The limits of self-commitment and private paternalism

Last registered on August 27, 2020

Pre-Trial

Trial Information

General Information

Title
The limits of self-commitment and private paternalism
RCT ID
AEARCTR-0006284
Initial registration date
August 27, 2020

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
August 27, 2020, 10:43 AM EDT

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

Locations

Region

Primary Investigator

Affiliation
Instituto Tecnologico Autonomo de Mexico

Other Primary Investigator(s)

PI Affiliation
UC Berkeley
PI Affiliation
ITAM

Additional Trial Information

Status
Completed
Start date
2012-08-06
End date
2013-08-13
Secondary IDs
Abstract
In the context of pawnbroker lending, we show that forcing people into commitment contracts with financial penalties for not paying on time decreases their (fee-including) financial cost by 9.5%, increases the likelihood of recovering their pawn by 25%, and increases the likelihood of repeat business by 55%. Using machine learning methods for estimating heterogeneous treatment effects, we find that more than 90% clients would reduce their financing cost with the commitment contract, however only 10% choose it. Relying on personal promises instead of fees as commitment triples take-up but has no effects on outcomes.
External Link(s)

Registration Citation

Citation
Bejarano, Enrique Seira, Isaac Meza and Joyce Sadka. 2020. "The limits of self-commitment and private paternalism." AEA RCT Registry. August 27. https://doi.org/10.1257/rct.6284-1.0
Experimental Details

Interventions

Intervention(s)
The experiment involved 5 arms, with randomization happening at the branch-day level. The first one involves a group of clients that were given only the status quo contract (henceforth the status-quo forcing group). The second arm (fee-forcing group) was given a monthly payment commitment contract, which stipulated that clients had to pay 1/3 of their loan plus accumulated interest in each of the 3 months of the term length. Failing to do this would result in a fee of 2% of the amount due that month. A third group (Fee-choice group) gave clients a choice between these two contracts.

To measure the causal effect of promises we implemented two other treatment arms. A promise-forcing arm exactly analogous to our fee-forcing arm except that instead of telling them that there is a fee if they don't pay on time, we made borrowers promise that they would pay one third of the amount in each of the 3 months, and told them that if they didn't they would have broken their promise and failed their word. Because salience is key we made them sign a (non-legally binding) personal promise and emphasized we were counting on them to keep their promise. A final arm (promise-choice arm) gave clients the opportunity to choose between the Promise contract and the Status-quo contract.
Intervention Start Date
2012-09-06
Intervention End Date
2013-08-13

Primary Outcomes

Primary Outcomes (end points)
(a) Financial cost, (b) probability of recovering their pawn
Primary Outcomes (explanation)
Financial cost is measured as the present value of servicing the loan, taking into account if the client lost the pawn.

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
We have 5 different experimental arms, which differ in which contracts were offered on a given branch-day.

1. Status-quo forcing arm: consisted of branch-days offering the status quo contract described in the paper and only this contract.
2. Fee-Forcing arm: consisted of branch-days offering only the monthly installment contract with a fee for late payment as described above.
3. Fee-Choice arm: consisted of branch-days where the client was given a choice between the monthly installment with fee contract, and the status quo contract.
4. Promise-Forcing arm: consisted of branch-days which offered the monthly payment contract only, but where there was no fee for paying late. Instead, the client was made to sign a paper which said ``I promise to pay every month the corresponding sum of _________, on the dates _________, _________, and_________. This is not a legal document and cannot be used in courts. It is just a personal promise. If I do not comply I would have broken my word.''. After signing, the promise was read to the client by the appraiser.
5. Promise-Choice arm: consisted on branch-days where the client was given a choice between the promise-forcing and the status quo contract.

By comparing the fee-forcing arm vs the status-quo arm we estimate the causal effect of mandatory installment frequency which acts as a incentive to pay earlier vis-a-vis the pay-when-you-want contract. On the other hand, having arms where clients can choose among these contracts allows us to study client's demand for commitment to pay earlier. A neoclassical consumer (with no intra-household concerns) would not voluntarily restrict her options of payment profiles (as in monthly payment contract) by being subject to fees, since she can always replicate the monthly payment profile herself under the status quo contract if she so desired. The status quo contract gives her extra flexibility to respond to negative income shocks. However, if clients have present biased preferences and are sophisticated about them, they may choose the frequent payment contract as it has commitment value. This value would be traded off against the value of flexibility.
This design also allows us to test if promises work in a natural field environment
Experimental Design Details
Randomization Method
Done in office by a computer
Randomization Unit
Branch-day
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
410
Sample size: planned number of observations
About 10,000
Sample size (or number of clusters) by treatment arms
close to 80 clusters per arm, with close to 2500 observations per cluster
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB

Institutional Review Boards (IRBs)

IRB Name
ITAMs IRB
IRB Approval Date
2020-08-26
IRB Approval Number
N/A

Post-Trial

Post Trial Information

Study Withdrawal

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Intervention

Is the intervention completed?
No
Data Collection Complete
Data Publication

Data Publication

Is public data available?
No

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials