The impact of informational nudges on saving: Evidence from Romania

How to cite

Walker, Sarah; Andries, Alin. 2021. "The impact of informational nudges on saving: Evidence from Romania." AEA RCT Registry. AEARCTR-0008247. https://doi.org/10.1257/rct.8247-1.0

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

Sarah Walker* , Alin Andries

* Primary investigator

Registry ID
AEARCTR-0008247
DOI
https://doi.org/10.1257/rct.8247-1.0
First registered
Last updated
Status
Completed
Trial dates
2017-11-01 – 2018-03-31
Countries
Romania ()
JEL codes
G40, G41, G51
Registry record
http://www.socialscienceregistry.org/trials/8247

Abstract

We implemented a randomized experiment in Romania to elucidate how messages, including general reminders and peer information, impact saving in an upper-middle income setting. We randomly selected 540 farmers from a rural region of Romania to participate in our study. Each participant was presented with an offer to join our experimental home-savings account and informed that after three months, the research team would revisit to observe the total accumulated savings and pay one percent interest on the balance. At baseline, all participants were asked whether they would like to set a savings goal and if so, how much they wished to save over the following months. In addition, participants were randomized into three groups. In the first group, participants received a general text message reminder to save. In a second group, participants received the general reminder and were also informed of the average savings goals of their peers. In a third, control group, participants received no text messages. Baseline was conducted in November 2017 and endline was conducted in March 2018. At endline, we measured and paid interest on the total savings in the experimental account. In our analysis, we compare savings outcomes across the three messaging treatment groups and conduct a heterogeneity analysis across the savings goal indicated at baseline.

Intervention

We randomly selected 540 farmers from a rural region of Romania to participate in our study. Each participant was presented with an offer to join our experimental home-savings account and informed that after three months, the research team would revisit to observe the total accumulated savings and pay one percent interest on the balance. At baseline, all participants were asked whether they would like to set a savings goal and if so, how much they wished to save over the following months. In addition, participants were randomized into three groups. In the first group, participants received a general text message reminder to save. In a second group, participants received the general reminder and were also informed of the average savings goals of their peers. In a third, control group, participants received no text messages.

Experimental design

We randomly selected 540 farmers from a rural region of Romania to participate in our study. Each participant was presented with an offer to join our experimental home-savings account and informed that after three months, the research team would revisit to observe the total accumulated savings and pay one percent interest on the balance. At baseline, all participants were asked whether they would like to set a savings goal and if so, how much they wished to save over the following months. In addition, participants were randomized into three groups. In the first group, participants received a general text message reminder to save. In a second group, participants received the general reminder and were also informed of the average savings goals of their peers. In a third, control group, participants received no text messages.

Design details

Randomization method
All randomizations were done in an office by a computer.
Randomization unit
Individual level randomization, stratified by location. The sample of 540 farmers were randomly selected from a sample frame of farmers who were enrolled in a European Union program to support rural agriculture. First, we selected a random sample of localities (called comuna) to ensure balance across particular geographic and historical factors. From a total of 41 comuna that met our geographic criteria, we randomly selected 27 for the experiment. Within each comuna, we then randomly selected 20 farmers to invite to the experiment, for a total target of 540 participants.
Clusters
No clusters.
Observations
540.
Arms
180 in control group; 180 in general reminder group; 180 in peer information group.
Primary outcomes
1) Whether the participant saved in the experimental account; and 2) how much the participant saved.

Cited by

No citation data available for this registration yet.

Keywords