Personalization Effect in Debt Collection via Regular Mail

How to cite

Saulitis, Andris. 2018. "Personalization Effect in Debt Collection via Regular Mail." AEA RCT Registry. AEARCTR-0003042. https://doi.org/10.1257/rct.3042-1.1

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

Andris Saulitis*

* Primary investigator

Registry ID
AEARCTR-0003042
DOI
https://doi.org/10.1257/rct.3042-1.1
First registered
Last updated
Status
Completed
Trial dates
2018-06-04 – 2018-07-09
Countries
Latvia ()
Registry record
http://www.socialscienceregistry.org/trials/3042

Abstract

The field experiment is carried out on 2,821 defaulted debtors with consumer debts in cooperation with a debt-collector enterprise in Latvia. The goal of the experiment is to examine the effect of three different regular mail debt reminder letters on paying back the debt. The control group receives a generic simple reminder, which resembles the usual reminder letter sent by the debt-collector at least once a year to each debtor. There are two treatment conditions. In both of them, a personalized letter is sent from a debt-collector's agent instead of a letter from the debt-collector's company. In one of the conditions the letter is sent and signed by a male agent, in the other one it is the female agent. Hence, the study has a factorial design with two factors: (1) personalization of the message and (2) gender of the debt-collector agent. The control group consists of 1,410 unique debtors while the two remaining groups consists of 706 (male agent) and 705 (female agent) individuals.

Experimental design

The control group receives a generic simple reminder, which resembles the usual reminder letter sent by the debt-collector at least once a year to each debtor. There are two treatment conditions. In both of them, a personalized letter is sent from a debt-collector's agent instead of a letter from the debt-collector's company. In one of the conditions the letter is sent and signed by a male agent, in the other one it is the female agent. Hence, the study has a factorial design with two factors: (1) personalization of the message and (2) gender of the debt-collector agent.

Design details

Randomization method
The randomization was done in office by a computer. Blocked randomization was implemented, using information on region, density of the living area, debt amount, debt due age, gender, debtor age and ethnicity on each debtor.
Randomization unit
Individual
Clusters
2,821 individuals with defaulted consumer debts
Observations
2,821 individuals with defaulted consumer debts.
Arms
The control group, which receives generic simple reminder, consists of 1,410 unique debtors. Personalized letter is sent to 1,411 unique debtors. Within these, 705 individuals receive the message from a female agent and 706 individuals from a male agent.
Primary outcomes
The variable of interest is called compliance rate. It is constructed from the information on the payment status in each case.

Papers from this trial

  1. Saulı̄tis, A. (2023). Nudging debtors with non-performing loans: Evidence from three field experiments. Journal of Behavioral and Experimental Finance, 37, 100776. https://doi.org/10.1016/j.jbef.2022.100776 https://doi.org/10.1016/j.jbef.2022.100776

Cited by

No citation data available for this registration yet.

Keywords