A Commitment Device to Reduce Spending: Evidence from a Choice Experiment on Banking Products

How to cite

Lent, Max van; Collewet, Marion; Knoef, Marike; Kessels, Roselinde. 2025. "A Commitment Device to Reduce Spending: Evidence from a Choice Experiment on Banking Products." AEA RCT Registry. AEARCTR-0016959. https://doi.org/10.1257/rct.16959-1.0

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

Max van Lent* , Marion Collewet , Marike Knoef , Roselinde Kessels

* Primary investigator

Registry ID
AEARCTR-0016959
DOI
https://doi.org/10.1257/rct.16959-1.0
First registered
Last updated
Status
In development
Trial dates
2025-10-06 – 2026-12-01
Countries
Private
JEL codes
D14, G41, G51
Registry record
http://www.socialscienceregistry.org/trials/16959

Abstract

Short run debt, financial problems, and problematic debt are ubiquitous. There is an increasing body of literature that studies savings commitments through defaults, automatic savings programs etc. as a possible solution for debts. The effectiveness of such interventions are mixed. In this project we study the demand for commitment devices that limit expenditures instead. We develop a choice experiment in which respondents make choices between different banking products that differ in price, the possibility to set an expenditure limit, the type of notification when reaching the expenditure limit, and the consequence of reaching this limit. In addition to this choice experiment we elicit risk and time preferences, and derive financial health and -literacy for our respondents. The results inform us about the demand for banking products that contain a self imposed expenditure limit, which is currently not widely available. In addition this informs us about whether demand is driven by those people who are likely to benefit from commitment the most.

Intervention

We randomly assign respondents to decide 10 times between two different banking products that contain options to self-impose an expenditure limit, and compare them to the current default where such an option is not available.

Experimental design

We develop and implement a vignette experiment on a representative sample of around 1450 Dutch people of at least 16 years old.
We ask the respondent several times to state which of two banking products they prefer. The banking products vary by whether it is possible to impose a self-chosen expenditure limit, the notification one receives when hitting the limit, the consequence of hitting the limit, as well as what the bank does in order to support you to not hit the limit.
Our experiment wants to study the demand for banking products that contain the option to self-choose an expenditure limit, as well heterogeneity analysis.

Design details

Randomization method
We randomize the vignettes that people get using a D-efficient design.
Randomization unit
We randomize at the individual level, not stratified.
Clusters
1450 respondents
Observations
1450 respondents
Arms
There is no control group: everyone receives vignettes in which they make a choice between different banking products.
Primary outcomes
The WTP for the different notifications, consequences of reaching the self-imposed expenditure limit, and the types of help offered to avoid reaching the limit.

Cited by

No citation data available for this registration yet.

Keywords