rct.6447-1.1

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

Jonathan Shaw* , Chris Burke , Karen Croxson , Laura Smart , Jonathan Reuter , Alberto Rossi , Francesco D'Acunto

* Primary investigator

Registry ID
AEARCTR-0006447
DOI
https://doi.org/10.1257/rct.6447-1.1
First registered
Last updated
Status
In development
Trial dates
2020-09-18 – 2022-12-31
Countries
United Kingdom of Great Britain and Northern Ireland (Nationwide)
JEL codes
G51, G53
Registry record
http://www.socialscienceregistry.org/trials/6447

Abstract

We have concerns that consumers can find it difficult to make good decisions across a wide range of financial products. This is particularly true in the context of debt management, where where through poor repayment and refinancing decisions, debts can increase and few consumers seek out advice until they are in serious financial difficulty. To address this situation, it is important to understand better how consumers make decisions and the extent to which well-timed and appropriate advice can help.

In this work, we will focus on the debt repayment and explore the potential for automated robo-advice to assist consumers to make better decisions. A number of firms are starting to provide these services. At present the majority of debt advice is provided through a mix of charitable and for-profit organisations but debt charities do not have sufficient capacity to meet consumer demand and consumers may be unwilling or unable to pay for for-profit assistance. Robo-advice has a potential to address this and may provide additional benefits when consumers feel uncomfortable discussing their debt issues in person with a human advisor. However, very little is known about the effectiveness of this advice, whether people would be willing to trust or take up automated advice and what they would be willing to pay for it.

We propose an online trial to investigate consumer outcomes and attitudes to different types of robo-advice for debt management. In this experiment, participants will face a number of hypothetical scenarios where they have a certain amount of money to allocate to the repayment of multiple debts. The objective of the participant is to allocate the repayments in such a way as to minimise the interest and fees they would pay.

Intervention

We propose an online trial to investigate consumer outcomes and attitudes to different types of robo-advice for debt management. In this experiment, participants will face a number of hypothetical scenarios where they have a certain amount of money to allocate to the repayment of multiple debts. In some of the scenarios they will be offered robo-advice, possibly with a training element. The objective of the participant is to allocate the repayments in such a way as to minimise the interest and fees they would pay.

Experimental design

Randomisation will take place at the individual level. Participants will be randomly allocated to the control group or one of the treatment arms.

Design details

Randomization method
Randomisation will be achieved using the randomizer built into the survey software we will use (Qualtrics).
Randomization unit
Individual
Clusters
4500 individuals
Observations
4500 individuals
Arms
900 individuals control, 900 individuals in each of four treatment arms (robo-advice with and without education and with and without willingness to pay)
Primary outcomes
Our primary outcome measure is the amount of hypothetical interest and fees the individuals saves relative to the optimal and most suboptimal repayment strategies.
Secondary outcomes
A secondary outcome measure is the length of time spent on each task.

Cited by

No citation data available for this registration yet.

Keywords