Enhancing the Saving for Every Child Program Through Behavioral Interventions

How to cite

Grinstein-Weiss, Michal. 2021. "Enhancing the Saving for Every Child Program Through Behavioral Interventions." AEA RCT Registry. AEARCTR-0008456. https://doi.org/10.1257/rct.8456-1.0

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

Michal Grinstein-Weiss*

* Primary investigator

Registry ID
AEARCTR-0008456
DOI
https://doi.org/10.1257/rct.8456-1.0
First registered
Last updated
Status
In development
Trial dates
2021-10-31 – 2022-09-30
Countries
Private
JEL codes
D10 Household Behavior: General; D14 Household Saving; Personal Finance; G51 Household Saving, Borrowing, Debt, and Wealth; I38 Welfare, Well-Being, and Poverty: Government Programs; Provision and Effects of Welfare Programs; D91 Micro-Based Behavioral Economics: Role and Effects of Psychological, Emotional, Social, and Cognitive Factors on Decision Making; C93 Field Experiments
Registry record
http://www.socialscienceregistry.org/trials/8456

Abstract

The Saving for Every Child Program (SECP) is a universal child savings account program in Israel administered by the National Insurance Institute of Israel (NII). The government opens a savings account for every newborn, and the government deposits approximately NIS 51 (New Israeli Shekel, equivalent to USD 15.5) per month to these accounts. Parents can transfer additional funds from a separate child allowance into the long-term savings account and select a specific savings account for their children.

This study will examine how the small changes in letters and text messages that the government sends to parents of each newborn in Israel will affect parents’ decisions to enroll in and engage with the SECP. We will randomly assign parents to the control group that will be exposed to the current SECP design or one of several treatment conditions, where the design of each condition will be informed by behavioral economics. We expect that programmatic changes in the SECP communication channels will encourage parents to make program-related decisions that will result in greater long-term savings for their children.

We expect that the study will have significant implications for the design and implementation of the current program as well as savings and asset-building programs more generally.

Intervention

As part of the experiment, we will randomly assign parents of newborns to the control or one of several treatment groups.

Parents in the control group will be exposed to the program in a business as usual way. These parents will receive the current communication regarding the Saving for Every Child Program (SECP) from the National Insurance Institute of Israel (NII). The current communication includes (i) a letter from the NII that informs parents about the eligibility for the SECP and (ii) two text messages sent to newborns’ mothers: the first is sent out 3 months after childbirth and the second is sent out 5.5 months after childbirth. The messages are sent out only to parents who have not made an active choice regarding their children’s savings accounts.

We expect to have a total of 10 treatment groups. Each group will receive a variation of a letter and a text message from the NII informing parents about the SECP. Depending on the group assignment, the letters will include the current letter’s text and design, simplified letter’s text and design, a letter containing graphical presentation of the program’s benefits, an attachment describing the program in Arabic, and a modified text and design of the envelope in which the letter is sent out. Furthermore, text messages will vary in their design, timing, and recipients depending on the group. Wording of the text messages will include either the current text or an updated text informed by the principles of behavioral economics. Timing of sending the text messages will include either the current timing (2 messages within the first 6 months since childbirth) or an updated timing (6 messages within the first 6 months since childbirth). Text messages will be sent out either only to newborns’ mothers or both parents/caretakers.

Experimental design

As part of the experiment, parents of newborns will be randomly assigned to the control or one of several treatment groups. As described above, each group will receive different combinations of SECP-related letters and text messages from the NII.

The sample will consist of mothers of nearly all children born in Israel between 10/31/2021 and 03/30/2022. We will exclude several groups of parents, such as those who have an outstanding balance to the NII, those who gave birth outside the hospital, and new immigrants with the children under the age of 6 months. In total, we expect that approximately 52,000 parents will participate in the experiment.

The study will include a total of 11 experimental groups: the control group and 8 treatment groups will include 4,000 parents per group, and 2 remaining treatment groups will include 8,000 parents each. These remaining groups include (i) an intervention that includes an attachment describing the program in Arabic (along with the current design of a text message) and (ii) an intervention that sends out text messages to both parents/caretakers of a newborn (along with the updated design of a letter).

We will use NII’s administrative data to analyze study outcomes. The outcomes will be tracked over a period of six months since childbirth, since this is the period during which parents can make active SECP decisions.

Administrative data will be supplemented with survey data of a random sample of experiment participants.

Design details

Randomization method
Randomization will be done by a computer.
Randomization unit
Individual level (mothers of eligible children).
Clusters
The treatment is not clustered.
Observations
The total expected number of observations will include 52,000 parents.
Arms
The control group and 8 treatment groups are expected to include 4,000 parents per group. Two remaining treatment groups are expected to include 8,000 parents per group.
Primary outcomes
The study’s primary outcomes will be related to the enrollment and participation in the program. In particular, we explore the following primary outcomes: (1) Active program enrollment: this means that rather than relying on the program’s defaults, parents make active choices regarding their children’s savings accounts. (2) Doubling the savings: this means that rather than receiving the default NIS 51 (USD 15.5) from the government, parents decide to transfer an additional NIS 51 (USD 15.5) from a separate child allowance into the children’s savings accounts. (3) Selection of savings account track: this means that parents actively decide to invest their children’s funds into one of available account tracks, such as high-, medium-, or low-yield investment accounts, one of several bank accounts, or one of two religious funds (Sharia or Halakhic). (4) Total amount of savings accumulated in SECP accounts in the first 6 months since childbirth.
Secondary outcomes
In addition to the primary outcomes, we will examine any change in the SECP enrollment choices for other children. This means that we will explore whether parents make a decision to double the amount of savings deposits and to change a savings account track for additional children (e.g., older siblings who were not the direct subjects of the intervention). Furthermore, we will use data from a follow-up survey to examine parents’ familiarity with the program and program features, attitudes towards the SECP, and the ways in which the SECP funds are affecting balance sheet characteristics and other economic outcomes for recipients. These outcomes may include liquid savings amounts, unsecured debt, college savings (outside the SECP), and life satisfaction/optimism. The survey is currently under development and the exact secondary outcomes will be described once the survey is developed.

Cited by

No citation data available for this registration yet.

Keywords